<%BANNER%>

Persistence of the Latin Accent in the Nominal System of Castilian, Catalan, and Portuguese


PAGE 484

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 3042 nupt ae, rum, f. plur. 1 3 3 HC1.L.L f npcies 33HC1.L.HCnupcias 22HC1.HC npcias 22HC1.HC 3043 unc a, ae, f. 1 3 3 HC1.L.L f una 22HC1.L onza 22HC1.L ona 22HV1.L 3044 orb ta, ae, f. 1 3 3 HC1.L.L f rbita 33HC1.L .L rbita 33HC1.L.L rbita 33HC1.L.L 3045 parv lus, a, um 1 3 3 HC1.L.L f prvula 33HC1 .L.L prvula 33HC1.L.L prvula 33HC1.L.L 3046 pasc a, ae, f. 1 3 3 HC1.L.L f pasqua 22HC1.L pascua 22HC1.L pscoa 33HC1.L.L 3047 pinn la, ae, f. 1 3 3 HC1.L.L f pnnula 33L1.L.L pnula 33L1. L.L pnula 33L1.L.L 3048 purp ra, ae, f. 1 3 3 HC1.L.L f prpura 33HC1.L.L pr pura 33HC1.L.L prpura 33HC1.L.L 304 9 pust la, ae, f. 1 3 3 HC1.L.L f pstula 3 3 HC1.L .L pstula 3 3 HC1.L.L pstula 3 3 HC1.L.L 3050 salv a ae, f. 1 3 3 HC1.L.L f slvia 3 3 HC1 .L.L salvia 2 2 HC1.L slvia 2 2 HC1.L 3051 sept mus, a, um 1 3 3 HC1.L.L f sptima 3 3 HC1 .L.L sptima 3 3 HC1.L.L stima 3 3 L1.L.L 3052 spong a o spong a, ae, f., = 1 3 3 HC1.L.L f esponja 3 2 HC1.L esponja 3 2 HC1.L esponja 3 2 HV1.L 3053 yll ba, ae, f., = 1 3 3 HC1.L.L f sllaba 3 3 HC1.L .L slaba 3 3 L1.L. L slaba 3 3 L1.L.L 3054 thaps a ae, f., del gr. 1 3 3 HC1.L.L f tpsia 3 3 HC1.L.L tapsia 2 2 HC1.L tpsia 2 2 HC1.L 3055 tert ae rum, f. 1 3 3 HC1.L.L f trcia 3 3 HC1 .L.L tercia 2 2 HC1.L trcia 2 2 HC1.L 3056 tess ra ae, f. 1 3 3 HC1.L.L f tssera 3 3 L1.L.L tsera 3 3 L1.L. L tssera 3 3 L1.L.L 3057 tipp la 1 3 3 HC1.L.L f tpula 3 3 L1.L. L tpula 3 3 L1.L.L tpula 3 3 L1.L.L 3058 transf ga ae, m./f. 1 3 3 HC1.L.L m/f trnsfuga 3 3 HC1 .L.L trnsfuga 3 3 HC1.L.L trnsfuga 3 3 HC1.L.L 3059 unc a ae, f., 1 3 3 HC1.L.L f una 2 2 HC1.L uncia 484 2 2 HC1.L una 2 2 HV1.L 3060 ung la ae, f. 1 3 3 HC1.L.L f ungla 2 2 HC.L ua 2 2 L1.L unha 2 2 L1.L 3061 valv la, dim. de valva 1 3 3 HC1.L.L f vlvula 3 3 HV1.L.L vlvula 3 3 HC1.L.L vlvula 3 3 HC1.L.L 3062 vict ma ae, f. 1 3 3 HC1.L.L f vctima 3 3 HC1 .L.L vctima 3 3 HC1.L.L vtima 3 3 L1.L.L 3063 virg la, dim. de virga 1 3 3 HC1.L.L f vrgula 3 3 HC1.L.L vrgula 3 3 HC1.L.L vrgula 3 3 HC1.L.L 3064 vesp ra ae, f. 1 3 3 HC1.L.L f vespre 2 2 HC1 .L vspera 3 3 HC1.L.L vspera 3 3 HC1.L.L HV1.HV.X (1) 3065 tr ch a ae, f. 1 3 3 HV1.HV.L f trquea 3 3 L1.L.L trquea 3 3 L1.L. L traqueia 3 2 L.L1.L HV1.L.L (107) 3066 f l a, ae, f. 1 3 3 HV1.L.L f filla 2 2 L1.L hija 2 2 L1.L filha 2 2 L1.L 3067 b c na (not bucc na), ae, f. 1 3 3 HV1.L.L f botzina 3 2 L.L1.L bocina 3 2 L.L1.L buzina 3 2 L.L1.L 3068 claus la ae. f. 1 3 3 HV1.L.L f clusula 3 3 HV 1.L.L clusula 3 3 HV1. L.L clusula 3 3 HV1.L.L 3069 c p a ae, f. 1 3 3 HV1.L.L f cpia 3 3 L1.L.L copia 2 2 L1.L cpia 2 2 L1.L 3070 c p la ae, f. 1 3 3 HV1.L.L f coble 2 2 L1.L copla 2 2 L1.L copla 2 2 L1.L 3071 c p la ae, f. 1 3 3 HV1.L.L f cpula 3 3 L1.L.L cpula 3 3 L1.L. L cpula 3 3 L1.L.L 3072 cr p la ae, f. 1 3 3 HV1.L.L f crpula 3 3 L1.L.L crpula 3 3 L1.L.L crpula 3 3 L1.L.L 3073 c r a ae, f. 1 3 3 HV1.L.L f cria 3 3 L1.L.L curia 2 2 L1.L cria 2 2 L1.L

PAGE 485

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 3074 aem lus a, um 1 3 3 HV1.L.L f mula 33L1.L.L mula 33L1. L.L mula 33L1.L.L 3075 r a, ae, f. 1 3 3 HV1.L.L f era 2 2 L1.L era 2 2 L1.L eira 2 2 HV1.L 3076 scr f lae, rum, f. 1 3 3 HV1.L.L f escrfula 4 2 L1. L.L escrfula 4 2 L1.L.L escrfula 4 2 L1.L.L 3077 th cus, a, um 1 3 3 HV1.L.L f tica 3 3 L1.L.L tica 3 3 L1.L.L tica 3 3 L1.L.L 3078 f br ca, ae, f. 1 3 3 HV1.L.L f fbrica 3 3 L1. L.L fbrica 3 3 L1.L.L fbrica 3 3 L1.L.L 3079 f b la, ae, f. 1 3 3 HV1.L.L f faula 2 2 HV1.L fbula 3 3 L1.L. L fbula 3 3 L1.L.L 3080 faec la (f c la), ae, f. 1 3 3 HV1.L.L f fcula 3 3 L1.L.L fcula 3 3 L1.L.L fcula 3 3 L1.L.L 308 1 f r a, ae, f. 1 3 3 HV1.L.L f fira 2 2 L1.L feria 2 2 L1.L feira 2 2 HV1.L 3082 f br ca, ae, f. 1 3 3 HV1.L.L f far ga 2 2 HC1.L fragua 2 2 L1.L frgua 2 2 L1.L 3083 gl r a, ae, f. 1 3 3 HV1.L.L f glria 3 3 L1.L.L gloria 2 2 L1.L glria 2 2 L1.L 3084 gr t a, ae, f. 1 3 3 HV1.L.L f grcia 3 3 L1.L.L gracia 2 2 L1.L graa 2 2 L1.L 3085 f b la, ae, f. 1 3 3 HV1.L.L f faula 2 2 HV1.L habla 2 2 L1.L fala 2 2 L1.L 3086 f m na, ae, f. 1 3 3 HV1.L.L f fembra 2 2 HC1.L hembra 2 2 HC1.L fmea 2 2 L1.L.L 3087 s p a ae, f., del gr. 1 3 3 HV1.L.L f spia 3 3 L1.L.L jibia 2 2 L1.L siba 2 2 L1.L 3088 l m na, ae, f. 1 3 3 HV1.L.L f lmina 3 3 L1. L.L lmina 3 3 L1.L.L lmina 3 3 L1.L.L 3089 l n a (l n a), a e, f. 1 3 3 HV1.L.L f lnia 3 3 L1.L.L lnea 3 3 L1. L.L linha 2 2 L1.L 3090 l n la, ae, f. 1 3 3 HV1.L.L f lnula 3 3 L1. L.L lnula 3 3 L1.L.L lnula 3 3 L1.L.L 3091 m ch na, ae, f. = 1 3 3 HV1.L.L f mquina 3 3 L1.L. L mquina 3 3 L1.L.L mquina 3 3 L1.L.L 3092 [ars] m tr ca < m tr cus, a, um = 1 3 3 HV1.L.L f mtrica 3 3 L1.L.L m trica 485 3 3 L1.L.L mtrica 3 3 L1.L.L 3093 m m cus, a, um, adj., = 1 3 3 HV1.L.L f mmica 3 3 L1.L.L mmica 3 3 L1.L.L mmica 3 3 L1.L.L 3094 naufr gus, a, um 1 3 3 HV1.L.L f nufraga 3 3 HV 1.L.L nufraga 3 3 HV1.L. L nufraga 3 3 HV1.L.L 3095 naus a, ae, f., = 1 3 3 HV1.L.L f nusea 3 3 HV1.L. L nusea 3 3 HV1.L.L nusea 3 3 HV1.L.L 3096 naut cus, a, um, adj., = 1 3 3 HV1.L.L f nutica 3 3 HV1.L.L nutica 3 3 HV 1.L.L nutica 3 3 HV1.L.L 3097 p g na, ae, f. 1 3 3 HV1.L.L f pgina 3 3 L1.L.L pgina 3 3 L1. L.L pgina 3 3 L1.L.L 3098 p n a, ae, f. 1 3 3 HV1.L.L f pinya 2 2 L1.L pia 2 2 L1.L pinha 2 2 HV1.L 3099 *pul ca


PAGE 486

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 3107 taen a ae, f., del gr. 1 3 3 HV1.L.L f t nia 33L1.L.L tenia 22L1.L tnia 33L1.L 3108 t b a, ae, f. 1 3 3 HV1.L.L f tbia 33 L1.L.L tibia 22L1.L tbia 33L1.L 3109 tr g la ae, f. 1 3 3 HV1.L.L f tralla 22L1.L tralla 22L1.L tralha 22L1.L 3110 v la, dim. de va 1 3 3 HV1.L.L f vula 33L1.L. L vula 33L1.L.L vula 33L1.L.L 3111 v p ra ae, f. 1 3 3 HV1.L.L f vibra 22L1.L vbora 33L1. L.L vbora 33L1.L.L 3112 v n a ae 1 3 3 HV1.L.L f vinya 22L1.L via 22L1.L vinha 22V1.L L1.L.X (64) 3113 aqu la 1 3 3 L1.L.L f guila 33L1.L. L guila 33 L1.L.L guia 3 3 L1.L.L 3114 l s, gr. 1 3 3 L1.L.L m loe 3 3 L1. L.L loe 3 3 L1.L.L alo 3 1 L.L.L1 3115 amph ra, y este del gr. 1 3 3 HC1.L.L f mfora 3 3 HC1.L.L nfora 3 3 HC1.L.L nfora 3 3 HC1.L.L 3116 ar a 1 3 3 L1.L.L m rea 3 3 L1.L. L rea 3 3 L1.L.L rea 3 3 L1.L.L 3117 LL bibl a ae, f., del gr. 1 3 3 L1.L.L f bbl ia 3 3 L1.L.L biblia 2 2 L1.L bblia 3 2 L1.L 3118 b r as, ae, m., del gr. 1 3 3 L1.L.L m brees 3 3 L1.L.HC breas 3 3 L1. L.HC breas 3 3 L1.L.HC 3119 c pr a ae, f. 1 3 3 L1.L.L f cbria 3 3 L1. L.L cabria 2 2 L1.L cbrea 3 3 L1.L.L 3120 c m ra (c -m ra), ae, f., = 1 3 3 L1.L.L f cambra 2 2 HC1.L c mara 3 3 L1.L.L cmara 3 3 L1.L.L 3121 sch d la, ae, f. 1 3 3 L1.L.L f cdula 3 3 L1.L.L cdula 3 3 L1.L. L cdula 3 3 L1.L.L 3122 c th ra ae, f., del gr. 1 3 3 L1.L.L f ctara 3 3 L1.L.L ctara 3 3 L1.L.L ctara 3 3 L1.L.L 3123 ch l ra ae, f., del gr. 1 3 3 L1.L.

Permanent Link: http://ufdc.ufl.edu/UFE0021181/00001

Material Information

Title: Persistence of the Latin Accent in the Nominal System of Castilian, Catalan, and Portuguese
Physical Description: 1 online resource (520 p.)
Language: english
Creator: Wohlmuth, Sonia
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: catalan, portguese, prosody, spanish
Romance Languages and Literatures -- Dissertations, Academic -- UF
Genre: Romance Languages thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The Latin Stress Rule is well known and is the object of centuries-long study through various theoretical prisms, including, most recently, generative phonology, autosegmental phonology, metrical theory, and optimality theory. The basic facts are that primary accent of stressable words is never word final; it falls on the penultimate syllable if and only if that syllable is heavy. However, the loss of quantitative differences in the vowel system in the transition from Latin to Romance necessitates a new basis for assignment of the primary accent of a word. In generative phonology terms the Latin Stress Rule is opaque because the required environment for application of the rule may or may not be present. Optimality theory (OT) provides a mechanism for the study of diachronic phenomena that is not based on rules or the existence of a particular environment to trigger change. Rather, OT establishes ranked constraints to account for the relationship between input and output forms. The input/output forms selected for this study are Latin nouns that have correspondents in the three major Romance languages of the Iberian Peninsula, Castilian, Catalan, and Portuguese. The tension between faithfulness to the input form and conformity to preferred constraints, often universal in scope, is reflected in the set of active constraints and their ranking. The link between primary accent and the segment that displays the effect of that accent (through duration, quality, and intensity) is rarely broken. However, major differences among the languages of this study are evident in the treatment of final unstressed vowels of second and third declension nouns. Deletion of the unstressed final vowel in such cases, giving rise to patterns of ultimate accent, is an important innovation in Ibero-Romance and corresponds to a constraint that rewards right alignment of word edge and head foot. Frequency of occurrence of this new accentual pattern follows an East to West gradient with the highest rate of frequency in Catalan and the lowest in Portuguese. This study also shows that universal retention of the final unstressed vowel of first declension nouns does not correspond to a facile morphological explanation; rather, it manifests the desirability of a trochaic foot at right word edge with the familiar pattern of duple rhythm. Such a pattern is obtained when the rightmost syllable contains an optimal peak, not subject to elision, as is the case of /a/. Limitation of this study to Castilian, Catalan, and Portuguese nouns with common etymon provides an opportunity to view the effect of language specific constraints on common input forms through comparison of resulting outputs. Within this reduced linguistic microcosm it is possible to examine the role of positional prominence, optimal syllable architecture, alignment with word edges, and rhythmic preferences as constraints that influence outcomes. The divergence of Catalan with regard to preference for a monosyllabic head foot in nouns from the second and third declensions is expected. However, the areas of coincidence are of greater import and correspond to linguistic universals: light syllables are generally preferred; alignment of the head foot is at the right word edge; the optimal trochee consists of two light syllables. The single heavy syllable that emerges as variant of the head foot may be viewed as a potential disyllable with an empty nucleus, a concept reinforced by paragoge in poetry and music.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Sonia Wohlmuth.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Pharies, David A.
Local: Co-adviser: Wiltshire, Caroline R.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0021181:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021181/00001

Material Information

Title: Persistence of the Latin Accent in the Nominal System of Castilian, Catalan, and Portuguese
Physical Description: 1 online resource (520 p.)
Language: english
Creator: Wohlmuth, Sonia
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: catalan, portguese, prosody, spanish
Romance Languages and Literatures -- Dissertations, Academic -- UF
Genre: Romance Languages thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The Latin Stress Rule is well known and is the object of centuries-long study through various theoretical prisms, including, most recently, generative phonology, autosegmental phonology, metrical theory, and optimality theory. The basic facts are that primary accent of stressable words is never word final; it falls on the penultimate syllable if and only if that syllable is heavy. However, the loss of quantitative differences in the vowel system in the transition from Latin to Romance necessitates a new basis for assignment of the primary accent of a word. In generative phonology terms the Latin Stress Rule is opaque because the required environment for application of the rule may or may not be present. Optimality theory (OT) provides a mechanism for the study of diachronic phenomena that is not based on rules or the existence of a particular environment to trigger change. Rather, OT establishes ranked constraints to account for the relationship between input and output forms. The input/output forms selected for this study are Latin nouns that have correspondents in the three major Romance languages of the Iberian Peninsula, Castilian, Catalan, and Portuguese. The tension between faithfulness to the input form and conformity to preferred constraints, often universal in scope, is reflected in the set of active constraints and their ranking. The link between primary accent and the segment that displays the effect of that accent (through duration, quality, and intensity) is rarely broken. However, major differences among the languages of this study are evident in the treatment of final unstressed vowels of second and third declension nouns. Deletion of the unstressed final vowel in such cases, giving rise to patterns of ultimate accent, is an important innovation in Ibero-Romance and corresponds to a constraint that rewards right alignment of word edge and head foot. Frequency of occurrence of this new accentual pattern follows an East to West gradient with the highest rate of frequency in Catalan and the lowest in Portuguese. This study also shows that universal retention of the final unstressed vowel of first declension nouns does not correspond to a facile morphological explanation; rather, it manifests the desirability of a trochaic foot at right word edge with the familiar pattern of duple rhythm. Such a pattern is obtained when the rightmost syllable contains an optimal peak, not subject to elision, as is the case of /a/. Limitation of this study to Castilian, Catalan, and Portuguese nouns with common etymon provides an opportunity to view the effect of language specific constraints on common input forms through comparison of resulting outputs. Within this reduced linguistic microcosm it is possible to examine the role of positional prominence, optimal syllable architecture, alignment with word edges, and rhythmic preferences as constraints that influence outcomes. The divergence of Catalan with regard to preference for a monosyllabic head foot in nouns from the second and third declensions is expected. However, the areas of coincidence are of greater import and correspond to linguistic universals: light syllables are generally preferred; alignment of the head foot is at the right word edge; the optimal trochee consists of two light syllables. The single heavy syllable that emerges as variant of the head foot may be viewed as a potential disyllable with an empty nucleus, a concept reinforced by paragoge in poetry and music.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Sonia Wohlmuth.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Pharies, David A.
Local: Co-adviser: Wiltshire, Caroline R.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0021181:00001


This item has the following downloads:


Full Text

PAGE 1

PERSISTENCE OF THE LATIN ACCENT IN THE NOMINAL S YSTEM OF CASTILIAN, CATALAN AND PORTUGUESE By SONIA RAMREZ WOHLMUTH A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2008 1

PAGE 2

2008 Sonia Ram rez Wohlmuth 2

PAGE 3

To the m emory of my parents, Manuel D. Ramrez and Estelle Lpez Ramrez. 3

PAGE 4

ACKNOWL EDGMENTS I would like to thank the Univ ersity of Florida for fundi ng through the University of Florida Alumni Graduate Fellowshi p initiative which enabled me to become a full time student while completing coursework toward the doctora te. My home institution, the University of South Florida, where I am a full time instructor, permitted flexible scheduling of my classes and reduced employment during the time I was taking classes. As I began the dissertation writing phase I received a Personnel Development Leave for spring semester 2008 at the University of South Florida which gave me the necessary impetu s to write the preliminary chapters and begin data analysis. Coursework at the University of Florida provided a good foundation for the work on my dissertation. I would like to ac knowledge perspectives gained th rough coursework in historical French linguistics with Jean Ca sagrande and George Diller, and in particular coursework taken with members of my dissertation committee: Hispanic lingu istics with David Pharies; phonological theory with Caroline Wiltshire; histor ical and Indo-European linguistics with Gary Miller; and applied linguistics w ith Joaquim Camps. Interaction with fellow students with whom I share common interests provided an opportunity to test ideas and pursue new directions. I am grateful for the ongoing dialog wi th Gary Baker and the path forged by his 2004 dissertation on palatal phenomena in Spanish phonology. I would like to express my appreciation to my husband, Enrique, for his unwavering support, encouragement, and acceptance of my c ontinued distraction as concomitant of my pursuit of educational goals. My son, Sebastia n, has been a willing participant in conversations on various phonological phenomena and made me rea lize that enthusiasm for my discipline can be contagious. 4

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........8 LIST OF FIGURES.......................................................................................................................14 ABSTRACT...................................................................................................................................20 CHAPTER 1 INTRODUCTION: TH E PROSODIC WORD.....................................................................22 Approaches to Word Level Accent.........................................................................................22 Metrical Theory and Its Precursors.........................................................................................23 Early Metrical Theory.......................................................................................................... ...24 Autosegmental Theory........................................................................................................... .33 Metrical Grid..........................................................................................................................35 Metrical Theory and Language Typology..............................................................................37 Moraic Theory.................................................................................................................39 Extrametricality............................................................................................................... 42 Optimality Theory.............................................................................................................. ....44 Architecture of the Syllable.............................................................................................46 Syllable onset...........................................................................................................47 Syllable coda............................................................................................................49 Syllable peak (nucleus)............................................................................................51 Onset and coda clusters............................................................................................52 Faithfulness and Repair Strategies..................................................................................53 Word Accent....................................................................................................................56 Prosodic Foot...................................................................................................................57 Alignment Constraints.....................................................................................................59 Optimality Theory, Multiple Outputs, and Analogy..............................................................60 2 THE NATURE OF THE LATIN ACCENT..........................................................................68 Polemic of Pitch versus Stress................................................................................................6 8 Evidence of Prosodic Change.................................................................................................68 Early Latin..............................................................................................................................73 Syncope and Accent........................................................................................................74 Iambic Shortening and Cretic Shortening.......................................................................81 Vowel Reduction.............................................................................................................85 Treatment of Words of Greek Origin..............................................................................87 Strategies of Moraic Preservation...........................................................................................90 Classical Latin Accent: From the Me trical Grid to the OT Tableau.....................................94 5

PAGE 6

Learnability of the Latin Accent.............................................................................................98 The Transition from Classical Latin to Late Latin.................................................................99 3 FROM QUANTITY SENSITIVE TO STRESS ACCENT..................................................101 Chronology of Latin.............................................................................................................101 Stress Accent as Catalyst of Phonological Change..............................................................103 Ambisyllabification and Mora Sharing.........................................................................119 Stop + liquid clusters in Romance.................................................................................123 Weakening of Consonants in Coda Position.................................................................125 Stop + liquid clusters in Romance.................................................................................127 Apocope.........................................................................................................................129 Hiatus and Onset Glides................................................................................................130 Motivation for Vowel Loss...........................................................................................133 4 ACCENTUAL PATTERNS IN THE NOMINAL SYSTEM OF LATIN: OUTCOMES OF THE FIRST DECLENSION..........................................................................................139 OT Constraints and the Latin Accent...................................................................................139 The Latin Nominal System...................................................................................................141 Construction of the Data Set.................................................................................................145 An Overview of the First Declension...................................................................................147 Two-Syllable Nouns.............................................................................................................147 Three-syllable Nouns............................................................................................................164 First Declension Trisyllables with Penultimate Accent................................................165 First Declensions Trisyllables with Antepenultimate Accent.......................................183 Four-syllable Nouns..............................................................................................................194 Four-syllable Nouns with Penultimate Accent..............................................................195 Four-syllable Nouns with Antepenultimate Accent......................................................213 Five-syllable Nouns..............................................................................................................221 Five-syllable Nouns with Penultimate Accent..............................................................221 Five-syllable Nouns with Antepenultimate Accent.......................................................227 Summary of Prosodic Patterns in No uns from the First Declension....................................230 5 VOWEL LOSS AND THE RISE OF ULTI MATE ACCENT IN IBERO-ROMANCE.....233 Synchronic and Diachronic Vowel Reduction.....................................................................233 Second/Fourth Declension Nouns........................................................................................234 Third Declension Nouns.......................................................................................................235 Two-syllable Nouns..............................................................................................................237 Second and Fourth Declension Disy llables with Penultimate Accent..........................238 Third Declension Disyllables with Penultimate Accent................................................250 Three-syllable Nouns............................................................................................................259 Trisyllables with Penultimate Accent............................................................................259 Trisyllables with accent on a pe nultimate HC type syllable..................................261 Trisyllables with penultima te accent on HV type syllable.....................................272 Trisyllables with Antepenultimate Accent....................................................................280 6

PAGE 7

Four-syllable Nouns ..............................................................................................................286 Four-syllable Nouns with Penultimate Accent..............................................................286 Four-syllable nouns with HC type ini tial syllable and HV type penultimate syllable................................................................................................................292 Four-syllable nouns with HV or L t ype initial syllable and HV type penultimate syllable............................................................................................295 Four-syllable Nouns with Antepenultimate Accent......................................................301 Five-syllable Nouns..............................................................................................................304 Five-syllable Nouns with Penultimate Accent..............................................................304 Five-syllable Nouns with Antepenultimate Accent.......................................................315 Summary of Effect of Apocope on Prosodi c Outcomes of Nouns from Declensions 2/4 and 3.....................................................................................................................321 6 PREFERRED PROSODIC TEMPL ATES: CONCLUSIONS............................................323 Patterns with Penultimate Accent and /-a/ Class Marker.....................................................323 Patterns with Antepenultimate Accent and /-a/ Class Marker..............................................331 Patterns with Penultimate Accent and /-o/ Class Marker or No Class Marker....................338 Patterns with Antepenultimate Accent and /o/ Class Marker or No Class Marker.............348 APPENDIX A APPENDIX PROBI..............................................................................................................3 67 B APPLICATION OF PERFECT GRID.................................................................................370 C VERSIFICATION: FROM METER TO RHYTHM..........................................................374 D DATABASE OF NOUNS IN CATALAN, CASTILIAN, AND PORTUGUESE WITH COMMON LATIN ETYMON.............................................................................................382 LIST OF REFERENCES.............................................................................................................503 BIOGRAPHICAL SKETCH.......................................................................................................520 7

PAGE 8

LIST OF TABLES Table page 1-1 Parameters for the construction of a metric foot................................................................37 1-2 Edge Marking Parameter in Latin......................................................................................44 1-3 Possible syllable types.................................................................................................... ...46 1-4 Typology of light and heavy syllables...............................................................................50 1-5 Sonority scale of phonological segments...........................................................................52 1-6 Ranking of ONSET and PARSE above FILL.........................................................................54 1-7 Foot form constraints for Latin..........................................................................................57 1-8 Alignment constraints in Latin...........................................................................................60 1-9 Free variation with unordered OT constraints...................................................................62 1-10 Foot pattern of hypocoristics in Portuguese, Spanish, and Catalan...................................65 2-1 Characteristics of stress-tim ed and syllable-timed languages...........................................70 2-2 Syllabic weight of the dactylic hexameter.........................................................................72 2-3 Early Latin syncope........................................................................................................ ...75 2-4 Pre-classical Latin word initial accent...............................................................................77 2-5 Early Classical Latin word accent vs. Classical Latin word accent...................................78 2-6 Iambic Shortening in Early Classical Latin.......................................................................83 2-7 Word Accent and cretic shortening....................................................................................85 2-8 Prosodic configurations in wo rds of two and three syllables............................................95 2-9 Parsing of three-sy llable end of word................................................................................97 3-1 Latin historical periods....................................................................................................101 3-2 Syncope of post-toni c short vowels in Latin....................................................................104 3-3 Cases of syncope in the Appendix Probi.........................................................................106 3-4 Changes in prosodic form as a result of syncope.............................................................108 8

PAGE 9

3-5 Vulgar Latin constraints on accent..................................................................................109 3-6 Evaluation of faithful and syncope candidates................................................................113 3-7 Post-tonic syncope......................................................................................................... ..114 3-8 Degree of difference in sonority be tween medial consonants (Geisler)..........................118 3-9 Resolution of medial consona nt clusters produced by syncope.......................................118 3-10 Frequency of occurrence of La tin medial clusters: stop + liquid.....................................123 3-11 Ibero-Romance reflexes of Latin stop + liquid clusters...................................................124 3-12 Diphthongization before stop + liquid clusters in Romance languages...........................128 3-13 Ibero-romance reflexes of medial consonant + yod sequences......................................132 3-14 Vowel loss and faithfulness constraints...........................................................................134 3-15 Post-tonic vowel loss in Early Western Romance...........................................................134 3-16 Latin words of 2 to 5 syllables marked for possible syncope..........................................137 4-1 Nominal inflection in Latin..............................................................................................14 1 4-2 Distribution of H/L syllable type in first syllable of Latin 2-syllable nouns of the first declension........................................................................................................................152 4-3 Contingency table for H/L in first syll able of Latin first declension disyllables.............152 4-4 Contingency table for H/L in first syllable of Catalan nouns from Latin first declension disyllables......................................................................................................154 4-5 Contingency table for H/L in first syll able of Castilian nouns from Latin first declension disyllables......................................................................................................155 4-6 Contingency table for H/L in first syll able of Portuguese nouns from Latin first declension disyllables......................................................................................................155 4-7 Preservation of HC initial syllables in Catalan, Castil ian, and Portuguese nouns from Latin first declension disyllables.....................................................................................155 4-8 Accented syllable in nouns from Latin first declension disyllables with HV initial syllable.............................................................................................................................160 4-9 Outcome of Latin firs t declension ('L) nouns...........................................................160 4-10 Duration of vowels in European Portuguese in stressed CV syllables............................162 9

PAGE 10

4-11 Heavy/Light (H/L) outcomes of accented syllables in Catalan, Castilian, and Portuguese based on nature of tonic syllable in Latin fi rst declension disyllables..........163 4-12 Contingency table for H/L outcomes in tonic syllable from (HC)('HC), (HV)('HC), and L('HC) first declension nouns..................................................167 4-13 Distribution of heavy/light syllables in out comes of first declension trisyllables with penultimate accent of the type (HC)('HC)................................................................167 4-14 Tableau for output of first declension tr isyllable with penultimate accent on HC type syllable.............................................................................................................................171 4-15 Tableau for output of first declension trisyllable with penultimate accent on HC type syllable with PARSEand ALIGNR constraints...............................................................172 4-16 Tableau for output of first declension trisyllable with penultimate accent on HV type syllable with PARSEand ALIGNR constraints...............................................................175 4-17 Distribution of heavy/light syllables in outcomes from first declension trisyllable patterns (HV)('HV) and (L)('HV)......................................................................176 4-18 Resolution of hiatus in output of Latin first declension trisyllables................................177 4-19 Tableau for output of first declension trisyllable with penultimate accent on HV type syllable with PARSEand ALIGNR constraints...............................................................180 4-20 Tableau for selection of hiat us or rising diphthong in Catalan........................................182 4-21 Tableau for selection of falling diphthong in Catalan.....................................................183 4-22 Prosodic templates of first declension trisyllables with antepenultimate accent.............183 4-23 Outcomes of syncope in first declen sion trisyllables with antepenultimate accent: ('HV)L type...............................................................................................................18 4 4-24 Output of first declension ('HV)L.............................................................................185 4-25 Output of first declension ('L.L)...............................................................................187 4-26 Output of first declension ('HV)L and ('L.L).....................................................188 4-27 Output of first declension ('HC)L.............................................................................190 4-28 Contingency table for H/L outcome in Catalan, Castilian, and Portuguese for the accented syllable from ('HC)L and ('HC)..........................................................192 4-29 Tableau to assign accent in Fijian loanwords..................................................................194 4-30 Prosodic templates for first declension four-syllable nouns............................................195 10

PAGE 11

4-31 Possible heavy/light syllable configurations in four-syllable first declension nouns with penultimate accent and HC1 penultimate syllable...................................................196 4-32 Nature of first three syllables in outcomes of first declension L.L.HC1.L and L.HV.HC1.L inputs..........................................................................................................197 4-33 Proto-Catalan sonority ranking........................................................................................19 7 4-34 Phonotactic restrictions for coda s in Catalan, Castilian, and Portuguese........................199 4-35 Tableau for first decl ension input type L.L.HC1.L.........................................................200 4-36 Possible heavy/light syllable configurat ions in 4-syllable first declension nouns with penultimate accent and HV1 penultimate syllable...........................................................201 4-37 Cases of syncope in 4-syllable first declension nouns with penultimate accent..............202 4-38 Nature of first three syllables in outcomes of first declension HC.HC.HV1.L...............204 4-39 Nature of the first three syllables in outcomes of first decl ension HC.HC.HV1.L and HC.L.L1.L........................................................................................................................206 4-40 Nature of the first three sylla bles in outcomes of first declension HV/L.HV/L.HV/L1.L......................................................................................................211 4-41 Nature of the first three syllables in out comes of all first declension tetrayllables with penultimate accent...........................................................................................................21 2 4-42 Syllable count in output of X.HC1.L.L............................................................................214 4-43 Outcomes of X.HC1.L.L first declension nouns by syllable type...................................215 4-44 Tableau for first declen sion input type HC.HC.L.L........................................................216 4-45 Outcomes of X.L1.L.L first declension nouns by type of accented syllable and word level syllable count..........................................................................................................2 17 4-46 Outcomes of five-syllable firs t declension nouns with penultimate accent.....................226 4-47 Output of first declension pentasyllables with penultimate accent by language and syllable count...................................................................................................................227 4-48 Outcomes of five-syllable first d eclension nouns with antepenultimate accent..............228 5-1 Vowel sonority hierarchy.................................................................................................237 5-2 Preservation of HC initial syll ables in Catalan, Castilian, and Portugue se nouns from Latin second and f ourth declension disyllables............................................238 11

PAGE 12

5-3 Prosodic outcomes of Latin disyllables of the second and fourth declension with initial heavy syllable........................................................................................................239 5-4 Differences in constraints govern ing word accent in Latin and Spanish.........................240 5-5 Constraints on complex codas in Catalan........................................................................241 5-6 Constraint against vowel deletion in Early Romance: R-ANCHOR-V..............................242 5-7 Proto-Catalan sonority ranking........................................................................................243 5-8 Nature of accent bearing syllables derived from Latin 2nd and 4th declension disyllables with L or HV initial syllable........................................................246 5-9 Accented syllable by type in the output of 3rd declension disyllabic nouns with initial HC syllable............................................................................................................250 5-10 Comparison of prosodic outcomes of Latin disyllables with HC initial syllable from the 3rd and 2nd/4th declensions.......................................................................................251 5-11 Accented syllable by type in the output of 3rd declension disyllabic nouns with initial HV/L syllable........................................................................................................25 4 5-12 Comparison of prosodic outcomes of Latin disyllables with HV and L initial syllable from 3rd and 2nd/4th declensions...................................................................................256 5-13 Constraints governing apocope in Ibero-Romance..........................................................260 5-14 Output of second/fourth declension (HC)('HC) by number of syllables and nature of accented syllable...............................................................................................261 5-15 Comparison of heavy/light syllables in initial and accent beari ng syllables in output of second/fourth declension (HC)('HC)....................................................................265 5-16 Tableau for output of declension 2/4 tr isyllable with penultimate accent on HC type syllable.............................................................................................................................266 5-17 Tableau for trisyllables with penultimate accent (declensions 2/4 and 3).......................267 5-18 Comparison of heavy/light syllables in initial and accent beari ng syllables in output of third declension (HC)('HC)..................................................................................268 5-19 Prosodic outcomes of HC('HV) in declension 2/4 nouns........................................274 5-20. Distribution of prosodic templates in output of third d eclension tetrasyllables with penultimate accent...........................................................................................................30 0 5-21 MAX/IO constraints related to the output of suffix -ione .................................................312 12

PAGE 13

5-22 Distribution of heavy/light syllables in pentasyllables with ione(m) suffix...................313 6-1 Distribution of initial heavy syllables in disyllables and trisyllables from first declension nouns..............................................................................................................327 6-2 Syllable count in outputs of first declension trisyllables with antepenultimate accent...331 6-3 Distribution of three and four-syllabl e outputs from first declension tetrasyllables with antepenultimate accent.............................................................................................333 6-4 Hierarchy of preferred vowels in non-prominent positions.............................................342 6-5 Distinctive features in a seven vowel system..................................................................343 6-6 OT tableau for final vowel deletion in Catalan................................................................344 6-7 OT tableau for vowel redu ction in seven vowel systems................................................345 6-8 Distribution and source of one and two-syllable nouns...................................................352 6-9 Distribution and source of thr ee-syllable nouns with penultimate accent.......................356 6-10 Distribution and source of threesyllable nouns with antepenultimate accent................358 6-11 Distribution and source of thr ee-syllable nouns with ultimate accent.............................360 6-12 Distribution and source of fou r-syllable nouns with penultimate accent........................361 6-13 Distribution and origin of four-s yllable words with antepenultimate accent..................363 B-1 Treatment of on the metrical grid................................................................................372 C-1 Duple rhythm in the Iberian ballad tradition....................................................................380 13

PAGE 14

LIST OF FI GURES Figure page 1-1 Branching trees with s/w nodes.........................................................................................25 1-2 Illicit configuration of a strong branch..............................................................................25 1-3 Bracketing of contiguous elements (Liberman).................................................................26 1-4 Metrical tree with labeling and assignment of the feature [stress]..................................26 1-5 Insertion of foot boundary.............................................................................................. ...27 1-6 Application of LCPR...................................................................................................... ...28 1-7 Grid alignment showing stress deletion through application of RPPR..............................29 1-8 Selkirks model of syllable constituents............................................................................31 1-9 Stress foot configurations for English (Selkirk)................................................................31 1-10 Prosodic hierarchy of the word..........................................................................................32 1-11 Application of prosodic category labels to differentiate word accent...............................32 1-12 Autosegmental repres entation of a prosodic word.............................................................34 1-13 Metrical grid.......................................................................................................................36 1-14 The revised pr osodic hierarchy..........................................................................................40 1-15 Tree structures for heavy and lig ht syllables based on moraic count................................40 1-16 Selkirks superfoot.............................................................................................................41 1-17 Latin prosodic word...........................................................................................................42 1-18 The intrinsic structure of the syllable (Hooper 1976)........................................................48 1-19 Universal strength hierarc hy for consonants (Hooper 1976).............................................49 1-20 Nested prosodic constituents..............................................................................................56 1-21 Overlapping constraints.................................................................................................. ...63 2-1 Typology of accent (van Coetsem)....................................................................................69 2-2 Alignment of scansion and word accent in Latin dactylic hexameter...............................71 14

PAGE 15

2-3 Iambic shortening........................................................................................................ .......83 2-4 Prosodic repair with REMOVE......................................................................................93 2-5 The prosodic hierarchy.................................................................................................... ..94 3-1 Parsing of candidates for syncope in Classical Latin and Vulgar Latin..........................107 3-2 Raddoppiamento sintattico as moraic compensation.......................................................121 3-3 Feature spreading in cons onants in contact with /j/.........................................................132 4-1 Likelihood of apocope by declension class and language...............................................145 4-2 Permissible syllable structures in Latin words of 2-syllables..........................................148 4-3 Extrasyllabic /s/ in Latin word initial and word final position........................................151 4-4 Outcomes of initial HC syllables from Latin first declension disyllables in Catalan, Castilian, and Portuguese.................................................................................................156 4-5 Coincidence of subsets of L type initial syllable in Catalan, Castilian, and Portuguese resulting from HC first syllable in Latin first declension disyllables..............................158 4-6 Distribution of heavy/light syllable s in tonic syllable of nouns from Latin first declension disyllables......................................................................................................164 4-7 Comparison of prosodic patte rns in three-syllable Latin nouns......................................165 4-8 Comparison of distribution of heavy/light in S3 and S2 from first declension trisyllables with penultimate and antepenultimate heavy syllables.................................169 4-9 Percentage of heavy syllables in word initial syllable and tonic syllable from first declension ('HC) and (HC)('HC)........................................................................173 4-10 Percentage distribution of heavy/light syllables in outcomes from first declension trisyllable patterns (HV)('HV) and (L)('HV).....................................................177 4-11 Comparison of 3 syllable and 2 syllable results of first declension ('HV)L input....186 4-12 Percentage distributi on of outcomes of first declension ('HV)L and ('L.L)......189 4-13 Percentage distribut ion of outcomes of first declension ('HC)L...............................190 4-14 Percentage distribut ion of outcomes of first declension ('HC)L...............................191 4-15 Percentage of heavy/light syllables in word initial syllable resulting from first declension ('HC)L, (HC)('HC), and ('HC)..................................................193 15

PAGE 16

4-16 Hasse diagram of OT constraints rela ted to primary accent in Fijian loanwords............193 4-17 Output of first declension HC.HC.HV1.L and HC.L.L1.L by syllable type...................205 4-18 Distribution of syllable types in the output of all first declension tetrasyllables with penultimate accent...........................................................................................................21 3 4-19 Heavy/light syllable configurations for first declension tetrasyllables with antepenultima te accent.....................................................................................................214 4-20 Distribution of syllable types in outcomes of first d eclension pentasyllables with penultimate accent. .........................................................................................................22 6 4-21 Distribution of syllable types in outcomes of first d eclension pentasyllables with antepenultima te accent.....................................................................................................230 5-1 Vowel quadrilateral for 7 vowel systems with [ ] in nonprominent positions................236 5-2 Nature of stressed syllable in Catalan, Castilian, and Portuguese nouns from Latin 2nd and 4th declension disyllables with L or HV initial syllable....................................248 5-3 Prosodic outcomes of all declensi on 1 and declensions 2/4 disyllables..........................249 5-4 Comparison of percent values of pr osodic outcomes of 2nd/4th and 3rd declension disyllables with HC initial syllable..................................................................................252 5-5 Comparison of percentage values of prosodic outcomes of declensions 2/4...................257 5-6 Percentage distribution of prosodic outcomes of all de clension 2/4 and declension 3 disyllables........................................................................................................................258 5-7 Prosodic templates for three-syllable nouns....................................................................259 5-8 Comparison of syllabic count and nature of the syllable with primary accent in outcomes of second/fourth declension (HC)('HC)....................................................263 5-9 Comparison of syllable count and nature of the syllable with primary accent in outcomes of third declension (HC)('HC)..................................................................269 5-10 Outcomes of HV('HC) and L('HC) in declension 2/4 nouns: Syllable count and nature of accented syllable........................................................................................270 5-11 Outcomes of HV('HC) and L('HC) in declension 3 nouns relative to syllable count and nature of accented syllable..............................................................................272 5-12 Outcomes of HC('HV) in declensi on 2/4 nouns: Syllable count and nature of accented syllable.............................................................................................................. 273 16

PAGE 17

5-13 Relative frequency of 3-syllable and 2syllable templates resulting from declension 2/4 input HC('HV).....................................................................................................275 5-14 Outcomes of HC('HV) in declen sion 3 nouns: Syllable count and nature of accented syllable.............................................................................................................. 276 5-15 Outcomes of HV('HV) and L('HV) in declension 2/4 nouns: Syllable count and nature of accented syllable........................................................................................278 5-16 Outcomes of HV('HV) and L('HV) in declension 3 nouns: Syllable count and nature of accented syllable........................................................................................279 5-17 Outcomes of HC.L.L declension 2/4 nouns: Syllable count and nature of accented syllable (n=68).................................................................................................................281 5-18 Outcomes of HC.L.L declension 3 nouns: Syllable count and nature of accented syllable.............................................................................................................................282 5-19 Prosodic templates from second and fourth declension HV.L.L and L.L.L..................283 5-20 Prosodic templates from third declension HV.L.L and L.L.L.......................................285 5-21 Prosodic outcomes of second declension te trasyllables with HC type initial and penultimate syllables........................................................................................................28 7 5-22 Prosodic outcomes of third declension te trasyllables with penultimate accent and HC type initial and penul timate syllables...............................................................................287 5-23 Prosodic outcomes of second/fourth declension tetrasyllables of the type HV.X.'HC.X and L.X.'HC.X with penultimate accent....................................................289 5-24 Prosodic outcomes of third declension tetrasyllables of the type HV.X.'HC.L and L.X.'HC.L (n=13) with penultimate accent.....................................................................291 5-25 Prosodic outcomes of second/fourth declension tetrasyllables of the type HC.X.'HV.L with penultimate accent..............................................................................293 5-26 Prosodic outcomes of third declension te trasyllables of the type HC.X.'HV.L with penultimate accent...........................................................................................................29 3 5-27 Prosodic outcomes of declension 2/4 te trasyllables of the type HV/L.HV/L.'HV.L with penultimate accent...................................................................................................296 5-28 Prosodic outcomes of third declension tetrasyllables of the type HV/L.HC.'HV.L with penultimate accent...................................................................................................298 5-29 Prosodic outcomes of third declension tetrasyllables of the type HV/L.HV/L.'HV.L with penultimate accent...................................................................................................299 17

PAGE 18

5-30 Prosodic outcomes of declension 2/4 te trasyllables with antepenultimate accent...........302 5-31 Prosodic outcomes of declension 3 tetr asyllables with antepenultimate accent..............303 5-32 Outputs of declension 2/4 pentasyllables with penultimate accent.................................307 5-33 Portuguese outputs of third declen sion pentasyllabic nouns with suffixes t te(m) and t d .................................................................................................................................308 5-34 Catalan and Castilian outputs of third decl ension pentasyllabic nouns with suffixes t te(m) and t d ............................................................................................................309 5-35 Realignment of syllables, moras, and segments in outputs of suffix -ione(m) in Catalan, Castilian, and Portuguese...................................................................................311 5-36 Outcomes of third declensi on pentasyllables with suffix -i ne(m) .................................313 5-37 Outcomes of third declensi on pentasyllables with suffix -re(m) ...................................314 5-38 Outcomes of second declension pentasyllables with antepenultimate accent and heavy initial syllable........................................................................................................317 5-39 Outcomes of second declension pentasyllables with antepenultimate accent and LH initial syllables.............................................................................................................. ...319 5-40 Outcomes of second declension pentasyllables with antepenultimate accent and all light syllables...................................................................................................................321 6-1 Prosodic outcomes of all 2-syllable and 3-syllable first declension nouns with penultimate accent...........................................................................................................32 3 6-2 Outcomes of all first declension tetrasyllables with penultimate accent.........................327 6-3 Outcomes of all first declension pe ntasyllables with penultimate accent........................329 6-4 Distribution of heavy and light syllables in tonic and word initial position in nouns derived from the first declension.....................................................................................330 6-5 Prosodic outcomes of all 3-syllable first declension nouns with antepenultimate accent ........................................................................................................................ ......332 6-6 Prosodic outcomes of all 4-syllable first declension nouns with antepenultimate accent ........................................................................................................................ ......334 6-7 Distribution of heavy/light syllables in initial and tonic positions of all nouns with antepenultimate accent from the first declension.............................................................337 6-8 Distribution of heavy/light syllables in initial and tonic positions of all nouns with penultimate accent from the first declension...................................................................338 18

PAGE 19

6-9 Outcomes of disyllablic nouns with penultimate acent in declensions 2, 3, 4.................339 6-10 Prosodic templates from declension 2/4 trisyllables with penultimate accent................347 6-11 Prosodic templates from declension 3 trisyllables with penultimate accent....................348 6-12 Prosodic templates from declension 2/4 trisyllables with antepenultimate accent..........349 6-13 Prosodic templates from declension 3 trisyllables with antepenultimate accent.............350 6-14 Preferred templates for nouns of one and two syllables..................................................351 6-15 Preferred templates for nouns of th ree syllables with penultimate accent.......................355 6-16 Preferred templates for nouns of thr ee syllables with antepenultimate accent................357 6-17 Preferred templates for nouns of th ree syllables with ultimate accent............................359 6-18 Preferred template for nouns of four syllables with penultimate accent.........................360 6-19 Preferred template for nouns of f our syllables with antepenultimate accent...................362 B-1 Skeleton to grid associat ion of a superheavy syllable.....................................................370 B-2 Perfect Grid L R (Cairene Arabic)..............................................................................371 B-3 QS and stress assignment (Latin).....................................................................................371 C-1 Metrical analysis of a line of Saturnian verse..................................................................375 19

PAGE 20

Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy PERSISTENCE OF THE LATIN ACCENT IN THE NOMINAL SYSTEM OF CASTILIAN, CATALAN AND PORTUGUESE By Sonia Ramrez Wohlmuth December 2008 Chair: David A. Pharies Cochair: Caroline R. Wiltshire Major: Romance Languages The Latin Stress Rule is well known and is the object of centu ries-long study through various theoretical prisms, including, mo st recently, generative phonology, autosegmental phonology, metrical theory, and optimality theory. The basic facts are that primary accent of stressable words is never word final; it falls on the penultimate syllable if and only if that syllable is heavy. However, the loss of quantita tive differences in the vowel system in the transition from Latin to Romance necessitates a new basis for assignment of the primary accent of a word. In generative phonology terms the Latin Stress Rule is opaque because the required environment for application of the rule may or may not be present. Optimality theory (OT) provides a mechan ism for the study of diachronic phenomena that is not based on rules or the existence of a particular environm ent to trigger change. Rather, OT establishes ranked constraints to account for the relationship be tween input and output forms. The input/output forms selected for this study are Latin nouns that have correspondents in the three major Romance languages of the Iberian Pe ninsula, Castilian, Catalan, and Portuguese. The tension between faithfulness to the input form and conformity to preferred constraints, often universal in scope, is reflected in the set of active constrai nts and their ranking. The link 20

PAGE 21

between prim ary accent and the segment that displa ys the effect of that accent (through duration, quality, and intensity) is rarely broken. However, major differences among the languages of this study are evident in the treatment of final uns tressed vowels of sec ond and third declension nouns. Deletion of the unstressed final vowel in such cases, giving rise to patterns of ultimate accent, is an important innovation in Ibero-Roman ce and corresponds to a constraint that rewards right alignment of word edge and head foot. Frequency of occurrence of this new accentual pattern follows an East to West gradient with the highest rate of fre quency in Catalan and the lowest in Portuguese. This study also shows th at universal retention of the final unstressed vowel of first declension nouns does not correspond to a facile morphological explanation; rather, it manifests the desirability of a trochaic foot at right word edge with the familiar pattern of duple rhythm. Such a pattern is obtained wh en the rightmost syllable contains an optimal peak, not subject to elision, as is the case of /a/. Limitation of this study to Castilian, Cata lan, and Portuguese nouns with common etymon provides an opportunity to view the effect of language specif ic constraints on common input forms through comparison of resulting outputs. W ithin this reduced linguistic microcosm it is possible to examine the role of positional promin ence, optimal syllable architecture, alignment with word edges, and rhythmic preferences as constraints that influence outcomes. The divergence of Catalan with regard to preference for a monosyllabi c head foot in nouns from the second and third declensions is expected. However, the areas of coin cidence are of greater import and correspond to linguistic un iversals: light syllables are generally preferred; alignment of the head foot is at the right word edge; the op timal trochee consists of two light syllables. The single heavy syllable that emerges as variant of the head foot may be viewed as a potential disyllable with an empty nucleus, a concept re inforced by paragoge in poetry and music. 21

PAGE 22

CHAP TER 1 INTRODUCTION: THE PROSODIC WORD Approaches to Word Level Accent Although it is difficult to find a single, unive rsal definition of accent, there is crosslinguistic recognition of the prominence of a syllable within the word or lexical unit. The basis of that prominence, however, is rarely unidimens ional. Rather, various factors including pitch, duration, and intensity coalesce to place a syllable in relief (Beckman 1986; Von Coetsem 1996; Crosswhite 2001; Kager 1995). Studies of word le vel accent, cross-linguistic as well as language specific, have offered insights into the typology of accent, its assignment at word level, including predictability, iterability of metric feet, and restricti ons imposed by morphology and phonology (Beckman 1986; Von Coetsem 1996; Delgado Martins 1982; Fox 2000; Harris 1992; Hayes 1995; Liberman and Prince 1977). This study examines the perseverance of the original locus of accent in nouns of Latin origin as they are reflected in Casti lian, Catalan, and Portuguese and the changing nature of that accent from a system of moraic trochees to one which derives its binarity from the presence of two-syllables de faulting to one heavy sy llable when it is not possible otherwise to build a prosodic foot at the right word edge. The selection of nouns for this study presents several advantages: they constitute a corpus of sufficient size (approximately 3,600 items that were found to have correspondences in all three Ibero-Romance languages) and are by and large unencumbered by morphological variants in the declension cases that pass into Ro mance, primarily accusative and occasionally nominative, that would alter co mputation of the primary accent. Analysis of the Latin input forms and the three-way output forms in Castilian, Catalan, and Portuguese is effected using core principles of metrical theory in the framework of optimality theory (hereafter OT) in order to describe the nature of the Latin accent as well as project differences in constraints and priorities 22

PAGE 23

of those constraints in the thre e m ajor Romance languages of the Iberian peninsula. Treatment of word accent prior to the rise of metrical theory is briefly examined followed by discussion of metrical and optimality theoretic approaches to word level assignment of accent. Since neither classical Latin nor the Romance languages form ally depend on tone for prominence of the primary accent the terms stress and accent are at times used interchangeably in the literature that discusses the prosody of these languages. Description of word level prosodic phenomena be gins as a footnote or short chapter in the works that form the foundation of modern phonologi cal theory. Wh ile the earliest attempts to describe the processes that determine primar y and secondary stress assignment are often incomplete and at times misleading they evince the realization that tr eatment of accent cannot follow the precepts of segmental phonology. Assi gnment of accent is obligatory for prosodic words and its placement corresponds to language-sp ecific criteria. These include among others, recognition of positions of prom inence within the word, avoida nce of sequential and equal accents, and the role of syllabic structure in attrac ting or repelling accent. A brief review of the literature provides insights into the different ap proaches that form the historical backdrop for optimality theoretic approaches to word prosody. Metrical Theory and Its Precursors Major advances in phonological theory, includ ing the nature of word level accent, were stimulated by the publication of The Sound Pattern of English (henceforth, SPE; Chomsky and Halle 1968). The rules governing assignment of word level accent in English, as seen in the SPE Main Stress Rule and its corollaries (Choms ky and Halle 1968, 240-245), are extraordinarily complex. The context portion of such rules at times conflates derivational and phonological processes in order to produce the desired output. Despite their unwieldiness the SPE stress rules demonstrate sensitivity to weight, the role of le xical categories, an awar eness of differing stress 23

PAGE 24

patterns in words of Latin and Germanic origin, the effect of morphem e boundaries, constraints regarding stress clash, and stress subordination. There is also a recognition that stress is unlike other distinctive features in that its domain ex tends over sequences that are longer than a word (Chomsky and Halle 1968, 68). However, absent in the SPE approach to word accent are formulations capable of capturing broad patterns of regularity such as which elements may bear stress, the direction of computation of stress, and the required distance between stressed elements. The several elements that work in tandem to determine st ress are dispersed through various rules such as the Main Stress Rule (Chomsky and Halle 1968, 240-245), the Alternating Stress Rule (1968, 77-79), and the Stress Adjustment Rule (1968, 84) as well as related rules that affect segments such as the Vo wel Reduction Rule (1968, 100-126). Early Metrical Theory The work of Liberman (1979, based on his 1975 dissertation) and Liberman and Prince (1977) represented an important de parture from SPE in that stress was considered to be a matter of RELATIVE prominence among syllables, rather than as a degree of absolute prominence attached to each vowel (Hayes 1981, 1). Libe rman (1979, 207-208) assigned metrical values, s (strong) and w (weak), to pairs of syllables in binary branching tree structures. The binary feature [stress] is then assigned to each vowel dominated by s or w According to Liberman (1979, 207-208) the distinction strong/weak is assigned to nodes wher eas the feature [stress] applies to segments. In its simplest form, the metric tree, labeled R for the rhythmic unit, results in undesirable opacity. Although one would expect [ ] for both weak syllables in Figure 1-1, the weak syllable of example A (bomba st) retains its full value of [] while in B (ballast) it has been reduced to [ ] (pronunciations from OED Online s. v. bombast, ballast). Liberman and Prince (1977:264-265) note that while many two-syll able words follow the pattern of Figure1-1 24

PAGE 25

B this perfect correlation cannot be m aintained in general, sin ce a [+stress] vowel may well be metrically weak, as words like gymnast, raccoon show. R R s w s w [ bm bst] [ bl. l st] + + + A B Figure 1-1. Branching trees with s/w nodes. A) Weak node with [+] segment. B) Weak node with [-] segment. The more important metrical considerati on can be captured forma lly through a rule of this type (Liberman and Prince 1977, 265): (1.1) Stress Rule: If a vowel is s then it is [+stress]. (Liberman and Prince 1977) Since there are abundant counterexamples to the expected if a vowel is w then it is [-stress], Liberman and Prince suggest that this rule be considered rather a prin ciple of well-formedness for metrical structures whose main purpose is to disallow the output seen in the tree structure in Figure 1-2. *s | V [-stress] Figure 1-2. Illicit configur ation of a strong branch. In order to process the segments that ha ve now been designate d as strong and weak Liberman (1979, 209-217) introduces two further con cepts to account for the output of metrical rules: bracketing and labeling. Bracketing of segments proceeds from the beginning of the word from left to right creating a seri es of binary groups in which the leftmost member is strong. The 25

PAGE 26

configuration of Figure 1-3 is representati ve of languages where words, unencumbered by derivational or morphological constraints, routinely have initial stress. s s s s x x x x x Figure 1-3. Bracketing of con tiguous elements (Liberman). Bracketing allows the constr uction of a tree to which labeling correctly assigns s and w labels, with the strong element always appearing as the left daughter of a constituent. Finally, the segmental feature [stress], whose primar y role is determining vowel-reduction, is assigned (Liberman 1979:210). This segmental feature is a carryover from the SPE approach to analysis of accent. For English, assignment of [stress] to a vowel indicates that it will most likely be realized as [ ]. In the example shown in Figure 14, the word testament is correctly rendered through bracketing, labeling, and assignment of the feature [stress]. The value [-] here generates [ ] for the two unstressed vowels, or [ t s.t .m nt]. R s w s w tes t m nt + Figure 1-4. Metrical tree with labeling and assignment of the feature [stress]. 26

PAGE 27

The vowel labeled [+stress] in Figure 1-4 is also in a closed or heavy syllable, an environment that Halle and Vergnaud assert attracts stress (S elkirk 1980, 575). Bracketi ng is also subject to conditions imposed by foot boundaries. The f oot corresponds to a metrical algorithm superimposed on disyllabic constituents, one of which must be stressable. For example, in a three-syllable word in English there are two allowa ble configurations of the metrical tree, seen in Figure 1-5, A and B. However, C violates the constraint on inse rtion of a foot boundary because it divides a metrical unit (Liberman 1979:211-213). | X X | X | | X | X X | *| X | X | X | A B C Figure 1-5. Insertion of foot boundary. A) Left-branching tree. B) Right-branching tree. C) Violation of well-formedness. It can be readily seen that words like testament correspond to pattern A, a common pattern for English nouns of Latin origin (cf. argument, firmament, instrument ). On the other hand, pattern B is representative of deverbal nouns from a later En glish derivational process that preserves the original stress of the verb as in agreement Because argue is also a verb and argument follows pattern A it can be readily seen th at neither morphologica l analysis nor stress assignment are transparent for many such polymorphemic words in English. Patterns A and B reflect broad patterns of regularity but cannot account for deviations from the expected word accent which may be the result of factors such as relative chronology, analogy with other lexical items, or faithfulness to original stress for borrowed words. Liberman (1979, 212) is concerned not only w ith a description of word accent but also with the broader metrical properties of language that extend beyond word level. He recognizes 27

PAGE 28

that stress rules are of a fundamentally differe nt type from segmental rules, and thus have inherently different properties, which include the misleading appear ance of cyclical application (1979, 232). Liberman proposes two foot-boundary insertion rules and assume s that the first is universal while the second, although specific to English, is probably valid for many other languages as well. The first rule (1.2) is pr ecursor to the alignment concept employed in optimality theory. The second rule (1.3) indicates that when there are twosyllables, the first of which is strong or stressed, a foot boundary must be inserted to the left. The third rule (1.4) provides a basis for determination of relative strength among metrical units. (1.2) All # boundaries are foot boundaries. (Liberman 1979) (1.3) A foot boundary is in serted in the environment __ / [+] [-]. (Liberman 1979) (1.4) In any lexical metrical constituent [M N], N is strong if and only if it is complex. (Liberman 1979) Complex is defined as dominating non-termin al material (Liberman 1979, 213). The final formulation of the principles s een in the three rules above is the Lexical Category Prominence Rule (LCPR). (1.5) LCPR: In the configuration [N1 N2], N2 is strong if and only if it branches. (Liberman and Prince 1977, 270). Application of the LCPR is illustrated in Figure 1-6 where N2 is complex because it dominates elements that are lower in the hierarchy. Wd N1 N2 s w s w [ ri m nt] Figure 1-6. Application of LCPR 28

PAGE 29

The word agreement in Figure 1-6 Liberman and Prince (1977, 270) corresponds to a familiar pattern of deverbal nouns formed with the Latin instrumental suffix -mentum (Miller 2006, 78-84). However, there are also verbs such as augment, cement, complement, document with word final accent. Liberman recognizes that there are many exceptions to accentual patterns based on grammatical function as is the case of words formed with mentum as well as those that reflect French versus Germanic origin (1979, 304). Visualization of word level accent assignment as an iterative process (Liberman and Prince 1977, 298-304) is enhanced by use of the metrical grid. The baseline of the me trical grid is built on a one-to-one correspondence to syllables (Lib erman and Prince 1977, 315), that is, the grids terminals reflect the relative st rength of the corresponding syllables numbered from left to right beginning with the lowest level. This is form alized as the Relative Prominence Projection Rule (Liberman and Prince 1977, 316). (1.6) Relative Prominence Projection Rule (RPPR) (Liberman and Prince 1977) In any constituent on which the strong-w eak relation is defined, the designated terminal element of its strong subcons tituent is metrically stronger than the designated terminal element of its weak subconstituent. Utilizing the same example as in Figure 1-6, application of RPPR is illustrated in Figure 1-7. 6 *4 *5 1 2 3 a.gree.ment | | | w s s s R A 4 1 2 3 a.gree.ment | | | w s w s R B Figure 1-7. Grid alignment show ing stress deletion through application of RPPR. A) Input scansion. B) Output scansion. 29

PAGE 30

The initial input, seen in Figure 1-7A contains violations such as clashing (stress bearing adjacen t syllables or terminal elements) in contrast to the preferred pattern of alternating stress. Avoidance of non-alternating stresse s licenses repair strategies such as stress retraction or stress deletion (Prince 1983, 21). Kiparsky (1979, 424) ascr ibes these adjustments to the Rhythm Rule which formalizes the universal principle of eurhythmicity, the altern ation of stressed and unstressed syllables in a word or phrase. The bidimensional representation in Figure 1-7 captures the linear a rrangement of the SPE approach in the horizontal displa y of segments in the word, labeled weak or strong, while the vertical dimension, the metrical gr id, shows stress assigned to those el ements that can bear stress. The Liberman/Prince (1979, 1977) approach correctly predicts the locus of primary and secondary stresses in words (and ph rases) but does not develop a fu ll-fledged metrical theory. The Prosodic Hierarchy The word level Lexical Category Prominen ce Rule (LCPR) of Liberman and Prince leaves some unresolved questions which are ad dressed in Selkirk (1980 ). Beginning with a discussion of the English syllable and generation of a tree-like struct ure of the syllabic template for the word flounce, Selkirk (1980, 568-569) also incorporates the use of s/w markers. However, in this case the s/w distinction reflects the relative sonority of the components of the onset and rhyme (nucleus and coda). At the first level, the s/w distincti on designates the onset as weak in terms of sonority, or [+cons], while the rh yme is [+son]; all other s/w labels also reflect a sonority hierarchy. The syllable can be wit hout onset and without c oda but the strongest element ( s terminal dominated by s terminal), the nucleus, must be present and constitutes the stress bearing element. In Figure 1-8 the nucleus is the only [+vocalic] element in the complex 30

PAGE 31

monosyllable. Therefore, it is ma rked as strong within a higher level strong constituent. The structure here offers an intere sting approach to the problem of the superheavy syllable. s w s w w s s w s w | | | | | | f l a w n s Figure 1-8. Selkirks model of syllable constituents. Role of the Stress Foot Selkirks (1980, 570) prosodic hier archy posits a stress foot ( ) as the next level above the syllable, which as seen above, has its own sonority hierarchy. Fo r English there are two basic configurations for the stre ss foot, monosyllabic and disyllabic, as well as a third type, the stress superfoot (stray foot adjunction in later metr ical theory). The constituents labeled as weak in B and C in Figure 1-9 are stressless. s w s w A B C Figure 1-9. Stress foot configurations for English (Selkirk). A) Monosyllabic. B) Disyllabic. C) Superfoot. Selkirk has effectively shown that below the level of the prosodic word there are constituents that have their own hierarchical principles of or ganization. This is a si gnificant enhancement to the Lexical Category Prominence Rule of Libe rman and Prince which relates only to the prosodic word level. The hierar chy proposed by Selkirk (Figure 1-10) culminates at word level 31

PAGE 32

but builds on the lower level prom inence principles that operate at the level of the stress foot and the syllable. (Word) | (Stress Foot) | (Syllable) Figure 1-10. Prosodic hierarchy of the word. The three tier system presented he re is able to effectively solve the dilemma of the word pair considered above, bombast and ballast The Liberman/Prince analysis is given in Figure 1-11 alongside Selkirks (1980, 565) prosodic hierarchy. The addition of the stre ss foot layer provides an explanation for the phonetic difference betwee n the unstressed final syllables in the two words. Since the final syllable in bombast forms a stress foot it r eceives a secondary accent while the strong member of the binary branch carries the primary accen t. R s w [ bm bst] + + | s w | | s w [ bm bst] R s w [ bl. l st] + | s w [ bl. l st] A A1BB1 Figure 1-11. Application of prosodic category labels to differentiate word accent. A, B (Liberman/Prince analysis). A1, B1 (Selkirk analysis). Selkirk (1980, 574-575) cites the work of Halle and Vergnaud as further support for dispensing with the segmental feature [stress] in favor of a hierarchical metrical system that computes word level accent on the basis of unive rsal principles of prosody as well as language 32

PAGE 33

specif ic rules. In the Halle/Vergna ud analysis a potential stress bearing syllable in English is one with a branching rhyme consisting of a tense or long vowel, or a vowel followed by a coda. The final syllable of ballast does branch yet it would be labele d weak in the Halle/Vergnaud metrical tree (cf. Figure 1-11B). The fa ilure to project on the metrical plane is not easily explained without reliance on the stress f oot (cf. Figure 1-11, A1). Autosegmental Theory Autosegmental theory represents a significant departure from the SPE approach to accent and early metrical theory accounts. In terms of visual or spa tial representation of the prosodic word autosegmental theory uti lizes a multidimensional representa tion in opposition to the linear sequence of SPE and the bidimensional, hierarchical representation of metric al trees and grids. The foundation of the prosodic word Goldsmith (1990, 48-50) construes as the skeletal tier. The skeletal tier is a string of units or slots that could be represented as C (consonant), V (vowel), or simply as X without specifying the nature of th e segment. The phonological features associate to the CV elements on the skeletal tier. However, other segments may also associate with the CV tier. These constitute autosegments. The auto segmental tier was initially devised as a way of representing tone; tone often affects several segments includi ng discontinuous segments. In similar fashion, an autosegmental tier can be used to visualize th e interaction between phonological segments, syllable structure, and as signment of word accent. Following the model given by Goldsmith (1990, 195), an autosegmental representation of the Latin word b n t tem, fem. sg. acc. (goodness), is provided in Figure 1-12. Proceeding from the syllable tier the solid lines represent the relationshi p between the phonological segmen ts and the syllable in the familiar hierarchy: syllable > onset, rhyme (> nucleus, coda). The syllable requires a vocalic segment in order to be well formed. The number of segments associated with the syllable appear in the skeletal tier. Thus, the last two-syllables are seen to branch. Goldsmiths autosegmental 33

PAGE 34

planes or tiers greatly facilita te the treatm ent of long vowels, diphthongs, geminate consonants, d tautosyllabic clusters. The interaction between planes is clear in the three dimensional representation. The metrical plane in Figure 1-12 marks all of the potential stress bearing elements with x on the syllable tier while indicati ng the metrical feet w ith parentheses on the foot tier; the final syllable, enclosed in angled brackets, is consid ered to be extrametrical and does not project a foot. On this level only the head of each foot is marked with x. The word level tier indicates the locus of primary stress. The one other foot on the foot tier, a left-head tr ochee, may be said to carry a secondary stress co rresponding to the initial syllable of the word. X X X X X X X X X X X X X < X > ( X ) ( X ) < X > ( 'X ) Figure 1-12. Autosegm ental repres entation of a prosodic word, Latin b n t tem f., acc. sg. Independence of the syllabic structure from the metrical grid is of critical importance as Halle (1998, 543) notes in a retrospective on the accent of English words: It was originally thought that feet are made up of syllables, but subsequent work has shown this to be incorrect ... It was therefore proposed in Halle and Vergnaud ( 1987) that feet are composed not of syllables, b n t tem Metrical Plane Syllable Foot Word Syllable Plane Skeletal Plane 34

PAGE 35

but of those pieces of a s yllable that may bear stre ss. Nevertheless, for Latin it is important to consider the syllable in its entire ty because the accented syllable, if it is penultimate, must also be bimoraic. Because autosegmental representa tions are not linear but multidimensional it is possible to see the constituents of a syllable on th e skeletal plane. The lines connecting these to the syllable have a branching rhyme for the last two-syllables whereas the first two do not. Metrical Grid The grid notation developed by Liberman (1979), Liberman and Prince (1977), Selkirk (1980), Hayes (1981), and others is formalized in Halle and Ver gnaud (1987) and Hayes (1995). For Hayes (1995, 27) the metrical gr id is a representation of a se quence of beats equally spaced (horizontal dimension) with in prescribed temporal limits. The beats vary in intensity and this is shown by the vertical dimension. These gene ral observations on construction of the grid establish that the metrical grid is (1) hierarchi cal in nature; (2) has evenly spaced intervals; and (3) observes downward implication, that is, a gr id mark on a higher level has a corresponding grid mark on all lower levels. The hierarchical nature of the metrical grid can be seen in its graphic representation. Using Selk irks (1980, 565) hierarchy of word, stress foot, syllable as a point of departure, the syllable level corresponds to line 0. Setting aside extrametricality for the moment it can be said that every member of line 1 (foot level) has a corresponding member in line 0, that is, line 1 is a subset of line 0; the same is true at word level and phrase level. In other words, construction of the vertical dimension must respect the princi ple known as Continuous Column Constraint (Hayes 1995, 34): (1.7) Continuous Column Cons traint (CCC) (Hayes 1995) A grid containing a column with a mark on layer n + 1 and no mark on layer n is ill-formed. Phonological rule s are blocked when they would create such a configuration. 35

PAGE 36

Because the principle of eurhythm icity rejects ad jacent stresses th e result is that the subsets decrease as higher level lines are constructed. This can be seen in Figure 1-13, an abstract representation of a four-syllable word consisting of a string of CV syllables with penultimate stress (foot type is trochaic). In this schematic only the head of the foot is marked with x on lines 1 and 2. The boxes on the first and thir d V columns in Figure 1-13 demonstrate CCC in determination of primary and secondary stress for a hypothetical four-syllable word. The construction of lines 2 and 3 will also depend on language specific criteria, for example, some languages require exhaustive parsi ng of the word whereas others construct only one word-level accent, typically at left or right word edge. Cons truction of the foot row ensures that only parsed segments will be ultimately designated as having a primary or other word level accent. x x x x x x x C V1 C V2 C V3 C V4 Line 2: Word level Line 1: Foot level Line 0: Syllable level Figure 1-13. Metrical grid. The construction of Row 1 (foot row) entail s recognition of the rela tionship between head and constituent. Halle and Vergnaud (1987, 8-28 ) assert that a relativ ely small number of parameters are able to describe this relations hip: [HT] and [BND]. The first parameter designates whether or not a constituent is head-terminal, that is, whether or not the head is aligned with one of the constituent boundaries; the second whether or not the head of the constituent is separated from the constituent bounda ry by more than one element. Additionally, [+HT] constituents can be either left headed or right headed. In Table 1-2 the conventions [+LH] and [+RH] will be used to indi cate location of foot head; --corresponds to unattested patterns 36

PAGE 37

and a per iod indicates non-head elements of the constituent. The resultin g possibilities, then, are the following: Table 1-1. Parameters for the construction of a metric foot. [-BND] [+BND] [+HT] [+HT] / [+LH] *. *. [+HT] / [+RH] .* .* [-HT] --.*. Because bounded feet have beginning and end it is now possible to construct bracketed grids. Hayes (1995, 55) departs from the Halle/Vergnaud parametric approach on several issues. He rejects non binary feet and assert s that the inventory can be lim ited to [+HT] and [BND]. This excludes ternary feet from any po ssible foot inventory leaving as possible foot types the four shaded cells in Table 1-1. Metrical Theory and Language Typology Hayess inventory of possible foot types and case studies builds on a long tradition. Liberman and Prince (1977) foresee that direction of scansion and type of foot provide important data for discussions of typology and in the same year Hymans (1977) survey of 444 languages, summarized in Fox (2000, 170) r ecognizes the following patterns: (a) languages with dominant initial stress1 114 (b) languages with dominant second-syllable stress 12 (c) languages with dominant penultimate stress 77 (d) languages with dominant final stress 97 (e) languages with non-dominant stress 144 In the list above the pr eference for stress at word edges is well attested and follows a long tradition in which stress serves a demarcative function. The third most frequent pattern, penultimate stress, on the other hand, is viewed by as a function of prefe rred intonational pattern, 1 Dominant stress gives prominence, through devices other than pitch, to a particular syllable while reducing the prominence of others (van Coetsem 1996, 43). A frequent byproduct of dominant stress is vowel reduction, present in both Catalan and Portuguese. 37

PAGE 38

that is, a con trast of high + low (Fox 2000, 170-171) Furthermore, the penultimate stress pattern of Latin may well have morphological grounding. In the case of a noun consisting only of stem + declensional suffixes it can be assumed that accent on the stem might be a more desirable outcome than accent on the inflection. The stem contains the new information and would, therefore, be treated in the same manner as infl ectional focus in the phrase. While this may well result in a high + low intonation pattern the mel ody is a byproduct of the word accent rather than a determinant. Hayess typology of accent (1995) is based on th e metrical foot. He proposes three basic bounded foot types (1995, 71): syllab ic trochee, moraic trochee, and iamb. The syllabic trochee (Hayes 1995, 63) is one which depends only on sy llable count. The moraic trochee (Hayes 1995, 69) takes into account the weight of the syllable ; accordingly, a moraic foot may consist of twosyllables of one mora each or a single bimoraic syllable. The addition of the moraic trochee is crucial for analysis of the Latin stress sy stem. Kager (1995, 370-373) uses the operative parameters that emerge from metric bracketi ng to describe foot typology and construction: (a) Foot typology a. Extension: Bounded/unbounded b. Dominance: Left headed/right headed c. Quantity-sensitivity: Quantity sensitive/ quantity insensitive/quantity-determined (obligatorily branching) (b) Foot construction a. Directionality i. Right to left/left to right ii. Bidirectionality: Noniterative foot assignment beginning at one edge and iterative foot assignment beginning at the opposite edge b. Iterativity: Iter ative/noniterative (c) Word tree parameters a. Dominance: Left-dominant and right-dominant b. Labeling: Strong (branc hing)/weak (nonbranching) (d) Word tree dominance a. Left dominant/right dominant 38

PAGE 39

W ith regard to extension, bounded syst ems are well known. These are the ones corresponding to the three basic foot types: moraic trochee, syllabic trochee, and iamb. Unbounded systems, as inventoried by Hayes (1995, 296-297), are thos e that favor heavy syllables at word edges. Arguably, the Indo-Eur opean accent could be considered representative of the type with a primary accent on the leftmost heavy syllable (otherwise leftmost syllable) (Halle, 1997). Absent in the Hayes inventory is any evidence for ternary feet, although he recognizes that there is limited evidence for tern ary feet in the case of Cayuvava, a linguistic isolate spoken in northeaster Bolivia (1995, 309316). In terms of syllabic structure Cayuvava has only light syllables and stress regularly falls on the antepenultima te syllable. In longer words there is also an accent on the sixth and ninth sy llables from the end of the word. However, Hayes (1995, 96-97) notes that even languages with established binary feet often manifest an initial dactyl effect. He cites as example a secondary stress pattern di scussed also in Harris (1983) and Roca (1986) in which an initial dactyl emerges in word s of three or more syllables with an odd number of syllables, such as gneratvo (primary accent indicated by acute and secondary by grave). In certain registers thes e words have an initia l dactylic foot which alternates with the usual alte rnating syllable pattern of sec ondary stress which would predict genratvo There are other cases of th e initial dactyl eff ect in languages as diverse as English, Indonesian, Swedish, and Polish (Hayes 1995, 97). It should be noted that the inventory of the worlds languages by accentual system is an ongoing effort. An example is Baileys (2007) Stress System Database which records information on location of stress, dire ctionality of stress, boundedness, and language specific conve ntions for syllable weight. Moraic Theory The question of quantity sensitivity as an i ndispensable element in defining foot typology appears in both Hayes (1995) and Kager (1995). The parameter of quant ity-sensitivity produces 39

PAGE 40

an additional layer in the m etrical tree, the mora, below the syllable level, as seen in Figure 1-14. Moraic theory assigns weight va lues to segments within the sy llable. Hayes (1989) and others assume that syllable-initial consonants are extramor aic; the moraic count be gins with the syllable nucleus and may also include post-nucleic segments. Word (Wd) | Foot (F) | Syllable ( ) | Mora ( ) Figure 1-14. The revised prosodic hierarchy A heavy syllable contains at least two moras, that is long vowel or vowel plus coda, and a light syllable contains no more than one. Tree representa tions of the syllable appear in Figure 1-15. Ewen and van der Hulst (2001, 151) distinguish between rhyme-weight languages (nucleus + coda) and nucleus-weight languages (nucleus only). | | | | | C V C V C C V C C VV light heavy light* heavy (*nucleus-weight languages only) A B B C Figure 1-15. Tree structures for heavy a nd light syllables based on moraic count. The first pattern, A, is universally recognized as light for languages w ith moraic trochees. What constitutes a heavy syllable, however, is not universal. Pattern B depends on a coda consonant for the two mora count. Fo r a nucleus-weight language, pattern B this same configuration, CVC, would be c onsidered a light syllable. Pattern C, where VV represents a long vowel, would be a heavy syllable for a nucleus-weight language. Latin displays both 40

PAGE 41

patterns B and C; these correspond to the fam iliar long by position and long by nature in traditional Latin grammars. Some languages also admit superheavy syllables (long vowel + coda consonant). The superheavy syllable does have a role in determining placement of accent in some varieties of Arabic (see Goldsmith1990, 197200). In these cases the superheavy foot is final which begs the question of extrametricality. In Latin, however, there are instances of non final superheavy feet such as claustra, n. pl. gate and maestus sad, as well as monosyllabic aut or and haec, f. sg. this. Selkirks (1980, 570) model of a superfoot, Figure 1-16, implies that the primary stress foot node dominates another stress foot containing a stray syllable. s w Figure 1-16. Selkirks superfoot. The stray syllable could be a non-canonical syllable, that is, without nucleus. Certainly when comparing haec to the full form haece this is a plausible explanation. Because /k/ can no longer function as onset of a fina l syllable it must attach itself to the preceding syllable. However, if the final consonant in a superheavy f oot is parsed as belong ing to another syllable, collocation of the primary accent in Latin becomes problematic (Steriade 1988). For example, if multiplex is treated as mul.ti.plek.s the penultimate syllable is now plek and, as a heavy syllable, would attract word stress. Description of the word level accent in Latin requires the mora layer in the prosodic hierarchy in order to distinguish between cases like c l mus reed and c l mus we convoke. Tree configurations for th e two are in Figure 1-17. Primary accent is indicated by the syllable in bold with a superscript 1. 41

PAGE 42

W d Ft 1 < > Wd Ft Ft 1 < > Figure 1-17. Latin prosodic word. A) Light penult. B) Heavy penult. Extrametricality Although Liberman and Prince (1977, 281-282, 294-298) include tangential discussions of extrametricality and adjunction in recognition of the fact that the metric al approach leaves residue that must be accounted for in some way, extrametricality cont inues to be a topic of study. Beasley and Crosswhite (2003) review previous approaches to extrametricality before proceeding with their own OT analysis. Hayes ( 1995, 56-60) utilizes Estonian as an example to show that CVC syllables, treated as light in final position but heavy in nonfinal syll ables, require a formal mechanism for exclusion from normal me tric parsing. He further notes that similar treatment of CVC as light occurs in English, Arabic dialects, a dial ect of Hindi, Spanish, Romanian, Ancient Greek, and Meno mini. Rather than adjusting metrical rules to the changing value of CVC final syllables such syllables coul d simply be excluded from the parsing operation. Halle and Idsardi (1995, 408), bu ilding on Idsardis previous work, propose the Edge-Marking Parameter. (1.8) Edge Marking Paramete r (Halle and Idsardi 1995) Place a parenthesis to the of the -most element in the string. right left right left right left 42

PAGE 43

In order to achieve the binary patterns proposed by Hayes (1995) another param eter is required which Halle and Idsardi (1995, 418) call Iterative Constituent Construction. Iterative Constituent Construc tion (Halle and Idsardi 1995) (1.9) Iterative Constituent Construc tion (Halle and Idsardi 1995) Insert a boundary for each pair of elements. right left As case in point Halle and Idsardi build a grid to show the correct assignment of stress in two Latin verb forms: r pr mitur and r pr muntur Row 0 begins with the Syllable Boundary Project parameter. As it applies to Latin, Buck ley (2000) expresses this parameter as follows: (1.10) Syllable Boundary Projection (Buckley 2000) Project the left boundary of a heavy syllable onto line 0. The Edge Marking Parameter (Edge), described above, is set at RLR, that is, place a right parenthesis to the left of the rightmost element in the string; Iterative Constituent Construction is L, insert a left boundary for each pair of elements. The prosodic foot is left headed. There is also an overriding constraint: Avoid x(x#. This constr aint disallows forms that do not meet the minimum word requirement of two moras but do es allow monosyllables with long vowel or vowel followed by coda consonant. While the cyclical application of the Edge Marking Parameter and Head assignment produces accurate results for Latin words monos yllabic, disyllabic and larger, as well as enclitics Buckley (2000, 271-275) argues that the Avoid x(x# cons traint seems to be language specific and does not correctly predict stress fo r Manam, an Oceanic Austronesian language (Papua New Guinea) which favors stress on a heavy final syllable. However, locus of stress changes with the addition of some classes of suffixes. With the AP suffixes, which show possession and beneficiary of an action, stress moves to a he avy penult; otherwise, the antepenultimate syllable. The Avoid x(x# parameter precludes final word stress and is thus 43

PAGE 44

unable to capture the accentual patt ern of for ms both with and without suffixes. What is needed is something resembling more closely the concept of extrametricality. Idsardi and Halle (Buckley 2000, 274) propose a modification of the Avoid parameter and suggest that the notation )x can be used to treat anything fo llowing the ) as extrametrical. Table 1-2. Edge Marking Parameter in Latin Line 0 Syllable Boundary Projection: L x x x x r pr mitur x x(x x r pr muntur Edge: RLR x x x)x r pr mitur x x(x) x r pr muntur Iterative Constituent Construction: L x (x x)x r pr mitur (x x(x) x r pr muntur Head: L x x (x x)x r pr mitur x x (x x(x) x r pr muntur Avoid x(x# applied Line 1 Edge: RRR x) x (x x)x r pr mitur x x) (x x(x) x r pr muntur Second application of Edge Marking Parameter Head: R x x) x (x x)x r pr mitur x x x) (x x(x) x r pr muntur Second assignment of head Line 1 in Table 1-2 illustrates that )x produ ces the desired effect for Latin, rendering the final syllable tur extrametrical in both cases. However, conventional use of extrametricality has been likened to simply using a diacritic to achieve what formal rules and parameters are unable to do. The Avoid parameter does not offer signi ficant improvements. On the other hand, OT presents some workable alternatives to extrametri cality and other issues ra ised by metrical theory through the mechanism of gradient constraints. Optimality Theory Prince and Smolensky (2004, 3) distinguish Optimality Theory (OT) from its predecessors as they formalize its operative principles, c onstraints that assess output configurations per se and those responsible for maintaining the faithful preservation of underlying structures in the 44

PAGE 45

output. The assessm ent of output structures is driven by wellformedness constraints derived from Universal Grammar. Further suppositions of Prince and Smolensky (1993, 2) include that constraints are often conflictiv e in appearance (for exampl e, use of opposite alignment constraints W/L and W/R in Jac obs 2003a, 275-276), that actual ou tput will violate one or more constraints, and that constraints are hierarchical with each constrai nt having priority over those at a lower level. Sherrard (1997) reviews some of the constraints that determine well formedness and reiterates that it is importa nt to remember that the evalua tion of outputs in OT does not require satisfaction of constraints but allows c onstraints to be disreg arded unless doing so would involve violating a more important constraint (Sherrar d 1997, 44). It can be assumed that some constraints are, in fact, inviolable. In brief, the mechanism for describing input and output relationships is the OT table or tableau. Construction of the tableau begins with an input form whose po ssible outputs (in rows) will be evaluated by a series of prioritized constraints that appear as heading of the columns. The configuration of the tableau represen ts the two core principles of OT: GEN and EVAL. The input form generates an unlimited number of possible candidate forms which are evaluated relative to constraints of well formedness and faithfulness. According to McCarthy and Prince (2001, 5) Shifting the explanatory burde n from input-driven rewrite rules to output constraints changes the way the input-outpu t pairing system must be se t up, particularly in phonology. Instead of taking an underlying form an input and transforming it deterministically step-by-step to its associated output, it is n ecessary to allow for the generation of a large set of candidate outputs. The candi date set of formal possibiliti es is submitted to evaluation by the system of well-formedness constraints, which selects the true output from among the candidates. Dominance of one constraint over another is indicated by the relative order of the columns and the use of broken or solid line s (These relationships can also be indicated in linear fashion in the form A>>B>>C). A violation by candidate A of a constraint to the left of candidate Bs 45

PAGE 46

violations autom atically disquali fies candidate A and is signaled by in the corr esponding cell of the tableau. A dotted vertical line divides constraints that do not have a hierarchical relationship. Lower level constraints that are violated but do not figure in computation are designated by shaded cells in the tableau. The optimal candida te, the one with no disqualifying (!) violations, is signaled by In the following sections two majo r aspects of OT are examined, the architecture of the syllable and metrical constrai nts. These have a dir ect relationship to the evolving accentual patterns of La tin and the Ibero-Romance languages. Architecture of the Syllable At the level of the syllable one can envisi on constraints that arise from the assumption that CV represents summum bonum. Possible sylla ble types are summarized in Table 1-3 (based on Prince and Smolensky 2004, 105). The shaded cel ls represent optional configurations with regard to onset (1B, 2B) and codas (2A, 2B) whereas the unshaded cell represents the ideal syllable, CV, which requires an onset and forbids a coda. The ranking of the constraints, relative to other constraints such as FAITH, faithfulness to input, is unpredictable and langua ge specific. Table 1-3. Possible syllable types Onsets Required Not required Forbidden 1 CV (C)V Codas Allowed 2 CV(C) (C)V(C) A B The distinctiveness of each language rests on th ese factorial typologies, that is, the relative ranking of constraints in order to produce the optimal candidate. With regard to the universal preference for CV syllable structure, Archangeli (1999, 534-535) suggests that there are four universal constraints that define the canonical syllable. The f act that violations of these constraints are widely attested bu t differ from language to language illustrates the efficacy of OT 46

PAGE 47

constraints in recognizing univers al preferences while adm itting more or less marked forms in language specific contexts. (1.11) ONSET: A syllable must have an onset. (Archangeli 1999) (1.12) NOCODA: A syllable must not have a coda. (Archangeli 1999) (1.13) PEAK: The nucleus is the most sonorous pa rt of the syllable (Archangeli 1999) (1.14) COMPLEX: Syllable margins (onsets, codas) contain at most one consonant. (Archangeli 1999) Syllable onset The CV syllable type, according to Prin ce and Smolensky (2004, 111) is universally optimal. Onsetless syllables for some language s invite repair strate gies. Thus, the well formedness constraints trigger violations of the other set of constraint s, those which promote faithfulness to input. Diachr onically it can be seen that the transition from Latin ( h)erba grass to Spanish ( h)ierba with alternate orthography yerba and fortition of initial glide to [ ], [ ], [ ], or [] represents an attempt to correct the violation of ONSET. Even the initial glide [j] would provide an acceptable onset because the high vocoid is less sonorous than the following vowel [e]. The idea of a rising/falling sonority contour with regard to the syllable is not new. In a discussion of the Sonority Sequencing Principl e Morelli (2003, 358-359) acknowledges that the concept of relative sonority wi thin the syllable has long been a topic of study for more than a century. In its simplest form Morelli de fines the SSP constraint as follows: (1.15) SONORITY SEQUENCING PRINCIPLE (SSP): In a syllable, sonority increases toward the peak and decreases toward the margins. (Morelli 2003) Hooper (1976, 199) also establishe s the symmetry of the syllable with possible onsets and codas as mirror images as seen in the schematic seen in Figure 1-18. A crucial difference in syllable onset and coda is indicated by the st rong/weak labels. For Hooper (1976, 197-207) the 47

PAGE 48

consonant in onset position is str ong as established by the universality of the CV syllable. In the case of Spanish, Hooper (1976, 209) notes th at ther e are no conditions on the strength value of the initial consonant while there are marked pref erences with regard to coda consonants favoring resonants and disfavoring obstruents. NUCLEUS MARGIN MARGIN obstruents nasals liquids glides vowels glides liquids nasals obstruents Least vowel-like Most vowel-like Less vowel-like STRONG WEAK WEAK Figure 1-18. The intrinsic structur e of the syllable (Hooper 1976). Further evidence for the sonority hierarchy, disc ussed in Kiparsky (2008) as an example of a constraint based universal sequencing preferen ce, comes from studies of language disorders where it is observed that the constraints governing se quencing must be represented in some multi-level, distributed fashion in the nervous syst em since they are impervious to even severe forms of brain damage (Buckingham and Christman 2008, 131). Hooper (1976, 206) also posits a universal strength hierarchy for consonants reproduced in Figure 1-19 (cf. Parker 2002, 210, who establishes a sonority hierarchy for Spanish; also Blevins 2004a, 159-172; Gordon 2007, 754-756; Kiparsky 2008, 49-52). She further speculates that weak consonants, those with lowe r numerical score, are more likel y to occur in the weak position of the syllable while the converse is true for th e strong (initial) position of the syllable. These concepts are validated by diach ronic studies, particularly in Romance linguistics (Cross 1934, Geisler 1992, Granda Gutirrez 1966, Jac obs 1994, Mayerthaler 1982, Vnnen 1967, Wireback 1996). The weakening and/or loss of syllable final and word final consonants in the transition from Latin to Romance is richly documented. 48

PAGE 49

glides liquids nasals voiced continuant voiceless continuant voiced stop voiceless stop 1 2 3 4 5 6 Figure 1-19. Universal strength hi erarchy for consonants (Hooper 1976). Another example of the resistance to onsetless syllables is seen in German. In this case less prominent positions in the word do not requ ire an onset but word initial position does. According to Fry (2003, 217-220) a glottal stop is inserted only when the left edge of the syllable coincides with the left e dge of a foot or prosodic word. In other words, in cases like Beamte civil servant both [ b .' am.t ] and [ b .'am.t ] are possible realizations but in the case of Idee idea only i.'de ] is possible. The aspiration of voicel ess stops in English in word initial position and the flapping of // discussed in Selkirk (1980) also seem to indicate that syllable position does impact the distributi on of consonants in onset and coda In brief, the subset of segments that can occur as onset is not equal to the subset of segments that can occur as coda. Furthermore, intervening factors such as accen t and positional prominence may affect outcomes. Syllable coda The existence of separate constraints that concern syllable onset, peak or nucleus, and coda presupposes that the syllable consists of separate constituents. The current model of the syllable is that of a binary, branching entity consisting of onset + rhyme. This is the view espoused by Blevins (1995, 212) who asserts in a discussion of sonority sequencing that coda clusters, if allowed, are not me rely a mirror image of occurring onset clusters. While Hooper (1976, 209) asserts that for Spanish there are no constraints on the stre ngth value of initial consonants the same cannot be said of coda cons onants. Modern Spanish has few occurrences in syllable final position of consonants that rate 6 on the strength scale with the exception of some marginal words: borrowings ( frac dress coat, tails), onomatopoeia ( tictac tick tock), and 49

PAGE 50

learn ed words such as aptitud aptitude. It should be reme mbered that the presence of a grapheme in standard orthography does not guara ntee the realization of a segment phonetically. Both aptitud and actitud can be rendered as [a.ti.'tu]. Consonants at the 5 level are also problematic, for example, the only voiced stop that occurs in word fina l position with regularity is /d/, subject to weakening and elision in many varieties of Spanish. Vo iceless continuants are represented by /s/ which has a morphological function ye t is also subject to aspiration or elision; /f/ occurs only in marginal words (/ /, limited to central and northern Spain, does not show signs of weakening or elision,). Additionally, there are many languages that disallow coda consonants altogether (see Blevins 2004b on loss of final consonants in Austronesian). Blevins (1995, 214-215) proposes a three way typology for syllable weight based on branching rhyme (nucleus + coda) or branching nucleus (long vowel). Within each type the weight of the nucleus and coda represent a gradient: light (no nbranching), heavy or heaviest. Which elements are computed in the weight resp ond to language-specific computation of weight (nucleus only or rhyme). Unattested forms are in dicated by --and X represents either V or C. Because type 1 languages do not admit coda consonant s the heaviest syllable type consists of a long vowel. The difference between heavy and heaviest syllables, for Type 3, is predicated on the sonority of the rhyme. A long vowel or a vowel plus resonant ar e more sonorous than a vowel followed by a non resonant coda consonant. Table 1-4. Typology of li ght and heavy syllables Light Heavy Heaviest Representative languages Type 1 C0V Nonbranching rhyme --C0VX Branching rhyme Sierra Miwok, Hausa Type 2 C0VC0 Nonbranching nucleus --C0VV Branching nucleus Huasteco, Hawaiian Type 3 C0V C0V Nonbranching rhyme C0VC1 C0VC1 Branching rhyme C0VV C0V{V, R} Branching nucleus Klamath, Yupik Creek 50

PAGE 51

It can be seen from Table 1-4 that the NOCODA constraint is regularly violated by Type 1 and Type 3 languages. According to Gordon (2002, 56) languages like Latin can be construed as Type 1 languages because they have two moras in the rhyme. All of the Ibero-Romance languages permit coda consonants albeit with a number of restrictions some of which have been discussed above with regard to Spanis h. However, the degree to which NOCODA is violated differs greatly among the three languages with Catalan at one extreme, allowing nearly all voiceless obstruents in c oda position, and Portuguese at th e other end of the spectrum allowing only liquids and /s/ (for Brazilian Portuguese only /r/ and /s/ because /l/ is regularly vocalized in syllable-final position). The dependence on coda for weight sensitive phenomena disappears in Ibero-Romance as the contexts which would allo w a syllable to be considered heavy become irrelevant. Bybee (2001, 16) not es that linguistic behaviors tend to coincide with the presumption of the NOCODA constraint as a universal indi cator of well formedness: If there is a constraint comparable to the no coda constraint of Optimality Theory, it is a result of the phonetic tendency to reduce and coarticulate coda consonants more than onset consonants. This tendency manifests itself in every instance of language use in languages that have coda consonants, reducing these c onsonants by very small degrees. Eventually, coda consonants are lost in such languages, leaving a language with a reduced number of coda consonants or none at all. Syllable peak (nucleus) Another approach to the question of prefe rred distribution of se gments in different syllable positions is to establish a constraint th at is based on the relativ e sonority of peaks and margins. Building on Prince and Smolensky (1993), Smith (2005, 55-58) proposes *PEAK/X: (1.16) *PEAK/X: For every segment a that is the head of some syllable x | a |>X where |y| is the sonority of segment y X is a particular step on the segmental sonority scale The sonority scale in Table 1-5 re presents an array of segments in order of acoustic prominence; the more sonorous elements constitute optimal peaks. 51

PAGE 52

Table 1-5. Sonority scal e of phonological segments High sonority Low sonority low vowels m id vowels high vowels/glides rhotics laterals nasals voiced obstruents voiceless obstruents Smith (2005, 56-59) acknowledges th at the above scale is not une quivocally universal, that is, there are some language-specific rerankings with in the scale. While rhotics>>laterals seems appropriate for English (see discussion in Zec 2003) this may not be the case for all languages. For example, in Brazilian Portuguese syllable-final /l/ is regularl y vocalized while /r/ may be a velar fricative [ ] or a glottal fricative [h] or [ ] (Silva 1999, 2211). If /r/ has, in fact, become a voiceless obstruent it has now moved to a position of lower sonority wi th regard to the la teral, /l/. While there is no evidence in Ibero-Romance of nasals or liquids as syllable peaks, vowel reduction in Catalan and Portugue se does produce less sonorous peaks. The status of [ ] in terms of the relative ranking of vowels is still unresolved. Smith (2005, 58) notes that it is sometimes classified below high vowels (cf. Gordon 2002; Blevins 2004, 159; Kiparsky 2008, 49-52). Onset and coda clusters The three languages under studyCatalan, Castilian, and Portugueseexemplify *COMPLEX, a constraint that limits onset clusters and, especially, coda clusters. Onset clusters, as is the case for most languages, have no prosodic role in Latin or Romance. In Ibero-Romance initial clusters cannot have mo re than two segments and the first segment (an obstruent) is always less sonorous than the second (liquid or rhotic; the status of glides requires further 52

PAGE 53

discussion). Additionally, initia l clusters with /l/ undergo pala talization in both western and central Ib ero-Romance and emerge as / /, / /, and / /. Coda clusters do occur in Latin and contribute to syllabic weight in the sense that a vowel is long by position when in a closed syllable. Latin does not distinguish heavy and superheavy syllables for purposes of computation of stress. Furthermore, nearly all superheavy syllables, both (C)VVC and (C)VCC, seem to involve /s/. The special properties of /s/ have been the subject of several studies (Paradis and Prunet 1991) and the extramet ricality of word final /s/ as the single consonant of a coda or as part of a cluster is an unresolved question. Faithfulness and Repair Strategies The syllable properties discussed up to this poi nt onset, peak, coda complex (onset or coda) concern well formedness, that is, adherence to a universal archetype of the ideal syllable. Faithfulness, in turn, is based on the premise that input equals output. However, when input fails to meet well formedness criteria tension arises be tween faithfulness to the input form and change to more nearly approximate the ideal CV syllabl e type. In their simplest form faithfulness constraints are devised to disall ow addition or deletion of segments from the input form. This proposes a dilemma for those cases where it might be desirable to add, delete, or modify a segment in order to reduce markedness. The onsetless syllable as in Latin ( h)erba grass which becomes Spanish hierba/yerba with word initial palatal cons onant is a good example. The alternate orthography, yerba, indicates clearly that the glide is now perceived as an onset rather than the nonsyllabic element of a rising diphthon g in morphophonemic alternation with /e/. Similarly, violation of the no coda constraint ca n be resolved by deleti on of a consonant, which in effect has happened with the Spanish word reloj watch where the markedness of word final /x/ has lead to its deletion, refl ected in the alternate orthography rel The online database of the 53

PAGE 54

Real Academ ia Espaola, Corpus de referencia del espaol actual (CREA) has 12 occurrences with the orthography rel and 5147 with the traditional orthography reloj Conservation of the coda in orthographic represen tation does not indicate that it is actually pronounced. Concepts of faithfulness can also be used in tandem with well formedness constraints as proposed by Prince and Smolensky (2004 (1993), 29-33, 106) who utilize FILL and PARSE to that effect. (1.17) FILL: Syllable positions are filled with segmental material. (Prince and Smolensky 2004 (1993)) (1.18) PARSE: Underlying elements must be parsed in to syllable structure. (Prince and Smolensky 2004 (1993)) For an underlying form /VCV/ the tableau in Table 1-6 (based on Prince and Smolensky 2004, 33) shows the effect of ranking ONSET and PARSE above FILL. Neither candidate violates onset although the onset is not realized on the surface. The symbol indicates an empty position; the onset slot is present but not occ upied in the case of candidate A. In the case of candidate B the extrametricality of the initial vowel makes the fo llowing syllable the first parsable element thus constituting a PARSE violation. Table 1-6. Ranking of ONSET and PARSE above FILL /VCV/ ONSET PARSE FILL A. V.CV. B. .CV. *! The PARSE and FILL constraints proposed by Prince and Smolensky (2004 (1993)) later take the form of MAX and DEP (McCarthy and Prince, 1995), cons traints that proscribe deletion or insertion of segments. (1.19) MAX: Every segment/feature of the input has an identical correspondent in the output. (McCarthy and Prince 1995) (1.20) DEP: Every segment/feature of the output has an identical correspondent in the input. (McCarthy and Prince 1995) 54

PAGE 55

Both MAX and DEP have consequences for prosodic system s that depend on heavy syllables. The deletion of a coda, for example, could result in loss of the context required by a quantity sensitive rule to assign word accent. The insert ion of a vowel to break up a medial cluster consisting of two consonants adds a syllable nucleus. Presumably one of the consonant segments will now parse with the anaptyct ic vowel as onset of the syllable. Differing repair strategies can be seen in Castilian, Catalan, and Portuguese in the treatment of medial clusters in words of Latin origin. Latin advoc tu(m) advocate enters the Ibero-Romance languages as Cast. abogado, Cat. advocat and Port. advogado. In terms of prosody, the absence of long vowels in Roman ce means that the accented syllable no longer attracts strength on a quantitative basis yet it rema ins as the head syllable of the head foot. Violation of MAX (loss of the nucleus in the final syll able) in Catalan has now created a heavy, accented, final syllable because cat has a coda consonant. Assuming that Catalan also has an alignment constraint preferring the right edge, viol ation of the faithfulness constraint has allowed preservation of right edge alignm ent. In the case of Portuguese advogado Brazilian Portuguese regularly inserts epenthetic /i/ to break up medi al clusters whose first member is a stop. Violation of DEP results in the addition of a vocalic nucle us with the result that the word now consists only of CV syllables (with the exception of the onsetless first syllable): [ a. i.vu.'ga.du ]. Faithfulness could in theory be applied to any feature of a segm ent voice, place of articulation, mode of articulationor to features not on the segmental plane such as word accent. Violations of faithfulness are common in both synchronic and diachronic processes. Syllable final /n/, for example, tends to assimilate to th e place features of a following obstruent or nasal. In the case of the indefinite article un in Castilian, the normally alve olar /n/ may become bilabial, interdental, dental, palatal, or velar depending upon the pl ace features of the following segment. 55

PAGE 56

For such cases it is necessary to ensure that the corresponding faithfulne ss constraint does not rule ou t the candidate [um] when the following segment is [p] or [b]. A similar case of assimilation can be seen diachronically in th e Castilian and Portuguese reflexes of Latin c m tem (acc. sg.) companion. Syncope eliminates the unaccented penultimate vowel and produces *komte. The nasal now assimilates to the following coronal producing conde in both Castilian and Portuguese. Catalan retain s the place of articulation in comte for which the standard pronunciation in Eastern Catalan is ['kom.t ]. Hualde (1992, 373) notes other cases of /m/ followed by an alveolar: himne hymn and premsa press. Since the latter involve front unrounded vowels any arguments for co-ar ticulation are invalidated. Word Accent In addition to the constraints that treat syllab le architecture OT has developed a conceptual framework to address the issue of word level stress. Prince and Smolensky (2004 [1993], 124) propose LX PR as a foundational element. (1.21) LX PR: Every Lexical Word must correspond to a Prosodic Word. (Prince and Smolensky 2004 [1993]) The prosodic word forms one level of the pr osodic hierarchy refine d by Selkirk (1980) and revised by Hayes (1995) to include the mora. The relationships are shown in the nested string in Figure 1-20. The mora, not a universal member of this set, is indicated by gray font. [ [ [ [ [ V(V(C)) ]Nuc/Coda ]Mora ]Syll ]PrFt ] PrW d Figure 1-20. Nested prosodic constituents. Every prosodic word contains a prosodic foot co nsisting of m inimally one syllable. Metrical theory admits the single syllabl e foot although there are often c onstraints with regard to the nature of the syllable. For example, the less sonorous peaks are not capable of bearing stress, as 56

PAGE 57

is the case of syllabic nasals and liquids in English (Zec 2003, 126). For languages with quantitative stres s such as Latin, there may be a minimal moraic requirement for what constitutes a prosodic foot, especially if it is the only foot in the word (Jacobs 2000, 2003a, 2003b, 2004, Mester 1994, Prince and Smolensky 2004 [1993]). Prosodic Foot At the level of the prosodic foot, OT constrai nts are often associated with the form of the prosodic foot. Thus Prince and Smolensky (2004 [1993], 71) propose for Latin the constraints FTBIN, RHTYPE = T, RHHRM, WSP and categorize them as violated or unviolated constraints. Table 1-7 gives the complete list of constraints with definitions and indicates the effect of the constraints. It is understood that unviolated means unviolated in the optimal forms of the language (Prince and Smolensky 2004, 70). Table 1-7. Foot form constraints for Latin Constraint Effect Status FTBIN Feet are binary at some level of analysis ( F = Unviolated RHTYPE = T Feet are either iambic or trochaic. Unviolated RHHRM An uneven trochee is disallowed. *(HL) Unviolated WSP Heavy syllables are prominent in foot structure and on the grid. *H Violated in Classical Latin Unviolated in Pre-cl assical shortening register The FTBIN constraint reflects the typology of metr ical systems also treated in Hayes who proposes only two basic binary foot types (1995, 71), iambs and trochees; tr ochees can be either syllabic or moraic. WSP does not permit a heavy sy llable (H) to be treated as if light (L). However, Latin offers many instances of heavy sy llables that do not attrac t word level stress. 57

PAGE 58

Since pars ing is not exhaustive once the head foot has been built at the right edge of the word (excluding the last foot) pre-tonic heavy syllable s are likely to be unrecognized in the prosodic hierarchy of the word. The one exception to FTBIN (Feet are binary) for languages with a nonmoraic rhythmic type is the light monosy llable. It can be argued that many light monosyllables actually meet a minimal length/ weight requirement, for example, Catalan p bread could be inte rpreted as underlying pan because the elided nasal reappears in the plural pans In Portuguese final nasal deletion has produced compensatory lengthening and nasalization through the forma tion of a nasal diphthong: po p w The nasal vowel / / is present in the plural form as well. Curiously, in the derived form padeiro baker the /n/ has disappeared without a trace, perhaps because primary word accent has moved to the suffix. RHTYPE = T (rhythmic type) and RHHRM (rhythmic harmony) work together to establish the well-formed trochaic foot, which for Latin is either a single heavy syllable (H) or two light syllables (LL). RHHRM excludes the uneven trochee, in bo th its manifestations, *(HL) and *(LH). Prince and Smolensky (2004 [1993], 70) note that such forms are known to be marked or even absent in trochaic systems. The *(HL ) type foot is firmly rejected by Mester (1994) who considers unparsed syllables to be a better alternative to permitting the uneven trochee as a canonical foot form. RHHRM also interacts with WSP (Weigh t-to-Stress Principle). Barring *(LH), for example, as a possible foot type invok es repair strategies designed to interpret the heavy syllable as a light syllable. Latin prov ides examples in iambic shortening and cretic shortening which have the effect of creating a mo re optimal word in prosodic terms. Iambic shortening converts word final sequences of the type HLH into HLH and cretic shortening converts both word final and word intern al sequences of the type LH into LH (Lahiri, Riad, and 58

PAGE 59

Jacobs 1999, 384). Examples of final heavy syll ables are abundant in Latin and are to be found in both verbal and nom inal systems where desi nences have either a long vowel or a vowel followed by a coda consonant thereb y creating a heavy syllable. Alignment Constraints The ALIGN constraint posited by Prince and Sm olensky (2004 [1993]) recognizes the interplay of phonology and morphology. A general form of this rule is given in their analysis of Lardil (2004 [1993], 127): (1.22) ALIGN: The final edge of a Morphological Wo rd corresponds to the final edge of a syllable (Prince a nd Smolensky 2004 [1993]). McCarthy and Prince (1996, 74) note that There is a strong connection between the edge-in character of redupli cative association and the sp ecial status of constrai nts aligning the edges of morphological and prosodic constituents. This sa me type of constraint can also be to assign word level accent. For Latin, Jacobs (2000, 2004) proposes an alignment constraint that favors the right edge of the word. (1.23) ALIGN (PRWD, R, FT, R): Align the right edge of a foot with the right edge of a prosodic word. However, since Latin does not allow primary acc ent on the ultimate syllable it is necessary to introduce a constraint that pr ecludes that outcome. Prin ce and Smolensky (2004 [1993], 62) propose NONFINALITY as a constraint capable of producing the desired effect: (1.24) NONFINALITY: No head of PrWd is final in PrWd. (Prince and Smolensky 2004 [1993]) Table 1-8 (adapted from Prince and Smolensky 2004 [1993], 69) shows how NONFINALITY, ALIGN, and PARSE(Syllables must be parsed) work toge ther in trisyllabic and quadrisyllabic Latin words to correctly pl ace word accent on the antepe nultimate syllable. 59

PAGE 60

Table 1-8. Alignm ent constraints in Latin NONFINALITY (F ; ) ALIGN ( ;R) PARSE/c p tem/ head (acc.) (c.pi)tem ca(p.tem) F /c p t lum/ head (dim.) ca(p.tu)lum ** (c.pi)(tu.lum) (ca.pi)(t.lum) F From Table 1-8 it is apparent th at the constraint ranks below NONFINALITY, but it will disfavor candidates that have stress beyond the three-syllable window counting from the right edge of the word even if doing so would place stress on a heavy syllable. Optimality Theory, Multiple Outputs, and Analogy That OT offers a viable approach to diachronic studies of phonological and prosodic phenomena is demonstrated by an examination of its basic principles, including recognition of a multiplicity of outcomes. This is an important aspect of the present study which is comparative in nature. A common input form often results in distinct outputs, not only among languages but within the same language. OT has the ability to deal with competing forms and multiple outcomes through the use of different, language-speci fic constraints as well as reranking of the same constraints to accommodate language prefer ences that can correctly predict outcomes. Additionally, OT is able to reflect the fact that language change is grad ual in nature and often produces competing forms. Differentiated outputs are expected across languages but also occur within the same language in the form of patrimonial or popular wo rds that contrast with learned or semi-learned words from a common etymon. OT addresses these issues through the use of gradient and output-output constraints. 60

PAGE 61

Gradient constraints (Hayes 2000, Boersm a and Hayes 2001) refl ect the fact that not all constraints can be viewed in binary terms, that is, the perception of well-formedness may constitute a continuum where Form A is preferre d over Form B which is preferred over Form C, ad infinitum. Over time, the distribution of competing forms comes to resemble a normal or Gaussian distribution. The zero point on the X-ax is marks the highest probability of occurrence whereas less likely outputs appear as data points to the left and right of the mean. The various constraints of any given language may appear at any given interval in a hypothetical horizontal continuum. If the range of Constraint 1 is far enough apart from Constraint 2, their distributions are unlikely to overlap. However, it is precisely the possibility of overlapping constraints that best expresses what occurs in language change As a case in point, outcomes of the Spanish future tense from tener+he (infinitive tener to have, to hold + auxiliary he lst pers. sg. must) are explored. Old Spanish provide s testimony of the coexistence of tern and tendr (the modern form) in the thirteenth Century (Davies 2002-2004, Corpus del espaol, s.v. terne/tendre). There are seventeen occurrences of the form with metathesis, and seven occurrences of the form with epenthesis. Both outcomes reflect the undesirab ility of /-n.r-/, the consonant sequence that results following loss of the pretonic vowel /e/. The fact that both are attested in the Corpus del espaol is indicative of free varia tion at a given time. This is the intent of overlapping constraints as posed by Boersma and Hayes (2001, 48): Over a longer se quence of evaluations, the overlapping ranges often yield an important observable effect: for forms in which C2 >> C3 yields a different output than C3 >> C2, one observes free varia tion, that is, multiple outputs for a single underlying form. For the example at hand the input form is posited as /tner/ because initial /h/ is lost; it is presumed that the initial syllable carries a se condary accent. The following constraints are 61

PAGE 62

utilized in th is case. It is not deemed necessary to distinguish primary and secondary accent for constraint 1.28. The intent of this constr aint is to ban intert onic vowels. (1.25) MAXI/O: A segment in the input must have a correspondent in the output. (1.26) DEPI/O: A segment in the output must have a correspondent in the input. (1.27) SYLLABLE CONTACT LAW (SYLCON): The final element of a syllable is not less sonorous than the initial element of an im mediately following syllable. (Wheeler 2007) (1.28) *V : A vowel may not occur between accented syllables in a word. (1.29) ROOT-IDENT: The root of a derived form must be identical to its base. (cf. Benua 2000, 19-22) Table 1-9. Free variation with unordered OT constraints /tner/ V SYLCON ROOT-IDENTDEPI/O MAXI/O A. tner *! B. tendr C. tern D. tenr *! High ranking of V prevents faithful but unattested tner from emerging as winner although it incurs only one violation of constraints. Similarly, SYLCON must be ranked above the faithfulness constraints in order to prevent tenr from emerging as preferable to the candidates that are actually attested, B and C. Another option is to view these constraints more abstractly in terms of faithfulness to root, and, as potentially overlapping. While candidate B satisfies ROOT-IDENT, it is less faithful than candidate A. Therefore it will be placed at the periphery of this constraint. Similarly, while candidate C satisfies DEPI/O, it is also less desirable in terms of faithfulness than candidate A and is placed in a marginal posi tion relative to this constraint. Since these two constraints are unranked they can be viewed as potentially overlapping with comp eting forms existing in free variation as seen in Figure 121. The ideal candidates are those that satisfy both contraints, 62

PAGE 63

DEPI/O and ROOT-IDENT. However, if there are no candida tes that satisfy the constraints (without violating higher ranking constraints), candidates B and C are able to occur in free variation until the constr aints become prioritized. This is seen in the Corpus del espaol where tern and tendr co-occur in documents throughout the thirteenth century. The eventual emergence of tendr (candidate B) as the wi nner is dependent on identity to the root and its promotion above DEPI/O. C B RT-IDENT DEPI/O Figure 1-21. Overlapping constraints. The role of root faithfuln ess in the selection of ten dr as the optimal candidate introduces a second important aspect of histori cal change. It has been seen that free variation and competing forms are well explained by OT constraints. In the case illustrated in Table 1-9 and Figure 1-21, the nature of violable constraint s is clear. The optimal candidate is not necessarily the one that incurs fewest violations of cons traints. Over time, these cons traints are promoted or demoted resulting in periods of free variation followed by emergence of a preferred form although variants may persist. Less preferred forms are often relegated to a re stricted semantic role The popular outputs of Latin cl vis, f. key, Cast. llave and Port. chave, both referring to the physical object, retain the primary meaning of the Latin etymon compared to the learned form clave in both languages which has the secondary and metaphorical uses of key such as musical key. Semi-learned Cat. clau designates both the physical obj ect, musical key, and other meta phorical uses. Cases such as the popular forms of clavis are important because they provide evidence of multiple outputs when a patrimonial or popular word shares a common etymon with a learned or semi-learned 63

PAGE 64

counterpart. Although many of these doublets m ay be the result of re-introduction of a Latinized form at a later chronological period, OT is able to demonstrate and predict the later outputs through demotion of constraints. Learned words nearly always show high rankin g of faithfulness constraints, although these are rarely as restrictive as the IDENT/IO type constraint. Segments present in the input are less likely to be deleted, and segments not present in the input are rarely added. In terms of the processes under study here, deletion of vocalic segments in particular can be viewed as obeying constraints that penalize the presence of pre-tonic and pos t-tonic vowels (deleted through processes of syncope and apocope). Likewise the presence of alignm ent constraints which reward alignment with left and right word edge s favor deletion of unstr essed final vowels, as well as the deletion of intertonic vowels in word internal position. The presence of prothetic vowels in the case of s+stop ini tial clusters, even in very recent loan words such as estop (from English stop), shows how some constraints can continue to be high ranking over a long period of time. The ROOT-IDENT constraint discussed above raises a nother important facet in historical change, the role of influences external to the inpu t form itself. In the case of the formation of the future tense, analogy is clearly a f actor in the outcome of Castilian venire + he > vendr I will come. Deletion of /i/ is far less frequent than de letion of the mid vowels /e/ and /o/. Epenthesis is extended by analogy to other verb stems ending in /n/. The role of analogy in assignment of word level accent has been studi ed as a synchronic phenomenon in both the context of first language acquisition (Eddington 2000) and second language acquisition (Apoussidou and Boersma 2003, 2004; Bullock and Lord 2003). Anal ogy is perhaps the best explanation for the few cases of stress shift that occu r in Ibero-Romance such as Spanish perodo period, where 64

PAGE 65

antepenu ltimate accent was once considered the normative pronunciation but the current edition of the dictionary of th e Real Academia (DRAE, s.v. perodo) now gives periodo with penultimate accent as an altern ate form. Comparison with peridico newspaper where the accented syllable has [jo] as nucleus, or periodista journalist and periodismo journalism both with diphthong [jo] in unstressed syllables, make s apparent the analogical pressures to shift the stress in perodo to the more optimal nucleus [o] whic h is more sonorant than the front high vowel [i]. The role of paradigmatic pressure is also seen in the outcomes of hypocoristic formation. Hypocoristic formation is an active process in Spanish (Lipski 1995, Pieros 2000), Catalan (Cabr 1998), and Portuguese (Gonalves 2001). Th e output form often results in an accentual pattern common to all three languages, that is, a binary troc hee realized as L .L or H .L (accented syllable underlined) as well as a less preferred heavy monosyllable, a pattern absent in Valencia according to Cabr (1998, 15-19). The monosylla bic hypocoristic is also present in both Portuguese and Spanish (for example, Portuguese Zus, Jus, and Spanish Chus for Jess ), though not a frequent pattern. Table 1-10. Foot pattern of hypocoristics in Portuguese, Spanish, and Catalan Full form of name Portuguese Spanish Eastern Catalan Valencian Foot form Josefa Zefa Chepa Pepa Pepa Pepa ( L .L) Domingo Mingo Mingo Mingo Mingo ( H .L) Isabel Bel Bela Bel Bela ( H ) aAlthough not given as the preferred output for Spanish, compounds such as Maribel < Mara Isabel suggest that Bel exists as a combining form. The forms illustrated in Table 1-10 correspond to faithfulness to a prosodic template (cf. Downing 2006, 48-64). They are built on the head foot of the word and are faithful to the prosodic structure of the input fo rm, although not necessarily to th e input segments as seen in Chepa and Pepa. However, there is an alternate patter n for all three languages that involves 65

PAGE 66

creation of a m inimal word (prefe rably a two-syllable word) from th e initial syllables of the input form. The origin of this alternate form is explai ned in part by reference to the idea of positional strength. Beckman (1998, 52) notes that One source of evidence for init ial-syllable positional privileg e may be found in the domain of lexical access and language processing. There is a considerable body of psycholinguistic research which indicates that word-initial mate rial, either spoken or written, plays a key role in lexical access, word r ecognition and speech production. Therefore, even if the initial syllable does not carry an accent its position renders it prominent and allows the creation of a trocha ic foot at the left edge of th e word. There are, however, cases where the initial syllable naturally ca rries a secondary accent, such as Federico which yields both Fede and Kiko in Spanish (cf. Spanish Sebasti >Sebas, Chano; Josefina > Jose, Pina ). It is perhaps through analogy with names in which the initial syllable bears a secondary accent that the new prosodic template arises fo r hypocoristic formation as in Port. Eduard>Edu, Cat. Montserrat>Montse It is noteworthy that the truncated forms that are built at the left word edge are always disyllabic trochees. Clearly, this is the prefer red prosodic template. The monosyllabic variants for the most part meet the minimal word requireme nt of two moras. One exception is Portuguese Z for Jos. If the vowel can be cons idered long, then this form would also meet the minimal word criteria. For the vowel /e/ Delga do Martins study on Po rtuguese phonetics (1988, 128132) shows that it has the longest duration of any of the stressed vowels in Portuguese. Because of the widespread use of apocope as a poetic de vice and the relative infrequence of monosyllabic content words that are not bimoraic it is not unlikely that words like Z are perceived as bimoraic. Future discussions of the various processes th at affect word level accent, specific to the languages under study, rely on OT to provide a way of describing crucial differences in syllable 66

PAGE 67

structure, su ch as constraints on codas, in so far as these affect assignment of the primary accent and its alignment in the word. The loss of sy llable codas reduces the coincidence of syllable weight and the assignment of word accent, either primary or secondary. The shift away from a quantity sensitive system of accent represents a ma jor change in the transition from Latin to Romance. Although there is some evidence that heavy syllables ar e favored as locus of primary word stress in Romance it cannot be said that WS P is still an operative constraint. Review of free variation and the role of analogy has shown that multiple outcomes are a real possibility, especially when the basis for assignment of accent is no longer transparent. The concept of extrametricality is no longer essential for correc t placement of primary word accent. All of the Romance languages under study allow syllable final accents. In some cases this is because desinences (p articularly in the verb system) ar e stress bearing; in other cases it is because apocope of unstressed final vowels ha s created a prosodic word that consists only of a bare stem without suffixation. The following chapters describe in detail the prosodic systems of Latin, Castilian, Catalan, and Portuguese through reference to a database comprised of Latin nouns and their reflexes in the three Ibero-Roman ce languages in order to substantiate or reject Lindsays assertion (1894, 164) that Latin accentu ation is retained with wonderful tenacity by the Romance languages. 67

PAGE 68

CHAP TER 2 THE NATURE OF THE LATIN ACCENT Polemic of Pitch versus Stress There is little doubt that the pattern of accentua tion in Latin is based on stress rather than pitch, one in which accent correlates with duration and intensity in the pronunciation of a syllable and which does not serve to distinguis h words (Baldi 1999, 268-269). Yet this runs contrary to the tradition of the Latin grammarians and subsequent European scholars. As late as 1869, Thurot (1869, 392) describes the Latin accent as musical, one which donnait lune des syllabes dun mot une intona tion plus aigu quaux autres, qui taient graves, et que, dans certaines conditions, la syllabe prononce dabord avec une intonation aigu tait acheve avec une intonation grave; cest ce quon appela it laccent circonflexe. When Baldi (1999, 268-269) asserts that early Latin must have had a stress accent he expresses a long-standing concern of scholars: while the Latin grammarians often speak of their accent in terms properly applicable only to a pitch-accent, a ll the features of thei r language point to its having been a stress-accent (Lin dsay 1894, 150). The terms usedgr ave, acute, and circumflex are calques from the Greek as are many other grammatical terms. The rules governing stress in Latin are clearly stated in the grammatical tradition albeit with the perplexing references to circumflex, acute, and grave. Juret (1921: 76) cites as example the rules found in Donatus (fourth century C.E.) where the operative accent terms are circumflectere acuere, and gravis pronuntiare. As in Greek, pitch was determined by the nature of the vowel. An acute accent (high pitch) corresponded to a short vowel wh ile a long vowel had a circumflex accent (falling pitch). Evidence of Prosodic Change According to Sturtevant (1921, 5-6) early desc riptions of Latin accent as pitch accent, as seen by the reference to Donatus, are highly cons istent until about the fourth century C.E. The 68

PAGE 69

co-occurrence of both pitch and stress as m arkers of the prominence of a syllable is consistent with modern theories regarding accentual systems. Use of the terms prominence (stress-accent) and nonprominence (pitch-accent) is suggested by Van Coetsem (1996, 36-43) to distinguish different types of accentual systems. Accent type may then be subdivided as in Figure 2-1 (based on van Coetsem 1996, 42). accent ____________ _____________ prominence nonprominence (stress accent) (pitch accent) ________ _________ dominance nondominance (characterized by (characterized vowel reduction) by pitch) Figure 2-1. Typology of accent (van Coetsem). The dominance accent gives prominence to a par ticular syllable while reducing the others and utilizes devices in addition to pitch to distinguish the pr ominent syllable (van Coetsem 1996, 43). In contrast, the nondominance accent type tends to rely heavily on pitch in placing the prominent syllable in relief. Another key difference between dominance and nondominance accent types is the concentration of energy on a part icular syllable versus an even distribution of energy. These terms of reference and their desc ription to some degree conflate issues of production and perception disc ussed in Beckman (1986, 13) who cites Joness (1950, 137) separation of stress as a subj ective activity on the part of th e speaker from prominence an effect perceived by the hearer. It is also important to note that the spli tting of the prominence accent into dominance and nondominance underscore s the degree to which giving one syllable prominence comes at the expense of others in the word. 69

PAGE 70

The contrast between dominance and nondomi nance corresponds in many ways to the traditional dichotomy of stress-timed versus syllable-timed languages (Abercrombie 1967, 97-98 and Pike 1947, 12-13). Syllabletimed languages are characteri zed by the regularity of the intervals at which syllables re cur; hence they are isochronous. Stress-timed languages, in contrast, depend on the occurrence of stress as a demarcative function. Since the number of syllables that occur between stresses is vari able, syllable length adjusts accordingly. While the underlying assumptions of syllable-timed versus stress-timed accentual systems have been eroded considerably in the last decades (see Ramus 2002 and Ramus, Nespor, and Mehler 1999 for recent acoustic studies), th e list of concomitant characteri stics of the timing patterns in Table 2-1 (from Crosswhite 2001, 173) contains valuable insights for the study of vowel reduction and syncope, evident at all stages in Latin. Table 2-1. Characteristics of st ress-timed and syllable-timed languages Stress-timed languages Syllable-timed languages Have vowel reduction Have full ar ticulation of unstressed vowels Tempo acceleration via compression of unstressed syllables Tempo acceleration via proportional compression of all syllables Complex syllable structure with relatively uncertain syllable boundaries Simple syllable structures with well-defined boundaries Tendency of stress to attract segmental material in order to build up heavy syllables No such tendency Native speakers are not good syllable-counters; presence of stress-c ounting versification Native speakers are better at syllable-counting; presence of syllable-counting versification That the accent of classical Latin was someth ing other than a stress accent seems to be supported by evidence from poetry. The discrepa ncy between ictus and word level stress, for example, seems to indicate that unstressed vowels were not so reduced th at they could not be assigned a musical beat in the scansion of a verse. For example, in the first line of Vergils Ecloga I, the independence of poetic meter from th e metrical word is readily apparent. In Figure 70

PAGE 71

2-2 the first line shows the poetic scansion in dactylic hexam eter; the line below the verse shows the location of the word stress marked by x. The preferred metric f oot is the dactyl, | |, and it is obligatory in the penultimate foot. The last foot has an obligatory pattern as well, the spondee, | |. The spondee may also substitute a dact yl in the first through fourth foot. In the first foot ictus and word accent coincide but the second foot has two word level accents, only the first of which coincides with the poetic meter. Th e third foot is headed by the long final syllable of pt lae although primary word level accent in Latin is proscribed for the final syllable of a word regardless of its weight. 1 2 3 4 5 6 | | | | | | Tityre tu patulae recubans sub tegmine fagi | | | | | | | x x x x x x x Figure 2-2. Alignment of scansion and wo rd accent in Latin dactylic hexameter Devine and Stephens (1993, 395-396) discuss the convention of line initial freedom in verse, that is, the initially chos en hypothesis can be set aside for a better one that emerges in the middle of the line of verse. In other words, e rrors at the beginning of the parsing sequences are less serious than deviations that occur once a pattern has been estab lished. If in a verse there is an attempt to incorporate an initial unaccented sy llable it is possible to set it aside if the predominant form is later seen to be a troch ee, for example. The initial syllable is now reinterpreted as anacrusis. This may be seen in the case of the text in Figure 2-2 where the second syllable of the fourth foot may be view ed as proclitic: it in troduces the grammatical constituent of which tegmine is the head. 71

PAGE 72

If the poetic m eter is quantity sensitive, the long vowel of t and the two short vowels in the first two-syllables of pt lae do not pose a problem in terms of satisfying the meter but the first syllable of pt lae cannot now project a metrical foot because its segments have been reassigned to the foot headed by t In constructing the dactylic hexameter the first element of every foot should contain a heavy nuc leus. Allen (1973, 123) notes that while some metric feet do not occur in Latin (the pyrrhic foot for example) it is the contrast between the first and following elements that makes the dactyl a suitable measure for poetry: the dactyl is admissible not on the basis of its greater length but because one of the 2-mora elements is differently constituted from the other. The preferred poetic foot, then, has a contrast (H=heavy syllable; L=light syllable): HLL, although the HH configuration of the spondee is also a licensed form. In moraic terms both contain four moras, or It can be said that in the context of poetry, not in the language at large, the phrase level metrical configur ation required to produce the desired sequence shown in Figure 2-2 overrides word level foot formation and accent assignment. Table 2-2. Syllabic weight of the dactylic hexameter H.L.L H.L.L H.L.L H.L.L H.L.L H .H Preferred foot structure H.H H.H H.H H.H Allowable altern ative foot structure for feet 1-4 1 2 3 4 5 6 Baldi (1999, 269, f. 11) asserts that the norms of Latin versification do not in themselves provide adequate evidence for two coexisting sy stems of word accent in the period between the first century B.C.E. and the fourth century C.E. It is well known that poetic license in the creation of oral or written compositions allows many uses of language that diverge from the norms of the linguistic community. Allen (1973, 15) also cautions against dependence on poetic forms: before making use of poetry in esta blishing the phonology [of a language], one must first determine the extent of convention and artifi ce in the poets usage. Dislocation of stress in 72

PAGE 73

poetry and song is well known in many m odern la nguages and it is unlikel y that anyone would rely on such texts exclusively to establish the prosodic characte ristics of a language (see, for example, Janda and Morgan 1988 for a study on disloc ation of lexical accent in song in Spanish, primarily, but with examples also from Italian, French, and Portuguese). Appendix B contains a discussion of popular poetry from the formativ e period of Ibero-Romance written in ballad meter, octosyllabic lines with stress on the firs t, third, fifth, and seventh syllables. The last stressed syllable is often, in fact, the last on e in the line although the frequent use of paragoge (Honsa 1962) reinforces the argument that the pref erred accentual pattern that ha s emerged is a left headed binary foot consisting of two-sy llables without regard to syllable weight. Early Latin The accentuation of the IE proto-language offers relevant insights into the unresolved question of competing accentuatual systems in La tin. Halle (1997, 276) suggests that the mobile stress system of Sanskrit and Ba lto-Slavic is the direct legacy of the proto-IE accent. The descendants of IE that no longer subscribe to this system of accen tuation fall into two categories: (1) languages with initial accent and (2) languages with accent on one of the last three syllables of the word. Other Italic langua ges are of the first kind while La tin belongs to the second group. Halle (1997, 276) further states that it is we ll known that among the IE languages, those with initial stress are the historical an tecedents of languages with stre ss on the last three syllables. There is much language-internal evidence to sup port an argument for a strong, initial accent in the formative period of Latin. To that end, the following processes are considered to be sources of evidence (Baldi 1999, 269-70): (1) syncope, (2) vowel reduc tion, (3) vowel raising in noninitial syllables, and (4) treatment of early Greek loanwords. 73

PAGE 74

Syncope and Accent Leumann (1963, 91) notes that the loss of unstr essed vowels calls into question the nature of the early Latin accent; furthermore, syncope of both pre-tonic and post-tonic vowels continues to be a phenomenon of classical Latin: Bei Synkope anderer Vokale kommt daneben auch die historische lat. Betonung in Betracht; jedenfal ls sind unter dem Dreisilbenakzent sowohl vorwie nachtonige Vokale untergegangen. He c ites many examples which are organized in the table below according to shape of input and the location of the unstressed vowel (1963, 92-3). Reduction of total word length is an obvious outcome of syncope but it will be seen that other important changes occur, particular ly with regard to rhythmic patterns. The examples in Table 2-2 offer many instances of change in moraic count, syllabic stru cture, and foot formation as a consequence of vowel loss. In the case of reddo I give back <*re-did for example the loss of the medial short vowel results in the following reconfigurat ion of moraic and syllabic structure: ., or L.L.H H.H (where L and H designate light or heavy syllables) In metrical terms the undesirable anapest has now become a spondee in c onformity with this principle of music cited in Prince (1983: 406): It is a fact of musical life that when a beat (or subbeat C of any grid position) is divided in two, the first half is felt to be naturally stronger than the second half. The loss of the medial light syllable (with short vowel) in the examples in Table 2-2 is best motivated by a strong initial accent. The c onsequence in prosodic terms is cl ear; the stressed syllable (in Classical Latin) is now heavy with few exceptions such as the compounds offic na
PAGE 75

Table 2-3. Early Latin syncope 1. Loss of 2nd syllable in 3 or 4-syllable words: ( ) ( ) (a) Between identical stops (in reduplicated forms): C1( V ) C1 reddo < *re-did ferre < *ferere reccid < re-cecid cont l < *con-tetul (b) Between non-identical consonants: C1( V ) C2 praeco < *praid(i)c o_ qu ndecim < qu nque + decem offic na < opific na (c) Between sonorants (N, L) L L N ( V ) N ulna < *olen nllus < *n(e)-oinelos domnus < dominus domna < domina (d) Between sonorants and obstruents C ( V ) L/N L/N ( V ) C surgo < *surrigo ardeo cf. ridus pno < *po-sin cf. positus 2. Loss of 3rd syllable in words of 4 or more syllables: ( ) ( ) (a) Between sonorants (N, L) L L N ( V ) N gemellus < *gemenelos ampulla < amporel cor lla < kor nel (b) Between sonorants and obstruents C ( V ) L/N L/N ( V ) C puerpara < *puero-para sepultus < *sepelitos (c) Between obstruents and obstruents C( V ) C aud cter < *aud ci-tero 3. Loss of vowel in pre-tonic position, or between primary and secondary stress (a) In compounds calfacio < cale + faci (b) In derivations discipl na cf. discipulus < *dis-kapelos 4. Syncope in word final syllables In disyllabic words ending in -s ars < *artis m ns < *mentis nox < *noktes sal s < *sal ars < *artis m ns < *mentis nox < *noktes sal s < *sal tis Cases of syncope, then, usually preserve, or, in some instances, create well-formed feet. If the second syllable of reddo is interpreted as extrametrical the word still contains one well 75

PAGE 76

for med foot, red-, which meets binarity criteri a on the basis of moraic count with the new word structure corresponding to ( )< >, where the head foot is a well formed moraic trochee. The primitive form opific na is somewhat problematic; it is attested in Plautus where it occurs in this line (Miles gloriosus III, iii, line 880, 1963, 151. The ictus of the poetic meter, iambic septenarius, is marked in this edition.): si ea in pificna nsciam at mala sse aut fradulnta The word has two accents, one on the first syllable and a second on the long vowel of the antepenultimate or fourth syllable: 1 4 The first is congruent with the theory of a strong initial accent in early Latin which Kent (1931, 189) calls the primitive Italic accent. The second accent, designated as bearing ictus in th e verse, corresponds to a long vowel in Classical Latin. Jacobs (2003b, 407) notes th at in the comedies of Plautus and Terence there was a strong tendency to harmonize verse ictus (the strong position in the vers e foot) and word accent. Jacobs (2003a, 270) maintains that in the transition from word initia l accent to right wordedge accent Latin passes through an intermediate stage in which words with more than four syllables obey the antepenultimate maximum. Yet the supposed antepenultimate accent of one of the examples that Jacobs (2003a, 276; 2003b, 411; 2004, 73) utilizes in three separate studies, the word opific na, is supported neither by evidence from Plautine poetic meter nor by the accentuation pattern of Classical Latin where penultimate stress coincides with the heavy syllable -c Nevertheless, the OT analyses put fo rward by Jacobs to describe the changing nature of Latin word accent provide a way of de scribing the word initial stress of Pre-classical Latin and the evolving change from left to right word edge seen in Early Classical Latin. The introduction of new constraints or the re-ranki ng of the same constraints combine to produce different optimal forms in each time period. Ad ditionally, the OT approach is best able to describe the environment predictive of vowel syncope, the weak position in a binary foot. 76

PAGE 77

The OT derivation of pre-classical initial stress in Latin presented by Jacobs (2003a, 275276; 2003b, 396-410) presum es the following undomina ted constraints not reflected in the OT tableau: FTBIN, RHTYPE (T), that is, feet are binary and troc haic. The other constraints that head the columns of the tableau (Jacobs 2003a, 275-276) are defined below: (2.1) NONF: A foot may not be final. (Jacobs 2003a) (2.2) W/L: Align the prosodic word and pros odic foot to the left. (Jacobs 2003a) (2.3) W/R: Align the prosodic word and pros odic foot to the right. (Jacobs 2003a) (2.4) H/L: Align the head foot to the left in the prosodic word. (Jacobs 2003a) (2.5) H/R: Align the head foot to the righ t in the prosodic word. (Jacobs 2003a) (2.6) PARSE: Syllables in the prosodic word ar e passed into a foot. (Jacobs 2003a) In Table 2-4 the first example, m l f c um is given in Jacobs; the second, opific na, has been added to compare the treatment of lexical accent in Pre-classica l Latin, Early Classical Latin, and Classical Latin. The locus of pr imary accent is indicated by bold and single underscore. The citation forms given in Table 24 include vowel length a ttributed to Classical Latin (s.v. m l f c um and ff c na, Lewis and Short), although according to Jacobs quantity sensitivity (QS) plays no role in pre-classical Latin. Table 2-4. Pre-classical Latin word initial accent NONF W/L W/R H/L H/R PARSEm .l .f .c .um misdeed / / A. ( ) ( ) B. ( ) ( ) *! C. ( ) *** D. ( ) *** p f c na workshop / / E. ( ) ( ) F. ( ) ( ) *! G. ( ) *** H. ( ) *** 77

PAGE 78

The optim al candidates A and E in Table 2-4 ha ve the primary accent in the leftmost foot and only the final syllable is left unparsed. This is achieved through the relative ranking of constraints, particularly the high ranking of NONF and W/L. The second word displayed in the tableau, p f c na, later undergoes syncope and is realized as ff c na in Classical Latin. However, here it is treated in parallel with m l f c um Candidates B and F are disqualified on the basis of nonfinality, that is, the formation of a foot coinciding with the right edge of the word. W/L and H/L insure that only candidates with in itial primary stress will be successful. For Early Classical Latin Jacobs (2003b, 411) reranks the H/L and H/R constraints. The higher ranking of H/R will cause the word leve l primary accent to move closer to the right edge of the word and thus block word initial stress of four-syllabl e words in which the first three syllables are light. Although Jacobs ( 2003a, 2003b, 2004) treats all the syllables of p f c na as light, the word is analyzed in Table 2-5 with a lo ng penultimate syllable in order to demonstrate the reranking of constraints in Classical Latin. Table 2-5. Early Classical Latin word accent vs. Classical Latin word accent Early Classical Latin NONF W/L W/R H/R H/L PARSEm l f c um misdeed /LLLL / A. (LL)( L L) B. ( L L)(LL) C. L( L L)(L ) !* D. (LL)L( L ) !* Classical Latin NONF W/R W/L H/R H/L PARSEp f c na workshop /LLLH / E. (LL)L( H ) ** F. L( L L)(H) ** G. ( L L)L(H) ** According to Steriade (1988: 376) the developm ent of penultimate syllable stress for those words with a heavy penult is a change that took pl ace during the fourth century B.C.E. In Table 78

PAGE 79

2-5 the results of the new constrain ts are seen. Primary accent is indicated by bold and underscore. Again, it is assumed that FTBIN and RHTYPE (T) are underlying, inviolable constraints. The OT analyses seen in Tables 2-4 and 2-5 now provide a vehicle to characterize the process of syncope in Latin as a mechanism to obtain the optim al foot configuration. Mester (1994, 38-43) notes that many cases of early sync ope occur in a specific environment, a light penultimate syllable between a heavy antepenultimate syllable and a heavy final syllable. The requirement for a heavy final syllable means that cases will be restricted by morphological properties of Latin, that is, the syllabic weight of the inflectional morpheme. However, the case of p f c na in Table 2-5 is a counter example. The unparsed syllable is not deleted; rather, the weak syllable of the first foot is elided to produce Classical Latin ff c na, or, (LL)L(H ) (H)L(H ) in prosodic terms. The resulting form incu rs fewer violations of H/L while avoiding violation of any of the high ranking constraints. Jacobs (2004, 75) not es that All cases of syncope, throughout the history of Latin, can be regarded as targeting the same structural context: the weak position of the foot. He recognizes, however, that his OT analysis results in syncope opacity in that it best predicts cases of pre-tonic syncope (2004, 84). One of the difficulties with Jacobs analysis is the resulting conflation of several different time periods. Reighard, on the other hand, althou gh limited by SPE type rules, describes three different types of syncope circumscribed by ch ronology and sources of evidence. The first, prehistorical syncope, consists of forms whose syncopated vo wel can only be reconstructed through comparative or internal evidence, for example praeco m. sg. nom. herald (cf. praedicare to proclaim). In such cases, the presyncope forms are non-occurrent in Latin. An OT analysis of these forms, then, would atte mpt to predict the known forms that would result from speculative input. Reighard characterizes the environment for prehistorical syncope as a 79

PAGE 80

short vowel in an in ternal open syllable; it is the same for historical syncope. Latin provides data for historical syncope, which Reighard (1977, 99 ) speculates occurred between the time of the early dramatists and the Classical period, thr ough textual examples of both syncopated and unsyncopated forms. An example of this type of alternation is captured in Tables 2-5 and 2-5 in the form p f c na, cited in Plautus, which is realized as ff c na in Classical Latin. The Plautine form corresponds to p fex nom. sg., worker. Certainly, in Classical Latin, the unsyncopated form was no longer an option. In light of the pr evious discussion one must also wonder if it is not a question of metrical convenience in Plautus. The third type of syncope, variable syncope, is best described as an ongoing change in fa vor of syncope: we know that the syncopated variants generally, by the Classical period and afterward, were normal, unaffected forms, and that the unsyncopated forms usually have someth ing of a pedantic or purist flavor to them, especially in the pos t-Classical period (R eighard 1977, 100). There are other syllable level adjustments th at co-occur with early syncope. The examples in Table 2-3 frequently involve the loss of a short vowel in the environment of a liquid or nasal. When the resulting clusters produce a more sonorous element in coda position followed by a less sonorous element, there is no change in the coda consonant, as in ulna, 1(c). However, when the following consonant is more s onorous, assimilation often occurs as in nllus, 1(c) and ampulla If the geminate /-ll-/ is consider ed to be ambisyllabic in the word nllus the first syllable is now superheavy. Ewen and van de r Hulst (2001: 184) note that in superheavy syllables the coda is restricted to a sonorant or a fricative, the coda sonorant does not display distinctive place contrast, and the favored place ca tegory is coronal--all of which are true in the case of nllus 80

PAGE 81

Iambic Sho rtening and Cretic Shortening The treatment of the final syllable in Latin has been the subject of prescriptive grammars, treatises on versification and phonol ogy, as well as more recent studi es on metrical theory and optimality theory. Niedermann (1953, 37) cites Quintilians cautions ab out pronunciation of the word in its entirety not neglecti ng the final elements: ne extr emae syllabae intercidant. The most extreme form of vowel reduction, vowel loss, seems to have occurred in the formative period of Latin, for exampl e, the demonstratives hic, haec, hoc with emphatic particle c(e) added to the root (Baldi 1999, 344) Apart from this case of hi storical apocope, vowel reduction in the word final syllable largely concerns iamb ic shortening rather than vowel deletion. Baldi (1999, 264) defines iambic shortening ( brevis brevians ) as the rule by which unstressed heavy syllables which immediately follow light syllable s in disyllabic words (i.e., those which are metrically ) are lightened by shortening the vowel of the second syllable, resulting in a pattern. He also notes that extensive use of iambic shortening in the lower registers in Plautus and Terence is indicative of the continuing influe nce of post-tonic weakening as the strong initial accent continues to exert some influence. As in the case of syncope Baldi (1999, 265) views iambic shortening as an optional process that is both lexically and morphologically conditioned, for example, there is vacillation with regard to words that ha ve a long vowel on a morphological basis: vol I wish, hom man. There is some evidence to suggest that vowel length in the final syllable was treated as [long], that is, it acquired the value needed to produce optimal syllabic weight and rhythmic configuration (cf. Juret 1921, 77-78: En latin la valeur des longues parat dpendre de la p nultime .. tout se passe comme si la finale vocalique longue, quelle quen soit la nature, avait deux units dintonation aprs une pnul time brve.) This ambiguity is also seen in the treatment of the vowel in imperatives. Lindsay (1894, 211-213) notes that there is evidence from scansion in which the long vowel of the first, second and fourth 81

PAGE 82

conjugations is at tim es short and at times long. Because the thematic vowel is often followed by a consonant, for example, the present tense second person singular /-s/ and third person singular /t/, vowel length does not determine syllable weight because the vowel is already considered to be long by position. In the case of the nominal system, however, the nature of the final vowel has both an important morphol ogical and prosodic function. Lindsay (1894: 220) notes that the etymological long vowel was, like other vowels, shortened in course of time. When final, perhaps only in iambic words originally ... but this shortening was extended to all Noms. Sing. of -stems and Noms. Accs. Plur. Neut ... Thus in the declension of -stems, the First Declension, final -a of the Nom. Sg. is even in the earliest poetry a short vowel. Although vow el length is important in a quantity sensitive system of accentuation, the final syllable in Latin never takes stress. The primitive stress accent favored the first syllable and the stress rule of Cla ssical Latin prescribed stress as falling on the penultimate or antepenultimate syllable depending on the weight of the penultimate syllable. Therefore, the weight of the final syllable wa s of little consequence be cause even as a heavy syllable it could not attract stress. Shortening of the vowel in final syllables is also seen in third declension nouns and adjectives of the type ending in -tor, -sor, -ter, -al, -ar (see Leumann 1963: 103). Here this seems to reflect an exchange of moras. The genitive l ct ris reader has a long vowel but since the syllable does no t end in a consonant the moraic count for that syllable is the same as the nominative, lector where final /-r/ is a syllable coda, thereby producing a count of two moras. Mester (1994) views the ear ly cases of iambic shorteni ng, well documented in Plautine Latin (ca. 200 BCE), as an optional process which resulted in the lexicalization of many forms as in Classical Latin while others remained as The examples that he cites (1994, 13), 82

PAGE 83

reproduced in Table 2-6, are all two-syllables wo rds in which the final vowel has no inflectional role un like the -stems mentioned by Lindsay. Table 2-6. Iambic Shorteni ng in Early Classical Latin or cit mod ben mal du fast only good bad two s r modic long timid amb late moderate long timid both Stephens (1985, 240) also notes that the shortening of to is first established in pronouns, lower cardinal numerals, discourse particles, semantically bleached verbs, and the like: they are all nonlexical words, as opposed to full lexical wo rds such as verbs, nouns, and adjectives. Despite the superficial similarities of words such as cit and s r iambic shortening has created two very different words in prosodic terms. Although both have a penul timate accent the latter has the structure HH (two quantitat ive trochees) and the former LL ( one quantitative trochee). Hayes (1995, 120-121) describes iambic shortening in terms of the metrical gr id as illustrated in Figure 2-3. The resulting bimoraic f oot corresponds to the LL pattern. (x .) (x .) m a l e m a l e Figure 2-3. Iambic shortening Mester (1994, 15-17) addresses the issue of extrametricality in his analysis of iambic shortening as part of a discussion on prosodic repair strategies which can have either of two forms: (2.8) SHORTEN: REMOVE(Mester 1994) (2.9) LENGTHEN: ADD(Mester 1994) 83

PAGE 84

These repa ir strategies interact with the constraint s related to final syllable extrametricality given below (Mester 1994, 17): (2.10) Final Syllable Extram etricality: In the sequence ]Wd: (a) avoid foot-head, (b) avoid footing. (Mester1994) Iambic shortening allows a word of the type to pr oduce a final foot, ( ), in violation of part (b) of the final syllable extrametricality constraint ; however, this is preferab le to the violation of part (a) that would result from a configuration of the type ( ) in which the only well formed foot would be the final syllable, leaving the first syllable unparsed, or in Mesters terms, producing initial trapping. Alt hough much has been written about the strength of the initial syllable (Kent 1931), it is also known to be vulnerable when it is pre-tonic. Cases of apheresis in the transition from Latin to Romance (Cross 1934 ), although infrequent, provide some evidence of the inherent weakness of an initial unaccented syllable, particul arly in polysyllables such as Cat. botiga Cast./Por. bodega (< p th ca, ae, f. storehouse). Cretic shortening, defined by Mester (1994, 30) as the m onomoraic scansion of heavy syllables at the end of words of the form [... ] Wd ... as [[ ]]. Unlike iambic shortening examples of cretic shortening emerge regularly in the imperative and other verb forms, including those whose final syllable is long by place (M ester 1994, 31). Application of the REMOVEconstraint results in a more optim al foot structure, that is, ( ) < >. Here the final syllable is treated as extrametrical thereby guaranteeing that stress will be assigned to the antepenultimate syllable of words such as d cit say. Using an OT tableau and the constraints established by Jacobs for determining Early Classical Latin word stress, Table 2-7 shows that failing to parse the last two-syllables (The firs t is medially trapped and the s econd is extrametrical.) does not result in disqualification of the footing propos ed by Mester. Again, it has been assumed that 84

PAGE 85

there are two underlying, invi olable constraints: FTBIN and RHTYPE(T). Heavy syllables satisfy FTBIN by having two moras. Table 2-7. Word Accent and cretic shortening NONF QS W/L W/R H/R H/L PARSE/HL / d cit say ( H )L ** (H)(L ) *! H (L ) *! Vowel Reduction Various weakening phenomena are indicative of the effect of Early Latin strong initial stress. Changes in vowel quality provide evidence of a pattern of strong initia l, particularly in derivation or compounding, where related lexical ite ms show alternations between the full value of the vowel and the reduced value. Reduction of unstressed vowels is a well documented process which occurs in many of the worlds la nguages. Crosswhite (2001: 21) views vowel reduction as a bipartite process which may result in prominence reduction or contrast enhancement. She has compiled typologies of vowel reduction based on a survey of 47 languages. For languages with five vowels, as Latin, 24 different patterns are attested (Crosswhite 2001: 218-222) although the particular pattern of reduction suggested in Janson (1979, 48) is not attested in Crosswhites inve ntory. For Latin, Janson gives the following cases of vowel reduction: e > i, o > i, a > i, u > i Steriade (1988: 376) deviates somewhat from Jansons account in that she adds as variants: a > e as in consecro (cf. sacrus ) in the environment of /r/ and o > u (no examples provided). Accord ing to Crosswhites classification the vowel reductions proposed here might be construed as prominence reduction, particularly when more open vowels such as /a/, /e/, /o/ are reduced. The motivation for /u/ (to /i/) is less apparent since both /i/ and /u/ are high vowels a nd, consequently, less prominent acoustically. 85

PAGE 86

However, /u/ has special phonetic features b ecau se it is rounded. The elimination of these special articulatory gestures in essen ce makes the vowel less prominent. Harris and Lindsey (2000, 190) note that a languages maximal vowel inventory is not sustained in all phonological cont exts. Although the reduced inventory is often a concomitant of stress, they caution against the construction of a cause and eff ect relationship: it would be far from the truth to assume that all cases of systematic contraction are c onditioned in this way. The interaction between phonology and morphology is evident in the case of Latin. Janson (1979: 48-52) points out that most examples of vow el reduction come from prefixation of verbs, for example, facio-inficio, rapio-corripio, capio-incipio, lego-deligo Assuming that such compounds were created during the formative peri od of Latin, word initial stress, then, might have been the catalyst of change. However, the examples cited in Janson raise another possibility, that of vowel harmony. The two sound changes seen here are a>i and e>i That /i/ should be the favored outcome is not surprising. According to Harris and Lindsey (2000, 189190) vowel systems cross-linguistically show a pred ilection for the triangular pattern, that is, /i, a, u/. For Latin, changes of the type a>i and e>i are well attested. That /a/ does not retain its articulation features may be owing to the influe nce of a following high front vowel evident in the first three examples although Harris and Lindsey (2000, 190-191) suggest that this is most likely to occur when the influencing vowel is in a dominant nucleus. There are many examples from Romance, however, that show the influence of a palatal element (non nucleic) on a preceding vowel such as vendimia < vind m a vintage. Here, it is the accen t bearing vowel that is raised by a palatal glide in the onset of the following syllable. The change o> i according to Janson (1979: 51-52) is poorly attested wi th the exception of the adverb ilico > in that very place = (cf. locus ) and the change u> i is equally infrequent, occurring in examples such as caput head 86

PAGE 87

(nom sg.), capitis (gen. sg.). Note that in both cases there is another high front vowel present either before or after the vowel that undergoes raising. The vowel reductions seen here do not affect the rhythmic structure of the syllables in questions because the moraic structure remains unchanged. However, loss of acoustic prominence for the stem vowel which is now unstressed, in the case of the prefixed verbs, is evident. Treatment of Words of Greek Origin There are two pertinent processes in the treatment of Greek borrowings in Latin: imposition of the Latin Stress Rule and vowel weakening. Lindsay (1894, 155-156) observes that prescriptive grammars recommend Greek accentuation for those cases where Latin and Greek accent patterns do not coincide. However, the avoidance of stress on the final syllable is well documented by the treatment of Greek words with final stress. These examples (Baldi 1999, 192-193) show that primary accent is now in conformity with the Latin Stress Rule. Latin Greek bardus stupid blennus blockhead h l rus cheerful moechus adulterer paed ggus pedagogue Vowel weakening in Greek loanwords, often a precursor to syncope is considered a chronological marker by Baldi (1990, 190) because the process in Latin was probably no longer productive by 250 B.C.E. However, the variab le treatment of vowels in Greek borrowings continues well beyond this time period and cannot be construed to be a reliable method for dating the entry into Latin of part icular words. An example is camera vault < where the reduction a>e is well attested in Latin (Garcea, Corpus Grammaticorum Latinorum (CGL), 87

PAGE 88

s.v. cam era2) yet the pronouncements from CGL are inc onsistent and suggest that the two forms camera/camar a may have co-existed for some time. Bi ville (1995, 92) sees another possibility: Cam(m)ara peut sinterprter comme un remprunt au grec, ou comme une forme vulgaire rsultant de louverture devant r du e de camera forme latinise, depuis longtemps acclimate. The outputs in Romance3 indicate that the form camera is the source of cambra in Catalan and Old Spanish. Under the prosodic constraints for Classical Latin, stress is assigned to the antepenultimate syllable of the word because the penultimate vowel is short. The penultimate, unstressed syllable is now vulnerable and a candida te for elision as attested by Catalan and Old Spanish cambra (with epenthetic /b/). However, only the variant cmara is found in Portuguese (cf. Spanish cmara and Catalan cmara, dubbed a castellanisme in DCVB, s.v. cmara). Vowel reduction without syncope is seen in a > i in Latin m ch na, ae f. machine < (Doric) (Baldi 1999, 190) as well as pt na ( p t na), ae f. broad, flat dish (< ) are both attested in Plautus (Lewis and Short, s.v. m ch na, pt na; Biville 1995, 380, 447). These cases are of interest because they are early borrowi ngs with a different outcome, vowel reduction without syncope. There are two other aspects of the vocalis m of Greek borrowings that should be mentioned: exceptions to vowel reduction and alteration of diphthongs. That vowel reduction is not exceptionless is seen in c r andrum or -dron, i, n.; also dru s, i, f. coriander (< ~ ~ ). The three reflexes in Castilian coriandro, cilantro, cilantro (DRAE)all show an unaltered post-t onic vowel as do Catalan variants coriandre and 2 The Corpus Grammaticorum Latinorum (CGL) cites Verrius Flaccus (ca. 55 BCE-20 CE) from Charisius (fl. 4th c. CE): camara dicitur, ut Verrius Flaccus adfirmat / non camera ; cf. Appendix Probi l. 84 camera non cammara 3 According to CORDE (s.v cambra; cmara), both cambra (which could only have come from camera ) and cmara are attested in Castilian in the mid twelfth century. For Catalan cambra is documented in the fifteen century (DCVB, s.v. cambra). In Davies Corp us do Portugus, the earliest attestation for cmara is from the fourteenth century. The form cambra is not reliably attested for Portuguese. 88

PAGE 89

celiandre However, Portuguese co entro suggests an alternate form with vowel reduction, *coriendrum On the other hand, the complete assimila tion of Greek words is seen in the fact that diphthongs also undergo alternations typical of native Latin words. The replacement of the diphthong /aj/ with long vowel as in ol va (< ) has as counterpart the alternations seen in the noun poena ae, f. (< ) and the related verb p n o (arch. poen o), vi or ii, tum, to punish (Biville 1995, 380). Many Greek loanwords correspond to a later pe riod chronologically and belong to very different contexts in terms of register. A ccording to Sedgwick (1949) Greek was not limited geographically to the eastern empire as it was widely spoken among slaves and freedmen in the west. Baldi (1999, 192-193) provides an extensive list of Greek loanword s related to everyday life such as canth rus, i, m ., < which enters Ibero-Romance as Cast. cntaro, Port. cntaro and Cat. cntir The Catalan form, as well as Italian, cntero seem to support the alternate form cantherus given in Biville (1995, 93). In this case the original antepenultimate accent has been preserved but post-tonic vowels are subject to reduction in some outputs. In the case of pr s tus < parasite results are variable with respect to assignment of primary accent. The Greek antepenultimate accent appears in learned Sp. parsito Cat. parsit (but Port. parasito with penultimate accent). Similar vaci llation in placement of primary accent is seen in the Ibero-Romance reflexes of Latin p l c nus/p l c nus, i, m. pelican (< ). Penultimate stress occurs in Ital. pellicano Cat. pelic Por. pelicano while Span. pelcano seems almost a case of hypercor rection (note that DRAE, s.v. pelcano also gives pelicano with penultimate accent, as a variant). A major impetus in the borrowing of Greek words was the rise of Christianity. Greek borrowings appear as proper names as well as l iturgical terms. Sedgwick (1949, 87) notes that 89

PAGE 90

Greek accen tuation is preserved in many personal na mes such as Ananas, Mara, Sopha which he attributes to unbroken ecclesiastical tradition. In liturgical Latin the following Greek words with antepenultimate accent ab y ssus, er mus, id lum are regularly scanned with short vowels in the penultimate syllable in order to avoi d the conflict generated by a quantity sensitive accentual system that assigns stress to a heavy penu lt. It is through the Greek loanwords that it is possible to establish two important facts relative to the Latin accent: first, that the early accent was word initial and stress-based. This is seen in the reduced vowel quality of Greek loans that follow the a>i, e>i pattern noted by Janson. Second, the quan titative nature of the Latin accent emerges in the reduction of penultimate vowels when they do not form the stress bearing foot. The change from word initial stress accent to right edge quantitative trochee is supported by evidence from both native Latin words and Greek loanwords. Strategies of Moraic Preservation Previous discussion has centered on vowel reduc tion and loss as a mechanism for prosodic repair. Mester (1994, 15) presents vowel lengthening as another prosodic repair strategy which he labels ADD. Vowel lengthening is operative in early Latin as seen in the case of pn < *po-sin cf. positus. Following loss of the pos t-tonic short vowel i -, the /s/ in coda position weakens and is lost. However, the preceding vowel ois lengthened. Such cases of vowel lengthening were common in Latin and often involv ed the loss of /s/ in coda position. Such changes are well attested cross-li nguistically. Kavitskaya (2002, 4) summarizes the process as follows: C1VC2 CV:. She notes that the process neit her adds nor removes syllables but the affected syllable now becomes an open syllab le with a long vowel. This type of vowel lengthening is more likely to occur in languages that already have a system of contrasting short and long vowels (Blevins 2004a, 154) : I suggest that the greater likelihood of compensatory 90

PAGE 91

lengthening to devlop in a language with pre-existing length contra sts than in a language without such contrasts reflects potential prim ing eff ects on the learner in the course of language acquisition. Leumann (1963: 103-10 5) provides other examples su ch as the reduction of the clusters -ns-, -nfs-. Lengthening before nasals is construed by Baldi (1999: 258) as evidence for the weakening of the nasal before a fricativ e. The well-known inscriptional variants c sul ~ consul and c sor ~ censor constitute further testimony of the relative weakness of the nasal before a fricative and support th e view that compensatory lengthening was an active repair strategy to insure perseverance of the optimal foot, the bimoraic trochee. In further support of the moraic trochee is the fact that vowel lengthening is also seen in monosyllabic words. Baldi (1999, 258) states th at Latin had a minimum moraic requirement of two moras per word. This requirement could be fulfilled by a heavy nuc leus, that is, a long vowel or diphthong, or a short vowel followed by a consonant. The alternating forms of prepositions such as /ex out of and /ab away from illustrate the prosodic exchange of long by nature and long by position. On the other han d, Lachmanns Law poses some problems with regard to strict bimoraicity. Lachmanns Law, the lengthening of the vow el in the formation of the past participle, according to Baldi (1998, 259262), has its origins in the case of verb roots ending in /d/ or /g/. Before the participial suffix -to, the final consonant of the verb root undergoes voice assimilation and the preceding vowel is lengthened, as in the following cases: V (1st pers. sg. present ) V (past participle) reg r ctus rule ed sus eat The lengthening seen in these ca ses contrasts with verbs like faci, factus do where there is no vowel lengthening. The SPE type rules of Kiparsky (1965) and King (1969) to describe Lachmanns Law, cited in Baldi (1999, 260), ha ve the disadvantage of overproduction. 91

PAGE 92

1. Vowel lengthening: [-cons] [+long] /__ +obst +obst +voice -voice 2. Voice ass imilation : [+obst] [ voice] /__ +obst voice However, without further constraints the above rule would produce th e unattested form gr x ( grex, gregis herd) and sim ilar constructions. Th e exceptions to Lachmanns Law are not limited to the nominal system. In the verb paradigm the opacity of Lach manns Law is seen in forms (Baldi 1999, 261) such as such as fund, f sus pour and find fissus split. This suggests that vowel lengthe ning was not a phonological rule or that it was a rule that was extant for a limited period of time. Baldi (1999, 261-262) refers to later studies, in particular Watkins (1970), to suggest that vowel le ngthening here is a product of analogy and morphological conditioning. Watkins (1970, 63) is mindful of the role of chronology. For example, he states that fund f sus and tund, t sus beat are late creations in Latin and also notes that t sus alternates with older t nsus. In the latter case one sees the familiar pattern of vowel lengthening to compensate for an elided nasal before a fricative. While Lachmanns law may not be phonologically grounded it does imp act the bimoraic limit of the quantitative trochee. In the case of the prepositions cited above, the alternate form with long vowel, for example /ab can be viewed as lengthening to preserve the minimum moraic count for a prosodic word. The lengtheni ng of the stem vowel in the case of the passive participle has the effect of producing in some instances a superheavy syllable as in reg r ctus. If the iconic root form for verbs is considered to be TeR, TeT, or TeRT, the formation of the past participle will always result in a syllable that is considered heavy because it has a coda. If the nucleus is a long vowel or a fall ing diphthong the resulting syllable now consists of three moras. The trimoraic syllable does occur in ot her env ironments in Latin, for example, maestus 92

PAGE 93

sorrowful. Allen (1973, 137) stat es that in the case of m edial s +stop, for example pestis plague, the syllable division separa tes the two consonants, that is pes.tis so that one forms the coda of the preceding syllable and the other is the onset of the following syllable (see also Devine and Stephens 19 77, 132). In the case of maestus, the resulting first syllable, the only parsable foot if extrametricali ty is assumed, is superheavy. The superheavy syllable in Latin is relatively infrequent and appears to be morphologically generated, as is the case of the sigmatic perfects as well as the past particple discussed above. Given the special properties of /s/ it is probable that the coda is viewed as extrametrical. In a strict bimoraic interpretation of the qua ntitative trochee as the basis of Latin prosody, discussed in Mester ( 1994, 29) who proposes (QT) as the parsing foot structure, the syllable maescannot project a foot because it is *( ). Mester (1994, 40) recognizes this as an issue in his discussion of syncope: In the output of syncope, long vowels in the preceding syllable are usually shortened, in conformity with the generally marked character of superheavy CVVC syllables in La tin. The case of jurg (
PAGE 94

Allen (1973, 66) suggests that what is im portant for Latin is to determine whether or not a syllable is heavy. He terms hypercharacterization the cases of long vowel in a closed syllable. Such syllables are already long by position and reduc tion of either the vowel or the consonant is seen in different chronological phases of Latin such as prehistoric am ntem > am ntem (Allen 1973, 141) and caussa, c ssus in Cicero and Vergil but later causa, c sus (Allen 1973, 66). The syllables in question remain heavy in prosodic term s. Loss of superheavy st atus does not affect footing because it plays no role in the metrical structure of Latin. There are well known cases where the superheavy foot impacts assignment of word level accent, for example, in Classical, Damascene, and Cairene Arabic, presented in Goldsmith (1990, 197-198) (further discussion in Appendix B). Classical Latin Accent: From the Metrical Grid to the OT Tableau The rise of metrical theory stimulated the examination of prosodic systems in many languages including Latin. The move to the metrical grid and away from tree structures offers many advantages. As McCarthy and Prince ( 1996, 1) note It is a commonplace of phonology that rules count moras ( ), syllables ( ), or feet (F) but never segments. The grid marks indicate stressable elements w ithin the prosodic word and cont ruct a base line of moras or syllables. Higher level rows build on the familia r prosodic hierarchy reproduced in Figure 2-5. Word (Wd) | Foot (F) | Syllable ( ) | Mora ( ) Figure 2-5. The prosodic hierarchy The grid representation may vary slightly depe nding on the prosodic properties of individual languages. For Latin the mora level is important in construction of the pr osodic foot but it also 94

PAGE 95

m eans that not all elements on the F Row can be pars ed if bimoraicity is strictly interpreted. Following Hayes (1995, 91-92), with modification fr om Mester (1994, 29), one can establish the precepts that govern construction of the grid as it applies to the accen t of Classical Latin. 1. Final syllable extrametricality: < > /___ ]word 2. Foot construction: a. Form a moraic trochee from R to L b. Trochee = min/max 2 3. Word layer construction End Rule Right Prosodic trapping is inevitable in the mapping of Latin words to the grid. As illustration, words of two and three syllables in Latin are examined. The conventional L(light) and H (heavy) differentiate syllables with short vowel and no coda from syllables with long vowel or vowel plus coda. As added information, syllables in which the nucleus is a vowel long by nature or a diphthong are coded HV and syllables that are heavy due to presence of a coda consonant are coded HC. Foot construction is problematic when initial or medial light syllables are stranded or trapped. Such is the case in 2A and 3C where the initial light syllable is stranded. Similarly, in 3B medial trapping occurs. Table 2-8. Prosodic confi gurations in words of two and three syllables Two-syllables 2A: L 2B: HV 2C: HC ac .tem cau .lem den .tem skin, f. cabbage, m. tooth, m. Three syllables 3A: L.L 3B: HV .L HC 3C: L. HV HC 3D: HC .HV HV 3E: HC .HC HV c .m .tem bm .l .tem bar .b.rem ad.l .rem ac .hor .tem vec.t .rem v.c .lem in.sec .tum d .fen .sam companion, m./f. soldier, m. tree, f. pain, m. court, f. bearer, m. vowel, f. insect, n. defense, f. aInitial light syllable is stranded. bMedial syllable is trapped. 95

PAGE 96

The light syllable b etween the initial heavy syllable and the extrametrical final syllable cannot form a foot because it lacks the minimu m moraic count. Mester (1994, 17) discusses trapping relative to iambic shortening and proposes an OT type solution: it is better to parse the final foot than to create an illicit trochee. The rules for constructing the metrical grid to assign stress (Hayes 1995 and Mester 1994), are readily rendered into OT constraints that (2.11) Define the prosodic word Lx Pr: A member of MCAT corresponds to a PrWD. (Jacobs 2000, 2003b) A member of the morphol ogical category MCAT (root, stem, lexical word) is a prosodic word. (2.12) Define foot form FtBin: Feet are binary at some level of analysis ( ) (Jacobs 2000, 2003b) RhType (T): Rhythm type is trochaic QT(2,2) : Feet are trochaic with mini mum and maximum of two moras. (Mester 1994) (2.13) Define quantity sensitivity WSP (Weight to Stress): heavy syllables are prominent in foot st ructure. (Jacobs 2000) QS (Quantity Sensitivity): heavy syllables are stressed. (Jacobs 2003b) Jacobs (2003a, 276 and 2004, 65-69) adapts thes e constraints to crea te tableaux that model the main stress of Classica l Latin. In its streamlined fo rm (Jacobs 2004, 65) the tableau appears in Table 2-9 with one constraint adde d to bar the uneven trochee, *(HL. Other constraints are defined below. Note that Jacobs has replaced the more familiar ALIGN family of constraints with W/R and H/R. (2.14) *(HL): A foot must not consist of a heavy and a light syll able. (Jacobs 2004) (2.15) W/R: Align the prosodic word and prosodic foot to the right. 96

PAGE 97

(2.16) H/R: Align the head foot to the right in the prosodic word. It is assumed that the top tier constraints, FTBIN and RHTYPE(T) are undominated. Since stress in Latin is dependent on the nature of the la st three syllables all possible Latin words are represented by the nine canonical sh apes in Table 2-9 (Expansion to the left is assumed.) The final extrametrical syllable is represented by ; the accented syllable is in bold and underlined. Words of two-syllables, always with penultimate accent, form a subclass of the nine forms. Monosyllables are minimally bimoraic and are an exceptional class in that some words cannot carry word level accent; that is, they are not prosodic words. The clitic pronouns in IberoRomance are a good example of unstressable elemen ts that cannot be assigned the status of a prosodic word. Table 2-9. Parsing of thr ee-syllable end of word NONFIN H/R W/R QS PARSE*(HL) 1 L.(HC ). 2 L.(HV ). 3 (HC).(HC ). 4 (HC).(HV ). 5 (HV).(HC ). 6 (HV).(HV ). 7 (L .L). 8 (HC ).L. 9 (HV ).L. avoided by extrametrical final avoided by extrametrical final and non parsing of trapped light syllables The first six forms in Table 2-9 have penul timate accent secured by the bimoraic nature of the penultimate syllable. Whether the preceding syllable is light or heavy (long vowel, falling diphthong, coda consonant) is immaterial. Howe ver, the light antepenultimate syllable is vulnerable if not word initial. In terms of OT constraints the unparsable light syllable incurs violations of PARSE, a low ranking constraint, as does the final extrametrical syllable. Yet early Latin syncope and late Latin syncope demons trate the inherent weakness of a pre-tonic or postonic medial short vowel. The fate of these light syllables in Late Latin and Romance has a 97

PAGE 98

profound effect on sy llabic count, syllabic weight and the relative position of primary accent within the word. Furthermore, forms 2, 4, and 6 in Table 2-9 depend on vowel length to assign stress to the penultimate syllable. What is the expected outcome of loss of quantitative difference between otherwise like vowels? Sy llables that are long by position also face quantitative challenges as the coda consonant is modified or elided entirely in the transition to Romance. If determination of word accent is dependent on those factors that create a heavy syllable, in their absence, the rules that govern computation of accent have become opaque. Learnability of the Latin Accent The questions posed above with regard to rule opacity are studi ed in Apoussidou and Boersma (2003, 2004). They survey the constrai nts developed by Prince and Smolensky (1993), Tesar and Smolensky (2000), and Jacobs (2000) and compare their efficacy in a simulation of the learnability of Latin stress. At issue in the study is not so much the nature of the constraints themselves but their effect on two acquisi tions patterns (Apoussidou and Boersma 2004, 107108): Error-Driven Constraint Demotion (EDCD) and the Gradual Learning Algorithm (GLA). EDCD presumes that error detectio n is a catalyst for reranking of constraints. Constraints are demoted if the learner perceives lack of congr uence between a perceive d correct form and a produced incorrect form. The process contin ues until the grammar has achieved a state of perfect balance between per ception and production. GLA depends on adjustments in the hierarchy in accord with evidence from input The adjustments are gradual and reflect vacillation in production of the l earner who may alternately revert to an old, incorrect form while occasionally producing the new, correct form. For purposes of the simulation Apoussidou and Boersma (2004) focus on the acquisition of primary stress in Latin. The virtual learners of this computer simulation were exposed to 10,000 to 40,000 tokens containing words of two to seven syllables. The GLA lear ners were overall far 98

PAGE 99

more successful in learning prim ary stress plac ement from the input forms. Curiously, the virtual learners (Apoussidou and Boersma 2003, 144) more easily assigned word accent with an uneven trochee than when constrained by strict bimoraicity. Jacobs (2000) admits the uneven trochee as a possibility a nd prefers to define the trochee in te rms of at least bimoraic. Using Mesters notation, Jacobs would favor QT(2,3). Although concerned primarily with language acquisition EDCD and GLA are also a promising approach for diachronic studies. The problems experienced by virtual learners may explain some of the vacillations seen in stress placement in the transition to Romance as well as competing form s that exist in the modern languages. As has been suggested earlier the musical nature of th e binary prosodic foot calls for a perceived contrast between the strong or accented element and the unstressed element. Apparently, this contrast is more easily mapped onto a configuration of HL as opposed to LL. The Transition from Classical Latin to Late Latin As is the case with Classical Latin, knowledge of lexical accent after the fall of the Roman Empire relies on documentary evidence of different types: the descriptions of grammarians, legal and testimonial documents, accidental evidence such as inscriptions and graffiti as well as metrical evidence. Cases of divergences have particular phonological environments: vowels in hiatus, medial clusters consisting of a voiced stop with /r/, verb compounds in which accent moves to the verb stem from the prefix, a nd stress shift in the pronunciation of some of the numerals, notably v ginti twenty and tr ginta thirty. The numerous cases of syncope and related phenomen a indicate the presence of a stress accent, perhaps limited by chronology and/or register. Loss of distinctive vowel length does not permit the continuation of a quantitativ e based accentual system because there would be far too many exceptions coming from cases of retention of original locus of accent on a heavy syllable which is now indistinguishable from a light syllable. Furthermore, th e descriptions of grammarians 99

PAGE 100

with regard to accent b egin to employ different vo cabulary and terms of reference. In describing a twelfth century treatise on pronun ciation from a Cistercian m onastery Thurot (1869: 392) remarks that on ne trouve aucune mention de s accents aigu, grave, ou circonflexe ... Il nest question que de laccent (accentus) en gnral. Thurot notes the ambiguity of the vocabulary which describes the contrast between stressed and unstressed syllables. Although words such as mora, instanter suggest duration rather than pitch, elevabitur and deprimentur could be construed as references to the classica l tradition which designated accent as acutus, gravis, circunflexus The transition from Latin to Romance marks a second major change in the stress system of Latin with significant consequences for vowel reduction/loss and syllable structure. The current chapter has reviewed the movement of word accent from alignment with the left word edge to alignment with the right wo rd edge excluding the final syllable. 100

PAGE 101

CHAP TER 3 FROM QUANTITY SENSITIVE TO STRESS ACCENT Chronology of Latin In the opening chapters of his study of La te Latin, Lfstedt (1959) struggles with questions of chronology, nomenclatu re, and definition. He acknowledges that the transition from Latin to Romance must have been gradual and imperceptible (1959, 14), beginning perhaps in the third century. Mill er (2006, xvii-xix) provides a detaile d division of Lati n, based primarily on periods of literary activity. Table 3-1 is an adaptation of this chronology (centuries are expreseed as C1 (=1st century). For the pres ent study, texts from the time periods labeled Late Latin and Medieval Latin are the most pertinent. These texts are often non literary but provide important evidence as to the state of the language at the time. Those cited in this chapter are indicated in Table 3-1. Table 3-1. Latin historical periods Period Dates Key Works 1 Archaic < -100 2 Ciceronian -100 to -43 3 Augustan -43 to +14 4 Silver Age 14 to 117 5 Antonine 117 to 192 6 192 to 337 7 337 to 430 8 430 to 636 Appendix Probi (C4) Egeria, Peregrinatio ad loca sancta (late C4) Claudius Hermeros, Mulomedicina Chironis (late C4/early C5) 9 Late Latin C7-8 CE Medieval Latin C9-13 CE Humanistic Latin C 4-16 CE Vulgar Latin, not reflected in the chronological grid, is an umbrella term, perhaps better used as plural, to describe varieties of popular language that co rrespond to an oral rather than written tradition. According to Hall (1974, 71) th e split between literary and popular Latin is a byproduct of the codification of language that oc curred as concomitant of a nascent tradition of 101

PAGE 102

written literature: Clas sical usage thenceforth became more and more static, while popular speech kept on developing. However, Vn nen (1967, 5) warns against the temptation dexagrer lopposition entre la langue parle et la langue crite; il existe notamment de nombreux points de contact entre le langage populaire et le styl e potique. These points of contact, in particular the refl ection of popular language in the written tradition, become valuable sources of information relative to ongoing language change. In subsequent discussions the imperfect term Vulgar Latin is the preferred de signation; as used here it is understood to be a reference to the popular stratum of the language that spans severa l centuries. This use follows the definition of Herman (1975, 16): nous appel ons latin vulgaire la langue parle des couches peu influences ou non influences par lenseignement scolaire et par les mod les littraires . Notre dfinition ne comporte aucu ne limitation chronologique. Sources of evidence of the popular stratum are of two types: direct and indirect (Hall 1974, 73-74). Examples of direct evidence are 1) literary works that include representation of popular speech, for example, early Latin dram atists Plautus (ca. 284-204 BCE) and Terence (185-50 BCE), and the C na Trimalchi nis from Petroniuss Satyricon (1st century CE); 2) Late Latin authors whose language refl ects the norms of the cultivat ed Latin of Christian authors together with elements of popular language as Egerias Peregrinatio ad loca sancta (late 4th century CE)4; 3) grammatical annotations of the type found in the often cited Appendix Probi (3rd or 4th century CE), a handwritten list of 227 common spelling/grammar errors added to a grammatical treatise, Instituta Artium attributed to a grammarian Probus (see Appendix A). While the Mulomedicina Chironis is of greater import for the study of medical and veterinary 4 In describing the polemic concerning Egerias use of language Wilkinson (1999, 2) asserts that although there is no evidence that she had studied the Latin classics, there is ample evidence that she was familiar with authors of the Church. Her rhetorical devices echo those of the Bible and of formal prayer. 102

PAGE 103

term inology, it is also characteri zed by imperfectness of language when the measuring stick is Classical Latin. Sihler (1995, 1617) notes that these Late Latin authors did not set about to write in a colloquial register rather they employ the conventions of Latin grammar to the best of their knowledge and frequently acknowledge deficiencies in self-deprecatory comments. However, he cautions that none of these documents can be trusted as a full and faithful representation of contemporar y speech (Sihler 1995, 17). Indirect evidence, on the othe r hand, is found outside of Latin proper. Words of Latin origin which entered the lexicon in other languag e families such as Celtic, Germanic, and Slavic, in the period of the early empire often aid in dating certain sou nd changes, for example, German Keller (
PAGE 104

and the consequences for segm ents that are not in prominent position. The first set of changes to be studied concerns vowel reduction in its most extreme, form, vowel loss. Syncope in Vulgar Latin It is possible to argue that widespread syncope in Western Romance corresponds to the desire to create a more desirable rhythmic st ructure within the new paradigm of a non quantity sensitive stress accent. Words with proparoxyton ic stress routinely lose the post-tonic vowel nucleus when that vowel is short as seen in th e words (cited in Jensen 1999, 138 as examples of syncope) in Table 3-2 in their Clas sical Latin and Vulgar Latin form s as well as their reflexes in Ibero-Romance. It should be noted, however, that many of the examples in Table 3-2 show alternations between disyllabic and trisyllabic va riants in Classical Lati n. In Lewis and Short disyllabic variants are documented as follows: caldus (s.v. c l dus) is the predominant form in the August period; postus (s.v. p no) occurs in Lucretius (99-55 BCE); soldus (s.v. s l dus) is attested in Horace (65 BCE-8 CE). The Cla ssical Latin forms in Table 3-2 represent the accusative singular with final /m/ in parent heses since the nasal was probably no longer pronounced by the end of the fifth century C.E. (Harris-Northall 1990, 38). Since the input forms are all of the type (C)V.CV.CV, syncope then implies also reso lution of newly created medial clusters. Table 3-2. Syncope of posttonic short vowels in Latin Classical Latin Vulgar Latin Castilian Catalan Portuguese 1 c l du(m) hot drink caldu caldo cald (Old Cat.) caldo 2 c l phu(m) blow colpu golpe colp golpe 3 c lu(m) eye oclu ojo ull olho 4 ps tu(m) position postu puesto post posto 5 s l du(m) coin soldu sueldo sou soldo 6 sp c lu(m) mirror speclu espejo espill espelho 104

PAGE 105

Clusters of the type liquid + stop and /s/ + st op appear to be stable. Because the liquid + stop groups are inadm issible as syll able onset they are syllabifie d as coda and onset of two contiguous syllables as in examples 1, 2, and 5. Example 5 for Catalan deserves some comment. In modern Catalan syllable final /l/ acquires its place of articulation from the following consonant (Hualde 1992, 398). The presence of Old Cat. cald as well as colp indicates that syllable final /l/ to does not re gularly undergo vocalization or de letion. Rather, the development of the glide /w/ from /d/ in coda position is consistent with the treatment of word final and syllable final voiced coronal stops as in peu foot < pedem (Wheeler 1988, 173). The preceding /l/ in this case is absorbed by /w/. The cluster /s/ + stop in example 4 is heterosyllabic because it is not a permitted onset in any of the modern Ibero-Romance languages. Additionally, the loss of /s / before another consonant in Fren ch suggests that it was a syllable coda rather than part of a complex onset. According to Pope ( 1966, 151) syllable final /s/ assimilates in voicing to followi ng voiced obstruents and is even tually lost; however, loss is usually accompanied by vowel lengthening. Comp ensatory lengthening suggests a device to restore syllabic weight that coul d only exist if the s + obstruent clusters were heterosyllabic. In metrical terms the input form in Table 3-2, (C)V.CV.CV represents L1.L2.L3 parsed as (L 1.L2) (accented syllable underlined; angled br ackets indicate extrametricality; and parentheses a prosodic foot). Jacobs (2004, 75) contends that th e vulnerability of L2 proceeds from its status as the weak member of a trochee. The complete list of cases of syncope in the Appendix Probi (Vnnen 1967, 254-257; repro duced in Appendix A) in Table 3-3 demonstrates that the common fact or is, indeed, a short vowel in post-tonic position. The items that appear in square brackets have been added to show the derived form usually a diminutive of the preferred Latin word, that has given rise to the Vulgar Latin fo rm. It is noteworthy that in 105

PAGE 106

severa l cases the original Latin word has become subject to confusion due to homonymic status or near homonymic status resulting from phonological changes. Line 83, auris ear is a good example; the diphthong /aw/ becomes /o / in Vulgar Latin and resulting ore is now identical to ore(m) acc. sg. mouth. The augmented form, oricla may therefore be lexically motivated. In any case, it also shows syncope in its development from the diminutive auricula Similarly, coda simplification in fax torch would result in homonymy with fas dicates of religion. Vulgar Latin, as seen in Table 3-3, attests to the fr equency of syncope in the spoken language. Table 3-3. Cases of sync ope in the Appendix Probi 3a speculum non speclum. 4 masculus non masclus. 5 vetulus non veclus. 6 vitulus non viclus. 7 vernaculus non vernaclus. 8 articulus non articlus. 9 baculus non vaclus. 10 angulus non anglus. 11 iugulus non iuglus. 29 avus non aus. 53 calida non calda. 54 frigida non fricda. 62 flavus non flaus. 83 auris non oricla. [
PAGE 107

one m ost propitious for syncope, the weak position of a foot seen in Figure 3-1 (The stressed syllable is underlined and the final, extr ametrical foot is in angled brackets.) Classical Latin Vulgar Latin Output after Syncope (t )bu (ta .bu) tbla (ca .li) (ca .li) cldu Vowel length is lost Stress assigned before syncope Figure 3-1. Parsing of candidates for sync ope in Classical L atin and Vulgar Latin. Mester (1994, 41) questions whethe r or not the accentual system of Vulgar Latin was, indeed, quantity sensitive at this time. He suggests that in the absence of length contrasts only a closed syllable could be construed as heavy, and, furthermore, only when in penultimate position. All other syllables were regarded as light (1994, 42). The parsings in Figure 3-2 illustrate the crucial difference between Classical and Vulgar Latin. The result of syncope in examples from the Appendix Probi in Table 3-4, demonstrates that in many but not all of the preferred Vulgar Latin forms, the primary accent now falls on a heavy syllable. This cannot be stated as an exceptionless constraint of the type QS (Quantit y Sensitivity: Heavy syllables are stressed) due to the complexities of syllabification of muta cum liquida medial clusters discussed below. As the comparative data of nouns of Latin origin in Castilian, Catalan, and Portuguese are examined in subsequent chapters it will become apparent that QS no longer plays a major role in prediction of stress. While the primary accent often coinci des with a heavy syllable, the heavy syllable in itself is not sufficient to attract the word accent. What is without exception in th ese cases of post-tonic syncope is that the loss of a vocalic nucleus has moved the word accent one syllable clos er to the right word edge. Antepenultimate stress in essence has been eliminated as a possibl e accentual pattern for the data set in Table 3-4. This is not generalizable for the entire corpus of Latin nouns that pa ss into Ibero-Romance, 107

PAGE 108

however. Except where indicated the item s in the Appendix Probi co rrespond to the citation form of nouns and adjectives that is, nominative singular. The final foot, considered extrametrical in Classical Latin, is coded as X because it cannot project a foot. In determining whether a syllable is light or heavy it will be assumed that onsets are maximized and codas are minimized. The stressed syllable is underlined. It should be remembered that for the input form (Classical Latin) the H type syllable may also correspond to a long vowel. A gloss in English has also been added to help identify the words as probable high freque ncy items because they relate to everyday objects and activities. Table 3-4. Changes in prosodic form as a result of syncope Prosodic Form Examples Words of 2 syllables: L X H H X H 29 62 96 avus non aus. grandfather flavus non flaus. golden yellow nubes non nubs. cloud Words of 3 syllables: L LX L X H LX H X 3 5 6 9 11 53 111 130 133 142 200 201 215 4 10 54 speculum non speclum. mirror vetulus non veclus. little old man vitulus non viclus. calf baculus non vaclus. staff iugulus non iuglus. collar-bone calida non calda. warm water oculus non oclus. eye tabula non tabla. board fax non facla. [
PAGE 109

How syncope actually af fects the assignment of word stress can be shown by placing templates for two, three, and four-syllable words fr om Table 3-4 in an OT tableau. For Classical Latin, Jacobs (2003b) suggests that there are two types of constraints to compute primary accent There are the undominated constrai nts which are assumed and not re flected in the OT tableau. Lx Pr: A lexical item must be prosodically analyzed. FTBIN: Feet are binary at some leve l of analysis (moras, syllables). RHTYPE (T): The rhythm type is trochaic; feet are trochaic. The other ranked constraints concern position of the accented foot and parsing of feet. NONFIN(NONFINALITY): A foot may not be final. W/L: Align prosodic word to the left; align feet to the left. W/R: Align prosodic word to the right; align feet to the right. H/L: Align head foot to the left ; align prosodic word to the left. H/R: Align head foot to the righ t; align prosodic word to the right. PARSE: Syllables must be parsed. It is noteworthy that Jacobs has not included in this analysis a QS (Quantity Sensitivity) parameter (cf. Prince and Smolensky 2004). The positioning constraints are ranked so as to favor placing primary accent close to the right edge but not in a foot aligned with the right edge, that is, NONFIN >> W/R >> W/L >> H/R >> H/L >>PARSE. The ranked constraints will thus prevent L(L L) from being a winning candidate even t hough its final foot is a perfect bimoraic trochee. Table 3-5 shows how the output forms of Table 3-4 would be evaluated by the ranked constraints. Underlining indicates the syllable with primary accent. Table 3-5. Vulgar Latin constraints on accent VL Forms Lx Pr NONFIN W/RW/LH/RH/L PARSE1 Words of 1 syllable (H ) 2 Words of 2 syllables (L X) 3 (H )L 4 Words of 3 syllables (H)(L X) 5 L(L X) 6 unattested (H )LX ** 7 unattested (L L)X 109

PAGE 110

Two for ms, unattested as outputs of syncope of four-syllable words in Table 3-4, have been added to Table 3-5 for purposes of discussi on. In comparing the forms that emerge as preferred in Table 3-4 and Table 35 with the constraint ranking for Classical Latin, it is apparent that the undominated constraints, FTBIN (Feet are binary) and RHTYPE(T) (Feet are trochaic), continue to preside in Vulgar Latin, assuming that the final syllable of a word cannot be considered heavy except in the ca se of monosyllables. Binarity and the left headed rhythmic type are also preferred patterns in modern Iber o-Romance and are consiste nt with the universals of prosody discussed in Hayes (1995). Most m onosyllabic lexical words in Castilian, Catalan, and Portuguese are minimally bimoraic, or under lyingly bimoraic as in the case of Catalan ca dog and pa bread where word final /n/ is regularly de leted in the singular but is realized in the plural. Corresponding words in Portuguese are overtly bimoraic because co and po have a nasal diphthong [ w ]. NONFIN (A foot may not be final), as seen from Table 3-5 (indi cated by !?), is problematic for Vulgar Latin, and for the modern Ibero-Roma nce languages, especially Catalan. When there is a single syllable, and hence, a single prosodic foot, the foot necessarily coincides with word edges both to the right and to the left. Leaving the single sy llable unparsed is not an option because it would violate Lx Pr (A lexical item must be prosodi cally analyzed). Therefore, it is necessary to include this higher ranking constraint in the tableau in order to demonstrate that violation of NONFIN is preferable to violations of Lx Pr. Syncope affecting medial vowels is not the only process that results in monosyllables in modern Romance; one must consider also widespread apocope in nouns of the third declension (as well as second for Catalan). In the case of th ird declension disyllables the resulting form in Ibero-Romance is often monosyllabic depending on satisfaction of sonority sequencing of 110

PAGE 111

com plex codas for Catalan and coda restrictions for Castilian and Portuguese. The Latin nouns in these examples appear in the accusative singular. pane(m) bread> Cast. pan Cat. pa Port. po site(m) thirst> Cast. sed Cat. set Port. sede (final e retained) pisce(m) fish> Cast. pez Cat. peix Port. peixe (final e retained) From the data in Table 3-5 it can be seen that violation of NONFIN should not automatically eliminate a potential candidate because NONFIN is subordinate to Lx Pr, FTBIN, and RHTYPE (T), all of which are satis fied by the monosyllables aus and flaus in Table 3-4. Avoidance of NONFIN would involve paragoge, a violation of faithfulness constrai nts of the type discussed in Pulleyblank (1997), MAX (Every segment/feature of the input has an identical correspondent in the output) and DEP (Every segment/feature of the output ha s an identical correspondent in the input). Therefore, NONFIN must be subordinated to Lx Pr (A lexical item must be prosodically analyzed) in order to allow heavy monosyllables to emerge as preferred candidates in some instances. The double sets of alignment constraints W/ L, W/R and H/L, H/R favor the bimoraic monosyllable because it does not incu r any violations other than NONFIN. The alignment constraints also favor a head foot located near but not coincident with the right word edge. Antepenultimate stress, no longer an operative stress pattern in the data set from the Appendix Probi is disfavored by the modification of the NONFIN constraint. H/R must also be modified so that it penalizes every syllable that intervenes betw een the head syllable of the head foot and the right edge of the prosodic word: (3.2) H/R: Align head syllable of the head foot to the right; align prosodic word to the right. The H/R constraint will always be violated except in the case of monosyllables. In words of two or more syllables either a foot or an unparsed syllable is at the righ t of the head syllable. 111

PAGE 112

Thus rtic lus cannot be preferred over artclus (< artculus ) even though the stress falls on a heavy syllable. Below these alignment constraints is the PARSEconstraint, whose low ranking will allow syllables to be unparsed without disqualifying a candidate. With regard to the monosyllables in Table 3-6 it should be noted that aus and flaus can be construed as bimoraic only if fi nal /s/ is treated as extrametri cal and does not project a mora. These monosyllabic outcomes fulfill RHTYPE (T) (The rhythm type is trochaic; feet are trochaic) when they are parsed as H. An alternative approach is to cons ider mora sharing. Rather than considering /s/ extrametrical, the final consonant shares a mora with the preceding more sonorous element (Broselow, Chen, and Huffman 1997). When syllables with a heavy nucleus and a coda are viewed this way all of the undominated constraints proposed by Jacobs (2003b) are satisfied: Lx P, FTBIN, RHTYPE (T). Therefore, these constr aints are not included in Table 3-6. The only candidates offered in Table 3-6 are the faithful (no change) candidate and the syncope candidate. Outcomes marked by ?! are discussed below. The candidates stablum, capiclum articlus, vernaclus present problems with regard to fulfillment of binarity under the Classical Latin system once vowel quantity ceases to be distinctive. If all candidates are parsed for minimum codas and maximum onsets, then sta .blum ca.pi .clum ar.ti .clus, ver.na .clus are the outpts. However, the underlined syllable, th e presumed accented syllable does not meet the minimum bimoraic requirement. It has therefor e been parsed with th e following syllable in violation of NONFIN. The creation of a foot at the right edge of the word may be the best interpretation of this transitional diachronic change, also reflected in the high ranking of the modified constraint H/R. In the nominal system, it is the accusative that passes into Romance and the final -m is lost at an early stage. Therefore, the proposed parsings for (sta .blum) 112

PAGE 113

ca(pi .clum ,), (ar)(ti .clus), (ver)(na .clus) when taking into considera tion the loss of quantity of the final syllable correspond to a trochee with two light syllables, c onstructed at right word edge. Loss of distinctive vowel length is a necessary step in the shift from the moraic trochee to the syllabic trochee. This transition also results in the demotion of NONFINALITY and its eventual elimination as an active constraint. Table 3-6. Evaluation of fa ithful and syncope candidates H/R NONFINH/L W/R W/L PARSEL X H / vus/ grandfather (aus) (a .vus) H X H /fl vus/ golden ( flaus ) (fla )vus L LX H X /c l da/ warm water (cal )da (ca .li)da H LX H X /ang lus/ angle (an )glus (an )gu.lus ** L LX L X /st b lum/ dwelling ?! (sta .blum) (sta .bu)lum LL LX LL X /c p t lum/ small head ?! ca(pi .clum) ca.(pi .tu)lum ** HL LX HL X /art c lus/ joint ?! (ar)(ti .clus) (ar)(ti .cu)lus HH LX HL X / vern c lus/ rabble ?! (ver)(na .clus) (ver)(na )cu.lus ** Jacobs (2004, 83-84), in examining the possibility of creating a single set of OT parameters that correctly predicts results for both pre-tonic and post-tonic syncope, finds that it is simply impossible to handle all th e syncope cases with one and th e same constraint hierarchy. A foot-based account is necessary for syncope af ter main stress, but fails for syncope in HL 113

PAGE 114

sequences b efore main stress and cannot account for syncope in soliculum cases and later leporem cases without additional assumptions or mach inery. The inherent opacity of syncope in Latin prompts Jacobs (2004, 77-88) to consider constraints based on faithfulness rather than well formedness. Reliance on Output-Output constraints, however, is not possible because this presumes the existence of an output form which syncope attempts to produce. Related lexical items such as disc pl na (cf. discipulina attested in Plautus, Lewis and Short s.v. disc pl na) disc p lus illustrate the nature of the problem. Sympathy theory builds on faithfulness to I/O constraints; the opaque candidate wins by more closely resembling the sy mpathy candidate than the candidate that best satisfies constraints (except for the sympat hy constraints). Even after the introduction of MAX-V-I/O (No vowel deletion) the desire d winning candidate does not emerge except through non-violation of PARSEwhich is not a faithfulness constraint. Evaluation at different levels would permit stress assignment firs t and then post-lexical application of the rules for syncope. Correct selection of the syncope candidate has been achieved in Table 3-6 by modification of two cons traints: H/R and NONFIN. Although the Appendix Probi examples all represent post-tonic syncope, Table 3-7 shows that the syncope ca ndidate still emerges as winner in cases of pre-tonic syncope. It has been a ssumed here that since vowel length is no longer a feature of Vulgar Latin that the vowel of the an tepenultimate syllable could be treated as short permitting formation of a foot (L L) at the right word edge. This also favors the later deletion of final /e/ because it is in the weak position of a foot. Table 3-7. Post-tonic syncope /v r t te(m)/ truth H/R H/L W/R W/L PARSE(ve.ri)(ta .te) ver(ta .te) 114

PAGE 115

Jacobss suggestion of syncope as a post-lexic al process does introduce the possibility of a faithfulness type constraint for future considera tion, correspondence of word accent in input and output (cf. HEADMAX in Lle 2003, 256-257). (3.2) ACCENT/IO: A segment that is assigned an accent in the input has a correspondent in the output. The ACCENT/IO constraint proposed above is able to de al with cases not covered in Tables 3-6 and 3-7 for which the constraints utilized there will be invalid: (1) preservation of antepenultimate stress, (2) penul timate stress when the final sylla ble is heavy (after apocope), and (3) ultimate stress when both penultimate and ultimate syllables are heavy. These cases emerge in all varieties of Ibero-Romance. Ta bles 3-6 and 3-7, then, pr esent a way of dealing with post-tonic and pre-tonic syncope at a particular ch ronological stage. What is learned from Vulgar Latin syncope is that it is not motivated by exhaustive parsing. Candidates (cal )da and (ca .li ) da are identical in terms of parsing; they each present one parsed trochaic foot and one unparsed syllable. The syncope candidate wi ns only because the gradient constraint H/R penalizes removal from right word edge. Furthermore, it cannot be stated unequivoc ally that the purpose of syncope is to create a heavy foot for main word accent. The syllabicity of stop + liquid clusters remains a subjec t of controversy and is discu ssed in detail in the following section. For purposes of analysis in Table 3-6 it has been assumed that codas, following syncope, are minimized and onsets are maximized. This corresponds to basic principles of well formedness with regard to sylla ble architecture. Vulgar Latin syncope has the effect of deleting vulnerable vowels. While this is a violation of a faithfulness constraint (MAX), the loss of the vowels implie s that they are no longer candidates for vowel reduction (The three modern languages that are the focus of this study differ substantively with regard to vowel reduction). Furthermor e, with regard to rhythmic 115

PAGE 116

criter ia the forms that result from vowel loss are superior. In the case of Classical Latin threesyllable words, if they are rhythmically 2/4 |, that is, a triplet in contrast to the stable disyllable, then preference for the unmarked rhythmic configuration is a motivating force for linguistic change (Miller 2007) The syncope candidate cal.da is attractive not only because the closed or heavy syllable coincides with the primary accent of the word, but also because it embodies the preferred duple rhythm. The former four-syllable words, now at three syllables, still conform to duple time. They are rhythmically 2/4 with the first syllable functioning as an upbeat or anacrusis to the downbeat of the following trochee. Furtherm ore, syncope supports Crosswhites (2001, 173) observation that stress-timed languages exhibit tempo acceleratio n via compression of unstressed syllables. Phonetic studies of vowel length in Spanish (Monroy Casas 1980, 44-47), Catalan (Recasens i Vives 1986, 149150), and Portuguese (Delgado Martins 1988, 128-132) support this viewpoint. More recent studies on the acoustic nature of linguistic rhythm (Ramus 2002 and Ramus, Nespor, and Mehler 1999, Grabe and Low 2002) view vowel length as pa rt of a complex issue involving rhythmic typology, vocalic and intervocalic in tervals as well as consonantal in tervals. It may not be so much vowel length per se but the contrast of vocalic and consonant al intervals that differentiate languages and accentual types. The forms tht resu lt from syncope in Table 3-6 show a shift in the proportion of acoustic space occupied by vo calic duration and consonantal duration. The reduction of vocalic interval as a stress accent emer ges is consistent with the findings of Ramus, Nespor and Mehler (1999, 272) whose study of eight languages s hows English with the lowest vocalic interval (sum of vocalic interval divide d by total duration of the sentence) at 40.1% and Japanese at 53.1%. Spanish a nd Catalan, also surveyed in this study, show 43.8% and 45.6% 116

PAGE 117

respec tively5 (The other languages studied are Dutch, Po lish, Italian, and French). Grabe and Lows study (2002) compares relative vowel dura tion and intervals between vowels in Spanish and Catalan. They find that Catalan, traditionally labeled a mixed language because it displays some degree of vowel reduction, differs from Spanish with regard to vowel duration but not on the intervocalic axis (time between vowels). In other words, vowel re duction does not affect vowel duration sufficiently to give the acousti c impression of a stress timed language. Spanish also shows differences in duration of accented and unaccented vowels (Monroy Casas 1980, 4447), although it is clearly not a stress timed language. Creation of Heavy Syllables Stress-timed accentual systems (Crosswhite 2001, 173), are also characterized by a tendency to build heavy syllables. Some of th e examples of syncope in the Appendix Probi show the creation of a heavy syllable which is now the syllable with the primary accent: aus, flaus, calda, and possibly stablum, capiclum, articlus, vernaclus The uncertain syllabification of this second group illustrates another of the characteristics of stress-timed languages: complex syllable structure with relative ly uncertain syllable boundaries. The treatment of the first consonant in cases such as stablum affects the assignment of stress even if quantity sensitivity is not an overt constraint, as seen in Table 3-4. The ability to build a trochee depends on the presence of either two light syllables or a heavy syllable. If vowel lengt h is no longer distinctive in Vulgar Latin, then a heavy syllable can onl y be formed by a falling diphthong or a syllable 5 The other languages studied are Dutch, Polish, French, and Italian. The various statis tical measures duration of vocalic intervals, duration of consonantal intervals, prop ortion of vocalic intervals, standard deviation of the duration of vocalic intervals, standard deviation of the duration of consonantal intervals resulted in three groupings. The first contained the stress-timed languages, Polish, English, and Dutch, although on some measures Polish seemed to set itself apart from the other languages in the group. The second group contained the syllabletimed languages, French, Spanish, Italia n, and Catalan; and the third, the only moraic-timed language, Japanese. 117

PAGE 118

with a coda. Syllabification of stablum as stab.lum would allow construction of an antepenultim ate foot without recourse to NONFINALITY. Geisler (1992, 48-49) has proposed a quantit ative method of predicting changes to syllable structure resulting from sy ncope in Latin. He relies on th e notion of a sonority scale for consonants: 4=stops, 3=fricatives, 2=nasals, 1= liquids. The possible combinations are then arranged as in Table 3-8 (adapted from Geisle r 1992, 48) to compute the degree of difference. Once the degree of difference has been computed it is possible to predict the most likely outcome of medial clusters that result from syncope. It is well known that in Romance such secondary clusters often differ in outcome from primary clusters that contai n the same segments. Table 3-8. Degree of difference in sonor ity between medial consonants (Geisler) 1 2 3 4 1 liquid-liquid 0 liqui d-nasal 1 liquid-fric ative 2 liquid-stop 3 2 nasal-liquid -1 nasal-nasal 0 n asal-fricative 1 nasal-stop 2 3 fricative-liquid -2 fricative-nasal 1 fricative-fricative 0 fricative-stop 1 4 stop-liquid -3 stop-nasal -2 stop-fricative -1 stop-stop 0 Geisler (1992, 49) considers other possiblities: ambisyllabici ty, a segment belonging to two-syllables, and tautosyllabicity, segments o ccurring within the same sy llable (Table 3-9). Table 3-9. Resolution of medial c onsonant clusters produced by syncope Difference Consonants in contact Change in syllable structure 3 liquid-stop no change 2 liquid-fricative nasal-stop no change 1 liquid-nasal nasal-fricative fricative-stop no change 0 liquid-liquid nasal-nasal fricative-fricative stop-stop no change ambisyllabification -1 nasal-liquid fricative-nasal stop-fricative ambisyllabification -2 fricative-liquid stop-nasal ambisyllabification -3 stop-liquid resyllabification 118

PAGE 119

When the difference is com puted as zero or less the most likely outcome is that the consonant sequence will now be ambisyllabic with one segm ent as both coda and onset. Resolution of unstable sequences in this manner presupposes total assimilation and a resulting geminate. The case of stop+liquid calls for resy llabification in Geislers schema and this seems to be supported by a few cases of accent shift, such as the often cited case of integru(m) whole with two sets of outputs corresponding to different re gisters and different chronologies as in Spanish patrimonial entero vs. learned ntegro Discrepancies with regard to placement of accent can best be explained by resyllabif ication of the medial -grsequence. Ambisyllabification and Mora Sharing Ambisyllabification often arises as the result of the creation of a ge minate cluster through progressive assimilation. However, of th e Romance languages today only Italian has a functional system of phonologically based geminates in addition to the well known raddoppiamento sintattico Saltarelli (2003, 60) describes th e Italian long consonant as a hetero-syllabic geminate straddling the coda an d the onset of contiguous syllables. He notes (2003, 61) that in recent acoustic studies it ha s been found there is no appreciable difference between geminates in medial positions and those th at arise between words; furthermore, there is no evidence of double articulatory movements. In moraic terms it is cl ear that the Italian geminate occupies two timing slots or two mora s. Catalan has many cases of geminates that arise from specific phonotactic situations but Hual de (1992, 384) suggests that only laterals and nasals (excluding / /) are underlying geminates as in fallera obsession [f l.lr ], espatlla back [ sp ], and cotna crackling [kn.n ]. Across morpheme boundaries (word internally) examples of total assimilation are abundant and encompass not only laterals and nasals but rhotics and stops. 119

PAGE 120

Bullock (2001, 188-190) proposes a c onstraint for Vulgar Latin, LCL, which forces a vowel followed by a stop + liquid cluster to be s canned as heavy. She finds evidence for this constraint in the historical deve lopment of geminates in Italian in the case of a labial consonant followed by liquids or glides such as labbro lips < l bru(m) or doppio double < duplu(m). These clusters are primary rather than seconda ry. Gemination does occur with secondary stop + liquid clusters when the stop is velar: occhio eye < oc lu(m) However, germination in Italian occurs in other environments as well reflect ing the interplay of long vowel/long consonant suggested by Page (2006; cf. Anderson 1984). One other piece of evidence for some type of compensatory lengthening is the case of medial stop + liquid clusters in Catalan where the stops /b/ and /g/ geminate in a stop + /l/ cluster after a stressed vowel as in poble people pronounced [ 'p b.bl or regla rule pronounced ['reg.gl ]. The geminate stop is realized as voiced in the Empord area (northeastern Catalonia) and as voiceless in Barcelona; in Western Cataloni a the stop is not geminate and is realized as approximant [ ] or [ ] (Hualde 1992, 411). There are a few lexical exceptions to the gemination which Hualde views as a process of fortition; these te nd to be learned words or late borrowings from Latin. The articulatory evidence cited above with regard to Italian and Catalan suggests that Bullocks interpretation of moraic compensation with regard to raddoppiamento sintattico (after an open final stressed syllable ) is correct. She illustrates (2001, 190-191) the case of citt vecchia [ it.tav.'v k.kja ] schematically as seen in Figure 3-2. Underlying assumptions are that primary stress in Italian is lexical rather than grammatical, that is, it ca nnot be predicted or accounted for by rules or constraints; vowels are not lengthened in stressed syllables; feet are quantity sensitive moraic trochees. Therefore, it is necessary for the final, stressed syllable of 120

PAGE 121

citt to project two m oras to satisfy the minimum foot requirement. Since vowel lengthening is not allowed the second mora is assigned to the following consonant. ( ) cit ta A | ( ) cit ta] vecchia B Figure 3-2. Raddoppiamento sintattico as moraic compensation. Tautosyllab ic versus heterosyllab ic medial clusters in Latin The syllabification of stop + liquid clusters has long been an object of interest in both Latin and Romance studies. Allen (1973, 137) relies on evidence from Latin verse and placement of accent in asserting that it is quite clear that a syllable containing a short vowel followed by such a sequence was regularly light in quantity. These sequences mu st therefore have functioned as complex releases of the following syllable: thus ten .brae, p .tris, p.plus The rationale for the tautosyllabic realization of the cluster as onset is often attributed to the admissibility of stop + liquid clusters in word initial position. However, Devine and Stephens (1980, 144) note that the patterning of medial clusters does not necessari ly depend on whether or not they can occur word initially, rather in many languages word-initial clusters are often heterosyllabic when they occur inside words. They further note that in the case of Latin the syllabi fication of stop + liquid often interacts with morphosynt actic boundaries. For a tautosy llabic interpretation, one must weigh the consequences of violation of two different constraints: NOCODA and *COMPLEX (No complex onsets). Neither Latin nor the Ibero-Romance languages have a *COMPLEX constraint with regard to stop + liquid clusters except fo r /tl/ and /dl/ (in native vocabular y). This is not surprising as 121

PAGE 122

they are infrequent com binations in the languag es of the world. A notable exception is Nahuatl where /tl/ exists as a single phoneme and has infl uenced pronunciation of /tl/ in Mexican Spanish where words like atlas are normally syllabified as at.las The /tl/ cluster is maintained in many toponyms of Nahuatl origin as well as comm on words of indigenous origin such as tlapalera hardware store (Hualde 1999, 171-172). Steriade (1982, 95-99) notes that onset sequences that are allowed or disallowed conf orm to language-specific criteria. This constitutes a major difference between Latin and Greek that can be resolved only when the minimum sonority distance (MSD) requirement of cont iguous elements for these two la nguages is allowed to differ: The MSD requirement can now be stated as 6 in tervals for Latin and as 4 intervals for Greek: this will allow all the stop-liquid clusters of Latin to be onsets except tl, dl (Steriade 1982, 98). In determining the viability of particular sequences of consonants it is useful to examine the frequency and distribution of such clusters. Devine and Stephens (1 977) have tabulated the relative frequency of Latin consonants as first or second member of a medial consonant cluster as well as the relative frequency of particular clusters. The quan titative analysis is based on a corpus of 49,488 words of continuous prose from Ciceros De Amicitia and De Senectute and Sallusts Catilinae Coniuratio and Bellum Iugurthinum. Their findings (1977, 181-182) show that medial clusters of the type stop + liquid are relatively infreque nt. As medial clusters in noncompounds, that is without an intervening mor pheme boundary, the stop + liquid clusters have the frequency indicated in Table 3-10 (% = percen tage of non-compound clusters in the corpus). The clusters /tl/, /dl/, /gl/ are non-occurring. With the exception of /-bl-/ overall the clusters with /l/ are less frequent. The relative infreque nce of /g/ as a phoneme (2.9316% of all medial consonants, the lowest percentage of any of the stops; Devine and Stephens 1977, 178) may also explain the absence of /-gl-/. Of the occurring stop + /l/ clus ters /-kl-/ shows the lowest 122

PAGE 123

frequency of occurrence. The inventory of occu rring stop + liquid consonan ts dem onstrates that as word internal clusters, not resulting from derivation, they are quite infrequent with /-tr-/ having the highest rate of occurrenc e and /-kl-/ the lowest It is not surprising, therefore, that stop + liquid clusters that arise from syncope might not be trea ted in like manner. In other words, there is not a pervasive pattern that would make tautosyllablic /-pr-/ a more desirable result than heterosyllabic /-p.r-/. Table 3-10. Frequency of occurrence of Latin medial clusters: stop + liquid pr 0.2882 % br 0.3871 % tr 1.4730% dr 0.0640% cr 0.3074 % gr 0.7685% pl 0.0830% bl 0.8838% tl --dl --cl 0.0448% gl --Stop + liquid clusters in Romance The relative chronology of syncope and lenition in Romance, as well as the question of spoken versus literary language, makes it difficu lt to propose all-encompassing rules for the treatment of original stop + liquid clusters, as well as those produced by syncope (secondary clusters). Table 3-11 displays the Ibero-Roman ce reflexes for stop + /l/ and stop + /r/ clusters, both original and secondary. There are some gaps, indicated by ---, where clusters are nonoccurring or reliable forms are unattested. A bla nk cell indicates that clus ters are non-occurring. Even with clusters of this type that do not arise from syncope, different accentual patterns are seen, as in the case of integrum, acc. sg. whole which enters Ibero-Romance with both penultimate stress in popular forms (Cast. entero Cat. enter Port. enteiro ) and antepenultimate stress in learned counterparts (Cast. ntegro, Cat. ntegre Port. ntegro ) despite the fact that in the antepenultimate pattern word accent falls on a heavy syllable but does not when stress is penultimate (with the exception of Port. enteiro ). The shift in accent is consistent with what is 123

PAGE 124

em erging as the preferred accentual pattern of the Vulgar Latin nominal system, penultimate accent. Table 3-11. Ibero-Romance reflexes of Latin stop + liquid clusters Primary Clusters Secondary Clusters -prVL capra -trVL petra -krVL lucru -prVL paupre -trVL utre -krVL socru Cast. cabra Cat. cabra Port. cabra Cast. piedra Cat. pedra Port. pedra Cast. logro Cat. llogre Port. logro Cast. pobre Cat. pobre Port. pobre Cast. odre Cat. odre Port. odre Cast. suegro Cat. sogre Port. sogro -plVL dupla -tl---klVL cochleare -plVL poplu -tlVL veclub -klVL apicla Cast. dobla Cat. dobla Port. dobra Cast. cuchar Cat. culler Port. colher Cast. pueblo Cat. poble Port. povoa Cast. viejo Cat. vell Port. velho Cast. abeja Cat. abella Port. abelha -brVL fibra -drVL quadru -grVL integrum -brVL labratorec -drVL (h)edera -gr--Cast. hebra Cat. fibra Port. fibra Cast. cuadro Cat. quadre Port. quadro Cast. entero Cat. enter Port. enteiro Cast. labrador Cat. llaurar Port. lavrador Cast. (h)iedra Cat. (h)eura Port. (h)edra -blVL publicu4 -dl---glLLat. siglad -blVL tabla -dl---glVL coaglu Cast. pblico Cat. pblic Port. pblico Cast. sigla Cast. tabla Cast. cuajo Cat. sigla Cat. taula Cat. quall Port. sigla Port. tabla Port. coalho aPortuguese povo shows deletion of intervocalic /l/, probably occurring before syncope. bThe usual reflex of tlis kl-; this is well attested in the Appendix Probi. cIn the case of labratore syncope is pre-tonic rather that post-tonic. dFor both publicu and sigla the reflexes in Ibero-Roman ce are learned or semi-learned. In the case of the clusters with /r/, ther e is little difference in treatment between the primary clusters and the secondary clusters. Both Castilian and Catalan show further lenition of /b/, /d/, /g/ which are usually rendered intervoca lically, including in inte rvocalic clusters, as [ ], [ ], [ ]. Exceptional treatment of the voiced stop is seen in Catalan in the case of the secondary clusters /-br-/ and /-dr -/. The velar + liquid clusters for all three languages deviate from the pattern predicted in Table 3-9, that is, resyllabification. The Ibero-Romance outcomes suggest that the ve lar stop, in weak posi tion, palatalizes as yod. Thus from -kl, as in VL oclu eye, result Cast. ojo (['oo] in Old Spanish), Cat. ull and 124

PAGE 125

Port. olho ; from -gl-, as in VL coaglu come Cast. cuajo (['kwao] in Old Spanish), Cat. quall and Port. Coalho. The palatalization here can be compared to the result of the kt cluster where lenition of /k/ to /j/ produces total assimilation resulting in [] in Cast. noche but yod remains as syllable coda in Old Cat. nuit (Mod. Cat nit) and Port. noite There is, however, another output of secondary cluster -gl for Catalan: VL tegula roof > Cat. teula (cf. regula rail > Cat. rella ). There are two questions that arise immediately from the data presented so far: the syllabification of the Vulgar Latin forms afte r syncope and the relativ e chronology of lenition and syncope. With regard to forms with falling diphthongs, li ke the examples presented in the discussion aboveCat. taula, teula Port. enteiro as well as peuma scaffolding from the Appendix Probi, Steriade (1988, 391) suggests that rath er than interpreting these as evidence of syllabification making the obstruent + liquid sequen ce ambisyllabic, that is -C.L-, it should be assumed that glide formation here results from a process of voicing (if th e stop is voiceless) and spirantization, that is, integrum *inte ru intejru ( j = palatal glide). Steriade further suggests that the voicing and sp irantization illustrate d above applied not onl y to the postvocalic member of an obstruent-liquid clus ter but also to intervocalic obs truents. The choice of yod or wau may be conditioned by the phonotactic environment. Weakening of Consonants in Coda Position However, the possibility of coetaneous developments in which stop + liquid clusters are both heterosyllabic and tautosyllabi c should not be ruled out. That is precisely th e situation in modern Catalan as described above where poble people can be realized as both [ pb.bl and ['po. le]. In the case of taula and teula, weakening of syllable final /b / and /g/ to /w/ fits well with other historical developm ents of Catalan. While there are some inherent dangers in 125

PAGE 126

com paring word internal syllable final and word fi nal positions, the treatment of /g/ in word final position seems to suggest that /w/ is not so much the result of leni tion of intervocalic /b/ and /g/ as it is a positional variant; cf. Lat. jugu(m) yoke > Cat. yugo Cat. jou Port. jugo. Similarly, Lat. activu(m) active results in Catalan actiu next to Cast./Port. activo Steriade (1988, 391) also notes a similar development in Romanian where staul results from stabulum dwelling. The Appendix Probi shows that syncope, alr eady well documented in Archaic Latin, has increased in frequency. However, syncope is only one of severa l reduction phenomena. This is in keeping with observations on reduction as a pr ocess in diachronic cha nge: These changes occur in real time as language is used. Words and phrases that are used more often have more opportunity to undergo these changes, just as other types of motor skills that are used more often become more compressed and efficient (Byb ee 2001, 58). Further evidence of reduction phenomena can be seen in cases of orthographic in certainty that indicate th at lenition in medial stops corresponds to the usual pronunciation. (L ine number on the left corresponds to Appendix A). Perception of medial consona nts as voiced is grounded in arti culatory and acoustic studies. Blevins (2004, 147) notes that lenition is marked acoustically by decreased duration: Decreased closure duration of intervocalic stops gives rise to a voiced percept in the absence of vocal-fold fibration, while stops w ith incomplete closure can be perceived as fricatives Speakers do not produce [k] as [g ] intervocalically as a result of gestural reduction. In prosodically weak positions, th e closure of [k] may be shortened, and linguopalatal contact may be reduced, resulti ng in things that sound like [g] and [ ] to the human ear. When these sounds are interpre ted and reproduced as [g]s and [ ]s, a leniting sound change has occurred. While orthographic norms of Latin do not reflect the reality of lenition there is evidence that it is, in fact, a much older phenomenon. Cravens (1996 [1991], 54-59) argues th at only an assumption of early lenition can fully explain some of the p/ b, t/d, k/g variants found on inscriptions, ostraca, and papyri throughout the Empire. In the Appe ndix Probi, the examples listed below (Line 126

PAGE 127

num ber of the left corresponds to Appendix A) show weakening of both voiced and voiceless stops, especially loss of intervoc alic voiced stops (line 12), voic ing of intervocalic voiceless stops (79, hypercorrection), lenition of voiced st ops (91, 93) and weakening of coda consonants to the point that spelling is no longer transparent (54, 85, 154, 185). References to combinations of stop + /s/ in coda position (line 60, 181, 184) reflect the assert ion of orthographic norms over long standing phonological realities. The varying treatments of these medial stops demonstrate that there are multiple processes underway that have staggered chronologies with outcomes that are far from uniform. 12 calcostegis non calcosteis. 54 frigida non fricda. 60 celebs non celeps. 79 digitus non dicitus. 85 pegma non peuma. 91 plebes non plevis. 93 tabes non tavis. 154 auctor non autor. 181 plebs non pleps. 184 c[a]elebs non celeps. 185 poples non poplex. [popleks] Stop + liquid clusters in Romance Further diachronic evidence for syllabification of stop + liquid cluste rs in Vulgar Latin comes from the reflexes of preceding toni c vowels. The diphthongization/nondiphthongization contrast of mid vowels in French is attributed by Steriade (1 988, 379) and Bullock (2001, 177) to the nature of the medial clusters that follo w the tonic vowel. Presumably the vowel only diphthongizes when it is in an open syllable (cf. Castilian which also diphthongizes when the tonic vowel is in a closed sylla ble). The data in Table 3-12 s how diphthongization in French in the two words with open syllables (VL pe.tra and fe.bre) and failure to diphthongize in the case of festa where /s/ is a coda, a supposition further strengthened by its loss. 127

PAGE 128

Table 3-12. Diphthongization be fore stop + liquid cluste rs in Rom ance languages Vulgar Latin Italian French Portuguese Catalan Castilian pe.tra stone pietra pierre pedra pedra piedra fes.ta feast festa fte festa festa fiesta fe.bre fever febbre fivre febre febre fiebre While the presence of a rising diphthong does not appreciably alter syllable structure in moraic terms it does attest to the syllabificat ion of the stop + liquid cluster for specific languages. The syllabification of these clusters was an open question even for Classical Latin where the general precept is that muta cum liquida does not make position either for purposes of accent assignment or for poetry. With the excepti on of velar stop + liquid clusters, the Romance languages support interpretation of the clusters as tautosyllabic, or, a preference for complex onset over stop as coda consonant. When a nucleus disappears, the segments to either side of that nucleus no longer have a hier archical structure to which to attach themselves. The choices (elision, heterosyllabicity, tautosyllabicity) rais e issues of positional faithfulness (Smith 2005) and sonority sequencing (Clements 1990). It is unlikely that a sy llable final stop would emerge as optimal for Late Latin/Early Romance. In a study of the resolution of stop+liquid clusters across word boundaries in modern varieties of Catalan and Occitan Pons Moll (2005, 3-5) finds that the undesirable sonority increase is resolved by total assimilation or regressive manner assimilation, whereas decreasing sonority transiti ons remain faithful to input. There are three exceptional cases: preservation of sibilant-nas al heterosyllabic clusters, nasal-rhotic heterosyllabic clusters, and liquidi -glide heterosyllabic clusters. If the rather different behavior of the velar stop + /l/ clusters is indicative of alternate syllabification, th e end result (a single palatal consonant as onset of following syllable) demonstrates the emergence of the unmarked in the resulting disyllables w ith initial accent, or L .L: Lat. grac la(m) > Cast. graja Cat. gralla, Port. gralha. The resulting forms are all legitimate bimoraic trochees. 128

PAGE 129

Apocope Apocope is further evidence of the weakening processes present in stress-tim ed languages. Cases of apocope in the nominal system are best seen in nouns of the third declension. Final /e/ is regularly retained in Eastern Romance and is regularly eliminated in Catalan, Occitan and French. In Spanish, the vowel is lost when the preceding consonant is coronal (Jensen 1999, 132). Examples for early Spanish show apocope occurring after /t/, /n/, /l/, /r/, /s/, / /. Examples for Catalan and Portuguese are provided as well. For Castilian, all outcomes have well-formed bimoraic trochees at right word edge consisti ng of a single heavy syllable, although this pattern is less harmonic that that seen in Portuguese idade or asse. Castilian Catalan Portuguese VL etate age > edad edat idade6 VL pane bread > pan pa po VL sale salt > sal sal sal VL mare sea = > mar mar mar VL luce light > luz lluu7 luz VL assem coin > as as asse The same is true of Catalan assuming that pa is underlyingly bimoraic as well as Old Catalan lluu The Portuguese reflex of Vn]wd is a bimoraic nasal diphthong. There are also a few cases in the Appendix Probi that suggest apocope as an ongoing process al though it is not clear to what degree the dispreferred forms might be better explained through analogy with historical Latin apocope (Sihler 1995, 70). 32 figulus non figel. 33 masculus non mascel. 36 barbarus non barbar. 6 Portuguese is the most restrictive of the three languages with regard to codas; note conservation of final e here and below in asse. 7 Coromines (1995, s.v. llum) discusses the possible coalescence of Old Catalan lluu< Lat. lucem and llum
PAGE 130

Hiatus and Onset Glide s Again, the Appendix Probi provides valuable information for language change that affects well formedness. The resistance to vowels in hiatus reflects the universal preference for a syllable onset. When lack of onset is resolved by glide formation, th ere is often an added benefit, movement of the primary accent to th e penultimate syllable. The unstable orthography with regard to the vowels /i/ and /e/ is indicative of vowel raisi ng, which can be a precursor to glide formation. In many instances, such as line 34, the vowel substitution appears to be a case of hypercorrection. Corresponding outputs in Castilian, Catalan, and Portuguese are given when available to support the case for vowel raising and/or glide formation. 34 lanius non laneo. 44 bravium non braveum.8 52 doleus non dolium.9 55 vinea non vinia. Cast. via, Cat. vinya, Port. vinha 61 ostium non osteum Cast. (ant.) uzo 63 cavea non cavia. Cast. gavia, Cat. gbia, Port. gaiva 65 brattea non brattia. 66 cochlea non coclia. cf. c cl ar (cochl-), ris > Cast. (ant.) cuchar, Cat. culler, Port. colher 68 palearium non paliarium.10 cf. Cast. pajar, Cat. pa llar/paller, Port. palhar 72 lancea non lancia. Cast. lanza, Cat. llana, Port. lana 80 solea non solia.11 Port. solha; cf. Cat. sola Cast. suela 81 calceus non calcius. cf. Cast. calza, Cat. cala Port. cala 113 alium non aleum. Cast. ajo Cat. all Port. alho 114 lilium non lileum. 117 tinea non . Cast. tia Cat. tinya Port. tinha 132 balteus non baltius.12 Cat. bal, Port. balso 8 Lewis and Short (s.v. br b um) gives both vocalisms, although corresponds to Greek 9 Lewis and Short (s.v. d l um) notes plural written DOLEA as inscriptional variant (Calend. ap. Orell. Inscr. II. p. 381), cf. Gr. 10 Although the DRAE (s.v. pajar) gives palear um as the etymon of Cast. pajar a more likely source is p l ris, e adj., given the fact that the three outp uts in Ibero-Romance all show truncation, which is more likely to occur if the input is third declension rather than second declension. Meyer-Lbke (1935,s.v. paleare) gives paleare as the source of Cast. pajar Galician pallar and Catalan paller The DCVB (s.v. pallar) gives pallar as a variant in Valencian. 11 Cat. sola and Cast. suela presuppose a variant form sola (Meyer-Lbke, s.v. solea). 12 Cast. balso appears to be a borrowing from Catalan (DRAE, s.v. balso). 130

PAGE 131

141 faseolus non fassiolus. Cat. fesol, Port. feijo 157 linteum non lintium. Cas. lienzo, Cat. llen, Port. leno 160 noxius non noxeus. 204 musium vel musivum non museum. 213 adon non adonius.13 As in the case of the stop + liquid cluste rs, these obstruent + glide clusters present questions about syllabification. Steriade (1988: 379-380) suggests that the operative principle in Romance is the maximization of syllable onsets. The glide formation resulting from a constraint of the type *V[+high].V (A high vowel nucleus cannot be followed by a vowel nucleus) causes the vowel in the weak position (unaccented) to beco me a glide. The Appendix Probi examples all show /i/+V combinations, even in the case of orthographic e The syllabification of a word like f l u(m) son as *fi-lyo not only fulfills the well formedness constraint, ONSET, but also creates a disyllabic trochee. As the di stinction of long and short vowels disappears, moraic count for a heavy syllable can only come from the presen ce of a coda glide (falling diphthong) or coda consonant. The first syllable in words like f l u(m) can no longer be considered to be heavy and the loss of -m the marker of the accusative, renders the final syllable light as well. The changes to f l u(m) in terms of prosodic struct ure can be summarized as (H )(LL) (L L), that is, two feet are reduced to one. Again, the preferred duple rhythm has emerged. The effects of glide formation in Ibero-Ro mance are far reaching. Not only is syllable count reduced but the various assimilations of consonants to a following or preceding palatal glide produce a rather different consonant inventory for Castilian (Lloyd 1987, Penny 2002), Catalan (Wheeler 1988), and Port uguese (Parkinson 1988) when compared to Latin. Table 3-13 shows the Ibero-Romance reflexes of stops + yod, s + yod, ss + yod, n + yod and l + j; for Late 13 This correction suggests that the shor tened form Adon (cf. Adonis, Adoneus < ;) is the preferred rendering of this name, misspelled in the Appendix Probi as Adonius though not necessarily pronounced w ith a high vowel. 131

PAGE 132

Latin it can be assumed that s = /z/ and ss = /s/. The secondary devel opments with regard to /ts/ and /dz/ occur at a later period (for Castil ian, mid 16th century according to Penny 2000, 98-99) The one gap in the table is for s + yod in Catalan. The lack of examples other than bas um kiss, of doubtful origin, is due to the wi de spread rhotacism that affected intervocalic /s/ in Latin. In the case of bas um the Catalan reflex shows metathesis of the s + yod sequence which resulted in raising of the preceding vowel to /e/: bes (cf. Cast. beso Port. beijo ). Table 3-13. Ibero-romance reflexes of medial consonant + yod sequences t + j k + j kk + j d + j g + j s + j ss + j n + j l + j Cast. > > > > > dz> ts> Cat. w, >s ts>s -->s, >z >s, >z ts>s Port. Although the outcomes of obstruent + yod, nasa l + yod, lateral + yod are either complex or palatal elements they are not geminates. Whethe r or not they were ambisyllabic in the earliest stages is unclear but in the case of /n/ + yod it may be possibl e to reconstruct the change from /nj/ to // on the basis of studies of the articulation of // in modern Catalan. If alveolopalatal consonants are treated as complex segments feat ure spreading of the f eature [-ant] can account for the change of /nj/ to // (Recasens, Fontdevila, Pallars 1995, 266) as seen in Figure 3-3. /j/ /n/ // Dorsal Coronal Coronal Dorsal Coronal [+hi] [-ant] [+ant] [+hi] [-ant] Figure 3-3. Feature spreading in consonants in contact with /j/. 132

PAGE 133

The authors of the study on Catalan refer to ab sorption of the glide as an extrem e case of palatalization (1995, 268). They further note that at this stag e the palatal nasal may acquire phonemic status and is no longer able to occur in free variation as [n j]. As an example they cite for Spanish the minimum pairs uranio uranium and hurao unsociable. In contrast, Majorcan Catalan realizes word final // as /j/ + a nasal stop when the following word begins with a consonant. Furthermore, the word final nasal now assimilates to the following consonant to derive its place of articulation; thus, any bo good year is realized as [ajm b] (Recasens, Fontdevila, Pallars 1995, 270). This illustrati on of segmental decomposition provides a possible explanation for the special stat us of consonant + yod sequences. Motivation for Vowel Loss The concept of Faithfulness in OT supposes that, under optimal conditions, generation of a candidate that best satisfies the constraints of a language will neither delete nor add segments in the output form that do not correspon d to the input form. Several pr ocesses that result in loss of segments, in particular syllabic nuclei, have been presented here. The question is how to best motivate the vocal loss with the theoretical und erpinnings of OT. Hartkemeyer (2000, 65) has suggested a general anti-vowel c onstraint, *V that would outrank MAX-IO-V. (3.3) *V: Avoid V segments in output forms. (Hartkemeyer 2000) (3.4) MAX-IO-V: Every input V segment should ha ve a corresponding V segment in the output. (Hartkemeyer 2000) The affect of these two ordered constraints can be readily seen in Table 3-14. When *V ranks above faithfulness (MAX-IO-V) both candidates A and B emerge as preferable to candidate C because the pre-tonic (candidate A) or post-t onic (candidate B) vowel has been deleted. However, if MAX-IO-V is the top ranking constraint the faithfulness candidate (C) will always win. 133

PAGE 134

Table 3-14. Vowel loss and faithfulness constraints /zino/ ass *V MAX-IO-V zno ** A. zin B. ** C. zino ***! Because *V can overproduce vowel deletion, its e ffects must be tempered by the addition of other constraints. Hartkemeyer (2000, 68) recognizes the difficulty with th is constraint and notes Of course, the constraint *V is always violated to some degree or another, since it is indeed rare cross-linguistically to fi nd completely vowel-less output forms. He also recognizes that vowel deletion almost never affects the prosodic head of the word; likewise, th e vowel of the initial syllable seems to be invulnerable to deletion (Apheresis is relativ ely infrequent in the transition from Latin to Ibero-Romance). Therefor e, he proposes an alignment constraint: L-ANCHOR-V. L-ANCHOR-V: Every leftmost V in the input stri ng has a V correspondent in the output string. In order to avoid overgeneraliza tion of apocope (not evident in the early stages of Western Romance) Harkemeyer adds a parallel constraint at the right word edge: (3.5) R-ANCHOR-V: Every rightmost V in the input string has a V correspondent in the output string. (Hartkemeyer 2006) Table 3-15 shows the mitigating effect of the adde d constraints; the candidate that has undergone apocope is eliminated. The same constraint s also account for pre-tonic vowel loss. Table 3-15. Post-tonic vowel loss in Early Western Romance /tbula/ R-ANC-V L-ANC-V *V MAX-IO-V tbla ** tbula ***! tbul *! ** Syncope in Ibero-Romance seems to routinely spare /a/. Hartkemeyer (2000, 79) proposes double Max-IO constraints to account for th is. In addition to the low ranking MAX-IO-V, there is now a new constraint MAX-IO-V[lo] which will be ranked above *V. 134

PAGE 135

(3.6) MAX-IO-V[lo]: If an input V segment bears a [lo] feature, then its output correspondent must also bear the feature. This added rule works well for Ibero-Romance where many surviving proparoxytones have posttonic /a/, for example pjaro bird and rbano radish. Hartkemeyer suggests that observance of the sonority contour may also be a factor in cases where sync ope does not occur but this does not seem to be the case in light of fraxinus ash (tree) resulting in fresno and the many cases of syncope of the type nomine name which becomes nombre via nomne (Such forms are attested in Old Spanish, particularly omne for hombre man.). The real shortcoming of Hartkemeyers constraints is that they do not treat vowel loss in terms of word prosody. The motivating factor in syncope seems to be observance of a maximu m number of syllables before or after the primary accent as a concomitant of stress accent. This is not really reflected in his constraints except in a peripheral way through the R-ANC-V, L-ANC-V constraints. The constraint *SCHWA is proposed by Wh eeler (2007, 8) to favor both syncope and apocope. It ranks above a gene ral faithfulness constraint, MAXV. (3.7) *SCHWA: A vowel must not be realized as [ ]. (Wheeler 2007) Although Wheeler does not couch the loss of schwa in terms of prosodic preference it is possible to motivate loss as avoida nce of the uneven trochee. Cr oss-linguistic studies of vowel reduction demonstrate that [ ] is often significantly shorter th an other vowels (Crosswhite 2004, 208). Therefore, if [ ] constitutes the peak of the ultimate syllable in a two-syllable word the presumption is that even in a language whic h does not have contrastive vowel length, the prosodic template is in fact ( H.L). This analysis is supported by a prominence constraint proposed by Crosswhite (2004, 220-221) for Bulg arian which requires that all monomoraic vowels be low in sonority: *MONOMORAIC/X. The value X is filled in by the vowel inventory of the language under study. For example, in a la nguage having only the five cardinal vowels 135

PAGE 136

ranking would be as follows: MONOMORAIC/a >> *MONOMORAIC/e, o >> MONOMORAIC/i, u. The more sonorous a vowel is, the stronger is the constraint to consider it as something other than monomoraic. To avoid violation of the c onstraint, the vowel could become long or undergo reduction. For German (Fry 2003, 213) it has been suggested that schwa is nonmoraic, although Crosswhite includes sc hwa as the lowest ranking member of the constraint *MONOMORAIC. In view of Ibero-Romance diachrony, it can be seen that elision is preferable to retention of the reduced vow el as long as coda constraints are not violated. With regard to the possibility of vowel lengthening Major (1992) su ggests that this has been the outcome in Brazilian Portuguese as it moves from syllabl e timing to stress timing. Tonic vowels are now distinguished by length as well as feature preservation. The nominal structure of Latin is such that words in excess of five syllables are quite uncommon. Primary word accent falls only on the penultimate or antepenultimate syllable. Therefore, it is possible to genera te an inventory of Cl assical Latin possible words of two to five syllables taking into account locus of accent and whether syllables are light or heavy. In Table 3-16 the syllable that carries the primary word ac cent is shaded and syllables that are vulnerable or likely to be deleted are in ital ic (not all possible forms are actua lly attested); the final syllable is indicated as X. Syncope targ ets unstressed, short vowels. In previous discussion it has been suggested that vulnerable vowels also occupy the second, unstressed position in a binary trochee; they also appear in trapped or unpa rsed syllables in word internal position. Vulnerable syllables, then, fall into two categories: the second syllable of a foot (L L) and an unparsed L between feet, or (F1)L(F2). In the latter case either F1 or F2 is an accented foot. In Table 3-16 the pertinent cases are: (L L) type deletion: 3c, 4a, 4e, 5a, 5b, 5e, 5i, 5k, 5l, 5m (F1)L(F2) type deletion: 3d, 4b, 4g, 4h, 5c, 5d, 5f, 5i, 5j, 5m, 5n, 5o, 5p 136

PAGE 137

Table 3-16. Latin words of 2 to 5 sy llables m arked for possible syncope 2 syllables 3 syllables 4 syllables 5 syllables a L X L H X L L HXL L L HX b H X H H X H L HXHL L HX c L L X L H HXL H L HX d H L X HH HXHH L HX e L L L XL L H HX f H L L XH L H HX g L H L XL HH HX h H H L XHHH HX i L L L L X j H L L L X k L H L L X l H H L L X m L L H L X n H L H L X o L H H L X p L H H H X 2 3 4 5 The constraints established in Table 3-6 are still viable. NONFIN allows a foot to be built on the right edge when it is the only possible foot in a word. Thus, L X can form a disyllabic trochee because there are no segments to the left of the left foot boundary. W/R, together with H/R, predicts that optimal candidates will build th e head foot as close to the right edge as possible. This favors syncope when it occurs in post-tonic position. Poss ible syncope candidates for words of 3, 4, and 5 syllables with antepenultim ate stress all fall into this category (3c-3d; 4e4h; 5i-5p). To a lesser degree H/L also rewards the removal of unparsed or unaccented syllables between the left word edge and the head foot (4a-5b; 5a-5f; 5i -5j; 5m-5n). Although there are a few words that do not appear to be learned words that pass into Ibero-Romance with antepenultimate stress such as Cast. pjaro birdand rbano radish it can still be asserted that in the nominal system the predominant accentu al pattern is penultimate stress with some exceptions for the third declension where apoc ope has produced ultimate accent. Syncope 137

PAGE 138

reinf orces this preferred patte rn of a final disyllabic troch ee by reducing the instances of antepenultimate accent. While no single process can be said to account for the changes that o ccur in the evolution of the quantity-based stress of Latin, it has b een shown that many different phenomena can be subsumed under the umbrella of development of an optimal rhythmic pattern, that is, duple time, for proto-Romance: syncope, vowel reduction, a pocope, and the metrical traditions of popular verse. Furthermore, these changes coincide in la rge part with the characteristics of stress accent systems noted above (Crosswhite 2001, 173), na mely vowel reduction, tempo acceleration via compression of unstressed syllables, complex syllable structure with relatively uncertain syllable boundaries, tendency of stress to attract segmental material in order to build up heavy syllables, and presence of stress-counting versification. The emerging accent favors the contrast of a twosyllable foot in words of two or more syllables with accent on the penultimate, that is, a left headed trochee. After the period of syncope and apocope of vulnerable syllable nuclei in Vulgar Latin, in which all three Ib ero-Romance languages particip ate, Catalan, Castilian, and Portuguese diverge significantly in terms of treat ment of unaccented vowels. Today they form a gradient: Portuguese shows extreme vowel reduction (elision), Catalan reduction in terms of inventory and sonority, and Castil ian no systematic reduction. 138

PAGE 139

CHAP TER 4 ACCENTUAL PATTERNS IN THE NOMINAL SYSTEM OF LATIN: OUTCOMES OF THE FIRST DECLENSION OT Constraints and the Latin Accent The traditional computation of the Latin accent, discussed in preceding chapters, predicts that in words of three or more syllables the primary accent falls on the penultimate syllable if the rhyme is bimoraic. If the penultimate syllable is not minimally bimoraic or heavy, then the primary accent is retracted to the antepenultimate syllable, whether light or heavy. Words of two-syllables have penultimate st ress by default regardless of th e quantity of the syllable that will be assigned primary accent. Monosyllabic word s in Latin are few and not all are subject to the constraint LX PR (Prince and Smolensky 2004, 124): Ev ery Lexical Word must correspond to a Prosodic Word. That is, function words in Latin, as in many languages, are not always accent bearing (cf. clitic pronouns in Ibero-Romance) It should be noted, however, that nearly all monosyllabic words in Latin are, in fact, bimoraic. This is true even of function words such as prepositions where phonotactic variations such as b, from suggest that Latin had a minimum moraic requirement for a word which applied to both prosodic words and to function words. Latin word prosody, as discussed in Chapters 2 and 3, can be well described using the constraints of optimality theory. A few basic principles are reviewed here. The first overriding constraint which defines the prosodic word is (4.1) LX PR: Every Lexical Word must correspond to a Prosodic Word. (Prince and Smolensky 2004) Optimality theory also defines the prosodic foot, which in the case of Latin is the binary trochee. (4.2) FTBIN: Feet are binary at some level of analysis ( ). (Jacobs 2000, 2003b) (4.3) RHTYPE (T): Rhythm is trochaic. (Jacobs 2000, 2003b) 139

PAGE 140

Note that Jacobs has selected a less specific s tatement of binari ty than, for example, Mester (1994) who defines the Latin foot as being mi nimally and maximally bimoraic. The prosodic foot is further defined as being left headed. RHTYPE (T) also disfavors placement of primary accent in the final syllable of a word b ecause it favors a two-syllable sequence ('L .L) in which the first syllable is accented and the second is not. However, the moraic trochee also admits ('H) as a viable foot. Unless extrametricality of the final foot is invoked, a moraic trochee could be built in the final foot of nouns ending in (V = any vowel) -Vs or Vm both frequent word final configurations for Latin nouns. Ranked positioning constraints are therefore introduced wh ich favor placement of the primary accent near the right word edge but with at least one intervening syllable. (4.4) NONFINALITY: A foot may not be final. (Jacobs 2000, 2004) NONFINALITY ranks above the alignment constraints (Jacobs 2003b) that are used to evaluate the placement of the head foot within the word. They are given in rank order. (4.14) W/R: Align prosodic word to the right; a lign feet to the right. (Jacobs 2003b) (4.14) W/L: Align prosodic word to the left; align feet to the left. (Jacobs 2003b) (4.14) H/R : Align head foot to the left ; al ign prosodic word to the right. (Jacobs 2003b) (4.14) H/L : Align head foot to th e left; align prosodic word to the left. (Jacobs 2003b) These positioning constraints rank above the lowest ranking constraint PARSEwhich requires parsing into feet in order to compute the placement of the primary accent: NONFIN >> W/R >> W/L >> H/R >> H/L >>PARSE. (4.14) PARSE: Parse syllables into feet. (Jacobs 2003b) Noticeably absent in the constraints discu ssed thus far is the question of quantity sensitivity although Jacobs (2000, 20 03b) has proposed for Latin two constraints that reflect the 140

PAGE 141

preference f or heavy syllables in prominent positi ons but he finds it unnecessary to incorporate them into the constraints he uses to generate optimal output for Latin. (4.14) WSP (Weight to Stress): heavy syllables ar e prominent in foot structure. (Jacobs 2000, 2003b) (4.14) QS (Quantity Sensitivity): heavy syllables are stressed. The preference for heavy syllables is reflec ted also in the inviolable constraint FTBIN. FTBIN disallows a foot of the type (L) but permits (H) as a well formed foot as well as ('L.L). Because PARSEis a low ranking constraint it is not necessary to admit ('H.L), an uneven trochee, as a well-formed foot. This is particularly true where th e light syllable is word fi nal as in ('H)L. The Latin Nominal System Latin nouns develop from two categories of Proto Indo-European nouns: root nouns, formed without derivational suffixes, and derived nouns, created by suffixation (Baldi, 1999, 300-303). Root nouns are relatively infrequent in Latin. The much larger class of derived nouns has characteristic or thematic vowels. In both in stances case endings are then added to the root or stem (root + suffix). Table 4-1 (based on Baldi, 310-330) shows the resulting nominal inflections in Latin for both nominative and accusative case. Table 4-1. Nominal inflection in Latin Vocalic stem Consonantal stems with thematic vowel Thematic vowel (declension class) (2nd) (1st) (3rd) (4th) Pure consonantal stems (3rd) Singular Nom. (non-neuter) -us -a -is -us -s Nom. (neuter) -um -e, Acc. (non-neuter) -um -am -em -um -em Acc. (neuter) -um -e, Plural Nom. (non-neuter) -ae s s s Nom. (neuter) -a -ia -ua -a Acc. (non-neuter) s s s, s s s Acc. (neuter) -a -ia -ua -a 141

PAGE 142

The accusative is th e nominal form that regularly passes into Romance with a few notable exceptions such as the genitive in the days of the week, some toponyms, and patronyms; the ablative in formation of adverbs (Lloyd 1987, 276-277); and the nominative in personal names such as Carlos as well as a few other nouns such as Cast. dios and Port. deus from Lat. d us, i m. god. Although the NONFINALITY constraint prevents construc tion of a foot in the last syllable in Latin it is worthwhile to note the mora icity of the inflectional endings associated with the nominal forms that enter Ibero-Romance. For all declensions except the first the case marker for nominative singular is -s ; the preceding vowel is also shor t in all cases. It has been effectively argued that s can be viewed as extrametrical a nd in the case of Attic Greek (Yip 1991, 67) permissible word final clusters have an extr ametrical s-slot; permissi ble coda clusters (-CC) are n, l+s or p, k+s. The Latin third declension consonant stems are similar to the obstruent + s cases as in princeps chief and dux (duk-s) leader. The accusative singular for all declensions consists of short vowel + -m The weakness of word final nasals is well attested: Inscr iptional evidence such as OLat. OINO one (for oinom) and AIDE building (for aidem ) indicates a weak or nonexistent pronunciation for -m (Baldi 1999, 277). In addition to the inscriptional eviden ce there is also the tr eatment of word final nasals in Latin poetry discusse d at length by the grammarians and summarized in Lindsay (1894, 61) as the curious usage of Latin poetry, by which a word ending with m elides its final syllable before an initial vowel or h, just as though it ended with a vowel. If m did not make position in poetry then it is unlikely that a final syllable consisting of short vowel + m would be perceived as having two moras. As for the first declension, the i nherited long vowel is systematically shortened in the singular whethe r or not a consonant follows. Baldi (1999, 318) notes that on comparative basi s the ending of the nominative should be reconstructed as 142

PAGE 143

Sihler (1995, 266-269) notes that the shortening of final *poses a dilemm a because it seems to apply only to the vowel /a/. Howeve r, the development of the accusative m in Proto-Italic does seem to provide paradigma tic support. In this case, undergoes the shortening that is characteristic of long vowels in final syllabl es before consonants with the exception of s It should also be noted that differentiation from (<-d), the emerging ending of the ablative may also be a factor in shortening the vowel of the nominative *. In the plural, is preserved in the accusative ending -s The accusative plural is characterized by long vowel + /s/ in all declension classes. If both final m and s are treated as non coda segments, then the crucial difference between the singular and plural stems (root + them atic vowel) of the accusative concerns vowel quantity, a measure that is lost in the transition fr om Classical Latin to Romance. Without the NONFINALITY constraint it is possible that a more complex system of variable word accent would have developed, as did indeed occur in third declension forms such as hmen man (nominative singular) but homnibus in the dative plural, with primary word accent alternating between the first and s econd syllables of the root; or, as in the case of the genitive plural of the first and second declensions between root and suffix. Thus, even in those words that have variable accent and permit the primary accent to occur in a suffix rather than a root, there is still a constraint agains t building a foot in the word fi nal syllable. In Ibero-Romance retention of the word final accusative marker s fulfills an important morphological function although concordance does provide another means for distinguishing singul ar/plural forms (Note that s is regularly elided in French). For Classical Latin, then, the resulting accusativ e singular and plural endings are displayed with an indication of the usual gender (masculin e/feminine) of words of that declension. It should be noted that the 5th declension, characterized by -, is an Italic innovation (Baldi 1999, 143

PAGE 144

334) and contains few words of which even fewer pass into Rom ance. Some, such as di s day, are so strongly identified with the first declension that they acquire its characteristic vowel as in Spanish and Portuguese da (dia) The 2nd and 4th declensions co alesce in Late Latin/Early Romance through phonological and analogical processe s. Nouns from the third declension in Latin could be of any gender and there are no di stinguishing morphological characteristics as can be seen in these imparisyllabics: ars, artis, f. skill, ens, entis, n. thing, and fons, fontis, m. fountain. The fate of the final unstressed vowel of the accusative in all declensions differentiates the three Ib ero-Romance languages under st udyCastilian, Catalan, and Portugueseand has important implications for f oot building and resulting metrical patterns. Singular Plural 1st declension -am s feminine 2nd declension -um s masculine 3rd declension -em s feminine, masculine 4th declension -um s masculine In Latin, neuter nouns are present in the second, third, and fourth declensions. The simplification of the three way gender distincti on to a two way system in Romance resulted in identification of the characteristic a suffix of the first declension with feminine gender and the characteristic o suffix of the second/fourth declension w ith masculine gender. Most neuter nouns came to be identified with the o category, perhaps due to th e polysemy of the Latin um suffix. The third declension lacks a pervasive ch aracteristic vocalic su ffix; many nouns have consonantal stems to which the -em suffix is applied in the accusative. The widespread syncope of early Western Roma nce, evident in Ibero-Romance, has been seen as an indicator of an accentual system th at produced weakening of atonic vowels. Syncope emerges as a force of change in the formativ e period of the Romance languages in the Iberian Peninsula where the earliest tes timonies of the emerging languages show vacillation with regard 144

PAGE 145

to pre-tonic and post-tonic vowels in word intern al position as we ll as word final vowel deletion or apocope. Syncope has already been shown to be a feature of popular Latin in the previous discussion of the Appendix Probi (Chapter 3). Wheeler (2006, 48) attempts to distinguish between early pan-Romance syncope, which he attribut es to a fairly small set of words, such as s l dus, i m. a gold coin where two-syllable *SOLDU appears to be the ancestor of Catalan sou, Castilian sueldo and Portuguese soldo Later syncope, and particular ly apocope, is less uniform in distribution. The degree of apocope that occurs in the transition from Latin to the different Ibero-Romance forms a gradient with regard to declension classes and geographic dispersion. (West) (East) Third declension Portuguese Castilian Catalan Second declension Likely Unlikely First declension Figure 4-1. Likelihood of apocope by declension class and language. The commonalities and differences that result from language-specific accentual systems are examined here beginning with first declension nou ns of two-syllables, th e noun class least likely to undergo apocope, characterized by the vowel a before the inflectional ending or case marker. Construction of the Data Set In order to analyze in detail the prosodic tend encies for Western Romance as they relate to the nominal systems of Castilian, Catalan, and Portuguese a dataset has been constructed of nouns of Latin origin (with a few Greek words that enter Romance th rough Latin) that are common to these three major Romance languages of the Iberian Peninsula. The point of departure is a set of Spanish nouns of Latin origin retrieved from Moliners Diccionario de uso del espaol (in future, DUE) on CD-ROM. The initial lexical sort specified three Boolean operators (etymology = Latin; function = noun; freque ncy = frequent). Infrequent use is defined in the introduction to the DUE (Moliner 1998, xxxi) as follows: Una acepcin es considerada no 145

PAGE 146

usual por diferentes m otivos: por tratarse de un sentido antiguo o dial ectal, por estar muy restringido a un mbito cientfico o profesional. The operators used in this initial search retrieved 6310 nouns. The form of the Latin et ymon was corroborated in the online version of Lewis and Short (Perseus Project). Corresponding noun forms, that is from the same etymon, were selected for Catalan and Portuguese using standard bilingual dictionaries: Diccionari avanat catal-castell/castellano-cataln (Vox), Gran diccionario espaol-portugus/ portugus-espanhol (Espasa), as well as online m onolingual dictionaries such as Gran diccionari de la llengua catalana (Enciclopdia Catalana, http://www.grec.net/home/cel/dicc.htm ), Diccionari catal-valenci-balear (A lcover and Moll, Editorial Moll, http://dcvb.iecat.net/ ), Dicionrio universal da lngua portuguesa on-line (Texto E ditora, http://www.priberam.pt/dicionarios.aspx ), Dicionrio de Lngua Portuguesa (Porto E ditora, http://www.infopedia.pt/default.jsp ). Items that did not have correspondents in the other two languages w ere discarded. The re sulting initial dataset of nouns of Latin origin common to Castilian, Catalan, and Portuguese contains 3605 nouns inventoried in Excel file format. Not all forms are incorporated in this study; therefore, the sum of individual tables by category is less than the original total. A numeric value appears for each subset that is presented. The database files (Appendix D) contain column headings for the Latin etymon, an indicator of its declension cla ss (Dcl), number of syllables ( ), locus of primary accent (Ac) counting from right to left where ultimate accent is 1, gender (Gen), nature of the last three syllables (HV=syllable with long vowel or diphthong, HC=syllable w ith coda consonant, L=light syllable; 1 is added after the tonic syllable), a nd the prosodic template (Temp). The information relative to syllable count, place of accent, and nature of the last th ree syllables is repeated for the corresponding nouns in Catalan, Ca stilian, and Portuguese. In the sample that follows outputs 146

PAGE 147

appear in shaded cells. Results are sortable by any of the colu m n headings: declension, number of syllables, locus of accent, prosodic template. For the etymon, gender is also indicated. Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Acc Temp POR Ac Temp 1 2 2 HV1.L f aula 2 2 HV1.L aula 2 2 HV1.L aula 2 2 HV1.L aula, ae 1 2 2 HV1.L f aura 2 2 HV1.L aura 2 2 HV1.L aura 2 2 HV1.L aura, ae 1 2 2 HV1.L f broma 2 2 L1.L bruma 2 2 L1.L bruma 2 2 L1.L br ma, ae 1 2 2 HV1.L f causa 2 2 HV1.L causa 2 2 HV1.L causa 2 2 HV1.L causa, ae c na, ae, 1 2 2 HV1.L f cena 2 2 L1.L cena 2 2 L1.L ceia 2 2 HV1.L An Overview of the First Declension Maintenance of the final vowel in first declension nouns has been assumed to be attributable to morpheme preservation. In this chapter a faithfulness constraint, MAXMORPH, is discussed as a possible explanat ion although a phonologically based c onstraint is presented in the discussion of second/fourth declension nouns in Chapter 5. First declension nouns are remarkably stable overall. Loss of internal syll ables is more often the result of glide formation rather than syncope. Within each subclass of nouns the preferred prosodic template favors light syllables. Preservation of the final vowel allows construction of preferre d foot form ('L.L) at right word edge. The preservation of proparoxyt onic accent in polysyllabl es does not negate the fact that the template ('X.L.L), where X may be either a light or heavy syllable, continues to be prosodically marked in th ree languages under study. Two-Syllable Nouns Nouns of two-syllables provide the simplest configuration of root and suffix. Since Romance nouns generally come from the accusa tive case it will be assu med that the basic template consists of a first syllable of variable weight and a final light syllable. The three possible templates, then, are 'L.L, 'HC.L, and 'H V.L where HC indicates a heavy syllable with a coda consonant and HV a heavy syllable with a long vowel or diphthong. Two of the patterns have the preferred branching nucleus whereas the third pattern has word initial accent by default 147

PAGE 148

and produces an undesirable st ructure that violates FTBIN since 'L.L, parsed as ('L), fails to meet any criteria for binarity. What are the possible permutations of syllable st ructure in Latin to create either a heavy or light syllable? It is assumed th at syllable onsets do not affect syllabic weight although an onset is desirable for the well-formedness of a syllable. However, there is no co nstraint that requires a syllable onset in Latin. Possible onsets consist of a single consonant s or clusters of up to three segments such as initial strin str ta, ae, f. paved road. Cser (2001, 175-176) inventories complex word initial onsets for Latin as follows: stop or /f/ + liquid /s/ + voiceless stop /s/ + /w/ /s/ + voiceless stop + liquid In Figure 4-2 these possibilities are represented as s (/s/), P (voiceless stop), R (liquid). Word Word Onset Rhyme Onset Rhyme | | (s) (P) (R) V (s) (P) (R) V Onset Rhyme Onset Rhyme | (s) (P) (R) V V (s) (P) (R) V C 'L.L 'H.L Figure 4-2. Permissible sy llable structures in Latin words of 2-syllables. As simple onsets, Latin may have fricatives /s/ or /f/, a voiced or voicele ss stop, or a sonorant. Complex onsets can be twofold, that is, s + P, or P + R, or threefold, s + P + R. It is understood that phonotactic restrictions in Latin do not permit all possible co mbinations; for example, /tl-/ and /sn-/ are disallowed as ini tial complex onsets. Knowing what are permissible onsets aids in 148

PAGE 149

the reso lution of complex internal clusters wh ere syllabification is uncertain and often obscured by contradictory evidence. An accent must be assigned in accordance with the constraint LX PR (Every Lexical Word must correspond to a Prosodic Word). The defa ult location is the penu ltimate syllable which avoids violation of the NONFINALITY constraint. However, 'L.L violates FTBIN (Feet are binary at some level of analysis) if it is assumed that the Latin metrical unit is a moraic trochee and the final syllable is extrametrical. Apocope, which occurs regularly in Catalan in second and third declension nouns, could then be viewed as a repair strategy because it would most likely create a monosyllable of the shape (C)VC. It should also be noted that th e loss of vowel quantity in Late Latin rendered light the first syllable of many nou ns of the 'HV.L type. Cases of 'HC.L that depend on a coda consonant for syllable weight would be expected to preserve the coda if bimoraicity of the accented syllable and extrametricality of the final syllable obtain. However, analysis of the database shows that this is not always the case. Since foot building occurs at the right edge of the word, pr eservation of the final vowel gives the appearance that the NONFINALITY constraint is still observed in Ibero-Romance (in the first declension). In fact, Ha rris (1991, 1992, 1995) has suggested that consideration of gender markers as extrametrical in modern Spanish is a way of computing prim ary accent by placing it on the first available syllable to the left of the edge marker ]. In or der to produce the desired results a slot for an unrealized vowel must be in serted at the right edge in words that have ultimate accent on an HC type syllable. Thus animal animal is construed as ani ml V for purposes of accent assignment. However, preservation of final -a in first declension nouns seems to be, rather, a function of the inherent st rength of the segment, that is the high degree of sonority of the vowel rather than a question of preserving a morpheme After all, second 149

PAGE 150

declension nouns also have a char acteristic vowel or class m arker which is routinely elided in Catalan. Analysis of the dataset begins with two-syllable words of the first declension that pass into the three major Romance languages of the Iberian Peninsula. At the two-syllable level, word shapes correspond to one of two types: 'L.L or 'H.L (with subtypes 'HC.L and 'HV.L). Heavy syllables labeled HC are those with a coda consonant and th ose labeled HV consist of long vowels or falling diphthongs. It should be noted th at HV type syllables for the most part become light with few exceptions, diphthongs /aw/ and /ew/ preserved in learned or semi-learned words, frequently of Greek origin. Syllables labeled HC are those with a coda consonant. They are the most likely to remain as heavy syllables in the transition to Romance. Ordinarily, treatment of word internal consonant sequents as tautosyllabic or heterosyllabic is straightford with the exception of /s/ + voiceless st op + liquid clusters as in f nestra, ae, f. window. In the case of f nestra the cluster must have been heterosyllabic at some point because the accent is paroxytonic. Devine (1980, 155) sugg ests, however, that the internal stcluster may have been tautosyllabic in colloquial Latin although classical prosody treats the cluster has heterosyllabic. A similar problem, the syllabification of internal obs truent + /r/ has been previously discussed in Chapter 3. In the case of monstrum, i n. monster, the initial sylla ble is clearly heavy whether or not it consists of mon or mon s-. Cser considers an alternative treatment of /s/, extrasyllabicity. Based on Goldsmith (1990) Cser proposes extrasyllabicity for /s/ in both word initial and word final positions. His word tree, modeled in Figure 4-3, shows bran ches for word initial /s/ in spes, sp i f. hope and word final /s/ in daps (d pis), dpis f. feast that descend directly from the Word level node. According to Goldsmith (1990, 107-108) a traditional view of sy llabification holds that every segment belongs 150

PAGE 151

to at leas t one syllable except fo r certain, language-specific word initial or word final segments which are part of the prosodic word but at word level rather than syllable level. These segments are instances of licensed extrasyllabicity. Word initial epenthesis in Late Latin and Romance could then be viewed as a repair strategy which provides a syllable nucleu s for the stray initial segment that is now a coda of the new syllable. Word Rhyme Onset Nucleus Coda /s/ Obs Son V(V) Son Obs /s/ Initial {s} p e e s Final d a p {s } Figure 4-3. Extrasyllabic /s/ in Latin word initial and word final position. Sim ilarly, the variant nomina tive singular forms of thir d declension nouns with and without a nucleus before s, as in daps (d pis) indicate the special status of /s/. Extrasyllabic status of final /s/ also obviat es the question of superheavy syll ables of the type VCs which are common in plurals in modern Catalan. Attachme nt of word final /s/ to a proximate onsetless syllable in some dialects of Spanish that regularly aspirate syllable-final /s / could be construed as an instance of contingent extrasyllabicity wh ich Goldsmith (1990, 108) describes as consonants that fail to become syllabified and are wai ting for a syllable to come along for them. The data set used for this anal ysis consists of 232 nouns of Latin origin present in all three languages, Catalan, Castilian, and Portuguese. They are distributed as follows in terms of weight 151

PAGE 152

of the first syllable of the Latin etymon, where HC is a heavy syllable with coda consonant, HV a heavy syllable consisting of a long vowel or diphthong, and L a light syllable. Table 4-2 shows that the number and percent of distribution of initial syllable types. The occurrence of undesirable (L).with a parsed but defective foot in the first syllable and unparsed or extrametrical last syllableis relatively limite d; however, it should be remembered that the HV type syllable becomes monomoraic to a large extend in Late Latin. Table 4-2. Distribution of H/L syllable type in first syllabl e of Latin 2-syllable nouns of the first declension Initial syllable type Number/percenta ge Heavy/light initial syllable HC 121 (52.2%) HV 71 (30.6%) 192 (82.8%) H L 40 (17.2%) 40 (17.2%) L n=232 In preparation of a contingency table to ev aluate the distributi on of heavy and light syllables in Latin first declension nouns, the first two categories, HC and HV are combined to produce two mutually exclusiv e categories, H or L. It is then assumed that for the data set of 232 nouns there is an equal chance that the accented sylla ble (first syllable) wi ll be heavy or light. The operative negative hypothesis for Table 43 is derived from the familiar WSP principle proposed for Latin (Prince and Smolensky 2004, 70-71) : Heavy syllables are prominent in foot structure and on the grid Heavy syllables are not prominent in foot structure and on the grid. The contingency table (Table 4-3), is then submitted to a chi-square goodness of fit test with these results: x2=99.6 and p<0.0001. The preponderance of heavy syllables, then, is significantly larger than expected. Table 4-3. Contingency table for H/L in first syllable of Latin first declension disyllables Syllable Type Actual number Expect ed number Expected percentage Heavy 192 116 50.0% Light 40 116 50.0% 152

PAGE 153

In the c ase of Latin two-syllable words the first syllable is accent bearing by default through the operation of NONFINALITY which does not allow the construction of a foot in the last syllable even when that syllable is heavy, as in the accusative plural of first declension nouns. Latin nouns that are heavy on the basis of long vow els or diphthongs are li kely to become light in early Romance with the exception of a few l earned or semi-learned words that retain the diphthong /aw/. Similarly, contingency tables can then be reconstructed for Catalan, Castilian, and Portuguese to determine whether or not syllabl e weight continues to b ear a relationship with placement of accent. It is expected that these tables will validate many of the following predictions with regard to the Latin etyma. OT constraints are able to account for cha nges from input to output and for differences among languages. The following constraints or fam ilies of constraints are the basis of evaluation of outputs, both those that are motivated by langua ge universals and those that respect language specific exigencies. 1. There is a family of overarchi ng faithfulness constraints, MAX: Every segment/feature of the input has an identical correspondent in the output. In terms of the dataset this means that nouns will not change with regard to the number of syllables, the nature of those syllables and their segments, nor will the locus of word accent be altered. 2. Initial HC syllables will remain unchanged: This prediction corresponds both to MAX and to WSP. However, this syllable type is in violation of a well-formedness constraint, that is, NOCODA. In the transition to Ibero-Romance codas that are obstruents (but not sibilants) are subject to lenition and deleti on, especially in the western languages. Therefore, the constraint might be specified as NOCODA[-strid]. Vocalization of an obstruent in coda position could be perceived as a mora pr eserving strategy but it is infrequent. 153

PAGE 154

3. Initial HV syllables will becom e light: All long vowels in La tin merge with the corresponding short vowels; diphthongs are reduced to monophthongs: ae> oe>e, aw>o. This is a violation of MAX but does not alter the syllabic count nor place of accent. However, the resulting trochee is now a suboptimal light foot (L). 4. Initial L syllables will remain unchanged. This reflects another faithfulness constraint, DEP: Every segment/feature of the output has an identical correspondent in the input. In this case it is assumed that no extra segments or syllable s will be added to create a more optimal foot structure of ('L.L) or ('H). The contingency tables compare the obser ved frequency for each of the three languages under study with the expected frequency based on the criteria enumerated above. The tables combine the HV and L categories of Table 4-2 an d project that the HC category will remain unchanged. The negative hypot hesis states: It is not expected that the proportion of HC type initial syllables in Ca talan/Castilian/Portuguese disyllable s inherited from the Latin first declension will differ from Latin. The results fo r all three languages ar e p<.0001, indicating that the actual distribution of H and L type syllables differs significantly from the input data. The number of light syllables is much larger than ex pected. This means that to a large degree word level accent does not now coincide with a heavy syllable. In Tables 4-4, 4-5, and 4-6 the expected number for heavy syllables corresponds to original HC type syllab les in Latin, that is, those which are heavy by virtue of a coda consonant which renders the syllable bimoraic. Table 4-4. Contingency table for H/L in first syllable of Catalan nouns from Latin first declension disyllables Syllable Type Observed Expected number Expected percentage Heavy 88 121 52.2% Light 144 111 47.8% The results for Table 4-4 (Catalan) are x2=18.8 and p<.0001. 154

PAGE 155

Table 4-5. Contingency table for H/L in first syllable of Castilian nouns from Latin first declension disyllables Syllable Type Observed Expected number Expected percentage Heavy 83 121 52.2% Light 149 111 47.8% The results for Table 4-5 (Castilian) are x2 = 24.9 and p<.0001. Table 4-6. Contingency table for H/L in first syllable of Portuguese nouns from Latin first declension disyllables Syllable Type Observed Expected number Expected percentage Heavy 86 121 52.2% Light 146 111 47.8% The results for Table 4-6 (Portuguese) are x2 = 21.2 and p<.0001. There is little variation among the three languages. Since H and L are mutually exclusive categories an increase in one implies a decrease in the other. A detailed examination of the original syllable types and their preservati on or modification provi des a context for the consideration of ranked constraint s that capture the basis of word level prominence and motivate assignment of primary accent. Table 4-7 shows the preservation of HC initial syllables by language and resulting syllable type, and in graphic form, in Figure 4-4. Table 4-7. Preservation of HC initial syllables in Catalan, Castilian, and Portuguese nouns from Latin first declension disyllables Resulting Syllable Type HC HV L Catalan 77 (63.6%) 3 (2.5%) 41 (33.9%) Castilian 76 (62.8%) 1 (0.8%) 44 (36.4%) Portuguese 63 (52.0%) 14 (11.6%) 44 (36.4%) n=121 The nearly identical count of initial heavy sy llables (both HC and HV) across languages is striking and the most salient divergence is read ily explained. The Port uguese nasal vowel is treated as a long vowel (HV), a view supported by both internal and extern al evidence (Paradis and Prunet 2000). 155

PAGE 156

W ord internal nasals in coda position are preserved as dis tinct segments in Catalan and Castilian. Therefore, the unexpectedly large number of HV cases (14; 11.6%) is unique to Portuguese. If the HV and HC segments are viewed as a whole (see Figure 4-4), it is apparent that in moraic terms the distribution of heavy and light syllables is similar across languages. For Portuguese heavy syllables are present in 63.6% of the words, the same percentage found in Catalan, and nearly the same as Castilian at 62.8%. Other sources of HV type syllables are less transparent. Vocalization and formation of a fa lling diphthong as the reso lution of an obstruent in coda position is a less frequent outcome. 77 76 63 3 1 14 41 44 44 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%Ca t alan Castilian P or t ugu es e L HV HC Figure 4-4. Outcomes of initial HC syllables from Latin first declension disyllables in Catalan, Castilian, and Portuguese. (n=121) The three cases of vowel + glid e that result in Catalan are from syllables of the type C(C)VC where the coda cons onant is a voiceless stop. Th e identical treatment of pacta across languages, given the semantic nature of the word, points to a semi-learned form where faithfulness constraints are more likely to hold. Lat. capsa, ae f. case Cat. caixa Cast. caja Port. caixa Lat. pacta, ae, f. pact Cat. pauta Cast. pauta Port. pauta 156

PAGE 157

Lat. tructa, ae, f. trout14 Cat. truita Cat. trucha Port. truta In the case of capsa the initial syllable in both Catalan and Portuguese outcomes can be rendered as both [ka] and [kaj] before the palatal [ ] but the glide is most likely an anticipation of the following consonant rather than an intent to preserve the bimoraicity of the Latin initial syllable /kap-/. In the absence of the palatal glide as a countable segment there may still be reasons to consider that / ka. a / is in moraic terms. Baker (2004, 4-19) has argued effectively that palatal consonants in Spanish because of their complex articulatory properties occupy two timing slots or two moras. Followi ng this line of reasoning only Portuguese truta shows a moraic loss in the first syllable15. As for the group of 40+ nouns where the first sy llable is now light, thei r source is primarily degemination of stops and to a lesser extent dege mination of nasals and liquids. While loss of vocalic quantity is a pan-Romance phenomenon, the loss of quantity of intervocalic consonants has important exceptions, namely the area to the south of a geographic line from La Spezia to Rimini in Italy. Cravens (2000, 48) reviews the extensive literature on degemination and lenition of intervocalic stops in Romance. He argues agai nst preservation of intervocalic geminate stops, /-pp-/, /-tt-/, -kk-/, as avoidance of merger with single intervocalic consonants because, as he notes, coalescence is a common phenomenon in di achronic sound change. For Ibero-Romance, the common output of the geminate stops is a sing le intervocalic stop. Degemination is also seen in liquids (/-ll-/) and nasals (/-mm-/, /-nn/). Palatalization of coronals /-ll-/ to / and /-nn-/ to / 14 The outcomes of Lat. tructa raise the question of an alternate form with a long vowel, cf. OE tr ht although Biville (1990, 79-80) notes that the tonic vowel of the supposed etymon, Greek is often rendered by u in initial syllables in Latin and that the outcomes of Latin / / and / / merge in Western Romance. However, the outputs of Lat. fr ctus, s, m. also seem to support a long vowel (see n. 15). 15 The reduction of uk.tu to uto is also seen in the reflexes of Lat. fr ctus, s, m. enjoyment in Spanish and Portuguese fruto in contrast with vocalization of syllable-final /k/ in Port. noite from Lat. nox, noctis, f. night. 157

PAGE 158

occurs in the case of patrimonial words, although there are a few instances of simplification to /l-/ and /-n-/ in learned words. In the case of palatalization once ag ain the issue of syllable weight emerges. The output of falla, ae f. trick, artifice, for example, is uniformly / fa a/ with a light first syllable. Unlike Cat./Port. caixa above there is no indication that the initial syllable is[faj]. The relationship of the three subsets by language is seen in the Venn diagram in Figure 4-5. Castilian n=44 Portuguese n=44 5 0 6 38 1 0 2 Catalan n=41 Figure 4-5. Coincidence of subs ets of L type initial syllable in Catalan, Castilian, and Portuguese resulting from HC first syllable in Latin fi rst declension disyllables. Of the original set of 121 first declension disy llables with HC type first syllable, roughly one third (41 in Catalan and 44 in Castilian a nd Portuguese) lose their bimoraic status. The shaded area indicates that 38 of the HC type syllables maintained in Castilian, Catalan, and Portuguese correspond to the same subset of nouns Castilian and Catalan share one additional item whereas 5 cases in Catalan, 6 in Portuguese and 2 in Catalan are unique to those languages. Moraic loss comes primarily from degemination of voiceless stops, liquids, and nasals. There are also isolated instances of rs->-ss ->-s (Catalan), nd->nn ->-n(Catalan), kt ->jt (with 158

PAGE 159

fronting of preced ing vowel, Portuguese); kt ->-jt ->-t(Castilian only); kt ->t (with fronting of preceding vowel, Catalan). The number of cases in which heavy first syllab les become light other than as a result of degemination is extremely limited. At this juncture it appears that rather than the interaction of well-formedness constraints (NOCODA) and faithfulness constraints (MAX/IO) one sees the result of what Weinreich (1958, 12) has termed der Quantittenskollaps. The complex relationship of geminate consonants, duration of preceding vowels, and syllabic boundaries is beyond the scope of the current study but Loporcaro (1996) has formulated a rule of vowel lengthening that demonstrates that for modern Italian the quantity relationship described by Weinreich for Latin (1958, 24) as the second stage in the dismantling of the quantitative system holds true: 1) long vowel + short consonant as in c pa, ae vat; 2) short vowel + l ong consonant as in b cca, ae cheek. Loporcaros vowel lengthening rule (1996, 162) follows: (4.14) Italian Vowel Lengthening: V [+long] / [ ] (Loporcaro 1996) _____ +stress Loporcaros rule should be viewed with cauti on. While words of two-syllables seem to correspond to the predicted template of VVC vs VCC (where duplication of V/C indicates long vowel/consonant), more recent studies demonstrat e that vowel length is also related to the number of syllables in a foot (assuming that trisyllables with an tepenultimate accent constitute a foot) as well as position within the word (Hajek, Stevens, and Webster 2000). Final vowels show both lengthening and shorteni ng, independent of their status as tonic or atonic vowels. The subset just examined contains the heavy sy llable type most likely to remain heavy, that is, a syllable with a coda consonant. The HV type, representing in Latin both long vowels and diphthongs, persists as a heavy sy llable only when a diphthong is retained. Table 4-8 indicates that preservation of a diphthong is the outcome in a small percentage of cases, corresponding 159

PAGE 160

mostly to /aw/. The resulting nouns are all disyllabic with the exception of Portuguese l wool < l na, ae, f. These outcom es are in contrast with those of Tabl e 4-7 in which over 60% of nouns in all three languages pres erve the moraic count of the input form (HC). Table 4-8. Accented syllable in nouns from Latin first declension disyllables with HV initial syllable HC HV L CAT 0 (0.0%) 7 (10.9%) 64 (90.1%) CAS 0 (0.0%) 6 (8.4%) 65 (91.6%) POR 0 (0.0%) 9 (12.7%) 62 (87.3%) n=71 While the resulting ('L) configuration is a suboptimal trochee (if the final syllable is extrametrical) it is reinforced as a template by the outcome of original Latin ('L) seen in Table 4-9. Even the five cases of prothesis in this group that resu lt in a three-syllable word do not represent an improvement of the prosodic co nfiguration which is now H('L). The head foot is separated from the right word edge by an unparsed syllable and la cks binarity whereas the initial syllable is bimoraic but not stressable due to constraints like HEAD-MAX-IO which requires that the head foot of the input form also be the head foot in the output. Table 4-9. Outcome of Latin first declension ('L) nouns HC. 'L (with prothesis) 'L.L 'HV.L 'HC.L 'HV CAT 5 (12.5%) 34 (85.0%) 1 (2.5%) 0 (0.0%) 0 (0.0%) CAS 5 (12.5%) 35 (87.5%) 0 (0.0%) 0 (0.0%) 0 (0.0%) POR 5 (12.5%) 33 (82.5%) 0 (0.0%) 0 (0.0%) 2 (5.0%) n=40 With the exception of Catalan aigua water < qua, ae, f. and Portuguese m millstone < m la, ae, f. and r prisoner (fem.) < r a, ae, f. all cases of original ('L.L) remain as disyllables with primary accent on the penultimate light sy llable. The exeptional cases merit a brief comment. Catalan aigua suggests the influence of an alte rnate etymon, seen in the Appendix Probi, l. 112 (Appendix A), acqua, with a presumptive pronunciation of ['ak.kwa]. The lenition of syllable final, word internal /k/ resulting in a palatal glide is well documented in the formative 160

PAGE 161

stages of Catalan, for exam ple, Old Catalan treita OPort. soa > soo > ModPort. so (Parkinson 1988, 136) pronounced [ 's ]. The vowel reflexes that occu r in the two cases in Table 4-9, [ ] and [ ], are only found in stress bearing syllables in modern European Portuguese. They are also relatively longer than other accented vowels according to Costa (2004) and Delgado Martins (1988, 128-132). Relevant data from Costa are presented in Table 4-10. The contrast in duration between the vowels in stressed pos ition and the absorbed final a, which would have been rendered as [ ] in the transitional stage, is striking. In pre-tonic position [ ] is even further reduced. It should also be noted that final a is maintained in Portuguese in cases where there is sufficient contrast between the preceding vowel and the atonic vow el realized as schwa as in via < v a, ae f. road and lua < l na, ae, f. moon. The final vowel is also maintained in adjective paradigms such as bom, boa good (masc. sg., fem. sg.); howev er, in this case the preceding vowel is [o] rather than the more open [ ]. Hiatus in Portuguese may have constraints that require a certain degree of contrast between the vowels. If that minimum contrast is not maintained, the vowel either forms a glide and becomes part of a falling diphthong with the precedi ng accent bearing vowel, or it is completely 161

PAGE 162

absorbed. A lternatively, insertion of a glide resolves the undesirable V.V sequence as in areia < (h) r na, ae f. sand ((Parkinson 1988, 136). In addition to the etymological evidence indicating that m and r originally had a branching rhyme, Table 4-10 reinforces the notion of increased prominence of non-high vowels, although [a ] is shorter than expected especially when compared with surprisingly long [ ], a contextual varian t of /a/ (and occasionally /e/) in pre-tonic and post-tonic position as well as in unstresse d final syllables (Mateus and dAndrade 2000, 1723). Table 4-10. Duration of vowels in Europ ean Portuguese in stressed CV syllables Duration in ms Relation to mean (128 ms) for all vowels [i] 111 0.87 [e] 125 0.97 [ ] 139 1.08 [u] 110 0.86 [o] 129 1.01 [ ] 139 1.08 [a] 132 1.03 [ ] 141 1.10 While it is justified to code Po rtuguese monosyllables as HV, th ey are not heavy as a result of an effort to maintain the moraicity of the i nput form; rather, their increased prominence results from a general lengthening process that applie s to all accented vowels in Portuguese (Delgado Martins (1988, 128-132). Furthermore, tonic mid vowels are qualitatively distinct from their unaccented counterparts. Vowel raising is eviden t in the realization of unstressed /e/ and / / as [ ] and unstressed /o/ and / / as [u]. By combining data from Tables 4-7, Table 4-8, and 4-9 it is po ssible to see to what degree the two-syllable template for the subclass of largely feminine nouns fr om the first declension favor a ('L). or ('H). pattern. Table 4-11 demonstrates the combined outcomes of the 162

PAGE 163

232 disyllab ic first declension Latin nouns. The rows indicate the nature of the accented syllable in Latin; the number of words in each subclass appears in the colu mn labeled n=. Output forms, distributed by language, appear in the columns headed H and L Table 4-11. Heavy/Light (H/L) outcomes of accen ted syllables in Catalan, Castilian, and Portuguese based on nature of tonic syllable in Latin fi rst declension disyllables Input n= Output H L HC 121 80 (66.1%) 41 (33.9%) Catalan HV 71 7 (9.9%) 64 (90.1%) L 40 1 (2.5%) 39 (97.5%) HC 121 77 (63.6%) 44 (36.4%) Castilian HV 71 6 (8.5%) 65 (91.5%) L 40 0 (0.0%) 40 (100.0%) HC 121 77 (63.5%) 44 (36.4%) Portuguese HV 71 9 (12.7%) 62 (87.3%) L 40 2 (5.0%) 38 (95.0%) Total n=232 Figure 4-6 shows that despite the hi gh rate of retention of syllable weight in the case of the HC input forms, the predominant output pattern of two-syllable nouns is a left headed binary foot whose binarity is dependent on the presence of twosyllables. This interpretation requires setting aside the notion of extrametricality, which is ce rtainly no longer operative in Ibero-Romance. Acceptance of a syllabic trochee recognizes that the alternative interpretation, a moraic trochee with extrametricality, is too cost ly as it would require licensing of a defective foot under LX PR (Every Lexical Word must correspond to a Prosodic Word, Prince and Smolensky 2004 [1993]) for the majority outcome, labeled 'L.L in Figure 4-6. The notion of extrametricality is reflected in the OT constraint NONFINALITY (No head of a prosodic word is final in a prosodic word (Prince and Smolensky 2004 [1993])) which precludes construction of a foot at the right word edge. 163

PAGE 164

Figure 4-6. Distribu tion of heavy/light syllable s in tonic syllable of nouns from Latin first declension disyllables (n=232). 'H 'H.L 'L.L CAT CAS 0.9% 37.1% 62.1% POR 0.0% 35.8% 64.2% 0.0% 37.9% 62.1% 'H 'H.L 'L.L70.0% 60.0% 50.0% 40.0% 30.0% 20.0% 10.0% 0.0% Monosyllables constitute an exceptional cla ss corresponding to the undom inated constrainted LX PR (Every Lexical Word must correspond to a Prosodic Word, Prince and Smolensky 2004 [1993]). As a pattern of ultimat e accent emerges, as will be s een in second/fourth and third declension nouns, there is no longer a justification for failing to parse the final syllable of first declension nouns. Three-syllable Nouns In the case of Latin three-syllable nouns two accentual patterns are extant, the familiar penultimate accent when that syllabic is bimoraic and a retraction to the antepenultimate syllable if the penultimate is light. In terms of distri bution of light and heavy sy llables the two patterns, in all possible permutations, are summarized in Figure 4-7. Shading indicates the syllable that carries the primary accent. Since penultimate accen t requires a heavy syllable it can be expected that only syllables with a coda consonant or glide (in the case of falling diphthongs) will remain as heavy syllables if MAXIO is a high ranking constraint. 164

PAGE 165

HC HC A) Penultimate Accent HV HV L B) Antepennultim ate Accent HC HV L L Figure 4-7. Comparison of prosodic patterns in thr ee-syllable Latin nouns. The first declension disyllables examined above show remarkable faithfulness across the three languages under study in terms of syllabic count. Preservation of input syllables is also high in the case of trisyllables although there are some instances of syncope in the case of nouns with antepenultimate accent. The chronology of vowel elision is uncerta in as seen in the discussion of the Appendix Probi (see Chapter 3). It is possible that in some cases the input form that was extant in the formative period of the Ibero-Romance languages was, in fact, disyllabic rather than trisyllabic. First Declension Trisyllables with Penultimate Accent Well formedness constraints such as NOCODA, if high ranking, lead to frequent violations of MAXIO. However, when assignment of stress, as is the case in Latin, depends on syllable weight the loss of a mora is problematic. The change from quantitative to qualitative vowel system which characterizes the transition from Classical to Late Latin results in loss of HV as a candidate syllable for locus of primary accent. Furthermore, the relative infrequency of falling diphthongs and their monophthongization is another contributing factor in the loss of syllable weight dependent on a glide in the coda. Only the nasalization of vowels in Portuguese before a consonant coda preserves moraic count if one cons iders nasality to be a separate segment. This argument is in line with Prunet and Paradis (2000 ) and supported by the appa rent spread of the feature nasality to other segments in the case of resulting nasal diphthongs as in the case of bem 165

PAGE 166

well rende red [ bj ] in European Portuguese. In the case of nasal vowels that are not part of a diphthong experimental data show that in an acce nted CV syllable they are on average 1.42 times longer than corresponding oral vow els (Costa 2004, [3]), an asser tion in marked departure from Delgado Martins study. In brief, it can be e xpected that the HV environment seen in form A) in Figure 4-7 will almost completely disapp ear in the formative period of Ibero-Romance with the exception of a few falling diphthongs pres erved in learned or semi-learned nouns such as Gr. via Lat. Centaurus, i, m. Centaur realized as Cat. centaure and Cas./Port. centauro On the other hand, one would expect faithfulness to the HC type syllab le if there is an operative principle of minimum bimo raic count in the syllable that constitutes the head foot. In the disyllables previously treated preservation of a coda (sometimes a glide) in initial HC syllables, the accent bearing syllables, occurred in 66.1% of the cases in Catalan, 63.6% in Castilian, and 63.6% in Portuguese. If these percentages are then us ed to calculate an expected value for three-syllable words with penultimate accent falling on HC type syllable it is possible to construct a contingency table to test the ne gative hypothesis that there is no difference in the outcome of HC type accented sy llables in words of two or three syllables from the first declension. The data set consists of i nput forms (HC)('HC), (HV)('HC), and L('HC). The results are provided in Table 4-12. Only for Portuguese, where p < 0.05, are the results significant, that is, a higher number of syllables than exp ected are light. This reflects the constraint, present in Portuguese toda y, that limits coda consonants to /l/, / / and /s/ (Mateus and dAndrade 2000, 52). 166

PAGE 167

Table 4-12. Contingency table for H/ L outcom es in tonic syllable from (HC)('HC), (HV)('HC), and L('HC) first declension nouns Syllable Type Observed Expected Expected number number percentage CATALAN Heavy 38 36.4 66.1% Light 17 18.6 33.9% x2=0.2080, p-value is 0.6484 CASTILIAN Heavy 31 35.0 63.6% Light 24 20.0 36.4% x2=1.2571, p-value is 0.2622 PORTUGUESE Heavy 26 35.0 63.6% Light 29 20.0 36.4% x2= 6.3643, p-value is 0.0116 n=55 The first of the data sets, that is (HC)('HC) input form as in cisterna, ae f. cistern, is small (n =18) but worth examining in detail beca use of the surprising differences in faithfulness to the HC type syllable present in both the penultimate accent-bearing syllable (S2) and the initial syllable (S3), presumably not prominent in the prosodic structure of the word. While the loss of syllabic quantity in S2 can be explained by constraints such as NOCODA, as well as the process of degemination, faithfulness to the initia l HC syllable merits further investigation. Table 4-13. Distribution of heavy/light sy llables in outcomes of first declension trisyllables with penultimate accen t of the type (HC)('HC) S3 S2 H L H L CAT 15 (83.3%) 3 (16.7%) 11 (61.1%) 7 (38.9%) CAS 14 (77.8%) 4 (22.2%) 8 (44.4%) 10 (55.6%) POR 14 (77.8%) 4 (22.2%) 6 (33.3%) 12 (66.7%) n=18 Table 4-13 shows a high rate of retention of syllable weight in S3 across languages and for Catalan also in S2. In previous discussions of the Latin accentual system the constraint WSP has been formulated but not utilized in construction of the OT table to generate Latin primary accent. 167

PAGE 168

(4.13) Weight-to-Stress Principle (WSP) (Pri nce and Smolensky 2004, 63): Heavy syllables are prominent in foot structure and on the grid. WSP is a low-ranking constraint, even in Clas sical Latin, and it is frequently violated. Extrametricality removes some of these violat ions; for example, suffixes consisting of long vowel or long vowel followed by a coda consonant occur in both nominal declensions and verb paradigms. However, in positions that cannot be construed as extrametrical there are also occurrences of a heavy syllable in non prom inent position. The ini tial heavy syllable of trisyllables with penultimate accent is such a case. In many instances in which placement of accent on the penult depends on a branching rhyme, the loss of phonemic vowel length, monophthongization, and the degemination of presumed long consonants such as -ss-, -ll-, -nnrender the penultimate syllable light yet it remains the locus of the primary accent. To further explore the unexpected faithfulness to the initial HC syllable, the small data set (18 items) in Table 4-13 (outputs of (HC)('HC)) is combined with the ou tput of (HC)('HV) as in the word fort na, ae f., fortune, an additional 57 word s, to produce the outcome seen in Figure 4-8 where n=75. All but one of the nouns in this data set of 75 are trisyllabic with penultimate accent. The one anomalous form is Portuguese irm (< Lat. germ na, ae, f. sister) which is now disyllabic with ultimate accent due to loss of intervocalic /n/. This outcome appears in the column headed by N in Figure 4-8 indicating that there is no syllable in the third slot counting from right to left. The loss of mo ras in the penultimate syllable should render that syllable no longer optimal if WSP, a presumed cons traint for Classical Lati n, is still active. Yet the tonic syllable, S2, is light in over 85% of the outputs across languages. Faithfulness to the prosodic head has been shown to be a high ranking constraint for Ibero-Romance (Catalan, Wheeler 2007; Castilian, Lle 2003) yet there is no evidence of retractio n of accent to the heavy initial syllable in outputs of the type HC.L.L from original (HC)('HC) or (HC)('HV). 168

PAGE 169

The new prosodic tem plate that has emerged fr om the original input form with two heavy syllables at the left word edge is (HC)('L.L) The near coincidence of lines for the three languages is notable. Figure 4-8. Comparison of distri bution of heavy/light in S3 and S2 from first declension trisyllables with penultimate and antepenultimate heavy syllables (n=75). A possible explanation for persistence of the initial HC syllable is found in a study of Old Spanish vowel deletion. Here Lle (2003) has util ized various sets of OT constraints to account for syncope and acope with relation to prosodic structure including a left edge alignment constraint, ALIGNL. The alignment constraint, ALIGNL, favors retention of the heavy first syllable because it requires the left edge of the wo rd to be aligned with a foot. The constraints adapted from Lle (2003, 256-257) are listed below. (4.14) FTTROCHEE: Feet are left-headed (either syll abic or moraic trochees). (Lle 2003) (4.15) STW: A stressed syllable is bimoraic. (Lle 2003) 100.0 % 90.0% 80.0% 70.0% 60.0% 50.0% 40.0% 30.0% 20.0% 10.0% 0.0% L NH L S2 S3 H 90.7% 9.3% 1.3% 20.0% 77.7% POR 89.3% 10.7% 0.0% 20.0% 80.0% CAS 85.3% 14.7% 0.0% 16.0% 84.0% CAT 169

PAGE 170

(4.16) ALIGNL: All prosodic words must have their left edge aligned with a foot. (Lle 2003) (4.17) HEADMAX: A stressed element in the input must have a stressed element as its output correspondent. (Lle 2003) (4.18) MAXMORPH: An input morpheme has a corres pondent in the output. (Lle 2003) In the analysis of three-syllable words with penultimate stress, Lle proposes that the pattern wsw (weak-strong-weak) can have one of two outcomes: wsw or ws that is, continuance of penultimate accent, or in the case of loss of final vowel (indicated by apostrophe but not seen in the first declension), ultimate accent. The weak/strong designations are not based on syllable weight but rather the susceptibil ity of a vowel to deletion. Lle s definition of trochaic foot admits both syllabic and moraic trochees (2003, 2 56). Additional prosodic contraints include a constraint regarding quantity sensitivity, ST W; however, it is low ranking. The faithful observance of HEADMAX has already been mentioned. In the case of the first declension trisyllables there is only one case of retracti on of the accent to the first syllable that does not involve vowels in hiatus, the outcome of Lat. m dulla, ae, f. marrow in Castilian where both mdula and medula appear in the current edition of DRAE (s.v. mdula), cf. popular meollo from a supposed VLat. medullum Retention of final /a/ is achieved through the MAXMORPH constraint as it is a cl ass and gender marker. Selecting items from the data set reflected in Figure 4-8, it is po ssible to show how OT constraints can generate placement of accent when the penultimate is/is not heavy. Lles definition of a trochaic foot (2003, 256) allo ws maximum flexibility and includes the uneven trochee, that is, all of the following would be possibl e realizations of a trocha ic foot: ('L.L), ('H), and ('H.L). The ALIGNL constraint favors systems with itera tive parsing beginning at the left word edge. In order to fulfill the constraint tris yllables with penultimate accent must be able to 170

PAGE 171

construc t a foot to the left of the trochee that constitutes the head foot. Preservation of the initial syllable when it is heavy, that is HC, satisfies ALIGNL. However, use of the constraint set listed in (4.14) to (4.18) presents a problem for selection of the optimal candidate when the input is (HC )('HC). In Table 4-14 (after Lle 2003, 259), both candidates A and B appear to satisfy all of the constraints but represent different parsing results for Latin cisterna, ae f. cistern16 which has identical output in all three languages. Table 4-14. Tableau for output of first declension trisylla ble with penultimate accent on HC type syllable CISTRNA FTTROCH HEAD MAX MAX MORPH STW ALIGNL a. (cis)(tr.na) b. (cis)(tr)na c. cis(tr.na) *! d. (cs)(ter.na) *! e. (cis.tr)na *! In the case of both candidates A and B the head foot is faithful to input and is trochaic; no morphemic elements have been lost (in particular the declension class/gender marker, a); and the left word edge coincides with a foot boundar y. Only the presence of a parsing constraint would cause candidate A to be selected over B. Lle (2003, 262) utilizes such a constraint in cases of syncope where a configuration s1s2 is preferred over S1WS2 (s/w indicate strong/weak syllables; numerals designate primary and sec ondary accent; upper and lower case distinguish Latin from Old Spanish). Two modifications will be made in Table 4-14 to demonstrate how the OT tableau can select th e desired output for words like cisterna as well as Lat. form ca The first is the inclusion of PARSE. 16Cisterna is attested for Castilian in CORDE s.v. cister na, ca. 1250 (Real Academia Espaola) and for Portuguese in Davies and Ferreira, Corpus do portugus s.v. cisterna, 15th century. For Catala n, the word is attested also in the 15th century according to DCBV, s.v. cisterna. It is likely to have been in continuous use in medieval Latin and early Romance, although perhap s displaced in some geographic areas in Spain by Arabic aljibe 171

PAGE 172

(4.19) PARSE: Syllables must be parsed into feet (cf. Prince and Smolensky 2004 [1993], 65 n. 36) A second constraint is needed to allow candidates that look like C, cis(tr.na), in Table 414 to be selected when the ante penultimate syllable is light and cannot be parsed into a foot. This could be accomplished by adding an alignment constraint to prefer the right word edge. (4.20) ALIGNR: All prosodic words must have th eir right edge ali gned with a foot. ALIGNR in Table 4-15 now rejects Candidate B b ecause it has an unparsed final syllable. Although PARSEis a low ranking constraint there is no compelling reason to render it inoperable here by assuming extrametricality for the final syllable. The preferred duple rhythm, which seems to ignore syllable weight, requires the presence of the final, unaccented syllable. Table 4-15. Tableau for output of first declension trisyllable with penultimate accent on HC type syllable with PARSEand ALIGNR constraints CISTRNA FTTROC H HEAD MAX MAX MORPH STW PARSEALIGNR ALIGNL a. (cis)(tr.na) b. (cis)(tr)na *! c. cis(tr.na) *! d. (cis)(trn) *! e. (cis.tr)na *! If preservation of the HC initial syllable in 80% (84% in Catalan) of the cases in figure 413 can no longer be attributed to AlignL it is necessary to look elsewhere for an explanation. Precisely because the heavy syll able occurs in word initial position the concept of positional faithfulness is attractive. If one labels the syllables of a first declension trisyllable with penultimate primary accent in gradient fashion with 1 to indicate pr imary accent, 2 secondary accent, and 3 no accent, the configur ation would be 2-1-3. The word initial syllable enjoys a 172

PAGE 173

higher deg ree of prominence because of its position rather than its weight17. There is ample evidence that the modern Ibero-Romance languages do indeed have secondary accent (Cabr and Prieto 2005, Serra 1997, Wheeler 2004; Daz Campos 1999, Harris 1991, 1992, Roca 1986, 1999, 2005; Abaurre, Galves, Mandel and Sandalo 2003, DAndrade 1997, and Vigrio 2003), although there is disagreement as to whether or not assignment of accent other than the primary accent is lexical or postlexical. Figure 4-9 shows that a syllable in word initial position (row labeled wd[H) is more likely to retain its moraic count in the case of a trisy llable than in a disyllable. Cases that retain the coda in the syllable with primary accent corres pond to the row labeled 'H, represented by the light gray line. The notable drop in this line indicates that the temp late (H)('L.L) is preferable to (H)('H.L) or (H)('H)L. 0 20 40 60 80 100 wd[H 66.163.663.683.377.877.8 'H 66.163.663.661.144.433.3 CATCASPORCATCASPOR 2 SYLL 3 SYLL Figure 4-9. Percentage of heavy syllables in word initial syllable and tonic syllable from first declension ('HC) and (HC)('HC). 17 Smith (2005, 49-52) discusses a series of M/str constraints (Markedness constraints th at make specific reference to strong position) that include HEAVY, that is, in strong positions a heavy syllable is preferable to a monomoraic syllable. The prominence of word initial position according to Beckman (1998, 52-57) is seen through both psycholinguistic and phonological evidence. The phonolog ical evidence, as Smith also suggests, often comes from immunity to well formedness constraints for segments in pr ivileged position. In the case under study the failure to repair the NoCoda violation of the initial HC syllable is a clear example of positional strength superseding the ideal open syllable type. Whether or not the initial syllable need be regarded as heavy is open to interpretation. In a study of Kashmiri stress Morn (2000, 367) notes that CVC syllables are only counted as heavy if it is the best locus of stress within the word. Otherwise it is considered to be light. 173

PAGE 174

In addition to the preferable m etric outcome, that is avoidance of an uneven trochee or an unparsed syllable, the higher rate of retention of coda consonant in the initial, unaccented syllable of trisyllables may indi cate that accented syllables have other measures of prominence such as vowel quality and/or quantity and do not rely on moraic c ount alone to place a syllable in relief or give it prominence. A similar high rate of preservation of coda consonants in the initial (unaccented) syllable is also seen in input ty pe (HC)('HV) where the coda consonant is retained in initial HC input syllables as follows (n=57): Catalan 48 (84.2%), Castilian 46 (80.1%), and Portuguese 39(68.4%). Portuguese also retains moraic count in resulting nasal vowels in an additional 6 (10.6%) cases; this yi elds a total of 45 heavy syllables (78.9%) for Portuguese. The efficacy of the OT constraints presente d above is demonstrated in Table 4-16 with the example Lat. form ca, ae, f. ant.> Cat., Old Cast., Port. formiga18. It is anticipated that selection of the output of form ca that is faithful to original place of accent, even when the accented syllable has lost a mora, will be faci litated by the low ranking of STW and high ranking of HEADMAX. By requiring alignment of a foot with the right word edge only two accentual patterns will be favored because they produce no alignment violations, penultimate accent 'L.L]wd and ultimate accent 'H]wd. Thus, in Table 4-16, candidates D and E are rejected although they satisfy STW. The only advantage A offers over B is that there are no unparsed syllables. Loss of a mora from input to output(HC)('HV) (HC)(L)L does not prevent assignment of primary accent to the light syllab le because STW ranks below the higher level constraints that require faithfulness to foot type and position of head foot. 18Formiga is attested for Castilian, Portuguese, and Catala n in the same sources cited above (s.v. formiga) with similar results for first date of attestation although Catalan formiga is documented for a slightly earlier period, 14th century. Again, because of th e popular nature of this word it was most likely used continuously in the trajectory from Latin to the modern languages. In modern Spanish the initial /f/ has been lost following aspiration to /x/ or /h/. 174

PAGE 175

Table 4-16. Tableau for output of first declensio n trisyllable with penultimate accent on HV type syllable with PARSEand ALIGNR constraints FORMCA FTTROCH HEAD MAX MAX MORPH STW PARSEALIGNR a. (for)('mi.ca) b. for('mi.ca) *! c. (for)( 'mic) *! d. ('fr)(mi.ca) *! e. ('for.mi.ca) *! f. (for. 'mi)ca *! The optimal output of Table 4-16, (H)('L.L), as a legitimate prosodi c pattern for threesyllable wordsin Lles terms wsw (weak-stron g-weak)is further reinforced by the outcome of two other patterns: (HV)('HV) and (L)('HV) . In these cases as well, the majority outcome is penultimate accent on a light syllable which must then combine with the final light syllable to form a syllabic, rather than moraic, trochee. The low ranking of PARSEwill still select a candidate with penultimate accent (for firs t declension nouns) even if it is not possible to parse the first syllable. The c onfiguration L('L.L) satisfies ALIGNR although there is now an unparsed initial syllable. Table 4-17 shows the output of the ( HV)('HV) and (L)('HV) templates. Syllables are counted from right to left; S2 is the penultimate and S3 the antepenultimate. The column heading N indicates that there is no syll able in that position. These exceptions arise primarily through the loss of a syllabic nucleus to resolve hiatus. The hiatus may be original or secondary, resulting from the loss of an inter vocalic consonant. Creat ion of a diphthong to resolve the hiatus is more likely to occur in Cas tilian than in either Catalan or Portuguese. As will be seen under discussion of antepenultimate accent patterns, creati on of penultimate accent by appropriating a syllabic nucleus as onset of a rising diphthong continues in Spanish today in 175

PAGE 176

cases such as perodo p eriod with alte rnate pronunciation periodo (penultimate accent); both forms appear in DRAE, s.v. perodo/periodo. Table 4-17. Distribution of heavy/light syllables in outcomes from first declension trisyllable patterns (HV)('HV) and (L)('HV) S3 S2 H L N H L CAT 4 (5.7%) 61(87.1.6%) 5 (7.1%) 2 (2.9%) 68 (97.1%) CAS 4 (5.7%) 57 (81.4%) 9 (12.9%) 2 (2.9%) 68 (97.1%) POR 4 (5.7%) 59 (84.3%) 7 (10.0%) 2 (2.9%) 68 (97.1%) n=70 The coincidence in outcomes of (HV)('HV)< L> and (L)('HV) patt erns across the three languages is displayed in graphic form in Figure 4-10. The results show the expected merger of L and HV syllable types and faithfulness to the head foot. The new prosodic template is L('L.L) in contrast with the (H)('L. L) pattern that emerged from original (HC)('HC) and (HC)('HV). The new L('L.L) template is probl ematic in that it does not allow for alignment with the left word edge since it is impossible to build a well-f ormed trochee over a single light syllable. Yet, this pattern is the clear choice in the 70 words that are reflected in Table 4-17 and Figure 4-10. With few exceptions the primary accent remains on the penultimate syllable, although there is one case of ultimate accent, Portuguese pag (
PAGE 177

H S3 L N H S2 L 0.0% 20.0% 40.0% 60.0% 80.0% 100.0% CAT 5.7%87.1%7.1%2.9%97.1% CAS 5.7%81.4%12.9%2.9%97.1% POR 5.7%84.3%10.0%2.9%97.1% H S3 LN H S2 L Figure 4-10. Percentage distribution of heavy/lig ht syllables in outcomes from first declension trisyllable patterns (HV)('HV) and (L)('HV) The first four cases of original hiatus in Table 4-18 involve words of Greek origin which entered Ibero-Romance through Latin. Secondary instances of hiatus result from lenition and total loss of intervocalic /k/ or /g/ and, in Portuguese, loss of intervocalic /l/. In these cases the vocalic nuclei in contact form falling diphthongs; or, in some cases the atonic vowel is elided. Table 4-18. Resolution of hiatus in output of Latin first declension trisyllables Original Hiatus Secondary Hiatus Latin input CAT CAS POR Latin input CAT CAS POR ta iota jota jota jota v g na sheath bina vina binha hyaena hyena hiena hiena c qu na kitchen cina d aeta diet dieta c l bra snake cbra m g a magic mgia s gitta arrow seta r na ruin ruina r g na queen rina rina rinha qu ta still hour queda queda queda t ra tiara tiara tiara Shaded cells indicate change in locus of accent; the accented syllable is indicated by an acute accent mark. With the exception of m g a in cases of original hiatus, the hiatus occurs between 177

PAGE 178

a high unaccented vowel in the in itial syllable and a long vowel or diphthong in the accented syllab le. In the case of jota the repair strategy in volves creating an onset by forming a palatal glide from the initial unaccented vowel. Glide formation followed by fortition of the palatal onset is uniform across languages (Velarization of the palatal glide is a later development in Castilian). Only Portuguese does not de velop a glide in the initial syllable of hiena The word is given a trisyllabic pronunciation in Gran diccionario espaol-portugus (s.v. hiena). This is also the case for both Catalan and Portuguese dieta, magia, ruina which are all trisyllabic with penultimate accent in contrast with disyllabic Ca stilian forms with a rising diphthong instead of vowels in hiatus. Loss of both the labializati on of latin /kw-/ and the following vowel nucleus characterize the output of qu t, queda which is uniform across languages and contrasts with learned quiet(o)/a19 also present in the three languages. In this case as well Castilian stands apart from the other two languages by having a rising dipht hong, [je] rather than [i.'e]. In the case of magia the undoing of hiatus has actually resulted in displacement of the accent to the initial syllable (originally antepenul timate). This constitutes a violation of the constraint HEADMAX because the output form does not have a correspondent for the accented vowel of the input form. Yet this pattern is pervasive in the cases of secondary hiatus. According to Lloyd (1987, 320) the elimination of hiatus becomes an agent of change in otherwise inexplicable developments in Ibero-Ro mance such as glide in sertion in Old Spanish trayo (modern traigo I bring) < Lat. tr h which should have yielded traho. Coalescence of vowels is an alternate strategy as seen in Portuguese lendo reading compared with Spanish leyendo < Lat. l gendu(m) The cases of accent shift seen in Table 4-27 involve displacement of the accent from a high accented vowel to a preceding lower vowel as in beina, vaina, bainha; 19 DCVB (s.v. quiet) also lists variant forms quet, queta for Catalan. 178

PAGE 179

reina, rainha (cf. Lloyd 1987, 320); or, coalescence of vow els subsequent to the stage seen in the f orms just cited, as in cobra, seta The high vowel in th e initial syllable of cuina is somewhat unexpected. The DCVB gives ['kuj.n ] as the most frequent pronunciation with ['kwi.n ] in Northern Catalan and ['kwi.na] in Valencian. Although both Castilian and Portuguese have a mid back vowel in the resulting trisyllabic forms ( cocina, cozinha ) the high back vowel in Catalan could be the result of assimilation to the he ight of the palatal glide. Additionally, [u] is the usual reflex of /o/ and // in atonic syllables. It should also be noted that rising diphthongs are historically less frequent in Catalan alt hough Cabr and Prieto (2004, 116) indicate that glide formation is evident as early as the seventeen th century and today regularly competes with maintenance of hiatus in Central Catalan (nor theastern Catalonia including Barcelona). What advantages does formation of a diphthong offe r? It has been stated that the emerging prosodic template is L('L.L). The head foot is a ligned at the right word e dge but not at the left. Lle (2003, 256-257) has proposed ALIGNL as an active constraint during the formative period of Castilian and it works well as an explanation fo r the preference of (H)(' H.L) over L('H.L) but creates other problems as discussed above such as the inability to select (H)('H.L) over (H)('H)L without a parsing constraint. Jacobs (2003a, 275-176) has proposed using both W/L and W/R (word left/right) alignment constraints. W ith both W/L and W/R as active constraints, monosyllables and disyllables (assuming the acceptabi lity of uneven trochees) are always able to satisfy both constraints. The case of trisyl lables with penultmate accent becomes more problematic. However, OT allows for violati on of certain constraints and will designate as winning candidate the form which does not violate critical constraints an d violates the fewest possible lower level constraints. 179

PAGE 180

The case of Lat. r g na is interesting in that one finds two dif ferent outcomes in Catalan, historical rehina or rena presumably [r .'i.n ], and modern reina ['rej.n ] (DCVB, s.v. reina). The historical form is correctly selected in Ta ble 4-19. The modern form candidate C, violates HEADMAX. The modern Ibero-Romance languages all show strategies to repair vowel/vowel sequences, including dislocation of original primary stress. Th e example of modern Spanish perodo> periodo with a change from antepenultimate to penultimate stress is well known. Table 4-19. Tableau for output of first declension trisylla ble with penultimate accent on HV type syllable with PARSEand ALIGNR constraints RE.'GI.NA FTTROCH HEAD MAX MAX MORPH STW PARSEALIGNR a. re('i.na) b. re('in) *! c. ('rej.na) *! d. (re.'i)na *! For Portuguese the situation is somewhat mo re complex. Mateus and dAndrade (1998) maintain that in the case of falling diphthongs both vowel and glide belong to the syllable nucleus. An argument to sustain this statemen t is the fact that, in nasal diphthongs, both segments are nasalised by the projection of the na sal autosegment to the nucleus. With regard to rising diphthongs as seen in the case of Spanish [.' j ] the authors note that although in informal speech the unstressed vowels /i/ and /u / before a vowel are reduced in duration and intensity they differ from the onset glides found in true rising diphthongs in Spanish. This is evident in the lack of spread of th e feature [+nasal] to the glide in criana child [ 'k j .s ] in contrast with the falling diphthong [ j ] in me mother [ m j ]. Curiously, in the case of Portuguese perodo the word is syllabified in Michael is (1998-2007, s.v. pe rodo) as pe.ro.do but peridico is syllabified as pe.ri..di.co (s.v. peridico). Fernandes (1996, s.v. perodo) 180

PAGE 181

provides a pronunciation of [p 'r i du] in a modified IPA transcription where a breve is used over a high vowel to indicate a glide. The preferen ce for a falling glide is consistent with the phonemic status of falling glides in Portuguese. Rising glides, on the other hand, emerge as free variants of vowels in hiatus in certain registers. The dictionary entry suggests, then, that the trisyllabic pronunciation is now the norm and no longer alternates with a tetrasyllabic pronunciation with vowels in hiatus A similar situation obtains in Catalan where rising diphthongs except after velar consonants emerge only as variants of vowels in hiatus. On the other hand, Cabr and Prieto (2004, 115-116) note that In Catalan falling sonority vocoid sequences present a rather different behavior than rising sonority sequences. Intramorphemic sequences are almost indefectibly pronounced with a diphthong (cf. mai ['maj] never, peu ['] foot, noi ['] boy, bou [' ] ox). The same authors also demonstrate how an ordered set of well formedness constraints favor glide formation. They include the familiar Onset (A syllable must have an onset) as well as the two related to peak prominence (Cabr and Prieto 2004, 133-134): (4.21) *M/V[-high]: A vowel at the margin of a syllabl e must not be [-high]. (Cabr and Prieto 2004) (4.22) *M/V[+high]: A vowel at the margin of a syllabl e must not be [+high]. (Cabr and Prieto 2004) The first constraint gives preference to a high vowe l in the margins of the syllable, that is, as coda or onset. The second constraint disfa vors diphtongs in general whether rising or falling. Table 4-20 (adapted from Cabr a nd Prieto 2004, 134) shows the selection of hiatus for the first example, ocens oceans, and glide formation for the second, avins airplanes. 181

PAGE 182

Table 4-20. Tableau for selection of hiatus or rising diphthong in Catalan oce'an+s *M/V[-high] ONSET *M/V[+high] a. (o.ce)('ans) b. o('ceans) *! avi'on+s a. (a.vi)('ons) *! b. a('vjons) The additional constraints below (Cabr and Prieto 2004, 136-139) reflect the prosodic pattern of Central Catalan at word level and, in particular, favor c onfigurations that result in the construction of a foot aligned with the left wo rd edge. Word initial prominence is a crosslinguistic phenomenon as discussed above. (4.23) *LAPSE: Pre-tonic vowels must be parsed in one foot. (Cabr and Prieto 2004) (4.24) MAXINIT: A word-initial mora in the input must be present in the output. (Cabr and Prieto 2004 (4.25) FT-LEFT: The left edge of a word must be aligned with a foot edge. (Cabr and Prieto 2004 The relative ranking of these constraints is *LAPSE >> MAXINIT / FT-LEFT >> ONSET >> *M/V[+high]. The two constraints that favor word initial position, MAXINIT and FT-LEFT, are unranked with respect to each other. Table 4-21 shows how these constraints can also select the optimal candidate diachronically. The absence of HEADMAX permits the selection of candidate A. This faithfulness constraint has been repl aced by one which favors prominence in word initial position, MAXINIT even if it means a departure from the pr imary accent of the input form. The historical form rena corresponding to candidate B, has an unparsed initia l syllable and consequently fails to align a foot with the left edge of the word. Inclusion of a parsing constraint would also exclude candidate B as long as it ranked above *M/V[+high]. 182

PAGE 183

Table 4-21. Tableau for selecti on of falling diph thong in Catalan RE.'GI.NA *LAPSE MAXINIT FT-LEFT ONSET *M/V[+high] a. ('rej.na) b. re('i.na) *! First Declensions Trisyllables wi th Antepenultimate Accent Nouns with antepneultimate accen t are characterized by light sy llables in both ultimate and penultimate syllables as in qu la, ae f. eagle. This is consistent with the Latin accent rule in which the final syllable is extrametrical and pl acement of the primary accent falls on the first bimoraic sequence to the left of the ultimate sylla ble. Because the word initial syllable carries primary accent by default its moraic count is no t a determining factor. The distribution of prosodic templates of first declension trisyllables is show in Table 4-22. Table 4-22. Prosodic templates of first declension trisyllables with antepenultimate accent Number Percent Type Example 64 39.5% ('L.L) r m ra, ae, f. hinderance 48 29.6% ('HV)L p g na, ae, f. leaf, page 50 30.9% ('HC)L vict ma, ae, f. beast for sacrifice, victim n=162 As with the previous data sets it can be predicted that the first two types will merge with a few exceptional cases of preservation of the diphthong /aw/. There are four such cases corresponding to learned or semi-learned words present in all three la nguages (Spanish and Portuguese with acute accent and Catalan with grave): clusula/clusula, nufraga/nufraga, nusea/nusea, nutica/nutica The remaining cases of ('HV)L input become indistinguishable from original ('L.L). It is possible, therefore to pro pose an input set of the type ('L.L) consisting of 112 items, only 4 of which are faithful to the initial syllable of original ('HV)L. With an input form of the shape ('L. L) it is apparent that ALIGNL/R cannot be satisfied. However, in Table 4-19 it is shown to be a relatively low ranking constraint. The final syllable cannot be elided because of MAX MORPH and both HEADMAX and MAXINIT 183

PAGE 184

protect the initial, accent bearing syllable. The only vulnerable sy llable, then, is the second from right word edge or the middle syllable. The Appendix Probi, discussed in Chapter 3, provides many examples of loss of syllabl e nucleus in this position. In the data set consisting of original (' HV)L (n=48) there are 10 cases of syncope common to all three languages with the exception of the output of v p ra, ae f. viper which in Castilian and Portuguese has three syllables. Ho wever, the lack of co rrespondence of the vowel in the input to the vowel of the output suggests su bsequent restoration of a lost vocalic nucleus. The resulting initial syll ables in Table 4-23 are not all heavy syllables (Unshaded cells are 'L.L; light gray cells with broken outline are 'HV.L; a nd dark gray cells with so lid outline are 'HC.L.). If the goal of syncope is to repair prosodic tr apping (as discussed in Me ster 1994) it must be assumed that the variable disyllables that result offer a metrical advantage. The unshaded cells in Table 4-23 clearly indicate that the majority outcome is a sequence of two light syllables or 'L.L, a well-formed trochee aligned with both left and right word edges. Table 4-23. Outcomes of syncope in first declen sion trisyllables with antepenultimate accent: ('HV)L type Latin Catalan Castilian Portuguese a. cpla, ae, f. tie coble copla copla b. f bla, ae, f. story faula habla fala c. f m na, ae, f. female fembra hembra fmeaa d. *pul ca< p lex, cis, m. flea pua pulga pulga e. r gla, ae, f. rule, model regla regla regra f. r gla, ae, f. ruler, staff rella reja relha g. s m ta, ae, f. path senda senda senda h. t gla, ae, f. roof-tile teula teja telha i. tr g la, ae, f. javelin tralla tralla tralha j. v p ra, ae, f. viper vibra vbora vbora a Portuguese regularly loses intervocalic /n/. The resulting vowels in hiatus are rendered as a rising diphthong: [j ]. Consequently, Portuguese fmea is also disyllabic although through a different process However, if the emerging trochee in Ibero-Romance is syllabic rather than moraic, then 'HV.L as well as 'HC.L are also well-formed trochees. There is little ev idence here to support 184

PAGE 185

STW (A stressed syllable is bimoraic) as an activ e constraint. Even in Classical Latin the one accentual pattern that regularly violates this constraint is the one under consideration, antepenultimate primary accent in trisyllables with all light syllables. Assuming that the trisyllables under question are primarily ideal CV syllables, then a sequ ence CV.CV.CV would result in CV[CC]V after syncope. The sequence in brackets could be syllabified as C.C or, dependent on the nature of the segments, both consonants could form the onset of the following syllable. The heterosyllabic sequence would pro duce a heavy initial syll able which would offer the advantage of satisfying MAXINIT by replacing the lost mora of the original HV syllable with a coda consonant (or glide). Howeve r, this is rarely the outcome as is seen in Table 4-24. This display of data shows that for ('HV)L input the majority outcome is the more faithful trisyllable, ('L.L)L, rather th an the innovative disyllable ('L.L). The trisyllabic output incurs a parsing violation (PARSE) as well as a violation of ALIGNR but these are both low ranking constraints as demonstrated in Table 4-19 above. Table 4-24. Output of fi rst declension ('HV)L ('HV)L 3 SYLLABLE 2 SYLLABLE HC HV L Subtotal HC HV L Subtotal CAT 0 (0.0%) 4 (8.3%) 26 (54.2%) 30 (62.5%) 3 (6.3%) 3 (6.3%) 12 (25.0%) 18 (37.5%) CAS 0 (0.0%) 4 (8.3%) 20 (41.7%) 24 (50.0%) 3 (6.3%) 0 (0.0%) 21 (43.8%) 24 (50.0%) POR 0 (0.0%) 4 (8.3%) 21 (43.8%) 25 (52.1%) 2 (4.2%) 7 (14.6%) 14 (29.2%) 23 (47.9%) n=48 The higher percentage of trisyl labic results in Catalan is due primarily to the preservation of vowels in hiatus in the input form. Portuguese is less inclined to preserve hiatus but does so to a far greater extent than Castilian. This accoun ts for the unusually high percentage of ('L.L) output for Castilian. Of the 21 cases of of ('L.L) in Table 4-24, 9 of them can be attributed to glide formation resulting in a disyllable as in sepia which is trisyllabic spia in both Catalan and 185

PAGE 186

Portuguese. If these cases are subtracted from th e ('L.L) results for Cast ilian the outcome would be idential to Catalan, 12 cases or 25% of th e total. This disparity by language is shown graphically in Figure 4-11, especially in the results of the 2-syllable output. 0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% CAT 0.0%8.3%54.2%6.3%6.3%25.0% CAS 0.0%8.3%41.7%6.3%0.0%43.8% POR 0.0%8.3%43.8%4.2%14.6%29.2% HCHVLHCHVL 3 syllables 2 syllables Figure 4-11. Comparison of 3 syllable and 2 sy llable results of first declension ('HV)L input. In Portuguese the higher percenta ge of ('HV.L) is attributable to nasal vowels, treated as bimoraic in this analysis, which result from loss of intervocalic /n/. Vocalization of syllable final /b/ in Catalan accounts for the 3 instances of ('HV.L) for th at language. Two examples, faula talk and teula tile are seen in Table 4-23. The distribution of outcomes from ('L.L) would be expected to parallel that of ('HV)L with the exception of word initial HV sylla bles in the trisyllabic results. This subset contains 64 words but the table s hows only a total of 59. A small set of words of Greek origin, learned and semi-learned in nature, show a preference for penultimate accent and have been excluded from Table 4-25: patena paten, patera flat dish, tisana tisane, viola viola, as well as the outputs of t n brae, rum f. darkness which are uniformly penultimate. Lewis and Short (s.v. t n brae) marks the penultimate vowel with bot h breve and macron. It has also been 186

PAGE 187

suggested th at such words may never have been pronounced with antepenultimate accent, perhaps because the consonant cluster was not viewed as tautosyllabic (Lloyd 1987, 115). Table 4-25. Output of fi rst declension ('L.L) ('L.L) 3 SYLLABLE 2 SYLLABLE HC HV L Subtotal HC HV L Subtotal CAT 1 (1.7%) 0 (0.0%) 40 (67.8%) 41 (69.5%) 4 (6.8%) 4 (6.8%) 10 (16.9%) 18 (30.5%) CAS 1 (1.7%) 0 (0.0%) 29 (49.2%) 30 (50.8%) 4 (6.8%) 0 (0.0%) 25 (42.4%) 29 (49.2%) POR 0 (0.0%) 0 (0.0%) 34 (57.6%) 34 (57.6%) 2 (3.4%) 4 (6.8%) 19 (32.2%) 25 (42.4%) n=59 The results of Table 4-25 are similar to t hose of Table 4-24. Again, Castilian has an unusually high number of disyllables in comparison to the other two languages. This number is inflated by the creation of rising diphthong to resolve hiatus which occurs in 10 of the 25 cases that emerge as ('L.L). There is one other unexp ected result, the presence of one 3-syllable word with heavy initial syllable (HC) in both Ca talan and Castilian. The outcome of Lat p l la, ae f. pill in both languages has an epenthetic consonant: Cat. pndola and Cast. pldora This suggests that there may have been an alte rnate form in popular speech. Portuguese plula is completely faithful to the input form. The 2syllable outcomes do produce some heavy initial syllables although the hypothetical pr ogression presented above, CV.CV.CV CVC.CV, is not widely attested and few examples are cross-linguistic with the except ion of the often cited c l da (calda), ae f. warm water. However, as the entry in Lewis and Short (s.v. c l dus) indicates the disyllabic form was already well established in the Augusta n period (27 BCE-14 CE); it is cited in the Appendix Probi (see Appendix C, l. 53). It is very unlikely, th en, that the trisyllabic form was ever present in popular speech in Ibero-Romance. The combined results of input forms ('HV)L< L> and ('L.L) are shown in Table 4-26 and Figure 4-12. There is a high degree of retention of syllable count of the imput form across 187

PAGE 188

languages, ranging from a low of 50.5% for Castilian to a high of 66.4% for Catalan. The lower percen tage for Castilian is primarily the result of glide formation to re solve vowels in hiatus. Table 4-26. Output of first decl ension ('HV)L and ('L.L) ('HV)L and ('L.L) 3 SYLLABLE 2 SYLLABLE HC HV L Subtotal HC HV L Subtotal CAT 1 (0.9%) 4 (3.7%) 66 (61.7%) 71 (66.4%) 7 (6.5%) 7 (6.5%) 22 (20.6%) 36 (33.6%) CAS 1 (0.9%) 4 (3.7%) 49 (45.8%) 54 (50.5%) 7 (6.5%) 0 (0.0%) 46 (43.0%) 53 (49.5%) POR 0 (0.0%) 4 (3.7%) 55 (51.4%) 59 (55.1%) 4 (3.7%) 11 (10.3%) 33 (30.8%) 48 (44.9%) n=107 While reduction of syllables is more frequent in Castilian there are some cases that are unique to Portuguese and result from loss of intervocalic /l/ and /n/. W ith the exception of the language specific cases discussed above it can be seen that over all the th ree languages have very similar outputs from the two antepenultimate patterns that merge in Late Latin due to loss of vocalic quantity. Only the frequency and type of the 2-syllable outcomes vary by language, largely for reasons not related to word prosody. Castilian has no heavy initial syllables of the HV type (fal ling diphthongs or nasal vowels) yet it has a significantly larger count of 2-syllable words owi ng to the aversion to onsetless syllables. The HV initial syllables that emerge in Catalan are primarily the outcome of vocalization of coda consonants that are the product of vowel elision and in Portuguese they result from loss of /n/ in coda position with nasalization of the preceding vowel. Nasal vowels are treated as bimoraic. The final data set in this group of trisylla bic first declension n ouns corresponds to the template ('HC)L. The initial syllable, which will remain as the syllable with primary accent, is also expected to retain its bimoraic status favored by positional prominence (MAXINIT: A word-initial mora in the input must be present in the output) rather than a need to fulfill STW (A 188

PAGE 189

stressed syllable is bim oraic). Maintenance of syllable count is higher cr oss linguistically than for nouns in which the initial syllabl e is light (see Figure 4-12). 0.0% 20.0% 40.0% 60.0% 80.0% 100.0% CAT 0.9%3.7%61.7%6.5%6.5%20.6% CAS 0.9%3.7%45.8%6.5%0.0%43.0% POR 0.0%3.7%51.4%3.7%10.3%30.8% HCHVLHCHVL 3 syllables 2 syllables Figure 4-12. Percentage distri bution of outcomes of first decl ension ('HV)L and ('L.L). Again, Castilian has a higher incidence of reduction to two-syllables owing primarily to formation of rising diphthongs in the case of vowels in hiatus. This occurs in 12 of the 22 cases with (('HC.L) output. When the data in Table 4-27 is graphed (see Figure 4-13) it appears that the output forms across languages have disparate results. Once again Castilian shows a proportionately higher number of disyllables. These are of the type ansia longing, bestia beast, branquia gill, etc. which are invariably disyllabi c in Castilian. The normativ e pronunciation for Catalan and Portuguese is trisyllabic although su bject to change in some cases in informal register or as a function of speech tempo (Badia Margarit 1973, Cabr and Prieto 2004, Mateus and dAndrade 1998). In a few instances the standard dictionary form for Portuguese is disyllabic rather than trisyllabic. Maintenance of hi atus or glide formation changes the syllabic count but does not affect the structure of the initial accented syllab le; it remains as a heavy syllable: CVC.CV.V CVC.CV.V or CVC.CGV. 189

PAGE 190

Table 4-27. Output of firs t declension (' HC)L ('HC)L 3 SYLLABLE 2 SYLLABLE HC HV L Subtotal HC HV L Subtotal CAT 30 (60%) 0 (0%) 6 (12%) 36 (72%) 12 (24%) 0 (0%) 2 (4%) 14 (28%) CAS 18 (36%) 0 (0%) 8 (16%) 26 (52%) 22 (44%) 0 (0%) 2 (4%) 24 (48%) POR 13 (26%) 4 (8%) 11 (22%) 28 (56%) 11 (22%) 9 (18%) 2 (4%) 22 (44%) n=50 The other point of incongruity, represented in Figure 4-13, is the differing representation of initial syllables of th e type HV. These are present onl y in Portuguese and correspond to bimoraic syllables with a nasal vowel as in juna chufa, sedge, lana lance, ona ounce, una Within each group of 2-syllable and 3-syllab le outcomes the proportion of heavy syllables seems quite high. Disregarding for the moment the question of number of syllables and collapsing HC and HV into a single category, H, the output of ('HC)L is surprisingly uniform across languages. 0.0% 20.0% 40.0% 60.0% 80.0% 100.0% CAT 60.0%0.0%12.0%24.0%0.0%4.0% CAS 36.0%0.0%16.0%44.0%0.0%4.0% POR 26.0%8.0%22.0%22.0%18.0%4.0% HCHVLHCHVL S3 S2 Figure 4-13. Percentage distri bution of outcomes of first decl ension ('HC)L: Place and nature of the primary accent (n=50). 190

PAGE 191

The difference in distribution of heavy and light syllables, seen in Fi gure 4-14, is greatest between Catalan, the m ost conservative in terms of loss of coda segments, and Portuguese, the most innovative and least likely to have a coda consonant; however, the variance is only 10%. 0.0%20.0%40.0%60.0%80.0%100.0% CAT CAS POR L 16.0%20.0%26.0% H 84.0%80.0%74.0% CATCASPOR Figure 4-14. Percentage distri bution of outcomes of first decl ension ('HC)L: Heavy vs. light tonic syllables. Faithfulness to the initial heavy syllable is high across languages and is very similar to the retention of HC initial syllables from first de clension (HC)('HC) where the distribution by language is Catalan 83.3%, Casti lian 77.8%, and Portuguese 77.8%. On the other hand, the retention of coda consonants in word initial tonic syllables is much lower for the output of disyllables with an initial heavy syllable. For ('HC) the percentage of initial heavy syllables that result is 66.1% for Catalan, 63.6 % for Castilian, and 63.6% for Portuguese. Whether or not these differences are significant can be shown in a contingency table (Table 4-28) in which the output s of ('HC)L and ('HC) ar e compared. These two groups are selected for comparison because in both tonic syllable and word initial syllable coincide. The operative negative hypothesis for Ta ble 4-28 is: The ratio of heavy to light initial syllables in three-syllable words with antepenultimate accent is not equivalent to the ratio of heavy to light initial syllables in disyllables w ith penultimate accent. The expected percentages, then, are those 191

PAGE 192

that correspond to two-syllable wo rds. At the confidence level of p<.01 only the results for Catalan a re significant while those for Castilian and Portuguese are within the parameters of expected results. Catalan shows the greatest difference in distribution of heavy and light syllables when they are in initial position and bear the primary accent: 42 heavy (84%) and 8 light (16%), in contrast w ith an expected distribution of 66.1% and 33.9% or 33 and 17. Table 4-28. Contingency table for H/L outco me in Catalan, Castilian, and Portuguese for the accented syllable from ('HC)L and ('HC) Syllable Type Observed Expected Expected number number percentage CATALAN Heavy 42 33 66.1% Light 8 17 33.9% x2=7.2192, p-value is 0.0072 CASTILIAN Heavy 40 32 63.6% Light 10 18 36.4% x2=5.5555, p-value is 0.0184 PORTUGUESE Heavy 37 32 63.6% Light 13 18 36.4% x2= 2.1701, p-value is 0.1407 n=50 It has been previously established that positional prominence rather than locus of primary accent is more likely to positively affect preservation of moras in the initial sy llable. It is useful, therefore, to compare outputs of ('HC)L, (HC)('HC), and ('HC) (Figure 4-15). The first two sets have trisyllabic inputs but di ffer with regard to accentual pattern yet the coincidence of these two sets with regard to dist ribution of heavy/light wo rd initial syllables is striking. If prominence of the initial syllable is construed to be a type of secondary accent the issue of stress clash is inevitable when it comes to the output of words with penultimate stress from (HC)('HC) input. 192

PAGE 193

0.00% 20.00% 40.00% 60.00% 80.00% 100.00% HC1.L.L 84.00%16.00%80.00%20.00%74.00%26.00% HC.HC1.L 83.30%16.70%77.80%22.20%77.80%22.20% HC1.L 66.10%33.90%63.60%36.40%63.60%36.40% HLHLHL CAT CAS POR Figure 4-15. Percentage of heavy/light syllable s in word initial syllable resulting from first declension ('HC)L, (HC)('HC), and ('HC). Stress clash can be outranked by other constrai nts as demonstrated in Kenstowiczs study of English loanwords in Fijian. Fijian regularly assigns primary accent by building a trochaic foot at the right word edge but it is also faithfu l to the locus of accent of the input form, in this case English, where many polysyllabic words ha ve syllable initial accen t. The top ranking, inviolable constraints are those that define foot form (Trochai c, Ft-Bin) and alignment (AlignRt). The faithfulness cons traint that governs loanword phonology is Max-Stress. The relationship of Max-Stress to cons traints dealing with syllable quantity and rhythmic form is shown in the Hasse diagram in Figure 4-16 (adapted from Kenstowicz 2007, 323). Max-Stress | Dep-Mora Prosodic Prom-2 | Clash | Lapse Prosodic Prom-1 Figure 4-16. Hasse diagram of OT constraints related to primary accent in Fijian loanwords. 193

PAGE 194

Prosodic Prom inence emerges as two constraints to reflect the gradient of vowel prominence: V < 'V < 'V. The most prominent vowel is one wh ich is long and accent bearing; the least prominent is short and atonic. Creation of a prominent vowel entails a faithfulness violation (Dep-Mora) but failure to create one to corres pond to original locus of accent violates MaxStress: Prosodic Prom-2 disfavors addition of both stress and vowel length whereas Prosodic Prom-1 involves a one step increase in prominence. The OT tableau in Table 4-29 demonstrates that Prosodic Prom-2 outranks Clash. Thus, the preferred candidate conforms to the normal accentual pattern of Fijian in that it assigns an accent to the rightmost foot (left-headed and bimora ic) but deviates in that it does not respect the pattern of alternating stress in or der to comply with Max-Stress. Candidate A is rejected because it violates PP-2; the final syllable has moved tw o steps in the vowel prominence scale. Although Fijian admits unparsed syllables candidate C vi olates Max-Stress and is also rejected. Table 4-29. Tableau to assign accent in Fijian loanwords /koloni/ colony Max-Stress PP-2 Clash a. (klo)(n) b. (k)(lni) c. ko(lni) *! If one assumes that Latin trisyllables with penultimate accent are also marked for word initial prominence, despite the apparent stress clash, th en preservation of that prominence is simply another instantiation of a constraint of the HEADMAX type. HEADMAX could be redefined to include both primary accent and seco ndary or word initial accent. Four-syllable Nouns First declension nouns with four syllables re present various possibilities for a prosodic template. In the case of antepenultimate stress the ultimate and penultimate syllables are both light. The antepenultimate syllable is unprescribed for moraicity and can be either light or 194

PAGE 195

heavy; the same situation holds for the preantepenultimate syllable. As expected the pattern of penultimate accent requires the accent bearing sy llable to be bimoraic. The two preceding syllables can be either light or heavy. The po ssible patterns for penultimate and antepenultimate accent are shown in Table 4-30. Syllables are labeled S1 to S4 counting from right to left and the syllable with primary accen t is indicated by shading. Table 4-30. Prosodic templates for first declension four-syllable nouns S4 S3 S2 S1 A. Penultimate Accent HC HV L HC HV L HC HV L B. Antepenultimate Accent HC HV L HC HV L L L Four-syllable Nouns with Penultimate Accent The class of nouns corresponding to pattern A in Table 4-30 (p enultimate accent), as in c t racta ae f. waterfall, contains several relatively small su bsets, each showing the various permutations of S3 and S4. If one pursues th e notion of positional prominence, S4 and S2 are prominent syllables while S3 and S1 are less so. However, it has been shown that there are no instances of apocope for this declension class. Presumably, deletion of S1 is blocked by MAXMORPH, although the applicability of this constraint is questioned in Chapter 5 in discussion of declensions 2, 3, and 4. Analysis of the development of this group of nouns begins with those in which S2 is a heavy syllable with coda consonant. There are 9 possible permutations for four-syllable words in which the last two-syllables are HC1.L. Not all patterns are represente d and expected mergers caused by loss of vowel quantity are i ndicated by arrows. Patterns at tested in the data set have a checkmark in the list in Table 4-31. 195

PAGE 196

Table 4-31. Possible heavy/ligh t syllable config urations in four-syllable first declension nouns with penultimate accent a nd HC1 penultimate syllable a. HC.HC.HC1.L d. HV.HC.HC1.L g. L.HC.HC1.L b. HC.HV.HC1.L L.HV.HC1.L e. HV.HV.HC1.L h. c. HC.L.HC1.L f. HV.L.HC1.L i. L.L.HC1.L = attested pattern For pattern A, with heavy syllables in S4, S3, and S2 there is only one case, a learned word from Greek via Latin, present in all three languages: exorcista, ae, m. exorcist. The suffix -ista is a productive suffix in the modern language s as well. The patterns represented as B and C can be treated together because the loss of vowel quantity would have made them identical in Late Latin. Only occasionally is the diphthon g /aw/ preserved, which does not occur in this data set. As with pattern A, patterns B and C are virtually unattested. There is only one case that originates from pattern C, HC.L.HC1.L, Late Latin offerenda offering (in Ducange 1710, s.v. offerenda). It is also the one of the few cases of syncope in this subgroup. The fact that vulnerable S3 is preserved in Portuguese suggests that the word entered Ibero-Romance as four syllables rather than three. Furthermore, a variant form oferena is given in DCVB (s.v. ofrena) as well as the more usual reflex in Catalan, ofrena. The Castilian form, ofrenda, also shows syncope. Because S2 is a light syllable trapped between two heavy syllables which can constitute moraic trochees in a configuration (HC)L('HC)L it is a likely candidate for elision. Furthermore, the coda consonant of S1 is able to form an onset cluster with the rhotic that is the onset of S3. The combined set of patterns H and I, L/ HV.L/HV.HC1.L contains 14 items in which all initial syllables are light. Th e nature of each syllable is summarized in Table 4-32. The emerging pattern in Table 4-32 is L.L.HC1.L, a lthough there are some cases (16.7% for Catalan and 33.3% for Castilian and Portuguese) in whic h the tonic syllable is also light producing 196

PAGE 197

L.L.L1.L. The prim ary cause of loss of moras is degemination of ss-. This orthography indicates a qualitative difference when compared with s rather than a quantitative difference, that is, two phonemes that contrast in voice, /s/ and /z/. In Castilian the vo icing contrast is lost by the sixteenth century (Penny 2002, 78). Table 4-32. Nature of first three syllables in outcomes of first declension L.L.HC1.L and L.HV.HC1.L inputs S4a S3 S2 HC HV L HC HV L HC HV L CAT 0 0 14 0 0 14 11 0 3 CAS 0 0 12 0 0 14 9 0 5 POR 0 0 14 0 0 14 3 6 5 n=14 a The total count is less than 14 for Castilian because there are two words with only three syllables, the product of glide formation: matriarca, patriarca Before proceeding to the optimal parse of a four-syllable sequence such as c t racta ae f. waterfall, it is important to establish that both Castilian and Portuguese display greater restrictions with regard to coda consonants than Catalan, and Portuguese is the more restrictive of the two. These various language-specific constrai nts can be captured by an umbrella NOCODA constraint with a subscript language designation. The constr aint relies on the fulfillment of languagespecific sonority sequencing and phonotactic constrai nts between syllables or at word edge. For example, Wheeler (2007, 4) has constructed a so nority ranking scale, reproduced in Table 4-33, for the formative stages of Catalan followi ng the precepts outlined in Clements (1990). Table 4-33. Proto-Ca talan sonority ranking 7 6 3 2 0 tap laterals, trill nasals obstruent voiced nonsibilant continuants sibilants plosives + non-sibilant voiceless continuant ts s l r m n v z b d g f p t k 197

PAGE 198

The scale is a determ inant of the well-formedness of syllables and it is an essential factor in the formulation of other constraints such as SONORITY SEQUENCE and SYLLABLE CONTACT LAW reproduced below. Wheelers (2007, 4) formulation of sonority largely coinci des with Clements (1990, 285) Sonority Sequencing Principle: Between any first or last member of a syllable and the syllable peak, only sounds of higher sonority rank are permitted. (4.26) SONORITY SEQUENCE (SONSEQ): Sonority must increase from the beginning of an onset to the nucleus of a syllable, and mu st decrease from the nucleus to the end of the syllable. (Wheeler 2007) Wheeler applies a further constrai nt on sequences that are heterosyllabic where one forms a coda and the other an onset of the following syllable: (4.27) SYLLABLE CONTACT LAW (SYLCON): The final element of a syllable is not less sonorous than the initial element of an im mediately following syllable. (Wheeler 2007) This formulation (Wheeler 2007, 4) does not ba n consonants of equal sonority, for example -ptacross a syllable boundary but it would prohibit abjecte abject which does, in fact, occur in Catalan as [ b kt ]. In general, despite these exceptiona l cases which are mostly late, learned additions to the lexicon, well-fo rmedness constraints interact w ith faithfulness constraints to produce the optimal sequence of segments for a sp ecific language. A well-formedness constraint is proposed here, NOCODACAT, CAS, POR (with subscripts for language specific restrictions), which allows a single constraint to be introduced in the OT tableau that app lies across languages. (4.28) NOCODACAT, CAS, POR: A syllable does not have a coda unless it satisfies languagespecific phonotactic constraints. A simplified list of coda conditions is given in Table 4-34. These coda conditions are selectively applied, that is, learned words may remain faithful to input bu t there may be another word with the same etymon which displays the resu lt of stricter coda conditions. However, the 198

PAGE 199

representation of a segm ent in standard ort hography is no guarantee that it is, in fact, pronounced. The retention of codas is a majo r point of difference between European and Brazilian Portuguese phonotactics (P arkinson 1988, 141), also reflecte d in standard orthography. Table 4-34 shows that for Catalan all stops are a llowed in coda position but they assimilate to voice of a following obstruent. The restriction on stops for Castil ian reflects the popular stratum where only the coronal stop is allowed in word fi nal position and word internal stops in coda position are strongly dispreferred. Again, favoring the restricti ons for the outcomes of popular words, it can be said that Portuguese allows no st ops in coda position either word internally or word finally. A line of dashes indicates that th ere are no particular rest rictions on segments as codas. Table 4-34. Phonotactic restrictions for c odas in Catalan, Castilian, and Portuguese Catalan Castilian Portuguese Stops voice] .[ voice/+cons -voice]wd *]syll +coronal]wd *]syll Sibilants voice] .[ voice voice] .[ voice voice] .[ voice -voice]wd Nasals place].[ place place].[ place *+cons/+nas]syll Liquids ---+coronal] syll, wd +coronal] syll, wd Glides ---------*Asterisk indicates that a segment is not allowed. Sibilants, liquids, and glides are permissibl e codas for all three languages. There are, however, language specific points of difference. Catalan preserves coda s but in casual speech stops assimilate to the place of articulation of a following consonantal segment (Hualde 1992, 384). Except in learned words and formal regiser of speech, coda stops are elided in Castilian and Portuguese. With regard to sibilants, although both Catalan and Portuguese distinguish /s/ and /z/ in intervocalic position, in coda position voice is depende nt on the following segment, as is also the case for Castilian. The phoneme /f/ is relatively infrequent and has been omitted from this table. It is disallowed word finally in Ca stilian and Portuguese. Catalan has no constraints 199

PAGE 200

on syllab le-final liquids. Both laterals, /l/ and / /, occur as codas but only /l/ shows assimilation and velarization to [ ] (Hualde 1992, 373, 396). Only in word final position does Catalan show variable deletion of rhotics, including / /, followed by the plural marker /s/. Deletion is blocked when the primary accent is penultimate or antepenultimate (Hualde 1992, 406-407). There are many exceptions, especially among monosyllables where / / is preserved, again supporting the idea of a minimal bimoraic count for words. It shoul d also be noted that the deletion of syllable final and word final rhotics is not unknown in synchronic varieties of Spanish and Portuguese (Lipski 1994, Penny 2000, 126-137, Cardoso 2005). Na sal consonants in coda position are disallowed for Portuguese whether syllable final or word final. The resulting nasal vowel renders the syllable heavy even if there is no glide in the coda; nasa l vowels are appreciably longer than non-nasal vowels in Portuguese (Costa 2004). Word final /n/ is deleted in Catalan when a stressed vowel preceeds and there are no other intervening segments; however, in the plural the final cluster /-ns/ is allowed in c ontrast with /-rs/ (Huald e 1992, 404). It is now possible to view the effect of the new constraint NOCODACAT, CAS, POR in the OT tableau in Table 435. Table 4-35. Tableau for first de clension input ty pe L.L.HC1.L c t racta waterfall FT TROCH NOCODACAT, CAS, POR HEADMAX MAXIO STW PARSEALIGNR a. CAT (ka.ta)('rak.ta) *CAS, POR b. (ka.ta)('rak)ta *CAS, POR c. CAS, POR (ka.ta)('ra.ta) Unless NOCODACAT, CAS, POR violates a constraint for the language specified for the candidate forms it is not marked with an asterisk. The two winning candidates do not incur 200

PAGE 201

violations specific to their language As expected, Catalan is more faithful in preservation of the coda consonant than Castilian and Portuguese. W hile both candidates A and B satisfy the major constraints for Catalan, candidate A is the pref erred candidate because it does not incur any parsing violations and the head f oot is aligned with the right word edge. Both candidates A and B violate the NOCODACAT, CAS, POR constraint for Castilian and Po rtuguese because stops in codas are generally not tolerated. Although candidate C violates STW because the syllable with primary accent is no longer bimo raic it is still the winner for Castilian and Portuguese because it satisfies NOCODACAT, CAS, POR. Satisfaction of NOCODACAT, CAS, POR requires violation of MAXIO. The two winning candidates both represent a sequence of two syllabic trochees, ( L .L)(' H .L) for Catalan and ( L .L)(' L .L) for Castilian and Portuguese. The other set of nouns with heavy penultimate syllable, HV type, is also relatively small. Its possible configurations are s hown in Table 436. The projected merger of HV and L syllable types is indicated by an arrow and the correspon ding new patterns are in shaded rows. Patterns corresponding to Latin inputs in th is data set have a checkmark. Assuming that HV to L mergers have occurred in most cases, Table 4-36 shows th at the nine original templates (in unshaded cells) are now reduced to four (The second a nd third columns result in identical forms), differentiated into two groups depending on the mo raic count of the word initial syllable. Table 4-36. Possible heavy/light syllable conf igurations in 4-syllable first declension nouns with penultimate accent a nd HV1 penultimate syllable a. HC.HC.HV1.L e. HV.HC.HV1.L j. L.HC.HV1.L b. HC.HC.L1.L f. L.HC.L1.L k. L.HC.L1.L b. HC.HV.HV1.L g. HV.HV.HV1.L l. L.HV.HV1.L c. HC.L.HV1.L HV.L.HV1.L L.L.HV1.L h. m. d. HC.L.L1.L i. L.L.L1.L n. L.L.L1.L 201

PAGE 202

The new patterns appear in the shaded rows of Table 4-36. N ote that the second and third columns effectively merge (Forms F, K and I, N ar e identical) due to the lo ss of vowel quantity. Following the mergers occasioned by loss of vow el quanity, four templates emerge as input types for Ibero-Romance. In all cases the penultimate accent bearing syllable is light. Because the idea of positional prom inence in the initial syllable is an object of study these four templates are grouped by the nature of the initial syllable or S4. Heavy S4 1. HC.HC.L1.L 2. HC.L.L1.L Light S4 3. L.HC.L1.L 4. LL.L1.L One would expect template 2 to be particularly vulnerable to loss of S3, however, there are relatively few cases of syncope and these are not always uniform across languages. When it does occur syncope follows the prediction and dele tes S3, the weakest syllable of the word. It occurs more frequently when S4 is heavy but si nce the data set (5 cases) is small it is not possible to establish a pattern. Table 4-37. Cases of syncope in 4-syllable first declension nouns with penultimate accent. Input Catalan Castilian Portuguese HC.L.L1.L a. ARBORTA grove arvoredoa arbreda arboleda b. CONSUTRA seam, stitchingb costura costura costura c. SEPTIMNA week setmana semana semana d. TERTINA tertian fever terana terciana ter L.L.L1.L e. POSITRA postureb postura postura postura a The citation form in Lewis and Short is arb r tum, i, n. The neuter plural in -a became the generalized form in Catalan and Castilian where it was reinterprete d as a feminine singular, a common development of neuter plurals. b Lloyd (1987, 206) suggests that syncope between /s/ an d /t/ occurred at a very early date. The fact that the outcome is uniform across languages seems to support this point of view. 202

PAGE 203

In the input for ms in Table 4-37 the locu s of primary accent is indicated by an acute accent. Because an initial heavy syllable is conc omitant to syncope in the first four cases the motivation for this change could be interpreted as a correction strategy for prosodic trapping, that is, a preference for (HC)(' L .L) over (HC)L(' L .L). Example D is somewhat different in that it represents a two stage process: 1) glide fo rmation, preserved in Castilian, followed by 2) palatalization of the preceding consonant with absorption of the glide, exemplified by Catalan and Portuguese. Prosodically, the result is the same as the cases of syncope previously presented. Portuguese loses intervocalic /n/ resu lting in two heavy syllables, the second of which bears the primary accent: ter or (HC)('HV). Arguably the vowel in the final syllable has been preserved although it has coalesced with th e vowel of the tonic syllable; thus MAXMORPH is not violated. However, prosodic repair does not apply in example E where the input form POSITRA contains two optimal trochees: ( L .L)(' L .L). The sibilant /s/ is an acceptable coda for the three languages studied here and should be viewed as a catalyst for atonic vowel deletion based on 1) historical evidence: plural formation in Catala n where addition of final /s/ without a preceding vowel except in the case of sibilants is the norm for the class of nouns inherited from second, third, and fourth declensions, and 2) synchronic evidence: severe vowel reduction and loss in the environment of /s/ in Andean and Mexican Sp anish (Lipski 1989, 802) an d European Portuguese (Mateus and dAndrade 1988). The universal result of POSITRA, postura is (H)(' L .L) in prosodic terms. While this parsing is e ssentially equivalent to the original ( L .L)(' L .L) there is one critical difference, the pos itional prominence afforded the initial unaccented syllable which is now heavy. 203

PAGE 204

Further analysis of the four-syl lable first declension nouns takes in to consideration first the two input patterns identified as havi ng a heavy initial syllable: H C .HC.L1.L and HC.L.L1.L. These two patterns represent native Latin nouns with a penultimate HV syllable as well as Greek nouns in which the original diphthong of the penultima te has been replaced with a vowel, treated as stress-bearing. Again, positional prominence should favor retention of coda consonants in the word initial syllable. The fi rst pattern, HC.HC.L1.L, corresponds to 6 items in the database whose outputs appear in Table 4-38. The word final syllable, S1, is light for all. Loss of moras in S4 is due primarily to degemination. In half the cases the heavy initial syllable remains heavy, either with a coda consonant or a long nasa l vowel when the input form has a nasal coda consonant. S3 is also faithful to input and re mains as a heavy syllable in 67% to 83% of the cases. Table 4-38. Nature of first th ree syllables in outcomes of first declension HC.HC.HV1.L S4 S3 S2 HC HV L HC HV L HC HV L CAT 3 0 3 5 0 1 0 0 6 CAS 3 0 3 4 0 2 0 0 6 POR 1 2 3 2 2 2 0 0 6 n=6 The second pattern with an ini tial heavy syllable, HC.L.L1.L is represented by 27 items in the database. Many of these words are learned or semi-learned words including a number of Greek borrowings. There are only a few cases that result in loss of a syllable, examples A-D given in Table 4-37, that is, arbr ta grove, cons t ra stitching, sept m na week, tert na tertian fever. For the 23 cases that remain as tetrasyllables all but tw o have an initial heavy syllable. Degemination accounts for the two instan ces of moraic loss wher e input forms /-ff-/ and /-rr-/ are now considered to be single consonants. 204

PAGE 205

Com bining the results of both groups with initial heavy syllab le (HC.HC.L1.L and HC.L.L1.L inputs) it is possible to create an aggregate profile of 29 items (excluding the four cases of syncope or syllable loss). Figure 4-17 shows the output by syllable type for S4, S3, and S2 (all ultimate syllables are light). As expected there is a high degree of retention of the coda consonant in S4, the initial syllable. The lowe r incidence of HC for Portuguese is compensated for by the presence of HV syllables, absent in th e other two languages. Faithfulness to input segments in the six cases of heavy S3 with coda consonant is somewhat surprising because this syllable is not in privileged position if one assumes a parsing of (H)(H)(' L .L) where S3 must be construed as a weak syllable to avoid stress clash. If trochees are treated as syllabic, S3 would be in a weak position metrically, (' H .H)( L .L). Figure 4-17. Output of first declension HC.HC.HV1.L and HC.L.L1.L by syllable type. For the six forms where the input has a heavy sy llable for S3, five in Catalan and four in Castilian and Portuguese (adding HC and HV togeth er) remain heavy, as seen in Table 4-39. The determining factor here appears to be th e morphemic status of S4, represented by the prefixes con- with, disnot, and innot. With the exception of ad- toward, these prefixes CAT CAS POR L HC S4 HV L HC S3 HV 0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% 100.0% HV S2 HC L 205

PAGE 206

retain their bim oraic status across languages. These prefixes can be considered to occupy a privileged position because they constitute morphemes and occur in the initial syllable of the prosodic word. Table 4-39. Nature of the first three sy llables in outcomes of first declension HC.HC.HV1.L and HC.L.L1.L S4 S3 S2 HC HV L HC HV L HC HV L CAT 22 0 7 5 0 24 0 0 29 CAS 22a 0 7 4 0 25 0 0 29 POR 12 10 7 2 2 25 0 0 29 n=29 a The count for Castilian includes three trisyllables, axioma axiom, cristiana christian (fem.), genciana gentian. They are treated as if they were aligne d left with the corresponding forms in the other two languages. A further argument for the special character of the prefix in the prosodic word is the blocking of glide formation in some word in itial syllables (Cabr and Prieto 2004, 2005; Chitoran and Hualde 2007). Many of these cases co rrespond to syllables that could be construed as morphemes or prefixes, such as biand triin Greek loanwords and scientific terminology. The avoidance of gliding here could be seen as a parallel to the gene ral resistence to glide formation across word boundaries ex cept in certain tempos and regist ers. If the four to five special cases reflected in the S3 columns are thus accounted for, then there is a single emerging pattern (head foot in bold): (H .L)('L .L). The optimal head troch ee is the universally preferred pattern, ('L .L), and the cross-linguistic outcome for al l cases represented in Table 4-39. The word initial trochee in nearly 80% of cases is the less optimal (H .L). The initial heavy syllable satisfies faithfulness by preserving segments or moras in the input and also gives prominence to the initial syllable de spite the vowel reduction attested in Catalan and Portuguese. It should be noted that some initial syllables do not show positional prominence, as is seen in vowel reduction in both Catala n and Portuguese in the case of comissura corner of ones mouth (
PAGE 207

corner of ones m outh is pronounced [u], the reduced vowel corresponding to phonemic /o/ and / /. Extreme vowel reduction or apheresis, elis ion of an initial unstressed vowel, is common in some varieties of Catalan (Bonet and Mascar 1997). When the onset that would result from apheresis is unacceptable, then deletion of the initial vowel is blocked. The blocking of apheresis in Catalan contrasts with Europ ean Portuguese where normally unacceptable onset clusters arise in informal, rapid speech due to elision of an unstressed initial vowel as in [ pa.su ] espao space. Omission of the syllable nucleus may extend across two or more syllables and thus create clusters of thr ee or four consonants (Mateus a nd dAndrade 2000, 43-44). Analysis of the two patterns with initial light syllable, L.HC.L1.L and L.L.L1.L will further demonstrate that for four-syllable words ( )(' ) is, indeed, the template of choice, two syllabic trochees. The first input pattern, L.HC.L1.L consists of 13 items, 7 of which are Greek words transmitted to Ibero-Romance through Latin. The first syllable in this group is light in input and remains so in all the output forms. The onl y heavy syllable, S3, loses moras in the case of geminate liquids cross-linguistically (with the exception of one case of geminate /l/ in Catalan) but in 10 words remains as a heavy syllable although Portuguese has replaced the mora of the coda with a long nasal vowel in one instance and with a falling di phthong in another. Portuguese also loses a syllable in the case of avel hazelnut (<[nux] Abell na, Lewis and Short s.v. Abella) due to deletion of intervocalic /n/. For this small set, then, the emerging pattern is faithful to the original in 77% of the cases. The sequence L.HC.L1.L presents a problem in terms of parsing. For Latin, the tonic syllable would have been a long vowel and in some cases the initial syllable as well. The two possible parsings for Latin, then, are: L(H)('H) and (H)(H)( 'H) In the first, the unparsed syllable would incur a violation of PARSE. However, in Table 4-35 the failure to 207

PAGE 208

parse is shown as a low ranking constraint. In the case of the second pa rsing with three heavy syllab les followed by the extrametrical final syllab le it can be assumed that the initial syllable probably had a secondary accent so that rhythmically there is a sequence of four syllables with alternating stress. Although the in itial light syllable looks particularly vulnerable, in the data set under consideration there are no cases of histori cal apheresis although ther e is vowel reduction in both Catalan and Portuguese as well as synchronic apheresis as discussed above. For Ibero-Romance it is necessary to move away from the concept of the moraic trochee. With loss of vowel length the input form must be construed as L.HC.'L. L. If trochees are constructed from right to le ft then the first well formed trochee consists of ('L .L). In accordance with HEADMAX (A stressed element in the input must have a stressed element as its output correspondent) the original locus of accent is maintained. A second trochee could also be formed from ('L .HC) if trochees are syllabic, that is, t he foot template simply counts syllables, ignoring their internal structur e (Hayes 1995, 63). An often anthologized example from 19th century Spanish poetry demonstrates the viability of this interpretation. Jos de Esproncedas (1808-1842) Cancin del pirata is written primaril y in octosyllabic verse but has a series of intercalated stanzas in pie quebrado or half lines consisting of four syllables. Navarro Toms (1968, 40) describes these as ritmo trocaico with accents on the first and third syllables. Using verses from the cited poem, syllables are coded as heavy or light taking into consideration the usual synalepha between words. The two cases of three-syllable lines ( versos agudos ending in a stressed syllable) are indicated by +1 for the empt y timing slot. Parsing of the line is indicated to the right. It is assumed that ('H) is a legitimate trochee beca use there must be a provision for stressable monosyllables. 208

PAGE 209

Parsing Frequency of trochaic patterns ('L.L) ('H.L)('H.H)('L.H) ('H) Veinte presas H L L H hemos hecho L L L L a despecho L H L L del ingls, L H H + 1 y han rendido H H L L sus pendones H H L H cien naciones L L L H a mis pies. L H H + 1 ('H.L) ('L.H) ('L.L) ('L.L) ('L.H) ('L.L) ('L.H) ('H) ('H.H) ('L.L) ('H.H) ('L.H) ('L.L) ('L.H) ('L.H) ('H) 0 2 1 0 1 0 1 0 1 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 1 0 1 1 0 1 1 1 0 0 0 1 0 0 0 1 Total frequency 5126 2 For any speaker of Spanish the rhythmical patte rn of these verses are transparent and there would be little hesitati on in reading the text in the intended pattern of alternating stress on the first and third syllables despite the fact these sy llables are most frequently not bimoraic while unstressed syllables are. The i nventory of trochaic patterns shows that the most and least harmonic patterns (Prince 1990, 367-373) ar e the most frequent, that is, (L .L) and (L .H). Having established the possibility that the most frequent and preferred outcome of input L.HC.'L.L is (L .H)('L .L) it now follows that the optimal output of LL.L1.L also cons ists of two harmonic trochees, (L .L)('L .L), with the rightmost as head foot. The data set corresponding to L.L.L1.L input consists of 37 items only one of which retains a heavy initial sy llable in the form of the falling diphthong /ew/, euph nia pleasant sound. Many members of this data set are clearly learned words which came to Romance via Latin. Placement of accent, although variable, gene rally follows the conventions of Latin rather than the referring to the original accent in Greek. The Greek diphthong becomes Latin by 209

PAGE 210

the 5th century CE (Biville 1990, 166) while Greek -> Latin -, is interpreted as a short vowel according to the rule uocalis ante uocalem corripitur cited in Biville (1990, 164) who notes that this is a key difference between Greek and Latin o lon ne peut, lp oque classique, trouver de voyelle longue devant une autre voyelle. Consequently, first declension nouns ending in a may reflect both (items 1-6 below) and (items 7-10 below) inputs as seen in the list below. Outputs with antepenultimate stress or derived from an antepenultimate input are underlined. All other ouputs have penultimate stress. Latin Greek Catalan Castilian Portuguese 1. pth a, ae f. apathy apatia apata apatia 2. l g a, ae f. elegy elegia elega elegia 3. r n a, ae f. irony ironia irona ironia 4. l t n a, ae f. litany lletania letana litania/ladainha 5. pr ph t a, ae f. prophecy profecia profeca profecia 6. pl t a, ae f. the State policia polica polcia 7. tr ph a, ae f. atrophy atrfia atrofia atrofia 8. euphnia, ae, f. pleasant sound eufonia eufona eufonia m l d a, ae, f. melody 9. melodia meloda melodia 10. th r a, ae f. theory. teoria teora teoria Although the quantity of the vowel of the penul timate syllable is variab le in the Latinized forms (from Lewis and Short), the accentual pattern indicates that these syllables were usually treated as heavy. Preference for antepenultimate accen t in the Latin input is seen infrequently: the outputs of tr ph a in Castilian (atrofia trisyllabic with troas the accented syllable) and Catalan atrfia (tetrasyllabic with troas the accented syllable). Additional examples of presumed antepenultimate input are polcia in Portuguese, as well as Old Catalan, polcia urbanity (DCVB, s.v. policia ). The popular output of l t n a in Portuguese, ladainha also suggests antepenultimate stress for the input form. 210

PAGE 211

In term s of syllabic loss, the L.L.L1.L nouns ar e extremely stable. There is only one case of syncope, ps t ra, ae f. posture. As indicated above lo ss of pre-tonic vowels seems to have occurred at an early date wh en the resulting heterosyllabic cluster contained /s/ as coda consonant. There are only two other types of syllabic loss, the first is apheresis in p th ca, ae f. storehouse where the initial syllable must have remained long enough for lentition of /p/ to occur. Forms with initial/b/ occur thr oughout eastern and western Romance, Ital. bottega Fr. boutique, Cat./Cast./Port. bodega. A similar example is Gr. >L. hemicrania which results in Cast. migraa (Biville 1990, 51) as well as Cat. migranya (DCVB, s.v. migranya). A second second type of syllabic loss, glide fo rmation which eliminates a syllabic nucleus, is more prevalent in Ca stilian. In the case of atrofia atrophy, euforia eurphoria and vesania madness stress is retracted to original S3. This result is not surprising in view of the uncertain vowel quantity of the vowel of the penultimate sy llable. Also, given the semantic category of these words this may also be a case of hype rcorrection since proparoxytonic stress is a characteristic of many learned words and scientific terms. Glide formation in pre-tonic position is also limited to Castilian in this data set in the word idiota idiot, but here the primary accent remains attached to the penultimate syllable. Table 4-40 demonstrates the regularity of outcomes for the 35 items in this data set (excluding trisyllabic bodega and postura ). Table 4-40. Nature of the first three sy llables in outcomes of first declension HV/L.HV/L.HV/L1.L S4 S3 S2 HC HV L HC HV L HC HV L CAT 0 1 34 0 0 35 0 0 35 CAS* 0 1 34 0 0 35 0 0 35 POR 0 1 34 0 0 35 0 0 35 n=35 *The count for Castilian treats the four trisyllables, atrofia, euforia, idiota, vesania as if they were aligned left with the corresponding forms in the other two languages. 211

PAGE 212

Table 4-41 shows aggregate da ta for all tetrasyllabic first declension nouns with penultim ate accent. The data set contains 78 noun s. With the exception of the small group of nouns with heavy first syllable with coda consonant (see Table 4-39), S4 is light in over 70% of cases. The number of light syllables in S3 (a weak position metrica lly) is 94-95%, and 86-88% in the case of S2. All final syllables (S1) ar e light and are not reflect ed in Table 4-41. Table 4-41. Nature of the first three syllables in outcomes of a ll first declension tetrayllables with penultimate accent S4 S3 S2 HC HV L HC HV L HC HV L CAT 22 (28.2%) 1 (1.3%) 55 (70.5%) 5 (6.4%) 0 (0.0%) 73 (93.6%) 11 (14.1%) 0 (0.0%) 67 85.9%) CAS 22 (28.2%) 1 (1.3%) 55 (70.5%) 4 (5.1%) 0 (0.0%) 74 (94.9%) 9 (11.5%) 0 (0.0%) 69 (88.5%) POR 12 11 55 2 2 74 3 (3.8%) 6 (7.7%) 69 (15.4%) (14.1%) (70.5%) (2.6%) (2.6%) (94.9%) (88.5%) n=78 The most frequent pattern for this class of nouns, th en, is all light syllables parsed as (L.L)('L.L). The second most frequent pattern has a heavy sy llable in the S4 position: (H.L)('L.L). The notion of positional prominence has been fully expl ored as a way of accoun ting for retention of moras in the word initial syllable. The distribution of syllable types, illustra ted in Figure 4-18, shows remarkable uniformity across languages with the exception of Portuguese where S4 (see Table 4-41) is frequently HV. The 22 cases of HC type syllables seen in Catala n and Castilian are evenly divided in Portuguese into those that retain the coda consonant and those that have either a falling diphthong or a nasal vowel when the coda consonant of the input is nasal. When the HC and HV categories are considered together the percentage distribution of heavy and light syllables is identical: 70.5% light and 29.5% heavy. In summary, it can be sa id that the most prev alent pattern for first declension paroxytonic tetrayllables is (L.L)('L.L). When the initial syllable is heavy in the input form it is likely to remain heavy; thus, (H.L)('L. L) is an alternate patte rn found in 28-29% of the 212

PAGE 213

78 cases rep resented in the data set. In both patterns the metrical commonplace is a trochaic head foot, of the preferred template ('L.L), built at the right edge of the word. This is indicated by the high frequency of light syllables (L) shown in Figure 4-18 in S1 and S2. Figure 4-18. Distribution of syllabl e types in the output of all firs t declension tetrasyllables with penultimate accent. CAT CAS PORHC S4 HV L HC S3 HV L HC S2 HV CAT 100.0% 90.0% 80.0% 70.0% 60.0% 50.0% 10.0% 20.0% 30.0% 40.0% 0.0%CAS PORL S1 LFour-syllab le Nouns with Antepenultimate Accent Antepenultimate accent patterns in Latin derive from the impossibility of building a foot with the penultimate syllable as head if that syllable is not bimoraic. Possible patterns for nouns in this class are shown in Figure 4-19. It has been established that the final syllable in first declension nouns does not undergo apocope. The vulnerable syllable, then, in terms of Mesters (1994) notion of prosodic trapping is S2. However, it will be seen that when S3 ends in a coda consonant there are virtually no case s of syncope. It is true that many tetrasyllables in this class are reduced to three syllables in Castilian, but th is corresponds to a constr aint against vowels in hiatus when the first vowel is [-low]. 213

PAGE 214

HC Antepenultimate Accent HV L HC HV L L L S4 S3 S2 S1 Figure 4-19. Heavy/light syllable configurations for first declension tetrasyllables with antepenultimate accent. The first data set in this cla ss corresponds to the input type X. H1C.L.L where X can be HC, HV, or L. These 44 nouns show distribution of heavy /light syllables in S4 of the input form as follows: HC (14; 31.8%), HV (7; 15.9%), L (23; 52.3%). It should be noted that in many cases an initial heavy syllable corresponds to a prefix. Positiona l prominence is further reinforced when the initial syllable also corresponds to a morpheme. Deletion or diminution of a syllable (such as prefixes ab-, con -, dis-, in/en -, prae -, sub -) is less likely when that syllable also carries a semantic value. Table 4-42 shows the outcome in terms of syllable count of the input type X.HC1.L.L. Castilian and Portuguese have widesp read reduction to three syllables, primarily as a result of glide formation in nouns ending in a such as ausencia (
PAGE 215

pronounced [ m ] and [ am ] (DCVB, s.v. ametlla ), and almendra in Castilian. Other than the inherent weakness of the post-tonic vowel, th ere is little motivation f or deletion of the nucleus in S2. The resulting consonant sequence is clearly suboptimal as can be seen in the repair strategies employed, assimilation in Catalan and epenthesis in Castilian. The nature of the syllables in three and four-syllable outputs of X.HC1.L.L is summarized in Table 4-43. The column headed by N indicates loss of a sylla bic nucleus through glide formation (and the one case of syncope discussed above). Table 4-43. Outcomes of Xa.HC1.L.L first declension nouns by syllable type Initial syllable Tonic syllable Post-tonic syllableb Final syllable HC HV L HC HV L HC HV L N HC HV L Catalan 13 1 30 44 0 0 0 0 41 3 0 0 44 Castilian 14 0 30 44 0 0 0 0 9 35 0 0 44 Portuguese 13 0 31 19 23 2 0 0 17 27 0 0 44 n=44 a X represents input syllables with the following distribution: 14 HC, 7 HV, 23 L. b Nouns that emerge as trisyllables are counted under the N column. The predominant patterns are L.'HC.L.L for Catalan, L.'HC.L for Castilian, and L.'HV.L and Portuguese. The Castilian/Portuguese patterns present little difficulty in a system of syllabic trochees built from right to lef t; parsing would be L('HC/HV.L). However, selection of this trisyllable as optimal output requires the addition of a constraint against vowels in hiatus. Previously discussed HEADMAX serves to maintain the original locus of accent. The well formedness constraint ONSET, *M/V[+high] (prohibits high vowels in the margins of the syllable), and FT-LEFT (requires formation of a disy llabic trochee at the begi nning of a word) are used by Cabr and Prieto (2004, 133-141) to account for acceptability of glide formation in different varieties of Catalan and in different word positions. The innovative variety of Catalan, which allows glide formation in some contexts (but no t word initial position), pa rallels the divergence of the three languages histor ically with regard to V[+high]V sequences. In the cases in question, 215

PAGE 216

none of the V[+high]V sequences occur at the beginning of th e word. Therefore, it is only the relative ranking of ONSET and *M/V[+high] that differentiates Castilian from Catalan and Portuguese. Table 4-44 displays the relevant rankings. Table 4-44. Tableau for first declension input type HC.HC.L.L infant a FT TROCH MAX MORPHHEADMAX ONSET *M/V[+high] PARSEALIGNR a. CAT in('fan.si)a **! ** b. POR ('f .si) **! ** c. POR ('f .sja) d. CAS in('fan. i)a **! ** e. CAS in('fan. ja) FT TROCH MAX MORPHHEADMAX *M/V[+high]ONSET PARSEALIGNR f. CAT in('fan.si)a ** ** in('fan.sja) *! g. CAT The pattern seen in candidate F incurs more violations than patterns C and E but none involve high ranking constraints. HEADMAX prevents displacement of the accent to the right which would allow the creation of two syllabic trochees and *M/V[+high] prevents glide formation which would also result in a right-aligned disyllabic trochee. Thes e constraints guarantee preservation of the original ternar y pattern. It is only in the forms dolena (
PAGE 217

Catalan (peripheral areas of Ce ntral Catalan as well as older speakers from Barcelona, according to Cabr and Prieto, 2004). For the gene ral variety of Catalan, in the case of historia history, part of this data set, a rising diphthong [j ] was preferred over hiatus for 85% of the subjects in Cabr and Prietos study (2004, 120). It can be pr edicted that over time the order of constraints needed to select candidate F will be revers ed resulting in selection of candidate G. The last data set in the category of tetrasyllabic first declension nouns with penultimate accent combines those in which the primary accent falls on a syllable of HV or L type. The resulting input for Ibero-Romance is then X.L1.L.L The nature of the initial syllable seems to have little effect on the outcome as seen in th e analysis of input form X.HC1.L.L above. The distribution of S4 in input is as follows (only diphthongs are c oded as HV): HC (43; 35.2%), HV (4; 3.3%), L (75; 61.5%). The predominant input pattern consists of four light syllables, or L.L1.L.L, parsed as L('L.L)L. The penultimate sylla ble is the weak syllable of the head foot and is unprotected by constraints such as MAXMORPH. Syncope is seen to occur more frequently when the input is L('L.L) as opposed to thos e templates in which the tonic syllable is heavy. Glide formation continues to be prevalent in both Castilian and Portuguese, especially when the final syllable lacks an onset and is preceded by a high vowel. The locus of accent and the nature of the accented syllable are displayed in Table 4-45. Table 4-45. Outcomes of Xa.L1.L.L first declension nouns by type of accented syllable and word level syllable count 4 syllables/accent on S3 3 syllables/accent on S2 HC HV L Total HC HV L Total Catalan 0 0 85 85 (69.7%) 3 1 33 37 (30.3%) Castilian 0 0 31 31 (25.4%) 3 0 88 91 (74.6%) Portuguese 0 1 35 36 (29.5%)3156886 (70.5%) n=122 aX corresponds to these inputs: 43 HC, 4 HV, and 75 L. 217

PAGE 218

In both 4 syllable and 3 syllable outcom es of X.L1.L.L the accented syllable is almost always light. In Chapter 3 it was suggested th at the changing accentual pattern of Late Latin corresponded in some ways to that of stresstimed languages where elimination of vulnerable nuclei often resulted in the buildup of segments in the accented syllable (Crosswhite 2001, 173). It was also observed that loss of posttonic syllab les also resulted in the creation of feet with preferred duple rhythm rather than ternary rhythm. It is clear from the data in Table 4-45 that preferred rhythmic type rather than preferred sy llable type motivates this change. The large number of HV type syllables that occur in Portuguese in the case of trisyllables with penultimate accent requires some explanation. There are a few cases that result from nasalization but most are reflexes of the suffix r a seen in r pr a, -ae f. river banks (cf. r pa, -ae f bank) which regularly becomes eira ( ribeira ) in Portuguese. The falling diphthong is generally attributed to anticipation of yod resulting in metathesis (Lloyd 1987, 264); in the case of Castilian the falling diphthong later undergoes monophthongization resulting in [e], as in ribera The cases of syncope or glide formation th at occur cross-lingui stically correspond to words of popular nature. They are an important subs et of the forms classed as syllables/accent on S2 in Table 4-45. Words that show reduction to three syllables in all three Ibero-Romance languages under study are listed below. It should be noted that in thes e cases there is little evidence of identifiable prefixes. Failure to parse the initial syllable, then, is not in conflict with a need to place it in re lief. Proof of the wea kness of the initial syllable is seen in cases of apheresis in modern Catalan when the first sylla ble contains unstressed schwa as nucleus. The word agulla needle in line 10 below is rendered as gulla in some varieties of Catalan (Bonet and Mascar 1997, 119). Similarly, a common pattern for hypocoristic form ation in all three languages retains only the final two-syllables, that is, the head foot, in trisyllables with 218

PAGE 219

penultim ate accent, for example, Chepa, Pepa (
PAGE 220

s+vowel+stop or liquid+vowel+stop. Elision of th e vowel nucleus in the latter results in a consonant sequence favored by the sonority sequence constraint (SONSEQ), that is, s + stop and liquid + stop. In the case of the diminutive su ffix the reflexes of Ibero-Romance suggest that loss of did not lead to resyllabifica tion (although /kl-/ as onset is also subject to palatalization in western Ibero-Romance) rather the velar st op suffers lenition and place assimilation. The resultant palatal glide leads to palatalization of the lateral in Catalan and Portuguese and palatalization and delateralization in Castilian. The reflexes of m r t ma tidelands also show the effect of SONSEQ in the change from /t/ to /s/ in c oda position following syncope. Those cases in which penultimate accent is retain ed cross-linguistically correspond to learned vocabulary including some Greek words, where proparoxytonic accent is more common. The few cases of dual output (learned words and popul ar words with the same etymon) demonstrate the variable ranking of MAX constraints that prohibit loss of segments. An example is pr bla, ae f. parable, speech which retains its prim ary meaning and original form in learned parbola (present in all three languages) and popular paraula/palabra/palavra which have the preferred accentual pattern and the secondary meaning of the word. From the two data sets analyzed here orig inal X.HC1.L.L and X.L 1.L.L it can be seen that for Castilian and Portuguese the preferred pr osodic template is X('X.L). Although Catalan shows a higher percentage of tetrasyllabic outco mes, X('X.L)L, there is evidence that when the last syllable lacks an onset and is preceded by a high vowel, glide formation is increasingly acceptable as a repair strategy for the well formed ness violation. Those words that remain as proparoxytones are lexically marked and stand apar t, both by their lexical register and prosodic configuration. The unparsed in itial syllable shows increased vulnerability when it does not 220

PAGE 221

coincide with a m orpheme or prefix. Elision is observed in all three languages in cases of synchronic and diachronic apheresis and hypocoristic formation. Five-syllable Nouns Polysyllabic nouns consisting of more than four syllables are less frequent than the patterns described heretofore. The subset of pentasyllabl es consists of 119 items, learned words (many of Greek origin) that show little departure from the original form. Reduction to four syllables occurs with some frequency in Castilian (especially in words w ith antepenultimate accent) and to a lesser degree in Portuguese. Lo ss of syllabic nuclei comes as a result of glide formation rather than syncope. Although the pattern is most pervasive in Castilian synchronic evidence demonstrates that vowels in hiatus consisti ng of an unaccented high vowel followed by a nonhigh vowel constitute an unstabl e sequence in both Catalan and Portuguese. Glide formation is not universal but a function of geographical vari ation and register. Because onsetless syllables are marked in most languages it can be predic ted that in the future, the rising diphthong will become the universal output for Catalan and Portuguese as well. Five-syllable Nouns with Penultimate Accent In the 42 nouns with penultimate stress the high front vowel of S3 forms a rising diphthong with the nucleus of the tonic syllable, S2. Ther e are 6 such cases in Castilian and only one in Catalan and Portuguese, the one word in this data set of obviously popular origin, *aculeata (pertica) goad, prod (Coromines 1995, s.v. agullada) > Cat. agullada Port. agulhada Cast. aguijada (
PAGE 222

The *LAPSE constraint is used in a study of tern ary rhythm by Elenbaas and Kager (1999, 282): Every weak beat must be adjacent to a str ong beat or the word edge. In binary systems words with an odd number of syllables will inev itably violate a constraint that requires the parsing of syllables such as PARSE. However, Elenbaas and Kage r suggest that there are two positions within the word in which unparsed syllables satisfy *LAPSE (1999, 309-310): at word edge and adjacent to a foot head. The preceding discussion of words of three and four syllables suggests that the three languages under study all provide evidence of a seconda ry accent at the left edge of a word in addition to the primary accent aligned at ri ght word edge. Therefore, in a five-syllable word primary (') and secondary ( ) accents occur as follows: ( ) (' ). Because the unparsed syllable is immediately to the left of the head of the rightmost foot *LAPSE is satisfied. Are there cases in which the initial syllable constitutes a foot on its own? Words with prefixes such as in -, dis -, com may fall into this category. An example which occurs in all three languages with identical syllabi c and prosodic outcomes is infanticida infanticide. Although Portuguese develops a nasal vowel in the environment VN]syll, wd (where N represents any nasal consonant) it has been previously proposed that nasal vowels are bimoraic. Therefore, Portuguese also satisfies a presumed bimoraic requirement for foot formation. Fudge (1984, 198) notes that stress-repelling pr efixes (those that re ject stress assignment and move main stress to the stem in suffixless words) such as ab-, con -, ex-, sub-, trans can in certain cases have a full vowel rather than schwa in Nort hern English (British) if they form a heavy syllable as in [ kn.'t nt content. This suggests that preservation of the periphe ral vowel, rather than the central vowel schwa, is a mechanism fo r assigning word initial prominence. 222

PAGE 223

For the lang uages in question, then, if the init ial syllable is perceived as constituting a foot (capable of carrying a secondary accent), the parsing would be ( ) (' ). The *LAPSE constraint is satisfied becase S4 follows a foot head, the initial syllable, and S3 precedes the head of the rightmost foot which carri es the primary accent in the word. There are few cases of this type in the subset of five-syllable nouns with pe nultimate accent. It should also be noted that the vocalism of the prefix con/com/col/cor in Catalan, for example, always has the reduced form of the vowel, [ku], in contrast w ith the two-syllable prefix contrawhich is [ 'k nt ]. A third prefix, enalso shows only the reduced vowel as in encobrir to hide pronounced [ ku ]. In Portuguese the same three pr efixes are realized as [ k knt ], and [ ], reflecting the normal realization of nasal vowels in pre-stressed position (Mateus and dAndrad e 2000, 21). The prefix des-(cf. dis -) shows vowel reduction in both Ca talan and Portuguese resulting in [ d s] and [ d ] (see Mateus and dAndrade 2000, 18, for discussion of oral vowels in pre-stressed position). The inconsistency in vowel reduction, even w ithin the same language, suggests a process of lexicalization in which disyllabic contra is perceived as constituting an independent prosodic element that can be parsed as a disyllabic tr ochee and given prominence within the word. The same does not hold for the monosyllabic prefix es. This discrepancy may result from lexicalization of com(and its allomorphs) as /kum-/, that is a stressless prefix. In the case of a disyllable with ultimate accent (Catalan has many resulting from 2nd/4th and 3rd declension nouns and adjectives.) the first syllable would be unaccented and unaccentable in order to avoid rhythmic clash. The following adje ctive pair is a case in point: [ kuntn ] content, [ ku fin ] confident. For the two-syllable word one is te mpted to posit an iambic foot. However, the 223

PAGE 224

blocking of schwa deletion s een in the fe minine [ kuntnt ], as well as in the case of clusters that violate SONSEQ like /kt/ in acte (
PAGE 225

morphem e. The second patte rn with initial foot ( ) presents some problems when one or both syllables are heavy as is the case in infanticida Because inis a recognizable prefix the preferred parsing is ( ) (' ). This is the only instance in the set of 44 five-syllable nouns with penultimate accent in which S5 and S4 are both heavy. Although there are five cases in which S5 is light and S4 is heavy this is not an obstacle for form ation of a syllabic trochee. The evidence from Castilian vesification cited under discussion of four-syllable words above provides evidence of a strong preference for an alternating pattern of accented and unaccented syllables. In the case of the pentasyllable with penultimate accent, the initial syllable would not bear an accent if syllables are labeled accented/unaccen ted from right to left. However, study of polysyllables in Castilian (Harris 1983, 85-86; Hayes 1995, 96-97; Roca 1986, 1999) supports the notion of prominence of the in itial syllable or the initial dactyl effect (Prince 1983, 49); in other words, a pattern of strong-weak-weak at the beginning of a wo rd is a desirable outcome. Both ( ) and ( ) conform to that rhythmic pattern because the unaccented syllable in a syllabic trochee is weak, as are unparsed syllables. As seen in Table 4-46 the most common output from the first declension pentasyllables with penultimate accent is faithful to the original in terms of syllabic count but with loss of contrasting vowel length there are few heavy syllables. The initial syllable has the highest percentage of heavy syllables and the tonic syllable the lowest Again, it can be seen that although the locus of accent has be en preserved the bimoraic re quirement of the Latin Stress Rule is no longer satisfied. C onsequently, in many cases the difference between this group of nouns and the next (pentasyllables with ante penultimate accent) is no longer transparent 225

PAGE 226

Table 4-46. Outcom es of five-syllable fi rst declension nouns with penultimate accent. S5/Initial Syll S4/Pre-tonic Syll S3/Pre-tonicSyll S2/Tonic Syll HC HV L HC HV L HC L Na HC L Catalan 11 2 29 5 0 37 4 37 1 2 40 26.2% 4.8% 69.0% 11.9% 0.0% 88.1% 7.1% 90.5% 2.4% 4.8% 95.2% Castilian 9 2 31 5 0 37 3 33 6 2 40 21.4% 4.8% 73.8% 11.9% 0.0% 88.1% 7.1% 78.6% 14.3% 4.8% 95.2% Portuguese 5 6 31 2 3 37 2 39 1 2 40 11.9% 14.3% 73.8% 4.8% 7.1% 88.1% 4.8% 92.8% 2.4% 4.8% 95.2% n=42 a N in the S3 column indicates that there is only one pre-tonic syllable due to glide formation. S1, the final syllable, is always light. Syllables that are metrically strong are indicated by shaded cells. Figure 4-20 shows the regularity of this subgrou p with regard to distribution of heavy and light syllables. Areas of divergence are indicate d by data point labels: (1) corresponds to the lower retention rate of coda consonants in Portuguese and the replacement of VN]syll with a nasal vowel; (2) indicates glide formation and loss of a syllable nuc leus in Castilian. However, these differences are relatively minor and do not detract from the overall regularity in the outcome of this noun class across languages. Figure 4-20. Distribution of syllable types in outcomes of first declension pentasyllables with penultimate accent. (1) Coda consonant loss and nasalization in Portuguese; (2) Glide formation in Castilian. 100.0% 80.0% 60.0% 40.0% 20.0% 0.0% S2/Tonic Syll L H V L HC H V L HCLNHC S4/Pretonic S yll S3/Pretonic Syll S5/Initial Syll HC 95.2% 4.8% 2.4% 92.8% 4.8% 88.1% 7.1% 4.8% 73.8% 14.3% 11.9% Portuguese 95.2% 4.8% 14.3% 78.6% 7.1% 88.1% 0.0% 11.9% 73.8% 4.8% 21.4% Castilian 95.2% 4.8% 2.4% 90.5% 7.1% 88.1% 0.0% 11.9% 69.0% 4.8% 26.2% Catalan 226

PAGE 227

Because the five-syllable words are m arked in terms of lexical register, the rate of syllable loss is far less than in the case of four-syllable words (Tables 4-43 and 4-45 above). Surprisingly, here it is Catalan and Portuguese th at pattern together in terms of retention of pre-tonic syllables. Only Castilian shows substantive loss at 14.3%. The most common syllabl e type for both tonic syllable and the three syllables that precede it is the light syllable. Distribution percentages range from a low of 69% in initial position in Catalan to a hi gh of 95.2% across languages on the tonic syllable. Five-syllable Nouns with Antepenultimate Accent This subset of pentasyllables, like the precedi ng one, consists primarily of learned words. A high percentage of these are de rived nouns that have the suffix -Vnt a (where V represents a thematic vowel). The learned reflexes of the su ffix in Catalan maintain the vowels in hiatus but glide formation occurs in both Castilian and Portuguese. The popular reflexes ( -na in Catalan and Portuguese, -nza in Castilian) are not observed ino this subset. Table 4-47 shows the expected divergence of Catalan and Castilian/ Portuguese. Although Portuguese often preserves hiatus in careful speech, in the case of -Vnt a the suffix seems to have been lexicalized as /-V sja/ with nasalization of the pre-to nic vowel and the glide as onset of the final syllable. Table 4-47. Output of first declensi on pentasyllables with penultimate accent by language and syllable count 5 Syllables 4 Syllables 3 Syllables Catalan 68 (88.3%) 2 (2.6%) 7 (9.1%) Castilian 4 (5.2%) 59 (76.6%) 14 (18.2%) Portuguese 8 (10.4%) 58 (75.3%) 11 (14.3%) n=77 Additionally, there is a small group of words that show changes typical of the popular stratum of the lexicon an d are actually reduced to three syllables in all three languages. Loss of 227

PAGE 228

the pre -tonic syllable as well as the post-tonic syllable through syncope or glide formation produces the trisyllabic outcomes: Catalan Castilian Portuguese 1 c r c us, a, um, adj. made of leather cuirassa coraza couraa 2 m n r us, a, um, adj. of or belonging to the hand manera manera maneira 3 p r r a, ae, f. woman worker obrera oberera obreira 4 qu -dr g s ma, ae, f. fortieth quaresma cuaresma quaresma 5 s l t r us, a, um, adj. alone, solitary soltera soltera solteira 6 tert r us, a, um, adj. third tercera tercera tercera 7 v r cund a, ae, f. shame vergonya vergenza vergonha Table 4-48 displays the overall distribution of place of accent and the nature of initial, pretonic, tonic, and post-tonic syllab les. Syllables that have posit ional prominence (tonic and word initial) are indicated by shaded cells; all final syllables are light and unaccented. A column headed by N indicates that an un accented vowel has been lost. When that vowel is post-tonic, the locus of accent moves one syllable toward the right word edge, that is, to the penultimate position. The contrast between Catalan and Casti lian/Portuguese is stark. Catalan has largely preserved the template of proparoxytonic accent (88.3%) while Castilian and Portuguese have shown a marked preference for the more familiar paroxytonic pattern which emerges in 94.8% of the words in Castilian and Portuguese (Portuguese includes 9 cases of accen t shift, 11.7% of the total). Table 4-48. Outcomes of five-syllable first declension nouns with antepenultimate accent. Initial Pre-tonic/non-initial Tonic Post-tonic/non-final HC HV L HC HV L N HC HV L L N Cat 29 1 47 10 0 60 7 44 0 33 68 9 37.7% 1.3% 61.0% 13.0% 0.0% 77.9% 9.1% 57.1% 0.0% 42.9% 88.3% 11.7% Cast 26 1 50 10 0 51 16 43 0 34 4 73 33.8% 1.3% 64.9% 13.0% 0.0% 66.2% 20.8% 55.8% 0.0% 44.2% 5.2% 94.8% Port 14 11 52 5 5 51 16 2 44 31 4 73 18.2% 14.3% 67.5% 6.5% 6.5% 66.2% 20.8% 2.6% 57.1% 40.3% 5.2% 94.8% n=77 228

PAGE 229

Surprisingly, Portuguese shows the highest pe rcentage of heavy tonic syllables due to for mation of falling diphthongs when a falling gl ide develops from the high vowel in the posttonic syllable, as in solteira unmarried woman (
PAGE 230

words represent Greek borrowings through Hum anistic Latin. The date of entry into Portuguese (15th century or later according to the Corpus do Portugus (Davies and Ferreira), together with the register of the words, accounts for the reten tion of Greek accent, in contrast with Catalan which prefers antepenultimate stress and Castil ian, where such words are reduced to four syllables through glide formation. 75 60 45 30 15 0 -15 L'LL ,HX'LLL ,LX'HLL ,LX'LLL H'LLL L'HLL ,LLX'LL ,HX'HL ,LX'HL ,LX'LL ,HX'LL H'HL L'HL H'LL ' ,H X'HLL CAT 19 5 23 20 1 0 0 10 10 014 2 CAS 000 30 1 0 12 21 18 6 454 3 571 2 7 9 21 30 1 9 12 00 0 PORFigure 4-21. Distribution of syllable types in outcom es of first declension pentasyllables with antepenultimate accent (n=77). Although there are small data sets for first de clension six-syllable (30 words) and sevensyllable (2 words) nouns the lexica l register of the words and thei r relative infrequence are not conducive for meaningful analysis. Nearly ar e all morphologically comp lex with identifiable prefixes and/or suffixes. Six of the nouns have penultimate accent in input and the remaining 24 have antepenultimate accent. As expected this latter group shows considerable syllable loss due to glide formation in Castilian a nd Portuguese (but not in Catalan) The end result for Castilian is that all now have penultimate accent while Portuguese retains antepenultimate accent in two cases. Summary of Prosodic Patterns in Nouns from the First Declension In the set of first declension nouns creation of a binary head foot with accent on the leftmost element is facilitated because they are st able with regard to preservation of the final 230

PAGE 231

unaccented syllab le. It is always maintained due to what has been assumed to be a MAXMORPH constraint although an alternate view is presented under discussion of second/fourth and third declension nouns. Loss of internal syllables is somewhat more frequent in the case of words with an odd number of syllables although this has been seen to be more a function of glide formation rather than syncope or a prosodic re pair. Within each class of nouns the emerging templates favor light syllables, a penultimate pr imary accent based on a head foot built at the right word edge, and a secondary accent based on a f oot built at the left wo rd edge, if required to avoid violation of LAPSE. Although proparoxytonic accent is pr eserved in some polysyllables it is a marked prosodic template in the languages under st udy. Furthermore, the lexical categories of the words that preserve original antepenultimate accent often correspond to specific semantic fields such as religion, philosophy, and aesthetics as well as scientific vocabulary. Therefore, these items are also more likely to be late addi tions to the language rath er than part of the patrimonial vocabulary. The presence of occas ional pairs with a common etymon demonstrates clearly the difference in treatment, for example cl v c la ae f. small key,, which gives rise to the learned form clavcula clavicle which preserves proparo xytonic accent in contrast with paroxytonic accent in patrimonial Cat. clavilla Cast. clavija and Port. cravelha bolt. Furthermore, the loss of distinctive quantity in both vowels and consonants eliminates STW as a principle for stress assignment. When th e final syllable is preserved, as is always the case in the first declension, the prototype templa te becomes the disyllabic trochee. The high ranking constraint HEADMAX (A stressed element in th e input must have a stressed element as its output correspondent) precludes movement of primary accent from the antepenultimate syllable to the penultimate for the purpose of producing a bi nary foot structure at the right word edge. There are very limited cases where this occurs and these can be ascribed to morphological 231

PAGE 232

interference or analogy. Exam ples are Castilian areola areole, aureola aureole, laureola crown of laurel in which the accent is attrac ted to the diminutive suffix perhaps through false analogy with cases like c rolla, ae f.litle crown (<*korono-la20) (see Sampson 1999, 45, for discussion on nasal assimilation). Furthermore, the vocalism of these words, that is, preservation of the diphthong /aw/ is indicative of their learned register (cf. Cat. oriol
PAGE 233

CHAP TER FIVE VOWEL LOSS AND THE RISE OF ULTI MATE ACCENT IN IBERO-ROMANCE Synchronic and Diachronic Vowel Reduction The existence of gradients of vowel reduction (No reduction Reduction Loss) in genetically related languages spoken in a cont iguous geographic area acc ounts for the relatively high incidence of ultimate stress in the Catalan nominal system as well as emerging phenomena related to suppression of unaccented vowels in European Portuguese. Of the three major Romance languages spoken in the Iberian Penins ula, only Portuguese has been described as stress-timed (Parkinson 1988, 141-142). It contrasts with phonological cons traints in Castilian and Catalan that seem to preclude the total deletion of a vocalic nucleus, although Catalan also exihibits vowel reduction. Vo wel reduction has some measur able phonetic correlates: fundamental frequency, relative intensity of F2 and F3 formants, relative overall intensity, and duration (Pickett 1999: 86). However, there are ot her characteristics that concern the syllable structure, weight and recogni tion of syllabicity that fall within the domain of phonology. It is tempting to conclude that vowel reduction only occurs in the context of stress-timing. However, Crosswhite (2000:173) ci tes studies that suggest that there are languages that exhibit characteristics of both timing systems. A study on French (Brakel 1985) opens the possibility of mixed systems. Brakel, using the framework of generative phonology, sugge sts that French is stress-timed in underlying structure yet syllable -timed in surface structure. The reduction and loss of vowels in French show th at the language is stress-timed, or at least was stress-timed at some stage in its development. Native speakers today are very much aware of the syllabic status of e muet. In careful speech, for example, to clar ify a word a regularly suppressed vowel may resurface as [ ]. In the metrical conventions of poe try and music, the vowel /e/ in final unstressed syllables is often restored. Crosswhite (2001: 173-184) suggest s that this apparent 233

PAGE 234

contradiction can be resolved by vie wing moraicity as distinct from syllabicity. This would allow a syllable containing a nonmoraic vowel to emerge as syllabic under certain conditions, such as careful speech styl e or metrical parsing. The previous discussion of first declension nouns has suggested that the final, morphemebearing syllable is widely preserved because the constraint MAXMORPH is high ranking. Retention of the final syllable is variable for second/fourth declen sion and third declension nouns. Catalan exhibits the most widespread el ision of the final vowel, followed by Castilian, and Portuguese. Although historically Port uguese limits coda consonants to to /l/, / / and /s/ (Mateus and dAndrade 2000, 52), vowel reduction in modern Eur opean Portuguese produces all manner of marked complex onsets and complex codas. This study, however, focuses on the formative period of the language when Portuguese shows a high ranking NOCODA constraint in contrast with limited retention of word fina l post-tonic vowels in Ca talan. The Catalan innovations create a prosodic template, frequent in the modern language, in which the prominent syllable of the head foot is now aligned at word edge. This new template, X ...('X), where X... represents an indeterminate number of preceding sy llables, is clearly a de parture from Latin as well as the other Ibero-Romance languages where ultimate accent occurs but is less frequent. Second/Fourth Declension Nouns Second and fourth declension nouns are tr eated together because they became indistinguishable in early Ibero-Romance. The accusative singular is characterized by o in the singular and os in the plural. Nouns in these declen sions are primarily masculine with the notable exception of m nus, s f. which in Castilian has the characteristic ending of masculine nouns, that is, mano while retaining feminine gender for purposes of concordance. Due to loss of /n/ neither Catalan nor Portuguese forms ( m and mo respectively) are readily identified with 234

PAGE 235

this declension class and fe minine gender is re tained for both, as is the case in Castilian extending even to derived forms such as diminutive manito treated as feminine despite the suffix. The differential treatment of the word final vowel sets Catalan apart from the other two languages along a dichotomy of final vowel dele tion/non-deletion. It raises a question with regard to retention of the final, unaccented vowel of the first declension, that is, whether or not retention is attributable to acoustic prominence or morphol ogical function. Third Declension Nouns The third declension characteristic vowel /e/, also present in the accusative plural of nouns, is not as readily identified as a marker of declension class as -afor the first declension and oin Late Latin for the second and fourth decl ensions. The accusative singular suffix is em Loss of final m results in a vulnerable unaccented e Third declension nouns, as those in the second, undergo apocope in the formative period of the m odern languages, to a high degree in Catalan and to a lesser degree in Cas tilian and Portuguese. The limite d cases in Portuguese depend on the acceptability of the resulting coda consonant which can only be /l/, / /, /n/ (which passes its nasality to the preceding vowel and then is ab sorbed) or /s/. On the other hand, final e is regularly lost in Catalan except when a resu lting single consonant or cluster violates NOCODA or consonant sequence constraints. There are, thus, two different factors that contribute to loss of word final -e in third declension nouns: its inherent acoustic wea kness, particularly when reduced to [ ], and its lack of essential functi on in the nominal morphological system. Examination of these differences sheds light on the status of word final vowels and the acceptability of a right aligned head foot in wh ich the prominent syllable is at word edge. A sonority hierarchy, applied to consonants and discussed above, can also be applied to vowels. Kenstowicz (1997, 158-164) suggests that for some languages two factors converge to 235

PAGE 236

determ ine the optimality of a syllabic peak. The first is vowel height which inversely corresponds to F1 as seen in Figure 5-1, repr esentative of the early Ibero-Romance vowel system, as well as modern Catalan. The second is peripherality; all the vowels labeled outside the vowel quadrilateral are periphe ral and considered more sonorous than interior vowels. In Catalan, the one central or non-peripheral vowel, [ ], emerges as a variant of /e/, / / and also of /a/ in unstressed or nonprominent positions. Figure 5-1. Vowel quadrilateral for 7 vowel systems with [ ] in nonprominent positions. A ranked order for vowels is proposed by de Lacy (2006, 286; also 2002, 55) and reproduced below as Table 5-1. De Lacy notes that in contrast to the hierarchies established for consonants (markedness, sonority), vowel sonority b ears a cause and effect relationship with the prosodic environment. The least marked vowels in a position of nonprominence are found to the left of the scale and the least marked vowels in a position of prom inence are to the right of the scale. The reduction of /e/ to [ ] is then a predictable outcome of vowel reduction in the third declension and the eventual elimination of this low sonority segment conforms to H/R. For third declension nouns total elimination of the syllable final nucleus is a cross-linguistic outcome in 236

PAGE 237

som e cases. The constraints that prevent loss are language specific and correspond to conditions of well-formedness rather than faithfulness. Table 5-1. Vowel sonority hierarchy m id high mid-high peripheral vowels high central vowels central vowels peripheral vowels mid-low peripheral vowels low vowels i y u e o a The third declension is far less regular than the first and second/four th declensions. It contains nouns of all genders; morphologi cally, there are both consonant stems and -i stems. Therefore, identity between vowel and declen sion class is not easily established. Second declension nouns, predominantly masculine and neuter, also undergo apocope and come to resemble consonant stems. Therefore, the absenc e of a final vowel cannot be equated to a zeromorpheme that represents both morphologica l class and gender in contrast with -a(m) the marker of singular accusative first declension nouns. In the first declension the thematic vowel comes to be associated with feminine gender due to the high percentage of feminine nouns, with the notable exception of Greek borrowings th at designate male occupations such as agr c la, ae, m. farmer and nauta, ae, m. sailor. Two-syllable Nouns Two-syllable nouns by default have the primar y accent on the initial sy llable. In input forms the initial syllable may correspond to all th ree-syllable types, HC HV, and L, although the last two are likely to coalesce in Late Latin. Si nce the weight of the initial syllable is not a determinant of placement of primary accent, reten tion of segments in th at syllable appears to correspond to the notion of promin ence rather than preservation of the conditions required by STW. Examination of the outputs of all disyllables in declensions 2/4 and 3 shows a high degree of faithfulness to initial heavy syllables. 237

PAGE 238

Second and Fourth Declension Dis yllables with Penultimate Accent Beginning with the set of two-sy llable Latin nouns that have a heavy first syllable with a coda consonant it is possible to construct a distribution table that shows the nature of the first syllable in outcomes across languages. It is not useful to build contingency tables similar to Tables 4-4, 4-5, and 4-6 because the loss of the post-tonic or final vowel which occurs in 110 nouns out of 167 in Catalan skews the results as seen in Table 5-2. There are no monosyllables in Castilian and only one in Portuguese, giz, from gypsum, i, n., < Gr. white lime, plaster. For example, of the 98.2% cases in Catalan with heavy syllables in Table 5-2, 65.8% correspond to monosyllables (see also Table 5-3). The resulting monosyllables in Catalan are not only of the type (C)(C)VC but also (C)(C)VCC. There are almost no monosyllables without a coda consonant and it can be argued that all monosyllables are minimally bimoraic. The one monosyllable in Portuguese, giz chalk, is CVC in structure and also bimoraic. From these data one may project as a well formedness constraint that the minimal prosodic word for all three languages consists of two moras. PrWdmin = Table 5-2. Preservation of HC initial syllables in Catalan, Ca stilian, and Portuguese nouns from Latin second and f ourth declension disyllables Resulting Syllable Type HC HV L Catalan 164 (98.2%) 0 (0.0%) 3 (1.8%) Castilian 114 (68.3%) 1 (0.6%) 52 (31.1%) Portuguese 84 (50.3%) 46 (26.5%) 37 (22.2%) n=167 Table 5-3 redisplays the data of Table 5-2 in metrical terms. For Catalan, the output of original ('H) results in a ve ry high percentage of well formed moraic trochees of the type ('H). For Castilian and Portuguese the resulting pr osodic templates are nearly all disyllabic. The less optimal ('H.L), however, still satisfies FTBIN (Feet are binary at some level of analysis) and is the prevalent pattern in this subset because, overall, the initial syllable remains heavy in a very 238

PAGE 239

high percentage of cases in both Castilian and Po rtuguese where there is almost no reduction to a single syllable. Additionally, a few l earned words in this subset such as campus have two heavy syllables but there is no com pelling reason to parse the word in any way other than a syllabic trochee, or ('H.H). For purposes of classification such cases will be treated as if they were part of the more represented type ('H.L). Table 5-3. Prosodic outcomes of La tin disyllables of the second and fourth declension with initial heavy syllable ('H). ('H) ('H.L) ('L.L) Catalan 110 (65.8%) 54 (32.3%) 3 (1.7%) Castilian 0 (0.0%) 115 (68.9%) 52 (31.1%) Portuguese 1 (0.6%) 129 (77.2%) 37 ((22.2%) n=167 Portuguese has a higher percentage of heavy tonic syllables (77.2%) compared to Castilian (68.9%) due to the pr esence of falling diphthongs and na sal vowels. In the limited cases where the initial syllable b ecomes light loss of moraic count can be attributed to various factors, among them: 1. Degemination of stops 2. Palatalization of -n.n3. Palatalization of -l.l4. Palatalization of the coda consonant in -k.t5. Loss of stop before stop as in -p.t6. Reduction of -n.s to s 7. Reduction of -s.s to s As previously discussed, palatal consonants that result from coda reduction may have a special value relative to perception of sy llable weight (Baker 2004, 228). W ith regard to the current data set this would result in the realignment of a fe w words from the ('L.L) column to the ('H.L) column for Castilian and Portuguese. However, the important difference that emerges in Table 5-3 is the tolerance of syllable /word final coda consonants in Ca talan in contrast to the more 239

PAGE 240

restrictive application of NOCODACAT, CAS, POR (A coda must satisfy language specific Sonority Sequence and other phonotactic cons traints) that is at work in Castilian and Portuguese (see Table 4-34). The loss of the word final vow el signals the demotion of NONFINALITY (No head of PrWd is final in PrWd) as a well-formedness constr aint in favor of H/R (Head foot aligned with the right word edge). Roca (2005, 356) has used these constraint s, together with FTBIN, to demonstrate differences in the prosody of Latin and Spanish non-verbs; they are reflected in Table 5-4. The constraints invoked here concer n position, primarily position of the head foot which is now able to coincide with the right word edge in Spanish, and the nature of binarity which is no longer dependent on the quantity of segments. Table 5-4. Differences in constraints governing word accent in Latin and Spanish Latin: Not allowed Spanish: Allowed Word final accent Constraint: Obligatory rightedge extrametricality (NONFINALITY) Constraint: Align head foot to the right; align prosodic word to the right (H/R) Light penults in words >2 syllables Constraint: FTBIN and NONFINALITY Constraint: H/R and FTBIN It will be seen in the case of third declension nouns that apocope also extends to Castilian and Portuguese although with more phonotactic constraints than in Catalan. In modern Catalan many of the final clusters th at appear in monosyllables are reduced to a single coda consonant but elided elements can be posited in underlying representation because of the reemergence of the final obstr uent in derived forms such as diminutives, for example /punt/ point alternately realized as ['pun] and diminutive [pun.'t t]. Herrick (2002, 70-72) proposes three optimality constraints to account for cl uster reduction of homorganic obstruents: (5.1) MAX(PLACE): Every input place feature has an output correspondent. (Herrick 2002) 240

PAGE 241

(5.2) MAX SEGMENT INPUT-OUTPUT (abbreviated as MAXIO): Every segment of the input has a correspondent in the output. (Herrick 2002) (5.3) *COMPLEXCODA (abbreviated as *COMP): Syllables must not have complex codas. (Herrick 2002) The tableau in Table 5-5 shows the relative ranking of the three constraints. Although the winning candidate violates MAXIO it emerges as the preferred fo rm because it does not violate the higher ranking constraint of MAX(PLACE). Conservation of the final cluster in cases like corb raven demonstrate that MAX(PLACE) outranks *COMP. The place features of the cluster consonants differ, [cor] and [bilabial]; therefore, both most be ma intained. It can be seen, then, that Catalan permits coda consonants and coda clusters under the lim itations described in constraints 5.1, 5.2, and 5.3. Table 5-5. Constraints on comp lex codas in Catalan: *COMP>>MAXIO /p u n t/ [cor] MAX(PLACE) *COMP MAXIO p u n | [cor] p u n t [cor] *! MAXIO is a violable constraint and in light of th e incidence of apocope visualized in gradient fashion in the previous chapter in Figure 4-1, it is violated to a lesser or greater degree in the three languages under study. The widespread vowel elision in Catalan, represented in the current data set, is the subject of two recent OT studies that address loss of atonic vowels in Western Romance. Hartkemeyer (2000) discusse s atonic vowel loss in Old French and Old Spanish while a more recent study by Wheeler (2007) examines the same phenomenon in Catalan. The relationship of apocope to word prosody is readily apparent and marks a significant 241

PAGE 242

departure from the foot building rule s of Latin. The head foot is now aligned with the right word edge and there is no extrametrical element (the class marker of the first declension). Hartkemeyer (2000, 65-72) proposes an undominated constraint HEAD-MAX-IO as well as three other ranked constraint s; all are listed below: (5.4) HEAD-MAX-IO: If is the prosodic head of the wo rd, then the output correspondent of is likewise the prosodic head of the word. (Hartkemeyer 2000) (5.5) R-ANCHOR-V: Every rightmost V in the input string has a V correspondent in the output string. (Hartkemeyer 2000) (5.6) *V: Avoid V segments in output forms. (Hartkemeyer 2000) (5.7) MAX-IO-V: Every input V segment should ha ve a corresponding V segment in the output. (Hartkemeyer 2000) The tableau in Table 5-6 shows how the constraints operate to retain Castilian and Portuguese unaccented final vowels. The example used is from the data set currently under consideration, Lat. truncus, i m. trunk > Cat. tronc Cast. tronco, Port. tronco R-ANCHOR-V also functions as a kind of morpheme pres erving constraint (cf. MAXMORPH described in above) because the final segment of Latin nouns is a class marker (resulting in early Romance as a = class 1, o = class 2, e /consonant =c lass 3). Table 5-6. Constraint against vowel deletion in Early Romance: R-ANCHOR-V /trun.ku(m)/ R-ANCHOR-V *V MAX-IO-V 'tron.ko ** 'tronk *! Demotion of R-ANCHOR-V for later stages of Catalan will allow elimination of final vowels. However, further constraints are needed to allow Catalan to retain some final vowels, albeit reduced to [ ], as has occurred in 49 instances out of 167 in the data set currently under consideration. The cases in point consist of heavy syllables w ith two consonants in the coda, occasionally three, that show increased sonority at the syllab le/word edge. Clements (1990, 303242

PAGE 243

315) em ploys the concept of the demisyllable to treat the question of sonority sequencing at syllable edges. The demisyllable is defined as a maximal sequence of tautosyllabic segments of the form Cm ... Cn V or VCm ... Cn, where n m 0. From this he is able to create a sequence of the unmarked order of segments in a demisyllable: ONLGV and its mirror image VGLNO where V = vowel, G = glide, L = liquid, N = nasa l, and O = obstruent. The numeric correlation of VGLNO is 4 3 2 1 0 with a vow el being the most sonorant at 4. Wheeler (2007, 4) has constructed a somewhat different sonority ranki ng scale, reproduced in Table 5-7, for the formative stages of Catalan following the precepts outlined in Clements. The scale is an indispensable foundation for the well-formedness constraints for syllables that follow, SONORITY SEQUENCE and SYLLABLE CONTACT LAW. Table 5-7. Proto-Ca talan sonority ranking 7 6 3 2 0 tap laterals, trill nasals obstruent voiced nonsibilant continuants sibilants plosives + non-sibilant voiceless continuant ts s l r m n v b d g f p t k z Wheelers (2007, 4) formulation of sonority sequencing echoes Clements (1990, 285) Sonority Sequencing Principle: Between a ny first or last member of a syllable and the syllable peak, only sounds of higher sonority rank are permitted. (5.8) SONORITY SEQUENCE (SONSEQ): Sonority must increase from the beginning of an onset to the nucleus of a syllable, and must decrease from the nucleus to the end of the syllable. (Wheeler 2007) There is a further constraint on sequences that are heterosyllabic where one forms a coda and the other an onset of the following syllable: 243

PAGE 244

(5.9) SYLLABLE CONTACT LAW (SYLCON): The final element of a syllable is not less sonorous than the initial element of an im mediately following syllable. (Wheeler 2007) The 49 words that retain a final vowel in Cata lan mostly do so because deletion of the final vowel would violate the preferred order of sonor ity sequencing seen in Clements formulation, VGLNO, with its proviso n m 0 as well as Wheelers similar SONORITY SEQUENCE constraint. Some of the words in the examples could be considered learned or semi-learned words and in a few cases there are popular counterparts with a different outcome such as Cat. tret, Cast. trecho both from Lat. tractum, i, n. tract. In the examples below, taken from the 49 exceptional cases, both Clements VGLNO notation and Wheelers numerical values are given. Twofold clusters NN (3, 3) ON (0, 3) OO (0, 0) himne hymn ritme rhythm tracte tract Threefold clusters NOL (3, 0, 7) OOL (2, 0, 7) centre center astre star Wheeler (2007, 5) notes that in the forms without apocope one observes the maintenance of the final vowel when the preceding consonants do not form an acceptable coda according to SONSEQ. It should also be noted that one of the re sulting heterosyllabic clusters is less than ideal in light of SYLCON, namely the stop/nasal sequence in ritme Selection of [ ] as a neutral final vowel is not unexpected. In Catala n it is the atonic allophone of /a/, / /, and /e/ and would be the usual output of first and third declension nouns where satisfaction of SONSEQ requires retention of the final vowel. As outcome of Latin ('H) the pattern ('L. L) is more prevalent in Castilian than in Catalan and Portuguese in view of the da ta in Table 5-2 and is motivated by NOCODA, a universal well-formedness constraint Retention of the mora of c oda consonants in the form of a 244

PAGE 245

glide accounts for the som ewhat lower percentage of cases of ('L.L) in Portuguese; and, in the case of Catalan, reduction and loss of the atonic final vowel creates a monosyllable that inevitably ends in a coda consonant as in lectus, i m. bed > Cat. llit Cast. lecho Port. leito Whether or not / / should be considered to occupy one or two mora slots has been previously discussed. For the moment it is considered to be onset of the fo llowing syllable although historically it results from the modification of an internal hetero syllabic cluster. It should be noted that // in Castilian tends to arise historically only after vowels with either primary or secondary accent (as in muchedumbre < multitudine(m), f., acc. sg. multitude and the above cited lecho ). For the three Ibero-Romance languages under consideration FTBIN needs to be redefined because insistence on ( ) as criterion for binarity will resu lt in a huge number of exceptions. Up to this point, only the outcomes of Latin tw o-syllable words with a heavy penult (ending in coda consonant) have been considered. The ex amination of two-syllable words from the same declension class in which the rhyme of the firs t syllable is either VV or VG (long vowel, vowel + glide) or V (short vowel) provides further eviden ce of the shifting prosodic paradigm. These three types of rhyme VV, VG, and Vesse ntially merge with the exception of a few learned/semi-learned forms that retain the La tin diphthong /aw/, for example, the outputs of austru(m) m. south wind, Cat. austre Cast./Port. austro In Table 5-8 the nature of the initial syllable in the Latin etymon is coded as L (shor t vowel) or HV (long vowel or diphthong). For the resultant Romance forms, the syllable with pr imary stress is coded HC when there is a coda consonant or HV when there is a falling diphthong. When the syllable type code is followed by m, the resulting word is either m onosyllabic or disyllabic with an initial, extrametrical syllable 245

PAGE 246

added through prothesis, as in Lat. str tum, i, n. bed-covering > Cat. es trat (cf. Span./Port. estrado). Table 5-8. Nature of accent bearing syllables de rived from Latin 2nd and 4th declension disyllables with L or HV initial syllable INPUT HC HCm HV HVm L L Catalan 0 31 0 8 19 HV Catalan 0 45 5 18 14 0 (0.0%) 76 (54.3%) 5 (3.6%) 26 (19.3%) 33 (23.6%) L Castilian 0 2 0 0 56 HV Castilian 0 1 4 0 77 0 (0.0%) 3 (2.1%) 4 (2.9%) 0 (0.0%) 133 (95.0%) L Portuguese 0 2 0 4 52 L Portuguese 0 0 5 6 71 0 (0.0%) 2 (1.4%) 5 (3.6%) 10 (7.1%) 123 (87.9%) N =140 (58 L, 82 HV) The data in Table 5-8 demonstrate two diverg ing patterns, the result of language-specific rankings of faithfulness constrai nts and H/R described above. Th e preference for coincidence of word edge and head foot is not surprising in view of studies on metrical typology (Halle 1997, Halle and Vergnaud 1987, Hyman 1977, van de r Hulst, Hendriks, and Weiter 1999) and positional strength (Beckman 1998, Crosswhite 2001, Smith 2005). The progression from word initial stress in Old Latin to moraic stress on penultimate or antepenultimate syllable has been discussed in Chapter 2. Word initial accent according to Hymans study (1977) is the most common pattern (114 languages) among the 444 languages surveyed. The second most common pattern (97 languages) is word final primary accent followed by penultimate accent (77 languages). Preference for the more common patte rn of word final accent is made manifest through the H/R constraint and lo w ranking of the faithfulness constraint which prohibits vowel deletion: MAXV (5.10) MAXV: A vowel in the input must have a corresponding vowel in the output. 246

PAGE 247

The effect of de motion of MAXV is seen in the columns labeled HCm and HVm, particularly in the case of Catalan. Apocope is the agent of change for the HCm group and to lesser extent for the HVm group. The HVm group in Catalan also results from vocalization of bilabial obstruents, thus creati ng a falling diphthong as in Lat. vum, i n. egg > Cat. ou [w ]. Formation of a diphthong also accounts for most of the cases of HVm in Portuguese but the catalyst here is loss of intervocalic /l/ and /n/. When the lateral is elided the vocal nuclei of the first and second syllables are now contiguous and glide formation provides an alternative to an onsetless final syllable as in Portuguese pau from Lat. plus, i m. stake (cf. Cat. pal Cast. palo ). Loss of intervocalic /n/ also produces a fi nal stressed syllable with a heavy nucleus but in this case the diphthong retains the nasal feature of the lost segment as in Portuguese gro from Latin gr num, i n. grain (cf. Cast. grano). However, in Catalan gra the loss of the final vowel is the trigger for loss of /n/ (Sampson 1999, 160-167) Because the nasal consonant reappears in the plural, grans it is considered to be present in underlying represen tation but does not surface due to a constraint prohibiting n]wd; thus gra is coded as a heavy syllable. Even without the supposition of underlying /n/, acoustic studies ha ve demonstrated that there is a substantive difference in both quality and duration of accent bearing vowels versus atonic vowels in Catalan (Aguilar et al.1997). The polarization by language in Figure 5-2 is striking. The L output, well represented in both Castilian and Portuguese, refl ects a high ranking faithfulness constraint that blocks vowel deletion. The coincidence of primary accent with h eavy syllable is 76.4% for Catalan in contrast wth 5% for Castilian and 12.1% for Portuguese. Elision of the final, unstressed vowel in Catalan, indicated by the column labeled N in Fi gure 5-2, creates a bimoraic monosyllable in 72.9% of the cases in this data set. The columns headed by HVm and HCm indicate 247

PAGE 248

monosyllabic outputs with a falling diphthong and t hose with a coda consonant. The treatm ent of monosyllables ending in a vowel as bi moraic has been discussed above. Figure 5-2. Nature of stressed sy llable in Catalan, Castilian, a nd Portuguese nouns from Latin 2nd and 4th declension disyllables with L or HV initial syllable (n=140). Alderete (1995, 3) proposes two constraints to account for fait hfulness to the prosodic head relying upon the precepts of McCarthy and Pr inces correspondence theory. The familiar concepts of MAX and IDENT are now used to insure faithfulness to the prosodic head. The first, MAXV, constrains against vowel deletion. High ranki ng of this constraint in Castilian and Portuguese differentiates these languages from Catalan. The second disfavors changes that would allow reduction to schwa of the tonic vowel in the head foot. However, IDENTV[+STRESS] is frequently violated historically because of loss of vowel quantity and monophthongization in the transition from Latin to Ibero-Romance. Only the HV column, containing 4 or 5 words out of 140, reflects preservation of the or iginal bimoraic nucleus. 100.0% 80.0% 60.0% 40.0% 20.0% 0.0% N L CAS 4.3% 0.7% 4.3% 90.7% 1.4% 7.1% 3.6% 87.9% POR 0.7% 0.0% 4.3% 95.0% 2.1% 0.0% 2.9% 95.0% 70.7% 3.6% 18.6% HV HVm HCm L HC HV Post-tonic Syllable Tonic Syllable 0.0% 11.4% 17.9% 54.3% 23.6% CAT(5.11) MAXV: A vowel in the input must have a corresponding vowel in the output. (5.12) IDENTV[+STRESS]: Accented vowels in the input must have an identical correspondent in the output. 248

PAGE 249

The locus of primary accent is predictably stable; it remains on the first syllable even when the final syllable is heavy as in learned campus common to all three languages. In this aspect the difference between Catalan and French is apparent. The widespread apocope that occurred in the latter resulted in a pattern si milar to Catalan, that is, the creation of a high percentage of monosyllables. However, the few cases like campus follow the standard H/R constraint for French and yield [ k.'pys ]. The faithfulness constraint proposed as undominated by Hartkemeyer ((2000, 67-72), HEAD-MAX-IO, is either demoted or inoperative in French. In summary, the two disyllabic data sets studied so far (outcomes of first declension disyllables and outcomes of second/fourth declension disyllables) demonstrate crucial differences between Latin and Ibero-Romance as well as differences among the three languages under study. Construction of the head foot is now at right word edge and the class marker is no longer construed as an extrametrical syllable. Constraints governi ng post-tonic unstressed vowels (although probably not as broad as *V ) delete final, atonic vowels in Catalan, occasionally in Castilian, and almost never in Portuguese providing SONSEQ is not violated. These differences emerge in nouns of the second and fourth decensions. Figure 5-3. Prosodic outcomes of all declensi on 1 and declensions 2/ 4 disyllables (n=539). 0 ('H.X) ('L.X) ('H.X) ('L.X) DECL 1 DECL 2/4 ('H ) 144 88 160 134 13 POR CAT 212 59 36 88 144 149 83 185 119 3 CAS50 100 150 200 250 249

PAGE 250

In contrast with the outcom es of nouns in decl ensions 2/4, the treatment of first declension nouns is very uniform across languges. Figure 5-3 high lights the points of divergence and coalescence; here the outputs of declensions 2/4 and declen sion 1 are shown side by side (n=539). Third Declension Disyllables with Penultimate Accent Table 5-9 shows the outcome of the subgroup of third declension disyllabic nouns with a heavy first syllable ending in coda consonant (R esulting monosyllables are indicated by m). Although this data set is relatively small (n= 57) the results are sim ilar to those for the comparable set of 2nd/4th declension nouns. Possi ble results of original third declension HC1.L are 'HC.L, 'HC(C), 'HV.L, 'HV, and 'L.L. Loss of syllable final -e in is prevented by high ranking NOCODA in both Castilian and Portuguese. The conditions under which a coda is allowed are language specific and less stringent for Castilian th an for Portuguese as seen by comparison of the HC and HV columns in Table 5-9. Only 24 (42.1%) of Portuguese disyllables have a coda consonant when the first syllable is heavy; an additional 21 (36.8%) have either a nasal vowel or a falling diphthong whereas Cast ilian has no heavy syllables based on a falling diphthong. In other words, vocalization of an obstr uent in coda position wo rd internally is an infrequent outcome in Castilian. Table 5-9. Accented syllable by type in the output of 3rd declension disyllabic nouns with initial HC syllable HC HCm HV HVm L CAT 13 (22.8%) 42 (73.7%) 0 (0.0%) 0 (0.0%) 2 (3.5%) CAS 41 (71.9%) 9 (15.8%) 0 (0.0%) 0 (0.0%) 7 (12.3%) POR 24 (42.1%) 3 (5.3%) 21 (36.8%) 1 (1.8%) 8 (14.0%) m indicates a monosyllable n=57 The one case of a Portuguese HV type monosyllable is p, resulting from pulvus, a competing form with pulvis, ris m. dust (cf. Castilian polvo, DRAE s.v. polvo). Catalan pols is attributed to Vulgar Latin pulvis which was apparently treated as a neuter noun with identical 250

PAGE 251

nom inative and accusative (Gran diccionari de la llengua catalana s.v. pols). In defense of HV status for this monosyllable, Portuguese, /o/ and / / are realized as mid vowels only in accented syllables. Costas study of relative vowel lengt h in European Portuguese (2004) shows that the duration of //, the vowel nucleus in p is relatively longer (1.08 ratio) than the mean of all vowels in stressed CV syllables. Increase in vowel duration is negatively correlated with vowel height, that is, low vowels are longer. To what degree prosodic weight is preserved in accented syllables is shown in Table 5-10. Here, prosodic faithfulness to the input form ('H). is related to language and declension class. The HC and HV columns from Table 59 are combined under ('H.L) and the HCm and HVm columns are combined under ('H). Both ('H.L) and ('L.L) are frequent outcomes in Castilian and Portuguese. The difference between declension classes is small in Castilian and Portuguese, while in Catalan the incidence of a pocope is more prevalen t in third declension nouns. Nouns derived from second/ fourth and third declension disylalbles with HC initial syllable also show in Catalan a high degree of fi nal vowel elision which re sults in a monosyllabic output. Table 5-10. Comparison of prosodic outcomes of Latin disyllables with HC initial syllable from the 3rd and 2nd/4th declensions ('H). Declension Language ('H) ('H.L) ('L.L) 3 Catalan 42 (73.7%) 13 (22.8%) 2 (3.5%) 2/4 Catalan 110 (65.8%) 54 (32.3%) 3 (1.7%) 3 Castilian 9 (15.8%) 41 (71.9%) 7 (12.3%) 2/4 Castilian 0 (0.0%) 115 (68.9%) 52 (31.1%) 3 Portuguese 4 (7.0%) 45 (78.9%) 8(14.0%) 2/4 Portuguese 1 (0.6%) 129 (77.2%) 37 ((22.2%) Declension 2/4, n=167 Declension 3, n=57 251

PAGE 252

W hen the word final vowel is retained as a result of phonotactic constraints such as SONSEQ, this vowel differs in both quality and du ration from its tonic counterpart (Recasens i Vives 1986, 149150). In Catalan, resulting ('L.L) words account for only 3.5% of cases from 3rd declension disyllables and 1.7% from 2nd/4th declension disyllables. When the data in Table 5-10 is displayed in a line graph (Figure 5-4) based on percentage values corresponding to the distribution of the three prosodic templates the areas of divergence by language are readily apparent, that is, the preference for a heavy mono syllable in Catalan and the emergence of ('L.L), a more harmonious trochee, in Castilian and Port uguese. Cross-linguistica lly, faithfulness to the prosodic head is high; the majority outcome for all three languages is a heavy initial syllable. 100.0% 80.0% 60.0% 40.0% 20.0% 0.0% POR CAT 2/4 DECL 14.0% 12.3% 3.5% 78.9% 71.9% 22.8% 7.0% 15.8% 73.7% 3 DECL 22.2% 0.0% 0.6% 32.3% 68.9% 77.2% 1.7% 31.1% CAS POR CAT CAS POR CAT CAS ('H.L) ('L.L) ('H) 65.8% Figure 5-4. Com parison of percent values of prosodic outcomes of 2nd/4th and 3rd declension disyllables with HC initial syllable. Divergence occurs precisely in the two templates that violate faithfulness constraints. In the case of ('H) one or more segments have been lost at the right word ed ge, notably the syllable nucleus and occasionally the onset of that same syllable. Loss of the syllable nucleus through apocope results in a complex c oda that violates language speci fic constraints of the type *COMPLEXCC]wd (cf. with the more general *COMPLEX constraint in Prince and Smolensky 1994, 252

PAGE 253

108). The final consonant of the CC cluster is also lost as seen in the outcom e of Lat. calx, calcis f. stone where only Catalan has a complex coda, cal pronounced [ k s], in contrast with Cast./Port. cal which are also heavy monosyllabl es but with a single post-nuclear consonant. Because complex codas occasionally em erge in word internal position as in the outputs of monstrum, i, n monster, word edge is specified. (5.13) *COMPLEXCC]wd : No more than one C may occur in the coda at the right word edge. Another strategy to avoid a complex cluster at th e end of a word is to retain the final vowel which occurs in some cases (22.8-32.3%) in Catalan such as centre center (
PAGE 254

requires that all elem ents of the input be presen t in the output. Reducti on of an initial heavy syllable to L is a result of either monophthongi zation or loss of distin ctive vowel quantity. Syllable weight is only occasionally preserved in cases of the diphthong /aw/. Nevertheless, the foot form ('L.L) contains two optimal syllables (no codas) and also corresponds to a preferred configuration for a syllabic trochee. Furthermore, this template is expected to be the product of a high percentage of nouns, at leas t in Castilian and Portuguese, re sulting from inputs ('L). and ('HV).. Since monopthongization of Latin /a j/ to /e/ and /aw to /o/ (for discussion of vowel quality and length see Lloyd 1987, 106-107) had already occurred in proto Ibero-Romance in the patrimonial vocabulary and vowel quantity was no longer phonemic it is appropriate to consider that these two templates had effec tively coalesced; therefore, they are combined in Table 5-11. Table 5-11. Accented syllable by type in the output of 3rd declension disyllabic nouns with initial HV/L syllable HC HCm HV HVm L Lm CAT 0 18 0 20 28 0 CAS 2 18 1 3 41 1 POR 0 12 1 7 45 1 m indicates a monosyllable n=66 The one faithful output in terms of an initial HV syllable is Castilian/Portuguese fraude < fraus, fraudis, f. fraud (cf. Catalan frau), a legal term. The monosyllable that emerges in both Castilian ( pie ) and Portuguese (p ) is the output of Latin p s, p dis, m. foot and is coded here as Lm although arguments could be made that the vowel of a monosyllabic stressable word differs both quantitatively and qualitatively from the same vowel in unaccented position (for Spanish see Quilis 1983, 243 and Monroy Casas 1980: 44-47; for Portuguese see Delgado Martins 1988, 128-132). The fact that in some cases the result is a falling diphthong, as in 254

PAGE 255

Catalan peu (

Port. exame As in Table 5-11, Table 5-12 compares the out put in prosodic terms of second/fourth and third declension disyllables with original HV or L t ype initial syllable. It is expected that there will be very few disyllabic outcomes with an initial heavy syllable and this is reflected in the fact that the ('H). column, the most faithful reflection of ('HV). input, is scarcely populated. Retention of diphthongs is limited only to a fe w learned or semi-learned words and vowels no longer have contrasting phonemic length. 255

PAGE 256

Table 5-12. Com parison of prosodic outcomes of Latin disyllables with HV and L initial syllable from 3rd and 2nd/4th declensions ('HV). ('L). Declension Language ('H) ('H). ('L). 3 Catalan 38 (57.5%) 0 (0%) 28 (42.4%) 2/4 Catalan 102 (72.9%) 5 (3.6%) 33 (23.6%) 3 Castilian 22 (33.3%) 3 (4.5%) 41 (62.1%) 2/4 Castilian 3 (2.1%) 4 (2.9%) 133 (95.0%) 3 Portuguese 20 (30.3%) 1 (1.5%) 45 (68.2%) 2/4 Portuguese 12 (8.6%) 5 (3.6%) 123 (87.9%) Declension 2/4, n=140 Declension 3, n=66 The unexpected difference in percentage of heavy monosyllables in Catalan can be explained by several factors. Th e first is that many of the thir d declension nouns are learned or semi-learned words that have faithfully retained th e two syllables of the Latin nominative. In the case of imparisyllabic nouns the nominative and accusative are quite different as seen in the case of Cat. llapis, Cast. lpiz and Port. lpis from Lat. l pis, dis m. stone. In other cases, the second syllable has a complex onset. Apocope would then result in a violation of the SONORITY SEQUENCE constraint. Therefore, in nouns such as Cat./Cast./Port. odre (
PAGE 257

100.0% 80.0% 60.0% 40.0% 20.0% 0.0% CAS POR POR Figure 5-5. Com parison of percen tage values of prosodic outcom es of declensions 2/4 (n=140) and 3 (n=66) with HV or L initial syllable. Figure 5-6 shows the emerging prosodic templa tes for Catalan, Castilian, and Portuguese nouns (N=430) from Latin second/fourth and thir d declension disyllable s. The association between syllable weight, applicable to both ('H .L) and ('H) templates, and locus of accent is very strong for Catalan and corresponds to 72.7% of th e nouns; for Castilian it is only 45.8%; and for Portuguese it is 50.5%. The conve rgent patterns of the first de clension disyllables, both with regard to number of syllables and preservation of heavy syllables from input forms do not carry over to declensions 2/4 and 3. In the case of first declension disyllables both Catalan and Portuguese show 37.9% of nouns with coincidenc e of primary accent and heavy syllable and 62.1% of nouns with a light accent bearing syllable. For Castilian the percentage of heavy syllables is slightly lower at 35.8% with a corresponding 64.2% light, accented syllables. In Figure 5-6 it can be seen that the percentage of light syllables with primary accent hovers at about 50% for both Castilian and Portuguese, that is the optimal troche ('L.L) is considerably less represented in these declensions than in the first. 2/4 DECL 72.9% 2.1% 8.6% 3.6% 2.9% 3.6% 23.6% 95.0% 87.9% 3 DECL 57.6% 33.3% 30.3% 0.0% 4.5% 1.5% 42.4% 62.1% 68.2% CAT CAT CAS POR CAT CAS ('H) ('H.L) (' L.L) 257

PAGE 258

Figure 5-6. Percentage distribution of prosodic outcom es of all declension 2/4 and declension 3 disyllables (n=430). Perhaps more important is the demotion of the constraint MAXMORPH which allows deletion of the word final vowel. High ranking of MAXMORPH in first declension nouns may be motivated by the fact that the morpheme in question is embodied by a vowel of high sonority. The final /a/ vowel is, in fact, th e only one that is systematically preserved in Catalan. The set of two-syllable input forms analyzed up to this point has shown two possible outcomes preservation of both syllables with accent by defa ult on the initial syllable, or deletion of the nucleus of the final syllable resulting in a h eavy monosyllable in near ly all cases. While reduction to a single syllable violates MAXIO through loss of segments it does maintain the minimum moraic requirement for a word and constitutes a well-formed trochee. The fact that word level accent coincides with a heavy syllable is by default rather than plan: every prosodic word must have a word-level accent. The anal ysis of words with th ree or more syllables demonstrates that coincidence of word accent with a heavy syllable is no longer a paramount consideration. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% CAS CAT POR 292 34 ('H) 37 163 ('H.L) 72 180 233 213 66 ('L.L) 258

PAGE 259

Three-syllable Nouns The outcomes of nouns of the second/fourth and third declensions present starkly diverging patterns in eastern Ibero-Romance and western Ibero-Romance as seen in the discussion of disyllables. Widesp read loss of final vowels is the norm for Catalan as long as the resulting coda does not violat e coda constraints or sonorit y sequencing. Castilian and Portuguese, as discussed above, are more restrictive with regard to acceptable word final codas with Portuguese being the least tolerant of c oda consonants. The possible accentual patterns for three-syllable nouns in input (Figure 5-7) are re produced here for convenience. Analysis will begin with the templates for penultimate accent. A) Penultim ate Accent HC HV L .HC HV . B) Antepennultim ate Accent HC HV L .L . Figure 5-7. Prosodic template s for three-syllable nouns. Trisyllables with Penultimate Accent Because of the importance of apocope in the outcome of input forms from the second and fourth declension classes, active constraints for word final consonants in each language are presented in Table 5-13. An asteri sk indicates prohibition of a particular type of coda and dashes indicate that there are no constraints that bar codas of th at type. It is understood that the constraints prevent consonants in the input from appearing in c oda position at the end of the word, or ]wd. It is not necessary to build constrai nts for all possible se gments. To predict historical apocope and syncope Wheeler (2007) utilizes SONSEQ to designate the preferred sequence of consonants at the sy llable right edge; these are co mbinations of s+stop, nasal+stop, 259

PAGE 260

liquid+stop, stop+s, nasal+s, liquid+s. Outcom es that violate th is constraint are thus blocked and result in retention of the fina l vowel, usually reduced to schwa and represented orthographically by e. Table 5-13. Constraints governing apocope in Ibero-Romance Catalan Castilian Portuguese Stops *VOICED *NON-CORONAL *CORONAL, -VOICE *STOPS Affricates *VOICED *AFFRICATES *AFFRICATES Sibilants *VOICED --*VOICED Fricatives (non-sibilant) *VOICED *LABIAL *LABIAL Nasals *CORONAL *LABIAL, PALATAL *NASALS Liquids *V R]wd 1 *PALATAL *PALATAL Glides ------Complex Codas SONSEQ MAX(C)/CONTR:F *C[nasal]C[stop] *COMPLEX *COMPLEX 1The deletion of final // is not uniform in Catalan. It occurs with regularity in certain morphological classes such as the infinitive. In the case of nouns elision is not entirely predicatable; Wheeler (2005, 34) suggests that final rhotic deletion is subject to dialect differences and is le xically conditioned. The clusters that arise during the formative pe riod of Catalan are not necessarily stable. In modern Catalan there is a general prohibi tion against homorganic cl usters consisting of nasal+stop. Clusters consisting of liquid+stop have variable re sults, with deletion occurring in some cases but not others, for example the DC VB (s.v. alt, falt) gives two pronunciations for alt old, both [ 'a t ] and [ 'a ] but only one for falt lacking, [ 'fa t ]. On the other hand, surface clusters of liquid+stop that de rive from an underlying voiced st op rarely show deletion of the final stop, for example [ 'su t ] sord, masc. deaf (cf. sorda, fem.). Ct (2004, 22-26) suggests that a constraint of the type MAX(C)/CONTR:F can account for the differential treatment in some cases. (5.14) MAX(C)/CONTR:F: Do not delete a consonant that contrasts in some feature F with an adjacent segment. (Ct 2004) 260

PAGE 261

For exam ple, in most varieties of Catalan font fountain is [ f n] but fort masc. strong is [ f t ]. Here the crucial feature would be an teriority, absent in the rhotic. MAX(C)/CONTR:F does not solve the dilemma of alt with and without the final stop21 but it does provide a good vehicle for differentiating the treatment of the final stop in homorganic nasal+stop cl usters from that in liquid+stop clusters. Trisyllables with accent on a penultimate HC type syllable Having established that constraints related to apocope vary from language to language the disparities seen in the analysis of the first data set from input (H C)('HC) are fairly predictable in terms of the fate of the final syll able. Since there are few constraints to prevent apocope in Catalan other than the well formedness constraint SONSEQ it is expected that there will be a high degree of word final vowel deletion if the preceding consonant or consonant sequence is a permissible coda. Table 5-14 provides confirmation of this expected outcome. Table 5-14. Output of second/ fourth declension (HC)('HC) by number of syllables and nature of accented syllable (HC)('HC) 3 SYLLABLE 2 SYLLABLE HC HV L Subtotal HC HV L Subtotal CAT 23 (28.0%) 0 (0.0%) 0 (0.0%) 23 (28.0%) 58 (70.7%) 0 (0.0%) 1 (1.2%) 59 (72.0%) CAS 57 (69.5%) 0 (0.0%) 23 (28.0%) 80 (97.6%) 1 (1.2%) 0 (0.0%) 1 (1.2%) 1 (2.4%) POR 43 (52.4%) 14 (17.1%) 23 (28.0%) 80 (97.6%) 1 (1.2%) 0 (0.0%) 0 (0.0%) 1 (2.4%) n=82 Apocope has occurred in 72% of the cases in Catalan resulting in a monosyllabic head foot aligned with the right word edge. The cases of non deletion reflected in the 28% of nouns that 21 Hualde (2007, 361) suggests that the treatment of final liquid+stop clusters functions on a register/dialect axis, for example, he notes that the final t of alt high (MASC) is pronounced in Valencia and Majorca, but not in Catalonia (Central Catalan), which does not allow this final cluster. In the plural alts high (MASC PL) the orthographic t is not pronounced in any dialect. On the other hand, in part part, the t is only omitted in informal speech in Catalonia, and consistently preserved in the other two geographical dialects, but in its plural parts the t is systematically deleted both in Catalonia and Majorca and pronounced only in Valencia. 261

PAGE 262

retain a vocalic nucleus at the right w ord edge contain the following internal clusters: -kt-, -pt-, -sm-, -ps-, -mpl-, and -mpt-. It can be readily seen how each would violate SONSEQ. The case of -psmerits some attention because its disquali fication here represents a degree of opacity in that it is found in surface clus ters such as the plural of cap head ( caps ). However, the presence of a morpheme boundary differentiates the two cases: *col laps vs. cap]morphs. Catalan is remarkable for its preservations of all coda consonants with the possible exception of transversum, i neut. traverse > travs (in all three languages) for which Lewis and Short gives an alternate form with a long vowel and no nasal consonant in coda position, tr versus. Although DCVB, DRAE, and the Dicionrio da lngua portuguesa all give transversum as etymon, it seems likely that the input form in Ibero-Romance had al ready lost the coda consonant. The expected similar patterning of Castilian and Port uguese is seen in Figure 5-8 with some deviation rela tive to the retention of moras in the accented syllable. Portuguese preserves the feature [+nasal] of a deleted nasal c oda consonant resulting in a nasal vowel that is appreciably longer than or al vowels, for example ( accentus, s m.>) acento accent realized as [ 's .tu ]. In a few instances a glide replaces a coda consonant creating a fa lling diphthong as in (d spectus, s, m.>) despeito spite. There are virtually no cases of apocope in Castilian and Portuguese other than the already cited travs traverse and (crystallum, i, n.>) cristal crystal (cf. final palatal // in Catalan). Therefore, nearly a ll output forms for these languages are trisyllables. The trisyllables in Catalan (28.0%) reflect blocking of vowel deletion by SONSEQ, that is, a configuration of (C)(C)VC1C2 in the final syllable in which C2 is more sonorous. 262

PAGE 263

0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% CAT 28.0%0.0%0.0%72.0%0.0%1.2% CAS 69.5%0.0%28.0%1.2%0.0%1.2% POR 52.4%17.1%28.0%1.2%0.0%1.2% HCHVLHCHVL 3 syllable 2 syllable Figure 5-8. Comparison of syllabi c count and nature of the sy llable with primary accent in outcomes of second/fourth declension (HC)('HC) (n=82). In the case of ('HC) input, treated previ ously, loss of the final syllable in Catalan produces a heavy monosyllable, a well-formed bimora ic foot. Loss of the vocalic nucleus of the ultimate syllable in the current trisyllabic data set should produce two heavy syllables. However, because the head syllable is aligned with the right word e dge the new pattern gives the appearance of an iamb consisting of two heavy syllables. The preferred form for the iamb, a quantity-sensitive right dominant foot (Hayes 1995, 75-76) is (L.'H). Given the nature of unaccented vowels in Catalan there may actually be a phonetic correlate that corresponds to the prosodic requirement of this foot type. The pred ominance of this new pattern, two syllables with stress on the rightmost, has led Wheeler (2004) to propose that the prosodic template for Catalan is iambic. He accomplishes this through insertio n of another element in the prosodic hierarchy between foot and word, the colon. For Catalan the colon is binary with right alignment. Wheeler (2004, 12) proposes the following optimal parsing (curly brackets indicate colon) for 263

PAGE 264

H.'H.L, parallel to the da ta set under discussion: {(H)}{(' H)L}. It is selected over {(H)('H.L} because the latter violates the constraint IAMB. The difficulty with Wheelers proposal becomes apparent in the case of three-syllable input with penultimate accent when the first syllable is li ght. In such cases it is necessary to admit ('L) as a well formed foot which Wheeler (2004, 7) ju stifies on the basis of monosyllables such as fe faith and pla flat. However, pla has a plural in plans and the quality of the vowel in accentbearing and non accent-bearing syllable s varies considerably. Thus, in Catalan the letter of the alphabet d is ['de] but the preposition de of, from is [d ]. While the rhyme of the former may not constitute a heavy syllable in the traditional sense there is a clear difference between accentbearing and non accent-bearing vowels in terms of th e extent and nature of the inventory. There are also differences in duration for those vowels which are undifferentiated in quality, namely /i/ and /u/, according to the study conducted by Agu ilar, et al. (1997) which found that on average stressed vowels in non-prepausal position were 12.5% longer than their unstressed counterparts and in prepausal position 8.9% longer. The vowel which can only appear in atonic syllables, [ ], was the shortest of all vowels with a mean duration of 94.3 ms. Its tonic counterparts include the vowels of greatest average duration, [ ] and [a], at 126.3 ms, and [e] at 119.6 ms. This study also shows that there is a relationshi p between vowel length and height. It can be argued, then, that a single light syllable, L, differs substantively in quality and quantity dependent on its prosodic prominence (strong/weak). Therefore, while (L)w may not be a well formed foot, (L)s, is. One other question of interest is the fate of the initial heavy syllable in (HC)('HC). In the first declension data set diffe ring rates of preservation of c oda consonant were observed for the initial syllable and for the syllable with primary accent. Wheelers broader WHC (Weight264

PAGE 265

to-Head) co nstraint calls for a heavy syllable to h ead a foot but it does not require assignment of primary accent to a heavy syllable. (5.15) WHC (Weight-to-Head Constraint): Every heavy syllable must be designated as the head of some foot. (Wheeler 2007) WHC as formulated here is well suited to the concept of positional prominence because it rewards construction of a foot at left word edge. Table 5-15 co mpares the incidence of heavy syllables in word initial positi on and the penultimate or accent bearing syllable resulting from (HC)('HC). Although the initial syllable does not bear the primary accent a high degree of preservation of its moraic count is expect ed in congruence with the idea of positional prominence. The constraint WHC also raises prob lems in parsing the output of trisyllabic words with penultimate accent such as (gr nlum, i, n.>) grnul granule. Clearly, the final syllable should not constitute a foot; rather, it is the weak syllable in a syllabic trochee of the form (' L .H). Table 5-15. Comparison of heavy/light syllables in initial and accent bear ing syllables in output of second/fourth declension (HC)('HC) (HC)('HC) (HC)('HC)L (HC)('HC) H L H L H L H L CAT 19 (82.6%) 4 (17.4%) 23 (100.0%) 0 (0.0%) 51 (86.4%) 8 (13.6%) 59 (100.0%) 0 (0.0%) CAS 61 (76.3%) 19 (23.8%) 57 (71.3%) 23 (28.8%) 1 (50.0%) 1 (50.0%) 2 (100.0%) 0 (0.0%) POR 57 (72.6%) 23 (27.5%) 53 (66.2%) 27 (33.8%) 1 (50.0%) 1 (50.0%) 2 (100.0%) 0 (0.0%) Note: Percentages are based on the number for each subgro up: 3 syllables (Catalan 23; Portuguese, Castilian 80), 2 syllables (Catalan 59; Portuguese, Castilian 2). n=82 The data for the two-syllable pattern are of diminished inte rest because the results are skewed by language; only Catalan is well repres ented for this pattern. Although Catalan is somewhat underrepresented in the three-syllable pattern it is still possible to observe and 265

PAGE 266

com pare outcomes. The three languages show small differences relative to retention of coda in the word initial syllable and relatively greater differe nces in the syllable with primary accent. It should be remembered that 14 of the 57 word ini tial syllables labeled hea vy in Portuguese result from nasal vowels and 19 of the 53 accented syllables are also heavy for the same reason. Overall, the maintenance of bimoraic status in th e word initial syllable is high and comparable to patterns previously observed in fi rst declension nouns. A large per centage of the initial syllables that are light in all three languages are the result of degemination as in the case of annexus and affixus The use of colon described above seems to be an unnecessary artifice to obtain the desired output from (HC)('HC). Table 5-16 attempts to correctly select the optimal candidate by replicating the constraints used with first declension nouns in Table 4-29. A small capital V is used for the vocalism of the final syllable nucle us. During the formative stage of Ibero-Romance there is little evidence of aspiration of /s/ in coda position so a candida te with aspiration or elision is not put forward. Accepting an uneven trochee ('H.L) as a possible foot, candidate A emerges as the winner. Although loss of coda in the syllable with primary accent does produce a better formed trochee, ('L.L) (Prince 1990, 367-373) it violates STW. Removal of STW would result in a tie for candidates A and C. Table 5-16. Tableau for output of declension 2/4 trisyllable with penultimate accent on HC type syllable ASPECT FTTROCH HEAD MAX MAX MORPH STW PARSEALIGNR a. as('pek.tV) b. ('as)(pe.tV) *! c. as( 'pe.tV) *! In order to account for the de viations seen in the three la nguages, that is, moraically faithful Catalan aspecte, coda-less tonic syllable in aspecto in Castilian informal register, and 266

PAGE 267

coda-less tonic syllable in aspecto in Portuguese despite the or thography (Brazilian Portuguese now accepts aspeto as standard orthography), it is necessa ry to develop faithfulness constraints that reflect language-specific phonot actics or criteria for a well-formed syllable. Previously introduced NOCODACAT, CAS, POR fulfills this function. Tables 4-34 and 5-13 provide languagespecific restrictions on coda consonants. Table 5-17 utilizes NOCODACAT, CAS, POR and eliminates STW, no longer required to achieve the desired result. MAXMORPH is clearly not operative in this data set in stark contrast to the first declension nouns. A more gene ral faithfulness constraint MAX (A mora which is present in the input is present in the output) is used to account for the effect of nasaliz ation and vocalization of /k/ in coda position which distinguish Portuguese from Castilian. This new constraint seeks to preserve moraic count but does not require identity of input and output segments. Castilian has two reflexes here, CAS1 and CAS2. The first indicates careful speech and the second a more popular register. For the moment, the vocalism of unstressed vowels is addressed by a constraint similar to NOCODACAT, CAS, POR. It takes the form V[-acct] CAT, CAS, POR. (5.16) V[-acct] cat, cas, por: A vowel which is not accent bearing is realized according to language specific constraints for vowels in nonprominent positions. Table 5-17. Tableau for trisyllables with penultimate accent (declensions 2/4 and 3) CONDUCTU FT TROCH HEADMAX NOCODACAT, CAS, POR V[-acct] CAT, CAS, POR MAX PARSEALIGNR a. kun('duk.t )CAT *CA2S, POR *CAS, POR b. kon('duk.to)CAS1 *POR, CAS2 *CAT, POR c. *CAT, POR *POR, CAS1 kon('du.to)CAS2 d. k('du.tu)POR *CAS, CAT ** The corresponding set, (HC)('HC) input, fro m the third declension is very small and consists of only 19 items. The results are also divided into two and three-syllable outcomes along the familiar east/west divide between Catalan and Castilian/Portuguese. In fact, in this set 267

PAGE 268

no disyllables resu lt for either Castilian or Port uguese. Distribution of outcomes in terms of syllable count and heavy/light syllables from orig inal HC syllables is displayed in Table 5-18. This small data set contains a high number of items with /-n.t-/ seque nce between the ultimate and penultimate syllable. This is an acceptable sequence in terms of SONSEQ but violates NOCODA in Portuguese. It has be en argued, however, that the process of vowel nasalization is mora preserving because the resulting nasal vo wel is appreciably longer than a corresponding oral vowel. However, loss of the word final vowel in Catalan triggers adjustment of the /-nt/ final word sequence because it violates *COMPLEX for Catalan. Complex codas are permitted in Catalan only to satisf y the higher ranking MAX(PLACE) constraint: Every input place feature has an output correspondent. Table 5-18. Comparison of heavy/light syllables in initial and accent bear ing syllables in output of third declension (HC)('HC) (HC)('HC) (HC)('HC)L (HC)('HC) H L H L H L H L CAT 4 (21.1%) 1(5.3%) 5 (26.3%) 0 (0.0%) 12 (63.2%) 2 (10.5%) 14 (73.7%) 0 (0.0%) CAS 14 (73.7%) 5 (26.3%) 18 (94.7%) 1 (5.3%) 0 (0.0%) 0 (0.0%) 0 (0.0%) 0 (0.0%) POR 13 (68.4%) 6 (31.6%) 18 (94.7%) 1 (5.3%) 0 (0.0%) 0 (0.0%)0 0 (0.0%) 0 (0.0%) n=19 Despite the skewed results in terms of th ree-syllable versus two-syllable outcomes, Figure 5-9 shows that the distribu tion of heavy/light syllables is quite similar. The accented syllable in this data set retains it s moraic count at a higher rate th an the word initial syllable. 268

PAGE 269

20 15 10 5 0 H L H L Figure 5-9. Com parison of syllabl e count and nature of the sy llable with primary accent in outcomes of third declension (HC)('HC) (n=19). However, due to the small size of this data set it is difficult to establish th at the pattern of coda retention in initial and accented syllables of thir d declension trisyllables with penultimate accent is appreciably different from that of second/fourth declension trisyllables with penultimate accent. The next data sets examine the outcomes of tr isyllables with penultimate accent where the accented syllable has a coda consonant but th e preceding initial syllable is light. The corresponding input forms are HV('HC) and L('HC) . Again it is assumed that nearly all HV type syllables become L syllables during the formative period of the Ibero-Romance languages. For example, the one HV syllable in tr isyllabic outcomes in Catalan is semi-learned autumne autumn with initial syllable /aw/ in contrast with Cast. otoo, Port. outono both with initial syllable /o /. The other HV initial syllables in Portuguese are secondary developments, nasal vowels and falling diphthongs which have their origins in coda consonants. For declensions 2/4 Catalan, predicta bly, shows a high incidence of apocope in this group of 76 nouns; there are 51 cases (65.8%), compared with only 2 in Castilian and 3 in Portuguese. CAT 4 1 5 0 12 2 14 0 CAS 14 5 18 1 0 0 0 0 POR 13 6 18 1 0 0 0 0 L HL H S3 S2 S2 S1 3 SYLLABLE 2 SYLLABLE 269

PAGE 270

Figure 5-10 shows the distribution of trisyllabic and disyllabic outcomes by language and nature of accented syllable. 100.0% 80.0% 60.0% 40.0% 20.0% 0.0% HC HV L HC HV LFigure 5-10. Outcom es of HV('HC) and L( 'HC) in declension 2/4 nouns: Syllable count and nature of accented syllable (n=76). Encircled numbers in Figure 5-10 mark the coincidence of heavy syllable and primary accent in the most common output patterns (exclu ding the two-syllable penultimate output). When these percentages are combined the aggr egate totals are Cata lan, 98.7%; Portuguese, 65.7%; and Castilian, 64.4%. The slightly highe r percentage for Portuguese results from one case, anel (cf. Cast. anillo < nellus, i, m. ring). The anomalous two-syllable outcomes with penultimate accent are the result of syncope (Cat. and Port. mestre < m gister, tri, m. master; Cast. sello Port. selo < s gillum, i, n. seal) and glide forma tion (Cast. duelo < duellum, i, n. war; Cast. diurno < diurnum, i, n daily portion). The extremely high association of heavy syllable and tonic syllabic arises in Catalan as a consequence of apocope. Loss of the nucleus of the final syllable renders the preceding syllable heavy even if a coda consonant has been lost motivated by a NOCODA constraint. The input output stri ngs in idealized form correspond to CAT 31.6% 0.0% 0.0% 65.8% 1.3% 0.0% 1.3% 0.0% CAS 61.8% 0.0% 31.6% 2.6% 0.0% 0.0% 1.3% 2.6% POR 44.7% 17.1% 31.6% 3.9% 0.0% 0.0% 1.3% 1.3% HC L 3 SYLL/PENULT 2 SYLL/ULTIMATE 2 SYLL/PENULT 270

PAGE 271

(C)V.' CVC2.C1V (C)V.'CVC1. Loss of C2 is compensated by the movement of C1 to coda position following the loss of the nucleus of the ultimate syllable. The corresponding set of third declen sion nouns from inputs HV('HC) and L('HC) is much smaller and consists of 29 items This set is uniform in having a light initial syllable in all outputs. As expect ed there is a high incidence of apocope in Catalan in contrast with only one case in Portuguese and none in Cast ilian. There are no case s of reduction to two syllables with resulting from syncope or glid e formation (with preservation of penultimate accent). The coincidence of primary accent with a heavy syllable is 100% in this subset. Figure 5-11 shows distribution of the tw o pervasive patterns: L('H.L) and L('H) and provides detail on the nature of the heavy syllable. Encircled numbers indicate majority outcome by language. Reduction to the two-syllable pattern corresponds to 65.5% of the outputs in Catalan and only 3.4% in Portuguese ( biso < b son, ontis, m. wild ox; cf. Cat. bis ). The one lexical item in this group not reflected in Figure 5-11 is Lat. d rectum, i n. straight line; la w and its outcomes in Ibero-Romance, Cat. dret Cast. derecho and Port. direito Outcomes in Ibero-Romance suggest competing etyma. The syncope of Cata lan and the vocalism of Castilian are best explained by positing an alternate form with a light first syllable and qualitatively different vowel. Palatalization of the medial consonant gr oup in Castilian and leniti on of syllable final /k/ (>/j/) in Portuguese indicate that this is a patrimonial wo rd, in contrast with learned directo in both languages. The patrimonial form is well atte sted for both languages: thirteenth century for Castilian according to the Corpus del Espaol (Davies 2002-2004, s.v. derecho) and fourteenth century for Portuguese according to the Corpus do Portugus (Davies and Ferreira 2004-2006, s.v. direito). 271

PAGE 272

Figure 5-11. Outcom es of HV('HC) and L( 'HC) in declension 3 nouns relative to syllable count and nature of accented syllable (n=29). Trisyllables with penultimate accent on HV type syllable Preservation of heavy syllables is high in the trisyllabic nouns reviewed to this point, both in the stress bearing syllable and in some cases the initial syllable as well. Both of these positions can be said to be prominent within the word although the prominence of the accented syllable is clearly greater than any secondary accent that may be accorded the initial syllable. Preservation of segments as a way of realizing pr ominence can be tested in the next set of nouns that are in essence a mirror image of those just re viewed; they consist of an initial heavy syllable and a penultimate light syllable. This analysis presumes that the original input form of HC('HV)L had probably become HC('L.L) in Late Latin with the excepti on of a handful of learned words; the one example in the subset of second/fourth declension nouns is Cat. centaure, Cast./Port. centaure centaur. Nouns from the second/forth declension corr esponding to input HC('L.L) number 26. All but one are reduced to two syllabl es in Catalan (the above cited centaure). Conversely, all 26 nouns remain trisyllabic in Castilian and 26 are tr isyllabic in Portuguese with the exception of 100.0% 80.0% 60.0% 40.0% 20.0% 0.0% CAT 34.5% 0.0% 0.0% 62.1% 3.4% 0.0% CAS 100.0%0.0%0.0%0.0%0.0%0.0% POR 24.1%72.4% 0.0% 0.0% 3.4% 0.0% HC HV L HC HV L 3 SYLL/PENULTIMATE 2 SYLL/ULTIMATE 272

PAGE 273

irmo broth er (< germ nus, i, m.), which like Catalan germ elides final /n/ following apocope of the final vowel. The distri bution of outputs by syllable count and nature of the initial and accented syllable appears in Figure 5-12. Percen tage within each category is a measure of incidence relative to total of 26. 100.0% 80.0% 60.0% 40.0% 20.0% 0.0% HC HV L HC HC HV Figure 5-12. Outcom es of HC('HV) in decl ension 2/4 nouns: Syllable count and nature of accented syllable (n=26). The points labeled 1 and 3 in Figure 5-12 i ndicate mora preservation of the unaccented initial syllable. When HC and HV types are added together for both three-syllable and twosyllable occurrences of heavy initial syllables Portuguese, at 57.6%, places the lowest in comparison with Catalan at 69.2% and Castilian at 65.4%. The gradient ranking reflects the high degree of tolerance of stops in word internal coda position in Catalan versus elision in some cases in Castilian and in nearly all cases in Portuguese. Preservation of the coda in initial position may serve to give that syllable prominen ce, especially if it co rresponds to a morpheme as exemplified by con/com-, dis-, ex, all represented in this data set. However, both Catalan and Portuguese show vowel reduction in precisely these same cont exts. This suggests that a mechanism other than duration, such as pitc h contour (Hay and Di ehl 2007, 121), accords CAT 3.8% 0.0% 0.0% 0.0% 3.8% 0.0% 65.4% 0.0% 30.8% 69.2% 26.9% 0.0% CAS 65.4% 0.0% 34.6% 0.0% 3.8% 96.2% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% POR 42.3% 11.5% 42.3% 0.0% 3.8% 92.3% 3.8% 0.0% 0.0% 0.0% 3.8% 0.0% HV L HC HV L L S3 S2 S2 S1 3 SYLL/PENULTIMATE 2 SYLL/ULTIMATE 273

PAGE 274

position al prominence to these prefixes. In the conservative template, the final unaccented syllable is preserved, data point 2, and in over 90% of the cases it is a light syllable. The innovative template, most represen ted in Catalan, always has a h eavy final syllable as indicated by data points 4 and 5. The prosodic templates that result from original HC('HV) (mostly HC('L) in late Latin) and their rate of occurrence by language app ear in Table 5-19. Threesyllable patterns are listed first followed by two-syllable pa tterns. Shading indicates that the tonic syllable is heavy. Table 5-19. Prosodic outcomes of HC('HV) in declension 2/4 nouns Catalan Castilian Portuguese a. HC('HV.L) 1 3.8% 1 3.8% 1 3.8% b. HC('L.L) 0 0.0% 15 57.7% 11 42.3% c. HV('L.L) 0 0.0% 0 0.0% 2 7.7% d. L('L.L) 0 0.0% 10 38.5% 11 42.3% e. HC('HC) 11 42.3% 0 0.0% 0 0.0% f. HC('HV) 6 23.1% 0 0.0% 1 3.8% g. L('HC) 7 26.9% 0 0.0% 0 0.0% h. L('HV) 1 3.8% 0 0.0% 0 0.0% Note: Coincidence of heavy syllable and primary accent is indicated by shaded cells. Without apocope one would expect the results se en in the columns for Castilian and Portuguese, that is, an occasional bimoraic syllable resulting from preservati on of a falling diphthong, and for Portuguese, secondary diphthongs resulting from loss of intervocalic /l/ and /n/. The incidence of apocope in trisyllables with penultimate accent has created a new prosodic template in which the head syllable of the head foot is now at ri ght word edge. Apocope also ensures that this syllable is always heavy. Loss of coda consonant s other than /n/ in Catalan and Portuguese is infrequent and the occasional deletion of / / is not consistent. There is some evidence that loss of these codas is compensated for by differences in quality and duration of the tonic vowel. The relative frequency of the templates in Table 5-20 is displayed graphically in Figure 5-13. The 274

PAGE 275

preference f or the two-syllable template in Catalan is readily apparent as the results by language are dichotomous with Castilian and Portuguese remaining trisylla bic in over 95% of cases and Catalan becoming disyllabic in over 95% of cases. 100.0% 80.0% 60.0% 40.0% 20.0% 0.0% HC('HV.L) HC('L.L) L('HC) Figure 5-13. Relative frequency of 3-syllable a nd 2-syllable templates resulting from declension 2/4 input HC('HV). (n=26) Data from the third declension, as well as cas es of proparoxytonic accent, should also be examined before pursuing discussion of trochaic versus iambic interpretations of the two-syllable templates. There are 69 thir d declension nouns with original input form HC('HV). If development parallels that of s econd and fourth declension nouns it can be expected that 60% to 70% of output forms will retain the initial hea vy syllable and there will be widespread apocope and concomitant reduction to two syllables in Ca talan. However, because the vocalism of the third declension is not associat ed with any morphological function, for example gender, it can also be predicted that apocope will occur at a high er rate in both Castilian and Portuguese as long as language-specific coda c onditions are not violated. The contrasts in outcomes displayed in Figures 5-13 and 5-14 are notable. In this data set all three languages participate in apocope and th e most frequent outcome is the two-syllable Catalan 3.8% 0.0% 0.0% 0.0% 42.3% 23.1% 26.9% 3.8% Castilian 3.8% 57.7% 0.0% 38.5% 0.0% 0.0% 0.0% 0.0% Portuguese 3.8% 42.3% 7.7% 42.3% 0.0% 3.8% 0.0% 0.0% HV('L.L) L(' L.L) HC('HC) HC('HV) L('HV) 2 SYLL 3 SYLL 275

PAGE 276

tem plate with ultimate accent. The split of thr ee-syllable to two-syllable results is 13 (18.8%) to 56 (81.2%) for Catalan, 15 (21.7%) to 54 (78.3%) for Castilian, and 17 (24.6%) to 52 (75.4%) for Portuguese. Not only are the proportions quit e similar but the distribution of heavy/light syllables seen in Figure 5-14 is ne arly identical. The slightly highe r rate of heavy final syllables with coda consonant in the two-syll able template in Castilian is a result of retention of final /n/ (elided in both Catalan and Portuguese) and retenti on of palatalized output of /k/ in coda position as in perdiz compared with vocalization in Catalan perdiu partridge ( in de clension 3 nouns: Syllable count and nature of accented syllable (n=26). Outcomes of third declension HC('HV) also evidence a high rate of retention of the heavy initial syllable. In the case of the threesyllable templates the retention rate is 100% (13/13) for Catalan, 93.3% (14/15) for Castilian, and 13/17 ( 76.5%; 3 of the 13 are HV type syllables). Therefore, the predom inant three-syllable pattern is H( 'L.L). This corresponds with the results for declensions 2/4 seen in Table 5-19 Retention of coda consonants in the initial syllable of two-syllable outputs is identical across th e three languages at 6 2.3% when type HC CAT 18.8% 0.0% 0.0% 0.0% 0.0% 18.8% 60.9% 1.4% 18.8% 60.9% 20.3% 0.0% CAS 20.3% 0.0% 1.4% 0.0% 0.0% 21.7% 60.9% 1.4% 15.9% 78.3% 0.0% 0.0% POR 14.5% 4.3% 5.8% 0.0% 0.0% 24.6% 55.1% 7.2% 13.0% 59.4% 15.9% 0.0% HC HV L HC HV L S3 S2 S2 S1 3 SYLL/PENULTIMATE 2 SYLL/ULTIMATE 276

PAGE 277

and HV syllables are combined. Again, the templa tic results coincide with those for nouns from declension 2/4, that is, H('H) is the most common output. The last data sets to be examined in the tr isyllabic group with penultimate accent consist of those with light or HV type initial syllable and HV type penultimate syllable. It is expected that the input form for Ibero-Romance is L.L1.L w ith few exceptions. Again, it is predicted that there will be a high occurrence of apocope in Ca talan in both declensi on 2/4 and declension 3 nouns. Furthermore, if is expected that for d eclension 3 nouns all thre e languages will exhibit a high degree of elision of the final vowel as seen in the previous set. Looking at the three-syllable sequence as idealized CV.'CV.CV, loss of the fina l vowel would inevitably result in a disyllable with heavy final syllable, CV.'CV C or CV.'CVV. The CVV type sy llable could result from glide formation and creation of a falli ng diphthong as in semi-learned datiu dative in Catalan, or from deletion of /-n/ in both Catalan and Portuguese as in Catalan padr and Portuguese padro (

and L('HV) are displayed in Figur e 5-15. There are very few cases in which an original diphthong is retained and a ll are learned words. In addition to tesaure cited above for 277

PAGE 278

Catalan ( tesauro in Castilian and Portuguese), the one other w ord which keeps an original diphthong is Cat. eunuc and Cast., Por. eunuco eunuch. 0.0% 20.0% 40.0% 60.0% 80.0% 100.0% CAT 0.0%5.9%1.5%4.4%1.5%1.5%91.2%63.2%30.9%0.0% CAS 1.5%95.6%1.5%95.6%0.0%0.0%2.9%2.9%0.0%0.0% POR 1.5%88.2%1.5%88.2%0.0%0.0%10.3%0.0%10.3%0.0% HVL HVLHCHVLHC HVL S3 S2 S2 S1 3 SYLL/PENULTIMATE 2 SYLL/ULTIMATE Figure 5-15. Outcom es of HV('HV) and L( 'HV) in declension 2/4 nouns: Syllable count and nature of accented syllable (n=68). Also of interest in the two-syllable outcomes is the fact that Catalan has a high incidence of HV type syllables. Catalan should coincide with Portuguese in cases of /n/ in word final position arising from apocope but Catalan has a much higher percentage seen at data point 3 in Figure 515. This is due to glide formation in Catalan wh ich occurs before apocope. The thematic vowel of the second/fourth declension combines with the preceding tonic vowel to form a falling diphthong as in natiu (
PAGE 279

n and Por. o, outcom es of Lat. ne(m ), a suffix indicating having a possessing or characteric meaning (Miller 2006, 76) but extended semantically to an augmentative function. The comparable third declension data set, also pres umed to consist of all light syllables in input with the exception of a few learned words, is somewhat larger with 101 items. An additional three items are not included in the data set because they seem to be late, learned additions to the lexicon and exhibit variable placem ent of primary accent: Cat., Cast. frenes, Port. frenesi(m) frenzy, Cat., Cast lquen Por. liquem lichen, and Cat. ugur Cast. augur and Por. ugure augur. It is expected that apocope will ex tend to all three languages, although with higher representation in Catalan because it has fewer coda constraints. Figure 5-16 confirms this prediction and shows little variat ion among languages other than th e expected higher occurrence of the two-syllable, ultimate accent pattern in Catalan. Data points 1 and 2 in Figure 5-16 indicate the two most comm on patterns, trisyllabic L('L.L) and disyllabic L('HC). Figure 5-16. Outcom es of HV('HV) and L( 'HV) in declension 3 nouns: Syllable count and nature of accented syllable (n=101). The distributions seen at data points 3 and 4 reflect constraints on coda consonants in word final position. Although data point 2 indicates that the 2 two-syllable template has a higher degree of representation in Cata lan, Castilan and Catalan have very similar outcomes with regard -20.0% 0.0% 20.0% 40.0% 60.0% 80.0% 100.0% HV HV HC L L L HC HV HC HV CAT 0.0% 13.9% 0.0% 13.9% 1.0% 0.0% 85.1% 74.3% 11.9% 0.0% 0.0% 0.0% 0.0% 0.0% CAS 0.0% 23.8% 0.0% 23.8% 1.0% 0.0% 71.3% 72.3% 0.0% 0.0% 2.0% 1.0% 1.0% 2.0% POR 0.0% 26.7% 0.0% 26.7% 1.0% 9.9% 59.4% 60.4% 9.9% 0.0% 0.0% 0.0% 0.0% 3.0% L L L HC S3 S2 S2 S1 S1 S2 S1 3 SYLL/PENULTIMATE 2 SYLL/ULTIMATE 2 SYLL/PENULTIMATE 1 S YLL 279

PAGE 280

to the nature of the final sylla ble, HC, in com parison with Portuguese. Surprisingly, Catalan has the highest percentage of HV type final syllables. Two processes c ontribute to this: deletion of final /n/ and glide formation of vowels in hiat us (penultimate and ultimate syllables) which precedes apocope of the final vowel. While both of these processes also occur in Portuguese, glide formation is more frequent in Catalan. Figures 5-15 and 5-16 represent very different outcomes with regard to preferred templates by language. The thematic vowel of the second/ fourth declension is largely preserved in Castilian and Portuguese but not in Catalan. However, when there is no discernible thematic vowel, as is the case of the thir d declension, that vowel is also elided. The most common reflex of accusative singular in the third declension is e(m). In the vowel sonority hierarchy (see Table 5-1), [ ] is ranked among the least sonorous vowels. It must be assumed that vowel reduction is a precursor to vowel deletion. Trisyllables with Antepenultimate Accent Apocope in the case of trisyllables with pr oparoxytonic accent creates a new template in which the disyllabic innovative forms are either H.'H or L.'H. The data set that gives rise to the former, H.'H, is examined first. It consists of trisyllables in which the initial HC type syllable is expected to remain heavy because the syllable occupies a prominent position within the word and is also the locus of primary accent. The exp ected high representation of the H.'H template in Catalan is indicated by data point s 3 and 4 in Figure 5-17. In contrast, preservation of the threesyllable input is predicted to be the usual out come for Castilian and Portuguese (data points 1 and 2) despite the fact that the antepenultimat e accent is a marked prosodic template. Words with antepenultimate accent are primarily learned words, and many are modern additons to the scholarly lexicon. Therefore, syncope, anothe r mechanism for reducing syllable count, is less likely to occur. 280

PAGE 281

100.0% 80.0% 60.0% 40.0% 20.0% 0.0% HC HC L L HC Figure 5-17. Outcom es of HC.L.L declension 2/4 nouns: Syllable count and nature of accented syllable (n=68). The data represented in Figur e 5-17 displays an unexpected outcome indicated by data point 5, the disyllabic word with a light final syllable. Although this syllable type occurs at approximately the same rate in the three language s it has different origins. In Castilian and Portuguese it represents conserva tion of the characteristic vowel of the second/fourth declension brought about by glide formation of the post-tonic vowel in the penultimate syllable. The glide palatalizes the preceding consonant a nd is then absorbed as in Cast. brazo and Por. brao arm (
PAGE 282

participate in apocope with Catalan at the highest rate, followed by Castilian and then Portuguese. Portuguese also shows the lowest incidence of HC type final syllable in twosyllab le outcomes. Because of the small size of this data set (31 items) it is more useful to view the results with numeric rather th an percentage values. In Figu re 5-18, data point 1 indicates a high degree of retention of initial HC type syllable across languages. 3025 20 15 10 Figure 5-18. Outcom es of HC.L.L declension 3 nouns: Syllable count and nature of accented syllable (n=32). The penultimate syllable remains light in 100% of cases (data point 2) Data points 3 and 4 indicate the relative frequency of the 'HC.HC template as outcome in the expected order Catalan, Castilian, Portuguese although at data point 4 Castilian has a slightly higher number of HC syllables than Catalan. This is due to th e retention of word final vowels (see data point 5) that follow consonant sequences resulting from syncope in cases such as angle (
PAGE 283

the HV type syllable in Ibero-Ro m ance, especially in the patrimonial lexicon. The data set corresponding to second/fourth declension nouns is numerous and consists of 169 items. An additional 7 items are not included because of vari able treatment of locus of accent. Most are scientific terms which did not enter Ibero-Roman ce until the eighteenth or nineteenth century. The one word which can be documented from an earli er period because of its use as a liturgical term is sp r tus, s m. spirit. In both Castilian and Portuguese antepenultimate accent is preserved ( espritu ). However, in Catalan the accent is on the final syllable. According to the DCVB (s.v. esperit) this is due to the influence of the dative case, sp r tu which would have had antepenultimate accent on the r syllable. Displacement of the accent is not a result of an attempt to align word primary accent with a heavy sy llable. In fact, data point 3 indicates that L.HC is the most common pattern for disy llabic outcomes in Catalan followed by L.L at data point 2. Figure 5-19. Prosodic templates fr om second and fourth declension HV.L.L and L.L.L (n=169). The high representation of L.L as outcom e in Catalan (data point 2) merits an explanation. Apocope is blocked when th ere is a violation of NOCODACAT, CAS, POR or SONSEQ. Beginning with an idealized representation of a three-sy llable word with penultimate accent, CV.CV.CV, a head foot consisting of two light syllables, ( L.L), could result if both ons et and nucleus of the final -20.0% 0.0% 20.0% 40.0% 60.0% 80.0% 100.0% CAT 0.0% 4.7% 1.2% 27.8% 1.2% 2.4% 2.4% 40.2% 0.0% 17.2% 3.0% CAS 3.0% 50.3% 0.6% 37.9% 2.4% 5.3% 0.0% 0.0% 0.0% 0.6% 0.0% POR 3.0% 51.5% 0.6% 30.2% 8.9% 4.7% 0.0% 0.0% 0.6% 0.0% 0.6% 'HV.L.L 'L.L.L. 'L.L.HC 'L.L 'HV.L. 'HC.L 'HV.HC 'L.HC'L.HV 'HC 'HV 3 SYL/ANTEPENULTIMATE L 2 SYLL/PENULTIMATE 1 S YLL 283

PAGE 284

syllable are lost. If the final syllable is onsetless, that is, CV.CV.V, and t he final syllable is elided, the result is a final, unstressed vowel, alt hough not the vowel of the or iginal final syllable. Alternatively, if there is coalescence of the nuclei in the penultimate and ultimate syllables through glide formation, a rising diphthong or falling diphthong will result depending on the nature of the vowels in hiatus. These divergent paths result in the same prosodic template but differentiate Catalan from Castilian/Port uguese. Catalan is not constrained by MAXMORPH and, therefore, elides the final unaccented vowel. Castilian and Portuguese, on the other hand, create rising diphthongs but retain the characteristic vowel of the second/fourth declension. An example in point is l b um, ii, n. lip which gives rise to Catalan llavi, Castilian/Portuguese labio All are disyllables with paroxytonic accen t but represent different paths to the same prosodic result. Word internal syncope accounts for the unexpect ed initial heavy syll able in the outputs indicated by data point 3 in Figure 5-19 and for Catalan data points 5 and 6. The relative chronology of syncope is difficult to establish and there is much ev idence that syncope existed in spoken registers of the language long before the Appendix Probi (3rd c. C.E.) in addition to its use as a poetic device (Coleman 1999, 38-40). Near ly all cases of syncope in this data set correspond to words which could be considered to be patrimonial and belonging to the popular register as in the example s l dus, i m. gold coin which entered Ibero-Romance as soldu and then gave rise to Catalan sou (pronunciation [s w] in DCVB, s.v. sou), Castilian sueldo and Portuguese soldo The last data set in this group, third declension nouns with antepenultimate accent, should demonstrate a high degree of like outcomes acro ss languages. Apocope is restrained only by 284

PAGE 285

NOCODACAT, CAS, POR or SONSEQ (in the case of forms with sync ope); therefore, outputs should show little distinction by language as confirmed by the results displayed in Figure 5-20. 35 30 25 20 15 10 5 0 'L.L.HC 'L.L 'L .HV 'HC Figure 5-20. Prosodic templates from third declension HV.L.L and L.L.L (n=36). The m inimal differences seen at data point s 1 and 5 in Figure 5-20 are complementary. Catalan has a higher incidence of apocope (see data point 5) because there are fewer coda restrictions. Castilian places restrictions on non-coronal obstrue nts and nasals while Portuguese admits only /s/ and liquids (see data point 1). Consequently, a gradient result is expected for retention of final vowel and elision of final vowel with Catalan showing the highest rate of apocope followed by Castilian and then Portuguese. This is confirmed by the numeric array at data point 1 and its corollaries at data points 2 and 5. The samples represented by data points 3, 4, and 6 are too small to analyze. They do s how a slightly higher incidence of apocope for Catalan and Castilian than for Portuguese, but only Portuguese has heavy final syllables with nasal vowels (data point 6). The anticipated resu lts, confirmed in Figure 5-20, concern the extent to which an unaccented final vowel can be delete d in all three languages when that vowel ranks low on the vowel sonority scale and has no apparent morphological function. CAT 7 2 7 2 4 13 0 1 CAS 10 2 7 2 5 10 0 0 POR 12 2 5 1 4 9 2 1 'L.L.L. 'HV.L. 'L .HC HC.L 3 SYLL/ANTEPENULT 2 SYLL/PENULTIMATE 1 SYLL 285

PAGE 286

Four-syllable Nouns Nouns of four syllables that undergo apocope present a new prosodic pattern. It has been previously suggested that word in itial position carries som e degree of prominence. In the case of nouns with penultimate accent, the prosodic structur e that results from apocope could then be represented as 2.1.3, where 3 is th e highest degree of prominence, the syllable with primary accent at word level. As expected, nouns of the second and fourth declension show widespread apocope only in Catalan but thir d declension nouns in all three languages show apocope albeit in varying degrees dependent on langua ge specific coda constraints. Four-syllable Nouns with Penultimate Accent The first data set contains tetrasyllables with penultimate accent from the second declension (There are no fourth declension nouns in this subset) that have HC type initial and penultimate syllables. Since both of the hea vy syllables represent locations of prominence within the word it is expected that the HC syllables will remain bimoraic. There are only 24 items corresponding to the parameters, HC.X. HC.X for second declension nouns. Figure 5-21 shows the prosodic outputs for this data set. Again, there is the expected split between loss of final vowel in Catalan and retention in Castil ian and Portuguese. The most common templates for four-syllable and three-syllable outcomes ar e marked as data points 1 and 2. These highly represented templates, H.L.' H.L and H.L.' H, show retention of the mo raic count of the initial syllable as well as the accented sy llable at the rate of approximately 67%. However, if all instances of heavy initial syllable and heavy accent ed syllable are aggregated the rates are much higher: Catalan, 100%; Castilian, 83.3%, Portug uese, 75.0%. The prevalence of apocope in Catalan obscures to what degree the coda consonant of the original accented penult has been preserved; therefore, the 100% retention of hea vy syllables in initial a nd accented syllables in Catalan may be an artifact. 286

PAGE 287

24 20 16 12 8 4 0 H.H.'H.X H.L.'H.X L.L.'H.X H.H.'L.X H.L.'L.X H.H.'H H.L.'H 3 4 CAT Figure 5-21. Prosodic outcom es of second declension tetrasyllables with HC type initial and penultimate syllables (n=24). The corresponding data set for thir d declension nouns is also re duced in size, consisting of 19 examplars in which both initial syllable a nd accented syllable are heavy. The two most frequent patterns, H.L. H.L and H.L.' H, are located at data points 1 and 3 in Figure 5-22. They coincide with the dominant templates in Figure 5-21. Because of the small size of this data set, distributions viewed as percentage of whole may be exaggerated. Figure 5-22. Prosodic outcomes of third declen sion tetrasyllables with penultimate accent and HC type initial and penultimate syllables (n=19). 00 01 16 17 2 CAS 11 201 16 1 POR 12 301 20 15 105 0 H.H.'H.X CAT 0 00 00 3 12 21 1 CAS 4 10 2 12 00 00 0 POR 4 84 12 00 00 0 H.L.'H.X L.L.'H.'H.X .X L.H H.'H.L H.H.'H H.L.'H L.H.'H L.L.'H H.'H ' 287

PAGE 288

Proceeding with this cav eat, when data are aggregated for each language the percentage of outcomes with heavy initial syllable as well as a heavy tonic syllable are 84.2% Catalan and Castilian, 73.7% Portuguese. Faithfulness in prom inent positions is high for all three languages. The trisyllabic template that results in Cas tilian and Portuguese (data point 2) and the one disyllabic example in Catalan (data point 4) have the same origin. They are the result of glide formation occurring between the initial syllable and the pre-tonic syllable in the present participles serv ns, -entis serving and amb ns, -entis surrounding which become nouns: Cat. servent (also sirvent ), Cast. sirviente, Por. servente servant; Cat. ambient Cast. and Por. ambiente ambience. In the case of the Ca talan and Portuguese outputs from serv ente it must be assumed that the high vocoid at first produces metaphony seen in the raised vowel of the Catalan alternate form sirvent The original vocalism is later restored in both languages through processes of analogy and regularization of the lexicon. The next data sets, consisting of words in which the accented penultimate syllable is HC type but the initial syllable is HV or L, assume an input template of the form L.X.'H.X. The pretonic syllable may be heavy or li ght but original HV type syllable s in initial position are probably light in Proto Ibero-Romance. Beginning with the data set of 56 items for second declension nouns, it is expected that the penul timate heavy syllable will retain its bimoraic nature and that an initial trochee can be built at left word edge with a preferred template of 'L.L. The expected outputs are the most frequent, indicated at data point 1 for Castilian and Portuguese and data point 6 for Catalan in Figure 5-23. Faithfulness to the moraicity of the accent bearing syllable is high for both the four-syllable outcomes and the th ree-syllable outcomes although in the case of the three-syllable outcomes it is likely that cons traints related to sonorit y sequencing or complex codas have in some cases superseded MAXIO so that the consonantal segments that appear in 288

PAGE 289

coda position m ay not be identical in either na ture or number to the input. The most highly represented output patterns, data points 1 and 6, demonstrate faithfulness to the moraic count of the accented syllable. However, data point 2 indicates that the coda consonant is lost in approximately one out of six cases in both Castilian (17.8%) and Po rtuguese (16.1%). 60 50 40 30 20 10 Figure 5-23. Prosodic outcom es of second/four th declension tetrasyllables of the type HV.X.'HC.X and L.X. 'HC.X with penultimate accent (n=56). While prominent syllables tend to observe MAXIO to a greater degree than non-prominent syllables, the sequence 'L.L is a preferable in stantiation of a trochaic foot. The template represented at data point 2 in Figure 5-23 consists of two ideal trochees, ( L.L) ('L.L). For Catalan, all tonic syllables are H, although in 15 of 56 cases (26.8 %) the tonic syllable is not rightmost. A nucleus is retained in the final syllable when the consona nt sequence that would result from apocope violates SONSEQ. Data points 3, 4, and 5 show reduction to three syllables in various cases in Castilian and Portuguese. The catalyst in ha lf the cases is syncope, and in the remaining cases, glide formation. Both processes result in loss of the pre-tonic nucleus. The pattern of syncope (data 0 CAT 13 2 0 0 0 0 37 2 2 CAS 37 3 10 2 4 0 0 0 0 POR 37 3 9 2 4 1 0 0 0 L.L.'H.X L.H.'H.X L.L.'L.X H.'L.X L.'H.X L.'L.X L.L.'HH.'HL.'H ' 289

PAGE 290

point 3) affects sem i-learned rath er than patrimonial words. Th is is seen by the absence of expected phonological processes such as deletion of syllable-final /p/ (cf. septem seven > Cast. siete, Por. sete ) and failure to palatalize /-kt-/ (cf. dictum having been said > Cast. dicho ). The two-syllable outputs in Catalan, H. H, are represented at data poi nt 7 in Figure 5-23. Catalan Castilian Portuguese c p tellum, i, n. head (dim.) cabdill caudillo caudilho m l dictum, i, n. curse maldit maldito maldito Data points 4 and 5 represent the products of glide formation th at result from the constraint against vowels in hiatus seen in Castilian and Portuguese. Thus, d lect cus, a, um dialectical maintains hiatus only in Catalan and becomes trisy llabic in the other two la nguages. Fortition of the onset of the initial syllable in hy cinthus/-os, i m. hyacinth eliminates the V.V sequence in all three languages producing Cat. jacint Por., Cast. jacinto Coalescence of the like vowels in c p rtum having been covered; cover (subst.) also results in reduction of the syllable count across languages: Cat. cobert Cast. cubierto Por. coberto These anomalous changes are indicated by data points 4 and 8 in Figure 5-23. Apocope in Castilian and Portuguese is unreprese nted in this data se t due to a constraint against complex codas: *COMP (Syllables must not have complex codas). Elimination of a final vowel after a heavy syllable with a coda consonant results in a complex coda: (C)VC.CV (C)VCC. Catalan is able to reduce the complex coda by ranking *COMP and MAX(PLACE) above MAX/IO. MAX(PLACE) (Every input place feature has an output correspondent) allows deletion of both the final vowel and the coronal obstruent be cause place of articulation is already marked by the nasal /n/. The corresponding data set for th ird declension nouns is very small and consists of only 13 words that come from input types HV.X.'HC.X and L.X.'HC.X. This small data set shows 290

PAGE 291

unif ormity in the nature of the first two syllables ; all but one are L.L. Nine of the words are nouns derived from the present participle so that the last two syllables ar e also uniform in input, that is, ente Because the tonic syllable has a nasal c oda in a large number of cases the results for Castilian and Portuguese will appear to be quite different unless HC and HV type syllables are aggregated. The aggregated view (HC and HV type syllables combined) is presented in Figure 5-24. Predictably, Castilian and Portugue se are aligned on one side (data point 1) and Catalan (data point 2) on the ot her. There are only two cases, uniforme and the scientific term epidermis, in which the final vowel is re tained in Catalan due to the SONSEQ constraint. Figure 5-24. Prosodic outcom es of third declen sion tetrasyllables of the type HV.X.'HC.L and L.X.'HC.L (n=13) with penultimate accent. The constraint on complex codas, *COMPLEX, bars apocope in Castilian and Portuguese. In a system of syllabic troc hees, the most common outputs (data points 1 and 2) have licit trochees in word initial positi on. The outputs with apocope, seen at the right in Figure 5-24, require further explanation in terms of parsing. Assuming that the metrical foot is the syllabic trochee and that it must be constr ucted at right word edge, it is necessary to accept the rightmost syllable as a degenerate foot. Th is is in line with a weak prohib ition on degenerate feet (Hayes 1995, 87) which allows a suboptimal foot in a strong position, in this case the right word edge. -1 1 3 5 7 9 11 13 L.H.'H.L L.L.'H.L L.'H.L CAT 02 0 1 10 CAS 19300 POR 19 3 00 L.H.'H L.L.'H 291

PAGE 292

For Catalan it can b e argued that the primary acc ent is aligned at the right word edge unless SONSEQ is violated or the final vowel is /a/. The constraint against deleti on of the low vowel /a/ could take the form MAX-IO-V[lo], as proposed by Hartkemeyer (2000, 79). Four-syllable nouns with HC type initial syllable and HV type penultimate syllable When the penultimate syllable is bimoraic based on vowel length, there is no coda consonant which could potentia lly create a violation of SONSEQ in the case of elision of a final vowel. It is expected, therefore, that the ra te of apocope in Catalan will be very high. Prominence of the initial syllable favors retentio n of the coda consonant. Moraic loss in initial syllables is due primarily to re duction of geminate obstruents. Figure 5-25 shows a light initial syllable in 7 of 31 cases in Catalan (22.6%) and 9 of 31 cases in both Castilian and Portuguese (29.0%). The difference among languages is due to the fact that Catalan re tains geminate /-l.l-/ which is simplified in Castilian and Portuguese. The few cases in which the final vowel is retained in Catalan represent the influence of SONSEQ. Because the penultimate syllable is light, consonant sequences arise at the ri ght word edge when the onset of the final syllable is complex. However, what SONSEQ prescribes for an onset is precisely what will be rejected in a coda, a sequence of less sonorous C1 followed by more s onorous C2 as in the case of Cat. canelobre (
PAGE 293

an uneven trochee due to its de creased duration and in tens ity in contrast with the preceding accented vowel (Crosswhite 2004, 208). 0 4 8 12 16 20 24 28 32 CAT 0310007142301 CAS 5165311000000 POR 4177101000010 H.H.'L.LH.L.'L.LL.L.'L.LL.H.'L.LH.'L.LL.'L.LH.H.'H H.L.'HL.H.'HL.L.'HH.'HL.'H ' Figure 5-25. Prosodic outcomes of second/four th declension tetrasyllables of the type HC.X.'HV.L with penul timate accent (n=31). Increased participation in apocope is notic eable for Castilian and Portuguese in third declension nouns (Figure 5-26). In the preceding data set of second/fourth declension nouns the final, unaccented vowel is retained by MAXMORPH which does not apply to th is declension class. The most prevalent pattern cross-linguistically is H.L.'H (data point 2 in Figure 5-26). 0 10 20 30 40 50 60 70 80 90 CAT 11066121042 CAS 11023809297 POR 114352027296 H.L.'H.LH.L.'L.XL.L.'L.XH.H.'HH.L.'HL.H.'HL.L.'HH.'HL.'H Figure 5-26. Prosodic outcom es of third declension tetrasyllables of the type HC.X.'HV.L with penultimate accent (n=87). 293

PAGE 294

There are a few cases of syncope in p a trimonial words that can be parsed ( H.L)('H). In this configuration the vulnerable syllable is the weak syllable of the foot constructed at left word edge. Syncope is relatively infrequent when th e first syllable is heavy because it is blocked by SONSEQ and SYLCON (Syllable Contact Law). Neverthele ss, the following patrimonial words show clear evidence of vowel loss in the pret onic syllable even though the resulting consonant sequences violate SONSEQ or SYLCON, or both. Repair strategies us ually involve elision of one of the consonants to produce an ac ceptable sequence. The case of arcione(m) differs in that the high vowel becomes a glide and is then absorbed by the preceding palatalized onset consonant. Portuguese conserves the pretonic vowel in the case of peitoral. Catalan Castilian Portuguese hosp t lis, is, m. guest hostal hostal hostal *arc o, nis, m.gullet (saddle) ar arzn aro acceptor ris, m. hawk astor azor aor pectr le, is, n. cinch (saddle) pitral petral peitoral The data in this third declension subset is somewhat skewed by the presence of a highly productive suffix, i ne(m) which always has the same outcome, Catalan i (disyllabic), Castilian in (monosyllabic), and Portuguese o (monosyllabic). This accounts for the discrepancies among languages seen at data poi nts 2 and 3 in Figure 5-26. Data point 1 illustrates the more restrictive c oda conditions of Portuguese that result in retention of the final vowel in the case of the productive suffix t te(m) The 20 items that undergo apocope in Portuguese, indicated at data point 2 in Figure 5-26, have /l/ and / / as coda consonants and one case of /s/. It should be noted that in Catalan the only two nouns that retain the final vowel are two learned words that probably entered th e language well beyond th e formative period, emfiteusi (emphyteusis, a legal term) and timpanitis (tympanites, a medical term). In other words, apocope can be said to occur in Catalan at the rate of 100% in the case of words that do 294

PAGE 295

not belong to a learned register when the penultim a te syllable is light and the final syllable has a simple onset. Four-syllable nouns with HV or L type initia l syllable and HV type penultimate syllable The input set described here is of the type HV/L.X.HV1.L. The second syllable is nearly always light in the subset of second/fourth declension nouns. There ar e only four nouns, all learned words, in which the second syllable is of the HC type; they are not included in the analysis here. The data set under consideration is considered to have as input L.L.HV1.L. With few exceptions all four syllables are light, although there are isol ated cases of conservation of original diphthong /aw/. There are also a few case s of syncope that result in an initial heavy syllable. If the input form is parsed as ( L.L1)('L.L2) there are two weak syllables, L1 and L2. It can be assumed that L2 is designated for deletion based on the preference in Catalan for right alignment of both head foot and head syllable. The resulting foot structure for Catalan is ( L.L1)('H). There are, nonetheless, three cases in Catala n in which the output reflects both syncope and apocope (data point 6 in Figure 5-27). These th ree exceptions all have desirable codas that are more sonorous that the following onsets. Sync ope does not have negati ve consequences in prosodic terms for Castilian and Portuguese (dat a points 2 and 3) as the parsing for these languages is ( H)('L.L), with the exception of Castilian cuidado where the first syllable is generally construed as [kwi]. Catalan Castilian Portuguese s l d tus, a, um paid (with solidi ) soldat soldado soldado c m t tus s, m. company comtat condado condado c g t tus s, m. act of thinking cuidat cuidado coitado 295

PAGE 296

W ith L.L.'L.L (exceptional cases of cons ervation of an original diphthong are very infrequent) as input, the presumed faithful outpu t will consist of four light syllables with penultimate accent which can be parsed as ( L.L)('L.L). This is the outcome of choice (data point 1) for both Castilian (81.4 %) and Portuguese (67.4%). Only one word in Catalan retains the post-tonic vowel, albeit reduced to [ ] and represented orthographically by e, simulacre (
PAGE 297

The nine cas es of ultimate accent in Portuguese (data point 5) coincide with the most productive pattern in Catalan. This ou tput, a mirror image of Castilian ( H)('L.L), that is ( L.L) ('H) at data points 4 and 5, reflects the common word final sequence eu(m). The vowels in hiatus become tautosyllabic producing a falling diphthong [ew]. The second most common output for Catalan (data point 4) is the more predictable rout e to the same prosodic result, ( L.L) ('H). In these cases the heavy final syllable has a coda consonant that results from apocope. The one Castilian word th at has this prosodic outcome is capitn captain (
PAGE 298

of dege mination of obstruents. The one instance of emergence of a heavy initial syllable (see data point 2) is Portuguese vontade (
PAGE 299

Figure 5-29. Prosodic outcom es of third declen sion tetrasyllables of the type HV/L.HV/L.'HV.L with penultimate accent (n=94). There is limited evidence of pre-tonic syncope seen in the emergence of heavy initial syllables in all three languages (F igure 5-29, data points 2, 7). Thes e words contrast with most in this subgroup in terms of lexical register. Because vowel deletion is uniform across languages the question of relative chronology emerges. It is possible that these words already exhibited a variant with syncope in the popul ar stratum of Latin that form ed the basis of proto IberoRomance. The words listed below are clearly popular in nature, since most designate common objects or concepts. Loss of the vowel in the pre-tonic syllable creates a coda consonant which sometimes violates SONSEQ. Although the bilabial obstrue nt is retained in Catalan cabdal glide formation is seen in Castilian a nd Portuguese outputs for this word and in all three languages in the case of c v tate(m ). Glide formation is also seen in the outputs of d g t le(m) where the yod that results from vowel loss im pacts vowel quality in Catalan didal Forms that do not undergo syncope are indicated in square brackets. Catalan Castilian Portuguese c p t lis e, adj. wealth cabdal caudal caudal 100 80 60 40 20 0 L.L.' L.L d g t lis e, adj. thim ble didal dedal dedal CAT 2 1 0 0 57 19 1 7 4 3 CAS 2 1 0 0 53 0 0 35 0 3 POR 25 1 4 2 30 2 2 9 19 0 L.L.'HV.L H.'L.L L.'L.L L.L.' HC L.L.'HV HV.L.'HV L.'HC L.'HV H.'H ' 299

PAGE 300

bon t s, tis, f. goodness bondat bondad bondade c v t s tis, f. city ciutat ciudad cidade m l t s, tis, f. evil maldat maldad maldade ver t s, tis, f. truth [veritat] verdad verdade Other than the collapse of the suffix ne(m) into a monosyllable in Castilian and Portuguese, reflected at data point s 5 and 6, the outputs of L.L.'L.L (with occasional instance of HV type syllables from the diphthong /aw/) ar e overall quite uniform across languages. Portuguese is by far more faithful to the original structure in terms of syllable type, syllable count, and parsing of the word as two disyllabic trochees with 25 out of 94 exemplars (26.6%). Nevertheless, Portuguese is also represented at data point 3, although to a lesser degree than Catalan and Castilian. By combining the outputs re presented at data points 3 and 4 it can be seen that the pattern ( L.L) ('H) is far more common in Catalan th an the other two languages. Table 520 summarizes the most common outputs by language The high degree of faithfulness recorded for Portuguese ensues from the prohibition on ap ocope which would violate language-specific coda restrictions. It also maintain s a preferred rhythm ic pattern of ( L.L) ('L.L). Table 5-20. Distribution of prosodic templates in output of third declension tetrasyllables with penultimate accent. Most faithful Most innovative no change syncope/syneresis apocope apocope & syncope or glide formation ' Catalan 3.2% 0.0% 81.9% 14.9% Castilian 3.2% 0.0% 56.4% 40.4% Portuguese 27.7% 6.4% 36.2% 29.8% Both Castilian and Portuguese have high re presentation in the mo st innovative column. The prosodic template for this class is problematic in that, if feet are tr ochaic and right-aligned, foot formation at the right edge of the word produces a monosyllabic trochee preceded by an unparsable light syllable. However, an examin ation of hypocoristic formation in Catalan shows 300

PAGE 301

that the proposed parsing, L(' H), is correct. Catalan has many monosylla bic nouns in its lexicon, nearly all of which are bimoraic. The heavy final syllable, parsed here as a monosyllabic trochee, could form a licit word leaving the first syllable stranded and subject to deletion. In the few proper names that match the L('H) template this is precisely the outcome: Josep > Pep, Nadal > Dal, Miquel > Quel, Remei > Mei (Cabr 1994, 4-5). Even when the first syllable is heavy as in Jaumet it is subject to deletion: Jaumet > Met Four-syllable Nouns with Antepenultimate Accent Representation of four-syllable nouns is uneve nly distributed between declensions 2/4 and 3. The data set for declensions 2/4 contains 144 nouns, only a small portion of which retain the four syllables of the input. Apocope accounts prin cipally for the loss of syllables in Catalan so that it can be predicted that a major output patte rn will now have penultimate stress due to the loss of the final unaccented syllable. This is th e case for 80% of the output in Catalan and major patterns are located at data point s 2 and 3. Retention of the four syllables of the input is high although not the majority pattern in Castilian (35.9%) and Portuguese (40.7%). Loss of syllables for these languages is usually the result of glide formation rather than syncope. The few patrimonial words in this subset often undergo both syncope and apocope which accounts for the unexpected reduction to one syllable of two words in Catalan: poll louse (< p dc lus, i, m.) and quall blood clot (c glum i, n.). The usual output results in two syllables indicated by the 15 exempl ars at data point 4 of which genoll knee (
PAGE 302

-30 0 30 60 90 120 150 Figure 5-30. Prosodic outcom es of declension 2/4 tetrasyllables with antepenultimate accent. (n=145) The outputs of Castilian and Portuguese show reduction to three syllables in high proportion (data point 2). For both faithful and i nnovative outputs the prefe rred pattern consists of all light syllables. Given the nature of the an tepenultimate accent in Latin it is expected that both post-tonic syllables are light. Even in cases of reduction to three syllables, mostly as a result of glide formation, the output still consists of all light syllables. A few cases of accent shift are evident in all three languages. This occurs primar ily in learned/semi-learned words, particularly borrowings from Greek. A case in point is chrysanth mum, i n. chrysanthemum which has shifted the accent one syllable to th e right (accented syllabl e underlined) in both Catalan cristantem and Castilian crisante mo while Portuguese crisn temo retains the antepenultimate accent. The data set for third declension nouns is sm all (42 items) and shows great regularity in outcomes. The register of the words accounts fo r the high degree of faith fulness to the foursyllable input: 45.2% in Catalan, 54.8% in Ca stilian and Portuguese. The two most common tetrasyllabic patterns are H('L.L.L), as in imm b lis (inm-), e adj. real estate, and the more CAT 0 0 4 0 0 103 614111 002 3 1 15 2 CAS 0 3 49 1 1 033 70 18050 0 0 00 POR 2 2 55 0 2 046 64 02330 1 0 00 L'HLX H'L.L.L L'LLL LH'LL LL'LL 'LHH 'L.L.L L'HL L'LLL'LH H'LL 'HL 'LL LH'HLL'H H'HL'H'H ' ' 302

PAGE 303

frequent L(' H.L.L) (data points 1 and 2), as in interpr s, tis m.&f. negotiator.. In Figure 5-31 the final syllable is designated as X because there is occasional retention of a final consonant due to the learned nature of the words. Typically that final consonant is /n/, /m/, or /s/. Arguments can be made that these particular co nsonants are, in fact, we ightless. It is not necessary to view the final syllable as extrametri cal in order to avoid c onstruing this syllable as heavy. Historical patterns of syncope, coupled w ith synchronic models fo r truncation processes, and the metrical conventions of poetry work togeth er to make this a plausible explanation. This is discussed in detail in the concluding section. 42 36 30 2418 12 6 0 -6 Figure 5-31. Prosodic outcomes of declension 3 tetrasyllables with antepenultimate accent. (n=42) Evidence of apocope is represented by the out comes at data points 3 and 6 in Figure 5-31 whereas the effect of syncope is seen at data point 5. The ma jor point of divergence in outputs occurs at data point 4. This corresponds to a small set of deadjectival nouns ending in bile such as imm b le(m) cited above. In both Castilian and Ca talan the penultimate vowel is elided, CAT 50 1 13 1 0 531 2 20 51 3 CAS 61 2 14 3 0 330 1 11 50 2 POR 91 1 12 2 1 210 1 06 40 2 H'LLX H'HLL L'HLL L'LLX 'LLX LL'LL H'LH H'LL H'HH L'HH L'LH L'LH L'LL LL'H L'H ' 303

PAGE 304

resulting in ble (Cat. immoble, Cast. inmueble ) b ut in Portuguese that vowel is retained and the final vowel is suppressed producing -vel ( imvel ). Five-syllable Nouns Polysyllabic nouns of more than three syllab les all display some degree of morphological complexity. They contain morphemes that are both productive and non-productive. For example, the suffix mentu(m) means, instrument, result (M iller 2006, 78) is used to create deverbal nouns as in Cat. sofriment, Cast. sufrimiento Por. sofrimento suffering. However, the outputs of temp r mentum i n. moderation (Cat. temperament Cast./Por. temperamento ) have a less transparent relationship to the relate d verbs that evince a more popular stratum of the lexicon through syncope of th e pre-tonic vowel: Cat. temprar, Cast. templar Old Port. temblar to tune a musical instrument. In the cas e of second declension nouns with antepenultimate accent the suffix r u(m) is highly represented. The thre e-syllable suffix is preceded by a disyllabic root or stem. As with the suffix common in third declension nouns, ione(m) the outputs of r u(m) are regular and language-specific. Th e usual reflexes in all three languages show effects of metathesis or metaphony but onl y Portuguese retains the glide. The Catalan output undergoes apocope while both Castilian and Portuguese retain the thematic vowel with resulting suffixes er, -ero, -eiro. Distribution of the set of fi ve-syllable nouns is very unequal with regard to declension cl ass and accentual patterns. Five-syllable Nouns with Penultimate Accent Nouns with penultimate accent are underrepresented by second/fourth declension nouns and overrepresented by third declension nouns larg ely due to the productivity of the suffix ione(m) The set of second/fourth de clension nouns consists of 25 items. There are two suffixes that predominate in this set. The first is mentu(m) discussed above, and the second is vu(m) an adjectival suffix present in some deadjectival nouns. Since the words in this data set are 304

PAGE 305

learn ed or semi-learned in nature, it is expected that the outputs will be faithful to the input. For Catalan, the projected prosodic output ( X.X.X)('H) groups the pretonic syllables in a single foot (the initial dactyl effect ), while the head foot, a monosyllabic trochee, is constructed at the right word edge. The head foot for Castilian and Portuguese has the form ('L.L). This template is problematic if the syllables that follow the head of the presumed dactylic foot are heavy. There are five such cases: Catalan jurisconsult jurist, anabaptisme anabaptism, superlatiu superlative, triumvirat triumvirate, episcopat episcopate and Castilian/Portuguese jurisconsulto, anabaptismo, superl ativo, triumvirato, episcopado Given the universality of word initial position as a position of prominence it seems counterintuitive to suggest that the first syllable shou ld be left unparsed or should be treated as the weak syllable in an iambic foot, thus violating what has been he retofore viewed as an inviolable constraint, RHTYPE = T (Feet are trochaic). However, that an in itial light syllable in a polysyllabic word is vulnerable is nevertheless attested by comparing episcopat (Cast./Port. episcopado) with the outcomes of piscpus, i, m.: Cat. bisbe Port. bispo, Cast. obispo. In two cases apheresis occurs, and the vocalism of Castilia n suggests that the initial sylla ble may also have been lost historically22 and then later replaced by a vowel that does not correspond to the etymon. There are few cases of apheresis that occur diachroni cally across all three la nguages. Another often cited case, also a Greek borrowi ng, is the first declension noun pth ca, ae f. > Cat. botiga Cast./Port. bodega. The proposed initial dactyl for the five words given above has three di fferent realizations in terms of heavy/light syllables, LHH, LLH, LH L, with the latter being the most frequent. Because placement of primary accent is no long er dependent on the moraicity of candidate 22 CORDE ( Corpus diacrnico del espaol) has 581 examples of bispo prior to 1500. 305

PAGE 306

syllables there is no reason to s uppose that the secondary accent would be attracted to a heavy syllable in a nonprom inent position and away from the initial syllable. Although this study does not consider words of greater than fi ve syllables, the invocation of a *LAPSE constraint (Every weak beat must be adjacent to a strong beat or the word edge) is obviated here because the weak syllables designated as L above all satisfy the requ irement of adjacency. It should be noted that *LAPSE is not fully supported by phonetic studies that demonstrate that sequences of unstressed syllables in speech are not, in fact, so narrowly circumscribed (Marotta 2003, 334). Furthermore, the prefix seen in superlatiu/superlativo in contemporary, casual speech has become a free morpheme that act s as intensifier (For all three languages examples of sper interessant(e)/ interesante can be located on web sites from Sp ain and Portugual). As a stand alone morpheme sper always has primary acce nt on the initial syllable although the ultimate syllable is heavy. Therefore, despite the unsuitab ility of LHH, LLH, LHL as dactyls or trochees followed by an unparsed foot, it is justifiable to in terpret the word in itial sequence as ( X.X.X) or ( X.X)X. Outputs of second and fourth declension pent asyllables with penultimate accent are expected to maintain two positions of prominence: the initial syllable and the accented syllable. In the input these syllables are separate d by two intervening, unaccented syllables: X.X.X. X.X. The two unaccented syllables seem to invite syncope yet in this small data set (23 items) there are no cases of loss of pretonic vowels. Adm ittedly, the words in this subgroup are lexically marked and belong to the register of academia, law, and medicine. O uputs with heavy initial syllable, with HLL being the optimal dactyl, are indicated at data points 1, 2, and 5 in Figure 532. Forms with all light syllables preceding the t onic syllable (In patrimonial words these would be candidates for syncope) are indica ted at data points 3 and 4. 306

PAGE 307

24 20 12 16 Figure 5-32. Outputs of declension 2/4 pentasyllables with penultimate accent (n=23). Due to the small size of this data set and the register of the examples it is difficult to see any changes in output to support ar guments for any special constrai nts that are operative in the treatment of pentasyllables. Even the intuitive notion that polysyllables should be reduced to no more than three syllables is not evinced by sync ope, haplology or other repair strategies to eliminate syllables and thus diminish the overall size of words. The third declension subset does provide evidence of apocope in Cata lan which reduces overall word length to four syllables. Pentasyllables with penultimate accent are mu ch better represented in third declension nouns. This is due to morphological factors, that is, the high productivity of certain denominal suffixes. The first group to be examined combines the outputs of nouns formed with the suffixes -td, nis and -ts, tis The accusative of -t s, -tte(m) seems to have influenced the outcome of nouns formed with the suffix t d. The accusative suffix -td ne(m) is discernible in very few outputs, for example, Cast. muchedumbre multitude (
PAGE 308

sam e data set are Cat. tat, Cast. dad (occasionally -tad ), Port. dade (rarely tade ). Since the expected form for Catalan, dat with lenition of intervocalic /t/ rarely surfaces (cf. Cat. maldat < m l tas, tis, f.) it can be assumed that the suffix has become lexicalized and no longer responds to the phonological processes that act on inter vocalic stops in the formative period of the language. It is the analogi cal effect of a highly produc tive suffix of the form tVt(e) that results in tut in Catalan and its correspondents, tud and tude in Castilian and Port uguese. In the data set under analysis, 55 of the 60 forms come from the suffix t te(m) and only five from t d. It should be noted also that the outputs -tut, tud and -tude have a more transparent relationship with the nominative than -dumbre in patrimonial Cast. muchedumbre The outputs of this subset of 60 nouns are best analyzed in two sets, the first containing only Portuguese and the second Castilian and Cata lan. The results are presented as Figure 5-33 (Portuguese) and Figure 5-34 (Cat alan and Castilian). When these two subgroups are matched for syllable types in the initial and accented syllable and then filtered by non-apocope/apocope, the distribution of outputs is vi rtually indistinguishable. 0 10 20 30 40 50 60 POR 153266721 ,H.L.L.'L.L,H.X.X.'L.L,L.L.L.'L.L,L.X.X.'L.L,L.L.'L.L,L.H.'L.L,H.'L.L Figure 5-33. Portuguese outputs of third de clension pentasyllabic nouns with suffixes t te(m) and t d (n=60) 308

PAGE 309

0 10 20 30 40 50 60 CAT 16233621 CAS 15232632 ,H.L.L.'H,H.X.X.'H,L.L.L.'H,L.X.X.'H,L.L.'H,L.H.'H Figure 5-34. Catalan and Castilian outputs of thir d declension pentasyllabic nouns with suffixes t te(m ) and t d (n=60) Data point 1 in Figures 5-33 and 5-34 shows an optimal dactyl ( H.L.L) constructed at left word edge with representation at about 25% for th e three languages. The most represented foot at left word edge (data points 2, 4 for Portuguese; data point 2 for Catalan and Castilian) contains all light syllables. Data point 3 for all three languages design ates a noun with initial light syllable followed by either heavy or light syllables in the two pretonic positions. At right word edge Portuguese maintains faithfulness to the ( L.L) head foot while Catalan and Castilian prefer ( H), which preserves the moraic count but deletes the weak syllable of the foot. Both outcomes are well-formed trochees. In addition to apocope which reduces syllabl e count for both Catalan and Castilian, there are other mechanisms that result in loss of syllables. The first is glide formation that occurs primarily in Castilian and Portuguese when vowels in hiatus involve a se quence of an unaccented high vowel with an accented (primary or seconda ry accent) mid or low vowel. The second is deletion of intervocalic /l/ or /n/ and the subseq uent coalescence of vocalic nuclei that occurs only in Portuguese. These processes account for output s of less than five sy llables in Castilian. 309

PAGE 310

Finally, in all thr ee languages s yncope occurs on a very limited s cale in patrimonial words such as *amicitate (cf. Cl. Lat. m c t a, ae, f. friendship) which results in Cat. amistat Cast. amistad and Port. amizade The various suffixes that function as nomina actionis, coalesce in Late Latin as ion(e) Ouputs of this derivational process have the highest representation in the subset of third declension pentasyllables with penultimate accent. As a suffix used to produce deverbal nouns, the last three syllables often took the form -ti-o-ne or -si-o-ne .23 The high vowel combines with the preceding coronal in popular output s of the suffix resulting in Catalan [ so], Castilian zn (originally [ on], but [ on] in modern Castilian), and Portuguese o (originally [ on] but [ sw ] in modern Portuguese). This output is scarcely represented in the data set here. The usual output for Catalan in le arned or semi-learned words is [-i. o], the most faithful to the input in terms of preserving input syllable st ructure. As expected, the unaccented vowel of the final syllable is elided as well as the resulting fi nal /n/. It has been prev iously asserted in this study that accented vowels at word edge should be treated as long vowels. For the two languages that show vowel reduc tion, Catalan and Portuguese, ther e are measurable differences in terms of vowel quantity and quality between accented and unaccented vowels. Therefore, in moraic terms, Catalan has preserved the structure of the input. That Catalan is the most faithful is apparent in viewing the changes to the suffix -ione(m) as realignment of syllables, moras, and segments illustrated in Figure 5-35. It is assu med that the input form is trisyllabic and each syllable corresponds to a mora, that is, /i. o.ne/. Although the original tonic vowel is long in 23 The Latin perfect passive participle is characterized by t -. The alternate form in s results from phonotactic constraints. When the root ends in a dental stop, assimilation occurs at the boundary between root and suffix, resulting in -ss-, shortened to -sif a consonant or long vowel precedes (Miller 2006, 9). The same alternations of t and s seen in the perfect passive participle occur in the case of the highly productive verbal abstract suffix *ti-, augmented by n (Miller 2006, 97). 310

PAGE 311

Classical Latin it is treated here as monom oraic. While Catalan loses one segment through apocope, the tonic syllable retains the mora of the original ultimate syllable through vowel lengthening. Therefore, there are still three moras in the resulting suffix i which now has two syllables reduced from three. syllabic tier | | | | | | | | | moraic tier i o jon w segmental tier (output) A) Catalan B) Castilian C) Portuguese Figure 5-35. Realignment of syllables, mo ras, and segments in outputs of suffix -ione(m) in Catalan, Castilian, and Portuguese. Vowels in hiatus, faithful to the input, violate a basi c constraint for a well-formed syllable, ONSET (A syllable must have an onset.). Castilian, as expected, does not maintain hiatus and the pretonic high vowel forms an edge of the penul timate syllable nucleus resulting in the loss of a mora (broken line in Figure 5-35). As in Catalan, apocope eliminates the final vowel but because the onset of that syllable is maintained as the coda of the penultimate syllable the mora is preserved. As noted above, in many cases the suffix ione(m) is preceded by /s/ or /t/. Thus, in Portuguese the high vowel /i/ is absorbed by th e preceding coronal consonant and is preserved only in exceptional cases such as poetic setentrio north, north wind (
PAGE 312

The constraints in Table 5-21 ar e given as if unranked. Max/IO shows the least violations and m ay well be the highest ranking. Assuming that the various outputs of ione are productive suffixes that carry the primary accent in the prosodi c word it can be argued that the head foot must be binary at some level (syllable, mora). In the case of Catalan, although the suffix has two syllables they cannot form a foot because RhT ype(T), which defines the rhythmic type as trochaic, would be violated. RhType(T) is considered to be an inviolable c onstraint. Loss of /n/, in word final position subsequent to apoc ope, produces compensatory lengthening of the preceding vowel in the tonic syllable and thus maintains the required bi moraic weight of the head foot. If constraints were ranked as give n in Table 5-21, the Cata lan suffix would be the preferred output because it has no violations of MAX/IO. Table 5-21. MAX/IO constraints related to the output of suffix -ione i. o.ne MAX/IO MAX/IO MAX/IOSEG Catalan i '. o ** Castilian jon ** ** Portuguese w ** *** Because the suffix -ione(m) itself occupies the last three syllables of the word, the possible syllable patterns for left word edge can be r eadily deduced: HH, HL, LH, and LL. Taking into consideration the nature of the suffixes by langua ge, all possible combinations of heavy/light syllables can be arrayed as in Table 5-22. The location of the primary accent in the suffix is indicated. Type 1 suffix is present only in Cata lan; type 2 suffix occurs primarily in Castilian and Portuguese, as well as a fe w patrimonial words in Catalan. Type A stem corresponds to 66 input forms in this data set. In only one case does the Catalan outcome (suffix type 1) show reduction to monosyllabi c ['so], patrimonial infan low ranking member of the nobility (cf. Cast. infanz Port. infano ). Type B and C stems are less well represented in this data set: 34 (B), 42 (C). There are 46 cases of stems with light syllables (D). 312

PAGE 313

Table 5-22. Distribu tion of heavy/light syllables in pentasyllables with ione(m) suffix A B C D Stems Suffixes H.H H.L L.L L .H 1. L.' H H.H.L. H H.L.L. H L.H.L. H L.L.L. H 2. H H.H. H H.L. H L.H.' H L.L. H Expected outcomes for template 1A (Catalan) are H.H.L. H, H.L.L. H, and L.H.L. H. Loss of moras comes primarily as the result of degemi nation of obstruents, nasals, and liquids with occasional exceptions such as [d.d], [n.n], and [l.l] preserved in Catalan addici [ ddisi'o], connexi [kunneksi'o], collecci [ku ksi'o]. These pronunciations, however, are normative and there is widespread reducti on to a single consonant (Wheeler 2005, 36). Portuguese shows in a few cases elision of a syll able final obstruent not reflected in the other two languages. If output forms remain faithful to input expe cted outcomes are H.H + suffix (66), H.L+ suffix (34), L.H + suffix (42), and L.L (46) + suffi x. In Figure 5-36 it is readily seen that the first category (H.H + suffix) is underrepresented and the last (L.L + suffix) overrepresented. 0 30 60 90 120 150 180 CAT 45394360100000 CAS 00004131427211 POR 00233831427011 H.H.L.'HH.L.L.'HL.H.L.'HL.L.L.'HH.H.'HH.L.'HL.H.'HL.L.'HH.'HL.'H Figure 5-36. Outcomes of third decl ension pentasyllables with suffix -i ne(m) (n=188). In all three languages moras in heavy syllables in the stem of the input form are lost, although Catalan shows a slightly higher rate of preservation of coda consona nts. Outputs of this set of 313

PAGE 314

188 nouns in Figure 5-36 show a high degree of correspondence between data points 1 through 4 and data points 5 through 8. The prim ary differen ce in outcomes of this se t of pentasyllables is the preferred form for the suffix, disyllabic in Catalan and monosyllabic in Castilian and Portuguese. There is one other productive suffix repres ented in the subset of third declension pentayllables with penultimat e accent; it is the agentive t r(em) Although the suffix is highly productive the five-syllable subset contains only 18 items. In this case the output of the suffix is uniform across languages since / /, which emerges in word fina l position following apocope, is an acceptable coda consonant for all three langu ages. Although the suffix is highly productive the five-syllable subset contains only 18 items. Distribution of syllable types in the three initial syllables is nearly identical in the th ree languages as seen in Figure 5-37. -3 0 3 6 9 12 15 18 CAT 126250110 CAS 126240012 POR 116241012 H.H.H.'HL.L.H.'HH.L.L.'HL.H.L.'HL.L.L.'HH.H.'HH.L.'HL.H.'HL.L.'H Figure 5-37. Outcomes of third decl ension pentasyllables with suffix -re(m) (n=18). 314

PAGE 315

There are few cases of reduction to three syllab le s. These are the result of glide formation in Castilian and Portuguese (data point 3) with the exception of one word which shows syncope in all three languages and probably belongs to a popular stratum of the language: Cat. llaurador Cast. labrador Port. lavrador In general, the outcomes are highl y faithful to the original input. Loss of moras occurs in the cas e of geminate obstruents but coda consonants are generally retained. The two most salient patterns are in dicated at data points 1 and 2. Data point 1 corresponds to the optimal dactylic foot. Data point 2, with three li ght syllables at the left edge, seems a prime candidate for syncope yet there is onl y one incidence of syncope in this subset. The uneven distribution of data for polysyllables makes it di fficult to draw conclusions about the modification of longer words to produce a more optimal rhythmic outcome. However, it can be noted that most of the examples are lexi cally marked, that is, th ey belong to particular registers of the language that do not correspond to everyday usage. Th ere are within each subclass a small number of examples which seem to indicate a different stratum of the language, the patrimonial words which do show syncope and other segmental modifications indicative of early entry into the various Ibero-Romance la nguages at a time when phonological changes such as lenition, syncope, and a pocope were active. Five-syllable Nouns with Antepenultimate Accent As with the subset of pentasyllables with penultimate accent, many nouns with antepenultimate accent are the product of stem + suffix derivations. Several suffixes are well attested in nouns of the second declension in contrast with the third declension where there are virtually no pentasyllables with antepenultimate accent. One of the most frequent is the denominal suffix r u(m) There are two outcomes from this suffix: (1) faithful Cat. ri Cast./Port. rio and (2) patrimonial Cat. r (normally pronounced without final / /), Cast. ero 315

PAGE 316

and Port. eiro The faithful outcomes all reduce the orig inal input from three syllables to two and from three moras to two. Catalan, as antici pated also shows apocope of the final vowel. Although the word final vowel is retained in Castilian and Port uguese glide formation produces the same results as Catalan, that is, two light sy llables at right word edge. Only the popular outcome in Portuguese preserves the moraic count of the original suffix through the creation of a falling diphthong under influence of the following high vocoid. The suffix occupies the three rightmost syllables in the input form; this leaves two syllables at the left edge of the word as stem All possible combinati ons of heavy and light syllables can be inventoried as H.H, H.L, L.H, and L.L. Analysis of outcomes begins with the first two templates, both having heavy initial syllable. Although not all five-syllable second declension nouns are derive d through suffixation of r u(m) they are treated together here because the outcomes are fairly uniform. The most common outcome is a four-syllable word which forms two syllabic trochees: ( X.X)( 'X.X). The head foot now has the familiar pattern of the predominant penultimate accent. There are very few cases in which the five syllables of the Latin correspondent are retained with the marked antepenultimate accent. The first subset, consisting of 53 items contai ns nouns with an initial heavy syllable. Retention of the initial heavy syllable occurs at the following rates: Catalan 46 (86.8%), Castilian and Portuguese 45 (84.9%). This is seen in the two predominant patterns at data points 1 and 2 in Figure 5-38, as well as data point 3 fo r Castilian and Portuguese. None of the highly represented patterns retain the an tepenultimate accent of the input fo rms. In fact, Catalan has no representation of the original pentasyllable with antepenultimate accent; in Castilian there are 6 (11.3%) cases and 8 (15.1%) in Portuguese. The input form is highly marked for two reasons: the accentual pattern and the length of the word. The three most common patterns seen in Figure 316

PAGE 317

5-35 have a unif ying factor, a syllabic trochee built at the right word edge. In the case of the suffix r u(m) this is accomplished through apocope in Catalan and glide formation in Castilian and Portuguese. These differing re pair strategies have the same result in prosodic terms. In a system of syllabic trochees it is possible to build initial trochees for both H.H and H.L at the left word edge although neither confi guration is optimal. 0 6 12 18 24 30 36 42 48 54 CAT 00011525333012 CAS 23111217125900 POR 24211216124900 ,H.H.'L.L. L ,H.L.'L.L. L ,L.L.'L.L. L ,H.H.'H.L,H.H.'L.L,H.L.'L.L,H.L.'X.X,L.H.'L.L,L.L.'L.L,H.L.'L,L.L.'H,H.'H ' Figure 5-38. Outcomes of second declension pe ntasyllables with an tepenultimate accent and heavy initial syllable (n=53). This subset contains one case of accent shift, Cast. amoniaco (
PAGE 318

incn blu(m ), analogic singular from inc n bla, rum, n., swaddling clothes). Vowel loss is not inevitable in Catalan treatment of this suffix; for example, mont c lus, i m. small mountain produces both monticle and less frequent variant montcol Table 5-23 shows that the variant suffix is the most faithful although the prosodi c output is less ha rmonious, that is, H. L.H versus H. L.L in the form that has undergone syncope. Table 5-23. MAX/IO constraints related to the output of suffix culu(m) MAX/IO MAX/IO MAX/IOSEG ku.lu -kl *! ** -kul There is a small group of nouns (26) in the set of pentasyllables that have an initial light syllable followed by a heavy syllable. While L.H is a less than optimal trochee it is still possible to build a foot at the left edge of the word. There is no evidence of deletion of either of the initial syllables as an attempt to produce a more harmonious foot structure. Only Portuguese mosteiro (m nast r um, ii, n. monastery) shows loss of the peninitial syllable but this is the result of deletion of intervocalic /n/. This same word in Catalan shows unusual development as it is one of the few cases in which final Vriu loses both the final vow el and the vowel of the penultimate syllable resulting in monestir pronounced [mun s ti]. Given the vocalism of the tonic syllable it can be assumed that metathesis of the high vowel in ariu accounts for both vowel raising and loss of expected wo rd final /i/. The predicted form monasteri (DCVB, s.v. monasteri) is a little used Latinism in modern Catalan. The rhythmic output of monestir is ( L.H)( H), two syllabic trochees as is that of variant monasteri ( L.H)( L.L). However, the more frequent form has the advantage of satisfying R-ALIGN for both foot and head. In this small set of nouns with initial L.H foot there is also limited evidence of syncope in Catalan: receptacle (
PAGE 319

reduction strategies seen he re, apocope (in the case of monestir ) and syncope, effectively create a m aximum word size of four syllables with no re presentation of antepenultimate accent in fivesyllable words. It appears to be the combina tion of marked prosodic te mplate together with dispreferred word length that prev ents the faithful output from being the preferred output. In the two cases of syncope cited here further reducti on is not possible because the resulting consonant sequences would violate SONSEQ. Outcomes from all three languages are seen in Figure 5-39 where data point 1 indicates pentasyllables for Castilian and Portuguese only. Data point 2 indi cates the most common output for all three languages influenced by the disy llabic outcomes of the originally trisyllabic suffix ariu Only Catalan shows instances of prim ary accent on the ultimate syllable (data point 3) as output from this same suffix, for example the case of monestir monastery discussed above, as well as taverner (< t bern r us, ii, m. shopkeeper), cf. Cast. tabernero Port. taberneiro All other outputs for this small data set have a disyllabic head foot constructed at right word edge. It should be noted that th e optimal trochee (L.L) em erges in 22 of 26 cases (84.6%) in Catalan, 25 in Castilian (96.2%), and 23 in Portuguese (88.5%). -5 0 5 10 15 20 25 CAT 0220200002 CAS 4018301000 POR 4016311100 ,L.H.'L.L. L ,L.H.'L.H,L.H.'L.L,L.L.'L.L,L.H'H.L,H.'L.L,H.'H.L,L.L.'L.L,L.H.'H ' Figure 5-39. Outcomes of second declension pent asyllables with antepe nultimate accent and LH initial syllables (n=26). 319

PAGE 320

The last subset (75 item s) of pentasyllables consists primarily of words with all light syllables. Exceptions are the outputs of aud t r um, ii n. lecture-hall, auditorium which retains the initial diphthong in a ll languages and outputs of praej d c um, ii, n. preceding judgement in Catalan and Castilian which show in fluence of an alternate prefix, per, not apparent in Portuguese. There is an expectation of increas ed frequency of apocope in Catalan since SONSEQ is less likely to force retention of a vocalic element to resolve an unacceptable consonant sequence. There are six words in Catalan reduced to two syllables that show the effect of both syncope and apocope. These words are clearly of a more popular nature than the other items in this subset as seen in the list of Ibero-Romance correspondents below. Catalan Castilian Portuguese F br r us, ii, m. febrer febrero fevereiro J n r us, i, m. January (VL genuariu) gener enero janeiro c m r r us, ii, m. chamberlain cambrer camarero camareiro p r r us, ii, m. laborer obrer obrero obreiro *semitariu pathway (Du Cange 1710, s.v. semitariu) sender sendero sendeiro s l t r us, a, um, adj. solitary (>bachelor ) solter soltero solteiro As seen in the previous set Catalan is never faithfull to the five syllable/antepenultimate accent of the original. In the current set of 75 items there are no words with antepenultimate accent in Catalan whereas both Castilian and Po rtuguese have a small percentage (18.7%) of learned words that are faithful to the input bot h in terms of word length and placement of primary accent (data point 1 in Figure 5-40). Th e favored outcome for all three languages is the four-syllable word with penultimate accent (data point 2) where the rate of occurrence is 65.3% for Catalan, 62.7% for Castilian, and 61.3% for Po rtuguese. In prosodic terms, the resulting ( L.L)('L.L) is optimal, that is, two well-formed syllabic trochees construc ted at left and right word edges. Setting aside the variab le nature of heavy and light syll ables at the left word edge it can be seen that the emergent pattern in all pentasyllables with antepenultimate accent (Figures 320

PAGE 321

5-38, 5-39, 5-40) is a four-syllabl e word with two syllabic troch ees. Surprisingly, there is no correspond ing set of pentasylla bles with antepenultimate accen t. This may be due to morphological factors, that is, the absence of a three-syllable accent-bearing suffix with accent on the first syllable. Consequently, there is no an alysis of the handful of Greek borrowings with five syllables and antepenultimate accent. -15 0 15 30 45 60 75 CAT 00249016001133 CAS 14314700306001 POR 14314630224000 ,LL.'LLL.'L.L. L ,H.L.'L.,L.L.'L.,L.L.'H.,L.L.'L.H.'L.LL.'H.LL.'L.L,L.L.'HH.'HL.'H ' Figure 5-40. Outcomes of second declension pent asyllables with antepe nultimate accent and all light syllables (n=75). Summary of Effect of Apocope on Prosodic Ou tcomes of Nouns from Declensions 2/4 and 3 Emergence of ultimate accent is a concomitant of apocope and is most highly represented in third declension nouns whose stem ends in a li quid or /n/. In those cases apocope applies across languages although the treatme nt of word final /n/ in both Catalan and Portuguese may at times obscure the nature of the input form. Apoc ope is blocked in Catalan only if a sequence of final consonants results that w ould violate sonority sequencing. Ca stilian requires that the coda consonant be coronal whereas Portuguese bloc ks final obstruents with exception of /s/. Therefore, the occurrence of apocope shows a gradient effect among the three languages. It has been suggested that the elision of final e is favored on bot h morphological and phonological grounds. The third declension is heterogenous morphologi cally and lacks a 321

PAGE 322

charac teristic thematic vowel (other than the i stems), but e(m) is the characteristic ending of the accusative singular. Furthermore, the third declension includes all genders masculine, feminine, and neuter. The association between declension class and gender assignment is, therefore, lacking. For this reason a morphologi cally based constraint has been suggested to block apocope, MAXMORPH. MAXMORPH proves to be a convenient de vice in the case of the homogenous first declension but it leaves a degree of dissatisfac tion with regard to the second and third declensions where apocope occurs with equal vigor in Catalan. It is necessary, therefore, to turn to phonological explanations. Final vowel inventories in languages with vowel reduction are usually limited in number and quality when compared to th e vowel inventory of privileged pos itions. The corner vowels: /i/, /a/, /u/ seem less vulnerable to processes of reduction, especially extreme reduction or loss. Survival of /a/ is a characteristic of first decl ension nouns. However, the vowels sonority makes it unsuitable as a peak in a non prominent positio n. Reduction of /a/ is a known feature of both Catalan and Portuguese. Both the vowels that are eliminated in Catalan fall outside of the parameters of corner vowels. Therefore, on phonological grounds they may be more likely candidates for apocope. The second declension vo wel (probably [o] in Proto Ibero-Romance) undergoes vowel raising in Portuguese and emerges as [u], a corner vowel. The question, then, is what constraint is rewarded by elimination of an unaccented final vowel. Both ('L.L) and ('H) are valid foot forms for the syllabic trochee. The monosyllabic variant offers as advantage alignment of the head syllable with the right word edge. If /a/ does not delete this is probably more in recognition of its sonority and suitability as a syllable peak rather than its value as a morphological marker. 322

PAGE 323

CHAP TER SIX PREFERRED PROSODIC TEMPL ATES: CONCLUSIONS Patterns with Penultimate Accent and /-a/ Class Marker Discussion begins with the group of nouns that have /-a/ class marker. Although usually associated with gender, preservation of word final /a/ seems to res pond equally to phonological and morphological criteria. Th e outcomes of all nouns resulting from two and three-syllable input forms with penultimate accen t are illustrated in Figure 61. The two monosyllables are anomalous (Portuguese l wool < l na, ae, f.; p spade < p la, ae, f.) and result from loss of intervocalic /n/ and /l/ peculiar to Portuguese. In terms of input, there are 232 disyllables, or 56% of the total, and 182 trisyllables, 44%. Figu re 6-1 shows that the two-syllable template has gained in strength, although slightly, and now comprises 57.4% of the total for Catalan, 57.9% for Castilian, and 58% for Portuguese. -40 0 40 80 120 160 200 240 CAT 0144884209578129 CAS 0149836205588526 POR 2144867115559023 'H'L.L'H.L'L.L'H.L'H.HH'H.LH'L.LL'L.LL'H.L < Figure 6-1. Prosodic outcomes of all 2-syllable and 3-syllable first declension nouns with penultimate accent (n=414). Despite a few language-specific deviations Figure 6-1 shows nearly identical outcomes for the three languages in terms of number of sylla bles, coincidence of primary accent with heavy 323

PAGE 324

syllable, prosodic tem plates, and percentage distribution for each template. It is noteworthy that initial heavy syllables occur with approximately the same frequency in both disyllables and trisyllables although in the case of trisyllables the initial syllable does not bear primary accent. This is more readily seen when distribution is expressed in percentages for each of the three languages. The dominant pattern in both disyllables and trisylla bles with penultimate accent has a light initial syllable (data points 1 and 4 in Figure 6-1). In the case of the disyllable, this results in the harmonious trochee ('L.L). The less frequent patterns with initial he avy syllable appear at data points 2 and 3. Portuguese offers one a dditional pattern with a unique representation, (H.'H), exemplified by irm sister (< germ na, ae, f.). Preservation of moras in the penultimate or accented syllable is poorly represented by the patterns H('H.L) and L('H.L), both with relative ly low percentages. This suggests that the preference for heavy syllables has more to do w ith positional prominence, that is, word initial position, rather than a residual appl ication of STW. Previous disc ussion has noted the effects of positional prominence at left word edge. In term s of the metrical grid, both the initial syllable and the penultimate syllable would receive marks at level 1. This assumes that there are reasons, perhaps morphological, to consider the initial heavy syllable to be a foot rather than an unparsed syllable. Finally, a further mark at level 2 is assigned only to the syllable with primary accent. This places it in the position of greatest prominen ce and avoids stess clash. The final syllable, the weakest, is assigned a grid mark only at level 0 where each syllable is recognized in the prosodic structure. Frequent coincidence of th e initial syllable with a morpheme or prefix provides further motivation for this secondary accent. The final parsing results in ( H)('H.L). It would be more difficult to make this case when the first syllable is light because (L) cannot constitute a binary foot; it lacks the minimum two moras or two syllables. 324

PAGE 325

Level 2: Assign to the head foot (* ) Level 1: Assign to the head of every foot (*) (* ) Level 0: Assign to every syllable Although L('H.L) does not seem to lend itself to an analogous interpretation it should be noted that there is anecdotal evidence of emphasis of initial light syllables in word plays. For example, an announcement for a brand of candy in Spain describes it as being relleno y rebueno (filled and delicious) The initial syllable of relleno with a simple stem and derived re-bueno is emphasized in both cases and thus constitu tes a secondary accent in a 2-3-1 pattern of intensity with 3 corresponding to the tonic syllable. An alternative appr oach to the LHL type word is to compare it to the wsw pattern of the amphibrach, the me ter selected to exemplify the use of metric verse in Spanish by the modernist poet Rubn Daro in Mar cha triunfal (Navarro Toms 1956, 515). As an example, the first two lines of the text are given with prosodic marking displayed to the right including division into syllables and underscoring of the head of the foot. Ya viene el cortejo! (ya.vie .nel) (cor.te .jo) Ya viene el cortejo! Ya se oyen los claros clarines. (ya.so .yen) (los.cla .ros) (cla.ri .nes) The amphibrach is a rhythmic template not unnatural to the languages under study and one which occurs frequently across word boundaries when a monosyllabi c definite/indefinite article is followed by a two-syllable noun with penultimate accent as in (los.cla .ros) above. There are examples of 6+6 verses (six syllables in each he mistich) in Catalan, Cast ilian, and Portuguese in the Pan-Hispanic Ballad Project where the appare nt accentual pattern matches the amphibrach. Furthermore, the wsw template is acquired in early ch ildhood for both Catalan and Spanish (Prieto 2006) although with some inst ances of elision of the weak in itial syllable in the earliest stages of acquisition. Suppression of the initial weak syllable is also at tested in hypocoristic formation in all three languages. Despite the hi gh rate of retention of coda consonants, The 325

PAGE 326

initial syllab le is still perceive d to constitute a sort of anacrusis whose main function is to serve as upbeat for the head foot. Morphological and/or semantic criteria can assign a foot to the initial syllable even if it is light. Howeve r, this is infrequent and context dependent. The relationship of amphibrach to more co mmon feet in poetry is described by Devine and Stephens (1993, 396). In metrical patterns, run unity implies that a se ries of amphibrachs will tend to be analyzed as dactyls or anapaests even if it begins with an amphibrach, provided the amphibrachs are presented continuously, that is, provided the amphibrach structure is not reinforced by demarcative pauses; the initial amphibrach is in terpreted as anacrustic. This ties in with what we know about language and about Greek metre. Languages rarely if ever have amphibrach stress feet. With relation to the current prob lem several assumptions need to be made. First, the short-longshort pattern of Greek can be replaced by unacce nted-accented-unaccented. It is also assumed that syllable weight is immate rial. Devine and Stephens ( 1993, 395-396) discuss the convention of line initial freedom in verse, that is, the init ially chosen hypothesis can be set aside for a better one that emerges in the middle of the line of verse. In other word s, errors at the beginning of the parsing sequences are less serious than deviations that occur once a pattern has been established. If in a verse there is an attempt to incorporate an initial unaccented syllable it is possible to set it aside if the predominant form is later seen to be a trochee, for example. The initial syllable is now reinterpreted as anacrusis. More revealing are the distribut ions of heavy and light syllables in prominent position (that is, initial) of two and three-syllable words. A brief review of the sources of heavy syllables is in order. Of the 232 nouns that begin as di syllables, 192 (82.8%) have a heavy initial syllable as input. However, only 121 (52.2%) have initia l syllables of the HC type with a coda consonant. Of the 182 trisyllables with penultim ate accent, 75 or 41.4% have HC type initial syllables. Together, these 196 cases, 47.3% of the data set of 414 words, are the best candidates 326

PAGE 327

for continuation of a heavy initia l syllable. Table 6-1 shows the actual outputs. The rate of retention of heavy initial syllable is relatively high (percentage of inputs realized as heavy syllables in parentheses) a ssum ing that 47.3% is the highest total percenta ge possible. Table 6-1. Distribution of initial heavy syllables in disyllables and trisyllables from first declension nouns Catalan Castilian Portuguese Catalan Castilian Portuguese Input 'H.L in 2 Syllable Outputs 'H.X.L in 3 Syllable Outputs 'HC.L (121) 88 (72.7%) 83 (68.6%) 86 (71.1%) 0 (0.0%) 0 (0.0%) 0 (0.0%) HC.'H.L (75) 2 (2.7%) 2 (2.7%) 2 (2.7%) 66 (88.0%) 63 (84.0%) 60 (80.0%) Subtotals 90 85 88 66 63 60 Totals 156 148 148 Total Percentage 37.7% 35.7% 35.7% Note: Total percentage is based on total number of 2 and 3 syllable inputs, 414 (including inputs with L or HV initial syllables. Words with an even number of syllables do not present this problem. Four-syllable words with penultimate accent are easily parsed in a preferred rhythmic pattern. Figure 6-2 shows a low level of syllabic loss in the three languages although Casti lian (14 of 110) has a greater incidence of reduction to three sy llables than do Portuguese (7 of 110) and Catalan (7 of 110). 100 8060 40 20 0 -20 (L.L)('L.L ) (H.L)('L.L) (L.H)('L.L) L('L.L ) L('H.L) Figure 6-2. Outcom es of all first declension te trasyllables with penultimate accent (n=110). CAT 45 20 13 3 21 1 052 00 CAS 4721 12 213 1 045 50 POR 47 21 12 2 19 1 142 01 (H.H)('L.L) (L.L)('H.L ) (H.H)('H.L) (H.L)('H.L) H('L.L) (L.L)('H) ,' ,' 327

PAGE 328

W ithin the subgroup that retains the four syll ables of the input two patterns emerge. The first (data point 1) has all light syllables, that is, two harmonious trochees: ( L.L)( L.L). The second most represented pattern cros s-linguistically (data point 2) also results in two trochees but retains a heavy initial syllable. Conservation of a coda consonant in a heavy initial syllable is about as frequent as in the toni c syllable (data points 3 and 5). With exception of the higher rate of reduction to three syllables in Castilian and the concomitant lower rate of four-syllable outputs, the three languages show remarkable convergence in outcomes of four-syllable input with penultimate accent. This has also been the case with input forms of two and three syllables (cf. Figure 6-1) where there is coincidence in preservation of syllables in input and template types. Within each group there is a notable preference for the most harmonious trochee, ( L.L). There is evidence that ('H) is an admissible foot form even in words of more than one syllable (data point 4) although in Figure 6-2 initial H type syllable is left unparsed. However, the H( L.L) type words differ somewhat from the templa te at data point 6 which corresponds to only one exemplar. In the case of the monosyllable, the most basic and inviolable OT constraint, Lx Pr (A lexical item must be pr osodically analyzed, Jacobs, 2003b) licenses a foot regardless of syllable type although most monosyllables are bimoraic with few exceptions. Therefore, the one exception to the (' L.L) type head foot can be considered licit, that is, ( L.L)('H), corresponding to Portuguese avel (
PAGE 329

The set of five-syllable nouns with penultim ate stress is small (42 items) and skewed in terms of lexical register. These words reflect the vocabulary of philosophy and religion as well as scientific terminology. Aside from the problematic aculeata (pertica)/aquileata goad, there are really no items that correspond to the popular lexicon. Surprisingly only two items in this set receive assignment of th e penultimate accent in input form due to an HC type syllable. In 40 of 42 cases the syllable that attracts th e primary accent under the STW principle is an HV type syllable which in Ibero-Romance becomes indistinguishable from a light syllable. In other words, the almost universal form of the head foot in this small class is ('L.L). In a few cases (5 in Castilian and 4 in Portuguese) there is a reduction to four syllabl es which produces the familiar pattern of two disyllabic trochees discussed above. Figure 6-3. Outcomes of all first declension pentasyllables with penultimate accent (n=42). For the most part, however, the five-syllable pa ttern persists and the reduced four-syllable type is present in only 2.4% of cases in Ca talan, 9.5% in Portuguese, and in Castilian, predictably the highest, 11.9%. Within the five-syllable template the two most common forms are L.L.L('L.L) and H.L.L('L.L) seen at data poi nts 2 and 1 in Figure 6-3. The former, all light 42 36 30 24 18 12 6 0 H.H.L( 'L.L) H.L.H('L.L) L.L.L('H.L) L.L('H.L) CAT 12 10 4 319 20 1 CAS 12 84 219 114 POR 11 94 220 113 H.L.L('L.L) L.H.L( 'L.L) L.L.H('L.L) L.L.L('L.L) L.L('L.L) ,' 329

PAGE 330

syllables, is the favored pattern in all languages and represents over 45% of the output for ms. Both of these highly represented templates exhibit the initial dact yl effect. Once again it can be seen that the coincidence of primary accent and h eavy syllable is lacking in all three languages under consideration. Preservation of the heavy initial syllable can be attributed to positional prominence rather than prosodic necessity although H.L.L is a more optimal dactyl. It should be remembered that while the dactyl is a common artifact of Greek and Latin verse it is not a prosodic foot of any language. Even as a poetic device the dactyl alternat es with the spondee. Figure 6-4 shows the degree to which there is a correspondence between syllable weight and positions of prominence. It is assumed that word initial position and the tonic syllable are both prominent in the structure of the word. The introduction of segments to increase prominence is not well attested in Ibero-Roman ce. Insertion of /b/ in the case of Castilian nombre name (
PAGE 331

Patterns w ith Antepenultimate Accent and /-a/ Class Marker The 157 nouns in trisyllabic repres entatives of this class conser ve syllabic count to a high degree in Catalan and to a lesser degree in Castilian and Portuguese Loss of a syllable nucleus alters the rhythmic template because there is no longer a sequence of an accented syllable followed by two unaccented syllables. As previously discussed these differences, displayed in Table 6-2, emerge more as a product of glide formation than syncope. Although proparoxytonic accent is a marked pattern in a ll three languages it is more stab le in Catalan even when the vowels in hiatus consist of a high unaccented vowel and a non high accented vowel. However, recent studies (Cabr and Prieto 2004, 2005) indica te that innovative varieties of Catalan show considerable erosion of conserva tion of hiatus which is necessary to maintain the proparoxytonic accent. In future the distribution of three-syll able and two-syllable outcomes may become more alike across languages. Table 6-2. Syllable count in outputs of first declension trisyllables with antepenultimate accent 3 Syllables 2 Syllables Catalan 107 68.2% 50 31.8% Castilian 80 51.0% 77 49.0% Portuguese 87 55.4% 70 44.6% In Figure 6-5 the distribution of the different prosodic templa tes is shown for all outputs of first declension trisyllables with antepenul timate accent. For both three-syllable and twosyllable outputs the most common template consists of all light syllables (data points 2 and 4). The input form exhibits variability in syllable we ight only in the initial syllable which may be heavy or light. With regard to the initial syllable, there is an expectation that only syllables which depend on a coda consonant for the second mora are likely to continue as heavy syllables in Ibero-Romance (although there are a few cases of conservation of the diphthong /aw/ in learned words). In the input data there are 98 words with heavy initial syllable (50 HC type, 48 331

PAGE 332

HV type), 62.4% of the item s in the data set. A dding together the ('H.L.L) and ('H.L) columns, indicated by data points 1 and 3, the differences among languages with regard to the heavy initial syllable are shown to be minor: Catalan 38.9% Castilian 33.1%, and Portuguese 35.7%. Figure 6-5. Prosodic outcom es of all 3-syllable first declension nouns with antepenultimate accent (n=157). Reduction to two syllables is the most not able difference among th e three languages. Antepenultimate accent is marked in all three language s. Therefore, reduction to two syllables is motivated by the desirability of the familiar trochaic template. Movement of the primary accent from the antepenultimate to the penultimate syllable is barred by HEADMAX (A stressed element in the input must have a stressed element as its output correspondent). In order to move the head foot closer to the right edge a syllable nucleus must be remove d. Apocope does not occur in any first declension noun so the default candidate for vowel elision beco mes the penultimate syllable. The usual catalyst is glide form ation which occurs at a higher ra te in Castilian and Portuguese, languages that disfavor vowels in hiatus. Elimination of the post -tonic syllable in words that reduce to two syllables occurs in 49% of cases in Castilian a nd 44.6% in Portuguese. The cases 0 20 40 60 80 100 120 140 160 'H.L.L CAT 35 72 26 24 CAS 23 57 29 48 POR 21 66 35 35 'L.L.L 'H.L 'L .L 332

PAGE 333

that coincide in the three la nguages involve syncope in patrim onial words (see Table 4-23 in Chapter 4). Antepenultim ate accent in the case of four-s yllable nouns leaves the initial syllable stranded. The initial syllable may have increased prominence when it is also a morpheme, that is, a prefix, but that occurs in only 18 out of 165 cases in this data set. Apart from those cases in which morphological considerations are clearly a determinant in assignment of a secondary accent to the initial syllable it is probable that onl y one foot is constructed at right word edge. Output results shown in Figur e 6-5 again demonstrate the polarizing effect of glide formation with Catalan having a high number of tetrasyllables with antepenultimate accent whereas Castilian and Portuguese a preponderance of trisyllables with penultimate accent. The distribution in number and percent of three a nd four-syllable outputs by language is shown in Table 6-3. Retention of all four syllables by and large means retention of locus of accent. Table 6-3. Distribution of three a nd four-syllable outputs from first declension tetrasyllables with antepenultimate accent 4 Syllables/Antepenultimate 3 Syllables/Penultimate Catalan 127 77.0% 38 23.0% Castilian1 42 25.5% 119 72.1% Portuguese2 47 28.5% 117 70.9% n=165 5. Total excludes 4 tetrasyllabl es with penultimate accent. 2. Total excludes 1 tetrasyllable with penultimate accent. This subset does contain a few cases in which th ere is an actual accent shift: However, this Castilian and Portuguese areola areole, Castilian only, aureola areole, laureola laurel crown, peona peony. These words all appear to have been introduced into the lexicon after the formative period of the languages under study. The four in Castilian according to CORDE enter the lexicon between the 13th and 15th cen turies. The Corpus do Portugus indicates a somewhat later date, 17th century, for areola in Portuguese. The shift in accent is best viewed as influence of fase analogy or hypercorrection rath er than conformity to a majority prosodic 333

PAGE 334

pattern. The m ost common template for the three languages consists of a ll light syllables (data points 4 and 8). The key question is whether the h ead foot is the familiar disyllabic trochee or the marked dactyl of the input form. The predom inant patterns for Catalan are all dactylic. 165 145 125 105 85 65 45 25 5 -15 H.'H.L.L H.'L.L.L L.'H.L.L H.'L.L L.'L.L Figure 6-6. Prosodic outcom es of all 4-syllable first declension nouns with antepenultimate accent (n=165). The most frequent pattern in Catalan (data point 4 in Figure 6-6) consists of all light syllables and has its counterpart at data point 8 for Castil ian and Portuguese, a trisyllable with penultimate accent and all light syllables. Init ial heavy syllables are represen ted at data points 1 and 2 for Catalan and data points 5 and 7 fo r Castilian and Portuguese. The i nput data for this set contains 63 out of 165 cases of an initial heavy syllable w ith coda consonant or falling diphthong likely to be preserved. These initial heavy syllables ar e preserved in 55 of 63 cases (87.3%) in Catalan and Castilian and 53 of 63 cases (84.1%) in Portuguese. In terms of the tonic syllable only 45 out of 165 are heavy with a coda consonant in the input data. However, in a fe w patrimonial words syncope creates a heavy toni c syllable after CAT 12 30 29 56 0 0 0 6 12 20 0 CAS 1 8 8 25 2 2 14 23 27 54 1 POR 0 9 10 28 0 1 13 29 28 47 0 L.'L.L.L ,H.L.'L.L ,L.L.'L.L H.'H.L L.'H.L H.'L ' 334

PAGE 335

deletion of a pretonic vowel. Thus, there are 49 cases where the ton ic syllable is potentially a heavy syllable. Retention is high in all three languages: 47 (95.9%) in Catalan, 46 (93.9%) in Castilian, and 44 (89.8%) in Portuguese (see data points 1, 3, 5, and 6). In addition, another 6 cases in Portuguese have a heavy tonic syllable due to the development of a falling diphthong as output of the suffix r a which becomes eira. It has already been seen in the case of outputs of original Latin disyllables and trisyllables that there is a tenuous relatio nship between the weight of a syllable and its ability to bear the primary accent of the word. Even those tonic syllables that began as bimoraic syllables and remain h eavy are far outnumbered by light tonic syllables within words of same syllable leng th and same declension class. The predominant pattern seen in the case of tetrasyllabic output, L( L.L.L), is not optimal. The inability to parse the initial syllable makes le ft alignment impossible. The question of right alignment is more complex. The active cons traint is assumed to take the form ALIGN-R(PrWd, Ft): The right edge of a word is aligned with the right edge of a foot. If the word is parsed as L( L.L.L) the alignment constr aint is satisfied but RHTYPE=T (Feet are trochaic) is violated. Deletion of the final vowel would satisfy TROCH (Feet are trochaic) but this never occurs in class /-a/ nouns. Historical syncope, the poetric tradition, and truncation processes converge in designating the penultimate vowel as the weak vowel. Piero s (2000a, 19) suggests that the weak vowel does not project a mo ra, that is, the constraint *MORA[V] is active. (6.1) MORA[V]: Do not associate a mora with a vowel. (Pieros 2002a) As further evidence, he cites hypocoristics fo rmed from names with proparoxytonic accent: Hiplito Polo, Aristbulo Tobo, Mlida Mela. Cases of historical syncope have been pr eviously observed and are exemplified by t b la, ae, f. plank > Cat. taula Cast. tabla, Port. tabla Only Catalan preserves the mora of the lost 335

PAGE 336

vowel in the falling diphthong of the word initial syllable. Syncope does not seem to be an active process synchroni cally and is documented primarily in patrimonial words. Further evidence that the weak penultimate syllable is weightless comes from the conventions used in metrical analysis in all three languages descri be where words with antepenultimate accent are described with the adjective Cat. exdrixol/a Cast. esdrjulo/a, Port. esdrxulo/a When a word of this type is aligned at the right edge of a verse or caesura, the penultimate syllable is not considered, either to maintain the syllable count of the verse or to esta blish rhyme (assonance of the last accented vowel and fina l vowel, if there is one). Ta ken as a whole, these various processes, both synchronic and diachronic, indicate that L( L.L.L) is the best parsing for tetrasyllables with antepenultimate accent with the caveat that the head foot should be considered to have two moras and not three. The interpretation of the trisyllabic head foot proposed here will also be required in the case of pentasyllables with penultimate accent. Th e major difference is that parsing of the three rightmost syllables in a foot doe s not leave the word initial sylla ble stranded in those cases in which the five syllables of the input are retained. However, onl y Catalan shows a high degree of retention of the five-syllable i nput. Castilian and Portuguese regularly reduce input to four syllables and occasionally three. The location of stress is the penultimate syllable which produces the familiar pattern of two parsable troch ees constructed at word edges. It is notable that this subset is one of the few instances of accent shift; 9 words of 77 re tain all five syllables in Portuguese but shift accent to the right. This configuration still allows the construction of two trochees at word edges leaving stranded an internal syllable. However this is still an acceptable outcome: ( ma.qui)na('ri.a) machinery. Only Catalan retains the faithful prosodic template: ( )(' ). If it is accepted that a trisyllabic foot can be formed at word edge and treated as if it 336

PAGE 337

were disylla bic, then the two most common outcomes are equivalent in prosodic terms. This is preferable to introducing the notion of extrametricality for th e final syllable. The previously presented evidence from poetic convention a nd hypocoristic formati on militate against extrametricality. Figure 6-7 shows the distribution of heavy and light syllables in prominent positions for all nouns derived from first declension nouns with antepenultimate accent. The results for two and five-syllable outputs (data points 1-3 and 4-6) are somewhat skew ed. The fact that there are output forms with two syllables from nouns with antepenultimate accent (requiring a minimum of three syllables) is often indi cative of syncope or glide forma tion, a process with the potential to create heavy syllables. The five-syllable subset is very small and uneven in distribution (67 exemplars in Catalan, 3 in Cast ilian, and 12 in Portuguese). Figure 6-7. Distribution of heavy/light syllables in initial and tonic positions of all nouns with antepenultimate accent from the first declension (n=1197). The distribution of heavy/light syllables in nouns of penultimate accent does not display the anomalies seen in Figure 6-7 although differen ces are also more pronounc ed in the case of polysyllables. The convergence of lines seen at data points 1, 2, 3 is attribut able to the fact that -20.0% 0.0% 20.0% 40.0% 60.0% 80.0% 100.0% CAT CAS POR CAS CAT CAS POR CAT CAS POR Tonic H 52.0% 37.7% 50.0% 27.6% 32.1% 34.2% 33.1% 40.4% 43.9% 62.7% 0.0% 0.0% Tonic L 48.0% 62.3% 50.0% 72.4% 67.9% 65.8% 66.9% 59.6% 56.1% 37.3% 100.0% 100.0% Word Initial H 52.0% 37.7% 50.0% 33.6% 34.0% 31.1% 33.1% 27.9% 28.6% 35.8% 0.0% 0.0% Word Initial L 48.0% 62.3% 50.0% 66.4% 66.0% 68.9% 66.9% 72.1% 71.4% 64.2% 100.0% 100.0% CAT POR 2 SYLL 3 SYLL 4 SYLL 5 SYLL 337

PAGE 338

in disyllables the tonic syllable a nd word initial syllable coincide. In the cas e of nouns of three, four, and five syllables, however, it can be seen that the ton ic syll able is more likely to be light than heavy and that the word init ial syllable is also more likely to be light than heavy. This preferred pattern of light tonic syllable and light word initial syllable is more prominent in polysyllables and in particular in the small subset (42 items) of pentasyllables (data points 4, 5, 6). For first declension nouns it is difficult to assert that there is any correspondence between syllable weight and prominent positions in the word. -20.0% 0.0% 20.0% 40.0% 60.0% 80.0% 100.0% Tonic H 37.9%35.8%37.1%22.0%23.6%21.4%20.0%17.3%20.0%4.8%4.8%4.8% Tonic L 62.1%64.2%62.9%78.0%76.4%78.6%80.0%82.7%80.0%95.2%95.2%95.2% Word Initial H 37.9%35.8%37.1%37.4%35.7%33.5%26.4%26.4%26.4%31.0%26.2%26.2% Word Initial L 62.1%64.2%62.9%62.6%64.3%66.5%73.6%73.6%73.6%69.0%73.8%73.8% CATCASPORCATCASPORCATCASPORCATCASPOR 2 SYLL 3 SYLL 4 SYLL 5 SYLL Figure 6-8. Distribution of heavy/light syllables in initial and tonic positions of all nouns with penultimate accent from the first declension (n=1698). Patterns with Penultimate Accent and /o/ Class Marker or No Class Marker Outputs of first declension nouns are extrem ely faithful to the input forms with the exception of those nouns that lose a syllabic nucle us through glide formation. Historically, the constraint against vowels in hi atus is higher ranking in Castilian and Portuguese although the emergence of glide formation in innovative vari eties of Catalan is er oding the distinction between treatment of vowel/vowel sequences in Eastern and Western Iber o-Romance. Figure 69 shows that the tonic syllable in words of two sylla bles is more likely to be light than heavy. The possibility of apocope dramatically change s the association between heavy syllable and 338

PAGE 339

locus of accent for Catalan. This is readily illustrate d through exam ination of disyllables from the second/fourth declension and third declension. The subsets selected correspond to those nouns in which the initial syllable is either light or contains a long vowel or diphthong. This type of syllable is less likely to retain its moraicity than a syllable with coda consonant. Figure 6-9 shows the output of th e 206 disyllabic nouns with penultimate accent. The predicted output for all (barring apocope) is ( L.L). This output is a desirable trochee and in the case of disyllabic words has the added advantage of satisfying both ALIGN-R and ALIGN-L. In other words, the word is equivalent to the most common foot form. -40 0 40 80 120 160 200 'H 102312382220 'H.L 545031 'L.L 33133123284145 CATCASPORCATCASPOR 2nd/4th decl 3rd decl Figure 6-9. Outcomes of disyllablic nouns with penultimate acent in declensions 2, 3, 4. (n=206) The foot type (L.L) is highly represented in declension 2/4 nouns in Castilian and Portuguese in contrast to Catalan where only 23.6% of nouns have this output fo rm. These results, especially for second/fourth declension nouns, ar e in stark contrast to those seen in Figure 6-9 where there 339

PAGE 340

is little variation am ong languages for any give n word type. In Figur e 6-9 declension 3 nouns show more evenly matched results, although Catala n still shows a much higher rate of apocope. Previous discussion of coincidence of heavy syllable with prominent positions shows that there is little reliance in Ibero-Romance on th e accumulation of segments or moras to call attention to a syllable that should be highlighted either because it is prominent in the structure or prosody of the word. It must be assumed, the n, that preference for th e alternate trochee ( H) responds to some other criteria unrelated to the need for primary accent to coincide with a heavy syllable. Deletion/non-deletion of final un accented vowels in nouns can be addressed morphologically as suggested by use of a constraint like MAXMORPH (An input morpheme has a correspondent in the output, Ll e 2003). However, if MAXMORPH is, indeed, the operative constraint one may well ask why it does not also preserve the characteristic vowel of the second declension which can serve as a morpheme that designates a particular noun class or gender. The traditional view of nominal morphology in Castilian and Portuguese has been that nouns are marked for gender [+fem, +masc] and asso ciated with characterist ic vowels /a/ and /o/; or unmarked for gender as in the clase of nouns ending in -e a consonant, or other uncharacteristic vowels such as -i, -u In Catalan nouns that ar e marked for gender may be viewed as having an overt morpheme that is [+ fem], with characteristic /a/, or a [+masc] morpheme that is most often unrealized, or [ ]. Evidence for this interpretation is provided by imparisyllabics as in the frequent pattern of adjectives of nationality, for example, angls (m.), anglesa (f.) English. Catalan also has, as in Castilian and Po rtuguese, nouns that are unmarked for gender. Because the nominal/adjectival mo rphological system has already licensed a null morpheme to indicate masculine, elision of the characteristic vowel of the second declension is merely an extension of that pattern. However, elision of vowels in Cata lan carries over to the 340

PAGE 341

verb system as well where verbs of the second and third conjuga tions regularly elide the final vowel of the third person singular in the present tense. The vowel is also suppressed before the s ending of the second person singular in the presen t tense. It is noteworthy that word final -e is also elided in 30% to 33% of the third declension bisyllab les in Castilian and Portuguese, although deletion in the verb system is infrequent and occurs more in Portuguese than Castilian with the exception of the monosyllabic imperativ es of high frequency verbs common to all three languages. Given the inadequacy of the morphological approach to motivate retention of final -a and elision of final -o and e one turns to phonological and prosod ic criteria to explain apocope. Languages that undergo vowel reducti on always have a smaller subset of reduced vowels than tonic vowels. Unaccented vowels that are regularly preserved in Catalan are /i/, /u/, /a/. Only /a/ occurs with any degree of frequency word fina lly and as a characteristic vowel of a Latin declension class. It is normally realized as [ ]. There are also a few words (mostly erudite later additions to the lexicon) in which final high vowels are preserved as in tribu tribe and tesi thesis. It is not surprisi ng that precisely these three vow els are the ones selected for nondeletion. They have special acoustic and articulatory properties (Crosswhite 2004, 194-195). Crosswhite proposes a constrai nt intended to enhance the co ntrast among reduced vowels: (6.2) LIC-NONCORNER/STRESS: Noncorner vowels are licensed only in stressed positions. (Crosswhite 2004) Such a constraint would preserve the corner vow els /i/, /u/, /a/. It is then possible to account for the reduction of /a/ to [ ] by relating the vowel sonority hierarchy to positions of prominence and non-prominence. Using a hier archy of vowels that are preferred in nonprominent positions (De Lacy 2006, 2002) it is po ssible to create a ranking limited to those vowels known to occur in Catalan and European Portuguese. In Table 6-4 the numbered cells 341

PAGE 342

show relative ranking and the row below indicate s in which languages the vowels occur. Vowels in ligh t gray cells occur only in non-prominent positions and those in dark gray only in prominent positions. The unshaded cells indicate vowels that occur in both unaccented and unaccented syllables although Portuguese [ ] in accented syllables is limited to occurrence in a specific environment, that is, preceding a nasa l or coronal consonant (Fikkert 2005, 9-10). Table 6-4. Hierarchy of preferred vowels in non-prominent positions > > > > i, u > e, o > a 1 2 3 4 5 6 7 Por Cat Por Cat/Por Cat/Por Cat/Por Cat/Por most preferred least preferred Using the constraint LIC-NONCORNER/STRESS, it can be seen that the vowels in cells 5 and 6 would be readily barred from appearing in non-pr ominent positions. In order to limit deletion at right word edge to noncorner vowels a new constrai nt is developed that builds on the concept of the noncorner vowel: (6.3) *V XVNC]PRWD: A vowel at the edge of a prosodic word that follows the accented vowel must be a corner vowel. X represents any segment(s) between the two peaks. The V XVNC]PRWD constraint promotes apocope when input forms do not contain the preferred final vowels. It is a na tural extension of the ALIGNR constraint that favors alignment of right word edge with the head syllable of the head fo ot. This new constraint disprefers any vowel after the vowel with primary accent. It is also necessary to consider another family of constraints to preclude /a/ in an unstressed position although it is a corner vowel and apparently licensed by the limitation to corner vowels in the V XVNC]PRWD constraint. The constraint *UNSTRESSED/a (Crosswhite 2004, 209-210) is based on the vowel sonority hierarchy (w hich can be read from left to right for 342

PAGE 343

position s of non-prominence and right to left for positions of prominence) effectively bars [a] from occurring in an unaccented position. (6.4) *UNSTRESSED/a: Low vowels are not found in unstressed position. (Crosswhite 2004) However, for purposes of the present study this constraint is modified to exclude all of the nonoccurring unaccented vowels in word final position. (6.5) *UNSTRESSED/>4: Vowels whose sonority is greater than 4 are not found in unstressed position. Finally, MAX constraints (cf. Crosswhite 2004, 211) are needed in a second step to ensure that all vowels allowed in word final position do not reduce to schwa sinc e it is the preferred syllabic peak in this context. This can be accomplished by adding constraints that reflect the salient features of the corner vowels such as tongue height retraction of tongue, and lip rounding. It is assumed that the feature specifications of the seven vowel system in Catalan (and Portuguese, although the neutral vowel is different) are those reflect ed in Table 6-5. The neutral vowel has been added to demonstrat e that it is featureless. Table 6-5. Distinctive featur es in a seven vowel system i e a o u high + + low + back + + + front + + + tense + + + + round + + + Table 6-6 illustrates retention of corner vowel s in Catalan. The constraints given in this tableau are not applicable to Castilian and Portuguese because *UNSTRESSED/>4 and *V XVNC]PRWD preclude the emergence of final unstressed mid vowels which are, in fact, realized in these languages. The higher ranking constraint, SONSEQ, prevents vowel deletion when an 343

PAGE 344

unacceptable final clu ster would emerge, that is a sequence CC in which the first consonant is less sonorous than the second. Table 6-6 includes MAX/IO to demonstrate that deletion is preferable to maintaining the vowel of the input in the case of the first four vowels. Maintaining faithfulness to the original input would result in a violation of the higher ranking constraint *V XVNC]PRWD. Table 6-6. OT tableau for fi nal vowel deletion in Catalan Final vowel in input SONSEQ *V XVNC]PRWDMAX/IO *UNSTRESSED/>4 1. -o 2. 3. * 4. -e 5. -a 6. -u 7. -i Additional constraints are n eeded to reduce /a/ to [ ] and to prevent reduction of /u/ and /i/ to [ ] as well. The proposed constraints are given below. Because rounding is characteristic of back vowels it is not necessary to add [back] as a MAX constraint. The presence of [front] will disallow reduction of /e/ or / / to [ ] because schwa does not share the specification for [front]. The inclusion of [high] also assu res that the corner vowels /i/ a nd /u/ continue to be preferred candidates over less sonorous [ ]. (6.6) MAX/round: A vowel in the input that has a specification for th e feature round has a correspondent in the output. (6.7) MAX /high: A vowel in the input that has a specification for the feature high has a correspondent in the output. (6.8) MAX /low: A vowel in the input that has a specification for th e feature low has a correspondent in the output. 344

PAGE 345

(6.9) Max/front: A vowel in th e input that has a specification for the feature front has a correspondent in the output. Table 6-7 demonstrates how these faithfulness constraints related to vowel features work to maintain the high vowels without change but reduce [a] to [ ] b ecause the input vowel violates *UNSTRESSED/>4. For the three acceptable candidate s that emerge from Table 6-6, two outcomes are posited in Table 6-7: no change and reduction to [ ], the preferred vowel in unstressed positions. Table 6-7. OT tableau for vowel reduction in seven vowel systems Final vowel Input Final vowel Output *UNSTRESSED/>4 MAX [round] MAX [high] MAX [low] MAX [front] -u -u -u *! -i -i -i *! -a -a *! -a Having established the various constraints that contribute to loss of final vowels in Catalan it is possible now to examine the effect that apoc ope has on the prosodic stru ctures established as models in discussion of first declension nouns. In the case of nouns of two syllables from declension 1 (Figure 6-8), 64.2% of nouns have a light initial/tonic syllable in Catalan. Apocope changes this ratio dramatically for nouns in the second/fourth and third declensions where percentages of heavy tonic syllables are now 72.9% and 57.6%. The new foot form ('H) is firmly established. There are now three highly attested trochees: the preferred disyllabic trochee ('L.L), the uneven trochee ('H.L), and the monosylla bic trochee ('H). Dist ribution of outcomes of second/fourth and third declension nouns is disp layed by language and type in Figure 6-8 above. 345

PAGE 346

The real innovation of Ibero-Rom ance begi ns with the outcomes of nouns of three syllables in which V XVNC]PRWD is an operative constraint in Catalan. Figure 6-10 shows the prosodic outcomes for trisyllables from the s econd and fourth declensions. The coincidence between heavy syllables and locus of primary accent in Catalan has now risen to 97.3%. In order to gauge the true effect of apocope in promo ting identity between heavy syllable and locus of accent it is necessary to cons ider the conditions that de termine placement of accent in penultimate position in a three-syllable word in Latin. The Latin Stress Rule requires that the tonic syllable be bimoraic. Therefore, there is an inherent bias toward the tonic syllable being also a heavy syllable. If the syllable is bimora ic on the bias of a long vowel it will not remain heavy in Ibero-Romance. In the case of the i nput data for Figure 6-10, 99 of 257 words have an HV type tonic syllable. Therefore, the likely can didates for preservation of the second mora are 158 of 257, or 61.4%. It can be seen that apocope in Catalan has increased the coincidence of primary accent with heavy syllable by over 50%. It should be noted that the num ber of light tonic syllables ha s also dropped dramatically in Castilian (56.0%) and Portuguese (50.6%) when compared with di syllabic nouns. Despite this bifurcation of outcomes on the basi s of apocope, Figure 6-10 does show remarkable similarity in the distribution of heavy/light syllables in othe r positions. Data point 1 represents the second most common pattern in Catalan (27.%). It is th e logical equivalent of the data represented at data points 3 and 5 where syncope is not operativ e. Similarly, the most common outcome for Catalan (data point 2) has its co rrespondent at data points 4 and 6. While it is tempting to view this pattern (L.'H) as an iamb, to do so would ignore the obvious relations hip that exists between imparisyllabics in all three languages, best exemp lified by adjectives of na tionality in the modern languages as well as agentives with the suffixes tor/a nd dor / a. 346

PAGE 347

Figure 6-10. Prosodic templates from declension 2/4 trisyllables with penultimate accent. (n=257) When V XVNC]PRWD operates across languages, as o ccurs in the third declension, only language specific NOCODA constraints prevent deleti on of the characteristic e vowel of the accusative singular of masculine/feminine nouns. In this data set 173 of 221 nouns depend on an HV type syllable for assignment of penultimate accent. Unless the post-tonic syllable has a complex onset apocope would rarely be blocked except for those cases in which Castilian and Portuguese NOCODA constraints are operative. The uniformity of results is testimony to the role of apocope in producing coincidence of heavy sylla ble and tonic syllable. In Figure 6-11 the two most salient prosodic templates (d ata points 1 and 2) have ultima te accent on a heavy syllable. The alternate trochee ('H) is aligned with right word edge and satisfies V XVNC]PRWD. Language specific NOCODA constraints operate in gr adient fashion to produce the highest degree of apocope in Catalan followed by Castilian and then Portuguese. Cases in which apocope is barred are predictable: those with a he avy syllable (HC type in the input ), represented at data points 3 and 4. Data point 5 reveals th e restrictive nature of the NOCODA constraint in Portuguese which permits final vowel deletion only if the re sulting coda is a liquid or /s/. 250 200 150 100 50 0 -50 H.'H L.'H L.' L.L 'L.L.L CAT 1 70 130 1 1 21 28 0 5 0 CAS 0 1 5 1 3 47 59 31 108 2 POR 0 2 11 2 2 49 63 27 98 3 'H 'H.L 'L.L H.'H.L L .'H.L H.'L.L ' 347

PAGE 348

Figure 6-11. Prosodic templates from declension 3 trisyllables with penultimate accent. (n=221) Although the end result is that locus of accen t coincides to a high degree with a heavy syllable (86.9% Catalan, 81% Ca stilian, 76.5% Portuguese) it is not possible to motivate apocope on this basis. That is, the determinant of apocope is not the creation of a heavy tonic syllable but rather an edge marking effect that aligns the he ad foot with the end of the prosodic word. An examination of patterns of antepenultimate accent de monstrates that the coincidence of primary accent and heavy syllables is not an endur ing quality. When the constraint V XVNC]PRWD is applied to trisyallables the result is often (' L.H). Discussions of the uneven trochee rarely include this alternative although (' H.L) is widely cited as an ex emplar. For the purposes of the current study the syllabic trochee is assumed to admit all of the following: ('L.L), ('H), ('H.L), ('L.H), ('H.H). Patterns with Antepenultimate Accent and /-o/ Class Marker or No Class Marker Figure 6-12 shows the prosodic templates th at result from declension 2/4 nouns with antepenultimate accent. The constraint V XVNC]PRWD has effectively eliminated the final syllable in 215 of 242 cases (88.8%) in Catalan. A sm all set of words (dat a points 5, 6) shows 220 -20 20 60 100 140 180 CAT 0 58 119 0 0 1 4 11 6 11 7 4 0 0 CAS 2 46 83 2 1 1 13 34 8 14 6 7 3 1 POR 3 45 79 3 0 0 12 33 11 31 3 0 0 1 'H H.'H L.'H 'L.H 'L.L L.L.'L H.'H.L L.'H.L H.'L.L L.'L.L H.'L.H L.'L.H L.'H.H 'L.L.L ' 348

PAGE 349

displacem ent of accent to the penul timate syllable. These are primarily words of Greek origin such as timp (
PAGE 350

that 50.6 % of the outputs of declension 2/4 trisylla bles with penultimate accent result in (L.'H) in Catalan it is somewhat surprising that forces of analogy do not cal l for accent shift. However, there is no evidence either diachronic or synchron ic that favors an argument that heavy syllables attract word level accent. The data set of third declension trisyllabl es with antpenultimate accent is considerably smaller than than for declension 2/4 nouns. A larger percentage of initial syllables are HC type (43.8%) which leads to an expectation of a higher representation of templa tes in which the tonic syllable is heavy. This is confirmed by the arra y of templates seen in Figure 6-13 (especially data points 1, 3, and 5). Differe nces by language in retention of the final, unaccented vowel are minimal but show a slightly higher rate of elision in Catalan (data points 1, 2). 0 12 24 36 48 60 72 CAT 3915167209111 CAS 181212101113141 POR 15131073116152 'H'H.H'L.H'H.L'L.LL.'HH.'H'H.L.L'L.L.XL.'L.L ' Figure 6-13. Prosodic templates from declension 3 trisyllables with antepenultimate accent. (n=73) Overall, coincidence of outputs in the set of third declension trisyllables is remarkable. Majority outputs in all three languages are disyllabic. Reduction to two syllables has two sources: syncope and apocope. Examples of syncope th at result in ('L.L) or ('H.L) are Cat. llebre Cast. 350

PAGE 351

liebr e Port. lebre (
PAGE 352

Disyllab les with ultimate accent, L.'H and H.'H, are products of the third declension with apocope cross linguistically. However, the second/fourth declension type nouns also provide a source for this template in Catalan. Both Fi gure 6-14 and following Tabl es 6-8 show that the second most frequent pattern for two-syllable word s in Catalan, after the ideal trochaic pattern ('L.L), is L('H) which has its origins in trisyl lables with penultimate accent in the second/fourth and third declensions. Despite the high representation of this pattern it does not appear to influence ('L.H), which has the lowest representation after ('H.H) in Catalan. Input data which give rise to the different templa tes displayed in Figure 6-14 appear in Tables 6-8. Table 6-8 is presented in two parts to improve legibility; summa ry data at the end of the continuation of Table 6-8 applies to both parts. Columns represent th e different prosodic temp lates that are found in nouns of one and two syllables in Catalan, Castil ian, and Portuguese. Rows indicate the inputs that give rise to the various templates and ar e ordered by number of syllables in the input, declension class, and locus of accent. Table 6-8. Distribution and source of one and two-syllable nouns. Source 'H 'L.L 'H.L 'H.H CAT CAS POR CAT CAS POR CAT CAS POR CAT CAS POR decl 1 0 0 2 144 149 144 88 83 86 0 0 0 decl 2/4 102 3 12 33 133 123 5 4 5 0 0 0 decl 3 38 22 20 28 41 45 0 3 1 0 0 0 decl 1 0 0 0 4 6 7 2 2 6 0 0 0 decl 2/4 1 0 0 6 3 1 1 1 2 0 0 0 decl 3 0 2 3 7 2 0 0 0 0 0 0 1 decl 1 0 0 0 24 48 35 26 29 35 0 0 0 decl 2/4 42 1 1 48 77 64 20 23 28 38 2 2 decl 3 3 1 1 7 10 7 0 0 0 9 8 5 decl 1 0 0 0 0 0 0 0 0 0 0 0 0 decl 2/4 0 0 0 0 0 0 0 0 0 0 0 0 decl 3 0 0 0 0 0 0 0 0 0 0 0 0 decl 2/4 0 0 0 0 0 0 0 0 0 0 0 0 decl 3 0 0 0 0 0 0 0 0 0 0 0 0 Total 186 29 39 301 469 426 142 145 163 47 10 8 352

PAGE 353

Table 6-8. Continued 'L.H L.'H H.'H CAT CAS POR CAT CAS POR CAT CAS POR decl 1 0 0 0 0 0 0 0 0 0 decl 2/4 69 0 0 0 0 0 0 0 0 decl 3 15 12 13 0 0 0 0 0 0 decl 1 0 0 0 0 0 0 0 0 0 decl 2/4 0 0 0 130 5 11 70 1 2 decl 3 0 0 0 119 83 79 58 46 45 decl 1 0 0 0 0 0 0 0 1 0 decl 2/4 0 0 0 5 0 0 3 0 0 decl 3 0 0 0 2 1 3 16 12 11 decl 1 0 0 0 0 0 0 0 0 0 decl 2/4 0 0 0 2 0 0 3 0 0 decl 3 0 0 0 2 8 6 5 29 29 decl 2/4 0 0 0 15 0 0 0 0 0 decl 3 0 0 0 3 2 2 0 0 0 Total 84 12 13 278 99 101 155 88 87 Monosyllables n=186 Catalan n=29 Castilian n=39 Portuguese Disyllables n=1007 Catalan n=994 Castilian n=837 Portuguese There are four templates with penultimate accent corresponding to nouns of two syllables. The first, with highest representation in all langu ages, is ('L.L), the pref erred pattern for syllabic trochees. The second is an uneven trochee, ('H.L), also present in Latin and in some ways preferable under the Latin system of prosody because removal of the final extrametrical syllable still leaves a bimoraic foot. The pattern with heavy initial (tonic) syllable was also the preferred pattern in Apoussidou and Boersmas studies (2003, 2004) on the learnability of the Latin Stress Rule. The two innovative patterns with heavy (unaccented) final syllable are outcomes of apocope of third declension nouns in all three la nguages, and of second/fourth declension nouns in Catalan only. Although attested in Castilia n and Portuguese the ('L.H) and ('H.H) are very poorly represented. There are 12 occurrences of ('L.H) in Castilian and 13 in Portuguese compared with 84 in Catalan. The template with two heavy syllables, ('H.H), occurs 10 times in Castilian and 8 in Portuguese comp ared with 47 in Catalan. There is no evidence of attraction of 353

PAGE 354

word accent to the heavy final syllable in thes e cases although a series of recent studies in Spanish have shown some correlation between assignment of accent to the final syllable in nonce words when that syllable is heavy. However, this preference does not extend to penultimate or antepenultimate syllables (E ddington 2004, 120-124). The importance of these undesirable trochees resides in the fact that their emergence as optimal candidates is indi cative of the high ranking constraints relative to foot form (trochee) a nd faithfulness to the original locus of accent. Only Catalan has high representation of th e three templates with ultimate accent('H), L('H), and H('H). This suggests that these pa tterns are dispreferred in Castilian and Portuguese and emerge only when constraints on the natu re of unaccented final vowels and the general prohibition on any vowel following th e syllable with the primary accent work in tandem to give the candidate with apocope the highest ranking. In terms of distribution of prosodic patterns in two-syllable words, L('H) accounts for 23.3% of disyllables in Catalan. Together with H('H) patterns with ultimate accent represent 30.7% of two-syllable words in Catalan. It should be remembered that 'H is an acceptable alternative to 'L.L as realization of a syllabic trochee. The advantage offered by 'H is satisfaction of ALIGN-R when that constraint specifies that both head foot and head syllable should be at the right edge of the word. Figure 6-15 represents all three-syllable words with penultimate accent and their sources. The most prevalent pattern for all three languages has all light syllables: L( 'L.L). The head foot is aligned right and corresponds to the preferred pa ttern for syllabic trochees The initial syllable remains unparsed. However, neither Latin nor the Ibero-Romance languag es require exhaustive parsing. With three syllables the possible dist ribution patterns for heavy and light syllables increases to eight. They are listed here with the last two syllable s parsed as a syllabic trochee: H('H.H), H('H.L), H('L.H), H('L.L), L('H .H), L('H.L), L.('L.H), L.('L.L). 354

PAGE 355

0 50 100 150 200 250 300 350 CAT CAS POR CAT CAS POR CAT CAS POR CAT CAS POR CAT CAS POR CAT CAS POR CAT CAS POR L'LLH'LLH'LHL'LHH'HLL'HLL'HH decl 1 decl 2/4 decl 3 decl 1 decl 2/4 decl 3 decl 1 decl 2/4 decl 3 Figure 6-15. Preferred templates for nouns of three syllables with penultimate accent. Tables 6-9A and 6-9B indicate the source of the prosodic templates in Figure 6-15. Of the eight possible patterns only H('H.H) is not repr esented in the dataset. It has already been noted that the pattern with all light syllables ranks highest for the three languages. The second most frequent pattern is more difficult to establish because H('L.L) is slightly higher for Catalan whereas L('H.L) appears to be more favored in Castilian and Portuguese Data for Figure 6-15 appears in Tables 6-9 with numeric summary for both tables at the end of the continuation of Table 6-9. The unequal representation of langua ges in the three-syll able pattern with penultimate accent is apparent. The total for Catalan is only 466 while the total for Castilian is 745 and for Portuguese 738. The difference is owing to two factors: reduc tion of four syllables to three in Castilian and Portugue se as a result of glide formation and conservation of unaccented final vowels in words of three syllables. Cata lan, on the other hand, ha s lost approximately 131 nouns from the three-syllable penultimate accen t pattern because they now appear as twosyllable nouns with ultimate accent. 355

PAGE 356

Table 6-9. Distribution a nd source of three-syllable nouns with penultim ate accent L'LL H'LL H'LH L'LH CAT CAS POR CAT CAS POR CAT CAS POR CAT CAS POR decl 1 81 85 90 57 58 55 0 0 0 0 0 0 decl 2/4 5 108 98 0 31 27 0 0 0 0 0 0 decl 3 11 14 31 6 8 11 7 6 3 4 7 0 decl 1 2 5 2 5 4 1 0 0 0 0 0 0 decl 2/4 0 4 5 0 5 3 0 0 0 0 0 0 decl 3 0 0 2 0 0 5 0 0 0 0 0 0 decl 1 20 54 47 12 27 28 0 0 0 0 0 0 decl 2/4 61 70 64 11 8 2 41 1 0 4 1 0 decl 3 5 5 4 3 3 1 5 4 8 0 1 6 Totals 185 345 343 94 144 133 53 11 11 8 9 6 Table 6-9. Continued H'HL L'HL L'HH CAT CAS POR CAT CAS POR CAT CAS POR decl 1 9 5 5 29 26 23 0 0 0 decl 2/4 21 47 49 28 59 63 0 0 0 decl 3 14 13 12 11 34 33 0 3 0 decl 1 0 0 0 0 0 0 0 0 0 decl 2/4 0 0 0 5 4 3 0 0 0 decl 3 0 0 0 0 4 3 0 0 0 decl 1 0 14 13 6 23 29 0 0 0 decl 2/4 0 0 0 3 3 6 0 0 0 decl 3 0 0 0 0 1 6 0 0 0 Totals 44 79 79 82 154 166 0 3 0 n=466 Catalan n=745 Castilian n=738 Portuguese With regard to the L('H.L) pattern it is rhythmically an amphibrach, a prosodic pattern found in poetry in all three languages (Navarro Toms 1956, 515; Serra i Bald 1932, 56; Ross 1992, 185-186). It is also a rhythmic pattern that arises in natural sp eech when taking into consideration monosyllabic function words followed by a disyllabic trochee. Note that this same pattern of function word + noun give s the appearance of iambic rhyt hm in Catalan in the case of monosyllabic nouns. The rhythmic pattern L'H is the basis of the dialog between man and god, Tirallonga de monosllabs (Monosyllabic litany) by Joan Oliv er, better known by the pseudonym Pere Quart. In a study on the use of monosyllables as a rhetorical device in Catalan (Garolera 1985, 179-180), the author, Pere Quart, refers to the wo rk as a curis exercici. 356

PAGE 357

While perhaps not representative of natural spe ech, the fact that a relatively long text can be written entirely with m onosyllabic prosodic wo rds displays a fundamental difference between Catalan and its sister languages and further reinforces the unequal representation of the ('H) type trochee among the three languages. The least favored uneven trochees, ('H.H) and ('L.H), have very low representation in the three languages al though there are more instances of ('L.H) in Catalan, again the product of a pocope, in this case from decl ension 2/4 tetrasyllables with antepenultimate accent. The marked nature of antepenultimate accent is readily seen by the reduced size of the inventory displayed in Figure 6-16. The preferred template is r eadily apparent: ('L.L.L). 0 40 80 120 160 200 CATCASPORCATCASPORCATCASPOR 'LLL 'HLL 'LHH decl 1 decl 2/4 decl 3 decl 2/4 decl 3 decl 2/4 decl 3 Figure 6-16. Preferred templates for nouns of th ree syllables with ante penultimate accent. Previous discussion has provided a basis for c ounting templates of this type as binary by considering that the penultimate syllable does not project a mora. The data that corresponds to Figure 616 appears in Table 6-10. The one anomalous form, ('LHH) corresponds to Catalan crisantem chrysanthemum pronounced [ k iz nt m ] and represents an ancient shift from the proparoxytonic pronunciation of the original Greek. The other for antepenultimate trisyllables is, 357

PAGE 358

predic tably, ('HLL). In order for stress to be retr acted to the initial syll able both the ultimate and penultimate must be light. Again Catalan has an appreciably lowe r representation in this subset because the largest input category for Castilia n and Portuguese consists of second/fourth declension tetrasyllables with antepenultimate accent, a class of nouns that undergoes apocope whenever permitted in Catalan. Table 6-10. Distribution and source of three-syllable nouns with antepenultimate accent 'LLL 'HLL 'LHH CAT CAS POR CAT CAS POR CAT CAS POR decl 1 72 57 66 35 23 21 0 0 0 decl 2/4 0 2 3 0 0 0 0 0 0 decl 3 0 1 1 0 0 0 0 0 0 decl 2/4 0 3 4 0 0 0 0 0 0 decl 3 1 3 2 0 0 0 0 0 0 decl 2/4 10 84 93 7 51 52 1 0 0 decl 3 11 14 15 9 13 16 0 0 0 Totals 94 164 184 51 87 89 1 0 0 n=146 Catalan n=251 Castilian n=273 Portuguese The one innovative class of third declension nouns is the subset with ultimate accent. This template is well represented only in Catalan. The tonic syllable is in all cases heavy and is the product of apocope in nouns from declensions 2/4 and 3 with the single exception of Portuguese avel hazelnut (
PAGE 359

0 20 40 60 80 100 120 140 160 180 CAT CAS POR CAT CAS POR CAT CAS POR CAT CAS POR LL'HHL'HH'HHLH'H decl 1 decl 2/4 decl 3 decl 2/4 decl 3 Figure 6-17. Preferred templates for nouns of three syllables with ultimate accent. Table 6-11 indicates in detail the origin of each template. Nearly all words in this subset are the product of apocope of nouns from the third declension and for Catalan, the second declension as well. Participation in the new template is not equal among languages. Although the number for Castilian is nearly 40% above that for Portuguese it is only 39% of the total for Catalan. Historically, apocope is most favored by the active constraints in Catalan, although it should be noted that modern European Portuguese exhibits extreme vowel reduction in both pretonic and postonic positions. Mateus and dAndrade (2000, 134) note that This is most remarkable in post-stressed and final position. Because this level of vowel reduction is a phenomenon of modern European Portuguese it ha s not been treated he re. This study is concerned primarily with the emergence of pros odic templates during the formative period of the Ibero-Romance languages and the subsequent borro wing from Latin in the erudite vocabulary of the early modern period. However, this repetition of historical processe s is worthy of further study. 359

PAGE 360

Table 6-11. Distribu tion and source of th ree-syllable nouns with ultimate accent LL'H HL'H LL'H LH'H CAT CAS POR CAT CAS POR CAT CAS POR CAT CAS POR decl 1 0 0 1 0 0 0 0 0 0 0 0 0 decl 2/4 78 1 9 31 1 1 2 0 0 8 0 0 decl 3 98 64 44 62 38 22 16 15 8 6 0 0 decl 2/4 0 0 0 0 0 0 2 0 0 0 0 0 decl 3 1 0 0 0 0 0 2 1 1 0 0 0 Totals 177 65 54 93 39 23 22 16 9 14 0 0 n=306 Catalan n=120 Castilian n=86 Portuguese This summary chapter has concentrated on nouns of one to three syll ables in Castilian, Catalan, and Portuguese because they are more re presentative of patrimonial vocabulary and are less likely to result from derivational processe s which may have influence on accentual patterns. However, four-syllable nouns are represente d here because there is less morphological interference compared with five-syllable nouns that are largely deri vational and show the productive nature of a few suffixes. Nevertheless, this subset is relatively small and reflects a more erudite vocabulary. The preferred temp late consists of all light syllables. 0 20 40 60 80 100 120 140 CAT CAS POR CAT CAS POR CAT CAS POR CAT CAS POR CAT CAS POR CAT CAS POR CAT CAS POR CAT CAS POR LL'LLHL'LLLH'LLHH'LLLL'HLHH'HLHL'HLLH'HL decl 1 decl 2/4 decl 3 decl 1 decl 2/4 Figure 6-18. Preferred template for nouns of four syllables with penultimate accent. The template (L.L)( 'L.L) allows construction of a disyllab ic trochee at the right word edge and an optional disyllabic trochee at left word edge. It should be noted that cross-linguistically 360

PAGE 361

input to this preferred template comes prim arily from the first d eclension. This can be readily seen by comparing the columns in Figure 6-18 as well as the first row of Table 6-12. Table 6-12 provides further detail on the source data fo r the tetrasyllables with penultimate accent. Numerical count is quite evenly distributed wh en input comes from the first declension. Otherwise, Catalan is underrepresented in comparison with Castilian and Portuguese. Examining the totals of each template, Catala n shows the lowest output for the disyllabic trochee. Output is higher for Castilian with mo re restrictions on admissible codas and in most cases Portuguese shows the highest preference fo r the disyllabic trochee at right word edge because it has the greatest restriction on coda cons onants. The summary totals show clearly the language-specific preferences : Portuguese 291 >> Castilian 254 >> Catalan 129. Table 6-12. Distribution and source of four-syllable nouns with penultimate accent LL'LL HL'LL LH'LL HH'LL CAT CAS POR CAT CAS PO R CAT CAS POR CAT CAS POR decl 1 45 47 47 20 21 21 13 12 12 3 2 2 decl 2/4 2 50 45 3 19 20 0 3 1 0 1 2 decl 3 2 2 28 1 1 14 1 1 0 0 0 0 decl 1 0 2 1 0 2 0 0 0 0 0 0 0 decl 2/4 0 1 2 0 1 0 0 0 0 0 0 0 Totals 49 102 123 24 44 55 14 16 13 3 3 4 Table 6-12. Continued LL'HL HH'HL HL'HL LH'HL CAT CAS POR CAT CAS PO R CAT CAS POR CAT CAS POR decl 1 12 13 19 1 1 1 0 0 1 0 0 0 decl 2/4 15 49 51 3 2 1 4 17 16 2 3 3 decl 3 1 2 2 0 0 0 1 1 1 0 1 1 decl 1 0 0 0 0 0 0 0 0 0 0 0 0 decl 2/4 0 0 0 0 0 0 0 0 0 0 0 0 Totals 28 64 72 4 3 2 5 18 18 2 4 4 n=129 Catalan n=254 Castilian n=291 Portuguese The final data set is very small and consists of four-syllable nouns with antepenultimate accent. The preferred pattern again c onsists of all light syllables. It is expected that the ultimate and penultimate syllables will be light when the accent is antepenultimate but the preceding 361

PAGE 362

syllab les in this case the antepenultimate and preantepenultimate/initial syllable could be either light or heavy. Yet Figure 6-19 shows that the pattern of all light syllables is the only one that has any level of represen tation among the three languages. What is interesting is that although the numerical differences are less striking (Catalan 77, Castilian 88, Portuguese 95) than in other data sets the input sources are quite different for the L('L.L.L) pattern. 0 10 20 30 40 50 60 70 80 90 100 CAT CAS POR CAT CAS POR CAT CAS POR CAT CAS POR H'HLLH'LLLL'HLLL'LLL decl 1 decl 2/4 decl 3 Figure 6-19. Preferred template for nouns of four syllables with antepenultimate accent. For Catalan the majority input for this antepe nultimate stress pattern comes from the first declension whereas for Castilian and Portuguese s econd declension nouns are major contributors. Furthermore, in examining the totals in Table 6-13 it can be seen that in total Catalan has a higher representation of tetrasyllables with proparoxytonic accent. This is due to its higher representation of the template H( 'L.L.L). In general, Catalan shows greater tolerance for coda consonants. Also, given the erud ite nature of many of the exem plars in this subset there are fewer pressures to produce optimal syllables by observing NOCODA constraints. In Castilian and Portuguese, NOCODA frequently takes precedence over faithfulness or MAX type constraints. Table 6-13 displays the distribution and origin of the template s represented in Figure 6-19. 362

PAGE 363

Table 6-13. Distribution and orig in of four-syllable words with antepenultim ate accent. H'HLL H'LLL L'HLL L'LLL CAT CAS POR CAT CAS POR CAT CAS POR CAT CAS POR decl 1 12 1 0 30 8 9 29 8 10 56 25 28 decl 2/4 0 0 0 0 3 2 0 0 2 4 49 55 decl 3 0 1 1 5 6 9 1 2 1 13 14 12 Totals 12 2 1 35 17 20 30 10 13 73 88 95 n=150 Catalan n=117 Catalan n=129 Portuguese This study has focused on the transition from an accentual system based on the moraic trochee to one which values rhythmic contrast between an accented head syllable and a following unaccented syllable in the case of the disyllabic tr ochee; or, alternatively has a final culminating heavy syllable aligned with the right word edge. Both developm ents bring the accentual system of Ibero-Romance in congruence with metrical systems that are cross-linguistically validated, that is, the preference for duple rhythm and the de marcative effect of building the head foot at word edge. The change from moraic trochee to syllabic trochee is motivated by the loss of contrastive differences in vowel le ngth. Although the result is rule opacity with regard to words consisting of three light syllbles, for example, the Ibero-Romance languages maintain with very few exceptions the original locus of accent. The three-syllable window remains in effect although accent is no longer assigned on basis of the moraic count of the penultimate syllable. There is to some degr ee a residual association between locus of primary accent and a heavy sylla ble but there are no constraints that require that a heavy syllable be prominent on the grid, or that it be assigned primary accent. Accents may still fall on the ultimate, penultimate, or antepe nultimate syllable. Because of the historical origins of antepenultimate accent the ultimate and penultimate syllables are almost always light. However, recent loanwords such as bdminton in Spanish retain the or iginal place of accent even though the ultimate and penultimate syllables are both heavy and there are analogic pressures to move the accent to the final syll able. Overall words with antepe nultimate do not show increased 363

PAGE 364

incidence of syncope. There are two constraints that act to reduce syllabic count, apocope in Catalan and glide form ation in Castilia n and to a lesser extent in Portuguese. The development of ultimate accent in the formative period of the Ibero-Romance languages perhaps made it easier to incorporat e words with oxytonic accent from the languages of the Americas such as these examples from Spanish: caimn caiman, tiburn shark, urub buzzard. Ultimate accent is the dividing line be tween East and West in Ibero-Romance. Only Catalan has as preferred output from second and third declension nouns a template with a word final heavy syllable, obviated only by constraint s on the acceptability of word final consonant clusters. When sonority sequencing is violated the final unaccented vowel remains. Although historically Portuguese shows a ve ry limited repertory of acceptabl e codas in word final position today it has surpassed Catalan in terms of vowel reduction. Vowels are not merely reduced in terms of the inventory; they are mostly elided in post-tonic position. As a result, modern European Portuguese today has marked consonant sequences in word initial, medial, and final positions. The comparison of aspects of vowel re duction in Catalan and Po rtuguese is topic of further investigation and beyond the scope of the present work. Over time relative positions of the three languages with regard to coda restrictions and sonor ity sequencing have shifted: Portuguese is now the most innovative; Catala n occupies a middle position; and Castilian continues to be the most conservative. The second major factor that differentiates th e three languages prosodically is treatment of vowels in hiatus. Again, recent investigations (Cabr and Prieto 2004, Mateus and dAndrade 1998) demonstrate that both Catala n and Portuguese show increased incidence of glide formation and will no longer contrast with Ca stilian with regard to the maintenance of vowels in hiatus. Glide formation results in elimination of a syllabl e reducing the syllable c ount of a word. It has 364

PAGE 365

been shown here that the datasets of polysyllables are significantly sm aller than those of two and three-syllable words. Uniform treatment of pol ysyllables is seen only in nouns of the first declension where there is little change in terms of number of syllables (except through glide formation) and locus of accent. In prosodic terms it can be seen that the th ree languages share a minimum bimoraic word requirement. The predominant foot type is the binary trochee which is overwhelmingly disyllabic in Castilian and Portuguese. However, the datasets reviewed here indicate that in the case of Catalan nearly a third of all two and three-syllable words have ultimate accent. It has been suggested that ultimate accent can be accounted for in Spanish by supposition of an unrealized final vowel (Harris 1995). If that argument were extended to the case of Catalan it could be readily seen that the ('H) type trochee is simply an alternate form of ('L.L), that is, the coda consonant is in anticipation of the unrealized nucleus. Most of the cases of final stressed vowels can be construed as vowel + coda where the coda has been deleted to conform to a specific NOCODA constraint involving rhot ics and coronal nasals. The primary templates seen here place in juxtaposition two preferences with regard to word level accent, a preference for a word final contrast of accented + unaccented versus a prominence effect at word edge. Vowel deletion has been shown to occur in accordance with recognition of different degrees of sonority that reflect more or less optimal syllable peak s. Elision of /a/ is almost unknown in this data set of mo re than 3,000 nouns. Vowel deletion also favors alignment of accent with word edges. Insertion of vowels, or anaptyxis, is not a prevalent feature of Catalan or the peninsular varieties of Spanish and Portugue se. On the other hand, prothesis in the case of word-initial s + consonant sequenc es, continues to the present day and is testimony to an active constraint against a particular t ype of complex cluster in word initial position. The epenthetic 365

PAGE 366

vowel, however, is never stress-bearing. This is a reflection of high ra nking faithfulness to the locus of accent in the in put form. The few cas es of accent shift discussed here reflect the influence of factors external to the input form itself and may be ascribed to the forces of analogy or faithfulness to output-output constraints. This study relies on an extensive inventory of nouns of Latin origin to provide evidence for the assertions made in this summary. Dive rgences among languages are explained in terms of OT constraints that readily allow for crosslinguistic comparison and reliance on linguistic universals with regard to preferred foot forms and tangential well-formedness constraints. It is not possible to provide a case history for every etymon and its outputs but construction and analysis of the dataset has made it possible to observe general tendencies and provide the groundwork for more extensive study. 366

PAGE 367

367 APPENDIX A APPENDIX PROBI Vnnen, Veikko. 1967. Introduction au latin vulgaire Paris: C. Klincksieck. (pp. 254-257) 1 Porphireticum marmor non purpureticum marmur. 2 tolonium non toloneum. 3 speculum non speclum. 4 masculus non masclus. 5 vetulus non veclus. 6 vitulus non viclus. 7 vernaculus non vernaclus. 8 articulus non articlus. 9 baculus non vaclus. 10 angulus non anglus. 11 iugulus non iuglus. 12 calcostegis non calcosteis. 13 septizonium non septidonium. 14 vacua non vaqua. 15 vacui non vaqui. 16 cultellum non cuntellum. 17 marsias non marsuas. 18 canilam nun cani{a}nus. 19 hercules non herculens. 20 columna non colomna. 21 pecten non pectinis. 22 aquaeductus non aquiductus 23 cithara non citera 24 crista non crysta 25 formica non furmica 26 musivum non museum. 27 exequae non execiae. 28 gyrus non girus. 29 avus non aus. 30 miles non milex. 31 sobrius non suber. 32 figulus non figel. 33 masculus non mascel. 34 lanius non laneo. 35 iuvencus non iuvenclus. 36 barbarus non barbar. 37 equs non ecus. 38 coqus non cocus. 39 coquens non cocens. 40 coqui non coci 41 acre non acrum. 42 pauper mulier non paupera mulier 43 carcer non car. 44 bravium non braveum. 45 pancarpus non parcarpus. 46 teofilus non izofilus. 47 homfagium non monofagium. 48 byzacenus non bizacinus. 49 capsesis non capressis 50 catulus ellus. 51 {catulus llus}. 52 doleus non dolium. 53 calida non calda. 54 frigida non fricda. 55 vinea non vinia. 56 tristis non tristus. 57 tersus non tertus. 58 umbilicus non imbilicus. 59 turma non torma. 60 celebs non celeps. 61 ostium non osteum 62 flavus non flaus. 63 cavea non cavia. 64 senatus non sinatus. 65 brattea non brattia. 66 cochlea non coclia. 67 cocleare non cocliarium. 68 palearium non paliarium. 69 primipilaris non primipilarius. 70 alveus non albeus. 71 glomus non glovus. 72 lancea non lancia. 73 favilla non failla. 74 orbis non orbs. 75 formosus non formunsus. 76 ansa non asa. 77 flagellum non fragellum. 78 calatus non galatus. 79 digitus non dicitus. 80 solea non solia. 81 calceus non calcius.

PAGE 368

82 iecur non iocur. 83 auris non oricla. 84 camera non cammara. 85 pegma non peuma. 86 cloaca non cluaca. 87 festuca non fistuca. 88 ales non alis. 89 facies non facis. 90 cautes non cautis. 91 plebes non plevis. 92 vates non vatis. 93 tabes non tavis. 94 suppellex non superlex. 95 apes non apis. 96 nubes non nubs. 97 suboles non subolis. 98 vulpes non vulpis. 99 palumbes non palumbus. 100 lues non luis. 101 deses non desis. 102 reses non resis. 103 vepres non vepris. 104 fames non famis. 105 clades non cladis. 106 syrtes non syrtis. 107 aedes non aedis. 108 sedes non sedis. 109 proles non prolis. 110 draco non dracco. 111 oculus non oclus. 112 aqua non acqua. 113 alium non aleum. 114 lilium non lileum. 115 glis non gliris. 116 delirus non delerus. 117 tinea non . 118 exter non extraneus. 119 clamis non clamus. 120 vir non vyr. 121 virgo non vyrgo. 122 virga non vyrga. 123 occasio non occansio. 124 caligo non calligo. 125 terebra non telebra. 126 effeminatus non imfimenatus. 127 botruus non butro. 128 grus non gruis. 129 anser non ansar. 130 tabula non tabla. 131 puella non poella. 132 balteus non baltius. 133 fax non facla. 134 vico capitis Africae non vico caput Africae. 135 vico tabuli proconsolis non vico tabulu proconsulis. 136 vico castrorum non vico castrae. 137 vico strobili non vico strobilu. 138 teter non tetrus. 139 aper non aprus. 140 amycdala non amiddula. 141 faseolus non fassiolus. 142 stabulum non stablum. 143 triclinium non triclinu. 144 dimidius non demidius. 145 turma non torma. 146 pusillus non pisinnus. 147 meretrix non menetris. 148 aries non ariex. 149 persica non pessica. 150 dysentericus non disintericus. 151 opobalsamum non ababalsamum. 152 mensa non mesa. 153 raucus non raucus (?) 154 auctor non autor. 155 auctoritas non autoritas. 156 157 linteum non lintium. 158 a... petre non ...tra. 159 terraemotus non terrimotium. 160 noxius non noxeus. 161 coruscus non scoriscus. 162 tonitru non tonotru. 163 passer non passar. 164 anser non ansar. 165 hirundo non harundo. 166 obstetrix non opsetris. 167 capitulum non capiclum. 168 noverca non novarca. 169 nurus non nura. 170 socrus non socra. 171 neptis non nepticla. 368

PAGE 369

369 172 anus non anucla. 173 tondeo non detundo. 174 rivus non rius. 175 imago non . 176 pauo{r} non pao{r}. 177 coluber non colober. 178 adipes non alipes. 179 sibilus non sifilus. 180 frustum non frustrum. 181 plebs non pleps. 182 garrulus non garulus. 183 parentalia non parantalia. 184 c[a]elebs non celeps. 185 poples non poplex. 186 locuples non locuplex. 187 robigo non rubigo. 188 plasta non blasta. 189 bipennis non bipinnis. 190 ermeneumata non erminomata. 191 tymum non tumum. 192 strofa non stropa. 193 bitumen non butumen. 194 mergus non mergulus. 195 myrta non murta. 196 zizipus non zizupus. 197 iunipirus non iuniperus. 198 tolerabilis non toleravilis. 199 basilica non bassilica. 200 tribula non tribla. 201 viridis non virdis 202 constabilitus non constabilitus(?) 203 sirena non serena. 204 musium vel musivum non museum. 205 labsus non lapsus. 206 orilegium non orolegium. 207 ostiae non {h}ostiae. 208 februarius non febrarius. 209 glatri non cracli. 210 allec non allex. 211 rabidus non rabiosus. 212 tintinaculum non tintinabulum. 213 adon non adonius. 214 grundio non grunnio. 215 vapulo non baplo. 216 necne non necnec. 217 passim non passi. 218 numquit non nimquit. 219 numquam non numqua. 220 nobiscum non noscum. 221 vobiscum non voscum. 222 nescioubi non nesciocobe. 223 pridem non pride 224 olim non oli. 225 adhuc non aduc. 226 idem non ide. 227 amfora non ampora.

PAGE 370

APPENDIX B APPLICATION OF PERFECT GRID In Arabic the superheavy foot only occurs in word final position and the surplus consonant is considered to ha ve secondary licensing, labeled (Goldsm ith 1990, 127). The configuration of a word final superheavy syllable a nd its relation to the metrical grid is shown in Figure B-1 (from Goldsmith 1990, 199) for the word kitaab book (R=rhyme, C=coda). x x x (x) R R C k i t a a b Figure B-1. Skeleton to grid asso ciation of a superheavy syllable. While kitaab contains only two syllables it occupies three slots on the metrical grid and potentially a fourth slot if the final consonant or appendix has a grid association. In Figure 2-4 the third grid mark (from left to right) is over the second elemen t of the long vowel or the coda. Placement of this grid mark is the result of Perfec t Grid, a rule formulated by Prince and cited in Goldsmith (1990, 194): scanning from left to right or from right to left assign a grid mark to every other grid mark on the immediately lower row. Perfect Grid may have peak first assignment, as in Arabic, where the placement of grid marks begins with the first available position. Alternatively, assignment of grid mark s may begin with the second position in which case it is designated trough first. The computation of word accent for, kitaab book, following the rule of Cairene Arabic which stresses the fi nal heavy or superheavy syllable (End Rule), is seen in Figure B-2 (after Goldsmith 1990, 200). B ecause Perfect Grid assigns a grid mark to the coda of the second syllable in kitaab an adjustment, weak mora st ress correction, is needed to move the place of prominence to the syllabic nucleus. 370

PAGE 371

x Row 2: End Rule (Final, Word) x x x Row 1: Perfect Grid: Assign x L to R, Peak first, to every other x in Row 0 x x x Row 0: Assign x to every mora k i t a a b Figure B-2. Perfect Grid L R (Cairene Arabic). The grid computation of Latin m jesttem fem. acc. sg. greatness, in Figure B-3, serves as an example of the adaptation of Quantity Se nsitivity (QS) to Perfect Grid (Goldsmith 1990, 197). QS places a grid mark on Row 1 (foot row) over any heavy syllable on Row 0 (mora row). Finally, End Rule selects the rightmost foot in th e word. Assignment of primary accent in Latin does not require exhaustive parsi ng but does require marking of th e first eligible heavy syllable counting from right to left. x Row 2: End Rule (Rightmost, Word) x Row 1: QS: Assign x R to L, Trough first, over first heavy syllable in Row 0 weak mora stress correction x x xx Row 0: Assign x to every mora m a je staa tem Figure B-3. QS and stress assignment (Latin). The status of syllable final consonants which would may or may not generate a mora in Row 0 can be clarified to some degree by examining the different types of quantity sensitivity that exist with regard to coda and appendix, especially in light of the fact that s+stop is no longer an acceptable word initial cluster in Western Romance. Historical deletion of syllable final /s/ in French fte as well as its fortition as [ ] in Portuguese festa (
PAGE 372

The superheavy syllable in Latin em erges most frequently in the case of monosyllables where the extra consonant is treated as an appendix, or as in auspex m. nom. sg. bird-seer where the final syllable is /-speks/. The treatmen t of the appendix in metrical terms is described in Table B-1 (after Goldsmith 1990, 207-208). Table B-1. Treatment of on the metrical grid 1) Totally quantity-insensitive: neither coda nor appendix licenses grid association onset rhyme nucleus coda C x Row 0 of grid 2) Totally quantity-sensitive: codas and appendices both license grid association onset rhym e nucleus coda C x x x Row 0 of grid 3) Codas license grid association, but the appendix does not onset rhym e nucleus coda C x x Row 0 of grid 4) The appendix licenses grid association, but the coda does not onset rhym e nucleus coda x x Row 0 of grid In Ibero-Romance cases which pose a dilemma for parsing into syllables also result internally from three or four segment cluster in which s+stop can no longer form the onset of a syllable, as in the case of monstrum i n. monster which results in Cast. monstruo, Port. 372

PAGE 373

monstro [ m t u] and C at. monstre The fortition of /s / in Portuguese is a clear indication that the consonant is in coda positi on and does not form an onset of the following syllable together with /tr/. Latin seems to corres pond to the third type described in Table B-1, that is, the coda is projected on the grid but not the appendix. With regard to the actual realization of a long vowel there is evidence that its duration is related to the presence or absence of a coda consonant. Such is the case in Levantine Arabic where there is significant shor tening of a long vowel before a consonant coda (Broselow, Chen, and Huffman 1997, 59-60); additionally there is evidence of shortening of coda consonants. The shorteni ng of both long vowels and coda consonants is interpreted as mora sharing. If the Latin superh eavy syllable is interpre ted in like manner, then, syllables of this type conform to the quantitat ive trochee (min/max=2) posited by Mester (1994). The first slot is occupied by the peak vowel while the non-nuclear vowel shares a mora with whatever element(s) may follow. 373

PAGE 374

APPENDIX C VERSIFIC ATION: FROM METER TO RHYTHM Although it cannot be categorically asserted that versification patterns accurately reflect the phonology, including accentual system, of a language the uniform abandonment of footbased meters in the three major Romance language s of the Iberian Peninsula is significant. Parsons (1999, 123) supports the notion that poeti c meters somehow favor the prosodic patterns of a language: I assume that meter is the styl ization of rhythm. A language will select a given metrical pattern because the pattern is inherently suitable to it. In other words, we should expect to find native rhythmic features in a language's me ters. It can, therefor e, be assumed that the transition from Latin to Romance verse forms refl ects fundamental changes in Latin. As Beare (1957, 56) states Latin had lost its old quantit ative rhythm, and the only principle, it is supposed, which could impose a new rhythm wa s the stress-accent. Furthermore, the quantitative meters of Classical Latin, with the possible exception of Saturnian verse, were of Greek origin (Beare 1957, 132). Assuming that Saturnian verse is somehow more representative of the prosodic properties of Latin, the basic features of this poetic form merit discussion. The origins of Saturnian verse are obscure and the extant cor pus is small (collected in Lu iselli 1967). Goldberg (1995, 60) notes that these examples of Saturnian verse reveal no normative pattern, either quantitative or accentual, on which to base a coherent metrical analysis, and vagaries of transmission often make essential details of their prosody uncomfo rtably problematic. Above all, Goldberg indicates the need to show that patterns of alternating heavy and light sy llables respond to meter rather than the fortuitous occurrence of such sequences in Latin words. Parsons (1999, 124) attempts to map Sa turnian verse to this hierarchy: L C || C A line consists of two cola separated by the principal caesura. C D | D | A colon consists of two dipodes. 374

PAGE 375

D F F Each dipode consists of two metrical feet. F PsPw Each foot consists of two metrical positions, the first strong and the second weak. P [ ( )] A metrical position is maximally a moraic trochee (but may be zero) To illustrate this proposed scansion, Parsons (1999, 125) uses an example from the Corpus Inscriptionum Latinorum an epitaph to Lucius Cornelius Scipio Barbatus (3rd century BCE). Figure 3-4 is an adapta tion of this diagram; moraic feet are enclosed in parentheses; extrametrical feet are in angled brackets, and un enclosed syllables are de generate or monomoraic feet. Two unfilled slots on the s/w row are indica ted by shading. According to Parsons, stress accent languages often have unrealized constituents after parsing for meter. In Classical Latin the inscription would read: s ubigit omnem Lucanam, obsidesque abducit (he subdued Lucania, and brought back hostages). L C C D D D D F F F F F F F F s w s w s w s w s w s w s w s w (subi) gi (t om) |(Lou) (ca) || (op) si (des) (ab) (dou) Figure C-1. Metrical analysis of a line of Saturnian verse. Despite the appealing symmetry of Parsons scansion, there is little motivation to support the creation of two feet to the left of the caesura when a more intuitive parsing is to have three feet corresponding to three words. Furthermore, the organization of the second hemistich seen in 375

PAGE 376

Figure C-1 also com es into doubt. It is as easily parsed as two trisyllabic feet corresponding to word stress and word boundaries: S W S | W S W. Accordingly, the first hemistich contains three feet; the first two are iambic, and the third amiphibrachic. Iambic f eet at the beginning of the line are subject to inversion and may appear as trochees as scanned below. This alternate scansion with uneven cola is shown below, repeat ing Parsons example; the first word is treated as a single foot, that is trip let timing as opposed to duple timing. In the second, hemistich the affix que and is treated as extrametrical. S W | S W | W S W || S W S | W S W subi git om ne | Lou ca nam || op si des ab dou cit This interpretation of Saturnian verse offers seve ral advantages. In a popular verse form there is more likely to be coincidence of ictus and accent, clearly not achieved in Parsons scanning in Figure C-1 but evident in the example above. Ther e is no compelling reason for hemistichs to be of equal length. The format F1 F2 F3 || F4 F5 (where F is a foot) is well attested without reliance on the triplet timing in the epitaph cited in Parsons, as in this example (Goldberg, 59): W S | W S | W S W || S W S | W S W da bunt ma lum Me tel l || Nae vi po tae the Metelli will give trouble to Naevius the poet Following the template established above the degree to which lines from the Cantar de mio Cid can be mapped onto the grid is remarkable (str essed syllables are under lined and it is assumed that e is joined to the following word, uos through synalepha). S W | S W | W S W || S W S | W S W vi .o | puer .tas | a .bier.tas || e u os sin | ca.a .dos he saw doors flung open and gates without bolts This preclassical model of Latin popular ve rse probably never disappeared entirely but attestations are ephemeral due to the nature and us es of this verse form (epitaphs, graffiti). As later Latin poets and song writers turn away fr om quantitative meter to a rhythmic meter, the 376

PAGE 377

stress-based poetic verse em erges again. B eare (1957, 206-207) quotes th e opening lines of an anonymous hymn cited in Bede (ca. 672-735 C.E.) as an example of modulatio (alternation of high and low sounds, defined in Martianus Cape lla, fl. 5th century C.E.) in contrast to ratio or proportion, which was only applie d to quantitative verse. apparebit repentina dies magna Domini, fur obscura velut nocte improvisos occupans. The great day of the Lord will appear suddenly as a thief in the dark night seizing the unwary. It is essentially built on trochaic feet as not ed by the 16th century musician Francisco Salinas (1577, 280): Item ad formam metri trochaici, rhythmum, qui canitur de die in diem per alphabetum (Also in the form of trochaic meter, rhythm, which is sung from day to day in alphabetical order1). Salinas uses the terms harmonia and rhythmum for the ratio / modulatio dichotomy. Using | to indicate foot boundaries and underlining to indicate primary word stress, these lines are rewritten below. A caesura has been added and the final syllable is considered to form part of a triplet timed f oot at the end of the verse. S W | S W | S W | S W || S W | S W | S W ap.pa | re .bit | re.pen | ti .na || di es | ma .gna | Do .mi.ni S W | S W | S W | S W || S W | S W | S W fur ob | scu .ra | ve .lut | noc .te) || im.pro.vi .sos | oc .cu.pans This pars ing assumes that at the time this hymn was written the trochee was syllabic rather than moraic because syllables labeled S are not necessarily bimoraic. 1 This refers to the so-called alphabetical or abecedarian hy mns in which each stanza begins with consecutive letters of the alphabet. 377

PAGE 378

The sim ilarities in the example above, the first hemistich of th e Saturnian verses previously presented, and phrase level rhythm in the modern tradition of popular verse and song such as nursery rhymes (Liberman 1979, Li berman and Prince 1977, Hayes 1995) are not surprising. Studies on music cognition (Lerda hl and Jackendoff 1996, Temperley 2001) have followed the theoretical framework of metrical theory to postulate well formedness constraints and metrical preferences. Lerdahl and Jackendoff (1996, 69-73), for example, propose the equivalent of the Continuous Column Constraint (Hayes 1995, 34) coupled with a constraint to avoid rhythmic clash. A well-formed metrical structure consists of several levels of beats, such that 1. Every beat at a given level must be a beat at all lower levels. 2. Exactly one or two beats at a given le vel must elapse between each pair of beats at the next level up. Building on the work of Lerdahl and Jackendoff, Temperley (2001, 38-39) adds two preference rules to reflect the favored status of duple rhythm. MPR 4 (Grouping Rule). Prefer to lo cate strong beats near the beginning of groups. MPR 5 (Duple Bias Rule). Prefer duple ove r triple relationships between levels. Working in tandem these two rules, Temperley s uggests, will place the in itial beat on the first note of a group and then place subsequent beats on every other note. It can be seen, then, that not only in language but in music, particularly uncultivated music, the preferred rhythmic pattern is duple rhythm with the strong beat on the initial memb er of a group. Examples from early Romance reflect the alte rnating stress pattern of Saturnian verse and the Late Latin hymn. They are also in duple rhyt hm; the length of the lin e in terms of syllable count may vary; and it is divided into two hemistichs each with two pr incipal stresses. The earliest period of Ibero-Romance is largely undo cumented because written texts continued to 378

PAGE 379

utilize Latin. However, W right (1982, 184) states that the presumed unattested popular verse of these centuries was patterned on stress. He goe s on to point out that the earliest ballads are often not isosyllabic and alludes to one from the cycle of the Romancero del Cid In the edition of Michalis de Vasconcellos, th e ballad known as La jura de Santa Gadea (The oath of Saint Agatha) appears in its en tirety. The irregularity of the syllable count is seen in these lines: Line 1 En Sancta Gadea de Burgos 9 syllables Line 9 Villanos te ma ten, Alonso 9 syllables Line 31 Si no dijeres la verdad 8 syllables + 1 Line 32 De lo que te fu ere preguntado 10 syllables The meter of the Spanish ballad is usually described as octosyllabic verse with alternating assonant rhyme. Variations in the syllable count of Spanish verse are designated (Baehr 1973, 22-23) as follows: the norm is verso llano; if a word with oxytonic st ress ends the line it is verso agudo and the syllable count is 7+1; if a word w ith proparoxytonic stress ends the line it is verso esdrjulo and the syllable count is 9-1. Although the existence of nomenclature that favors an exact number of syllables and labels variants im plies that the template requires isosyllabism, Baehr (1973, 37) notes that El isosilabismo, sin embargo, no es condicin previa ni necesaria para que exista un verso completo desde el punto de vista rtmico. Baeh r builds on a theory put forth by Henrquez Urea (1961, 19-24) that the basis of popular poetry, in particular the romance was rhythmic rather than syllabic. Henrquez Urea also notes that the earlier romances tend to be more irregular because with the passage of time the octosyllabic verse emerged as an art form. The three parallel texts in Table 316 are versions of the Balla d of the prisoner in Castilian, Catalan (Aguil y Fuster 1888, 79), an d Portuguese. The texts of the Castilian and Portuguese versions are from the Pan-Hispanic Ba llad Project. Because of the relative regularity of the meter these ballads are probably from a later period than La jura de Santa Gadea but 379

PAGE 380

they all correspond to a continu ing oral tradition in the Iberian Peninsula. Cases of synalepha are underlined. Occasionally, hiatus is maintained as in l. 6 of th e Portuguese version. All lines begin with an accented syllable which is the head of a trochee. Although there are many cases where word accent and ictus do not coincide (l. 6, Catalan; only the seventh syllable would be accented in natural discourse) it is possible for these lines to be recited with an alternating stress pattern beginning with the first sylla ble and ending with the seventh. Table C-1. Duple rhythm in the Iberian ballad tradition. Castilian Catalan Portuguese 1 sino yo triste, cuidado, 8 No hi vaig pas jo, mesquinet, 7+1 2 que vivo e n esta prisin, 7+1 3 que ni s cundo e s de da, 8 4 ni cuando las noches son 7+1 5 sino por una a vecilla 8 6 que me cantaba a l albor. 7+1 tancat ac e n greu pres 7+1 hont no conech si e s de dia, 8 si e s de dia ni si e s fosch, 7+1 sino per tres aucellets 7+1 que cantant voltan la tor. 7+1 S eu sou u m triste, coitado, 8 que a qui e stou nesta priso: 7+1 no sei quando de dia, 8 nem quando a rraia o sol, 7+1 se no so trs passarinhos 8 que me cantam no alvor. 7+1 The crucial accents are the first and the last In duple rhythm a strong beat at the beginning of the rhythmic group would be expected. This is reflected in the frequent inversion of verb and clitic pronoun, exemplified by the Romance de don Roldn (Clavera 2004, 124127). An acute accent mark indicates the primary accent of the first word. l. 111 Quitle luego las armas l. 121 djole que lo llevase l. 135 llrale toda la corte l. 138 furase para las tiendas l. 149 Recibile con mucha honra l. 158 hzole su capitn l. 187 Hcese una batalla l. 241 Conocironse entrambos l. 247 vnse con mucho amor Except in the last foot, clitic pronouns can be combined with verbs to produce the desired pattern of alternating rhythm as well as avoidanc e of an unstressable word at the beginning of a verse. In the example above the la st foot is either ('X.X) or ('X) setting aside for the moment the question of moraicity. It does not suffice for the last syllable to be heavy; it must also be capable 380

PAGE 381

of bearing an accent. T herefore, func tion words such as definite article un and preposition con, por, en could not constitute a phrase final foot. That ('X.X) is the preferred form is shown by the frequent use of paragogic -e in the Romance del Conde Dirlos (Clavera 2004, 3-40). Although ('X) may be binary in moraic terms if it consists of a single hea vy syllable, it lacks the preferred pattern of alternating a ccented and unaccented syllables. 381

PAGE 382

382 APPENDIX D DATABASE OF NOUNS IN CATALAN, CA STILIAN, AND PORTUGUESE WITH COMMON LATIN ETYMON

PAGE 383

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 1 altus, a, um 2 2 2 HC1.L m alt 11HC1 alto 22HC1.L alto 22HC1.L 2 annus, i, m. 2 2 2 HC1.L m any 11HC1 ao 22L1.L ano 22L1.L 3 VL bassus, a, um 2 2 2 HC1.L m baix 11HC1 bajo 22L1.L baixo 22HV1.L 4 barbus i, m. 2 2 2 HC1.L m barb 11HC1 barbo 22HC1.L barbo 22HC1.L 5 celtolat. bardus i, m. 2 2 2 HC1.L m bard 11HC1 bardo 22HC1.L bardo 22HC1.L 6 beccus, i, m. Gallic 2 2 2 HC1.L m bec 11HC1 pico 22L1.L pico 22L1.L 7 bulbus (bulb s ), i, m. 2 2 2 HC1.L m bulb 11HC1 bulbo 22 HC1.L bolbo 2 2 HC1.L 8 burgus i, m. 2 2 2 HC1.L m burg 1 1 HC1 burgo 2 2 HC1.L burgo 2 2 HC1.L 9 bustum i, n. 2 2 2 HC1.L m bust 1 1 HC1 busto 2 2 HC1.L busto 2 2 HC1.L 10 callum i, n. 2 2 2 HC1.L m call 1 1 HC1 callo 2 2 L1.L calo 2 2 L1.L 11 campus i, m. 2 2 2 HC1.L m camp 1 1 HC1 campo 2 2 HC1.L campo 2 2 HC1.L 12 cardus, i, m. 2 2 2 HC1.L m card 1 1 HC1 cardo 2 2 HC1.L cardo 2 2 HC1.L 13 cippus (c pus ), i, m. 2 2 2 HC1.L m cep 1 1 HC1 cepo 2 2 L1.L cepo 2 2 L1.L 14 circus i, m. 2 2 2 HC1.L m circ 1 1 HC1 circo 2 2 HC1.L circo 2 2 HC1.L 15 lat. v. coxus, a, um 2 2 2 HC1.L m coix 1 1 HC1 cojo 2 2 L1.L coxo 2 2 L1.L 16 collum i, n. 2 2 2 HC1.L m coll 1 1 HC1 cuello 2 2 L1.L colo 2 2 L1.L 17 cunnus i, m. 2 2 2 HC1.L m cony 1 1 HC1 coo 383 2 2 L1.L conho 2 2 L1.L 18 corvus i, m. 2 2 2 HC1.L m corb 1 1 HC1 cuervo 2 2 HC1.L corvo 2 2 HC1.L 19 crispus a, um 2 2 2 HC1.L m cresp 1 1 HC1 crespo 2 2 HC1.L crespo 2 2 HC1.L 20 damnum, i, n. 2 2 2 HC1.L m dany 1 1 HC1 dao 2 2 L1.L dano 2 2 L1.L 21 discus i, m., del gr. 2 2 2 HC1.L m disc 1 1 HC1 disco 2 2 HC1.L disco 2 2 HC1.L 22 dorsum i, n. 2 2 2 HC1.L m dors 1 1 HC1 dorso 2 2 HC1.L dorso 2 2 HC1.L 23 falsus, a, um 2 2 2 HC1.L m fals 1 1 HC1 falso 2 2 HC1.L falso 2 2 HC1.L 24 factus, a, um 2 2 2 HC1.L m fe t 1 1 HC1 hecho 2 2 L1.L feito 2 2 HV1.L 25 factum, i, n. 2 2 2 HC1.L m fe t 1 1 HC1 fecho 2 2 L1.L feito 2 2 HV1.L 26 fiscus, i, m. 2 2 2 HC1.L m fisc 1 1 HC1 fisco 2 2 HC1.L fisco 2 2 HC1.L 27 fictus, a, um 2 2 2 HC1.L m fit 1 1 HC1 hito 2 2 L1.L fito 2 2 L1.L 28 flaccus, a, um 2 2 2 HC1.L m fl ac 1 1 HC1 flaco 2 2 L1.L fraco 2 2 L1.L 2 2 2 HC1.L m fons 1 1 HC1 fondo 2 2 HC1.L fundo 2 2 HV1.L 2 2 2 HC1.L m fruit 1 1 HC1 fruto 2 2 L1.L fruto 2 2 L1.L 2 2 2 HC1.L m furt 1 1 HC1 hurto 2 2 HC1.L furto 2 2 HC1.L 2 2 2 HC1.L m gall 1 1 HC1 gallo 2 2 L1.L galho 2 2 L1.L 1 2 2 HC1.L m gat 1 1 HC1 gato 2 2 L1.L gato 2 2 L1.L

PAGE 384

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 34 VL grassus,a,um
PAGE 385

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 2 2 2 HC1.L m roig 1 1 HC1 rojo 2 2 L1.L roxo 2 2 L1.L 65 rhombus, i, m., 2 2 2 HC1.L m rom 1 1 HC1 rombo 2 2 HC1.L rombo 2 2 HV1.L 66 rhombus, i, m., = 2 2 2 HC1.L m rumb 1 1 HC1 rumbo 2 2 HC1.L rumbo 2 2 HV1.L 67 saccus, i, m., = 2 2 2 HC1.L m sac 1 1 HC1 saco 2 2 L1.L saco 2 2 L1.L 68 psalmus, i, m., = 2 2 2 HC1.L m salm 1 1 HC1 salmo 2 2 HC1.L salmo 2 2 HC1.L 69 sanctus, a, um 2 2 2 HC1.L m sant 1 1 HC1 santo 2 2 HC1.L santo 2 2 HV1.L 70 sargus, i, m., = 2 2 2 HC1.L m sarg 1 1 HC1 sargo 2 2 HC1.L sargo 2 2 HC1.L 71 servus, i, m. 2 2 2 HC1.L m serf 1 1 HC1 siervo 2 2 HC1.L servo 2 2 HC1.L 72 somnus, i, m. 2 2 2 HC1.L m son 1 1 HC1 sueo 2 2 L1.L sono 2 2 L1.L 73 thallus i, m., = 2 2 2 HC1.L m tall 1 1 HC1 tallo 2 2 L1.L talho 2 2 L1.L 74 tantus, a, um 2 2 2 HC1.L m tant 1 1 HC1 tanto 2 2 HC1.L tanto 2 2 HV1.L 75 ternus, a, um 2 2 2 HC1.L m tern 1 1 HC1 terno 2 2 HC1.L terno 2 2 HC1.L 76 testu (testum), i, n. 2 2 2 HC1.L m te st 1 1 HC1 tiesto 2 2 HC1.L testo 2 2 HC1.L 77 thyrsus, i, m., = 2 2 2 HC1.L m tirs 1 1 HC1 tirso 2 2 HC1.L tirso 2 2 HC1.L 78 turdus i, m. 2 2 2 HC1.L m to rd 1 1 HC1 tordo 2 2 HC1.L tordo 2 2 HC1.L 79 tornus i, m. 2 2 2 HC1.L m torn 1 1 HC1 torno 2 2 HC1.L torno 2 2 HC1.L 80 tortus a, um 2 2 2 HC1.L m tort 1 1 HC1 tuerto 385 2 2 HC1.L torto 2 2 HC1.L 81 transtrum i, n. 2 2 2 HC1.L m trast 1 1 HC1 trasto 2 2 HC1.L trasto 2 2 HC1.L 82 truncus i, m. 2 2 2 HC1.L m tronc 1 1 HC1 tronco 2 2 HC1.L tronco 2 2 HV1.L 83 unctum, i, n. 2 2 2 HC1.L m unt 1 1 HC1 unto 2 2 HC1.L unto 2 2 HV1.L 84 villus i, m. 2 2 2 HC1.L m vell 1 1 HC1 vello 2 2 L1.L velo 2 2 L1.L 85 ventus i, m. 2 2 2 HC1.L m vent 1 1 HC1 viento 2 2 HC1.L vento 2 2 HV1.L 86 verbum i, n. 2 2 2 HC1.L m verb 1 1 HC1 verbo 2 2 HC1.L verbo 2 2 HC1.L 87 ciccus, i, m., = 2 2 2 HC1.L m xic 11HC1 chico 22L1.L chico 22L1.L 88 ulmus, i, f. 2 2 2 HC1.L m om (ant. olm) 1 1 HC1 olmo 2 2 HC1.L olmo 2 2 HC1.L 176 furnus, i, m.. 2 2 2 HC1.L m forn 1 1 HC1 horno 2 2 HC1.L forno 2 2 HC1.L HC1.X (4th decl) (21) 89 cantus s, m. 4 2 2 HC1.L m cant 1 1 HC1 canto 2 2 HC1.L canto 2 2 HV1.L 90 census s, m. 4 2 2 HC1.L m cens 1 1 HC1 censo 2 2 HC1.L censo 2 2 HV1.L 91 corn s, n. 4 2 2 HC1.L m corn 1 1 HC1 cuerno 2 2 HC1.L corno 2 2 HC1.L 92 cursus s, m. 4 2 2 HC1.L m cors 1 1 HC1 corso 2 2 HC1.L corso 2 2 HC1.L 93 cursus s, m. 4 2 2 HC1.L m cs 1 1 HC1 coso 2 2 L1.L cosso 2 2 L1.L 94 cursus s, m. 4 2 2 HC1.L m curs 1 1 HC1 curso 2 2 HC1.L curso 2 2 HC1.L

PAGE 386

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 95 fluxus, s, m. 4 2 2 HC1.L m flux 11HC1 flujo 22L1.L fluxo 22HC1.L 96 gestus, s, m. 4 2 2 HC1.L m gest 11HC1 gesto 22HC1.L gesto 22HC1.L 97 gustus, s, m. 4 2 2 HC1.L m gust 11HC1 gusto 22HC1.L gosto 22HC1.L 98 partus, s, m. 4 2 2 HC1.L m part 11HC1 parto 22HC1.L parto 22HC1.L 99 passus, s, m. 4 2 2 HC1.L m pas 11HC1 paso 22L1.L passo 22L1.L 100 pastus, s, m. 4 2 2 HC1.L m past 11HC1 pasto 22HC1.L pasto 22HC1.L 101 planctus, s, m. 4 2 2 HC1.L m plany 11HC1 llanto 22HC1.L pranto 22HV1.L 102 puls us, s, m. 4 2 2 HC1.L m pols 1 1 HC1 pulso 2 2 HC1.L pulso 2 2 HC1.L 103 portus, s, m. 4 2 2 HC1.L m port 1 1 HC1 puerto 2 2 HC1.L porto 2 2 HC1.L 104 textus s, m. 4 2 2 HC1.L m text 1 1 HC1 texto 2 2 HC1.L texto 2 2 HC1.L 105 tractus, s, m. 4 2 2 HC1.L m tret 1 1 HC1 trecho 2 2 L1.L treito 2 2 HV1.L 106 arcus, s, m. 2 2 2 HC1.L m arc 1 1 HC1 arco 2 2 HC1.L arco 2 2 HC1.L 107 cinctus s, m. 2 2 2 HC1.L m cint 1 1 HC1 cinto 2 2 HC1.L cinto 2 2 HV1.L 108 sessus, s, m. 2 2 2 HC1.L m ses 1 1 HC1 sieso 2 2 L1.L sesso 2 2 L1.L 109 versus, s, m. 2 2 2 HC1.L m vers 1 1 HC1 verso 2 2 HC1.L verso 2 2 HC1.L HC1.X (prothesis) (6) 110 scincos or -us, i, m. 2 2 2 HC1.L m e scinc 2 1 HC1 escinco 3 2 HC1.L sengue 2 2 HV1.L 111 spartum (-ton), i, n., = 2 2 2 HC1.L m espart 2 1 HC1 esparto 3 2 HC1.L esparto 3 2 HC1.L 112 sponsus, a, um 2 2 2 HC1.L m es ps 2 1 HC1 esposo 386 3 2 L1.L esposo 3 2 L1.L 113 schistos, a, on, adj. 2 2 2 HC1.L m esquist 2 1 HC1 esquisto 3 2 HC1.L xisto 2 2 HC1.L 114 stannum (stagnum), i, n., 2 2 2 HC1.L m estany 2 1 HC1 estao 3 2 L1.L estanho 3 2 L1.L 115 strictus, a, um 2 2 2 HC1.L m estret 2 1 HC1 estrecho 3 2 L1.L estreito 3 2 HV1.L HC1.X (35) 116 amplus, a, um 2 2 2 HC1.L m ample 2 2 HC1.L ancho 2 2 HC1.L ancho 2 2 HV1.L 117 antrum, i, n., = 2 2 2 HC1.L m antre 2 2 HC1.L antro 2 2 HC1.L antro 2 2 HV1.L 118 astrum, i, n., cf. 2 2 2 HC1.L m astre 2 2 HC1.L astro 2 2 HC1.L astro 2 2 HC1.L 119 bombus i, m. 2 2 2 HC1.L m bombo 2 2 HC1.L bombo 2 2 HC1.L bombo 2 2 HV1.L 120 cactus i, m. 2 2 2 HC1.L m cactus 2 2 HC1.L cacto 2 2 HC1.L cacto 2 2 HC1.L 121 c l dum (caldum ), i, n. 2 2 2 HC1.L m caldo 2 2 HC1.L caldo 2 2 HC1.L caldo 2 2 HC1.L 122 carrus i, m. (
PAGE 387

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 125 cervus, i, m. 2 2 2 HC1.L m cervo 22HC1.L ciervo 22HC1.L cervo 22HC1.L 126 claustrum, i, n. 2 2 2 HC1.L m claustre 22HC1.L claustro 22HC1.L claustro 22HC1.L 127 conger, gri, m., = 2 2 2 HC1.L m congre 22HC1.L congrio 22HC1.L congro 22HV1.L 128 dexter, tra, trum 2 2 2 HC1.L m destr e 22HC1.L diestro 22HC1.L destro 22HC1.L 129 oestrus, i, m. 2 2 2 HC1.L m estre 22HC1.L estro 22HC1.L estro 22HC1.L 130 ferrum, i, n. 2 2 2 HC1.L m ferro 22HC1.L hierro 22HC1.L ferro 22HC1.L 131 fulcrum, i, n. 2 2 2 HC1.L m fulcre 22HC1.L fulcro 22HC1.L fulcro 22HC1.L 132 hy mnus, i, m., = 2 2 2 HC1.L m himne 2 2 HC1.L himno 2 2 HC1.L hino 2 2 L1.L 133 isthmus, i, m. 2 2 2 HC1.L m istme 2 2 HC1.L istmo 2 2 HC1.L istmo 2 2 HC1.L 134 lapsus, a, um 2 2 2 HC1.L m lapse 2 2 HC1.L lapso 2 2 HC1.L lapso 2 2 HC1.L 135 limbus, i, m. 2 2 2 HC1.L m limbe 2 2 HC1.L limbo 2 2 HC1.L limbo 2 2 HV1.L 136 lustrum, i, n. 2 2 2 HC1.L m lustre 2 2 HC1.L lustro 2 2 HC1.L lustro 2 2 HC1.L 137 membrum, i, n. 2 2 2 HC1.L m membre 2 2 HC1.L miembro 2 2 HC1.L membro 2 2 HV1.L 138 monstrum, i, n. 2 2 2 HC1.L m monstre 2 2 HC1.L monstruo 2 2 HC1.L monstro 2 2 HC1.L 139 nexus a, um 2 2 2 HC1.L m nexe 2 2 HC1.L nexo 2 2 HC1.L nexo 2 2 HC1.L 140 nimbus, i, m. 2 2 2 HC1.L m nimbe 2 2 HC1.L nimbo 2 2 HC1.L nimbo 2 2 HV1.L 141 pactum, i, n. 2 2 2 HC1.L m pacte 2 2 HC1.L pacto 2 2 HC1.L pacto 2 2 HC1.L 142 pl ctrum, i, n. 2 2 2 HC1.L m plectre 2 2 HC1.L plectro 387 2 2 HC1.L plectro 2 2 HC1.L 143 rastrum, i, n. 2 2 2 HC1.L m rastre 2 2 HC1.L rastro 2 2 HC1.L rastro 2 2 HC1.L 144 rectus, a, um 2 2 2 HC1.L m recte 2 2 HC1.L recto 2 2 HC1.L recto 2 2 HC1.L 145 regnum, i, n. 2 2 2 HC1.L m regne 2 2 HC1.L reino 2 2 HV1.L reino 2 2 HV1.L 146 rhythmos o -us, i, m., = 2 2 2 HC1.L m ritme 2 2 HC1.L ritmo 2 2 HC1.L ritmo 2 2 HC1.L 147 r strum, i, n. 2 2 2 HC1.L m rostre 2 2 HC1.L rostro 2 2 HC1.L rosto 2 2 HC1.L 148 signum, i, n. 2 2 2 HC1.L m signe 2 2 HC1.L signo 2 2 HC1.L signo 2 2 HC1.L 149 s strum, i, n., = 2 2 2 HC1.L m sistre 2 2 HC1.L si stro 2 2 HC1.L sistro 2 2 HC1.L 150 templum i, n. 2 2 2 HC1.L m temple 2 2 HC1.L templo 2 2 HC1.L templo 2 2 HV1.L HC1.X (4th decl) (6) 151 cultus s, m. 4 2 2 HC1.L m culte 2 2 HC1.L culto 2 2 HC1.L culto 2 2 HC1.L 152 luxus, s, m. 4 2 2 HC1.L m luxe 2 2 HC1.L lujo 2 2 L1.L luxo 2 2 L1.L 153 raptus, s, m. 4 2 2 HC1.L m rapte 2 2 HC1.L rapto 2 2 HC1.L rapto 2 2 HC1.L 154 sexus, s, m. 4 2 2 HC1.L m sexe 2 2 HC1.L sexo 2 2 HC1.L sexo 2 2 HC1.L 155 tactus s, m. 4 2 2 HC1.L m tacte 2 2 HC1.L tacto 2 2 HC1.L tacto 2 2 HC1.L 156 tractus s, m. 4 2 2 HC1.L m tracte 2 2 HC1.L tracto 2 2 HC1.L treito 2 2 HV1.L

PAGE 388

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp HC1.X (prothesis) (2) 157 spasmus, i, m., = 2 2 2 HC1.L m espasme 32HC1.L espasmo 32HC1.L espasmo 32HC1.L 158 spectrum, i, n. 2 2 2 HC1.L m espectre 32HC1.L espectro 32HC1.L espectro 32HC1.L HC1.X (3) 159 campus i, m. 2 2 2 HC1.L m campus 22HC1.HC campus 22HC1.HC campus 22HV1.HC 160 cancer cri m. 2 2 2 HC1.L m cncer 22HC1.HC cncer 22 HC1.HC cncer 22HV1.HC 161 phallus, i, m., = 2 2 2 HC1.L m fallus 22HC1.HC falo 22L1.L falo 22L1.L HC1.X > (4th decl) (1) 162 ictus, s, m. 4 2 2 HC1.L m ictus 2 2 HC1.HC ictus 2 2 HC1.HC icto 2 2 HC1.L HC1.X (4th decl) (1) 163 actus, s, m. 4 2 2 HC1.L m acte 2 2 HC1.L acto 2 2 HC1.L acto 2 2 HC1.L HC1.X (3) 164 porrum, i, n., 2 2 2 HC1.L m porro 2 2 L1.L porro 2 2 L1.L porro 2 2 L1.L 165 porrum, i, n. ; porrus, i, m. 2 2 2 HC1.L m porro 2 2 L1.L puerro 2 2 L1.L porro 2 2 L1.L 166 cirrus i, m. 2 2 2 HC1.L m cerro 2 2 L1.L cerro 2 2 L1.L cerro 2 2 L1.L HV1.X (37) 167 br tus a, um 2 2 2 HV1.L m brut 1 1 HC1 bruto 388 2 2 L1.L bruto 2 2 L1.L 168 caecus, a, um 2 2 2 HV1.L m cec 1 1 HC1 ciego 2 2 L1.L cego 2 2 L1.L 169 caelum, i, n., 2 2 2 HV1.L m cel 1 1 HC1 cielo 2 2 L1.L cu 1 1 HV1 170 cl rus a, um 2 2 2 HV1.L m clar 1 1 HC1 claro 2 2 L1.L claro 2 2 L1.L 171 c nus i, m. < 2 2 2 HV1.L m con 1 1 HC1 cono 2 2 L1.L cone 2 2 L1.L 172 c lus i, m. 2 2 2 HV1.L m cul 1 1 HC1 culo 2 2 L1.L cu 1 1 HV1 173 f tum, i, n. 2 2 2 HV1.L m fat 1 1 HC1 hado 2 2 L1.L fado 2 2 L1.L 174 faenum (f n-), i. n. 2 2 2 HV1.L m f enc 1 2 HC1 heno 2 2 L1.L feno 2 2 L1.L 175 f lum, i. n. 2 2 2 HV1.L m f il 1 1 HC1 hilo 2 2 L1.L fio 2 2 L1.L 177 f mus, i, m. 2 2 2 HV1.L m fum 1 1 HC1 humo 2 2 L1.L fumo 2 2 L1.L 178 gr tus, a, um 2 2 2 HV1.L m grat 1 1 HC1 grado 2 2 L1.L grado 2 2 L1.L 179 gr mus, i, m. 2 2 2 HV1.L m grum 1 1 HC1 grumo 2 2 L1.L grumo 2 2 L1.L 180 l mus, i, m. 2 2 2 HV1.L m llim 1 1 HC1 limo 2 2 L1.L limo 2 2 L1.L 181 M ius, i, m. 2 2 2 HV1.L m maig 1 1 HC1 mayo 2 2 L1.L maio 2 2 L1.L 182 m mus, i, m., = 2 2 2 HV1.L m mim 1 1 HC1 mimo 2 2 L1.L mimo 2 2 L1.L 183 m lus, i, m. 2 2 2 HV1.L m mul 1 1 HC1 mulo 2 2 L1.L mulo 2 2 L1.L 184 m rus, i, m. 2 2 2 HV1.L m mur 1 1 HC1 muro 2 2 L1.L muro 2 2 L1.L 185 m tus a, um 2 2 2 HV1.L m mut 1 1 HC1 mudo 2 2 L1.L mudo 2 2 L1.L

PAGE 389

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 186 n nus i, m., 2 2 2 HV1.L m nan 11HC1 enano 32L.L1.L ano 21L.HV1 187 n pus, i, m. 2 2 2 HV1.L m nap 11HC1 nabo 22L1.L nabo 22L1.L 188 aurum, i, n. 2 2 2 HV1.L m or 11HC1 oro 22L1.L ouro 22L1.L 189 p lus, i, m. 2 2 2 HV1.L m pal 11HC1 palo 22L1.L pau 11HV1 190 paucus, a, um 2 2 2 HV1.L m poc 11HC1 poco 22L1.L pouco 22L1.L 191 p mum, i, n. 2 2 2 HV1.L m pom 11HC1 pomo 22L1.L pomo 22L1.L 192 pr tum, i, n. 2 2 2 HV1.L m prat 11HC1 prado 22L1.L prado 22L1.L 2 2 2 HV1.L m prim 1 1 HC1 primo 2 2 L1.L primo 2 2 L1.L 2 2 2 HV1.L m ram 1 1 HC1 ramo 2 2 L1.L ramo 2 2 L1.L 2 2 2 HV1.L m ras 1 1 HC1 raso 2 2 L1.L raso 2 2 L1.L 196 r mus, i, m. 2 2 2 HV1.L m re m 1 1 HC1 remo 2 2 L1.L remo 2 2 L1.L 197 s cus (succus), i, m. 2 2 2 HV1.L m suc 1 1 HC1 jugo 2 2 L1.L suco 2 2 L1.L 198 t tus a, um 2 2 2 HV1.L m to t 1 1 HC1 todo 2 2 L1.L todo 2 2 L1.L 199 tphus i, m. 2 2 2 HV1.L m tuf 1 1 HC1 tufo 2 2 L1.L tufo 2 2 L1.L 200 v sum & v sus i, m / n. 2 2 2 HV1.L m vas 1 1 HC1 vaso 2 2 L1.L vaso 2 2 L1.L 201 v lum i, n. 2 2 2 HV1.L m vel 1 1 HC1 velo 2 2 L1.L veio 2 2 L1.L 202 v tum i 2 2 2 HV1.L m vot 1 1 HC1 voto 389 2 2 L1.L voto 2 2 L1.L 203 z lus i, m. 2 2 2 HV1.L m zel 1 1 HC1 celo 2 2 L1.L zelo 2 2 L1.L HV1.X (4th decl) (4) 204 c sus, s, m. 4 2 2 HV1.L m cas 1 1 HC1 caso 2 2 L1.L caso 2 2 L1.L 205 fl tus, s, m. 4 2 2 HV1.L m flat 1 1 HC1 flato 2 2 L1.L flato 2 2 L1.L 206 f sus, s, m. 4 2 2 HV1.L m fus 1 1 HC1 huso 2 2 L1.L fuso 2 2 L1.L 207 sus s, m. 4 2 2 HV1.L m s 1 1 HC1 uso 2 2 L1.L uso 2 2 L1.L HV1.X (prothesis) (4) 208 scr tum, i, n. 2 2 2 HV1.L m escrot 2 1 HC1 escroto 3 2 L1.L escroto 3 2 L1.L 209 sc tum, i, n. 2 2 2 HV1.L m escut 2 1 HC1 escudo 3 2 L1.L escudo 3 2 L1.L 210 sp tum, i, n. 2 2 2 HV1.L m esput 2 1 HC1 esputo 3 2 L1.L esputo 3 2 L1.L 211 str tum, i, n. 2 2 2 HV1.L m estrat 2 1 HC1 estrado 3 2 L1.L estrado 3 2 L1.L HV1.X (prothesis) (4th decl) (1) 212 str tus, s, m. 4 2 2 HV1.L m estrat 2 1 HC1 estrato 3 2 L1.L estrato 3 2 L1.L HV1.X > HV1 (18) 213 cl vus i, m. 2 2 2 HV1.L m clau 1 1 HV1 clavo 2 2 L1.L cravo 2 2 L1.L 214 d num i, n. 2 2 2 HV1.L m do 1 1 HV1 don 1 1 HC1 dom 1 1 HV1 215 LL feudum (Germ.) 2 2 2 HV1.L m f eu 1 1 HV1 feudo 2 2 HV1.L feudo 2 2 HV1.L

PAGE 390

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 216 fr num or fraenum, i, n. 2 2 2 HV1.L m fre 11HV1 freno 22L1.L freio 22L1.L 217 gr num, i, n. 2 2 2 HV1.L m gra 1 1 HV1 grano 2 2 L1.L gro 1 1 HV1 218 j gum, i, n. 2 2 2 HV1.L m jou 1 1 HV1 yugo 2 2 L1.L jugo 2 2 L1.L 219 l num, i, n. 2 2 2 HV1.L m lli 1 1 HV1 lino 2 2 L1.L linho 2 2 L1.L 220 n dus i, m. 2 2 2 HV1.L m ni u 1 1 HV1 nido 2 2 L1.L ninho 2 2 L1.L 221 vum, i, n. 2 2 2 HV1.L m ou 1 1 HV1 huevo 2 2 L1.L ovo 2 2 L1.L 222 pl nus a, um 2 2 2 HV1.L m pl a 1 1 HV1 llano 2 2 L1.L cho 1 1 HV1 223 pl num, i, n 2 2 2 HV1.L m pla 1 1 HV1 plano 2 2 L1.L plano 2 2 L1.L 224 pl nus, a, um 2 2 2 HV1.L m ple 1 1 HV1 lleno 2 2 L1.L cheio 2 2 L1.L 225 pl nus, a, um, adj. 2 2 2 HV1.L m pl e 1 1 HV1 pleno 2 2 L1.L pleno 2 2 L1.L 226 r vus, i, m. 2 2 2 HV1.L m ri u 1 1 HV1 ro 2 2 L1.L rio 2 2 L1.L 227 s bum, i, n. 2 2 2 HV1.L m su 1 1 HV1 sebo 2 2 L1.L sebo 2 2 L1.L 2 2 2 HV1.L m va 1 1 HV1 vano 2 2 L1.L vo 1 1 HV1 229 v num i, n. 2 2 2 HV1.L m vi 1 1 HV1 vino 2 2 L1.L vinho 2 2 L1.L 230 vivus, a, um 2 2 2 HV1.L m vi u 1 1 HV1 vivo 2 2 L1.L vivo 2 2 L1.L HV1.X (4th decl) (1) 231 p nus, s & i, f. 4 2 2 HV1.L m pi 1 1 HV1 pino 2 2 L1.L pinho 2 2 L1.L HV1.X (4) 232 auster tri, m. 2 2 2 HV1.L m austr e 2 2 HV1.L austro 390 2 2 HV 1.L austro 2 2 HV1.L 233 c prum i, n. 2 2 2 HV1.L m cour e 2 2 HV1.L cobre 2 2 L1.L cobre 2 2 L1.L 234 Faunus, i, m. 2 2 2 HV1.L m f aune 2 2 HV1.L fauno 2 2 HV 1.L fauno 2 2 HV1.L 235 taurus i, m. 2 2 2 HV1.L m taure 2 2 HV1.L taur o 2 2 HV1.L tauro 2 2 HV1.L HV1.X (2) 236 l tos and l tus, i, f., = 2 2 2 HV1.L m lotus 2 2 L1.HC loto 2 2 L1.L loto 2 2 L1.L 237 v rus i, n. 2 2 2 HV1.L m viru s 2 2 L1.HC virus 2 2 L1.HC vrus 2 2 L1.HC HV1.X (4th decl) (1) 238 f tus (foet-), s, m. 4 2 2 HV1.L m fetus 2 2 L1.HC feto 2 2 L1.L feto 2 2 L1.L HV1.X (8) 239 c drus i, f. < 2 2 2 HV1.L m cedre 2 2 L1. L cedro 2 2 L1.L cedro 2 2 L1.L 240 qu drus, a, um 2 2 2 HV1.L m quadr e 2 2 L1.L cuadro 2 2 L1.L quadro 2 2 L1.L 241 d vus (d us ), i, m. 2 2 2 HV1.L m di vo 2 2 L1.L divo 2 2 L1.L divo 2 2 L1.L 242 d plus a, um, adj. 2 2 2 HV1.L m dup le 2 2 L1.L duplo 2 2 L1.L duplo 2 2 L1.L 243 l crum, i, n. 2 2 2 HV1.L m lucre 2 2 L1.L lucro 2 2 L1.L lucro 2 2 L1.L 244 maurus, a, um 2 2 2 HV1.L m mo ro 2 2 L1.L moro 2 2 L1.L mouro 2 2 L1.L 245 n dus i, m. 2 2 2 HV1.L m nod e 2 2 L1.L nodo 2 2 L1.L nodo 2 2 L1.L

PAGE 391

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 246 taurus i, m. 2 2 2 HV1.L m toro 22L1.L toro 22L1.L touro 22L1.L HV1.X (4th decl) (2) 247 cho, s, f., = 4 2 2 HV1.L m eco 22L1. L eco 22L1.L eco 22L1.L 248 r tus, s, m. 4 2 2 HV1.L m ritu 22 L1.L rito 22L1.L rito 22L1.L HV1.X (prothesis) (1) 249 st prum, i, n. 2 2 2 L1.L m estupre 3 2 L1.L estupro 3 2 L1.L estupro 3 2 L1.L L1.X (28) 250 bl tum i, n. 2 2 2 L1.L m blet 1 1 HC1 bledo 2 2 L1.L bredo 2 2 L1.L 251 ch rus, i, m., = 2 2 2 L1.L m cor 1 1 HC1 coro 2 2 L1.L coro 2 2 L1.L 252 c bus, i, m. 2 2 2 L1.L m cub 1 1 HC1 cubo 2 2 L1.L cubo 2 2 L1.L 253 LL cucum < coccum, i, n., = 2 2 2 L1.L m cuc 1 1 HC1 cuco 2 2 L1.L cuca 2 2 L1.L 254 LL d lum
PAGE 392

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp L1.X (4th decl) (1) 278 l cus s, m. 4 2 2 L1.L m llac 11HC1 lago 22L1.L lago 22L1.L L1.X (prothesis) (1) 279 st lus, i, m. 2 2 2 L1.L m estil 21HC. HC1 estilo 32L1.L estilo 32L1.L L1.X (prothesis) (4th decl) (1) 280 st tus, s, m. 4 2 2 L1.L m estat 21 HC1 estado 32L1.L estado 32L1.L L1.X (4) 281 d us i, m. 2 2 2 L1.L m du 11HV1 dios 11HC1 deus 11HC1 282 *nudus < n dus, i, m. 2 2 2 L1.L m nu (nuu) 11HV1 nudo 2 2 L1.L n 2 2 HV1 283 r us, i, m. 2 2 2 L1.L m reu 1 1 HV1 reo 2 2 L1.L ru 1 1 HV1 284 t nus i, m., = 2 2 2 L1.L m to 1 1 HV1 tono 2 2 L1.L tono 2 2 L1.L L1.X (4th decl) (3) 285 m nus, s, f. 4 2 2 L1.L f m 1 1 HV1 mano 2 2 L1.L mo 1 1 HV1 286 gr dus, s, m. 4 2 2 L1.L m grau 1 1 HV 1 grado 2 2 L1.L grado 2 2 L1.L 287 s nus, s, m. 4 2 2 L1.L m si 1 1 HV1 seno 2 2 L1.L seio 2 2 L1.L L1.X (prothesis) (1) 288 LL sclavus
PAGE 393

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 303 n ger, gra, grum 2 2 2 L1.L m negr e 22L1.L negro 22L1.L negro 22L1.L 304 siclus, i, m. (Heb. shekel ) 2 2 2 L1.L m sicle 22L1.L siclo 22L1.L siclo 22L1.L 305 tr plus a, um 2 2 2 L1.L m triple 2 2 L1.L triplo 2 2 L1.L triplo 2 2 L1.L 306 v gus a, um 2 2 2 L1.L m v ague 2 2 L1.L vago 2 2 L1.L vago 2 2 L1.L L1.X (4th decl) (1) 307 tr bus s, f. 4 2 2 L1.L f tribu 2 2 L1. L tribu 2 2 L1.L tribo 2 2 L1.L HC.HC1.X (31) 308 adiunctus, a, um 2 3 2 HC.HC1.L m adjunt 2 1 HC.HC1 adjunto 3 2 HC.HC1 .L adjunto 3 2 HC.HV1.L 309 arbustum, i, n. 2 3 2 HC.HC1.L m arbust 2 1 HC.HC1 arbusto 3 2 HC.HC1.L arbusto 3 2 HC.HC1.L 310 asbest s i, m. 2 3 2 HC.HC1.L m asbest 2 1 HC. HC1 asbesto 3 2 HC.HC1.L asbesto 3 2 HC.HC1.L 311 asphaltus, i, m. = 2 3 2 HC.HC1.L m asfalt 2 1 HC.HC1 as falto 3 2 HC.HC1.L asfalto 3 2 HC.HC1.L 312 castellum i, n. 2 3 2 HC.HC1.L m caste ll 2 1 HC.HC1 castillo 3 2 HC.L1.L castelo 3 2 HC.L1.L 313 cultellus i, m. 2 3 2 HC.HC1.L m coltell 2 1 HC.HC1 cuchillo 3 2 L.L1.L cutelo 3 2 L.L1.L 314 confessus a, um 2 3 2 HC.HC1.L m confs 2 1 HC.HC1 confeso 3 2 HC.HC1.L confesso 3 2 HV.HC1.L 315 contentus a, um 2 3 2 HC.HC1.L m content 2 1 HC.HC1 content o 3 2 HC.HC1.L contente 3 2 HV.HV1.L 316 conversus, a, um 2 3 2 HC.HC1.L m convers 2 1 HC.HC1 converso 3 2 HC.HC1.L converso 3 2 HV.HC1.L 317 crystallum i, n. 2 3 2 HC.HC1.L m crista ll 2 1 HC.HC1 cristal 393 2 1 HC.HC1 cristal 2 1 HC.HC1 318 diphthongus i, f., = 2 3 2 HC.HC1.L m diftong 2 1 HC.HC1 dipt ongo 3 2 HC.HC1.L ditongo 3 2 L.HV1.L 319 incensum, i, n. 2 3 2 HC.HC1.L m encens 2 1 HC. HC1 incienso 3 2 HC.HC1.L incenso 3 2 HV.HC1.L 320 expertus a, um 2 3 2 HC.HC1.L m expert 2 1 HC.HC1 experto 3 2 HC.HC1.L experto 3 2 HV.HC1.L 321 fermentum, i, n. 2 3 2 HC.HC1.L m ferment 2 1 HC.HC1 fermento 3 2 HC.HC1.L fermento 3 2 HC.HC1.L 322 fragmentum, i, n. 2 3 2 HC.HC1.L m fragment 2 1 HC.HC1 fragmento 3 2 HC.HC1.L fragmento 3 2 HC.HV1.L 323 infartus, a, um 2 3 2 HC.HC1.L m infart 2 1 HC.HC1 infarto 3 2 HC.HC1.L infarto 3 2 HV.HC1.L 324 infernum, i, n. 2 3 2 HC.HC1.L m infern 2 1 HC.HC1 infierno 3 2 HC.HC1.L inferno 3 2 HV.HC1.L 325 intentus, a, um 2 3 2 HC. HC1.L m intent 2 1 HC.HC1 intent o 3 2 HC.HC1.L intento 3 2 HV.HC1.L 326 internus, a, um 2 3 2 HC.HC1.L m intern 2 1 HC.HC1 interno 3 2 HC.HC1.L interno 3 2 HV.HC1.L 327 inventum, i, n. 2 3 2 HC.HC1.L m invent 2 1 HC.HC1 invento 3 2 HC.HC1.L invento 3 2 HV.HC1.L 328 lentiscum, i, n. 2 3 2 HC. HC1.L m llentisc 2 1 HC.HC1 lentisco 3 2 HC.HC1.L lentisco 3 2 HV.HC1.L 329 molluscum, i, n. 2 3 2 HC.HC1.L m mollusc 2 1 HC.HC1 molusco 3 2 L.HC1.L molusco 3 2 L.HC1.L 330 nocturnus a, um 2 3 2 HC.HC1.L m nocturn 2 1 HC.HC1 nocturno 3 2 HC.HC1.L nocturno 3 2 HC.HC1.L 331 pistillum, i, n., pistillus, i, m. 2 3 2 HC.HC1.L m pestell 2 1 HC.HC1 pes tillo 3 2 HC.L1.L pestilo 3 2 HC.L1.L 332 pigmentum, i, n. 2 3 2 HC.HC1.L m pigment 2 1 HC.HC1 pigmento 3 2 HC.HC1 .L pigmento 3 2 HC.HV1.L 333 pistillum, i, n. 2 3 2 HC.HC1.L m pistil 2 1 HC.HC1 pistilo 3 2 HC. L1.L pistilo 3 2 HC.L1.L

PAGE 394

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 334 portentus, a, um 2 3 2 HC. HC1.Lm portent 21HC.HC1 portent o 32HC.HC1.Lportento 32HC.HV1.L 335 sarmentum, i, n. 2 3 2 HC.HC1.Lm sarment 21HC.HC1 sarmiento 32HC.HC1.Lsarmento 32HC.HV1.L 336 segmentum, i, n. 2 3 2 HC.HC1.Lm segment 21HC.HC1 segmento 32HC.HC1.Lsegmento 32HC.HV1.L 337 torm ntum, i, n. 2 3 2 HC.HC1.Lm turment 21 HC.HC1 tormento 32HC.HC1 .Ltormento 32HC.HV1.L 338 unguentum, i 2 3 2 HC.HC1.Lm ungent 21 HC.HC1 ungento 32HC.HC1.Lunguento 32HV.HV1.L HC.HC1.X (4th decl) (18) 339 abscessus, s, m. 4 3 2 HC.HC1.Lm abscs 21HC.HC1 absceso 32HC.L1.L abcesso 32HC.L1.L 340 accentus, s, m. 4 3 2 HC.HC1.L m accent 2 1 HC.HC1 acento 3 2 L.HC1.L acento 3 2 L.HV1.L 341 accessus, s, m. 4 3 2 HC.HC1.L m accs 2 1 HC.HC1 acceso 3 2 HC.L1.L acesso 3 2 L.L1.L 342 complexus, s, m. 4 3 2 HC.HC1.L m complex 2 1 HC.HC1 complejo 3 2 HC.L1.L complexo 3 2 HV.HC1.L 343 concursus s, m. 4 3 2 HC.HC1.L m concurs 2 1 HC.HC1 c oncurso 3 2 HC.HC1.L concurso 3 2 HV.HC1.L 344 congressus s, m. 4 3 2 HC.HC1.L m congrs 2 1 HC.HC1 congreso 3 2 HC.L1.L c ongresso 3 2 HV.L1.L 345 conjunctus s, m. 4 3 2 HC.HC1.L m conjunt 2 2 HC.HC1 conjunto 3 2 HC.HC1.L conjunto 3 2 HV.HC1.L 346 contextus s, m. 4 3 2 HC.HC1.L m context 2 1 HC.HC1 contexto 3 2 HC.HC1.L contexto 3 2 HV.HC1.L 347 conventus s, m. 4 3 2 HC.HC1.L m convent 2 1 HC.HC1 convento 3 2 HC.HC1.L convento 3 2 HV.HC1.L 348 discursus s, m. 4 3 2 HC.HC1.L m discurs 2 1 HC.HC1 discurso 3 2 HC.HC1.L discurso 3 2 HC.HC1.L 349 excessus, s, m. 4 3 2 HC.HC1.L m excs 2 1 HC.HC1 exceso 3 2 HC.L1.L excesso 3 2 HC.L1.L 350 impulsus, s, m. 4 3 2 HC.HC1.L m impuls 2 1 HC.HC1 impulso 3 2 HC.HC1.L impulso 3 2 HC.HC1.L 351 incestus, s, m. 4 3 2 HC.HC1.L m incest 2 1 HC.HC1 incesto 394 3 2 HC.HC1.L incesto 3 2 HV.HC1.L 352 indultus, s, m. 4 3 2 HC.HC1.L m indult 2 1 HC.HC1 indulto 3 2 HC.HC1.L indulto 3 2 HV.HC1.L 353 ingressus, s, m. 4 3 2 HC.HC1.L m ingrs 2 1 HC.HC1 ingreso 3 2 HC.L1.L ingresso 3 2 HV.L1.L 354 instinctus, s, m. 4 3 2 HC.HC1.L m instint 2 1 HC.HC1 in stinto 3 2 HC.HC1.L instinto 3 2 HV.HV1.L 355 successus, s, m. 4 3 2 HC.HC1.L m succs 2 1 HC.HC1 suceso 3 2 L.L1.L sucesso 3 2 L.L1.L 356 adventus, s, m. 4 3 2 HC.HC1.L m advent 2 1 HC.HC1 adviento 3 2 HC.HC1.L advento 3 2 HC.HV1.L 399 d spectus s, m. 4 3 2 HC.HC1.L m despit 2 1 HC.HC1 despecho 3 2 HC.L1.L despeito 3 2 HC.HV1.L HC.HC1.X (prothesis) (1) 357 scalpellum, i, n. 2 3 2 HC.HC1.L m escalpel 3 1 HC.HC1 escalpelo 4 2 HC.L1.L escalpelo 4 2 HC.L1.L HC.HC1.X (10) 358 addictus, a, um 2 3 2 HC.HC1.L m addicte 3 2 HC.HC1.L adicto 3 2 L.HC1.L adicto 3 2 L.HC1.L 359 baptismus i, m. 2 3 2 HC.HC1.L m baptisme 3 2 HC.HC1.L baptismo 3 2 HC.HC1.L baptismo 3 2 HC.HC1.L 360 collapsus, a, um 2 3 2 HC.HC1.L m collapse 3 2 HC.HC1.L colapso 3 2 L.HC1.L colapso 3 2 L.HC1.L 361 compactum, i, n. 2 3 2 HC.HC1.L m compacte 3 2 HC.HC1.L compacto 3 2 HC.HC1.L compacto 3 2 HV.HC1.L 362 conductum i, n. 2 3 2 HC.HC1.L m conducte 3 2 HC.HC1.L conducto 3 2 HC. HC1.L conduto 3 2 HV.L1.L 363 districtus a, um 2 3 2 HC.HC1.L m districte 3 2 HC.HC1.L distrito 3 2 HC. L1.L distrito 3 2 HC.L1.L 364 exemplum i, n. 2 3 2 HC.HC1.L m exemple 3 2 HC.HC1.L ejemplo 3 2 L.HC1.L exemplo 3 2 L.HV1.L

PAGE 395

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 365 extractus, a, um 2 3 2 HC.HC1.Lm extracte 32HC.HC1.Lextracto 32HC.HC1.Lextracto 32HC.HC1.L 366 insectum, i, n. 2 3 2 HC.HC1.Lm insecte 32HC.HC1.Linsecto 32HC.HC1.Linsecto 32HV.HC1.L 367 sarcasmos, i, m., = 2 3 2 HC.HC1.Lm sarcasme 32HC.HC1.Lsarcasmo 32HC.HC1.Lsarcasmo 32HC.HC1.L HC.HC1.X (4th decl) (9) 368 conceptus s, m. 4 3 2 HC.HC1.Lm concepte 32HC.HC1.Lconcepto 32HC.HC1.Lconceito 32HV.HV1.L 369 conflictus s, m. 4 3 2 HC.HC1.Lm conflicte 32HC.HC1.L conflicto 32HC.HC1.L conflito 32HV.L1.L 370 contactus s, m. 4 3 2 HC.HC1.Lm contacte 32HC.HC1 .Lcontacto 32HC.HC1.Lcontacto 32HV.HC1.L 371 contractus s, m. 4 3 2 HC.HC1.L m contracte 3 2 HC.HC1 .L contrato 3 2 HC.L1.L contrato 3 2 HV.L1.L 372 olfactus, s, m. 4 3 2 HC.HC1.L m olfacte 3 2 HC.HC1.L olfato 3 2 HC.L1.L olfacto 3 2 HC.HC1.L 373 subjectus s, m. 4 3 2 HC.HC1.L m subjecte 3 2 HC. HC1.L sujeto 3 2 L.L1.L sujeito 3 2 L.HV1.L 374 transcursus s, m. 4 3 2 HC.HC1.L m transcurs 3 2 HC.HC1.L transcurso 3 2 HC.HC1.L transcurso 3 2 HC.HC1.L 375 aspectus, s, m. 2 3 2 HC.HC1.L m aspecte 3 2 HC.HC1 .L aspecto 3 2 HC.HC1.L aspeito 3 2 HC.HV1.L 376 objectus, s, m. 4 3 2 HC.HC1.L m objecte 3 2 HC. HC1.L objeto 2 2 HC.L1.L objecto 2 2 HC.L1.L HC.HC1.X (7) 377 transversus, a, um 2 3 2 HC.HC1.L m travs 2 1 L.HC1 travs 2 1 L.HC1 travs 2 1 L.HC1 378 annexus, a, um 2 3 2 HC.HC1.L m anex 2 1 L.HC1 anejo 3 2 L.L1.L anexo 3 2 L.L1.L 379 annexus, a, um 2 3 2 HC.HC1.L m anex 2 1 L.HC1 anexo 3 2 L.HC1.L anexo 3 2 L.HC1.L 380 pigmentum, i, n. 2 3 2 HC.HC1.L m piment 2 1 L.HC1 pimiento 395 3 2 L.HC1.L pimento 3 2 L.HV1.L 381 possessus, a, um 2 3 2 HC.HC1.L m posss 2 1 L.HC1 poseso 3 2 L.L1. L possesso 3 2 L.L1.L 382 suffixus, a, um 2 3 2 HC.HC1.L m sufix 2 1 L.HC1 sufijo 3 2 L.L1.L sufixo 3 2 L.HC1.L 383 b. lat. vassallus 2 3 2 HC.HC1.L m vassall 2 1 L.HC1 vasallo 3 2 L.L1.L vassalo 3 2 L.L1.L 407 affixus, a, um 2 3 2 HC.HC1.L m afix 2 1 L.HC1 afijo 3 2 L.L1. L afixo 3 2 L.HC1.L HC.HC1.X (1) 384 assumptus, a, um 2 3 2 HC.HC1.L m assumpte 3 2 L.HC1.L asunto 3 2 L.HC1 .L assunto 3 2 L.HV1.L HC.HC1.X (4th decl) (2) 385 affectus, s, m. 4 3 2 HC.HC1.L m afecte 3 2 L.HC1.L afecto 3 2 L.HC1.L afecto 3 2 L.HC1.L 386 effectus s, m. 4 3 2 HC.HC1.L m efecte 3 2 L.HC1.L efecto 3 2 L.HC1.L efeito 3 2 L.HV1.L 387 respectus, s, m. 4 3 2 HC.HC1.L m respecte 3 2 L.HC1.L respecto 3 2 L.HC1.L respeito 3 2 L.HV1.L HC.HV1.X (5) 388 cost tus, a, um 2 3 2 HC.HV1.L m costat 2 1 HC.HC1 costado 3 2 HC.L1. L costado 3 2 HC.L1.L 389 discr tus, a, um 2 3 2 HC.HV1.L m discret 2 1 HC.HC1 discreto 3 2 HC.L1. L discreto 3 2 HC.L1.L 390 extr mus a, um 2 3 2 HC.HV1.L m extrem 2 1 HC.HC1 extremo 3 2 HC.L1. L extremo 3 2 HC.L1.L 391 mand tum, i, n. 2 3 2 HC.HV1.L m mandat 2 1 HC.HC1 mandato 3 2 HC.L1.L mandat o 3 2 HV.L1.L 392 int ger, t gra, t grum 2 3 2 HC.HV1.L m ent er 2 1 HC.HC1 entero 3 2 HC.L1.L inteiro 3 2 HV.HV1.L

PAGE 396

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp HC.HV1.X (4th decl)(7) 393 concr tus s, m. 4 3 2 HC.HV1.Lm concret 21HC.HC1 concreto 32HC.L1.L concreto 32HV.L1.L 394 inc sus, s, m. 4 3 2 HC.HV1.Lm incs 21HC.HC1 inciso 32HC.L1.L inciso 32HV.L1.L 395 merc tus, s, m. 4 3 2 HC.HV1.Lm mercat 21HC. HC1 mercado 32HC.L1.L mercado 32HC.L1.L 396 substr tus, s, m. 4 3 2 HC.HV1.Lm s ubstrat 21HC.HC1 sustrato 32HC.L1.L substr ato 32HC.L1.L 397 transl tus, s, m. 4 3 2 HC.HV1.Lm trasllat 21HC.HC1 traslado 32HC.L1.L traslado 32HC.L1.L 398 tract tus s, m. 4 3 2 HC.HV1.Lm tractat 21 HC.HC1 tratado 32L.L1. L tratado 32L.L1.L HC.HV1.X (6) 400 act vus, a, um 2 3 2 HC.HV1.L m actiu 2 1 HC.HV1 activo 3 2 HC.L1.L activo 3 2 HC.L1.L 401 arc nus, a, um 2 3 2 HC.HV1.L m arc 2 1 HC. HV1 arcano 3 2 HC.L1.L arcano 3 2 HC.L1.L 402 arch vum, del gr. 2 3 2 HC.HV1.L m arxiu 2 1 HC.HV1 archivo 3 2 HC.L1. L arquivo 3 2 HC.L1.L 403 aest vum [tempus] 2 3 2 HC.HV.L m estiu 2 1 HC.HV1 esto 3 2 HC.L1.L esto 3 2 HC.L1.L 404 germ nus, i, m. 2 3 2 HC.HV1.L m germ 2 1 HC.HV1 hermano 3 2 HC.L1.L irmo 2 1 HC.HV1 405 Silv nus (Sylv-), i, m. 2 3 2 HC.HV1.L m silv 2 1 HC.HV1 silvano 3 2 L. L1.L silvano 3 2 L.L1.L HC.HV1.X (1) 1 406 Centaurus i, m. 2 3 2 HC. HV1.L m centaure 3 2 HC. HV1.L centauro 3 2 HC.HV 1.L centauro 3 2 HV.HV1.L HC.HV1.X (6) 408 nh lus, a, um 2 3 2 HC.HV1.L m anhel 2 1 L.HC1 anhelo 3 2 L. L1.L anelo 3 2 L.L1.L 409 2 3 2 HC.HV1.L m cunyat 2 1 L.HC1 cuado 396 3 2 L.L1.L c unhado 3 2 L.L1.L 410 2 3 2 HC.HV1.L m pecat 2 1 L.HC1 pecado 3 2 L.L1.L pecado 3 2 L.L1.L 411 2 3 2 HC.HV1.L m safir 2 1 L.HC1 za firo 3 2 L.L1.L safira 3 2 L.L1.L 412 2 3 2 HC.HV1.L m terreny 2 1 L.HC1 terreno 3 2 L.L1.L terreno 3 2 L.L1.L 413 4 3 2 HC.HV1.L m aflat 2 1 L.HC1 af lato 3 2 L.L1.L aflato 3 2 L.L1.L 414 4 3 2 HC.HV1.L m grunyit 2 1 L.HC1 gr uido 3 2 L.L1.L grunhido 3 2 L.L1.L HC.HV1.X (1) 415 2 3 2 HC.HV1.L m passiu 2 1 L.HV1 pasivo 2 2 L.L1.L pasivo 2 2 L.L1.L HV.HC1.X (1) 416 d rectus, a, um 2 3 2 HC.HV1.L m dret 1 1 HC1 derecho 3 2 L.L1.L direito 3 2 L.HV1.L HV.HC1.X (1) 417 eun chus, i, m 2 3 2 HV.HV1.L m eunuc 2 1 HV. HC1 eunuco 3 2 HV.L1. L eunuco 3 2 HV.L1.L HV.HC1.X (1) 418 autumnus, i, m. 2 3 2 HV.HC1.L m autumne 3 2 HV.HC1.L otoo 3 2 L.L1.L Outono 3 2 L.L1.L

PAGE 397

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp HV.HC1.X (14) 419 Augustus 2 3 2 HV.HC1.Lm agost 21L.HC1 agosto 32L.HC1.L Agosto 32L.HC1.L 420 b ryllus, i, m. 2 3 2 HV.HC1.Lm beril 21L.HC1 berilo 32L. L1.L berilo 32L.L1.L 421 caementum i, n. 2 3 2 HV.HC1.Lm cement 21L.HC1 cemento 32L.HC1 .L cemento 32L.HV1.L 422 d sertus a, um 2 3 2 HV.HC1.Lm desert 21 L.HC1 desierto 32L.HC1.L deserto 32L.HC1.L 423 d gestus a, um 2 3 2 HV.HC1.Lm digest 21 L.HC1 digesto 32L.HC1.L digesto 32L.HC1.L 424 l mentum i, n. 2 3 2 HV.HC1.Lm lament 21 L.HC1 lamento 32L.HC1 .L lamento 32L.HV1.L 425 l bertus, a, um 2 3 2 HV.HC1.Lm llibert 21 L.HC1 liberto 32 L.HC1.L liberto 3 2 L.HC1.L 426 praefixus, a, um 2 3 2 HV.HC1.L m prefix 2 1 L.HC1 prefijo 3 2 L. L1.L prefixo 3 2 L.HC1.L 427 praetextus, a, um 2 3 2 HV.HC1.L m pretext 2 1 L.HC1 pretexto 3 2 L.HC1.L pretexto 3 2 L.HC1.L 428 pr fectus, a, um 2 3 2 HV.HC1.L m profit 2 1 L. HC1 provecho 3 2 L.HC1.L proveito 3 2 L.HV1.L 429 p pillus, i, m. 2 3 2 HV.HC1.L m pupil 2 1 L.HC1 pupilo 3 2 L. L1.L pupilo 3 2 L.L1.L 430 r sponsum, i, n. 2 3 2 HV.HC1.L m respons 2 1 L.HC1 responso 3 2 L.HC1.L responso 3 2 L.HV1.L 431 t mentum i, n. 2 3 2 HV.HC1.L m toment 2 2 L.HC1 tomento 3 2 L.HC1.L tomento 3 2 L.HV1.L 432 v burnum i, n. 2 3 2 HV.HC1.L m viburn 2 1 L.HC1 viburno 3 2 L.HC1.L viburno 3 2 L.HC1.L HV.HC1.X (4th decl) (4) 433 d cursus s, m. 4 3 2 HV.HC1.L m decurs 2 1 L.HC1 decurso 3 2 L.HC1.L decurso 3 2 L.HC1.L 434 d cessus s, m. 4 3 2 HV.HC1.L m decs 2 1 L.HC1 deceso 3 2 L.L1. L decesso 3 2 L.L1.L 435 d functus s, m. 4 3 2 HV.HC1.L m difunt 2 1 L.HC1 difunto 397 3 2 L.HC1.L defunto 3 2 L.HV1.L 436 pr gressus, s, m. 4 3 2 HV.HC1.L m progrs 2 1 L. HC1 progreso 3 2 L.L1.L progresso 3 2 L.L1.L HV.HC1.LX (8) 437 dictum i, n. 2 3 2 HV.HC1.L m edicte 3 2 L.HC1.L edicto 3 2 L.HC1.L edicto 3 2 L.HC1.L 438 lectrum i, n. 2 3 2 HV.HC1.L m electre 3 2 L.HC1.L electro 3 2 L.HC1.L electro 3 2 L.HC1.L 439 praeceptum, i, n. 2 3 2 HV.HC1.L m precepte 3 2 L.HC1.L precepto 3 2 L.HC1.L preceito 3 2 L.HV1.L 440 praefectus, i, m. 2 3 2 HV.HC1 .L m prefecte 3 2 L.HC1.L prefecto 3 2 L.HC1.L prefeito 3 2 L.HC1.L 441 pr ductus, a, um 2 3 2 HV.HC1.L m producte 3 2 L.HC1.L prod ucto 3 2 L.HC1.L produto 3 2 L.L1.L 442 pr jectus, a, um 2 3 2 HV.HC1.L m projecte 3 2 L. HC1.L proyecto 3 2 L.HC1.L projecto 3 2 L.HC1.L 443 r scriptum, i, n. 2 3 2 HV.HC1.L m rescripte 3 2 L.HC1.L rescripto 3 2 L.HC1.L rescrito 3 2 L.L1.L 444 t nesmos i, m. 2 3 2 HV.HC1.L m tenesme 3 2 L. HC1.L tenesmo 3 2 L.HC1.L tenesmo 3 2 L.HC1.L HV.HC1.LX (4th decl) (2) 445 d fectus s, m. 4 3 2 HV.HC1.L m defecte 3 2 L.HC1.L defecto 3 2 L.HC1.L defeito 3 2 L.HV1.L 446 praecinctus, s, m. 4 3 2 HV.HC1.L m precinte 3 2 L.HC1.L precinto 3 2 L.HC1.L precinto 3 2 L.HV1.L HV.HV1.X (10) 447 c r tus a, um 2 3 2 HV.HV1.L m curat 2 1 L.HC1 curato 3 2 L.L1.L curato 3 2 L.L1.L 448 d cr tum i, n. 2 3 2 HV.HV1.L m decret 2 1 L. HC1 decreto 3 2 L.L1.L decreto 3 2 L.L1.L 449 d v tus, a, um 2 3 2 HV.HV1.L m devot 2 1 L.HC1 devoto 3 2 L. L1.L devoto 3 2 L.L1.L

PAGE 398

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 450 hys pum (hysso-), i, n., = 2 3 2 HV.HV1.Lm hisop 21L.HC1 hisopo 32L.L1.L hissopo 32L.L1.L 451 l g tus, i, m. 2 3 2 HV.HV1.Lm llegat 21L.HC1 legado 32L. L1.L legado 32L.L1.L 452 prael tus, a, um 2 3 2 HV.HV1.Lm prelat 21L.HC1 prelado 32L. L1.L prelado 32L.L1.L 453 qu dr tus a, um 2 3 2 HV.HV1.Lm quadrat 21L.HC1 cuadrado 32L.L1. L quadrado 32L.L1.L 454 s cr tum, i, n. 2 3 2 HV.HV1.L m secret 2 1 L.HC1 secreto 3 2 L.L1. L secreto 3 2 L.L1.L 455 th saurus, i, m., del gr. 2 3 2 HV.HV1.L m tres or 2 1 L.HC1 tesoro 3 2 L. L1.L tesouro 3 2 L.HV1.L 456 v n tum i, n., infl. v n a 2 3 2 HV.HV1.L m vinyet 2 1 L.HC1 viedo 3 2 L.L1.L vinhedo 3 2 L.L1.L HV.HV1.X (4th decl) (4) 457 c n tus s, m. 4 3 2 HV.HV1.L m conat 2 1 L. HC1 conato 3 2 L.L1. L conato 3 2 L.L1.L 458 m g tus, s, m. 4 3 2 HV.HV1.L m mugit 2 1 L. HC1 mugido 3 2 L.L1. L mugido 3 2 L.L1.L 459 pr m tus, s, m. 4 3 2 HV.HV1.L m primat 2 1 L. HC1 primado 3 2 L.L1. L primado 3 2 L.L1.L 460 v g tus s, m. 4 3 2 HV.HV1.L m vagit 2 1 L.HC1 vagido 3 2 L.L1. L vagido 3 2 L.L1.L HV.HV1.X (7) 461 d v nus a, um, adj. 2 3 2 HV.HV1.L m div 2 1 L.HV1 divino 3 2 L.L1.L divino 3 2 L.L1.L 462 h m nus, a, um 2 3 2 HV.HV1.L m hum 2 1 L.HV1 humano 3 2 L.L1. L humano 3 2 L.L1.L 463 M s um, i, n., = 2 3 2 HV.HV1.L m museu 2 1 L.HV1 museo 3 2 L.L1.L museu 2 1 L.HV1 464 n t vus, a, um 2 3 2 HV.HV1.L m natiu 2 1 L.HV1 nativo 398 3 2 L. L1.L nativo 3 2 L.L1.L 2 3 2 HV.HV1.L m pag 2 1 L.HV1 pagano 3 2 L.L1.L pago 2 1 L.HV1 2 3 2 HV.HV1.L m plebeu 2 1 L.HV1 plebeyo 3 2 L.L1.L plebeu 2 1 L.HV1 467 v c nus a, um 2 3 2 HV.HV1.L m ve 2 1 L. HV1 vecino 3 2 L.L1.L vizinho 3 2 L.L1.L HV.HV1.X (1) 468 th saurus, i, m. = 2 3 2 HV.HV1.L m tesaurus 3 2 L.HV1.HC tesauro 3 2 L.HV1.L tesauro 3 2 L.HV1.L HV.HV1.X (1) 469 d tr tus s, m. 4 3 2 HV.HV1.L m detritus 3 2 L.L1. HC detrito 3 2 L.L1.L detrito 3 2 L.L1.L L.HC1.X (1) 470 m gister, tri, m. 2 3 2 L.HC1.L m mestre 2 2 HC1.L maestro 3 2 L.HC1.L mestre 2 2 HC1.L L.HC1.X (26) 471 acanthus,, m. 2 3 2 L.HC1.L m acant 2 1 L. HC1 acanto 3 2 L.L1. L acanto 3 2 L.L1.L 472 amentum 2 3 2 L.HC1.L m ament 2 1 L. HC1 amento 3 2 L.HC1.L amento 3 2 L.HC1.L 473 an llus 2 3 2 L.HC1.L m anell 2 1 L.HC1 anillo 3 2 L.L1.L anel 2 1 L.HC1 474 vernus a, um 2 3 2 L.HC1.L m avern 2 1 L. HC1 averno 3 2 L.HC1.L averno 3 2 L.HC1.L 475 b cillum i, n. 2 3 2 L.HC1.L m bacil 2 1 L.HC1 bacilo 3 2 L. L1.L bacilo 3 2 L.L1.L 476 b tillum, i, n. 2 3 2 L.HC1.L m badil 2 1 L.HC1 badil 2 1 L.HC1 badil 2 1 L.HC1

PAGE 399

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 477 b sextus i, m. 2 3 2 L.HC1.L m bixest 2 1 L.HC1 bisiesto 3 2 L.HC1.L bissexto 3 2 L.HC1.L 478 c pillus i, m. 2 3 2 L.HC1.L m cabell 2 1 L.HC1 cabello 3 2 L. L1.L cabelo 3 2 L.L1.L 479 c tarrhus, i, m. 2 3 2 L.HC1.L m cadarn 2 1 L.HC1 catarro 3 2 L.HC1.L catarro 3 2 L.HC1.L 2 3 2 L.HC1.L m cavall 2 1 L.HC1 c aballo 3 2 L.L1.L cavalo 3 2 L.L1.L 481 c lossus i, m. < 2 3 2 L.HC1.L m cols 2 1 L.HC1 co loso 3 2 L.L1.L colosso 3 2 L.L1.L 482 d urnus a, um 2 3 2 L.HC1.L m dirn 2 1 L.HC1 diurno 2 2 HC1.L diurno 2 2 HC1.L 483 duellum, i, n. 2 3 2 L.HC1.L m due l 2 1 L.HC1 duelo 2 2 L1.L duelo 2 2 L1.L 484 le nchus i, m. < 2 3 2 L.HC1.L m elenc 2 1 L.HC1 elenco 3 2 L.HC1.L elenco 3 2 L.HC1.L 485 fl gellum, i, n. 2 3 2 L.HC1.L m flagel 2 1 L.HC1 flagelo 3 2 L.L1.L flagelo 3 2 L.L1.L 486 l bellus, i, m. 2 3 2 L.HC1.L m libel 2 1 L.HC1 libelo 3 2 L.L1.L libelo 3 2 L.L1.L 487 pr fessus, a, um 2 3 2 L.HC1.L m profs 2 1 L.HC1 profeso 3 2 L.L1. L professo 3 2 L.L1.L 488 qu terni ae, a, adj. pl. 2 3 2 L.HC1.L m quadern 2 1 L.HC1 cuaderno 3 2 L.HC1.L caderno 3 2 L.HC1.L 489 *rep llu 2 3 2 L.HC1.L m reboll 2 1 L.HC1 rebollo 3 2 L.L1.L rebolo 3 2 L.L1.L 490 r versus, a, um 2 3 2 L.HC1.L m revers 2 1 L.HC1 reverso 3 2 L.HC1.L reverso 3 2 L.HC1.L 491 r versus, a, u m 2 3 2 L.HC1.L m revs 2 1 L.HC1 revs 2 1 L.HC1 revs 2 1 L.HC1 492 s gillum, i, n. 2 3 2 L.HC1.L m sege ll 2 1 L.HC1 sello 2 2 L1.L selo 2 2 L1.L 493 s cundus, a, um 2 3 2 L.HC1.L m segon 2 1 L. HC1 segundo 3 2 L.HC1.L segundo 3 2 L.HC1.L 494 s gillum, i, n. 2 3 2 L.HC1.L m sigil 2 1 L.HC1 sigilo 399 3 2 L.L1.L sigilo 3 2 L.L1.L 495 t lentum i, n. = 2 3 2 L.HC1.L m talent 2 1 L.HC1 ta lento 3 2 L.HC1.L talento 3 2 L.HC1.L 496 v tellus i, m. 2 3 2 L.HC1.L m vitel 2 1 L.HC1 vitelo 3 2 L.L1.L vitelo 3 2 L.L1.L L.HC1.X (4th decl) (6) 497 pr cessus, s, m. 4 3 2 L.HC1.L m procs 2 1 L.HC1 proceso 3 2 L.L1.L processo 3 2 L.L1.L 498 r cessus, s, m. 4 3 2 L.HC1.L m recs 2 1 L.HC1 receso 3 2 L.L1.L recesso 3 2 L.L1.L 499 r cursus, s, m. 4 3 2 L.HC1.L m recurs 2 1 L.HC1 recurso 3 2 L.HC1.L recurso 3 2 L.HC1.L 500 r flexus, s, m. 4 3 2 L.HC1.L m reflex 2 1 L.HC1 reflejo 3 1 L.L1.L reflexo 3 2 L.HC1.L 501 r gressus, s, m. 4 3 2 L.HC1.L m regrs 2 1 L.HC1 regreso 3 2 L.L1.L regresso 3 2 L.L1.L 502 t multus s, m. 4 3 2 L.HC1.L m tumult 2 1 L.HC1 tumulto 3 2 L.HC1.L tumulto 3 2 L.HC1.L L.HC1.X (1) 503 tyrannus, i, m., = 2 3 2 L.HC1.L m tir 2 1 L.HV1 tirano 3 2 L.L1.L tirano 3 2 L.L1.L L.HC1.X (10) 504 lumnus, a, um 2 3 2 L.HC1.L m alumne 3 2 L.HC1.L alumno 3 2 L.HC1.L aluno 3 2 L.L1.L 505 c pistrum i, n. 2 3 2 L.HC1.L m cabestre 3 2 L. HC1.L cabestro 3 2 L.HC1.L cabresto 3 2 L.HC1.L 506 c lostrum, ii, n 2 3 2 L.HC1.L m calostre 3 2 L. HC1.L calostro 3 2 L.HC1.L colostro 3 2 L.HC1.L

PAGE 400

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 507 cylindrus dri, m., = 2 3 2 L.HC1.L m cilindre 32L.HC1.L c ilindro 32L.HC1.L cilindro 32L.HC1.L 508 cynismus i, m., = 2 3 2 L.HC1.L m cinisme 32L.HC1.L ci nismo 32L.HC1.L cinismo 32L.HC1.L 509 c rymbus i, m., del gr. 2 3 2 L.HC1.L m corim be 32L.HC1.L corimbo 32L.HC1.L corimbo 32L.HC1.L 510 m nister tri, m. 2 3 2 L.HC1.L m ministre 32 L.HC1.L ministro 32L.HC1.L ministro 32L.HC1.L 511 r gestum, i, n. 2 3 2 L.HC1.L m registre 32L. HC1.L registro 32L.HC1.L registro 32L.HC1.L 512 s pulcrum, i, n. 2 3 2 L.HC1.L m sepulcre 32L. HC1.L sepulcro 32L.HC1.L sepulcro 32L.HC1.L 513 s nister, tra, trum 2 3 2 L.HC1.L m sinistre 32 L.HC1.L siniestro 32L.HC1.L sinestro 32L.HC1.L L.HC1.X (4th decl) (2) 514 prospectus, s, m. 4 3 2 L.HC1.L m prospecte 3 2 L.HC1.L prospecto 3 2 L.HC1.L prospecto 3 2 L.HC1.L 515 r tractus, s, m. 4 3 2 L.HC1.L m retracte 3 2 L.HC1.L retracto 3 2 L.HC1.L retracto 3 2 L.HC1.L L.HV1.X (1) 516 c r brum, i, n. 2 3 2 L.HV1.L m cervell 2 1 HC.HC1 cerebro 3 2 L. L1.L crebro 3 3 L1.L.L L.HV1.X (22) 517 pr cus, a, um 2 3 2 L.HV1.L m abric 2 1 L.HC1 abrigo 3 2 L. L1.L abrigo 3 2 L.L1.L 518 am cus 2 3 2 L.HV1.L m amic 2 1 L.HC1 amigo 3 2 L.L1.L amigo 3 2 L.L1.L 519 sylum, i, n., = 2 3 2 L.HV1.L m asil 2 1 L.HC1 asilo 3 2 L.L1.L asilo 3 2 L.L1.L 520 be tus 2 3 2 L.HV1.L m beat 2 1 L. HC1 beato 400 3 2 L.L1. L beato 3 2 L.L1.L 521 c m lus, i, m.= 2 3 2 L.HV1.L m camell 2 1 L.HC1 camello 3 2 L.L1.L camelo 3 2 L.L1.L 522 c l nus i, m. 2 3 2 L.HV1.L m colon 2 1 L.HC1 colono 3 2 L.L1.L colono 3 2 L.L1.L 523 d c rum i, n. 2 3 2 L.HV1.L m decor 2 1 L.HC1 decoro 3 2 L. L1.L decoro 3 2 L.L1.L 524 ph bus i, m., = 2 3 2 L.HV1.L m efeb 21L.HC1 efebo 32L.L1.L efebo 32L.L1.L 525 f t rus, a, um 2 3 2 L.HV1.L m futur 21 L.HC1 futuro 32L.L1.L futuro 32L.L1.L 526 m r tus, i, m. 2 3 2 L.HV1.L m marit 2 1 L.HC1 marido 3 2 L.L1.L marido 3 2 L.L1.L 527 m n tus, a, um 2 3 2 L.HV1.L m menut 2 1 L.HC1 menudo 3 2 L. L1.L mido 3 2 L.L1.L 528 m n tus, a, um 2 3 2 L.HV1.L m minut 2 1 L.HC1 minuto 3 2 L. L1.L minuto 3 2 L.L1.L 529 n v tus a, um 2 3 2 L.HV1.L m nevat 2 1 L.HC1 nevado 3 2 L. L1.L nevado 3 2 L.L1.L 530 p r tus, a, um 2 3 2 L.HV1.L m perit 2 1 L.HC1 perito 3 2 L. L1.L perito 3 2 L.L1.L 531 r c mus, i, m. 2 3 2 L.HV1.L m ram 2 1 L.HC1 racimo 3 2 L.L1.L racimo 3 2 L.L1.L 532 r l tus, a, um 2 3 2 L.HV1.L m relat 2 1 L.HC1 relato 3 2 L.L1.L relato 3 2 L.L1.L 533 r s tus, a, um 2 3 2 L.HV1.L m rosat 2 1 L.HC1 rosado 3 2 L.L1.L rosado 3 2 L.L1.L 534 s cr tus, a, um 2 3 2 L.HV1.L m sagrat 2 1 L.HC1 sagrado 3 2 L.L1. L sagrado 3 2 L.L1.L 535 s b cus, i, f. 2 3 2 L.HV1.L m sac 2 1 L.HC1 saco 3 2 L. L1.L saco 3 2 L.L1.L 536 s l tus a, um 2 3 2 L.HV1.L m solut 2 1 L.HC1 soluto 3 2 L.L1.L soluto 3 2 L.L1.L 537 tr b tum i, n. 2 3 2 L.HV1.L m tribut 2 1 L.HC1 tributo 3 2 L.L1. L tributo 3 2 L.L1.L

PAGE 401

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 538 pyl rus, i, m., = 2 3 2 L.HV1.L m plor 22L1.HC ploro 33L1.L.L piloro 32L.L1.L L.HV1.X (4th decl) (4) 539 b sus, s, m. 4 3 2 L.HV1.L m abs 21 L.HC1 abuso 32L.L1. L abuso 32L.L1.L 540 m tus, s, m. 4 3 2 L.HV1.L m meat 21L. HC1 meato 32L.L1. L meato 32L.L1.L 541 occ sus, s, m. 4 3 2 L.HV1.L m ocs 21 L.HC1 ocaso 32L. L1.L ocaso 32L.L1.L 542 s n tus, s, m. 4 3 2 L.HV1.L m senat 21L. HC1 senado 32L.L1. L senado 32L.L1.L L.HV1.X (prothesis) (1) 543 st t tum, i, n. 2 3 2 L.HV1.L m estatut 31L. HC1 estatuto 4 2 L.L1.L estatuto 4 2 L.L1.L L.HV1.X (14) 544 c n nus a, um 2 3 2 L.HV1.L m can 2 1 L.HV1 canino 3 2 L. L1.L canino 3 2 L.L1.L 545 c m num (cym), i, n., = 2 3 2 L.HV1.L m com 2 1 L.HV1 comino 3 2 L.L1.L cominho 3 2 L.L1.L 546 d t vus, i, m. 2 3 2 L.HV1.L m datiu 2 1 L.HV1 dativo 3 2 L. L1.L dativo 3 2 L.L1.L 547 d c nus i, m. 2 3 2 L.HV1.L m deg 2 1 L.HV1 decano 3 2 L. L1.L decano 3 2 L.L1.L 548 iei num 2 3 2 L.HV1.L m jej 2 1 L.HV 1 yeyuno 3 2 L.L1.L jejum 2 1 L.HV1 549 Lyc um (alt. Lyc um), i, n., = 2 3 2 L.HV1.L m liceu 2 1 L.HV1 liceo 3 2 L.L1.L liceu 2 1 L.HV1 550 m r nus, a, um 2 3 2 L.HV1.L m mar 2 1 L.HV1 marino 3 2 L.L1. L marinho 3 2 L.L1.L 551 m l num, i, n. 2 3 2 L.HV1.L m mol 2 1 L.HV1 molino 401 3 2 L. L1.L moinho 3 2 L.L1.L 552 *patr nus 2 3 2 L.HV1.L m padr 2 1 L.HV1 padrino 3 2 L.L1.L padrinho 3 2 L.L1.L 553 p tr nus, i, m. 2 3 2 L.HV1.L m padr 2 1 L.HV1 padrn 2 1 L.HC1 padro 2 1 L.HV1 554 p tr nus, i, m. 2 3 2 L.HV1.L m patr 2 1 L.HV1 patrn 2 1 L. HC1 patro 2 1 L.HV1 555 s p num, i, n. 2 3 2 L.HV1.L m sup 2 1 L.HV1 supino 3 2 L. L1.L supino 3 2 L.L1.L 556 tr b nus i, m. 2 3 2 L.HV1.L m trib 2 1 L.HV1 tribuno 3 2 L.L1. L tribuno 3 2 L.L1.L 557 tr paeum o tr phaeum, i, n., = 2 3 2 L.HV1.L m trofeu 2 1 L.HV1 trofeo 3 2 L.L1.L trofu 3 2 L.HV1 L.HV1.X (prothesis) (1) 558 sc l nus, a, um = 2 3 2 L.HV1.L m escal 3 1 L.HV1 escaleno 4 2 L.L1.L escaleno 4 2 L.L1.L L.HV1.X (1) 559 h tus, s, m. 2 3 2 L.HV1.L m hiatus 3 2 L. L1.HC hiato 3 2 L.L1. L hiato 3 2 L.L1.L L.HV1.X (1) 560 th trum i, n., = 2 3 2 L.HV1.L m teatre 3 2 L.L1.L teatro 3 2 L.L1.L teatro 3 2 L.L1.L L.L1.X (1) 561 p r plus, i, m., = 2 3 2 L.L1.L m periple 3 2 L.L1.L per iplo 3 2 L.L1.L priplo 3 3 L1.L.L

PAGE 402

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp HC1.L.X (7) 562 baln um i. n. = 2 3 3 HC1.L.L m bany 11HC1 bao 22L1.L banho 22L1.L 563 bracch um, ii, n. 2 3 3 HC1.L.L m bra 11HC1 brazo 22L1.L brao 22L1.L 564 lint um, i, n. 2 3 3 HC1.L.L m llen 11HC1 lienzo 22HC1.L leno 22HC1.L 565 mall us, i, m. 2 3 3 HC1.L.L m mall 11HC1 mallo 22L1.L malho 22L1.L 566 Mart us, a, um 2 3 3 HC1.L.L m mar 11HC1 marzo 22HC1.L maro 22HC1.L 567 mort us, a, um 2 3 3 HC1.L.L m mort 11HC1 muerto 22HC1.L morto 22HC1.L 568 tert us a, um, adj. 2 3 3 HC1.L.L m ter 11HC1 tercio 22 HC1.L tero 2 2 HC1.L HC1.L.X (1) 569 tymp num, i, n., = 2 3 3 HC1.L.L m timp 2 1 HC.HV1 tmpano 3 3 HC1.L.L tmpano 3 3 HV1.L.L HC1.L.X (1) 570 burr cus, i, m. 2 3 3 HC1.L.L m borrec 2 1 L.HC1 borrico 3 2 L.L1.L borrego 3 2 L.L1.L HC1.L.X (27) 571 ang lus, i, m., = 2 3 3 HC1.L.L m ngel 2 2 HC1.HC ngel 2 2 HC1.HC anjo 2 2 HC1.L 572 bals mum i, n., = 2 3 3 HC1.L.L m blsam 2 2 HC1.HC blsamo 3 3 HC1.L.L blsamo 3 3 HC1.L.L 573 canth rus i, m., = 2 3 3 HC1.L.L m cnter 2 2 HC1.HC cntaro 3 3 HC1.L.L cntaro 3 3 HC1.L.L 574 cant cum i, n. 2 3 3 HC1.L.L m cntic 2 2 HC1 .HC cntico 402 3 3 HC1.L.L cntico 3 3 HC1.L.L 575 cymb lum i, n., = 2 3 3 HC1.L.L m cmbal 2 2 HC1.HC cmbalo 3 3 HC1.L.L cmbalo 3 3 HC1.L.L 576 cing lum i, n. 2 3 3 HC1.L.L m cngol 2 2 HC1.HC cngulo 3 3 HC1.L.L cngulo 3 3 HC1.L.L 577 comp tus i, m. 2 3 3 HC1.L.L m cmput 2 2 HC1.HC cmputo 3 3 HC1.L.L cmputo 3 3 HC1.L.L 578 emb lus i, m., = 3 3 HC1.L.L m mbol 2 2 HC1.HC mbolo 3 3 HC1.L.L mbolo 3 3 HC1.L.L 579 pharm cus, i, m., = 2 3 3 HC1.L.L m frmac 2 2 HC1.HC f rmaco 3 3 HC1.L.L frmaco 3 3 HC1.L.L 580 flosc lus, i, m. 2 3 3 HC1.L.L m flscul 2 2 HC1.HC flsculo 3 3 HC1.L.L flsculo 3 3 HC1.L.L 581 Phosph rus, i, m. = 2 3 3 HC1.L.L m fsfor 2 2 HC1.HC f sforo 3 3 HC1.L.L fsforo 3 3 HC1.L.L 582 inc bus, i, m. 2 3 3 HC1.L.L m ncub 2 2 HC1 .HC ncubo 3 3 HC1.L.L ncubo 3 3 HC1.L.L 583 max mus, a, um 2 3 3 HC1.L.L m mxim 2 2 HC1 .HC mximo 3 3 HC1.L.L mximo 3 3 HC1.L.L 584 org num, i, n. = 2 3 3 HC1.L.L m rgan 22HC1.HC rgano 33HC1.L.L rgo 22HC1.HV 585 osc lum, i, n. 2 3 3 HC1.L.L m scul 22HC1. HC sculo 33HC1.L.L sculo 33HC1.L.L 586 parv lus, a, um 2 3 3 HC1.L.L m prvul 22HC1.HC prvulo 33HC1.L.L prvulo 33HC1.L.L 587 pend lus, a, um 2 3 3 HC1.L.L m pndul 22HC1.HC pndulo 33HC1.L.L pndulo 33HC1.L.L

PAGE 403

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 588 plast cus, a, um = 2 3 3 HC1.L.L m plstic 2 2 HC1.HC pls tico 3 3 HC1.L.L pl stico 3 3 HC1.L.L 589 prox mus, a, um 2 3 3 HC1.L.L m prxim 2 2 HC1.HC prjimo 3 3 L1.L.L prximo 3 3 HC1.L.L 590 pulp tum, i, n. 2 3 3 HC1.L.L m plpit 2 2 HC1.HC plpito 3 3 HC1.L.L plpito 3 3 HC1.L.L 591 symb lus, i, m. ; symb lum, i, n. = o 2 3 3 HC1.L.L m smbol 2 2 HC1.HC smbolo 3 3 HC1.L.L smbolo 3 3 HC1.L.L 592 synd cus, I, m., = 2 3 3 HC1.L.L m sndic 2 2 HC1.HC sndi co 3 3 HC1.L.L sndico 3 3 HC1.L.L 593 techn cus I, m., = 2 3 3 HC1.L.L m tcnic 2 2 HC1.HC t cnico 3 3 HC1.L.L tcnico 3 3 HC1.L.L 594 torc lum I, n 2 3 3 HC1.L.L m trcul 2 2 HC1 .HC trculo 3 3 HC1.L.L trculo 3 3 HC1.L.L 595 tox cum I, n. 2 3 3 HC1.L.L m txic 2 2 HC1. HC txico 3 3 HC1.L.L txico 3 3 HC1.L.L 596 ult mus a, um 2 3 3 HC1.L.L m ltim 2 2 HC1 .HC ltimo 3 3 HC1.L.L ltimo 3 3 HC1.L.L 597 Vand lus, -i 2 3 3 HC1.L.L m vndal 2 2 HC1.HC vndalo 3 3 HC1.L.L vndalo 3 3 HC1.L.L HC1.L.X (4th decl) (3) 598 amb tus s, m. 4 3 3 HC1.L.L m mbit 2 2 HC1.HC mbito 3 3 HC1.L.L mbito 3 3 HC1.L.L 599 ex tus, s, m. 4 3 3 HC1.L.L m xit 2 2 HC1 .HC xito 3 3 HC1.L.L xito 3 3 HC1.L.L 600 port cus, s, f 4 3 3 HC1.L.L m prtic 2 2 HC1.HC prtico 3 3 HC1.L.L prtico 3 3 HC1.L.L HC1.L.X (prothesis) (2) 601 scept cus, i, m. / Scept ci, rum, m., = 2 3 3 HC1.L.L m escptic 3 2 HC1.HC e scptico 403 4 3 HC1.L.L cptico 3 3 HC1.L.L 602 scand lum, i, n. = 2 3 3 HC1.L.L m escndol 3 2 HC1.HC escndalo 4 3 HC1.L.L escndalo 4 3 HC1.L.L HC1.L.X (1) 603 conc vus a, um 2 3 3 HC1.L.L m cncau 2 2 HC1.HV cncavo 3 3 HC1.L.L cncavo 3 3 HC1.L.L HC1.L.X (14) 604 ang lus, i, m. = 2 3 3 HC1.L.L m angle 2 2 HC1.L ngulo 3 3 HC1.L.L ngulo 3 3 HC1.L.L 605 bronch a rum, n., = 2 3 3 HC1.L.L m bronqui 2 2 HC1.L bronquio 2 2 HC1.L brnquio 2 2 HC1.L 606 Sci Lat. calcium < calx, calcis 2 3 3 HC1.L.L m calci 2 2 HC1.L calcio 2 2 HC1.L clcio 2 2 HC1.L 607 circ lus i, m. 2 3 3 HC1.L.L m cercle 2 2 HC1.L crculo 3 3 HC1.L.L crculo 3 3 HC1.L.L 608 comp tus i, m. 2 3 3 HC1.L.L m conte 2 2 HC1.L cuento 2 2 HC1.L conto 2 2 HC1.L 609 orph nus, i, m., = 2 3 3 HC1.L.L m orfe 2 2 HC1.L hurfano 3 3 HC1.L.L rfo 2 2 HC1.HV 610 ind cum, i, n. 2 3 3 HC1.L.L m indi 2 2 HC1 .L ndigo 3 3 HC1.L.L ndigo 3 3 HC1.L.L 611 masc lus, a, um 2 3 3 HC1.L.L m mascle 2 2 HC1.L macho 2 2 L1.L macho 2 2 L1.L

PAGE 404

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 612 menstr us, a, um 2 3 3 HC1.L.L m menstru 22HC1.L menstruo 22HC1.L mnstruo 22HC1.L 613 myst cus, a, um, adj., = 2 3 3 HC1.L.L m mstic 22HC1.L m stico 3 3 HC1.L.L mstico 3 3 HC1.L.L 614 musc lus, i, m. 2 3 3 HC1.L.L m muscle 2 2 HC1 .L msculo 3 3 HC1.L.L msculo 3 3 HC1.L.L 615 term nus i, m. 2 3 3 HC1.L.L m terme 2 2 HC1.L trmino 3 3 HC1.L.L termo 2 2 HC1.L 616 vinc lum o vinclum i, n. 2 3 3 HC1.L.L m vincle 2 2 HC1.L vnculo 3 3 HC1.L.L vnculo 3 3 HC1.L.L 617 gangl on, i, n. 2 3 3 HC1.L.L m gangli 2 2 HC1.L ganglio 2 2 HC1.L gnglio 2 2 HC1.L HC1.L.X (6) 618 arb ter, tri, m. 2 3 3 HC1.L.L m rbitre 3 3 HC1.L .L rbitro 3 3 HC1.L.L rbitro 3 3 HC1.L.L 619 cent plus, a, um 2 3 3 HC1.L.L m cntuple 3 3 HC1.L .L cntuplo 3 3 HC1.L.L cntuplo 3 3 HC1.L.L 620 Ex dus, i, f. 2 3 3 HC1.L.L m xode 3 3 HC1.L.L xodo 3 3 HC1.L.L xodo 3 3 HC1.L.L 621 quint plus 2 3 3 HC1.L.L m quntuple 3 3 HC1.L.L quntuplo 3 3 HC1.L.L quntuplo 3 3 HC1.L.L 622 sept plum, i, n. 2 3 3 HC1.L.L m sptuple 3 3 HC1.L.L sptuplo 3 3 HC1.L.L sptuplo 3 3 HC1.L.L 623 sext plus 2 3 3 HC1.L.L m sxtuple 3 3 HC1.L.L sxtuplo 3 3 HC1.L.L sxtuplo 3 3 HC1.L.L HC1.L.X (4th decl) (1) 624 imp tus, s, m. 4 3 3 HC1.L.L m mpetu 3 3 HC1.L .L mpetu 3 3 HC1.L.L mpeto 3 3 HC1.L.L HC1.L.X (3) 625 Att cus a, um = 2 3 3 HC1.L.L m tic 2 2 L1.HC tico 4043 3 L1.L.L tico 3 3 L1.L.L 626 class cus a, um 2 3 3 HC1.L.L m clssic 2 2 L1.HC clsico 3 3 L1.L.L clssico 3 3 L1.L.L 627 *succ bus, cf. inc bus 2 3 3 HC1.L.L m scub 2 2 L1.HC scubo 3 3 L1.L.L scubo 3 3 L1.L.L HC1.L.X (1) 628 frax nus, i, f. 2 3 3 HC1.L.L m freixe 2 2 L1.L fresno 2 2 HC1.L freixo 2 2 L1.L HV1.HV.X (1) 629 d lum or -on, i, n., = 2 3 3 HV1.HV.L m dol 2 2 L1.HC dolo 3 3 L1.L.L dolo 3 3 L1.L.L HV1.L.X (9) 630 b s um ii, n., 2 3 3 HV1.L.L m bes 1 1 HC1 beso 2 2 L1.L beijo 2 2 HV1.L 631 p p lus i, f. 2 3 3 HV1.L.L m xop 1 1 HC1 chopo 2 2 L1.L choupo 2 2 HV1.L 632 fr g dus, a, um 2 3 3 HV1.L.L m fred 1 1 HC1 fro 2 2 L1.L frio 2 2 L1.L 633 gr c lus (gracc-), i, m. 2 3 3 HV1.L.L m grall 1 1 HC1 grajo 2 2 L1.L gralho 2 2 L1.L 634 f l us, ii, m. 2 3 3 HV1.L.L m fill 1 1 HC1 hijo 2 2 L1.L filho 2 2 L1.L 635 l cus i, m. 2 3 3 HV1.L.L m laic 1 1 HC1 laico 2 2 HV1.L laico 2 2 HV1.L 636 l cus, a, um = 2 3 3 HV1.L.L m llec 1 1 HC1 lego 2 2 L1.L leigo 2 2 HV1.L 367 l c us, ii, m. 2 3 3 HV1.L.L m llu 1 1 HC1 lucio 2 2 L1.L lcio 2 2 L1.L 638 p d tum, i, n. 2 3 3 HV1.L.L m pet 1 1 HC1 pedo 2 2 L1.L peido 2 2 HV1.L

PAGE 405

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp HV1.L.X (1) 639 cl r cus i, m., = 2 3 3 HV1.L.L m clergue 22HC1.L cl rigo 33L1.L.L clrigo 33L1.L.L HV1.L.X (4) 640 caust cus a, um = 2 3 3 HV1.L.L m custic 22HV1.HC cus tico 33HV1.L.L custico 33HV1.L.L 641 l d num o l d num, i, n., = 2 3 3 HV1.L.L m ludan 22HV1.HC l udano 33HV1.L.L ludano 33HV1.L.L 642 naufr gus, a, um 2 3 3 HV1.L.L m nufrag 22 HV1.HC nufrago 33HV1.L. L nufrago 33HV1.L.L 643 naut lus o naut l s, i, m., = 2 3 3 HV1.L.L m nutil 2 2 HV1.HC nautilo 3 2 HV.L1.L nutilo 3 3 HV1.L.L HV.L.X (1) 644 h l tus, s, m. 4 3 3 HV1.L.L m hlit 2 2 L1.HC hlito 3 3 L1.L. L hlito 3 3 L1.L.L HV.L.L (prothesis) (1) 645 sp r tus, s, m. 4 3 3 HV1.L.L m es perit 3 1 L.HC1 espritu 4 3 L1.L.L esprito 4 3 L1.L.L HV1.L.X (2) 646 aur us i, m. 2 3 3 HV1.L.L m auri 2 2 HV 1.L ureo 3 3 HV1.L. L ureo 3 3 HV1.L.L 647 glauc on, ii, n., = 2 3 3 HV1.L.L m glauci 2 2 HV1.L glaucio 2 2 HV1.L glucio 2 2 HV1.L HV1.L.X (18) 648 b b lus i, m.= 2 3 3 HV1.L.L m bbal 2 2 L1.HC bbalo 405 3 3 L1.L.L bbalo 3 3 L1.L.L 649 b f lus i, m. 2 3 3 HV1.L.L m bfal 2 2 L1.HC bfalo 3 3 L1.L. L bfalo 3 3 L1.L.L 650 c m cus i, m. = 2 3 3 HV1.L.L m cmic 2 2 L1.HC cmico 3 3 L1.L.L cmico 3 3 L1.L.L 651 cr d t um i, n. 2 3 3 HV1.L.L m crdit 2 2 L1.HC crdito 3 3 L1.L.L crdito 3 3 L1.L.L 652 d b tum i, n. 2 3 3 HV1.L.L m dbit 2 2 L1.HC dbito 3 3 L1.L. L dbito 3 3 L1.L.L 653 aem lus a, um, cf. and 2 3 3 HV1.L.L m mul 2 2 L1.HC mulo 3 3 L1.L.L mulo 3 3 L1.L.L 654 gr n lum, i, n. 2 3 3 HV1.L.L m grnul 2 2 L1.HC grnulo 3 3 L1.L. L grnulo 3 3 L1.L.L 655 l qu dus a, um 2 3 3 HV1.L.L m lquid 2 2 L1.HC lquido 3 3 L1. L.L lquido 3 3 L1.L.L 656 m s cus, i, m. 2 3 3 HV1.L.L M msic 2 2 L1.HC msico 3 3 L1.L.L msico 3 3 L1.L.L 657 n d lus i, m. 2 3 3 HV1.L.L m ndul 2 2 L1.HC ndulo 3 3 L1. L.L ndulo 3 3 L1.L.L 658 ov lu m < vum, i, n. (Sci Lat) 2 3 3 HV1.L.L m vul 2 2 L1.HC vulo 3 3 L1.L.L vulo 3 3 L1.L.L 659 p b lum, i, n. 2 3 3 HV1.L.L m pbul 2 2 L1.HC pbulo 3 3 L1. L.L pbulo 3 3 L1.L.L 660 p lyp s, i, m. = 2 3 3 HV1.L.L m plip 2 2 L1.HC plipo 3 3 L1.L.L plipo 3 3 L1.L.L 661 p m lum, i, n. 2 3 3 HV1.L.L m pmul 2 2 L1.HC pmulo 3 3 L1. L.L pmulo 3 3 L1.L.L

PAGE 406

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 662 p bl c s, i, m. 2 3 3 HV1.L.L m pblic 22L1.HC pblico 33L1.L. L pblico 33L1.L.L 663 r g lus, i, m. 2 3 3 HV1.L.L m rgul 22L1.HC rgulo 33L1.L. L rgulo 33L1.L.L 664 s s mum, i, n., = 2 3 3 HV1.L.L m ssam 22L1.HC s samo 33L1.L.L ssamo 33L1.L.L 665 tr glyphus i, m., = 2 3 3 HV1.L.L m trglif 2 2 L1.HC tr iglifo 3 2 L.L1.L trglifo 3 3 L1.L.L HV1.L.X (prothesis) (1) 666 scr p lus, i, m. 2 3 3 HV1.L.L m escrpol 3 2 L1.HC escrpulo 4 3 L1. L.L escrpulo 4 3 L1.L.L HV1.L.X (13) 667 caes us a, um 2 3 3 HV1.L.L m cesi 2 2 L1.L cesio 2 2 L1.L csio 2 2 L1.L 668 c r us i, m. 2 3 3 HV1.L.L m ci ri 2 2 L1.L cirio 2 2 L1.L crio 2 2 L1.L 669 c l cus a, um, = 2 3 3 HV1.L.L m clic 2 2 L1.L c lico 3 3 L1.L.L clica (f.) 3 3 L1.L.L 670 l l um, i, n. 2 3 3 HV1.L.L m lliri 2 2 L1. L lirio 2 2 L1.L lrio 2 2 L1.L 671 t um, i, n. 2 3 3 HV1.L.L m oci 2 2 L1.L ocio 2 2 L1.L cio 2 2 L1.L 672 praed um, i, n. 2 3 3 HV1.L.L m predi 2 2 L1. L predio 2 2 L1.L prdio 2 2 L1.L 673 praem um, ii, n. 2 3 3 HV1.L.L m premi 2 2 L1.L premio 2 2 L1.L prmio 2 2 L1.L 674 pr pr -um, i, n. 2 3 3 HV1.L.L m propi 2 2 L1.L propio 2 2 L1.L prprio 2 2 L1.L 675 saec lum, i, n. 2 3 3 HV1.L.L m segle 2 2 L1.L siglo 2 2 L1.L sculo 3 3 L1.L.L 676 s m us, ii, m. 2 3 3 HV1.L.L m simi 2 2 L1.L simio 406 2 2 L1.L smio 2 2 L1.L 677 taed um ii, n. 2 3 3 HV1.L.L m tedi 2 2 L1.L tedio 2 2 L1.L tdio 2 2 L1.L 678 tr d um i, n. 2 3 3 HV1.L.L m tr idu 2 2 L1.L triduo 2 2 L1.L trduo 2 2 L1.L 679 v tr us a, um, 2 3 3 HV1.L.L m vidre 2 2 L1.L vidrio 2 2 L1.L vidro 2 2 L1.L HV1.L.X (1) 680 qu dr plus, a, um 2 3 3 HV1.L.L m qudruple 3 3 L1.L.L cudruple 3 3 L1.L.L qudruplo 3 3 L1.L.L L1.L.X (18) 681 al um, i, n. 2 3 3 L1.L.L m all 1 1 HC1 ajo 2 2 L1.L alho 2 2 L1.L 682 VL *col pus
PAGE 407

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 691 m l um, ii, n. 2 3 3 L1.L.L m mill 11HC1 mijo 22L1.L milho 22L1.L 692 pl c tum, i, n. 2 3 3 L1.L.L m plet 11 HC1 pleito 22HV1.L pleito 22HV1.L 693 p lyp s, i, m. 2 3 3 L1.L.L m polp 1 1 HC1 pulpo 2 2 HC1.L polvo 2 2 HC1.L 694 p s tus, a, um 4 3 3 L1.L.L m post 1 1 HC1 puesto 2 2 HC1.L posto 2 2 HC1.L 695 p d um, ii, n., = 2 3 3 L1.L.L m puig (dial.pui) 1 1 HC1 poyo 2 2 L1.L poio 2 2 L1.L 696 r d us, ii, m. 2 3 3 L1.L.L m raig 1 1 HC1 rayo 2 2 L1.L raio 2 2 L1.L 697 trib lum 2 3 3 L1.L.L m trill 1 1 HC1 trillo 2 2 L1.L trilho 2 2 L1.L 698 c l us, i, m. 2 3 3 L1.L.L m ull 1 1 HC1 ojo 2 2 L1.L olho 2 2 L1.L L1.L.X (4th decl) (1) 699 c tus, s, m. 4 3 3 L1.L.L m coit 1 1 HC1 coito 2 2 HV1.L coito 2 2 HV1.L L1.L.X (prothesis) (1) 700 sp c lum, i, n. 2 3 3 L1.L.L m espill 2 1 HC.HC1 espejo 3 2 L1.L espelho 3 2 L1.L L1.L.X (prothesis) (1) 701 sp t um, ii, n. 2 3 3 L1.L.L m espai 2 1 HV1 espacio 3 2 L1.L espao 3 2 L1.L L1.L.X (4) 702 b d us a, um 2 3 3 L1.L.L m bai (cat.ant. baig, bai del castell?) 1 1 HV1 bayo 2 2 L1.L baio 2 2 L1.L 703 p t us, i, m. 2 3 3 L1.L.L m pou 1 1 HV1 pozo 2 2 L1.L poo 2 2 L1.L 704 pr t um, ii, n. 2 3 3 L1.L.L m preu 1 1 HV1 precio 407 2 2 L1.L preo 2 2 L1.L 705 s l dus, i, m. 2 3 3 L1.L.L m sou 1 1 HV1 sueldo 2 2 HC1.L soldo 2 2 HC1.L L1.L.X (2) 706 cath tus, i, f., = 2 3 3 L1.L.L m catet 2 1 L.HC1 cateto 3 2 L.L1.L cateto 3 2 L.L1.L 707 p pyrus, i, m. and f., and p py-rum, i, n., = 2 3 3 L1.L.L m papir 21L.HC1 pap iro 32L.L1.L papiro 32L.L1.L L1.L.X (2) 708 LL trep num (< tryp num < ) 2 3 3 L1.L.L m trep 21L.HV1 trpano 33L1.L.L trpano 33L1.L.L 709 r c nus, i, m. 2 3 3 L1.L.L m ric 21 L.HV1 ricino 3 2 L.L1. L rcino 3 3 L1.L.L L1.L.X (1) 710 fl dus, a, um 2 3 3 L1.L.L m fluid 2 1 L1.HC fluido 2 2 L1.L fluido 2 2 L1.L L1.L.X (3) 711 dec mus 2 3 3 L1.L.L m delme 2 2 HC1.L diezmo 2 2 HC1.L dzimo 3 3 L1.L.L 712 m n chus, i, m., = 2 3 3 L1.L.L m monjo 2 2 HC1.L monje 2 2 HC1.L monge 2 2 HC1.L 713 r t lus, i, m. 2 3 3 L1.L.L m rotlle 2 2 HC1.L rollo 2 2 L1.L rolo 2 2 L1.L L1.L.X (38)

PAGE 408

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 714 b cus, i, m. 2 3 3 L1.L.L m bac 22L1.HC baco 33L1. L.L baco 33L1.L.L 715 ac rus, m. 2 3 3 L1.L.L m car 22L1. HC caro 33L1.L.L caro 33L1.L.L 716 ac dus 2 3 3 L1.L.L m cid 22L1.HC cido 33L1.L.L cido 33L1.L.L 717 an mus, y este del gr. 2 3 3 L1.L.L m ni m 22L1.HC nimo 33L1.L.L nimo 33L1.L.L 718 t mus a, um, 2 3 3 L1.L.L m tom 22L1.HC tomo 33L1.L. L tomo 33L1.L.L 719 b c lum i, n. 2 3 3 L1.L.L m bcul 22L1.HC bculo 33L1.L. L bculo 33L1.L.L 720 b l nus i, f., rarely m. = 2 3 3 L1.L.L m blan 22L1.HC blano 33L1.L.L blano 33L1.L.L 721 big mus [vo x hibrida, bis] 2 3 3 L1.L.L m bgam 2 2 L1.HC b gamo 3 3 L1.L.L b gamo 3 3 L1.L.L 722 c l mus i, m. 2 3 3 L1.L.L m clam 2 2 L1.HC clamo 3 3 L1.L.L clamo 3 3 L1.L.L 723 cr t cus a, um, adj. 2 3 3 L1.L.L m crtic 2 2 L1.HC crtico 3 3 L1.L.L crtico 3 3 L1.L.L 724 cr t lum i, n. 2 3 3 L1.L.L m crtal 2 2 L1.HC crtalo 3 3 L1.L. L crtalo 3 3 L1.L.L 725 c m lus i, m 2 3 3 L1.L.L m cmul 2 2 L1. HC cmulo 3 3 L1.L.L cmulo 3 3 L1.L.L 726 d c mus, a, um 2 3 3 L1.L.L m dcim 2 2 L1.HC dcimo 3 3 L1.L. L dcimo 3 3 L1.L.L 727 d g tus i, m. 2 3 3 L1.L.L m dgit 2 2 L1.HC dgito 3 3 L1.L. L dgito 3 3 L1.L.L 728 h b nus ( b nus), i. f. = 2 3 3 L1.L.L m eben 2 2 L1.HC bano 3 3 L1.L.L bano 3 3 L1.L.L 729 tymon, i, n. = 2 3 3 L1.L.L m tim 22L1.HC timo 33L1.L. L timo 33L1.L.L 730 f m lus, i, m. 2 3 3 L1.L.L m fmul 2 2 L1.HC fmulo 408 3 3 L1.L.L fmulo 3 3 L1.L.L 731 phys cus, i, m. = 2 3 3 L1.L.L m fsic 2 2 L1.HC fsico 3 3 L1.L.L fsico 3 3 L1.L.L 732 gl b lus, i, m. 2 3 3 L1.L.L m glbul 2 2 L1.HC glbulo 3 3 L1.L. L glbulo 3 3 L1.L.L 733 m rus (h m rus), i, m. 2 3 3 L1.L.L m hmer 2 2 L1.HC hmero 3 3 L1. L.L mero 3 3 L1.L.L 734 l g cus, a, um, adj., = 2 3 3 L1.L.L m lgic 2 2 L1.HC lgi co 3 3 L1.L.L lgico 3 3 L1.L.L 735 man c us 2 3 3 L1.L.L m mnec 2 2 L1.HC mango 2 2 HC1.L mango 2 2 HC1.L 736 m r tum, i, n. 2 3 3 L1.L.L m mrit 2 2 L1.HC mrito 3 3 L1.L.L mrito 3 3 L1.L.L 737 m d lus, i, m. 2 3 3 L1.L.L m mdul 2 2 L1.HC mdulo 3 3 L1.L. L mdulo 3 3 L1.L.L 738 p lus, i, m. 2 3 3 L1.L.L m pal 2 2 L1.HC palo 3 3 L1.L. L palo 3 3 L1.L.L 739 pan cus, a, um = 2 3 3 L1.L.L m pnic 2 2 L1.HC pn ico 3 3 L1.L.L pnico 3 3 L1.L.L 740 p l gus, i, n. = 2 3 3 L1.L.L m plag 2 2 L1.HC pi lago 3 3 L1.L.L plago 3 3 L1.L.L 741 pl t nus, i, f. = 2 3 3 L1.L.L m pltan 2 2 L1.HC pl tano 3 3 L1.L.L pl tano 3 3 L1.L.L 742 prol gus, i, m., = 2 3 3 L1.L.L m prleg 2 2 L1.HC prlogo 3 3 L1.L.L prlogo 3 3 L1.L.L 743 s tyrus, i, m., = 2 3 3 L1.L.L m stir 2 2 L1.HC s tiro 3 3 L1.L.L stiro 3 3 L1.L.L 744 s l dus a, um 2 3 3 L1.L.L m slid 2 2 L1.HC slido 3 3 L1.L.L slido 3 3 L1.L.L

PAGE 409

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 745 th l mus i, m. = 2 3 3 L1.L.L m tlem 22L1.HC tlamo 3 3 L1.L.L tlamo 3 3 L1.L.L 746 t t lus, i, m. 2 3 3 L1.L.L m ttol 2 2 L1.HC ttulo 3 3 L1.L.L ttulo 3 3 L1.L.L 747 ton cus, a, um 2 3 3 L1.L.L m tnic 2 2 L1.HC tnico 3 3 L1.L. L tnico 3 3 L1.L.L 748 tr p cus a, um = 2 3 3 L1.L.L m trpic 2 2 L1.HC trpi co 3 3 L1.L.L trpico 3 3 L1.L.L 749 t m lus i, m. 2 3 3 L1.L.L m tmul 2 2 L1.HC tmulo 3 3 L1.L.L tmulo 3 3 L1.L.L 750 t rus i, m. 2 3 3 L1.L.L m ter 2 2 L1.HC tero 3 3 L1. L.L tero 3 3 L1.L.L 751 Z phyrus i, m. 2 3 3 L1.L.L m zfir 2 2 L1.HC cfiro 3 3 L1.L.L zfiro 3 3 L1.L.L L1.L.X (4th decl) (5) 752 c b tus s, m. 4 3 3 L1.L.L m cbit 2 2 L1. HC cbito 3 3 L1.L.L cbito 3 3 L1.L.L 753 h b tus, s, m. 4 3 3 L1.L.L m hbit 2 2 L1.HC hbito 3 3 L1.L. L hbito 3 3 L1.L.L 754 b tus, s, m. 4 3 3 L1.L.L m bit 2 2 L1. HC bito 3 3 L1.L.L bito 3 3 L1.L.L 755 r d tus, s, m. 4 3 3 L1.L.L m rdit 2 1 L1. HC rdito 3 3 L1.L.L rdito 3 3 L1.L.L 756 v m tus s, m. 4 3 3 L1.L.L m vmit 2 2 L1. HC vmito 3 3 L1.L.L vmito 3 3 L1.L.L L1.L.X (prothesis) (3) 757 sp c lum, i, n. 2 3 3 L1.L.L m espcul 3 2 L1.HC espculo 4 3 L1.L. L espculo 4 3 L1.L.L 758 st m lus, i, m. 2 3 3 L1.L.L m estmul 3 2 L1.HC estmulo 4 3 L1.L. L estmulo 4 3 L1.L.L 759 st m chus, i, m., = 2 3 3 L1.L.L m estmac 3 2 L1.HC estmago 409 4 3 L1.L.L estmago 4 3 L1.L.L L1.L.X (pro) (4th dcl) (1) 760 str p tus, s, m. 4 3 3 L1.L.L m estrpit 3 2 L1.HC estrpito 4 3 L1.L.L estrpito 4 3 L1.L.L L1.L.X (32) 761 p um ii, n 2 3 3 L1.L.L m api 2 2 L1.L apio 2 2 L1.L pio 2 2 L1.L 762 s nus i, m. 2 3 3 L1.L.L m ase 2 2 L1.L asno 2 2 HC1.L asno 2 2 HC1.L 763 tr um ii, n. 2 3 3 L1.L.L m atri 2 2 L1.L atrio 2 2 L1.L trio 2 2 L1.L 764 c l um ii, n. 2 3 3 L1.L.L m cili 2 2 L1.L cilio 2 2 L1.L clio 2 2 L1.L 765 *cod cus, der. regres. de codic lus 2 3 3 L1.L.L m codi 2 2 L1.L c digo 3 3 L1.L.L cdigo 3 3 L1.L.L 766 LL cranium (gr. ) 2 3 3 L1.L.L m crani 2 2 L1.L crneo 3 3 L1.L.L crnio 2 2 L1.L 767 LL *f c tum, i, n. 2 3 2 L1.L.L m fet ge 2 2 L1.L hgado 3 3 L1. L.L fgado 3 3 L1.L.L 768 f l um, ii, n. 2 3 3 L1.L.L m foli 2 2 L1.L folio 2 2 L1.L folho 2 2 L1.L 769 g n us, i, m. 2 3 3 L1.L.L m geni 2 2 L1.L genio 2 2 L1.L gnio 2 2 L1.L 770 gr m um, ii, n 2 3 3 L1.L.L m gremi 2 2 L1.L gremio 2 2 L1.L grmio 2 2 L1.L 771 l b um ii, n. 2 3 3 L1.L.L m llavi 2 2 L1.L labio 2 2 L1.L lbio 2 2 L1.L 772 m d cus, i, m. 2 3 3 L1.L.L m metge 2 2 L1.L mdico 3 3 L1.L. L mdico 3 3 L1.L.L 773 m n um, ii, n. 2 3 3 L1.L.L m mini 2 2 L1.L minio 2 2 L1.L mnio 2 2 L1.L

PAGE 410

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 774 n cl us (n c l us), i, m. 2 3 3 L1.L.L m nucli 2 2 L1.L ncleo 3 3 L1.L. L ncleo 3 3 L1.L.L 775 *nov us 2 3 3 L1.L.L m nuvi 2 2 L1.L novio 2 2 L1.L noivo 2 2 HV1.L 776 b lus, i, m., del gr. 2 3 3 L1.L.L m bol 2 2 L1.L bolo 3 3 L1.L.L bolo 3 3 L1.L.L 777 d um, ii, n. 2 3 3 L1.L.L m odi 2 2 L1.L odio 2 2 L1.L dio 2 2 L1.L 778 l um, i, n. 2 3 3 L1.L.L m oli 2 2 L1. L leo 3 3 L1.L.L leo 3 3 L1.L.L 779 p um o -on, i, n. 2 3 3 L1.L.L m opi 2 2 L1.L opio 2 2 L1.L pio 2 2 L1.L 780 pl g um, ii, n. 2 3 3 L1.L.L m plagi 2 2 L1.L plagio 2 2 L1.L plgio 2 2 L1.L 781 p p lus, i, m. 2 3 3 L1.L.L m poble 2 2 L1.L pueblo 2 2 L1.L povo 2 2 L1.L 782 p d um, ii, n. 2 3 3 L1.L.L m podi 2 2 L1.L podio 2 2 L1.L pdio 2 2 L1.L 783 r d us, ii, m. 2 3 3 L1.L.L m radi 2 2 L1.L radio 2 2 L1.L rdio 2 2 L1.L 784 r ph nus, i, m. 2 3 3 L1.L.L m rave 2 2 L1.L rbano 3 3 L1.L. L rbano 3 3 L1.L.L 785 s p dus, a, um 2 3 3 L1.L.L m savi 2 2 L1.L sabio 2 2 L1.L sbio 2 2 L1.L 786 s c us a, um 2 3 3 L1.L.L m soci 2 2 L1.L socio 2 2 L1.L scio 2 2 L1.L 787 s cer, ri, m. 2 3 3 L1.L.L m sogre 2 2 L1.L suegro 2 2 L1.L sogro 2 2 L1.L 788 s l um ii, n. 2 3 3 L1.L.L m soli 2 2 L1.L solio 2 2 L1.L slio 2 2 L1.L 789 tr v um i, n. 2 3 3 L1.L.L m trivi 2 2 L1.L trivio 2 2 L1.L trvio 2 2 L1.L 790 tr v um i, n. 2 3 3 L1.L.L m trivi 2 2 L1.L trvium 2 2 L1.L trvio 2 2 L1.L 791 v t um ii 2 3 3 L1.L.L m vici 2 2 L1.L vicio 410 2 2 L1.L vcio 2 2 L1.L 792 v d us a, um 2 3 3 L1.L.L m vidu 2 2 L1.L viudo 2 2 L1.L vivo 3 2 L.L1.L L1.L.X (prothesis) (3) 793 st b lum, i, n. 2 3 3 L1.L.L m estable 3 2 HC.L1.L establo 3 2 L1.L estbulo 4 3 L1.L.L 794 st d um, ii, n. 2 3 3 L1.L.L m estadi 3 2 HC.L1.L estadio 3 2 L1.L estdio 3 2 L1.L 795 st d um, ii, n. 2 3 3 L1.L.L m estudi 3 2 HC.L1.L estudio 3 2 L1.L estudo 3 2 L1.L L1.L.X (2) 796 l b rum i, n. 2 3 3 L1.L.L m lbarum 3 3 L1.L.HC lbaro 3 3 L1.L. L lbaro 3 3 L1.L.L 797 t t nus i, m. = 2 3 3 L1.L.L m ttanus 3 3 L1.L.HC ttanos 3 3 L1.L.HC ttanos 3 3 L1.L.HC L1.L.X (7) 798 d c plus a, um 2 3 3 L1.L.L m dcuple 3 3 L1.L.L dcuplo 3 3 L1.L. L dcuplo 3 3 L1.L.L 799 m d um, i, n. 2 3 3 L1.L.L m mdium 3 3 L1.L.L mdium 2 2 L1.L mdium 2 2 L1.L 800 m th dus and m th dos, i, f., = 2 3 3 L1.L.L m mtode 3 3 L1.L.L mtodo 3 3 L1.L.L mtodo 3 3 L1.L.L 801 n m rus, i, m. 2 3 3 L1.L.L m nmero 3 3 L1.L.L nmero 3 3 L1. L.L nmero 3 3 L1.L.L 802 r pr bus, a, um 2 3 3 L1.L.L m rprobe 3 3 L1.L.L rprobo 3 3 L1.L. L rprobo 3 3 L1.L.L 803 syn dus, i, f., = 2 3 3 L1.L.L m snode 3 3 L1.L. L snodo 3 3 L1.L.L snodo 3 3 L1.L.L 804 trimeter (-etros) m 2 3 3 L1.L.L m trmetre 3 3 L1.L.L trmetro 3 3 L1.L.L trmetro 3 3 L1.L.L

PAGE 411

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp HC.HC1.X (1) 805 p limpsestus, i, m. 2 4 2 HC.HC1.Lm palimpsest 31HC.HC1 palimpsesto 4 2 HC.HC1.L palimpsesto 4 2 HC.HC1.L HC.HC1.X (4) 806 exorcismus, i, m. 2 4 2 HC.HC1.L m exorcisme 4 2 HC.HC1.L exorcismo 4 2 HC.HC1.L exorcismo 4 2 HC.HC1.L 807 interdictum, i, n. 2 4 2 HC.HC1.L m interdicte 4 2 HC.HC1.L interdicto 4 2 HC.HC1.L interdito 4 2 HC.L1.L 808 r collectus, a, um 2 4 2 HC.HC1.L m recollecte 4 2 HC.HC1.L recoleto 4 2 L.L1.L recolecto 4 2 L.HC1.L 809 s n pismus, i, m., = 2 4 2 HC.HC1.L m sinapisme 4 2 HC.HC1.L sinapismo 4 2 HC.HC1.L sinapismo 4 2 HC.HC1.L HC.HC1.L (4th decl) (1) 810 intellectus, s, m. 4 4 2 HC.HC1.L m intellecte 4 2 HC.HC1 .L intelecto 4 2 L.HC1.L intelecto 4 2 L.HC1.L HC.HC1.X (1) 811 interregnum, i, n. 2 4 2 HC.HC1.L m interregne 4 2 L. HC1.L interregno 4 2 L.HC1.L interregno 4 2 L.HC1.L HC.HV1.X (3) 812 intervallum, i, n. 2 4 2 HC.HV1.L m interval 3 1 HC.HC1 intervalo 4 2 HC. L1.L intervalo 4 2 HC.L1.L 813 p rall lus, a, um 2 4 2 HC.HV1.L m parallel 3 1 HC.HC1 paralelo 4 2 L. L1.L paralelo 4 2 L.L1.L 814 pl bisc tum, i, n. 2 4 2 HC.HV1.L m plebiscit 3 1 HC. HC1 plebiscito 4 2 HC.L1.L plebiscito 4 2 HC.L1.L HC.HV1.X (4th decl) (1) 815 m gistr tus, s, m. 4 4 2 HC.HV1.L m magistrat 3 1 HC.HC1 magistrado 4 2 HC.L1.L magistrado 4 2 HC.L1.L HC.HV1.X (9) 816 adiect vus, a, um 2 4 2 HC.HV1.L m adjectiu 3 1 HC.HV1 adjetivo 411 4 2 L.L1. L adjectivo 4 2 HC.L1.L 817 attract vus 2 4 2 HC.HV1.L m atractiu 3 1 HC.HV1 atractivo 4 2 HC.L1.L atractivo 4 2 HC.L1.L 818 collect vus, a, um 2 4 2 HC.HV1.L m colectiu 3 1 HC.HV1 colectivo 4 2 HC.L1.L coletivo 4 2 L.L1.L 819 effect vus 2 4 2 HC.HV1.L m efectiu 3 1 HC.HV1 efectivo 4 2 HC.L1.L efectivo 4 2 HC.L1.L 820 incent vum, i, n. 2 4 2 HC.HV1.L m incentiu 3 1 HC.HV1 incentivo 4 2 HC.L1.L incentivo 4 2 HC.L1.L 821 intest num, i, n. 2 4 2 HC.HV1.L m intest 3 1 HC. HV1 intestino 4 2 HC.L1.L intestino 4 2 HC.L1.L 822 *sacristanus 2 4 2 HC.HV1.L m sagrist, sacrist 3 1 HC.HV1 sacris tn 3 1 HC.HC1 sacristo 3 1 HC.HV1 823 subjunct vus, a, um 2 4 2 HC.HV1.L m subjuntiu 3 1 HC. HV1 subjuntivo 4 2 HC.L1.L subjuntivo 4 2 HV.L1.L 824 vespert nus 2 4 2 HC.HV1.L m vespert 3 1 HC.HV1 vesperti no 4 2 HC.L1.L vespertino 4 2 HC.L1.L HV.HC1.X (20) 825 arg mentum i, n. 2 4 2 HV.HC1.L m argument 3 1 L.HC1 argumento 4 2 L1.HC 1.L argumento 4 2 L1.HC1.L 826 compl mentum i, n 2 4 2 HV.HC1.L m complement 3 1 L. HC1 complemento 4 2 L.HC1.L complemento 4 2 L.HV1.L 827 compr missum i, n. 2 4 2 HV.HC1.L m comproms 3 1 L.HC1 compromiso 4 2 L.L1.L compromisso 4 2 L.L1.L 828 cond mentum i, n. 2 4 2 HV.HC1.L m condiment 3 1 L. HC1 condimento 4 2 L.HC1.L condimento 4 2 L.HV1.L 829 d cr mentum i, n. 2 4 2 HV.HC1.L m decrement 3 1 L.HC1 decremento 4 2 L.HC1.L decremento 4 2 L.HV1.L 830 excr mentum, i, n. 2 4 2 HV.HC1.L m excrement 3 1 L.HC1 excremento 4 2 L.HC1.L excremento 4 2 L.HV1.L

PAGE 412

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 831 firm mentum, i, n. 2 4 2 HV.HC1.Lm firmament 31L. HC1 firmamento 42L.HC1.L firmamento 42L.HV1.L 832 fund mentum, i, n. 2 4 2 HV.HC1.Lm fonament 31 L.HC1 fundamento 42L.HC1.L fundamento 4 2 L.HV1.L 833 incr mentum, i, n. 2 4 2 HV.HC1.L m increment 3 1 L.HC1 incremento 4 2 L.HC1.L incremento 4 2 L.HV1.L 834 ind mentum, i, n. 2 4 2 HV.HC1.L m indument 3 1 L. HC1 indumento 4 2 L.HC1.L indumento 4 2 L.HV1.L 835 instr mentum, i, n. 2 4 2 HV.HC1.L m instrument 3 1 L.HC1 inst rumento 4 2 L.HC1.L instrumento 4 2 L.HV1.L 836 j r mentum i, n. 2 4 2 HV.HC1.L m jurament 3 1 L.HC1 juramento 4 2 L.HC1.L juramento 4 2 L.HV1.L 837 l g mentum, i, n. 2 4 2 HV.HC1.L m lligament 3 1 L.HC1 ligamento 4 2 L.HC1.L ligamento 4 2 L.HV1.L 838 l n me ntum, i, n. 2 4 2 HV.HC1.L m liniment 3 1 L. HC1 linimento 4 2 L.HC1.L linimento 4 2 L.HV1.L 839 orn mentum, i, n. 2 4 2 HV.HC1.L m ornament 3 1 L.HC1 ornamento 4 2 L.HC1.L ornamento 4 2 L.HV1.L 840 p v mentum, i, n. 2 4 2 HV.HC1.L m paviment 3 1 L.HC1 pavimento 4 2 L.HC1 .L pavimento 4 2 L.HV1.L 841 r d mentum, i, n. 2 4 2 HV.HC1.L m rudiment 3 1 L.HC1 rudimento 4 2 L.HC1.L rudimento 4 2 L.HV1.L 842 s cr mentum, i, n. 2 4 2 HV.HC1.L m sagrament 3 1 L.HC1 sacramento 4 2 L.HC1.L sacramento 4 2 L.HV1.L 843 suppl mentum, i, n. 2 4 2 HV.HC1.L m suplement 3 1 L. HC1 suplemento 4 2 L.HC1.L suplemento 4 2 L.HV1.L 844 test mentum i, n. 2 4 2 HV.HC1.L m testament 3 1 L. HC1 testamento 4 2 L.HC1.L testamento 4 2 L.HV1.L HV.HC1.X (4th decl) (1) 845 r tr cessus, s, m. 4 4 2 HV.HC1.L m retrocs 3 1 L.HC1 retroceso 4 2 L.L1.L retrocesso 4 2 L.L1.L HV.HC1.X (3) 846 pr pismus, i, m., = 2 4 2 HV.HC1.L m priapisme 4 2 L.HC1.L pr iapismo 4 2 L.HC1.L priapismo 4 2 L.HC1.L 847 aquaeductus 2 4 2 HV.HC1.L m aqeducte 4 2 L. HC1.L acueducto 412 4 2 L.HC1.L aqueduto 4 2 L.L1.L 848 c t chism us i, m. 2 4 2 HV.HC1.L m catecisme 4 2 L.HC1 .L catecismo 4 2 L.HC1.L catecismo 4 2 L.HC1.L HV.HC1.X (2) 849 d tr mentum i, n. 2 4 2 HV.HC1.L m detriment 3 1 L.HC1 detrimento 4 2 L.HC1.L detrimento 4 2 L.HV1.L HV.HV1.L (3) 850 pr r tus, s, m. 4 4 2 HV.HV1.L m priorat 3 1 L.HC1 priorat o 3 2 L.L1.L priorato 4 2 L.L1.L 851 tr b n tus s, m. 4 4 2 HV.HV1.L m tribunat 3 1 L. HC1 tribunado 4 2 L.L1. L tribunado 4 2 L.L1.L 852 r qu s tus, a, um 2 4 2 HV.HV1.L m requisit 3 1 L.HC1 requisito 4 2 L.L1. L requisito 4 2 L.L1.L HV.HV1.X (8) 853 th naeum i, n., y este del gr. 2 4 2 HV.HV1.L m ateneu 3 1 L.HV1 ateneo 4 2 L.L1.L ateneu 3 1 L.HV1 854 inf n tus, a, um 2 4 2 HV.HV1.L m infinit 3 1 L. HC1 infinito 4 2 L.L1.L infinito 4 2 L.L1.L 855 Maus l um, i, n. 2 4 2 HV.HV1.L m mausoleu 3 1 L.HV1 mausoleo 4 2 L. L1.L mausolu 3 1 L.HV1 856 n g t vus, a, um 2 4 2 HV.HV1.L m negatiu 3 1 L.HV1 negativo 4 2 L.L1. L negativo 4 2 L.L1.L 857 p l t nus, a, um 2 4 2 HV.HV1.L m palat 3 1 L.HV1 palatino 4 2 L.L1.L palatino 4 2 L.L1.L 858 p tr n tus, s, m. 4 4 2 HV.HV1.L m patronat 3 1 L.HC1 patronato 4 2 L.L1.L patronato 4 2 L.L1.L 859 praem t rum, i, n. 2 4 2 HV.HV1.L m prematur 3 1 L.HC1 prematur o 4 2 L.L1. L prematuro 4 2 L.L1.L

PAGE 413

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 860 r l t vus, a, um 2 4 2 HV.HV1.Lm relatiu 31L. HV1 relativo 42L.L1.L relativo 42L.L1.L 861 v c t vus a, um 2 4 2 HV.HV1.Lm vocatiu 31L. HV1 vocativo 42L.L1.L vocativo 42L.L1.L HV.HV1.X (1) 862 oct g n s, i, adj. 2 4 2 HV.HV1.Lm octgon 32L1.HC octgono 43L1.L. L octgono 43L1.L.L HV.HV1.X (1) 863 cand l brum i, n. 2 4 2 HV.HV1.Lm canelobre 42 L.L1.L candelabro 42L.L1. L candelabro 42L.L1.L L.HC1.X (3) 864 *capitellus 2 4 2 L.HC1.L m cabdill 21HC1 caudillo 32L1.L caudilho 32L1.L 865 m l dictus, a, um 2 4 2 L.HC1.L m maldit 2 1 HC1 maldito 3 2 L1.L maldito 3 2 L1.L 866 c pertus 2 4 2 L.HC1.L m cobert 2 1 HC1 c ubierto 3 2 L.HC1.L coberto 3 2 L.HC1.L L.HC1.X (2) 867 hyacinthus (-os) m. 2 4 2 L.HC1.L m jacint 2 1 L. HC1 jacinto 3 2 L.L1.L jacinto 3 2 L.L1.L 868 m n mentum, i, n. 2 4 2 L.HC1.L m moment 2 1 L.HC1 momento 3 2 L.HC1.L momento 3 2 L.HC1.L L.HC1.X (27) 869 l mentum i 2 4 2 L.HC1.L m aliment 3 1 L.HC1 alimento 4 2 L.HC1.L alimento 4 2 L.HC1.L 870 m antus i, del gr. 2 4 2 L.HC1.L m amiant 3 1 L.HC1 amianto 3 2 HC1.L amianto 3 2 HC1.L 871 Antichristus, y este del gr. 2 4 2 L.HC1.L m anticrist 3 1 L.HC1 antic risto 4 2 L.HC1.L anticristo 4 2 L.HC1.L 872 b s liscus i, m., y este del gr. 2 4 2 L.HC1.L m basilisc 3 1 L.HC1 basilisco 413 4 2 L.HC1.L basilisco 4 2 L.HC1.L 873 lat. vulg. cascabellus 2 4 2 L.HC1.L m cascavell 3 1 L.HC1 cascabel 3 1 L.HC1 cascavel 3 1 L.HC1 874 c r bellum i, n. 2 4 2 L.HC1.L m cerebel 3 1 L.HC1 cerebelo 4 2 L.L1.L cerebelo 4 2 L.L1.L 875 c d cillus, i, m. 2 4 2 L.HC1.L m codicil 3 1 L.HC1 codicilo 4 2 L.L1. L codicilo 4 2 L.L1.L 876 b. lat.contrapunctus 2 4 2 L.HC1 .L m contrapunt 3 1 L.HC1 contrapunt o 4 2 L.HC1.L contraponto 4 2 L.HC1.L 877 crucifixus, a, um 2 4 2 L.HC1.L m crucifix 3 1 L.HC1 crucifijo 4 2 L. L1.L crucifixo 4 2 L.L1.L 878 d c mentum, i, n. 2 4 2 L.HC1.L m document 3 1 L. HC1 documento 4 2 L.HC1.L documento 4 2 L.HC1.L 879 l mentum i, n. 2 4 2 L.HC1.L m element 3 1 L.HC1 elemento 4 2 L.HC1.L elemento 4 2 L.HC1.L 880 b. lat. filamentum 2 4 2 L.HC1 .L m filament 3 1 L.HC1 filament o 4 2 L.HC1.L filamento 4 2 L.HC1.L 881 hipp campus o -os, i, m., = 2 4 2 L.HC1.L m hipocamp 3 1 L.HC1 hipocampo 4 2 L.HC1.L hipocampo 4 2 L.HC1.L 882 l byrinthus i, m., del gr. 2 4 2 L.HC1.L m laberi nt 3 1 L.HC1 laberinto 4 2 L. HC1.L labirinto 4 2 L.HC1.L 883 l pardus, i, m. 2 4 2 L.HC1.L m lleopard 3 1 L. HC1 leopardo 4 2 L.HC1.L leopardo 4 2 L.HC1.L 884 m n festus, a, um 2 4 2 L.HC1.L m manifest 3 1 L.HC1 manifiesto 4 2 L.HC1.L manifesto 2 L.HC1.L 885 manus+scriptus 2 4 2 L.HC1.L m manuscrit 3 1 L.HC1 manuscrito 4 2 L.L1. L manuscrito 4 2 L.L1.L 886 m n mentum, i, n. 2 4 2 L.HC1.L m monument 3 1 L. HC1 monumento 4 2 L.HC1.L monumento 4 2 L.HC1.L 887 b liscus, i, m. 2 4 2 L.HC1.L m obelisc 3 1 L. HC1 obelisco 4 2 L.HC1.L obelisco 4 2 L.HC1.L 888 orth doxus, a, um, adj. 2 4 2 L.HC1.L m ortodox 3 1 L.HC1 ortodoxo 4 2 L.HC1 .L ortodoxo 4 2 L.HC1.L

PAGE 414

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 889 b.lat. protocollum, del gr. 2 4 2 L.HC1.L m prot ocol 31L.HC1 protocolo 42 L.L1.L protocol o 42L.L1.L 890 s d mentum, i, n. 2 4 2 L.HC1.L m sediment 31L. HC1 sedimento 42L.HC1.L sedimento 42L.HC1.L 891 t g mentum, i, n. 2 4 2 L.HC1.L m tegument 31L. HC1 tegumento 42L.HC1.L tegumento 42L.HC1.L 892 t r binthus i, f., del gr. 2 4 2 L.HC1.L m terebint 3 1 L.HC1 terebinto 4 2 L.HC1 .L terebinto 4 2 L.HC1.L 893 n versum i, n. 2 4 2 L.HC1.L m univers 3 1 L.HC1 universo 4 2 L.HC1.L universo 4 2 L.HC1.L 894 vert cillus i 2 4 2 L.HC1.L m verticil 3 1 L. HC1 verticilo 4 2 L.L1. L verticilo 4 2 L.L1.L 895 r g mentum, i, n. 2 4 2 L.HC1.L m regiment 3 1 L.HC1 regimiento 4 2 L.HC1.L regimento 4 2 L.HC1.L L.HC1.X (2) 896 m m randus, a, um 2 4 2 L.HC1.L m memorndum 4 2 L.HC1 .HC memorndum 4 2 L.HC1.HC memorandum 4 2 L.HC1.HC 897 r f rendum, i 2 4 2 L.HC1.L m referndum 4 2 L.HC1 .HC referndum 4 2 L.HC1.HC referendum 4 2 L.HV1.HV L.HC1.X (12) 898 aph r smus (
PAGE 415

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 916 m l tum, i, n. 2 4 2 L.HV1.L m amulet 3 1 L.HC1 amuleto 4 2 L. L1.L amuleto 4 2 L.L1.L 917 appar tus 2 4 2 L.HV1.L m aparat 3 1 L.HC1 aparato 4 2 L.L1.L aparato 4 2 L.L1.L 918 attr b tum i, n. 2 4 2 L.HV1.L m atribut 3 1 L.HC1 atributo 4 2 L. L1.L atributo 4 2 L.L1.L 919 cand d tus, i, m. 2 4 2 L.HV1.L m candidat 3 1 L.HC1 candidat o 4 2 L.L1.L candidato 4 2 L.L1.L 920 card m mum i, n. 2 4 2 L.HV1.L m cardamom 3 1 L.HC1 cardamomo 4 2 L.L1. L cardamomo 4 2 L.L1.L 921 n m cus i, m. 2 4 2 L.HV1.L m enemic 3 1 L.HC1 enemigo 4 2 L.L1. L inimigo 4 2 L.L1.L 922 r d tus, a, um 2 4 2 L.HV1.L m erudit 3 1 L.HC1 erudito 4 2 L. L1.L erudito 4 2 L.L1.L 923 h l caust um, i, n., = 2 4 2 L.HV1.L m holocaust 3 1 L.HC1 holoc austo 4 2 L.HC1.L holocausto 4 2 L.HC1.L 924 n d rus, a, um 2 4 2 L.HV1.L m inodor 3 1 L.HC1 inodoro 4 2 L.L1. L inodoro 4 2 L.L1.L 925 inst t tum, i, n. 2 4 2 L.HV1.L m institut 3 1 L.HC1 institut o 4 2 L.L1.L inst ituto 4 2 L.L1.L 926 litt r tus, a, um 2 4 2 L.HV1. L m literat 3 1 L.HC1 literat o 4 2 L.L1.L liter ato 4 2 L.L1.L 927 mete rus, y este del gr. 2 4 2 L.HV1.L m meteor 3 1 L.HC1 meteoro 4 2 L.L1.L meteoro 4 2 L.L1.L 928 myst g gus, i, m., = 2 4 2 L.HV1.L m mistagog 3 1 L.HC1 mi stago go 4 2 L.L1.L mistagogo 4 2 L.L1.L 929 pant m mus, i, m. 2 4 2 L.HV1.L m pantomim 3 1 L.HC1 pantomimo 4 2 L. L1.L pantomimo 4 2 L.L1.L 930 p r d s us, i, m. 2 4 2 L.HV1.L m parads 3 1 L.HC1 paraso 4 2 L.L1. L paraso 4 2 L.L1.L 931 paed g gus, i, m., = 2 4 2 L.HV1.L m pedagog 3 1 L.HC1 pedagogo 4 2 L.L1.L pedagogo 4 2 L.L1.L 932 p rist l um, ii, n., = 2 4 2 L.HV1.L m peristil 3 1 L.HC1 per istilo 415 4 2 L.L1.L peristilo 4 2 L.L1.L 933 pl v sus, a, um, adj. 2 4 2 L.HV1.L m pluvi s 3 1 L.HC1 pluvioso 3 2 L1.L pluvioso 3 2 L1.L 934 pr st t tus, a, um 2 4 2 L.HV1.L m prostitut 3 1 L. HC1 prostituto 4 2 L.L1.L prostituto 4 2 L.L1.L 935 subst t tus, a, um 2 4 2 L.HV1.L m substitut 3 1 L. HC1 sustituto 4 2 L.L1.L substituto 4 2 L.L1.L 936 ung l tus a, um 2 4 2 L.HV1.L m ungulat 3 1 L.HC1 ungulado 4 2 L.L1. L ungulado 4 2 L.L1.L L.HV1. X (4th decl) (1) 937 princ p tus, s, m. 4 4 2 L.HV1.L m pr incipat 3 1 L.HC1 principado 4 2 L.L1.L principado 4 2 L.L1.L 938 p g l tus, s, m. 4 4 2 L.HV1.L m pugilat 3 1 L.HC1 pugilato 4 2 L.L1.L pugilato 4 2 L.L1.L L.HV1.X / HC1.L.X (1) 939 cinn m mum, cinn mum, o cinn mon, i, n. 2 4 2 L.HV1.L m cnamom 3 3 L1.L.HC cinamomo 4 2 L.L1.L cinamomo 4 2 L.L1.L L.HV1.X (21) 940 abl t vus, i, m. 2 4 2 L.HV1.L m ablatiu 3 1 L.HV1 ablativo 4 2 L. L1.L ablativo 4 2 L.L1.L 941 p g us a, um 2 4 2 L.HV1.L m apogeu 3 1 L.HV1 apogeo 4 2 L. L1.L apogeu 3 1 L.HV1 942 c p t nus, i, m. 2 4 2 L.HV1.L m capit 3 1 L.HV1 capitn 3 1 L.HC1 capito 3 1 L.HV1 943 c ryphaeus i, m., del gr. 2 4 2 L.HV1.L m co rifeu 3 1 L.HV1 corifeo 4 2 L.L1.L corifeu 3 1 L.HV1

PAGE 416

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 944 Christ nus a, um 2 4 2 L.HV1.L m cristi 31L.HV1 cristiano 32L1.L cristo 21HV1 945 g n t vus, a, um 2 4 2 L.HV1.L m genitiu 31 L.HV1 genitivo 42L.L1.L genitivo 4 2 L.L1.L 946 Hymenaeus, a, um = 2 4 2 L.HV1.L m himeneu 3 1 L.HV1 himeneo 4 2 L.L1.L Himeneu 3 1 L.HV1 947 hyp gaeum, i, n. 2 4 2 L.HV1.L m hipogeu 3 1 L.HV1 hipogeo 4 2 L. L1.L hipogeu 3 1 L.HV1 948 inqu l nus, a, m. & f. 2 4 2 L.HV1.L m inquil 3 1 L.HV1 inquilino 4 2 L.L1. L inquilino 4 2 L.L1.L 949 j b laeus, i, m. 2 4 2 L.HV1.L m jubileu 3 1 L.HV1 jubileo 4 2 L.L1.L jubileu 3 1 L.HV1 950 b. l. locativus, -a, -um 2 4 2 L.HV1.L m loca tiu 3 1 L.HV1 locativo 4 2 L.L1.L locativo 4 2 L.L1.L 951 p r gr nus, i, m. 2 4 2 L.HV1.L m pelegr 3 1 L.HV 1 peregrino 4 2 L.L1.L peregrino 4 2 L.L1.L 952 p l c nus, i, m. 2 4 2 L.HV1.L m pelic 3 1 L.HV1 pelcano 4 3 L1.L. L pelicano 4 2 L.L1.L 953 lat. tard. Pergam num (Perg m nus, a, um) 2 4 2 L.HV1.L m pergam 3 1 L.HV1 per gamino 4 2 L.L1.L pergaminho 4 2 L.L1.L 954 p r naeon & p r n on, i, n. 2 4 2 L.HV1.L m perineu 3 1 L.HV1 perin 3 1 L.L1 perneo 4 3 L1.L.L 955 p s t vus, a, um 2 4 2 L.HV1.L m positiu 3 1 L.HV1 positivo 4 2 L.L1.L positivo 4 2 L.L1.L 956 pr m t vus, a, um 2 4 2 L.HV1.L m primitiu 3 1 L.HV1 primitivo 4 2 L.L1.L primitivo 4 2 L.L1.L 957 pr pylaeum (-laeon), i, n., = 2 4 2 L.HV1.L m propileu 3 1 L.HV1 propileo 4 2 L.L1.L propileu 3 1 L.HV1 958 p bl c n us, i, m. 2 4 2 L.HV1.L m public 3 1 L.HV1 publicano 4 2 L. L1.L publicano 4 2 L.L1.L 959 *super nus 2 4 2 L.HV1.L m sobir 3 1 L. HV1 soberano 416 4 2 L.L1. L soberano 4 2 L.L1.L 960 v t r nus a, um 2 4 2 L.HV1.L m veter 3 1 L. HV1 veterano 4 2 L.L1.L veterano 4 2 L.L1.L L.HV1.X (1) 961 ultim tum 2 4 2 L.HV1.L m ultimtum 4 2 L.L1.HC ultimtum 4 2 L.L1.HC ultimato 4 2 L.L1.L L.HV1.X (3) 962 ambul crum 2 4 2 L.HV1.L m ambu lacre 4 2 L.L1.L ambulacro 4 2 L.L1.L ambulacro 4 2 L.L1.L 963 inv l crum, i, n. 2 4 2 L.HV1.L m involucre 4 2 L.L1.L involucro 4 2 L.L1. L invlucro 4 3 L1.L.L 964 s m l crum i, n. 2 4 2 L.HV1.L m simulacre 4 2 L.L1.L simulacr o 4 2 L.L1.L simulacro 4 2 L.L1.L HC1.L.X (1) 965 commerc um, ii, n. 2 4 3 HC1.L.L m comer 2 1 HC1 comercio 3 2 HC1.L comrcio 3 2 HC1.L HC1.L.X (2) 966 chrysanth mum o -mon, i, n., = 2 4 3 HC1.L.L m crisantem 3 1 HC.HC1 cris antemo 4 2 HC.L1.L crisntemo 4 3 HC1.L.L 967 tr umvir, v ri, m. 2 4 3 HC1.L.L m triumvir 3 3 HC.HC1 triunviro 3 2 HC.L1.L trinviro 4 3 HC1.L.L HC1.L.X (8) 968 adulter, ri 2 4 3 HC1.L.L m adlter 3 2 HC1.HC adltero 4 3 HC1.L.L adltero 4 3 HC1.L.L 969 apost lus, del gr. 2 4 3 HC1.L.L m apstol 3 2 HC1.HC apstol 3 2 HC1.HC apstolo 3 2 HC1.L.L 970 archang lus, del gr. 2 4 3 HC1.L.L m arcngel 3 2 HC1.HC arcngel 3 2 HC1.HC arcanjo 3 2 HC1.L 971 f namb lus, i, m. 2 4 3 HC1.L.L m funmbul 3 2 HC1.HC funmbulo 4 3 HC1.L.L funmbulo 4 3 HC1.L.L

PAGE 417

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 972 h rosc pus, i. m. 2 4 3 HC1.L.L m horscop 32HC1 .HC horscopo 43HC1.L.L horscopo 43HC1.L.L 973 incogn tus, a, um 2 4 3 HC1.L.L m incgnit 32HC1. HC incgnito 43HC1.L.L incgnito 43HC1.L.L 974 praeamb lus 2 4 3 HC1.L.L m prembul 32HC1.HC prembulo 43HC1.L.L prembulo 43HC1.L.L 975 pr gn st con o -um, i, n. = 2 4 3 HC1.L.L m pronstic 32HC1.HC pronstico 43HC1.L.L prognstico 43HC1.L.L HC1.L.X (4th decl) (1) 976 exerc tus s, m. 4 4 3 HC1.L.L m exrcit 32HC1.HC ejrcito 43HC1.L.L exrcito 43HC1.L.L HC1.L.X (prothesis) (1) 977 sch last cus, i, m. 2 4 3 HC1.L.L m escolstic 42HC1.HC escolstico 5 3 HC1.L.L escolstico 5 3 HC1.L.L HC1.L.X (24) 978 adverb um, i 2 4 3 HC1.L.L m adverbi 3 2 HC1.L adverbio 3 2 HC1.L advrbio 3 2 HC1.L 979 compend um ii, n. 2 4 3 HC1.L.L m compendi 3 2 HC1.L compendio 3 2 HC1.L compndio 3 2 HC1.L 980 consort um ii, n. 2 4 3 HC1.L.L m consorci 3 2 HC1 .L consorcio 3 2 HC1.L consrcio 3 2 HC1.L 981 corpusc lum i, n. 2 4 3 HC1.L.L m corpuscle 3 2 HC1.L corpsculo 4 3 HC1.L.L corpsculo 4 3 HC1.L.L 982 crepusc lum, i, n. 2 4 3 HC1.L.L m crepuscle 3 2 HC1 .L crepsculo 4 3 HC1.L.L crepsculo 4 3 HC1.L.L 983 dispend um ii, n. 2 4 3 HC1.L.L m dispendi 3 2 HC1.L dispendio 3 2 HC1.L dispndio 3 2 HC1.L 984 d vort um, ii, n. 2 4 3 HC1.L.L m divorci 3 2 HC1 .L divorcio 3 2 HC1.L divrcio 3 2 HC1.L 985 exord um, ii, n. 2 4 3 HC1.L.L m exordi 3 2 HC1.L exordio 3 2 HC1.L exrdio 3 2 HC1.L 986 g rund um, ii, n. 2 4 3 HC1.L.L m gerundi 3 2 HC1.L gerundio 417 3 2 HC1.L gerndio 3 2 HC1.L 987 dyll um or dyll um, ii, n. 2 4 3 HC1.L.L m idilli 3 2 HC1.L idilio 3 2 L1.L idlio 3 2 L1.L 988 incend um, i, n. 2 4 3 HC1.L.L m incendi 3 2 HC1.L incendio 3 2 HC1.L incndio 3 2 HC1.L 989 insomn um, i, n. 2 4 3 HC1.L.L m insomni 3 2 HC1.L insomnio 3 2 HC1.L insnia 3 2 L1.L 990 m tall cus, a, um 2 4 3 HC1.L.L m metllic 3 2 HC1.L metlico 4 3 L1.L.L metlico 4 3 L1.L.L 991 perc lum, i, n 2 4 3 HC1.L.L m opercle 3 2 HC1.L oprculo 4 3 HC1.L.L oprculo 4 3 HC1.L.L 992 pusc lum, i, n. 2 4 3 HC1.L.L m opuscle 3 2 HC1.L opsculo 4 3 HC1.L.L opsculo 4 3 HC1.L.L 993 pr verb um, i, n. 2 4 3 HC1.L.L m proverbi 3 2 HC1.L proverbio 3 2 HC1.L provrbio 3 2 HC1.L 994 r nunc lus, i, m. 2 4 3 HC1.L.L m ranuncle 3 2 HC1.L rannculo 4 3 HC1.L.L rannculo 4 3 HC1.L.L 995 rect ang lum, i, n. 2 4 3 HC1.L.L m rectangle 3 2 HC1 .L rectngulo 4 3 HC1.L.L rectngulo 4 3 HC1.L.L 996 sestert um, i, n. 2 4 3 HC1.L.L m sesterci 3 2 HC1 .L sestercio 3 2 HC1.L sestrcio 3 2 HC1.L 997 s lent um, ii, n. 2 4 3 HC1.L.L m silenci 3 2 HC1.L silencio 3 2 HC1.L silncio 3 2 HC1.L 998 suburb um, ii, n. 2 4 3 HC1.L.L m suburbi 3 2 HC1.L suburbio 3 2 HC1.L subrbio 3 2 HC1.L 999 tr ang lus i, m. 2 4 3 HC1.L.L m triangle 3 2 HC1 .L tringulo 4 3 HC1.L.L tringulo 4 3 HC1.L.L 1000 tr syll bus a, um, adj. 2 4 3 HC1.L.L m trisllab 3 2 HC1.L trislabo 4 3 L1.L. L trisslabo 4 3 L1.L.L 2 4 3 HC1.L.L m tubercle 3 2 HC1.L tubrculo 4 3 HC1.L.L tubrculo 4 3 HC1.L.L HC1.L.X (prothesis) (1) 1002 st pend um, ii, n. 2 4 3 HC1.L.L m estipendi 4 2 HC1.L estipendio 4 2 HC1.L estipndio 4 2 HC1.L

PAGE 418

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp HC1.L.X (2) 1003 synd ton i, n., y este del gr. 2 4 3 HC1.L.L m asndeton 43HC1.L.HCa sndeton 43HC1.L.HCassndeto 43HC1.L.HC 1004 hyperb ton, i, n., = 2 4 3 HC1.L.L m hiprbaton 43HC1.L.HChiprbato 43HC1.L.L hiprbato 43HC1.L.L HC1.L.X (3) 1005 b enn um ii, n. 2 4 3 HC1.L.L m bienni 32L1.L bienio 32L1.L binio 32L1.L 1006 d cenn um ii, n. 2 4 3 HC1.L.L m decenni 32L1.L decenio 32HC1.L decnio 32HC1.L 1007 quinquenn um, ii, n. 2 4 3 HC1.L.L m quinquenni 32L1.L quinquenio 32L1.L quinqunio 32L1.L HV1.L.X (3) 1008 J n us, a, um 2 3 3 HV1.L.L m juny 11HC1 junio 22L1.L junho 22L1.L 1009 p d c lus, i, m. 2 4 3 HV1.L.L m po ll 1 1 HC1 piojo 2 2 L1.L piolho 2 2 L1.L 1010 c g lum i, n. 2 4 3 HV1.L.L m qua ll 1 1 HC1 cuajo 2 2 L1.L coalho 3 2 L1.L HV1.L.X (17) 1011 arm r um ii, n. 2 4 3 HV1.L.L m armer 2 1 HC1 armero 3 2 L1.L armeiro 3 2 HV1.L 1012 caps r us ii, m. 2 4 3 HV1.L.L m caixer 2 1 HC1 cajero 3 2 L1.L caixeiro 3 2 HV1.L 1013 cald r um ii, n. 2 4 3 HV1.L.L m calder 2 1 HC1 caldero 3 2 L1.L caldeiro 3 2 HV1.L 1014 c n c lus i, m. 2 4 3 HV1.L.L m conill 2 1 HC1 conejo 3 2 L1.L coelho 3 2 L1.L 1015 VL *cordarius, der. de cordus 2 4 3 HV1.L.L m corder 2 1 HC1 co rdero 418 3 2 L1.L cordeiro 3 2 HV1.L 1016 d n r us a, um, adj. 2 4 3 HV1.L.L m diner 2 1 HC1 dinero 3 2 L1.L dinheiro 3 2 HV1.L 1017 fast d um, i, n. 2 4 3 HV1.L.L m fastig 2 1 HC1 fastidio 3 2 L1.L fastio 3 2 L1.L 1018 ferr r us, ii, m., 2 4 3 HV1.L.L m ferrer 2 1 HC1 herrero 3 2 L1.L ferreiro 3 2 HV1.L 1019 mort r um, ii, n. 2 4 3 HV1.L.L m morter 2 1 HC1 mortero 3 2 L1.L morteiro 3 2 HV1.L 1020 p r c lum, i, n. 2 4 3 HV1.L.L m peril l 2 1 HC1 peligro 3 2 L1.L perigo 3 2 L1.L 1021 pl m r us, a, um 2 4 3 HV1.L.L m plomer 2 1 HC1 plumero 3 2 L1.L plumeiro 3 2 HV1.L 1022 port r us, ii, m. 2 4 3 HV1.L.L m porter 2 1 HC1 portero 3 2 L1.L porteiro 3 2 HV1.L 102 3 pr n men, nis, n. 2 4 3 HV1.L.L m pronom 2 1 HC1 pronombre 3 2 HC1.L pronome 3 2 L1.L 1024 terr r us, i, m. (lat. med.) 2 4 3 HV1.L.L m te rrer 2 1 HC1 terrero 3 2 L1.L terreiro 3 2 HV1.L 1025 vivar um 2 4 3 HV1.L.L m viver 2 1 HC1 vivero 3 2 L1.L viveiro 3 2 HV1.L 1026 extr n us a, um 2 4 3 HV1.L.L m estrany 2 1 HC1 extrao 3 2 L1.L estranho 3 2 L1.L 1027 susp r um ii, n. 2 4 3 HV1.L.L m sospir 2 1 HC1 suspiro 3 2 L1.L suspiro 3 2 L1.L HV1.L.X (1) 1028 fast d um, i, n. 2 4 3 HV1.L.L m fstic (ant. fa sti) 2 2 HC1.HC hasto 3 2 L1.L fastio 3 2 L1.L HV1.L.X (2) 1029 r g num / -on, i, n. 2 4 3 HV1.L.L m oren ga 3 2 HC1.L organo 4 3 L1.L.L orgo 3 2 L1.HV 1030 p sc pus, i, m., del gr. 2 4 3 HV1.L.L m bisbe 2 2 HC1.L obispo 3 2 HC1.L bispo 2 2 HC1.L

PAGE 419

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp HV1.L.X (1) 1031 p l t um, ii, n. 2 4 3 HV1.L.L m palau 21HV 1 palacio 32L1.L palcio 32L1.L HV1.L.X (3) 1032 VL desid um
PAGE 420

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 1059 d r um ii, n. 2 4 3 HV1.L.L m diar i 32L1.L diario 22L1.L dirio 22L1.L 1060 fals r us, ii, m. 2 4 3 HV1.L.L m falsari 32L1.L falsario 32L1.L falsrio 32L1.L 1061 gloss r um, ii, n. 2 4 3 HV1.L.L m glossari 3 2 L1.L glosario 3 2 L1.L glossrio 3 2 L1.L 1062 herb rium, ii, n. 2 4 3 HV1.L.L m herbari 3 2 L1.L herbario 3 2 L1.L herbrio 3 2 L1.L 1063 h r r um, i, n. 2 4 3 HV1.L.L m horari 3 2 L1.L horario 3 2 L1.L horrio 3 2 L1.L 1064 l d br um, ii, n. 2 4 3 HV1.L.L m ludibri 3 2 L1.L ludibrio 3 2 L1.L ludbrio 3 2 L1.L 1065 m n br um, ii, n. 2 4 3 HV1.L.L m manubri 3 2 L1.L manubrio 3 2 L1. L manbrio 3 2 L1.L 106 6 mars p um, ii, n., = 2 4 3 HV1.L.L m marsupi 3 2 L1.L marsupio 3 2 L1.L marspio 3 2 L1.L 1067 m c n um, ii, n., = 2 4 3 HV1.L.L m meconi 3 2 L1.L meconio 3 2 L1.L mecnio 3 2 L1.L 1068 myst r um, ii, n., = 2 4 3 HV1.L.L m misteri 3 2 L1.L mi sterio 3 2 L1.L mistrio 3 2 L1.L 1069 n g t um (n g c um), ii, n. 2 4 3 HV1.L.L m negoci 3 2 L1.L negocio 3 2 L1.L negcio 3 2 L1.L 1070 n t r us ii, m. 2 4 3 HV1.L.L m notar i 3 2 L1.L notario 3 2 L1.L notrio 3 2 L1.L 1071 n v c us (lb. lat. t us ), a, um 2 4 3 HV1.L.L m novici 3 2 L1.L novicio 3 2 L1.L novio 3 2 L1.L 1072 obst c lum, i, n. 2 4 3 HV1.L.L m obstacle 3 2 L1.L obstculo 4 3 L1.L. L obstculo 4 3 L1.L.L 1073 oppr br um (obp-), i. n. 2 4 3 HV1.L.L m oprob i 3 2 L1.L oprobio 420 3 2 L1.L oprbrio 2 L1.L m 1074 r c lum, i, n. 2 4 3 HV1.L.L m oracle 2 2 L1.L orculo 4 3 L1.L. L orculo 4 3 L1.L.L 1075 v r um, ii, n. 2 4 3 HV1.L.L m ovari 3 2 L1.L ovario 3 2 L1.L ovrio 3 2 L1.L 1076 p c l um, ii, n. 2 4 3 HV1.L.L m peculi 3 2 L1.L peculio 3 2 L1.L peclio 3 2 L1.L 1077 perj r um, ii, n. 2 4 3 HV1.L.L m perjuri 3 2 L1.L perjurio 3 2 L1.L perjrio 3 2 L1.L 1078 pinn c lum, i, n. 2 4 3 HV1.L.L m pinacle 3 2 L1. L pinculo 4 3 L1.L.L pinculo 4 3 L1.L.L 1079 portat cum, i, n. 2 4 3 HV1.L. L m portatge 3 2 L1.L portazgo 3 2 HC1.L portagem 3 2 L1.HV 1080 prael d um 2 4 3 HV1.L.L m preludi 3 2 L1. L preludio 3 2 L1.L preldio 3 2 L1.L 1081 praep t um, ii, n. 2 4 3 HV1.L.L m prepuci 3 2 L1.L prepucio 3 2 L1.L prepcio 3 2 L1.L 1082 praes g um, ii, n. 2 4 3 HV1.L.L m presagi 3 2 L1.L presagio 3 2 L1.L pressgio 3 2 L1.L 1083 praet r us, ii, m. 2 4 3 HV1.L.L m pretori 3 2 L1.L pretorio 3 2 L1.L pretrio 3 2 L1.L 1084 pr oem um, ii, n., = 2 4 3 HV1.L.L m proemi 3 2 L1.L proemio 3 2 L1.L promio 3 2 L1.L 1085 proscaen um (prosc n-), ii, n.,= 2 4 3 HV1.L.L m prosceni 3 2 L1.L proscenio 3 2 L1.L proscnio 3 2 L1.L 1086 r s r us, a, um 2 4 3 HV1.L.L m rosari 3 2 L1.L rosario 3 2 L1.L rosrio 3 2 L1.L 1087 s cr r um, ii, n. 2 4 3 HV1.L.L m sagrari 3 2 L1.L sagrario 3 2 L1.L sacrrio 3 2 L1.L 1088 s l r um, ii, n. 2 4 3 HV1.L.L m salari 3 2 L1.L salario 3 2 L1.L salrio 3 2 L1.L 1089 psalt r um, i, n., = 2 4 3 HV1.L.L m salteri 3 2 L1.L sa lterio 3 2 L1.L saltrio 3 2 L1.L

PAGE 421

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 1090 sens r um, ii, n. 2 4 3 HV1.L.L m sensori 32L1. L sensorio 32L.L1.L sensrio 32L.L1.L 1091 s c r us, i, m. 2 4 3 HV1.L.L m sicari 32L1.L sicario 32L1.L sicrio 32L1.L 1092 s l r um, ii, n. 2 4 3 HV1.L.L m solari 32L1.L solario 32L1.L solrio 32L1.L 1093 s d r um, ii, n. 2 4 3 HV1.L.L m sudari 32L1.L sudario 32L1.L sudrio 32L1.L 1094 summ r um i, n. 2 4 3 HV1.L.L m sumari 32 L1.L sumario 32L1.L sumrio 32L1.L 1095 tentac lum 2 4 3 HV1.L.L m tentacle 32L1.L t entculo 43L1.L.L tentculo 43L1.L.L 1096 tern r us a, um 2 4 3 HV1.L.L m ternari 32 L1.L ternario 32L1.L tenreiro 32HV1.L 109 7 tr cl n um ii, n. 2 4 3 HV1.L.L m triclini 3 2 L1.L triclinio 3 2 L1.L triclnio 3 2 L1.L 1098 umbr c lum i, n. 2 4 3 HV1.L.L m umbracle 3 2 L1.L umbrculo 4 3 L1.L. L umbrculo 4 3 L1.L.L 1099 v n b lum i, n. 2 4 3 HV1.L.L m venabl e 3 2 L1.L venablo 3 2 L1.L venbulo 3 2 L1.L 1100 vest g um ii, n. 2 4 3 HV1.L.L m vestigi 3 2 L1.L vestigio 3 2 L1.L vestgio 3 2 L1.L 1101 v c r us a, um 2 4 3 HV1.L.L m vicari 3 2 L1.L vicario 3 2 L1.L vigrio 3 2 L1.L 1102 v c b lum i, n. 2 4 3 HV1.L.L m vocable 3 2 L1.L vocablo 3 2 L1.L vocbulo 4 3 L1.L.L HV1.L.X (prothesis) (2) 1103 scaen r us, a, um, adj. 2 4 3 HV1.L.L m escenari 4 2 L1.L escenario 4 2 L1.L cenrio 3 2 L1.L 1104 spect c lum, i, n. 2 4 3 HV1.L.L m espectacle 4 2 L1.L espectculo 5 3 L1.L.L espectculo 5 3 L1.L.L L1.HV.X (1) 1105 d m tros, i, f. 2 4 3 L1.HV.L m dimetre 4 3 L1.L.L dimetro 3 3 L1.L.L dimetro 4 3 L1.L.L L1.L.X (11) 1106 arb tr um (alt. arb t r um), ii, n. 2 4 3 L1.L.L m albir 2 1 HC1 albedr o 4214 2 L.L1.L alvedrio 4 2 L.L1.L 1107 art c lus i, m., dim. de artus 2 4 3 L1.L.L m artell 2 1 HC1 ar tejo 3 2 L1.L artelho 3 2 L1.L 1108 conc l um ii, n. 2 4 3 L1.L.L m conse ll 2 1 HC1 concejo 3 2 L1.L concelho 3 2 L1.L 1109 cons l um ii, n. 2 4 3 L1.L.L m consell 2 1 HC1 consejo 3 2 L1.L conselho 3 2 L1.L 1110 ex g um ii, n. 2 4 3 L1.L.L m assaig 2 1 HC1 ensayo 3 2 L1.L ensaio 3 2 L1.L 1111 gymn s um, ii, n. 2 4 3 L1.L.L m gimns 2 1 HC1 gimnasio 3 2 L1.L ginsio 3 2 L1.L 1112 g n c lum, i, n. 2 4 3 L1.L.L m geno ll 2 1 HC1 hinojo 3 2 L1.L joelho 3 2 L1.L 1113 lat. vulg. manuc lus 2 4 3 L1.L.L m manoll 2 1 HC1 manojo 3 2 L1.L manolho 3 2 L1.L 1114 p n c um, ii, n. 2 4 3 L1.L.L m pans 2 1 HC1 panizo 3 2 L1.L paino 2 2 HV1.L 1115 r p s tus, a, um 2 4 3 L1.L.L m rebost 2 1 HC1 repuesto 3 2 HC1.L reposte 3 2 HC1.L 1116 *vincic lum 2 4 3 L1.L.L m vencill 2 1 HC1 vencejo 3 2 L1.L vencilho 3 2 L1.L L1.L.X (4th decl) (2) 1117 circ tus, s, m 4 4 3 L1.L.L m circuit 2 1 HC1 circuito 3 2 L1.L circuito 3 2 HV1.L 1118 intr tus, s, m. 4 4 3 L1.L.L m introit 2 1 HC1 introito 3 2 HV1.L intrito 3 2 HV1.L L1.L.X (1) 1119 archa cus, y este del gr. 2 4 3 L1.L.L m arcaic 2 1 HC.HC1 arcaico 3 2 HV1.L arcico 3 2 HV1.L

PAGE 422

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp L1.L.X (3) 1120 c n n cus i, m. 2 4 3 L1.L.L m canonge 32HC1.L cannigo 43L1.L.L cnego 33L1.L.L 1121 d b lus, i, m., del gr. 2 4 3 L1.L.L m diab le 32HC1.L diablo 22L1.L diabo 22L1.L 1122 tr enn um ii, n. 2 4 3 L1.L.L m trienni 32HC1.L trienio 32L1.L trinio 32L1.L L1.L.X (2) 1123 *alen tus, por anhel tus 2 4 3 L1.L.L m al 21HV1 aliento 32HC1.L alento 32HC1.L 1124 LL. mosa cum, i, n. 2 4 3 L1.L.L m mosaic 21 HV1 mosaico 32HV1.L mosaico 32HV1.L L1.L.X (1) 1125 retabulus
PAGE 423

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 1145 aut m tus um (os, on), del gr. 2 4 3 L1.L.L m autmat 3 2 L1.HC autmata 4 3 L1. L.L autmato 4 3 L1.L.L 1146 b ryt n s n, y este del gr. 2 4 3 L1.L.L m barton 3 2 L1.HC bar tono 4 3 L1.L.L bar tono 4 3 L1.L.L 1147 c p t lum i, n. 2 4 3 L1.L.L m captol 3 2 L1.HC captulo 4 3 L1.L.L captulo 4 3 L1.L.L 1148 carn v rus a, um 2 4 3 L1.L.L m carnvor 3 2 L1.HC carnvoro 4 3 L1.L. L carnvoro 4 3 L1.L.L 1149 cosm gr phus i, m., del gr. 2 4 3 L1.L.L m cosmgraf 3 2 L1.HC cosm grafo 4 3 L1.L.L co smgrafo 4 3 L1.L.L 1150 d c l gus i, m., del gr. 2 4 3 L1.L.L m decl eg 3 2 L1.HC declogo 4 3 L1. L.L declogo 4 3 L1.L.L 1151 dec b tus 2 4 3 L1.L.L m decbit 3 2 L1.HC decbito 4 3 L1.L.L decbito 4 3 L1.L.L 115 2 d p s tum, i, n. 2 4 3 L1.L.L m depsit 3 2 L1.HC depsito 4 3 L1. L.L depsito 4 3 L1.L.L 1153 d l gus, i, m., del gr. 2 4 3 L1.L.L m dil eg 3 2 L1.HC dilogo 3 3 L1. L.L dilogo 3 3 L1.L.L 1154 disc b lus, i, m., del gr. 2 4 3 L1.L.L m di scbol 3 2 L1.HC discbolo 4 3 L1.L.L discbolo 4 3 L1.L.L 1155 oec n mus i, m., del gr. 2 4 3 L1.L.L m ecn om 3 2 L1.HC ecnomo 4 3 L1. L.L ecnomo 4 3 L1.L.L 1156 m t cus a, um, del gr. 2 4 3 L1.L.L m emt ic 3 2 L1.HC emtico 4 3 L1. L.L emtico 4 3 L1.L.L 1157 p l gus, i, m. del gr. 2 4 3 L1.L.L m ep leg 3 2 L1.HC eplogo 4 3 L1.L.L eplogo 4 3 L1.L.L 1158 p th ton, i, n. 2 4 3 L1.L.L m eptet 3 2 L1.HC epteto 4 3 L1.L.L epteto 4 3 L1.L.L 1159 aequ v cus, a, um, adj. 2 4 3 L1.L.L m equvoc 3 2 L1.HC equvoco 4 3 L1.L. L equvoco 4 3 L1.L.L 1160 exp s tus a, um 2 4 3 L1.L.L m expsit 3 2 L1.HC expsito 4 3 L1.L.L exposto 2 2 HC1.L 1161 phaen m non, i, n. 2 4 3 L1.L.L m fenomen 3 2 L1.HC fenmeno 423 4 3 L1.L. L fenmeno 4 3 L1.L.L 1162 f l lum, i, n. 2 4 3 L1.L.L m folol 3 2 L1.HC fololo 4 3 L1.L. L fololo 4 3 L1.L.L 1163 hept g nos, on, adj.= 2 4 3 L1.L.L m heptgon 3 2 L1.HC hept gono 4 3 L1.L.L heptgono 4 3 L1.L.L 1164 hex g num, i, n., = 2 4 3 L1.L.L m hexgon 3 2 L1.HC he xgono 4 3 L1.L.L hexgono 4 3 L1.L.L 1165 hippodr mos, i, m., = 2 4 3 L1.L.L m hipdrom 3 2 L1.HC hip dromo 4 3 L1.L.L hi pdromo 4 3 L1.L.L 1166 h l gr phus, a, um, adj., = 2 4 3 L1.L.L m holgraf 3 2 L1.HC hol grafo 4 3 L1.L.L hol grafo 4 3 L1.L.L 1167 m n p lus, i, m. 2 4 3 L1.L.L m manpul 3 2 L1.HC manpulo 4 3 L1.L.L manpulo 4 3 L1.L.L 1168 m ch n cus, a, um = 2 4 3 L1.L.L m mecnic 32L1.HC mec nico 43L1.L.L mecnico 43L1.L.L 1169 m n g mus, i, m., = 2 4 3 L1.L.L m mongam 32L1.HC m ongamo 43L1.L.L m ongamo 43L1.L.L 1170 n phytus, a, um = 2 4 3 L1.L.L m nefit 32L1.HC nef ito 43L1.L.L nef ito 43L1.L.L 1171 pneum t cus a, um, del gr. 2 4 3 L1.L.L m neumtic 32L1.HC neumt ico 43L1.L.L pneum tico 43L1.L.L 1172 omn v rus, a, um, adj. 2 4 3 L1.L.L m omnvor 32L1.HC omnvoro 4 L1.L.L omnvoro 4 L1.L.L

PAGE 424

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 1173 paragr phus 2 4 3 L1.L.L m pargraf 32L1.HC par grafo 43L1.L.L pargrafo 43L1.L.L 1174 p r s tus, i, m., = 2 4 3 L1.L.L m parsit 3 2 L1.HC par sito 4 3 L1.L.L par asito 4 2 L.L1.L 1175 p t b lum, i, n. 2 4 3 L1.L.L m patbul 3 2 L1.HC patbulo 4 3 L1.L. L patbulo 4 3 L1.L.L 1176 pragm t cus, i, m., = 2 4 3 L1.L.L m pragmtic 3 2 L1.HC pragmtico 4 3 L1. L.L pragmtico 4 3 L1.L.L 1177 praet r tus, a, um 2 4 3 L1.L.L m pretrit 3 2 L1.HC pr etrito 4 3 L1.L.L pr etrito 4 3 L1.L.L 1178 pr p s tum, i, n. 2 4 3 L1.L.L m propsit 3 2 L1.HC propsito 4 3 L1.L. L propsito 4 3 L1.L.L 1179 pr st b lum, i, n. 2 4 3 L1.L.L m pr ostbul 3 2 L1.HC prostbulo 4 3 L1.L.L prost bulo 4 3 L1.L.L 118 0 s cr l gus, a, um 2 4 3 L1.L.L m sacrleg 3 2 L1. HC sacrlego 4 3 L1.L.L sacrlego 4 3 L1.L.L 1181 sarc ph gus, i, m. 2 4 3 L1.L.L m sarcfag 3 2 L1.HC sarcfago 4 3 L1.L. L sarcfago 4 3 L1.L.L 1182 syc m rus, i, f.< 2 4 3 L1.L.L m sicmor 3 2 L1.HC sic moro 4 3 L1.L.L sicmoro 4 3 L1.L.L 1183 somn fer ra, rum 2 4 3 L1.L.L m somnfer 3 2 L1.HC somnfero 4 3 L1.L.L sonfero 4 3 L1.L.L 1184 supp s tus, a, um 2 4 3 L1.L.L m supsit 3 2 L1.HC supuesto 3 2 HC1.L suposto 3 2 HC1.L 1185 tard gr dus a, um 2 4 3 L1.L.L m tardgrad 3 2 L1. HC tardgrado 4 3 L1.L.L tardgrado 4 3 L1.L.L 1186 th l gus i, m., del gr. 2 4 3 L1.L.L m tel eg 3 2 L1.HC telogo 4 3 L1. L.L telogo 4 3 L1.L.L 1187 t r b lum (th r), i, n. 2 4 3 L1.L.L m turbul 3 2 L1.HC turbulo 4 3 L1.L. L turbulo 4 3 L1.L.L 1188 vest b lum i, n. 2 4 3 L1.L.L m vestbul 3 2 L1.HC vestbulo 424 4 3 L1.L. L vestbulo 4 3 L1.L.L 1189 zodi cus, y este del gr. 2 4 3 L1.L.L m zod ac 3 2 L1.HC zodiaco 3 2 L1.L zodaco 3 2 L1.L L1.L.X (64) 1190 amphib us, y este del gr. 2 4 3 L1.L.L m amfi bi 3 2 L1.L anfibio 3 2 L1.L anfbio 3 2 L1.L 1191 artic lus 2 4 3 L1.L.L m article 3 2 L1.L artculo 4 3 L1.L.L artculo 4 3 L1.L.L 1192 augur um 2 4 3 L1.L.L m auguri 3 2 L1. L agero 3 2 L1.L augrio 3 2 L1.L 1193 aug r um ii, n. 2 4 3 L1.L.L m auguri 3 2 L1.L augurio 3 2 L1.L augrio 3 2 L1.L 1194 ausp c um ii, n. 2 4 3 L1.L.L m auspici 3 2 L1.L auspicio 3 2 L1.L auspcio 3 2 L1.L 1195 auxil um 2 4 3 L1.L.L m auxili 3 2 L1. L auxilio 3 2 L1.L auxlio 3 2 L1.L 1196 coen b um ii, n., y este del gr. 2 4 3 L1.L.L m cenobi 3 2 L1.L cenobio 3 2 L1.L cenbio 3 2 L1.L 1197 c l c um ii, n 2 4 3 L1.L.L m cilici 3 2 L1.L cilicio 3 2 L1.L cilcio 3 2 L1.L 1198 cym t um o -on ii, n., del gr. 2 4 3 L1.L.L m cimaci 3 2 L1.L cimacio 3 2 L1.L cimcio 3 2 L1.L 1199 collyr um ii, n., del gr. 2 4 3 L1.L.L m co lliri 3 2 L1.L colirio 3 2 L1.L colrio 3 2 L1.L 1200 conc l um ii, n. 2 4 3 L1.L.L m concili 3 2 L1.L concilio 3 2 L1.L conclio 3 2 L1.L 1201 cons c us a, um 2 4 3 L1.L.L m consoci 3 2 L1.L consocio 3 2 L1.L conscio 3 2 L1.L 1202 cons cer, ri, m. 2 4 3 L1.L.L m consogre 3 2 L1. L consuegro 3 2 L1.L consogro 3 2 L1.L 1203 cont n us i, m. 2 4 3 L1.L.L m continu 3 2 L1.L continuo 3 2 L1. L contnuo 3 2 L1.L

PAGE 425

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 1204 disc p lus i, m. 2 4 3 L1.L.L m deixeble 32 L1.L discpulo 43L1.L. L discpulo 43L1.L.L 1205 d l qu um ii, n. 2 4 3 L1.L.L m deliqui 32L1.L deliquio 32L1.L delquio 32L1.L 1206 d l v um ii, n. 2 4 3 L1.L.L m diluvi 32L1.L diluvio 32L1.L dilvio 32L1.L 1207 effl v um ii, n. 2 4 3 L1.L.L m efluvi 32L1.L efluvio 32L1.L eflvio 32L1.L 1208 l g um ii, n. 2 4 3 L1.L.L m elogi 32 L1.L elogio 32L1.L elogio 32L1.L 1209 emp r um, ii, n. 2 4 3 L1.L.L m empori 32 L1.L emporio 32L1.L emprio 32L1.L 1210 r c us, ii, m. 2 4 3 L1.L.L m erio 32L1.L erizo 32L1.L ourio 32L1.L 121 1 armen us [mus] 2 4 3 L1.L.L m ermini 3 2 L1.L armio 3 2 L1.L arminho 3 2 L1.L 1212 ex l um, ii, n. 2 4 3 L1.L.L m exili 3 2 L1.L exilio 3 2 L1.L exlio 3 2 L1.L 1213 fasc c lus, i, m. 2 4 3 L1.L.L m fascicle 3 2 L1.L fascculo 4 3 L1.L. L fascculo 4 3 L1.L.L 1214 foll c lus, i, m. 2 4 3 L1.L.L m follicle 3 2 L1.L folculo 4 3 L1.L. L folculo 4 3 L1.L.L 1215 g r n on o -um, ii, n. 2 4 3 L1.L.L m gerani 3 2 L1.L geranio 3 2 L1.L gernio 3 2 L1.L 1216 hosp t um, i, n. 2 4 3 L1.L.L m hospici 3 2 L1.L hospicio 3 2 L1.L hospcio 3 2 L1.L 1217 impl v um, i, n. 2 4 3 L1.L.L m impluvi 3 2 L1. L impluvio 3 2 L1.L implvio 3 2 L1.L 1218 ind c um, i, n. 2 4 3 L1.L.L m indici 3 2 L1.L indicio 3 2 L1.L indcio 3 2 L1.L 1219 ing n um, ii, n. 2 4 3 L1.L.L m ingeni 3 2 L1.L ingenio 3 2 L1.L engenho 3 2 L1.L 1220 n t um, i, n. 2 4 3 L1.L.L m inici 3 2 L1. L inicio 3 2 L1.L incio 3 2 L1.L 1221 j d c um ii, n. 2 4 3 L1.L.L m judici 3 2 L1.L juicio 2 2 L1.L juzo 3 2 L1.L 1222 l t g um, i, n. 2 4 3 L1.L.L m litigi 3 2 L1.L litigio 425 3 2 L1.L litgio 3 2 L1.L 1223 martyr um, i, n., = 2 4 3 L1.L.L m martiri 3 2 L1.L mart irio 3 2 L1.L martrio 3 2 L1.L 1224 Merc r us, ii, m. 2 4 3 L1.L.L m mercuri 3 2 L1.L mercurio 3 2 L1.L mercrio 3 2 L1.L 1225 m n c lus, i, m. 2 4 3 L1.L.L m monocle 3 2 L1.L monculo 4 3 L1.L. L monculo 4 3 L1.L.L 1226 mont c lus, i, m. 2 4 3 L1.L.L m monticle 3 2 L1.L montculo 4 3 L1.L.L montculo 4 3 L1.L.L 1227 naufr g um, ii, n. 2 4 3 L1.L.L m naufragi 3 2 L1.L naufragio 3 2 L1.L naufrgio 3 2 L1.L 1228 obs qu um, ii, n. 2 4 3 L1.L.L m obsequi 3 2 L1.L obsequio 3 2 L1.L obsquio 3 2 L1.L 1229 off c um, ii, n. 2 4 3 L1.L.L m ofici 3 2 L1.L oficio 3 2 L1.L ofcio 3 2 L1.L 1230 p tr c us, a, um, adj. 2 4 3 L1.L.L m patrici 3 2 L1.L patricio 3 2 L1.L patrcio 3 2 L1.L 1231 b. lat. petr l um 2 4 3 L1.L.L m petroli 3 2 L1.L pet rleo 4 3 L1.L.L petrleo 4 3 L1.L.L 1232 praes d um, ii, n. 2 4 3 L1.L.L m presidi 3 2 L1.L presidio 3 2 L1.L presdio 3 2 L1.L 1233 praestig um, ii, n. 2 4 3 L1.L.L m prestigi 3 2 L1.L prestigio 3 2 L1.L prestgio 3 2 L1.L 1234 princ p um, ii, n. 2 4 3 L1.L.L m principi 3 2 L1.L principio 3 2 L1.L princpio 3 2 L1.L 1235 pr d g um, ii, n. 2 4 3 L1.L.L m prodigi 3 2 L1.L prodigio 3 2 L1. L prodgio 3 2 L1.L 1236 qu dr v um ii, n. 2 4 3 L1.L.L m quadrivi 3 2 L1.L cuadrivio 3 2 L1.L quadrvio 3 2 L1.L 1237 qu dr v um ii, n. 2 4 3 L1.L.L m quadrivi 3 2 L1.L quadrvium 3 2 L1.HC quadrvio 3 2 L1.L 1238 r f g um, ii, n. 2 4 3 L1.L.L m refugi 3 2 L1.L refugio 3 2 L1. L refgio 4 3 L1.L.L

PAGE 426

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 1239 r s d us, a, um 2 4 3 L1.L.L m residu 32L1.L residuo 32L1.L resduo 32L1.L 1240 r t c lum, i, n. 2 4 3 L1.L.L m reticle 32 L1.L retculo 43L1.L.L retculo 43L1.L.L 1241 serv t um, ii, n. 2 4 3 L1.L.L m servici 32 L1.L servicio 32L1.L servio 32L1.L 1242 solst t um ii, n. 2 4 3 L1.L.L m solstici 32L1.L solsticio 32L1.L solstcio 32L1.L 1243 subs d um, ii, n. 2 4 3 L1.L.L m subsidi 32 L1.L subsidio 32L1.L subsdio 32L1.L 1244 suffr g um ii, n. 2 4 3 L1.L.L m sufragi 32L1.L sufragio 32L1.L sufrgio 32L1.L 1245 suppl c um, ii, n. 2 4 3 L1.L.L m suplici 32 L1.L suplicio 32L1.L suplcio 32L1.L 124 6 test c lus i, m. 2 4 3 L1.L.L m testicle 3 2 L1.L testculo 4 3 L1.L. L testculo 4 3 L1.L.L 1247 t paz us i, f., del gr.; t paz on ii, m., 2 4 3 L1.L.L m topazi 3 2 L1.L t opacio 3 2 L1.L topzio 3 2 L1.L 1248 tr f l um ii, n. 2 4 3 L1.L.L m trifoli 3 2 L1.L trifolio 3 2 L1.L triflio 3 2 L1.L 1249 b. lat. trisag um, del gr. bizant. 2 4 3 L1.L.L m trisagi 3 2 L1.L tr isagio 3 2 L1.L trisgio 4 3 L1.L.L 1250 t g r um, ii, n. 2 4 3 L1.L.L m tuguri 3 2 L1. L tugurio 3 2 L1.L tugrio 3 2 L1.L 1251 v h c lum i, n. 2 4 3 L1.L.L m vehicle 3 2 L1. L vehculo 4 3 L1.L. L veculo 4 3 L1.L.L 1252 ventr c lus, i, m. 2 4 3 L1.L.L m ventricle 3 2 L1.L ventrculo 4 3 L1.L.L ventr culo 4 3 L1.L.L 1253 vers c lus i, m. 2 4 3 L1.L.L m versicle 3 2 L1.L versculo 4 3 L1.L. L versculo 4 3 L1.L.L L1.L.X (1) 1254 VL iinip rus < j n p rus, i, f. 4 4 3 L1.L.L m ginebre 3 2 L1. L enebro 426 3 2 L1.L zimbro 2 2 L1 L1.L.X (1) 1255 curr c lum i, n. 2 4 3 L1.L.L m currculum 4 3 L1.L.HC currculo 4 3 L1. L.L currculo 4 3 L1.L.L L1.L.X (6) 1256 hex m ter, tri, m., = 2 4 3 L1.L.L m hexmetre 4 3 L1.L.L he xmetro 4 3 L1.L.L hexmetro 4 3 L1.L.L 1257 oct dros, i, m. and f., = 2 4 3 L1.L.L m octedre 4 3 L1.L.L octaedro 4 2 L. L1.L octaedro 4 2 L.L1.L 1258 pent m ter, tri, m. 2 4 3 L1.L.L m pentmetre 4 3 L1.L.L pentmetro 4 3 L1.L.L pentmetro 4 3 L1.L.L 1259 p r m tros, i, f. 2 4 3 L1.L.L m permetre 4 3 L1.L.L permetro 4 3 L1.L.L permetro 4 3 L1.L.L 1260 p r dus, i, f. 2 4 3 L1.L.L m perode 4 3 L1.L.L perodo 4 3 L1.L. L perodo 4 3 L1.L.L 1261 qu dr p dus, a, um 2 4 3 L1.L.L m quadrpede 4 3 L1.L.L cuadrpedo 4 3 L1.L. L quadrpede 3 3 L1.L.L HC.HC1.X (1) 1262 j risconsultus i, m. 2 5 2 HC.HC1.L m jurisconsult 4 1 HC.HC1 jurisconsulto 5 2 HC.HC1.L jurisconsulto 5 2 HC.HC1.L HC.HC1.X (1) 1263 anabaptismus, del gr. 2 5 2 HC.HC1.L m anabaptisme 5 2 HC.HC1.L anabaptismo 5 2 HC.HC1.L anabaptismo 5 2 HC.HC1.L HV.HC1.X (4) 1264 exp r mentum, i, n. 2 5 2 HV.HC1.L m experiment 4 1 L.HC1 experimento 5 2 L.HC1.L experimento 5 2 L.HC1.L

PAGE 427

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 1265 imp d mentum (inp-), i, n. 2 5 2 HV.HC1 .Lm impediment 41L.HC1 impediment o 52L.HC1.L impedimento 52L.HC1.L 1266 praed c mentum, i, n. 2 5 2 HV.HC1.Lm predicament 41L.HC1 predicamento 52L.HC1.L predicamento 52L.HC1.L 1267 temp r mentum i, n. 2 5 2 HV.HC1.L m temperament 4 1 L.HC1 temperamento 5 2 L.HC1.L temperamento 5 2 L.HC1.L HV.HC1.X (1) 1268 d tismus, i, m. 2 5 2 HV.HC1.L m idiotisme 5 2 L.HC1.L idiotismo 4 2 L.L1.L idiotismo 4 2 L.L1.L HV.HV1.X (6) 1269 dim n t vus 2 5 2 HV.HV1.L m diminutiu 4 1 L.HV1 diminutivo 5 2 L.L1.L diminutivo 5 2 L.L1.L 1270 imp r t vus, a, um, adj. 2 5 2 HV.HV1.L m imperatiu 4 1 L.HV1 imperativo 5 2 L. L1.L imperativo 5 2 L.L1.L 1271 inf n t vus, a, um 2 5 2 HV.HV1.L m infinitiu 4 1 L.HV1 infinitivo 5 2 L.L1.L infinitivo 5 2 L.L1.L 1272 n m n t vus a, um, adj. 2 5 2 HV.HV1.L m nominatiu 4 1 L.HV1 nominativo 5 2 L. L1.L nominativo 5 2 L.L1.L 1273 praed c t vus, a, um 2 5 2 HV.HV1.L m predicatiu 4 1 L. HV1 predicativo 5 2 L.L1. L predicativo 5 2 L.L1.L 1274 s perl t vus, i, m. 2 5 2 HV.HV1.L m superlatiu 4 1 L. HV1 superlativo 5 2 L.L1.L superlativo 5 2 L.L1.L L.HC1.X (2) 1275 additamentum 2 5 2 L.HC1.L m additament 4 1 L.HC1 aditamento 5 2 L.HC1.L aditamento 5 2 L.HC1.L 1276 emolumentum 2 5 2 L.HC1.L m emolument 4 1 L.HC1 emolumento 5 2 L.HC1.L emolumento 5 2 L.HC1.L L.HC1.X (2) 1277 ph l s phaster, tri, m. 2 5 2 L.HC1.L m filosofastre 5 2 L. HC1.L filosofastro 5 2 L.HC1.L filosofastro 5 2 L.HC1.L 127 8 rheum tismus, i, m., = 2 5 2 L.HC1.L m reumatisme 4 2 L.HC1.L r eumatismo 4 2 L.HC1.L reumatismo 4 2 L.HC1.L L.HV1.X (2) 1279 content sus a, um 2 5 2 L.HV1.L m contencis 4 1 L.HC1 contencioso 427 4 2 L1. L contencioso 4 2 L1.L 2 5 2 L.HV1.L m triumvirat 4 1 L.HC1 tr iunvirato 4 2 L.L1.L tr iunvirato 5 2 L.L1.L L.HV1.X (4) 1281 accusat vus 2 5 2 L.HV1.L m acusatiu 4 1 L.HV1 acusativo 5 2 L.L1.L acusativo 5 2 L.L1.L 1282 aperit vus 2 5 2 L.HV1.L m aperitiu 4 1 L.HV 1 aperitivo 5 2 L.L1.L aperitivo 5 2 L.L1.L 1283 m r d nus, a, um 2 5 2 L.HV1.L m meridi 4 1 L.HV1 meridiano 4 2 L1.L meridiano 4 2 L1.L 1284 p r t n um, i, n., = 2 5 2 L.HV1.L m peritoneu 4 1 L.HV1 peritoneo 5 2 L.L1.L peritoneu 4 1 L.HV1 L.HV1.X (1) 1285 amphithe trum, i, n. 2 5 2 L.HV1.L m amfiteatre 5 2 L.L1.L anfiteatro 5 2 L.L1.L anfit eatro 5 2 L.L1.L L.HV1.X (1) 1286 d s d r tus a, um 2 5 2 L.HV1.L m desidertum 5 2 L.L1.HC desidertum 5 2 L.L1.HC desiderato 5 2 L.L1.L HC1.L.X (2) 1287 m n syll bus, a, um = 2 5 3 HC1.L.L m monosllab 4 2 HC1.HC monoslabo 5 3 L1.L.L m onosslabo 5 3 L1.L.L 1288 oct syll bus, a, um 2 5 3 HC1.L.L m octosllab 4 2 HC1 .HC octoslabo 5 3 L1.L. L octosslabo 5 3 L1.L.L HC1.L.X (7)

PAGE 428

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 1289 rheubarb rum 2 5 3 HC1.L.L m ruibarbre 32HC1.L ruibarbo 32HC1.L ruibarbo 32HC1.L 1290 con-t bern um ii, n. 2 5 3 HC1.L.L m contuberni 42 HC1.L contubernio 42HC1.L contubrnio 42HC1.L 1291 aequ noct um, i, n. 2 5 3 HC1.L.L m equinocci 42HC1.L equinoccio 42HC1.L equincio 42L1.L 1292 internunt us, i, m. 2 5 3 HC1.L.L m internunci 42HC1.L internuncio 42HC1.L internncio 42HC1.L 1293 l t fund um, ii, n. 2 5 3 HC1.L.L m latifundi 42HC1.L latifundio 4 2 HC1.L latifndio 4 2 HC1.L 1294 s m circ lus, a, um 2 5 3 HC1.L.L m semicercle 4 2 HC1.L semicrculo 5 3 HC1.L.L semicrculo 5 3 HC1.L.L 1295 p r ost on, i, n., = 2 5 3 HC1.L.L m periosti 4 2 HC1.L periostio 3 2 HC1.L peristeo 4 3 HC1.L.L HV1.L.X HC1 (8) 1296 camarar us, camerar us 2 5 3 HV1.L.L m cambrer 2 1 HC1 camarero 4 2 L1.L camareiro 4 2 HV1.L 1297 F br r us, ii, m 2 5 3 HV1.L.L m Febrer 2 1 HC1 febrero 3 2 L1.L Fevereiro 4 2 HV1.L 1298 lat. vulg. jenuar us, lat. J n r us, a, um 2 5 3 HV1.L.L m gener 2 1 HC1 enero 3 2 L1.L Janeiro 3 2 HV1.L 1299 p r r us, ii, m. 2 5 3 HV1.L.L m obrer 2 1 HC1 obrero 3 2 L1.L obreiro 3 2 HV1.L 1300 *p s t c us, a, um, de pos tus 2 5 3 HV1.L.L m posts 2 1 HC1 postizo 3 2 L1.L postio 3 2 L1.L 1301 s m t r us, a, um 2 5 3 HV1.L.L m sender 2 1 HC1 sendero 3 2 L1.L sendeiro 3 2 HV1.L 1302 s l t r us a, um 2 5 3 HV1.L.L m solter 2 1 HC1 soltero 3 2 L1.L solteiro 3 2 HV1.L 1303 tert r us, a, um 2 5 3 HV1.L.L m tercer 2 1 HC1 tercero 3 2 L1.L terceiro 3 2 HV1.L HV1.L.X (4) 1304 mill n r us a, um 2 5 3 HV1.L.L m milenar 3 1 HC1 milenario 428 4 2 L1.L milenrio 4 2 L1.L 1305 m nast r um, ii, n., = 2 5 3 HV1.L.L m monestir 3 1 HC1 monas terio 4 2 L1.L mosteiro 3 2 HV1.L 1306 s l n r us, a, um 2 5 3 HV1.L.L m saliner 3 1 HC1 salinero 4 2 L1.L salineiro 4 2 HV1.L 1307 t bern r us a, um, adj. 2 5 3 HV1.L.L m taverner 3 1 HC1 tabernero 4 2 L1. L taberneiro 4 2 HV1.L HV1.L.X (4) 1308 aest r um, i, n. 2 5 3 HV1.L.L m estuari 4 2 L1.L estuario 3 2 L1.L esturio 3 2 L1.L 1309 prompt r um, i, n. 2 5 3 HV1.L.L m promptuari 4 2 L1.L prontuario 3 2 L1.L pronturio 3 2 L1.L 1310 sr us a, um 2 5 3 HV1.L.L m usuari 4 2 L1.L usuario 3 2 L1.L usurio 3 2 L1.L 1311 infort n um, i, n. 2 5 3 HV1.L.L m infortuni 4 2 L1. L infortunio 4 2 L1.L infortnio 4 2 L1.L HV1.L.X (1) 1312 cons l r us i, m. 2 5 3 HV1.L.L m consiliari 5 2 L. L1.L consil iario 4 2 L1.L concilirio 4 2 L1.L HV1.L.X (2) 1313 c t ch m nus i, m., del gr. 2 5 3 HV1.L.L m catecumen 4 2 L1.HC ca tecmeno 5 3 L1.L.L catecmeno 5 3 L1.L.L 1314 d r t cus a, um, del gr. 2 5 3 HV1.L.L m dirt ic 4 2 L1.HC diurtico 4 3 L1. L.L diurtico 4 3 L1.L.L HV1.L.X (73) 1315 act r us, i, m. 2 5 3 HV1.L.L m actuari 4 2 L1. L actuario 3 2 L1.L acturio 3 2 L1.L

PAGE 429

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 1316 adiutor um 2 5 3 HV1.L.L m adj utori 42L1.L adjutorio 42L1.L adjutrio 42L1.L 1317 ant qu r us a, um 2 5 3 HV1.L.L m antiquari 42L1.L anticuario 42L1.L antiqurio 4 2 L1.L 1318 aud t r um ii, n. 2 5 3 HV1.L.L m auditori 4 2 L1.L auditorio 4 2 L1. L auditrio 4 2 L1.L 1319 baln r us, a, um 2 5 3 HV1.L.L m balneari 4 2 L1.L balneario 4 2 L1.L balnerio 4 2 L1.L 1320 baptist r um ii, n., del gr. 2 5 3 HV1.L.L m baptisteri 4 2 L1.L baptisterio 4 2 L1.L baptistrio 4 2 L1.L 1321 best r us ii, m. 2 5 3 HV1.L.L m bestiari 4 2 L1.L bestiario 3 2 L1.L bestirio 3 2 L1.L 1322 br v r um ii, n. 2 5 3 HV1.L.L m breviari 4 2 L1.L breviario 3 2 L1.L brevirio 3 2 L1.L 132 3 K lend r um (Cal), ii, n. 2 5 3 HV1.L.L m cale ndari 4 2 L1.L calendario 4 2 L1.L calendrio 4 2 L1.L 1324 C p t l um ii, n., 2 5 3 HV1.L.L m capitoli 4 2 L1.L capitolio 4 2 L1.L capitlio 4 2 L1.L 1325 chart l r us ii, m. 2 5 3 HV1.L.L m cartulari 4 2 L1.L cartulario 4 2 L1. L cartulrio 4 2 L1.L 1326 coem t r um ii, n., del gr. 2 5 3 HV1.L.L m cementiri 4 2 L1.L ce menterio 4 2 L1.L cemitrio 4 2 L1.L 1327 cent n r us a, um 2 5 3 HV1.L.L m centenari 4 2 L1.L centenario 4 2 L1. L centenrio 4 2 L1.L 1328 c lumb r um ii, n. 2 5 3 HV1.L.L m columbari 4 2 L1.L columbario 4 2 L1.L columbrio 4 2 L1.L 1329 b. lat. commissar us 2 5 3 HV1.L.L m comissari 4 2 L1. L comisario 4 2 L1.L comissrio 4 2 L1.L 1330 consist r um ii, n. 2 5 3 HV1.L.L m consistori 4 2 L1.L consistorio 4 2 L1.L consistrio 4 2 L1.L 1331 consultor us 2 5 3 HV1.L.L m consultori 4 2 L1.L consultorio 4 2 L1.L consultrio 4 2 L1.L 1332 c roll r um ii, n. 2 5 3 HV1.L.L m corollari 4 2 L1.L corolario 429 4 2 L1.L corolrio 4 2 L1.L 1333 l t r um ii, n. 2 5 3 HV1.L.L m elateri 4 2 L1.L elaterio 4 2 L1.L elatrio 4 2 L1.L 1334 miss r us ii, m. 2 5 3 HV1.L.L m emissari 4 2 L1.L emisario 4 2 L1.L emissrio 4 2 L1.L 1335 aequ l br um, ii, n. 2 5 3 HV1.L.L m equilibri 4 2 L1.L equilibrio 4 2 L1. L equilbrio 4 2 L1.L 1336 fr tr c d um, ii, n. 2 5 3 HV1.L.L m fratricidi 4 2 L1.L fratricidio 4 2 L1. L fratricdio 4 2 L1.L 1337 fr ment rius, ii, m. 2 5 3 HV1.L.L m frumentari 4 2 L1.L frumentario 4 2 L1.L frumentrio 4 2 L1.L 1338 gent l t us, a, um 2 5 3 HV1.L.L m gentilici 4 2 L1.L gentilicio 4 2 L1. L gentilcio 4 2 L1.L 1339 h b t c lum, i, n. 2 5 3 HV1.L. L m habitacle 4 2 L1.L habitcul o 5 3 L1.L.L habit culo 5 3 L1.L.L 1340 h misphaer um, i, n. 2 5 3 HV1.L.L m hemisferi 4 2 L1. L hemisferio 4 2 L1.L hemisfrio 4 2 L1.L 1341 h m c d um, i, n. 2 5 3 HV1.L.L m homicidi 4 2 L1. L homicidio 4 2 L1.L homicdio 4 2 L1.L 1342 h n r r um, i, n. 2 5 3 HV1.L.L m honorari 4 2 L1.L honorario 4 2 L1.L honorrio 4 2 L1.L 1343 inc n b la, rum, n. 2 5 3 HV1.L.L m incunable 4 2 L1. L incunable 4 2 L1.L incunbulo 5 3 L1.L.L 1344 ind v d us, a, um 2 5 3 HV1.L.L m individu 4 2 L1.L individuo 4 2 L1.L indivduo 4 2 L1.L 1345 invent r um, i, n. 2 5 3 HV1.L.L m inventari 4 2 L1. L inventario 4 2 L1.L inventrio 4 2 L1.L 1346 l p d r us, ii, m. 2 5 3 HV1.L.L m lapidari 4 2 L1.L lapidario 4 2 L1.L lapidrio 4 2 L1.L 1347 m tr m n um, ii, n. 2 5 3 HV1.L.L m matrimoni 4 2 L1.L matrimonio 4 2 L1.L matrimnio 4 2 L1.L 1348 merc n r us, a, um 2 5 3 HV1.L.L m mercenari 4 2 L1.L mercenario 4 2 L1.L mercenrio 4 2 L1.L 1349 m n t r us, a, um 2 5 3 HV1.L.L m monetari 4 2 L1.L monetario 4 2 L1. L monetrio 4 2 L1.L

PAGE 430

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 1350 m n p l um, ii, n. = 2 5 3 HV1.L.L m monopoli 42L1.L m onopolio 42L1.L monoplio 42L1.L 1351 n t l tius, a, um 2 5 3 HV1.L.L m natalici 42L1.L natalicio 42L1. L natalcio 42L1.L 1352 n t t r um, i, n 2 5 3 HV1.L.L m natatori 4 2 L1.L natatorio 4 2 L1.L natatrio 4 2 L1.L 1353 n m r r us, ii, m. 2 5 3 HV1.L.L m numerari 4 2 L1.L numerario 4 2 L1.L numerrio 4 2 L1.L 1354 offert r um, ii, n. 2 5 3 HV1.L.L m ofertori 4 2 L1.L ofertorio 4 2 L1.L ofertrio 4 2 L1.L 1355 p r r us, ii, m. 2 5 3 HV1.L.L m operari 4 2 L1.L operario 4 2 L1.L operrio 4 2 L1.L 1356 r t r um, ii. n. 2 5 3 HV1.L.L m oratori 4 2 L1.L oratorio 4 2 L1.L oratrio 4 2 L1.L 135 7 ord n r us, i, m. 2 5 3 HV1.L.L m ordinari 4 2 L1.L ordinario 4 2 L1.L ordinrio 4 2 L1.L 1358 ost r us, ii, m. 2 5 3 HV1.L.L m ostiari 4 2 L1.L ostiario 3 2 L1.L ostirio 3 2 L1.L 1359 parr c d um, ii, n. 2 5 3 HV1.L.L m parricidi 4 2 L1.L parricidio 4 2 L1.L parricdio 4 2 L1.L 1360 p tr m n um, ii, n. 2 5 3 HV1.L.L m patrimoni 4 2 L1. L patrimonio 4 2 L.L1. L patrimnio 4 2 L.L1.L 1361 p t t r um, ii, n. 2 5 3 HV1.L.L m petitori 4 2 L1.L petitorio 4 2 L1.L petitrio 4 2 L1.L 1362 pl g r us, ii, m. 2 5 3 HV1.L.L m plagiari 4 2 L1.L plagiario 3 2 L1.L plagirio 3 2 L1.L 1363 pl n t r us, ii, m. 2 5 3 HV1.L.L m planetari 4 2 L1.L planetario 4 2 L1. L planetrio 4 2 L1.L 1364 pl n l n um, ii, n. 2 5 3 HV1.L.L m pleniluni 4 2 L1.L plenilunio 4 2 L1.L plenilnio 4 2 L1.L 1365 pr v l g um, ii, n. 2 5 3 HV1.L.L m privilegi 4 2 L1. L privilegio 4 2 L1.L privilgio 4 2 L1.L 1366 pr l t r us, i, m. 2 5 3 HV1.L.L m pr oletari 4 2 L1.L proletario 4 2 L1.L proletrio 4 2 L1.L 1367 purg t r us, a, um 2 5 3 HV1.L.L m purgatori 4 2 L1.L purgatorio 430 4 2 L1.L purgatrio 4 2 L1.L 1368 qu tern r us a, um 2 5 3 HV1.L.L m quaternari 4 2 L1.L cuaternario 4 2 L1. L quaternrio 4 2 L1.L 1369 r cept c lum, i, n. 2 5 3 HV1.L.L m receptacle 4 2 L1.L receptculo 5 3 L1.L. L receptculo 5 3 L1.L.L 1370 r fect r us, a, um 2 5 3 HV1.L.L m refectori 4 2 L1.L refectorio 4 2 L1.L refeitrio 4 2 L1.L 1371 r pert r um, ii, n. 2 5 3 HV1.L.L m repertori 4 2 L1.L repertorio 4 2 L1.L repertrio 4 2 L1.L 1372 r spons ria, rum, n. 2 5 3 HV1.L.L m responsori 4 2 L1.L responsorio 4 2 L1.L responsrio 4 2 L1.L 1373 s cerd t um, ii, n. 2 5 3 HV1.L.L m sacerdoci 4 2 L1. L sacerdocio 4 2 L1.L sacerdcio 4 2 L1.L 1374 s gitt r us, ii, m. 2 5 3 HV1.L.L m sagitari 4 2 L1.L sagitario 4 2 L1.L sagitrio 4 2 L1.L 1375 sanct r um, ii, n. 2 5 3 HV1.L.L m santuari 4 2 L1.L santuario 3 2 L1.L santurio 3 2 L1.L 1376 secretar us< s cr t r um, ii, n. 2 5 3 HV1.L.L m secretari 4 2 L1.L se cretario 4 2 L1.L secretrio 4 2 L1.L 1377 s m n r um, ii, n. 2 5 3 HV1.L.L m seminari 4 2 L1.L seminario 4 2 L1.L seminrio 4 2 L1.L 1378 sept n r us, a, um 2 5 3 HV1.L.L m septenari 4 2 L1.L septenario 4 2 L1. L septenrio 4 2 L1.L 1379 s l t r us a, um 2 5 3 HV1.L.L m solitari 4 2 L1.L solitario 4 2 L1. L solitrio 4 2 L1.L 1380 sustent c lum i, n. 2 5 3 HV1.L.L m sustentacle 4 2 L1.L sustentculo 5 3 L1.L.L sustent culo 5 3 L1.L.L 1381 t bern c lum i, n. 2 5 3 HV1.L.L m tabernacle 4 2 L1. L tabernculo 5 3 L1.L.L tabernculo 5 3 L1.L.L 1382 t n br r us a, um, adj. 2 5 3 HV1.L.L m tenebrar i 4 2 L1.L tenebrario 4 2 L1.L tenebrrio 4 2 L1.L 1383 tert r um ii, n. 2 5 3 HV1.L.L m terciari 4 2 L1.L terciario 3 2 L1.L tercirio 3 2 L1.L 1384 terr t r um ii, n. 2 5 3 HV1.L.L m territori 4 2 L1.L territorio 4 2 L1.L territrio 4 2 L1.L

PAGE 431

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 1385 test m n um ii, n. 2 5 3 HV1.L.L m testimoni 4 2 L1.L testimonio 4 2 L1. L testemunho 4 2 L1.L 1386 v lunt r us a, um 2 5 3 HV1.L.L m voluntari 4 2 L1.L voluntario 4 2 L1.L voluntrio 4 2 L1.L 1387 v m t r us a, um 2 5 3 HV1.L.L m vomitori 4 2 L1.L vomitorio 4 2 L1.L vomitrio 4 2 L1.L L1.L.X (1) 1388 h r l g um, i, n. 2 5 3 L1.L.L m rellotge 3 2 L1.L reloj 2 1 HC1 relgio 3 2 L1.L L1.L.X (19) 1389 ammoni cum, del gr. 2 5 3 L1.L.L m amonac 4 2 L1.HC amoniaco 4 2 L1. L amonaco 5 3 L1.L.L 1390 anthr p ph gus i, m. 2 5 3 L1.L.L m antropfag 4 2 L1.HC antropfago 5 3 L1.L.L antropfago 5 3 L1.L.L 1391 post l cus a, um 2 5 3 L1.L.L m apostlic 4 2 L1.HC apostlico 5 3 L1.L. L apostlico 5 3 L1.L.L 1392 b n pl c tus a, um 2 5 3 L1.L.L m beneplcit 4 2 L1.HC beneplcito 5 3 L1.L. L beneplcito 5 3 L1.L.L 1393 cathedrat cus 2 5 3 L1.L.L m catedrt ic 4 2 L1.HC catedrtico 5 3 L1. L.L catedrtico 5 3 L1.L.L 1394 cyn c ph lus i, m., del gr. 2 5 3 L1.L.L m cinocfal 4 2 L1.HC cinoc falo 5 3 L1.L.L ci nocfalo 5 3 L1.L.L 1395 nerg m nos i, m., = energoumenos 2 5 3 L1.L.L m energumen 4 2 L1.HC ener gmeno 5 3 L1.L.L ener gmeno 5 3 L1.L.L 1396 fr g r f cus, a, um 2 5 3 L1.L.L m frigorfic 4 2 L1. HC frigorfico 5 3 L1.L.L frigorfico 5 3 L1.L.L 1397 h g ogr phus, i, m. 2 5 3 L1.L.L m hagigraf 4 2 L1. HC hagigrafo 4 3 L1.L.L hagigrafo 4 3 L1.L.L 1398 hipp p t mus, i, m., = 2 5 3 L1.L.L m hipoptam 4 2 L1.HC hi poptamo 5 3 L1.L.L hi poptamo 5 3 L1.L.L 1399 m th m t cus, a, um, adj., = 2 5 3 L1.L.L m matemtic 4 2 L1.HC matem tico 4315 3 L1.L.L matemtico 5 3 L1.L.L 1400 p n gyr cus, i, m. 2 5 3 L1.L.L m panegric 4 2 L1. HC panegrico 5 3 L1.L.L panegrico 5 3 L1.L.L 1401 p r lyt cus, i, m 2 5 3 L1.L.L m paraltic 4 2 L1. HC paraltico 5 3 L1.L.L paraltico 5 3 L1.L.L 1402 p tr nym cus, a, um 2 5 3 L1.L.L m patronmic 4 2 L1.HC patronmico 5 3 L1.L.L patron mico 5 3 L1.L.L 1403 p r d cus, a, um 2 5 3 L1.L.L m peridic 4 2 L1. HC peridico 4 3 L1.L.L peridico 4 3 L1.L.L 1404 pr m g n tus, a, um, adj. 2 5 3 L1.L.L m primognit 4 2 L1.HC primognito 5 3 L1.L. L primognito 5 3 L1.L.L 1405 qu dr l t rus a, um, 2 5 3 L1.L.L m quadrilter 4 2 L1. HC cuadriltero 5 3 L1.L.L quadriltero 5 3 L1.L.L 1406 s p r fer, f ra, f rum 2 5 3 L1.L.L m soporfer 4 2 L1.HC s oporfero 5 3 L1.L.L soporfero 5 3 L1.L.L 1407 n g n tus a, um 2 5 3 L1.L.L m unignit 4 2 L1.HC unignito 5 3 L1.L. L unignito 5 3 L1.L.L L1.L.X (40) 1408 adminic lum 2 5 3 L1.L.L m adminicle 4 2 L1.L adm inculo 5 3 L1.L.L adminculo 5 3 L1.L.L 1409 adulter um, i 2 5 3 L1.L.L m adulteri 4 2 L1. L adulterio 4 2 L1.L adultrio 4 2 L1.L 1410 adversar us, a, um 2 5 3 L1.L.L m adversari 4 2 L1.L adversario 4 2 L1.L adversrio 4 2 L1.L 1411 armistitium 2 5 3 L1.L.L m armistici 4 2 L1.L armisticio 4 2 L1.L armistcio 4 2 L1.L 1412 art f c um ii, n. 2 5 3 L1.L.L m artifici 4 2 L1.L artificio 4 2 L1.L artifcio 4 2 L1.L 1413 b n f c um, ii, n. 2 5 3 L1.L.L m benefici 4 2 L1.L beneficio 4 2 L1.L benefcio 4 2 L1.L 1414 c n t ph um ii, n., del gr. 2 5 3 L1.L.L m cenotafi 4 2 L1.L c enotafio 4 2 L1.L c enotfio 4 2 L1.L

PAGE 432

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 1415 circumloqu um 2 5 3 L1.L.L m circumloqui 42L1.L circunloquio 42L1.L circunlquio 42L1.L 1416 con-disc p lus i, m. 2 5 3 L1.L.L m condeixeble 42L1. L condiscpulo 53L1.L.L condiscpulo 53L1.L.L 1417 d m c l um ii, n. 2 5 3 L1.L.L m domicili 42 L1.L domicilio 42L1.L domiclio 42L1.L 1418 aed f c um i, n. 2 5 3 L1.L.L m edifici 42L1.L edificio 42L1.L edifcio 42L1.L 1419 p t ph um, ii, n. 2 5 3 L1.L.L m epitafi 42 L1.L epitafio 42L1.L epitfio 42L1.L 1420 vang l um (eua-), ii, n. 2 5 3 L1. L.L m evangeli 42L1.L evangelio 42L1.L Evangelho 42L1.L 1421 exerc t um ii, n. 2 5 3 L1.L.L m exercici 42L1.L ejercicio 42L1. L exerccio 42L1.L 142 2 h m cyclium, i, n. 2 5 3 L1.L.L m hemicicle 4 2 L1. L hemiciclo 4 2 L1.L hemiciclo 4 2 L1.L 1423 h mist ch um, i, n. 2 5 3 L1.L.L m hemistiqui 4 2 L1. L hemistiquio 4 2 L1.L hemistquio 4 2 L1.L 1424 impr p r um, i, n. 2 5 3 L1.L.L m impr operi 4 2 L1.L improperio 4 2 L1.L improprio 4 2 L1.L 1425 interm d us, a, um 2 5 3 L1.L.L m intermedi 4 2 L1.L intermedio 4 2 L1.L intermdio 4 2 L1.L 1426 interst t um, i, n. 2 5 3 L1.L.L m interstici 4 2 L1.L intersticio 4 2 L1.L interstcio 4 2 L1.L 1427 l tr c n um, i, n. 2 5 3 L1.L.L m lladronici 4 2 L1.L latrocinio 4 2 L1.L latrocnio 4 2 L1.L 1428 m gist r um, ii, n. 2 5 3 L1.L.L m magisteri 4 2 L1. L magisterio 4 2 L1. L magistrio 4 2 L1.L 1429 m l f c um, i, n. 2 5 3 L1.L.L m malefici 4 2 L1. L maleficio 4 2 L1.L malefcio 4 2 L1.L 1430 m nist r um, ii, n. 2 5 3 L1.L.L m ministeri 4 2 L1.L ministerio 4 2 L1.L ministrio 4 2 L1.L 1431 m n c p um, ii, n. 2 5 3 L1.L.L m municipi 4 2 L1.L municipio 4 2 L1.L municpio 4 2 L1.L 1432 r f c um, ii, n. 2 5 3 L1.L.L m orifici 4 2 L1.L orificio 4 2 L1.L orifcio 4 2 L1.L 1437 part c p um, ii, n. 2 5 3 L1.L.L m participi 4 2 L1.L participio 432 4 2 L1.L particpio 4 2 L1.L 1438 p tr c n um, i, n. 2 5 3 L1.L.L m patrocini 4 2 L1. L patrocinio 4 2 L1.L patrocnio 4 2 L1.L 1439 praej d c um, ii, n. 2 5 3 L1.L.L m perjudici 4 2 L1.L perjuicio 3 2 L1.L prejuzo 4 2 L1.L 1440 p lyp d um (on), ii, n. 2 5 3 L1.L.L m polipodi 4 2 L1.L polipodio 4 2 L1.L polipdio 4 2 L1.L 1441 praec p t um, i, n. 2 5 3 L1.L.L m precipici 4 2 L1.L precipicio 4 2 L1.L precipcio 4 2 L1.L 1442 presbyt r um, i, n.= 2 5 3 L1.L.L m presbiteri 4 2 L1.L pr esbiterio 4 2 L1.L presbitrio 4 2 L1.L 1443 pr munt r um (pr mon-), i, n. 2 5 3 L1.L.L m promontori 4 2 L1.L pr omontorio 4 2 L1.L promontrio 4 2 L1.L 1444 p erp r um, ii, n. 2 5 3 L1.L.L m puerperi 4 2 L1.L puerperio 3 2 L1.L puerprio 3 2 L1.L 1445 r fr g r um, ii, n. 2 5 3 L1.L.L m refrigeri 4 2 L1. L refrigerio 4 2 L1.L refrigrio 4 2 L1.L 1446 s cr f c um, ii, n. 2 5 3 L1.L.L m sa crifici 4 2 L1.L sacrificio 4 2 L1.L sacrifcio 4 2 L1.L 1447 s cr l g um, ii, n. 2 5 3 L1.L.L m sacrilegi 4 2 L1.L sacrilegio 4 2 L1.L sacrilgio 4 2 L1.L 1448 s l l qu um, ii, n. 2 5 3 L1.L.L m soliloqui 4 2 L1.L soliloquio 4 2 L1.L solilquio 4 2 L1.L 1449 subterf g um, i, n. 2 5 3 L1.L.L m subterfugi 4 2 L1.L subterfugio 4 2 L1.L subterfgio 4 2 L1.L 1450 tensil a um, n., pl. n. de utens lis 2 5 3 L1.L.L m utensili 4 2 L1.L utensilio 4 2 L1.L utenslio 4 2 L1.L 1447 vituper um 2 5 3 L1.L.L m vituperi 4 2 L1.L vituperio 4 2 L1.L vituprio 4 2 L1.L

PAGE 433

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp HC.HC1.X (1) 1448 f d commissum, i, n. 2 6 2 HC.HC1.L m fidecoms 5 1 L. HC1 fideicomiso 5 2 L.L1.L fideicomisso 5 2 L.L1.L L.HC1.X (1) 1449 p rall logrammus, a, um, adj. 2 6 2 L.HC1.L m parallelogram 5 1 L.HC1 paralelogramo 6 2 L.L1.L paralelogramo 6 2 L.L1.L HC1.L.X (1) 1450 eccl s ast cus i, m., del gr. 2 6 3 HC1.L.L m eclesistic 5 2 HC1.HC eclesi stico 5 3 HC1.L.L eclesistico 5 3 HC1.L.L HC1.L.X (1) 1451 interc lumn um, i, n. 2 6 3 HC1.L.L m intercolumni 5 2 HC1.L intercolumnio 5 2 HC1.L intercolnio 5 2 L.L1.L HV1.L.X (2) 1452 r p s t r um, ii, n. 2 6 3 HV1.L.L m reboster 3 1 HC1 repostero 4 2 L1.L reposteiro 4 2 L1.L 1453 h r d t r us, a, um 2 6 3 HV1.L.L m hereter 3 1 HC1 heredero 4 2 L1.L herdeiro 3 2 HV1.L HV1.L.X (2) 1454 conc l b lum i, n. 2 6 3 HV1.L.L m concilibul 5 2 L1.HC concilibulo 5 3 L1.L. L concilibulo 5 3 L1.L.L 1455 P r p t t cus, i, m. 2 6 3 HV1.L.L m peripattic 5 2 L1.HC per ipattico 6 3 L1.L.L per ipattico 6 3 L1.L.L HV1.L.X (17) 1456 b c d r um, i, n., 2 6 3 HV1.L.L m abecedari 5 2 L1.L abecedario 5 2 L1.L abecedrio 5 2 L1.L 1457 ann vers r us a, um 2 6 3 HV1.L.L m aniversari 5 2 L1.L aniversario 433 5 2 L1.L aniversrio 5 2 L1.L 1458 conservator us 2 6 3 HV1.L.L m conservatori 5 2 L1.L c onservatorio 5 2 L1.L conservatrio 5 2 L1.L 1459 b. lat. dictionarium 2 6 3 HV1.L.L m diccionar i 5 2 L1.L diccionario 4 2 L1.L dicionrio 4 2 L1.L 1460 f d c r us, a, um 2 6 3 HV1.L.L m fiduciari 5 2 L1.L fiduciario 4 2 L1.L fiducirio 4 2 L1.L 1461 infant c d um, i, n. 2 6 3 HV1.L.L m infanticidi 5 2 L1. L infanticidio 5 2 L1.L infanticdio 5 2 L1.L 1462 t n r r um, i, n. 2 6 3 HV1.L.L m itinerari 5 2 L1.L itinerario 5 2 L1. L itinerrio 5 2 L1.L 1463 n n g n r us a, um 2 6 3 HV1.L.L m nonagenari 5 2 L1.L nonagenario 5 2 L1.L nonagenrio 5 2 L1.L 1464 oct g n r us, a, um, adj. 2 6 3 HV1.L.L m octogenari 5 2 L1.L octogenario 5 2 L1. L octogenrio 5 2 L1.L 1465 propr t r us, ii, m. 2 6 3 HV1.L.L m propietari 5 2 L1.L propietario 4 2 L1.L proprietrio 4 2 L1.L 1466 quaest n r us ii, m. 2 6 3 HV1.L.L m qestionari 5 2 L1.L cuestionario 4 2 L1.L questionrio 4 2 L1.L 1467 r cl n t r um, i, n. 2 6 3 HV1.L.L m re clinatori 5 2 L1.L reclinatorio 5 2 L1.L reclinatrio 5 2 L1.L 1468 sex g n r us, a, um 2 6 3 HV1.L.L m sexagenari 5 2 L1.L sexagenario 5 2 L1.L sexagenrio 5 2 L1.L 1469 supp s t r us, a, um 2 6 3 HV1.L.L m supositori 5 2 L1.L supositorio 5 2 L1.L supositrio 5 2 L1.L 1470 tyrannicid um 2 6 3 HV1.L.L m tiranicidi 5 2 L1.L tiranicidio 5 2 L1.L tiranicdio 5 2 L1.L 1471 veterin rius, der. de veter nae 2 6 3 HV1.L.L m veterinari 5 2 L1.L veterinario 5 2 L1.L veterinrio 5 2 L1.L 1472 b. ll. vocabul rium 2 6 3 HV1.L.L m vocabulari 5 2 L1.L vo cabulario 5 2 L.L1.L vocabulrio 5 2 L.L1.L

PAGE 434

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp L1.L.X (1) 1473 arch presbyter ri, m. 2 6 3 L1.L.L m arxiprest 31HC1 ar cipreste 42HC1.L arcipreste 42HC1.L L1.L.X (1) 1474 aphrodisi cus (< ) 2 6 3 L1.L.L m afrodisac 52L1.HC afrodisiaco 52L. L1.L afrodis aco 63L1.L.L L1.L.X (2) 1475 p th l m um, ii, n. 2 6 3 L1.L.L m epitalami 52 L1.L epitalamio 52L1.L epitalmio 52L1.L 1476 r t c n um, ii, n. 2 6 3 L1.L.L m raciocini 52 L1.L raciocinio 42L1.L raciocnio 42L1.L HV1.L.X (3) 1477 extr ord n r us a, um 2 7 3 HV1.L.L m extraordinari 62L1.L extraordinario 6 2 L1.L extraordinrio 6 2 L1.L 1478 sept g n r us, a, um 2 7 3 HV1.L.L m s eptuagenari 6 2 L1.L septuagenario 5 2 L1.L septuagenrio 5 2 L1.L 1479 s fruct r us ii, m. 2 7 3 HV1.L.L m usufructuari 6 2 L1.L usufrutuario 5 2 L1.L usufruturio 5 2 L1.L HC1 (3) 1480 fel, fellis, n. 3 1 1 HC1 m/f fel 1 1 HC1 hiel 1 1 HC1 fel 1 1 HC1 1481 m l, mellis, n. 3 1 1 HC1 f mel 1 1 HC1 miel 1 1 HC1 mel 1 1 HC1 1482 p s, p ris, n. 3 1 1 HC1 m pus 1 1 HC1 pus 1 1 HC1 pus 1 1 HC1 HC1.X (42) 1483 ars, artis, f. 3 2 2 HC1.L m/f art 1 1 HC1 arte 2 2 HC1.L arte 2 2 HC1.L 1484 as assis, m. 3 2 2 HC1.L m as 1 1 HC1 as 434 1 1 HC1 s 1 1 HC1 1485 calx calcis, f. 3 2 2 HC1.L f cal 1 1 HC1 cal 1 1 HC1 cal 1 1 HC1 1486 c ro carnis, f. 3 2 2 HC1.L f carn 1 1 HC1 carne 2 2 HC1.L carne 2 2 HC1.L 1487 cors, cortis, f. 3 2 2 HC1.L f cort 1 1 HC1 corte 2 2 HC1.L corte 2 2 HC1.L 1488 corpus ris, n. 3 2 2 HC1.L m cos 1 1 HC1 cuerpo 2 2 HC1.L corpo 2 2 HC1.L 148 9 dens, dentis 3 2 2 HC1.L m dent 1 1 HC1 diente 2 2 HC1.L dente 2 2 HC1.L 1490 dulcis e, adj. 3 2 2 HC1.L m dol 1 1 HC1 dulce 2 2 HC1.L doce 2 2 L1.L 1491 axis, is, m. 3 2 2 HC1.L m eix 1 1 HC1 eje 2 2 L1.L eixo 2 2 HV1.L 1492 ens, entis, n. 3 2 2 HC1.L m ens 1 1 HC1 ente 2 2 HC1.L ente 2 2 HC1.L 1493 falx, falcis, f. 3 2 2 HC1.L f fal 1 1 HC1 hoz 1 1 HC1 fouce 2 2 HV1.L 1494 fascis, is, m. 3 2 2 HC1.L m feix 1 1 HC1 haz 1 1 HC1 feixe 2 2 L1.L 1495 fons, fontis, m. 3 2 2 HC1.L f f ont 1 1 HC1 fuente 2 2 HC1.L fonte 2 2 HC1.L 1496 fortis, e, adj. 3 2 2 HC1.L m fort 1 1 HC1 fuerte 2 2 HC1.L forte 2 2 HC1.L 1497 frons, frontis, f. 3 2 2 HC1.L f front 1 1 HC1 frente (ant. fruente) 2 2 HC1.L fronte 2 2 HC1.L 1498 fustis, is, m. 3 2 2 HC1.L m fust 1 1 HC1 fuste 2 2 HC1.L fuste 2 2 HC1.L 1499 gens, gentis, f. 3 2 2 HC1.L f gent 1 1 HC1 gente 2 2 HC1.L gente 2 2 HC1.L 1500 glans, glandis, f. 3 2 2 HC1.L f gland 1 1 HC1 glande 2 2 HC1.L glande 2 2 HC1.L 1501 grandis, e, adj. 3 2 2 HC1.L m gran 1 1 HC1 grande 2 2 HC1.L grande 2 2 HC1.L

PAGE 435

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 1502 hostis, is, m.& f. 3 2 2 HC1.L m/fhost 11HC1 hueste 22HC1.L hoste 22HC1.L 1503 lens, tis, f. 3 2 2 HC1.L f lent 11HC1 lente 22HC1.L lente 22HC1.L 1504 lynx, lyncis, com., = 3 2 2 HC1.L m/flinx 11HC1 lince 22HC1.L lince 22HC1.L 1505 lac, lactis, n. 3 2 2 HC1.L f llet 11HC1 leche 22L1.L leite 22HV1.L 1506 mens, mentis, f. 3 2 2 HC1.L f ment 11HC1 mente 22HC1.L mente 22HC1.L 1507 mensis, is, m. 3 2 2 HC1.L m mes 11HC1 mes 11HC1 ms 11HC1 1508 messis, is, f. 3 2 2 HC1.L f messes [pl] 11HC1 mies 11HC1 messe 22L1.L 150 9 mons, tis, m. 3 2 2 HC1.L m mont 1 1 HC1 monte 2 2 HC1.L monte 2 2 HC1.L 1510 mors, tis, f. 3 2 2 HC1.L f mort 1 1 HC1 muerte 2 2 HC1.L morte 2 2 HC1.L 1511 nox, noctis, f. 3 2 2 HC1.L f nit 1 1 HC1 noche 2 2 L1.L noite 2 2 HV1.L 1512 pars, partis, f. 3 2 2 HC1.L f part 1 1 HC1 parte 2 2 HC1.L parte 2 2 HC1.L 1513 piscis, is, m. 3 2 2 HC1.L m peix 1 1 HC1 peje 2 2 L1.L peixe 2 2 HV1.L 1514 piscis, is, m. 3 2 2 HC1.L m peix 1 1 HC1 pez 1 1 HC1 peixe 2 2 HV1.L 1515 pellis, is, f. 3 2 2 HC1.L f pell 1 1 HC1 piel 1 1 HC1 pele 2 2 L1.L 1516 pulvus, por pulvis, ris, m./f. 3 2 2 HC1.L m pols 1 1 HC1 polvo 2 2 HC1.L p 1 1 L1 1517 pons, ntis, m. 3 2 2 HC1.L m pont 1 1 HC1 puente 2 2 HC1.L ponte 2 2 HC1.L 1518 postis, is, m. 3 2 2 HC1.L m post 1 1 HC1 poste 2 2 HC1.L poste 2 2 HC1.L 1519 serpens, entis, f. 3 2 2 HC1.L f serp 1 1 HC1 sierpe 435 2 2 HC1.L serpe 2 2 HC1.L 1520 sors, tis, f. 3 2 2 HC1.L f sort 1 1 HC1 suerte 2 2 HC1.L sorte 2 2 HC1.L 1521 tempus, -oris, n. 3 2 2 HC1.L m temps 1 1 HC1 tiempo 2 2 HC1.L tempo 2 2 HC1.L 1522 tussis is, f. 3 2 2 HC1.L f tos 1 1 HC1 tos 1 1 HC1 tosse 2 2 L1.L 1523 urbs urbis, f. 3 2 2 HC1.L f urbs 1 1 HC1 urbe 2 2 HC1.L urbe 2 2 HC1.L 1524 valles o vallis, is, f. 3 2 2 HC1.L m vall 1 1 HC1 valle 2 2 L1.L vale 2 2 L1.L HC1.X (prothesis) (2) 1525 stans, -antis 3 2 2 HC1.L m estant 2 1 HC.HC1 estante 3 2 HC1 .L estante 3 2 HC1.L 1526 stirps, stirpis, f. 3 2 2 HC1.L f estirp 2 1 HC.HC1 estirpe 3 2 HC1.L estirpe 3 2 HC1.L HC1.X (6) 1527 tl s antis 3 2 2 HC1.L m atles 2 2 HC1 .HC atlas 2 2 HC1.HC atlas 2 2 HC1.HC 1528 forceps, c pis, m.&f. 3 2 2 HC1.L m frceps 2 2 HC1.HC frceps 2 2 HC1.HC frceps 2 2 HC1.HC 1529 germen, nis, n. 3 2 2 HC1.L m germen 2 2 HC1.HC germen 2 2 HC1.HC grmen 2 2 HC1.HV 1530 herpes, tis, m. 3 2 2 HC1.L m herpes 2 2 HC1 .HC herpes 2 2 HC1.HC herpes 2 2 HC1.HC 1531 nectar, ris, n., = 3 2 2 HC1.L m nctar 2 2 HC1.HC nctar 2 2 HC1.HC nctar 2 2 HC1.HC 1532 unguis is 3 2 2 HC1.L m unguis 2 2 HC1.HC unguis 2 2 HC1.HC nguis 2 2 HC1.HC HC1.X (1) 1533 sphincter, ris, m.= 3 2 2 HC1.X m esfnter 3 2 HC1.HC esfnter 3 2 HC1.HC esfncter 3 2 HC1.HC

PAGE 436

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp HC1.X (5) 1534 orbis, is, m. 3 2 2 HC1.L m orbe 22HC1.L orbe 22HC1.L orbe 22HC1.L 1535 pelvis, is, f. 3 2 2 HC1.L f pelvis 22HC1.L pelvis 22HC1.L plvis 22HC1.L 1536 pestis, is, f. 3 2 2 HC1.L f pesta 22HC1.L peste 22HC1.L peste 22HC1.L 1537 venter, -tris, m. 3 2 2 HC1.L m ventre 22HC1.L vientre 22HC1.L ventre 22HC1.L 1538 virgo nis, f. 3 2 2 HC1.L f virgo 22HC1.L virgo 22HC1.L Virgo 22HC1.L HC1.X (2) 1539 classis, is, f. 3 2 2 HC1.L f classe 22L1.L clase 22L1.L classe 22L1.L 1540 turris is, f. 3 2 2 HC1.L f torre 2 2 L1.L torre 2 2 L1.L torre 2 2 L1.L HV1.X (9) 1541 caulis, is, m. 3 2 2 HV1.L f col 1 1 HC1 col 1 1 HC col 1 1 HC 1542 cr men nis, n. 3 2 2 HV1.L m crim 1 1 HC1 crimen 2 2 L1.HC crime 2 2 L1.L 1543 cr nis is, m. 3 2 2 HV1.L m/f crin 1 1 HC1 crin 1 1 HC1 crina 2 2 L1.L 1544 d s tis, f. 3 2 2 HV1.L m/f dot 1 1 HC1 dote 2 2 L1.L dote 2 2 L1.L 1545 fl s, ris, m. 3 2 2 HV1.L f flor 1 1 HC1 flor 1 1 HC1 flor 1 1 HC1 1546 gr men, nis, n. 3 2 2 HV1.L m/f gram 1 1 HC1 grama 2 2 L1.L grama 2 2 L1.L 1547 l men, nis, n. 3 2 2 HV1.L m/f llum 1 1 HC1 lumbre 2 2 HC1.L lume 2 2 L1.L 1548 lux, l cis, f. 3 2 2 HV1.L f lux 1 1 HC1 lux 1 1 HC1 lux 1 1 HC1 1549 r te, is, n. 3 2 2 HV1.L m/f ret 1 1 HC1 red 436 1 1 HC1 rede 2 2 L1.L 1550 s l, s lis, m. 3 2 2 HV1.L m sol 1 1 HC1 sol 1 1 HC1 sol 1 1 HC1 1551 tr m s tis, m. 3 3 3 HV1.L m tram 1 1 HC1 tramo 2 2 L1.L tramo 2 2 L1.L HV1.X (11) 1552 cl vis, is, f. 3 2 2 HV1.L f clau 1 1 HV1 clave 2 2 L1.L clave 2 2 L1.L 1553 f nis, is, m.(&f.) 3 2 2 HV1.L m fi 1 1 HV1 fin 1 1 HC1 fim 1 2 HV1 155 4 fraus, fraudis, f. 3 2 2 HV1.L m/f f rau 1 1 HV1 fraude 2 2 HV1.L fraude 2 2 HV1.L 1555 lex, l gis, f. 3 2 2 HV1.L f llei 1 1 HV1 ley 1 1 HV1 lei 1 1 HV1 1556 cl vis, is, f. 3 2 2 HV1.L f clau 1 1 HV1 llave 2 2 L1.L chave 2 2 L1.L 1557 n vis, is, f. 3 2 2 HV1.L f nau 1 1 HV1 nave 2 2 L1.L nave 2 2 L1.L 1558 p nis, is, m. 3 2 2 HV1.L m pa 1 1 HV1 pan 1 1 HC1 po 1 1 HV1 1559 pax, p cis, f. 3 2 2 HV1.L f pau 1 1 HV1 paz 1 1 HC1 paz 1 1 HC1 1560 rex, r gis, m. 3 2 2 HV1.L m rei 1 1 HV1 rey 1 1 HV1 rei 1 1 HV1 1561 s des, is, f. 3 2 2 HV1.L f seu 1 1 HV1 sede 2 2 L1.L sede 2 2 L1.L 1562 vox v cis, f. 3 2 2 HV1.L f veu 1 1 HV1 voz 1 1 HC1 voz 1 1 HC1 HV1.X (10) 156 3 tr ceps c p tis, adj. 3 2 2 HV1.L m trceps 2 2 L1. HC trceps 2 2 L1.HC tricpite 4 3 L1.L.L 1564 b lis is, f. 3 2 2 HV1.L f bilis 2 2 L1.HC bilis 2 2 L1.HC bilis 2 2 L1.HC

PAGE 437

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 1565 cl max cis, f. 3 2 2 HV1.L m/f clmax 2 2 L1.HC clmax 2 2 L1.HC clmax 2 2 L1.HC 1566 c dex cis, m. 3 2 2 HV1.L m cdex 2 2 L1.HC cdex 2 2 L1.HC cdex 2 2 L1.HC 1567 phoenix, cis, m. 3 2 2 HV1.L m fnix 2 2 L1.HC fnix 2 2 L1. HC fnix 2 2 L1.HC 1568 ris, is, f. 3 2 2 HV1.L f iris 2 2 L1.HC iris 2 2 L1.HC ris 2 2 L1.HC 1569 n men, nis, n. 3 2 2 HV1.L m numen 2 2 L1.HC numen 2 2 L1.HC nmeno 3 2 L1.L.L 1570 p nis, is, m. 3 2 2 HV1.L m pen is 2 2 L1.HC pene 2 2 L1.L pene 2 2 L1.L 1571 s men, nis, n. 3 2 2 HV1.L m semen 2 2 L1.HC semen 2 2 L1.HC smen 2 2 L1.HV 157 2 th rax cis, m. 3 2 2 HV1.L m trax 2 2 L1.HC trax 2 2 L1.HC trax 2 2 L1.HC HV1.X (3) 1573 ter, tris, m. 3 2 2 HV1.L m odre 2 2 L1.L odre 2 2 L1.L odre 2 2 L1.L 1574 pr les, is, f. 3 2 2 HV1.L f prole 2 2 L1.L prole 2 2 L1.L prole 2 2 L1.L 1575 m ter, tris, f. 3 2 2 HV1.L f ma re 2 2 L1.L madre 2 2 L1.L me 1 1 HV1 1579 n men nis, n. 3 2 2 HV1.L.L m nom 1 1 HC1 nombre 2 2 HC1.L nome 2 2 L1.L L1.X (3) 1576 c p t, tis, n. 3 2 2 L1.L m cap 1 1 HC1 cabo 2 2 L1.L cabo 2 2 L1.L 1577 L r, L ris, m. 3 2 2 L1.L m llar 1 1 HC1 lar 1 1 HC1 lar 1 1 HC1 1578 m re, is, n. 3 2 2 L1.L m/f mar 1 1 HC1 mar 1 1 HC1 mar 1 1 HC1 1580 p r, p ris, f. 3 2 2 L1.L m par 1 1 HC1 par 437 1 1 HC1 par 1 1 HC1 1581 plebs, -bis, f. 3 2 2 L1.L f plebs 1 1 HC1 plebe 2 2 L1.L plebe 2 2 L1.L 1582 sal, s lis, m. 3 2 2 L1.L f sal 1 1 HC1 sal 1 1 HC1 sal 1 1 HC1 1583 s tis, is, f. 3 2 2 L1.L f set 1 1 HC1 sed 1 1 HC1 sede 2 2 L1.L L1.X (9) 1584 c nis, c nis, m.& f. 3 2 2 L1.L m ca 1 1 HV1 can 1 1 HC1 co 1 1 HV1 1585 vis is, f. 3 2 2 L1.L f au 1 1 HV1 ave 2 2 L1.L ave 2 2 L1.L 1586 br vis is, m. (lat tard) 3 2 2 L1.L m breu 1 1 HV1 breve 2 2 L1.L breve 2 2 L1.L 1587 b s b vis, m. 3 2 2 L1.L m bou 1 1 HV1 buey 1 1 HV1 boi 1 1 HV1 1588 crux cis, f. 3 2 2 L1.L f creu 1 1 HV1 cruz 1 1 HC1 cruz 1 1 HC1 1589 gr vis, e 3 2 2 L1.L m greu 1 1 HV 1 grave 2 2 L1.L grave 2 2 L1.L 1590 nix n vis, f. 3 2 2 L1.L f neu 1 1 HV1 nieve 2 2 L1.L neve 2 2 L1.L 1591 nux, n cis, f. 3 2 2 L1.L f nou 1 1 HV1 nuez 1 1 HC1 noz 1 1 HC1 1592 p s, p dis, m. 3 2 2 L1.L m peu 1 1 HV1 pie 1 1 L1 p 1 1 L1 L1.X (8) 1593 pex cis, m. 3 2 2 L1.L m pex 2 2 L1.HC pex 2 2 L1.HC pex 2 2 L1.HC 1594 b ceps (
PAGE 438

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 1596 f mur, ris, n. 3 2 2 L1.L m fmur 22L1.HC fmur 22L1.HC fmur 22L1.HC 1597 ibis, is & dis, f. 3 2 2 L1.L m ibis 22 L1.HC ibis 22L1.HC bis 22L1.HC 1598 l pis, dis, m. 3 2 2 L1.L m llapis 22L1.HC lpiz 22L1.HC lpis 22L1.HC 1599 l tex, cis, m. 3 2 2 L1.L m ltex 22L1.HC ltex 22L1.HC ltex 22L1.HC 1600 s lex, cis, m.& f. 3 2 2 L1.L m slex 22 L1.HC slex 22L1. HC slex 22L1.HC L1.X (7) 1601 b sis is, f., y este del gr. 3 2 2 L1. L f base 22L1.L base 22L1.L base 22L1.L 1602 cr sis is, f. 3 2 2 L1.L f crisi 22 L1.L crisis 22 L1.HC crise 2 2 L1.L 1603 f bris, is, f. 3 2 2 L1.L f febre 2 2 L1.L fiebre 2 2 L1.L febre 2 2 L1.L 1604 phr sis, is, f. 3 2 2 L1.L f frase 2 2 L1.L frase 2 2 L1.L frase 2 2 L1.L 1605 p ter, tris, m. 3 2 2 L1.L m pare 2 2 L1.L padre 2 2 L1.L padre 2 2 L1.L 1606 th sis is, f., del gr. 3 2 2 L1.L f tesi 2 2 L1.L tesis 2 2 L1.HC tese 2 2 L1.L 1607 tigris, idis, m., del gr. (acc, tigrem en V.) 3 2 2 L1.L m tigre 2 2 L1.L tigre 2 2 L1.L tigre 2 2 L1.L HC.HC1.X (13) 1608 archon, ntis, y este del gr. 3 3 2 HC.HC1.L m arcont 2 1 HC.HC1 ar conte 3 2 HC.HC1.L arconte 3 2 HC.HC1.L 1609 tlantis dis 3 3 2 HC.HC1.L m atlant 2 1 HC.HC1 atlante 3 2 HC.HC1.L atlante 3 2 HC.HC1.L 1610 consors, sortis, m.& f. 3 3 2 HC.HC1.L m/f consort 2 1 HC.HC1 consor te 438 3 2 HC.HC1.L consorte 3 2 HC.HC1.L 1611 constans, -antis 3 3 2 HC. HC1.L f constant 2 1 HC.HC1 const ante 3 2 HC.HC1.L constante 3 2 HC.HC1.L 1612 infans, fantis, 3 3 2 HC. HC1.L m/f infant 2 1 HC.HC1 infant e 3 2 HC.HC1.L infante 3 2 HC.HC1.L 1613 instans, -antis 3 3 2 HC. HC1.L m instant 2 1 HC.HC1 instant e 3 2 HC.HC1.L instante 3 2 HC.HC1.L 1614 pendens, entis 3 3 2 HC.HC1.L m/f pendent 2 1 HC.HC1 pendiente 3 2 HC.HC1.L pendente 3 2 HC.HC1.L 1615 serpens, -entis, f. 3 3 2 HC.HC1.L f serpent 2 1 HC.HC1 serpiente 3 2 HC.HC1.L serpente 3 2 HC.HC1.L 1616 solvens, -ntis 3 3 2 HC.HC1.L m solvent 2 1 HC.HC1 solve nte 3 2 HC.HC1.L solvente 3 2 HC.HC1.L 1617 tangens, -entis 3 3 2 HC. HC1.L m tangent 2 1 HC.HC1 tangent e 3 2 HC.HC1.L tangente 3 2 HC.HC1.L 1618 currens, -entis 3 3 2 HC.HC1.L f corrent 3 2 HC.HC1 corriente 3 2 HC.HC1.L corrente 3 2 HC.HC1.L 1619 sextans, antis, m. 3 3 2 HC.HC1.L m sextant 3 2 HC.HC1 sext ante 3 2 HC.HC1.L sextante 3 2 HC.HC1.L 1620 nascens, -entis 3 3 2 HC.HC1.L m naixent 2 2 HC.HC1 naciente 3 2 HC1.L nascente 3 2 HC1.L HC.HC1.X (4) 1621 ellipsis, is, f., del gr. 3 3 2 HC.HC1.L f ellip se 3 2 HC.HC1.L elipse 3 2 L.HC1.L elipse 3 2 L.HC1.L 1622 ellipsis, is, f., del gr. 3 3 2 HC.HC1.L f ellip si 3 2 HC.HC1.L elipsis 3 2 L.HC1.HC elipse 3 2 L.HC1.L 1623 syllepsis, is, f., = 3 3 2 HC.HC1.L f sillepsi 3 2 HC.HC1.L s ilepsis 3 2 L.HC1.HC silepse 3 2 L.HC1.L 1624 syntaxis, is, f., = 3 3 2 HC.HC1.L f sintaxi 3 2 HC.HC1.L sintaxis 3 2 HC.L1.HC sintaxe 3 2 HC.L1.L

PAGE 439

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp HC.HC1.X (2) 1625 torrens, -entis 3 3 2 HC.HC1.L m torrent 2 2 L.HC1 torrente 3 2 L.HC1.L torrente 3 2 L.HC1.L 1626 September, bris, m. 3 3 2 HC.HC1.L m setembre 3 2 L.HC1.L septiembre 3 2 HC.HC1.L setembro 3 2 L.HC1.L HC.HV1.X (30) 1627 actor, ris 3 3 2 HC.HV1.L m actor 2 1 HC.HC1 actor 2 1 HC.HC1 actor 2 1 HC.HC1 1628 albor, ris 3 3 2 HC.HV1.L m albor 2 1 HC.HC1 albor 2 1 HC.HC1 albor 2 1 HC.HC1 1629 alt re, -ris 3 3 2 HC.HV1.L m altar 2 1 HC.HC1 altar 2 1 HC.HC1 altar 2 1 HC.HC1 1630 ardor ris, m. 3 3 2 HC.HV1.L m ardor 2 1 HC.HC1 ardor 2 1 HC.HC1 ardor 2 1 HC.HC1 1631 candor ris, m. 3 3 2 HC.HV1.L m candor 2 1 HC.HC1 candor 2 1 HC.HC1 candor 2 1 HC.HC1 1632 castor, ris, m. 3 3 2 HC.HV1.L m castor 2 1 HC. HC1 castor 2 1 HC.HC1 castor 2 1 HC.HC1 1633 censor, ris, m. 3 3 2 HC.HV1.L m censor 2 2 HC. HC1 censor 2 2 HC.HC1 censor 2 2 HC.HC1 1634 cursor, ris 3 3 2 HC.HV1.L m cursor 2 1 HC.HC1 cursor 2 1 HC.HC1 cursor 2 1 HC.HC1 1635 dent le, is, n. 3 3 2 HC.HV1.L m dental 2 1 HC.HC1 dental 2 1 HC.HC1 dental 2 1 HC.HC1 1636 doctor ris, m. 3 3 2 HC.HV1.L m doctor 2 1 HC. HC1 doctor 2 1 HC.HC1 doutor 2 1 HV.HC1 1637 error, ris, m. 3 3 2 HC.HV1.L m error 2 1 HC. HC1 error 2 1 HC.HC1 error 2 1 HC.HC1 1638 factor, ris, m. 3 3 2 HC.HV1.L m factor 2 1 HC. HC1 factor 2 1 HC.HC1 factor 2 1 HC.HC1 1639 fervor, ris, m. 3 3 2 HC.HV1.L m fervor 2 1 HC.HC1 hervor 2 1 HC.HC1 fervor 2 1 HC.HC1 1940 fisc lis, e, adj. 3 3 2 HC.HV1.L m fiscal 2 1 HC.HC1 fiscal 439 2 1 HC.HC1 fiscal 2 1 HC.HC1 1641 fulgor, ris, m. 3 3 2 HC.HV1.L m fulgor 2 1 HC. HC1 fulgor 2 1 HC.HC1 fulgor 2 1 HC.HC1 1642 gent lis, e, adj. 3 3 2 HC.HV1.L m gentil 2 1 HC.HC1 gentil 2 1 HC.HC1 gentil 2 1 HC.HC1 1643 gestor, ris, m. 3 3 2 HC.HV1.L m gestor 2 1 HC.HC1 gestor 2 1 HC.HC1 gestor 2 1 HC.HC1 1644 lector, ris, m. 3 3 2 HC.HV1.L m lector 2 1 HC. HC1 lector 2 1 HC.HC1 leitor 2 1 HV.HC1 1645 lictor, ris, m. 3 3 2 HC.HV1.L m lictor 2 1 HC.HC1 lictor 2 1 HC.HC1 lictor 2 1 HC.HC1 1646 magnas, tis, m. 3 3 2 HC.HV1.L m magnat 2 1 HC. HC1 magnate 3 2 HC.L1. L magnate 3 2 HC.L1.L 1647 magnes, tis, m., = 3 3 2 HC.HV1.L m magnet 2 1 HC.HC1 m agneto 3 2 HC.L1.L magneto 3 2 HC.L1.L 1648 mant le (mant le), is, n. 3 3 2 HC.HV1.L m mantell 2 1 HC.HC1 mantel 2 1 HC. HC1 mantel 2 1 HC.HC1 1649 mort lis, e 3 3 2 HC.HV1.L m mortal 2 1 HC.HC1 mortal 2 1 HC.HC1 mortal 2 1 HC.HC1 1650 palm ris, e, adj. 3 3 2 HC.HV1.L m palmar 2 1 HC.HC1 palmar 2 1 HC.HC1 palmar 2 1 HC.HC1 1651 pastor, ris, m. 3 3 2 HC.HV1.L m pastor 2 2 HC.HC1 pastor 2 2 HC.HC1 pastor 2 2 HC.HC1 1652 raptor, ris, m. 3 3 2 HC.HV1.L m raptor 2 1 HC. HC1 raptor 2 1 HC.HC1 raptor 2 1 HC.HC1 1653 rector, ris, m. 3 3 2 HC.HV1.L m rector 2 1 HC. HC1 rector 2 1 HC.HC1 reitor 2 1 HV.HC1 1654 sector, ris, m. 3 3 2 HC.HV1.L m sector 2 1 HC. HC1 sector 2 1 HC.HC1 sector 2 1 HC.HC1 1655 Sext lis, e 3 3 2 HC.HV1.L m sextil 2 1 HC.HC1 sextil 2 1 HC.HC1 sextil 2 1 HC.HC1 1656 tensor, ris 3 3 2 HC.HV1.L m tensor 2 1 HC.HC1 tensor 2 1 HC.HC1 tensor 2 1 HC.HC1

PAGE 440

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 1657 vector, ris 3 3 2 HC.HV1.Lm vector 21HC.HC1 vector 21HC.HC1 vector 21HC.HC1 1658 virtus, tis 3 3 2 HC.HV1.Lf virtut 21HC. HC1 virtud 21HC.HC1 virtude 32HC.L1.L HC.HV1.X (prothesis) (1) 1659 splendor, ris, m. 3 3 2 HC.HV1.Lm esplendor 31HC. HC1 esplendor 31HC.HC1 esplendor 31HC.HC1 HC.HV1.X (10) 1660 actrix, cis 3 3 2 HC.HV1.Lf actriu 21HC.HV 1 actriz 21HC.HC1 actriz 21HC.HC1 1661 carbo, nis, m. 3 3 2 HC.HV1.Lm carb 21HC. HV1 carbn 21HC.HC1 carvo 21HC.HC1 1662 cento nis, m. 3 3 2 HC.HV1.Lm cent 21HC. HV1 centn 21HC.HC1 cento 21HC.HV1 1663 delph n nis, m., del gr. 3 3 2 HC.HV1.L m delf 2 1 HC.HV1 delfn 2 1 HC.HC1 delfim 2 1 HC.HV1 1664 falco, nis, m 3 3 2 HC.HV1.L m falc 2 1 HC. HV1 halcn 2 1 HC.HC1 falco 2 1 HC.HV1 1665 perdix, cis, m.& f. 3 3 2 HC.HV1.L f perdiu 2 1 HC. HV1 perdiz 2 1 HC.HC1 perdiz 2 1 HC.HC1 1666 ponto, nis, m. 3 3 2 HC.HV1.L m pont 2 1 HC. HV1 pontn 2 1 HC.HC1 ponto 2 1 HC.HV1 1667 pulmo, nis, m. 3 3 2 HC.HV1.L m pulm 2 1 HC. HV1 pulmn 2 1 HC.HC1 pulmo 2 2 HC.HV1 1668 salmo, nis, m. 3 3 2 HC.HV1.L m salm 2 1 HC.HV 1 salmn 2 1 HC.HC1 salmo 2 1 HC.HV1 1669 sermo, nis, m. 3 3 2 HC.HV1.L m serm 2 1 HC.HV1 sermn 2 1 HC.HC1 sermo 2 1 HC.HV1 HC.HV1.X (1) 1670 alb go, nis, f. 3 3 2 HC.HV1.L f albugo 3 2 HC. HV1.L albugo 3 2 HC.HV1. L albugo 3 2 HC.HV1.L HC.HV1.X (7) 1671 abd m n, nis, n. 3 3 2 HC.HV1.L m abdomen 3 2 HC.L1.HC abdomen 440 3 2 HC.L1. HC abdmen 3 2 HC.L1.HC 1672 alb men, nis, n. 3 3 2 HC.HV1.L m albumen 3 2 HC. L1.HC albumen 3 2 HC.HV1. L albmen 3 2 HC.HV1.L 1673 arthr tis dis, f., y este del gr. 3 3 2 HC.HV1.L f artritis 3 2 HC.L1.HC artritis 3 2 HC.L1.HC artrite 3 2 HC.L1.L 1674 cert men nis, n. 3 3 2 HC.HV1.L m certamen 3 2 HC.L1.HC certamen 3 2 L1.HC certmen 3 2 L1.HC 1675 dict men, nis, n. 3 3 2 HC.HV1.L m dictamen 3 2 HC.L1.HC dictamen 3 2 HC.L1.L ditame 3 2 L.L1.L 1676 ex men, nis, n. 3 3 2 HC.HV1.L m examen 3 2 HC.L1.HC examen 3 2 HC. L1.HC exame 3 2 L.L1.L 1677 vex men nis, n. 3 3 2 HC.HV1.L m vexamen 3 2 HC.L1.HC vejamen 3 2 L1.HC vexame 3 2 L1.L HC.HV1.X (4) 1678 concl ve is, n. 3 3 2 HC.HV1.L m conclave 3 2 HC. L1.L cnclave 3 3 HC1.L.L conclave 3 2 HC.L1.L 1679 embl ma tis, n., del gr. 3 3 2 HC.HV1.L m emblem a 3 2 HC.L1.L emblema 3 2 HC. L1.L emblema 3 2 HC.L1.L 1680 lumb go, nis, f. 3 3 2 HC.HV1.L f lumbago 3 2 HC. L1.L lumbago 3 2 HC.L1.L lumbago 3 2 HC.L1.L 1681 Oct ber, bris, adj. 3 3 2 HC.HV1.L m octubre 3 2 HC.L1.L octubre 3 2 HC.L1. L Outubro 3 2 HV.L1.L 1682 torp do nis, f. 3 3 2 HC.HV1.L m torpede 3 2 HC. L1.L torpedo 3 2 HC.L1. L torpedo 3 2 HC.L1.L HC.HV1.X (1) 1683 conf nis, e, adj. 3 3 2 HC.HV1.L m conf 2 1 HC.HV1 confn 2 1 HC.HC1 confim 2 1 HV1

PAGE 441

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp HC.HV1.X (1) 1684 auctor, ris 3 3 2 HC.HV1.Lm autor 21HV.HC1 autor 21HV.HC1 autor 21HV.HC1 HC.HV1.X (8) 1685 lat. vulg. *currale 3 3 2 HC.HV1.Lm corral 2 1 L.HC1 corral 2 1 HV.HC1 curral 2 1 HV.HC1 1686 abbas, tis, m., del gr. 3 3 2 HC.HV1.L m abat 2 1 L.HC1 abad 2 1 L.HC1 abade 3 2 L.L1.L 1687 coll ris e, adj. 3 3 2 HC.HV1.L f collar 2 1 L.HC1 collar 2 1 L.HC1 colar 2 1 L.HC1 1688 horror, ris, m. 3 3 2 HC.HV1.L m horror 2 1 L. HC1 horror 2 1 L.HC1 horror 2 1 L.HC1 1689 b. ll. missalis 3 3 2 HC.HV1.L m missal 2 1 L.HC1 misal 2 1 L.HC1 missal 2 1 L.HC1 1690 *pugn le 3 3 2 HC.HV1.L m punyal 2 1 L.HC1 pual 2 1 L.HC1 punhal 2 1 L.HC1 1691 lat. tard. s gn le 3 3 2 HC.HV1.L m/f senyal 2 1 L.HC1 seal 2 1 L.HC1 sinal 2 1 L.HC1 1692 terror ris, m. 3 3 2 HC.HV1.L m terror 2 1 L. HC1 terror 2 1 L.HC1 terror 2 1 L.HC1 HC.HV1.X (3) 1693 aff nis 3 3 2 HC.HV1.L m/f af 2 1 L.HV 1 afn 2 1 L.HC1 afim 2 1 L.HV1 1694 glutto, nis, m. 3 3 2 HC.HV1.L m glot 2 1 L. HV1 glotn 2 1 L.HC1 gluto 2 1 L.HV1 1695 comm nis, e, adj. 3 3 2 HC.HV1.L m com 2 2 L.HV1 comn 2 1 L.HC1 comum 2 2 L.HV1 HV.HC1.X (3) 1696 p nens, -entis 3 3 2 HV.HC1.L m ponent 2 1 L. HC1 poniente 3 2 L.HC1.L ponente 3 2 L.HC1.L 169 7 praesens, -entis 3 3 2 HV.HC1.L m present 2 1 L. HC1 presente 441 3 2 L.HC1.L presente 3 2 L.HC1.L 1698 s mentis, is, f. 3 3 2 HV.HC1.L m/f sement 2 1 L. HC1 simiente 3 2 L.HC1.L semente 3 2 L.HC1.L HV.HC1.X (1) 1699 s mestris, e 3 3 2 HV.HC1.L m semestre 3 2 L.HC1 .L semestre 3 2 L.HC1.L semestre 3 2 L.HC1.L HV.HV1.X (1) 1700 quaestor, ris 3 3 2 HV.HV1.L m qestor 2 1 HC.HC1 cuestor 2 1 HC.HC1 questor 2 1 HC.HC1 HV.HV1.X (1) 1701 augur, ris 3 3 2 HV.HV1.L m ugur 2 1 HV.HC1 augur 2 1 HV.HV1 ugure 2 1 HV.HV1 HV.HV1.X (23) 1702 cl mor, ris, m. 3 3 2 HV.HV1.L m clamor 2 1 L. HC1 clamor 2 1 L.HC1 clamor 2 1 L.HC1 1703 aetas tis, f. 3 3 2 HV.HV1.L f edat 2 1 L. HC1 edad 2 1 L.HC1 idade 3 2 L.L1.L 1704 aed lis is, m. 3 3 2 HV.HV1.L m edil 2 1 L.HC1 edil 2 1 L.HC1 edil 2 1 L.HC1 1705 f tor (faet-, foet-), ris, m. 3 3 2 HV.HV1.L m fetor 2 1 L.HC1 hedor 2 1 L.HC1 fedor 2 1 L.HC1 1706 f n lis, e, adj. 3 3 2 HV.HV1.L m final 2 1 L.HC1 final 2 1 L.HC1 final 2 1 L.HC1 1707 mor (not hu-), ris, m. 3 3 2 HV.HV1.L m humor 2 1 L.HC1 humor 2 1 L.HC1 humor 2 1 L.HC1 1708 aequ lis, e, adj. 3 3 2 HV.HV1.L m igual 2 2 L.HC1 igual 2 2 L.HC1 igual 2 2 L.HC1 1709 m jor, ris 3 3 2 HV.HV1.L m major 2 1 L.HC1 mayor 2 1 L.HC1 maior 2 1 L.HC1 171 0 m r lis, e 3 3 2 HV.HV1.L m/f moral 2 1 L.HC1 moral 2 1 L.HC1 moral 2 1 L.HC1

PAGE 442

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 1711 m tor, ris, m. 3 3 2 HV.HV1.Lm motor 21 L.HC1 motor 21L.HC1 motor 21L.HC1 1712 m r lis, e 3 3 2 HV.HV1.L m mural 2 1 L.HC1 mural 2 2 L.HC1 mural 2 2 L.HC1 1713 n t lis, is, m. 3 3 2 HV.HV1.L m natal 2 1 L.HC1 natal 2 1 L.HC1 natal 2 1 L.HC1 1714 poen lis, e, adj. 3 3 2 HV.HV1.L m penal 2 1 L.HC1 penal 2 2 L.HC1 penal 2 2 L.HC1 1715 pl r lis, e, adj. 3 3 2 HV.HV1.L m plural 2 1 L.HC1 plural 2 2 L.HC1.L plural 2 2 L.HC1.L 1716 praen men, nis, n. 3 3 2 HV.HV1.L m prenom 2 1 L. HC1 prenombre 3 2 L.HC1 .L prenome 3 2 L.L1.L 1717 praetor, ris, m. 3 3 2 HV.HV1.L m pretor 2 1 L.HC1 pretor 2 1 L.HC1 pretor 2 1 L.HC1 171 8 pr mas, tis, m.& f. 3 3 2 HV.HV1.L m primat 2 1 L. HC1 primate 3 2 L.L1. L primata 3 2 L.L1.L 1719 pr m ris e 3 3 2 HV.HV1.L f primor 2 1 L.HC1 primor 2 1 L.HC1 primor 2 1 L.HC1 1720 r v lis, e 3 3 2 HV.HV1.L m rival 2 1 L.HC1 rival 2 1 L.HC1 rival 2 1 L.HC1 1721 r mor, ris, m. 3 3 2 HV.HV1.L m rumor 2 1 L.HC1 rumor 2 1 L.HC1 rumor 2 1 L.HC1 1722 s dor, ris, m. 3 3 2 HV.HV1.L m suor 2 1 L.HC1 sudor 2 1 L.HC1 suor 2 1 L.HC1 1723 t tor ris, m. 3 3 2 HV.HV1.L m tutor 2 2 L.HC1 tutor 2 2 L.HC1 tutor 2 2 L.HC1 1724 visor, ris 3 3 2 HV.HV1.L m visor 2 1 L.HC1 visor 2 1 L.HC1 visor 2 1 L.HC1 1725 v c lis is, f. 3 3 2 HV.HV1.L f vocal 2 1 L.HC1 vocal 2 1 L.HC1 vogal 2 1 L.HC1 HV.HV1.X (9) 1726 furo, nis 3 3 2 HV.HV1.L m fur 2 1 L.HV1 hurn 2 1 L.HC1 furo 2 1 L.HV1 1727 l tro nis, m. 3 3 2 HV.HV1.L m lladr 2 1 L.HV1 ladrn 2 1 L.HC1 ladro 2 1 L.HV1 1728 m trix, cis, f. 3 3 2 HV.HV1.L f matriu 2 1 L.HV1 matriz 442 2 1 L.HC1 matriz 2 1 L.HC1 1729 p vo, nis, m. 3 3 2 HV.HV1.L m pa 2 1 L. HV1 pav n 2 1 L.HC1 pavo 2 1 L.HV1 1730 s po nis, m. [Germ.] 3 3 2 HV.HV1.L m sab 2 2 L.HV1 jabn 2 1 L.HC1 sabo 2 1 L.HV1 1731 s pho, nis, m., = 3 3 2 HV.HV1.L m sif 2 1 L.HV1 sifn 2 1 L.HC1 sifo 2 1 L.HV1 1732 lat. tard. t lo, nis 3 3 2 HV.HV1.L m tal 2 1 L.HV1 taln 2 1 L.HC1 talo 2 1 L.HV1 1733 t m nis, m. 3 3 2 HV.HV1.L m tim 2 1 L. HV1 timn 2 1 L.HC1 timo 2 1 L.HV1 1734 T tan, nis, m. 3 3 2 HV.HV1.L m tit 2 1 L.HV1 titn 2 1 L.HC1 tit 2 1 HV1 HV.HV1.X (4) 1735 d cl ve, is, n. 3 3 2 HV.HV1.L m declivi 3 2 L.L1.L declive 3 2 L. L1.L declive 3 2 L.L1.L 1736 d pl ma tis, n., del gr. 3 3 2 HV.HV1.L m diploma 3 2 L.L1.L diploma 3 2 L. L1.L diploma 3 2 L.L1.L 1737 n trix (notrix Quint. 1, 4, 16), cis, f. 3 3 2 HV.HV1.L f nodrissa 3 2 L.L1. L nodriza 3 2 L.L1.L nutriz 2 2 L.HC1 1738 S r n, nis, f., = 3 3 2 HV.HV1.L f sirena 3 2 L.L1.L siren a 3 2 L.L1.L sirena 3 2 L.L1.L HV.HV1.X (1) 1739 l ch n, nis, m., = 3 3 2 HV.HV1.L m liquen 2 2 L1.HC liquen 2 2 L1.HC lquen 2 2 L1.HC L.HC1.X (1) 1740 N vember, bris, adj., 3 3 2 L.HC1.L m novembre 3 2 HC1.L noviembre 3 2 HC1.L Novembro 3 2 HC1.L

PAGE 443

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp L.HC1.X (14) 1741 agens, -entis 3 3 2 L.HC1.L m agent 32 L.HC1 agente 32L.HC1.L agente 32L.HC1.L 1742 amans, -antis 3 3 2 L.HC1 .L m/famant 21L.HC1 amante 32L.HC1.L amante 32L.HC1.L 1743 basaltes is, m. 3 3 2 L.HC1.L m basalt 21L.HC1 basalto 32L.HC1.L basalto 32L.HC1.L 1744 cl ens, -entis, m. 3 3 2 L.HC1.L m client 2 1 L.HC1 cliente 3 2 L.HC1.L cliente 3 2 L.HC1.L 1745 c hors, -rtis, f. 3 3 2 L.HC1.L f cohort 2 1 L. HC1 cohorte 3 2 L.HC1.L coorte 3 2 L.HC1.L 1746 d cens, -entis 3 3 2 L.HC1.L m docent 2 1 L. HC1 docente 3 2 L.HC1.L docente 3 2 L.HC1.L 1747 G g s, antis, m. 3 3 2 L.HC1.L m gegant 2 1 L. HC1 gigante 3 2 L.HC1.L gigante 3 2 L.HC1.L 1748 g rens, ntis 3 3 2 L.HC1.L m gerent 2 1 L.HC1 gerente 3 2 L.HC1.L gerente 3 2 L.HC1.L 1749 p rens, entis, m. & f. 3 3 2 L.HC1.L m/f paren t 2 1 L.HC1 pariente 3 2 L.HC1.L parente 3 2 L.HC1.L 1750 p tens, entis 3 3 2 L.HC1.L m patent 2 1 L. HC1 patente 3 2 L.HC1.L patente 3 2 L.HC1.L 1751 r gens, entis, m. 3 3 2 L.HC1.L m regent 2 1 L.HC1 regente 3 2 L.HC1.L regente 3 2 L.HC1.L 1752 tr dens entis, m. 3 3 2 L.HC1.L m trident 2 1 L.HC1 tridente 3 2 L.HC1.L tridente 3 2 L.HC1.L 1753 v cans, -antis 3 3 2 L.HC1.L f vacant 2 1 L. HC1 vacante 3 2 L.HC1.L vacante 3 2 L.HC1.L 1754 v dens entis, m. 3 3 2 L.HC1.L m vident 2 1 L.HC1 vidente 3 2 L.HC1.L vidente 3 2 L.HC1.L 1755 v lans antis 3 3 2 L.HC1.L m volant 2 1 L.HC1 volante 3 2 L.HC1.L voante 3 2 L.HC1.L L.HC1.X (9) 1756 ch racter ris, m. 3 3 2 L.HC1.L m carcter 3 2 L.HC1.L carcter 3 2 L.HC1.L carcter 3 2 L.HC1.L 1757 d cember bris, m. 3 3 2 L.HC1.L m desembre 3 2 L. HC1.L diciembre 443 3 2 L.HC1.L Dezembro 3 2 L.HC1.L 1758 ecl psis is, f., del gr. 3 3 2 L.HC1.L m eclipsi 3 2 L.HC1.L eclipse 3 2 L.HC1.L eclipse 3 2 L.HC1.L 1759 ph lanx, angis, f. = 3 3 2 L.HC1.L f falange 3 2 L.HC1.L falange 3 2 L.HC1.L falange 3 2 L.HC1.L 1760 f rensis, e, adj. 3 3 2 L.HC1.L m forense 3 2 L. HC1.L forense 3 2 L.HC1.L forense 3 2 L.HC1.L 1761 siccans, -antis 3 3 2 L.HC1.L m secante 3 2 L.HC1.L secante 3 2 L.HC1.L secante 3 2 L.HC1.L 1762 synopsis, is, f., = 3 3 2 L.HC1.L f sinopsi 3 2 L.HC1.L sinopsis 3 2 L.HC1.HC sinopse 3 2 L.HC1.L 1763 tr mestris e, adj. 3 3 2 L.HC1.L m trimestre 3 2 L.HC1.L trimestre 3 2 L.HC1 .L trimestre 3 2 L.L1.L 1764 b mestris e 2 3 2 L.HC1.L m bimestre 3 2 L.HC1 .L bimestre 3 2 L.HC1.L bimestre 3 2 L.HC1.L L.HC1.X (1) 176 5 b son ontis, m. 3 3 2 L.HC1.L m bis 2 1 L.HV1 bisonte 3 2 L.HC1.L biso 2 1 L.HV1 L.HV1.X (3) 1766 l men, nis, n. 3 3 2 L.HV1.L m alum 2 1 HC1 alumbre 3 2 HC1.L alume 3 2 L1.L 1767 l g men, nis, n. 3 3 2 L.HV1.L m llegum 2 1 HC1 legumbre 3 2 HC1.L legume 3 2 L.L1.L 1768 v l men nis, n. 3 3 2 L.HV1.L m volum 2 1 HC1 volumen 3 2 L1.HC volume 3 2 L1.L L.HV1.X (45) 1769 pr lis, is, m. 3 3 2 L.HV1.L m abril 2 1 L.HC1 abril 2 1 L.HC1 abril 2 1 L.HC1 1770 amor, ris 3 3 2 L.HV1.L m amor 2 1 L.HC1 amor 2 1 L.HC1 amor 2 1 L.HC1

PAGE 444

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 1771 b t men nis, n. 3 3 2 L.HV1. L m betum 2 1 L.HC1 betn 2 1 L.HC1 betume 3 2 L.L1.L 1772 b r mis is, f. 3 3 2 L.HV1.L f birrem 2 1 L. HC1 birreme 3 2 L.L1.L birreme 3 2 L.L1.L 1773 c lor ris, m. 3 3 2 L.HV1.L m calor 2 1 L.HC1 calor 2 1 L.HC1 calor 2 1 L.HC1 1774 c n lis is, m. 3 3 2 L.HV1.L m canal 2 1 L.HC1 canal 2 1 L.HC1 canal 2 1 L.HC1 1775 b. lat. cas lis 3 3 2 L.HV1.L m casal 2 1 L.HC1 casal 2 1 L.HC1 casal 2 1 L.HC1 1776 Cyclops, pis (acc. pem or pa), m., = 3 3 2 L.HV1.L m ciclop 2 1 L.HC1 cclope 3 3 L1.L.L ciclope 3 2 L.L1.L 1777 c lor, ris, m. 3 3 2 L.HV1.L m color 2 1 L.HC1 color 2 1 L.HC1 cor (f.) 1 1 HC1 177 8 d lor ris, m. 3 3 2 L.HV1.L m dolor 2 1 L.HC1 dolor 2 1 L.HC1 dor 1 1 HC1 1779 d lis e, adj. 3 3 2 L.HV1.L m dua l 2 1 L.HC1 dual 2 1 L.HC1 dual 2 1 L.HC1 1780 f vor, ris, m. 3 3 2 L.HV1.L m favor 2 1 L.HC1 favor 2 1 L.HC1 favor 2 1 L.HC1 1781 f d lis, is, m. 3 3 2 L.HV1.L m fidel 2 2 L.HC1 fiel 1 1 HC1 fiel 1 1 HC 1782 fl or, ris, m. 3 3 2 L.HV1.L m fluor 2 1 L.HC1 flor 2 2 L2.HC fluor 2 1 L.HC1 1783 foc re 3 3 2 L.HV1.L m fogar 2 1 L.HC1 hogar 2 1 L.HC1 fogar 2 1 L.HC1 1784 fr gor, ris, m. 3 3 2 L.HV1.L m fragor 2 1 L. HC1 fragor 2 1 L.HC1 fragor 2 1 L.HC1 1785 f ror, ris, m. 3 3 2 L.HV1.L m furor 2 1 L.HC1 furor 2 1 L.HC1 furor 2 1 L.HC1 1786 h nor or h nos, ris, m. 3 3 2 L.HV1.L m honor 2 1 L. HC1 honor 2 1 L.HC1 honor 2 1 L.HC1 1787 l quor, ris, m. 3 3 2 L.HV1.L m licor 21L. HC1 licor 444 2 1 L.HC1 licor 2 1 L.HC1 1788 l bor, ris, m. 3 3 2 L.HV1.L m llavor 2 1 L.HC1 labor 2 1 L.HC1 lavor 2 1 L.HC1 1789 l c lis, e, adj. 3 3 2 L.HV1.L m local 2 1 L.HC1 local 2 1 L.HC1 local 2 1 L.HC1 1790 minor, ris 3 3 2 L.HV1.L m/f menor 2 1 L.HC1 menor 2 1 L.HC1 menor 2 1 L.HC1 1791 myops, pis, adj.= 3 3 2 L.HV1.L m/f miop 2 1 L. HC1 miope 2 2 L1.L mope 3 3 L1.L.L 1792 m l ris, is, m. 3 3 2 L.HV1.L m molar 2 1 L.HC1 molar 2 1 L.HC1 molar 2 1 L.HC1 1793 n gror, ris, m. 3 3 2 L.HV1.L m negror 2 1 L.HC1 negror 2 1 L.HC1 negror 2 1 L.HC1 1794 n c lis e, adj. 3 3 2 L.HV1.L m nogall 2 1 L.HC1 nogal 2 1 L.HC1 nogal 2 1 L.HC1 1795 lor, ris, m. 3 3 2 L.HV1.L m olor 2 2 L. HC1 olor 2 2 L.HC1 olor 2 2 L.HC1 1796 p d lis, e, adj. 3 3 2 L.HV1.L m pedal 2 1 L.HC1 pedal 2 1 L.HC1 pedal 2 1 L.HC1 1797 pr or (alt. pr us), ris, adj. 3 3 2 L.HV1.L m prior 2 1 L.HC1 prior 1 1 HC1 prior 2 1 L.HC1 1798 p dor, ris, m. 3 3 2 L.HV1.L m pudor 2 1 L. HC1 pudor 2 1 L.HC1 pudor 2 1 L.HC1 1799 r pax, cis, m.& f. 3 3 2 L.HV1.L m rapa 2 1 L.HC1 rapaz 2 1 L.HC1 rapaz 2 1 L.HC1 1800 r gor, ris, m. 3 3 2 L.HV1.L m rigor 2 1 L.HC1 rigor 2 1 L.HC1 rigor 2 1 L.HC1 1801 r bor, ris, m. 3 3 2 L.HV1.L m rubor 2 1 L.HC1 rubor 2 1 L.HC1 rubor 2 1 L.HC1 1802 s por, ris, m. 3 3 2 L.HV1.L m sabor 2 1 L. HC1 sabor 2 1 L.HC1 sabor 2 1 L.HC1 1803 s l s, tis, f. 3 3 2 L.HV1.L f salut 2 1 L. HC1 salud 2 1 L.HC1 sade 3 2 L.L1.L 1804 s quax, cis, adj. 3 3 2 L.HV1.L m/f sequa 2 1 L.HC1 secuaz 2 1 L.HC1 sequaz 2 1 L.HC1

PAGE 445

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 1805 s por, ris, m. 3 3 2 L.HV1.L m sopor 21 L.HC1 sopor 21L.HC1 sopor 21L.HC1 1806 t p te is, n. 3 3 2 L.HV1.L m tapet 21L.HC1 tapete 32L. L1.L tapete 32L.L1.L 1807 t mor, ris, m. 3 3 2 L.HV1.L m temor 21 L.HC1 temor 21L.HC1 temor 21L.HC1 1808 t nor ris, m. 3 3 2 L.HV1.L m tenor 22 L.HC1 tenor 22L.HC1 tenor 2 2 L.HC1 1809 tr -r mis e, adj. 3 3 2 L.HV1.L m trirrem 2 1 L.HC1 trirreme 3 2 L.L1.L trirreme 3 2 L.L1.L 1810 t mor ris, m. 3 3 2 L.HV1.L m tumor 2 1 L. HC1 tumor 2 1 L.HC1 tumor 2 1 L.HC1 1811 v lor ris, m. 3 3 2 L.HV1.L m valor 2 1 L.HC1 valor 2 1 L.HC1 valor 2 1 L.HC1 181 2 v por, ris, m. 3 3 2 L.HV1.L m vapor 2 1 L. HC1 vapor 2 1 L.HC1 vapor 2 1 L.HC1 1813 v gor ris, m. 3 3 2 L.HV1.L m vigor 2 1 L.HC1 vigor 2 1 L.HC1 vigor 2 1 L.HC1 L.HC1.X (prothesis) (2) 1814 sch l ris, e, adj. 3 3 2 L.HV1.L m escolar 3 1 L.HC1 escolar 3 1 L.HC1 escolar 3 1 L.HC1 1815 st por, ris, m. 3 3 2 L.HV1.L m estupor 3 1 L. HC1 estupor 3 1 L.HC1 estupor 3 1 L.HC1 L.HC1.X (2) 1816 Dr co nis, m., del gr. 3 3 2 L.HV1.L m drag 2 1 L.HV1 dragn 2 1 L.HC1 drago 2 1 L.HV1 1817 l o, nis, m. 3 3 2 L.HV1.L m lle 2 1 L.HV1 len 2 1 L.HC1 leo 2 2 L.HV1 L.HC1.X (2) 1818 b. lat. scriba, nis, del lat. scriba, -ae 3 3 2 L.HV1.L m escriv 3 1 L.HV1 escribano (ant. escrivn) 4 2 L.L1.L escrivo 3 1 L.HV1 L.HC1.X (1) 1819 transf s o nis, f. 3 5 2 L.HV1.L f transfusi 4 1 L. HV1 transfusin 445 3 1 HC1 transfuso 3 1 HV1 L.HC1.X (11) 1820 phr n sis, is, f., = 3 3 2 L.HV1.L f frenes 3 1 L.L.L1 frenes 3 1 L.L.L1 frenesi 3 2 L.L1.L 1821 gr v men, nis, n. 3 3 2 L.HV1.L m gravamen 3 2 L. L1.HC gravamen 3 2 L.L1. HC gravame 3 2 L.L1.L 1822 nephr tis, dis, f., = 3 3 2 L.HV1.L f nefritis 3 2 L.L1.HC nefritis 3 2 L.L1.HC nefrite 3 2 L.L1.L 1823 c d ver ris, n. 3 3 2 L.HV1.L m cadver 3 3 L.L1.HC cadver 3 3 L.L1.HC cadver 3 3 L.L1.HC 1824 g des, is, m. 3 3 2 L.HV1.L m geoda 3 2 L.L1.L geoda 3 2 L. L1.L geode 3 2 L.L1.L 1825 l b do, nis, f. 3 3 2 L.HV1.L f libido 3 2 L. L1.L libido 3 2 L.L1.L libido 3 2 L.L1.L 1826 necr sis, is, f., = 3 3 2 L.HV1.L f necrosi 3 2 L.L1.L nec rosis 3 2 L.L1.HC necrose 3 2 L.L1.L 1827 p ma, tis, n. 3 3 2 L.HV1.L m poema 3 2 L.L1.L poema 3 2 L.L1. L poema 3 2 L.L1.L 1828 probl ma, tis, n.= 3 3 2 L.HV1.L m problema 3 2 L.L1.L pr oblema 3 2 L.L1.L problema 3 2 L.L1.L 1829 vir go, nis 3 3 2 L.HV1.L f virago 3 2 L.L1.L virago 3 2 L.L1.L virago 3 2 L.L1.L 1830 c m n, nis, n. 3 3 2 L.HV1.L m acumen 3 2 L1.HC acumen 3 2 L1.HC acume 3 2 L1.L

PAGE 446

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp HC1.L.X (2) 1831 sanguis, nis, m. 3 3 3 HC1.L.L m/fsang 11HC1 sangre 22HC1.L sangue 22HC1.L 1832 pectus, ris, n. 3 3 3 HC1.L.L m pit 11HC1 pecho 22L1.L peito 22HV1.L HC1.L.X (10) 1833 aster, ris, m., = 3 3 3 HC1.L.L m ster 22HC1.HC ster 22HC1.HC ster 22HC1.HC 1834 calc lus i, m. 3 3 3 HC1.L.L m clcul 22HC1 .HC clculo 33HC1.L.L clculo 33HC1.L.L 1835 carcer ris, m. 3 3 3 HC1.L.L f crcer 22HC1. HC crcel 22HC1.HC crcere 33HC1.L.L 1836 index, d cis, m.& f. 3 3 3 HC1.L.L m ndex 22HC1. HC ndice 33HC1.L.L ndice 33HC1.L.L 1837 martyr, yris, m.& f. = 3 3 3 HC1.L.L m/f mrtir 2 2 HC1.HC mrtir 2 2 HC1.HC mrtir 2 2 HC1.HC 1838 pollen, nis, n. 3 3 3 HC1.L.L m pollen 2 2 HC1.HC polen 2 2 L1.HC plen 2 2 L1.HC 1839 princeps, c pis, m. 3 3 3 HC1.L.L m prncep 2 2 HC1.HC prncipe 3 3 HC1.L.L prncipe 3 3 HC1.L.L 1840 rept lis, e, adj. 3 3 3 HC1.L.L m rptil 2 2 HC1.HC reptil 2 1 HC.HC1 rptil 2 1 HC.HC1 1841 vertex, cis 3 3 3 HC1.L.L m vrtex 2 2 HC1.HC vrtice 3 3 HC1.L.L vrtice 3 3 HC1.L.L 1842 vortex, cis 3 3 3 HC1.L.L m vrtex 2 2 HC1.HC vrtice 3 3 HC1.L.L vrtice 3 3 HC1.L.L HC1.L.X (10) 1843 arbor, ris 3 3 3 HC1.L.L m arbre 2 2 HC1.L rbol 2 2 HC1.HC rvore 3 3 HC1.L.L 1846 consul, lis, m. 3 3 3 HC1.L.L m cnsol 2 2 HC1.L cnsul 2 2 HC1.L cnsul 2 2 HC1.L 1847 hosp s, tis, m. 3 3 3 HC1.L.L m hoste 2 2 HC1.L husped 446 2 2 HC1.HC hspede 3 3 HC1.L.L 1561 aspis dis, f., del gr 3 3 3 HC1.L.L m jaspi 2 2 HC1.L jaspe 2 2 HC1.L jaspe 2 2 HC1.L 1862 marmor, ris, n. 3 3 3 HC1.L.L m marbre 2 2 HC1 .L mrmol 2 2 HC1.HC mrmore 3 3 HC1.L.L 1863 margo, nis, m.& f. 3 3 3 HC1.L.L m marge 2 2 HC1 .L margen 2 2 HC1.L margem 2 2 HC1.HV 1864 nunt us, a, um, adj. 3 3 3 HC1.L.L m nunci 2 2 HC1.L nuncio 2 2 HC1.L nncio 2 2 HC1.L 1865 ordo, nis, m. 3 3 3 HC1.L.L m/f orde 2 2 HC1.L orden 2 2 HC1.HC ordem 2 2 HC1.HV 1866 pallium, i, n. 3 3 3 HC1.L.L m pal li 2 2 HC1.L palio 2 2 L1.L plio 2 2 L1.L 1867 virgo, nis 3 3 3 HC1.L.L f verge 2 2 HC1.L virgen 2 2 HC1.HC virgem 2 2 HC1.HV HC1.L.X (9) 1853 complex pl cis, adj. 3 3 3 HC1.L.L m/f cmplice 3 3 HC1.L.L cmplice 3 3 HC1.L.L cmplice 3 3 HC1.L.L 1854 coniux, gis, m./f. 3 3 3 HC1.L.L m/f cnjuge 3 3 HC1.L .L cnyuge 3 3 HC1.L.L cnjuge 3 3 HC1.L.L 1855 cuspis dis, f. 3 3 3 HC1.L.L f cspide 3 3 HC1.L.L cspide 3 3 HC1.L.L cspide 3 3 HC1.L.L 1856 emph sis is, f., del gr. 3 3 3 HC1.L.L f mfasi 3 3 HC1.L.L nfasis 3 3 HC1.L.HC nfase 3 3 HC1.L.L 1857 ind les, is, f 3 3 3 HC1.L.L f ndole 3 3 HC1.L .L ndole 3 3 HC1.L.L ndole 3 3 HC1.L.L 1858 omn bus 3 3 3 HC1.L.L m mnibus 3 3 HC1.L.L mnibus 3 3 HC1.L.L nibus 3 3 L1.L.L 1859 pyxis, dis, f. = 3 3 3 HC1.L.L f pxide 3 3 HC1.L.L pxide 3 3 HC1.L.L pxide 3 3 HC1.L.L 1860 synth sis, is, f., = 3 3 3 HC1.L.L f sntesi 3 3 HC1.L.L snt esis 3 3 HC1.L.HC sntese 3 3 HC1.L.L

PAGE 447

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 1861 syst le < 3 3 3 HC1.L.L f sstole 3 3 HC1.L.L s stole 3 3 HC1.L.L sstole 3 3 HC1.L.L HC1.L.X (1) 1862 foss lis, e, adj. 3 3 3 HC1.L.L m fss il 2 2 L1.HC fsil 2 2 L1.HC fssil 2 2 L1.HC HC1.L.X (1) 1863 cann bis is, f. 3 3 3 HC1.L.L f cnnabis 3 3 L1.L.L cannabis 3 2 L. L1.L canabis 3 2 L.L1.L HV1.L.X (1) 1864 c mex cis, m. 3 3 3 HV1.L.L m xinxe 2 2 HC1.L chinche 2 2 HC1.L chinche 2 2 HC1.L HV1.L.X (2) 1865 h ros, is, m.= 3 3 3 HV1.L.L m heroi 2 1 L.HV1 hroe 3 3 L1.L.L heri 2 1 L.HV1 1866 j dex cis, com. 3 3 3 HV1.L.L m ju (also HV 1) 2 1 L.HV1 juez 1 1 HC1 juiz 2 1 L.HC1 HV1.L.X (6) 1867 d b lis, e, adj. 3 3 3 HV1.L.L m/f dbil 2 2 L1.HC dbil 2 2 L1.L dbil 2 2 L1.L 1868 aether, ris, m. 3 3 3 HV1.L.L m ter 2 2 L1.HC ter 2 2 L1.HC ter 2 2 L1.HC 1869 gl ten, nis, n. 3 3 3 HV1.L.L m gluten 2 2 L1.HC gluten 2 2 L1.L glten 2 2 L1.L 1870 l m s, tis, m. 3 3 3 HV1.L.L m lmit 2 2 L1.HC lmite 3 3 L1.L. L limite 3 3 L1.L.L 1871 p bes & p ber, ris, m. 3 3 3 HV1.L.L m pber 2 2 L1.HC pber 2 2 L1. HC pbere 3 3 L1.L.L 1872 trames, tis 3 3 3 HV1.L.L m trmit 2 2 L1.HC trmite 3 3 L1.L.L trmite 3 3 L1.L.L HV1.L.X (3) 1873 m b lis, e 3 3 3 HV1.L.L m moble 2 2 L1.L mueble 447 2 2 L1.L mvel 2 2 L1.HC 1874 n b lis is, m. 3 3 3 HV1.L.L m nob le 2 2 L1.L noble 2 2 L1.L nobre 2 2 L1.L 1875 pauper, p ris, m.& f. 3 3 3 HV1.L.L m/f pobr e 2 2 L1.L pobre 2 2 L1.L pobre 2 2 L1.L L1.L.X (1) 1876 v r dis e 3 3 3 L1.L.L m verd 1 1 HC1 verde 2 2 HC1.L verde 2 2 HC1.L L1.L.X (3) 1877 c lix cis, m. 3 3 3 L1.L.L m calze 2 2 HC1.L cliz 2 2 L1.HC clice 3 3 L1.L.L 1878 c ms, tis, m.&f. 3 3 3 L1.L.L m comte 2 2 HC1.L conde 2 2 HC1.L conde 2 2 HC1.L 1879 gener, gen ri 3 3 3 L1.L.L m gendre 2 2 HC1.L yerno 2 2 HC1.L genro 2 2 HC1.L L1.L.X (2) 1880 aer, ris, m. 3 3 3 L1.L.L m aire 2 2 HV1.L aire 2 2 HV1.L ar 1 1 HC1 1881 rheuma, tis, n., = 3 3 3 L1.L.L m reuma 2 2 HV1.L reuma 2 2 HV1.L reuma 2 2 HV1.L L1.L.X (1) 1882 v rix cis, m. & f. 3 3 3 L1.L.L f varia 3 2 L.L1.L variz 2 1 L.HC1 variz 2 1 L.HC1 L1.L.X (8) 1883 c non nis, m., del gr. 3 3 3 L1.L.L m cno n 2 2 L1.HC canon 2 2 L1.HC cnon 2 2 L1.HC

PAGE 448

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 1884 Hymen, nis, m., = 3 3 3 L1.L.L m himen 22L1.HC himen 22L1.HC hmen 22L1.HC 1885 L m res, um, m. 3 3 3 L1.L.L m lmur 2 2 L1.HC lmur 2 2 L1.HC lmure 3 3 L1.L.L 1886 nyx, ychis, m. & f. 3 3 3 L1.L.L m nix 2 2 L1.HC nice 3 3 L1.L.L nix 2 2 L1.HC 1887 pr cer, ris, m. 3 3 3 L1.L.L m prcer 2 2 L1.HC prcer 2 2 L1. HC prcer 2 2 L1.HC 1888 p gil, lis, m. 3 3 3 L1.L.L m pgil 2 2 L1.HC pgil 2 2 L1.HC pgil 2 2 L1.HC 1889 r g men nis, n 3 3 3 L1.L.L m rgim 2 2 L1. HC rgimen 3 3 L1.L.HC regime 3 2 L.L1.L 1890 s m le, is, n. 3 3 3 L1.L.L m smil 2 2 L1.HC smil 2 2 L1. HC smil 2 2 L1.HC L1.L.X (4) 1891 cil a, pl. de c l um ii, n. 3 3 3 L1.L.L f cella 2 2 L1.L ceja 2 2 L1.L celha 2 2 L1.L 1892 h mo, nis, m. 3 3 3 L1.L.L m home 2 2 L1.L hombre 2 2 HC1.L homem 2 2 1893 j v nis is, adj.&subst. 3 3 3 L1.L.L m/f jov e 2 2 L1.L joven 2 2 L1.L jovem 2 2 L1.HV 1894 l pus, ris, m. 3 3 3 L1.L.L m llebre 2 2 L1.L liebre 2 2 L1.L lebre 2 2 L1.L L1.L.X (3) 1895 ar es, tis 3 3 3 L1.L.L m/f ries 3 3 L1. L.HC aries 2 2 L1.HC ries 2 2 L1.HC 1896 sp c men, nis, n. 3 3 3 L1.L.L m espcimen 4 3 L1. L.HC espcimen 4 3 L1.L. HC espcimen 4 3 L1.L.HC L1.L.X (3) 1897 C c ro nis, m. 3 3 3 L1.L.L m ccero 3 3 L1. L.L ccero 3 3 L1.L. L Ccero 3 3 L1.L.L 1898 ag pe, y este del gr. 3 3 3 L1.L.L m gape 3 3 L1.L.L gape 3 3 L1.L.L gape 3 3 L1.L.L 1899 dyas dis, f., del gr. 3 3 3 L1.L.L f d ade 3 3 L1.L.L dada 448 3 3 L1. L.L dada 3 3 L1.L.L 1900 g nus, ris, n. 3 3 3 L1.L.L m gnere 3 3 L1. L.L gnero 3 3 L1.L.L gnero 3 3 L1.L.L 1901 g n sis, is, f. 3 3 3 L1.L.L f gnesi 3 3 L1. L.L gnesis 3 3 L1.L. HC gnesis 3 3 L1.L.HC 1902 h lix, cis, f. 3 3 3 L1.L.L f hlice 3 3 L1. L.L hlice 3 3 L1.L.L hlice 3 3 L1.L.L 1903 s lex, cis, m.& f. 3 3 3 L1.L.L f slice 3 3 L1.L.L slice 3 3 L1. L.L slice 3 3 L1.L.L 1904 tr p s p dis, m., del gr. 3 3 3 L1.L.L m tr pode 3 3 L1.L.L trpode 3 3 L1. L.L trpode 3 3 L1.L.L HC.HC1.X (6) 1905 ascendens, -entis 3 4 2 HC.HC1.L m ascendent 3 1 HC.HC1 ascendente 4 2 HC.HC1.L ascendente 4 2 HC.HC1.L 1906 assistens, -entis 3 4 2 HC.HC1.L m assistent 3 1 HC.HC1 asistente 4 2 HC. HC1.L assistente 4 2 HC.HC1.L 1907 adstringens, -entis 3 4 2 HC.HC1.L m astring ent 3 1 HC.HC1 astringente 4 2 HC.HC1.L adstringente 4 2 HC.HC1.L 1908 contingens, -entis 3 4 2 HC.HC1 .L m contingent 3 1 HC.HC1 continge nte 4 2 HC.HC1.L contingente 4 2 HC.HC1.L 1909 d linquens, -entis 3 4 2 HC.HC1.L m/f delinqent 3 1 HC.HC1 delincuente 4 2 HC.HC1.L delinquente 4 2 HC.HC1.L 1910 intendens, -entis 3 4 2 HC.HC1.L m intendent 3 1 HC.HC1 intendente 4 2 HC.HC1.L intendente 4 2 HC.HC1.L HC.HV1.X (24) 1911 ascensor, ris 3 4 2 HC.HV1.L m ascensor 3 1 HC.HC1 ascensor 3 1 HC.HC1 ascensor 3 1 HC.HC1 1912 collector, ris 3 4 2 HC.HV1.L m colllector 3 1 HC.HC1 colector 3 1 HC.HC1 coletor 3 1 L.HC1 1913 exactor, ris, m. 3 4 2 HC.HV1.L m exactor 3 1 HC. HC1 exactor 3 1 HC.HC1 exactor 3 1 HC.HC1 1914 exemplar, ris, n. 3 4 2 HC.HV1.L m exemplar 3 1 HC.HC1 ejemplar 3 1 HC.HC1 exemplar 3 1 HC.HC1

PAGE 449

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 1915 facultas, tis 3 4 2 HC.HV1.Lf facultat 31HC.HC1 facultad 31HC.HC1 faculdade 42HC.L1.L 1916 impostor, ris, m. 3 4 2 HC.HV1.Lm impostor 31HC. HC1 impostor 31HC.HC1 impostor 31HC.HC1 1917 infractor, ris, m. 3 4 2 HC.HV1.Lm infractor 31HC. HC1 infractor 31HC.HC1 infractor 31HC.HC1 1918 l bertas, tis, f. 3 4 2 HC.HV1.Lf llibertat 31HC. HC1 libertad 31HC.HC1 liberdade 31HC.HC1 1919 m jestas, tis, f. 3 4 2 HC.HV1.L f majestat 3 1 HC.HC1 majestad 3 1 HC.HC1 majestade 4 2 HC.L1.L 1920 maxill ris, e 3 4 2 HC.HV1.L m maxillar 3 1 HC. HC1 maxilar 3 1 L.HC1 maxilar 3 1 L.HC1 1921 offensor, ris, m. 3 4 2 HC.HV1.L m ofensor 3 1 HC.HC1 ofensor 3 1 HC.HC1 ofensor 3 1 HC.HC1 192 2 p testas, tis, f. 3 4 2 HC.HV1.L f potestat 3 1 HC.HC1 potestad 3 1 HC.HC1 potestade 4 2 HC.L1.L 1923 praeceptor, ris, m. 3 4 2 HC.HV1.L m preceptor 3 1 HC. HC1 preceptor 3 1 HC.HC1 preceptor 3 1 HC.HC1 1924 praecursor, ris, m. 3 4 2 HC.HV1.L m precursor 3 1 HC. HC1 precursor 3 1 HC.HC1 precursor 3 1 HC.HC1 1925 pr ductor, ris, m. 3 4 2 HC.HV1.L m productor 3 1 HC. HC1 productor 3 1 HC.HC1 produtor 3 1 L.HC1 1926 p bertas, tis, f. 3 4 2 HC.HV1.L f pubertat 3 1 HC. HC1 pubertad 3 1 HC.HC1 puberdade 4 2 HC.L1.L 1927 qu drant lis e, adj. 3 4 2 HC.HV1.L m quadrantal 3 1 HC.HC1 cuadrantal 3 1 HC.HC1 quadrantal 3 1 HV.HC1 1928 r ceptor, ris, m. 3 4 2 HC.HV1.L m receptor 3 1 HC. HC1 receptor 3 1 HC.HC1 receptor 3 1 HC.HC1 1929 r demptor, ris, m. 3 4 2 HC.HV1.L m redemptor 3 1 HC. HC1 redentor 3 1 HC.HC1 redentor 3 1 HV.HC1 1930 s cerdos, tis, m.& f. 3 4 2 HC.HV1.L m sacerdot 3 1 HC. HC1 sacerdote 4 2 HC.L1. L sacerdote 4 2 HC.L1.L 1931 s nect s, tis,f. 3 4 2 HC.HV1.L f senectut 3 1 HC.HC1 senectud 3 1 HC.HC1 senectude 4 2 HC.L1.L 1932 tempestas tis, f. 3 4 2 HC.HV1.L f tempestat 3 1 HC.HC1 tempestad 3 1 HC.HC1 tempestade 4 2 HC.L1.L 1933 tr ductor ris, m. 3 4 2 HC.HV1.L m traductor 3 1 HC.HC1 traductor 449 3 1 HC.HC1 traductor 3 1 HC.HC1 1934 voluntas, tis 3 4 2 HC.HV1.L f voluntat 3 1 HC.HC1 voluntad 3 1 HC.HC1 vontade 3 2 HV.L1.L HC.HV1.X (1) 1935 n mencl tor, ris, m. 3 4 2 HC.HV1.L m nomencltor 4 2 HC.L1.HC nomencltor 4 2 HC.L1.HC nome nclador 4 1 L.HC1 HC.HV1.X (1) 1936 acceptor ris, m. 3 4 2 HC.HV1.L m astor 2 1 HC1 azor 2 1 HC1 aor 2 1 HC1 HC.HV1.X (7) 1937 assessor, ris 3 4 2 HC.HV1.L m assessor 3 1 L.HC1 asesor 3 1 L.HC1 assessor 3 1 L.HC1 1938 oppressor, ris, m. 3 4 2 HC.HV1.L m opressor 3 1 L.HC1 opresor 3 1 L.HC1 opressor 3 1 L.HC1 1939 percussor, ris, m. 3 4 2 HC.HV1.L m percussor 3 1 L.HC1 percusor 3 1 L.HC1 percussor 3 1 L.HC1 1940 pr fessor, ris, m. 3 4 2 HC.HV1.L m professor 3 1 L. HC1 profesor 3 1 L.HC1 professor 3 1 L.HC1 1941 r pressor ris, m. 3 4 2 HC.HV1.L m repressor 3 1 L. HC1 represor 3 1 L.HC1 repressor 3 1 L.HC1 1942 successor, ris, m. 3 4 2 HC.HV1.L m successor 3 1 L. HC1 sucesor 3 1 L.HC1 sucessor 3 1 L.HC1 1943 b enn lis e, adj. 3 4 2 HC.HV1.L m biennal 3 1 L.L.HC1 bienal 2 1 L.HC1 bienal 3 1 L.HC1 HV.HC1.X (5) 1944 adhaerens, -entis 3 4 2 HV.HC1.L m adherent 3 1 L.HC1 adherente 4 2 L.HC1.L aderente 4 2 L.HC1.L 1945 conf dens entis 3 4 2 HV.HC1.L m/f c onfident 3 1 L.HC1 confidente 4 2 L.HC1.L confidente 4 2 L.HC1.L 1946 mend cans, antis 3 4 2 HV.HC1.L m/f mendicant 3 1 L. HC1 mendicante 4 2 L.HC1.L mendicante 4 2 L.HC1.L

PAGE 450

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 1947 praec dens, -entis 3 4 2 HV.HC1.Lm precedent 31L.HC1 precedente 4 2 L.HC1.L precedente 4 2 L.HC1.L 1948 ves cans, -antis 3 4 2 HV.HC1.L m vesicant 3 1 L. HC1 vesicante 4 2 L.HC1.L vesicante 4 2 L.HC1.L HV.HC1.X (20) 1949 adi tor, ris 3 4 2 HV.HV1.L m adjutor 3 1 L.HC1 adjutor 3 1 L.HC1 adjutor 3 1 L.HC1 1950 aud tor, ris 3 4 2 HV.HV1.L m auditor 3 1 L.HC1 auditor 3 1 L.HC1 auditor 3 1 L.HC1 1951 dict tor, ris 3 4 2 HV.HV1.L m dictador 3 1 L.HC1 dictador 3 1 L.HC1 ditador 3 1 L.HC1 1952 aequ tor, ris 3 4 2 HV.HV1.L m equador 3 1 L.HC1 ecuador 3 1 L.HC1 equador 3 1 L.HC1 1953 r tor, ris, m. 3 4 2 HV.HV1.L m orador 3 1 L. HC1 orador 3 1 L.HC1 orador 3 1 L.HC1 1954 lat. vulg. p l t re
PAGE 451

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp L.HC1.X (18) 1975 acc dens, entis, n. 3 4 2 L.HC1.L m accident 31 L.HC1 accidente 42L.HC1 .L acidente 42L.HV1.L 1976 amb ens, -entis 3 4 2 L.HC1.L m ambient 31 L.HC1 ambiente 32HC1.L ambiente 3 2 HV1.L 1977 cons quens entis, n. 3 4 2 L.HC1.L m consegent 3 1 L. HC1 consiguiente 4 2 L.HC1.L conseguinte 4 2 L.HV1.L 1978 cons quens entis, n. 3 4 2 L.HC1.L m conseqent 3 1 L.HC1 consecuente 4 2 L.HC1.L consequente 4 2 L.HV1.L 1979 VL. diamas, -antis < lat. ad mas, -antis, m. del gr. 3 4 2 L.HC1.L m diamant 3 1 L.HC1 diam ante 3 2 L.HC1.L di amante 3 2 L.HV1.L 1980 el phas, -antis, y este del gr. 3 4 2 L.HC1.L m elefant 3 1 L.HC1 elef ante 4 2 L.HC1.L elefante 4 2 L.HV1.L 1981 ind gens, entis 3 4 2 L.HC1.L m indigent 3 1 L.HC1 indigente 4 2 L.HC1.L indigente 4 2 L.HV1.L 1982 ind lens, -entis 3 4 2 L.HC1.L m indolent 3 1 L.HC1 indolente 4 2 L.HC1.L indolente 4 2 L.HV1.L 1983 inn cens, entis, adj. 3 4 2 L.HC1.L m innocent 3 1 L.HC1 inocente 4 2 L.HC1 .L inocente 4 2 L.HV1.L 1984 ins lens, ntis, adj. 3 4 2 L.HC1.L m insolent 3 1 L. HC1 insolente 4 2 L.HC1.L insolente 4 2 L.HV1.L 1985 occ dens, -entis, adj. 3 4 2 L.HC1.L m occident 3 1 L.HC1 occidente 4 2 L.HC1 .L ocidente 4 2 L.HV1.L 1986 r ens, entis, adj. 3 4 2 L.HC1.L m orient 3 1 L.HC1 oriente 3 2 HC1 .L oriente 3 2 HV1.L 1987 p t ens, entis 3 4 2 L.HC1.L m/f pacient 3 1 L.HC1 paciente 3 2 HC1.L paciente 3 2 HV1.L 1988 paen tens, entis 3 4 2 L.HC1.L m penitent 3 1 L. HC1.L penitente 4 2 L.HC1.L pen itente 4 2 L.HV1.L 1989 praes dens, entis 3 4 2 L.HC1.L m president 3 1 L.HC1 presidente 4 2 L.HC1.L presidente 4 2 L.HV1.L 1990 ref rens, -entis 3 4 2 L.HC1.L m referent 3 1 L.HC1 referente 451 4 2 L.HC1.L referente 4 2 L.HV1.L 1991 res dens, -entis 3 4 2 L.HC1.L m resident 3 1 L.HC1 residente 4 2 L.HC1.L residente 4 2 L.HV1.L 1992 trans ens, -seuntis 3 4 2 L.HC1.L m/f t ransent 3 1 L.HC1 transente 4 2 L.HC1.L transeunte 4 2 L.HV1.L L.HC1.X (1) 1993 p dermis, dis, f., del gr. 3 4 2 L.HC1.L f epidermis 42 L.HC1.HCepidermis 42L.HC1.HCepiderme 42L.HC1.L L.HC1.X (1) 1994 n formis e 3 4 2 L.HC1.L m uniforme 42L.HC1.L uniforme 4 2 L.HC1.L uniforme 4 2 L.HC1.L L.HV1.L (13) 199 5 *bell tas, tis, de bellus 3 4 2 L.HV1.L f beutat (del occitan) 2 1 HC1 beldad 2 1 HC1 beldade 3 2 L1.L 1996 bon tas, tis 3 4 2 L.HV1.L f bondat 2 1 HC1 bondad 2 1 HC1 bondade 3 2 L1.L 1997 brachi lis 3 4 2 L.HV1.L m braal 2 1 HC1 brazal 2 1 HC1 braal 2 1 HC1 1998 c p t lis e, adj. 3 4 2 L.HV1.L m cabdal 2 1 HC1 caudal 2 1 HC1 caudal 2 1 HC1 1999 c v tas tis, f. 3 4 2 L.HV1.L f ciutat 2 1 HC1 ciudad 2 1 HC1 cidade 3 2 L1.L 2000 c cl ar (cochl-), ris (c cl r -um, ii), n. 3 4 2 L.HV1.L m coller 2 1 HC1 cuchar 2 1 HC1 colher 2 1 HC1 2001 d g t lis e, adj. 3 4 2 L.HV1.L m didal 2 1 HC1 dedal 2 1 HC1 dedal 2 1 HC1 2002 gr t t do, nis, f. 3 4 2 L.HV1.L f gratitud 3 1 HC1 gratitud 3 1 HC1 gratido 3 1 HV1 2003 hosp t lis, is, m. 3 4 2 L.HV1.L m hostal 2 1 HC1 hostal 2 1 HC.HC1 hostal 2 1 HC.HC1

PAGE 452

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 2004 m l tas, tis, f. 3 4 2 L.HV1.L f maldat 21HC1 maldad 21HC1 maldade 42L1.L 2005 p r s, tis, m. 3 4 2 L.HV1.L f paret 2 1 HC1 pared 2 1 HC1 parede 3 2 L1.L 2006 pect r le, is, n. 3 4 2 L.HV1.L m pitral 2 1 HC1 petral 2 1 HC1 peitoral 3 1 L.HC1 2007 s n or, ris, adj. 3 4 2 L.HV1.L m senyor 2 1 HC1 seor 2 1 HC1 senhor 2 1 HC1 L.HV1.L (7) 2008 *arc o, nis 3 4 2 L.HV1.L m ar 2 1 HV1 arzn 2 1 HC1 aro 2 1 HV1 2009 cant o nis, f. 3 4 2 L.HV1.L f can 2 1 HV1 cancin 2 1 HC1 cano 2 1 HV1 2010 l. vg. coleo, -onis de cull us, -i 3 4 2 L.HV1.L m coll 2 1 HV1 cojn 2 1 HC1 colho 2 1 HV1 2011 pr hens o (prens o), nis, f. 3 4 2 L.HV1.L f pres 2 1 HV1 prisin 2 1 HC1 priso 2 1 HV1 2012 r t o, nis, f. 3 4 2 L.HV1.L f ra 2 1 HV1 razn 2 1 HC1 razo 2 1 HV1 2013 s t o, nis, f. 3 4 2 L.HV1.L f sa 2 1 HV1 sazn 2 1 HC1 sazo 2 1 HV1 2014 lect o, nis, f. 3 4 2 L.HV1.L f lli 2 1 HV1 leccin 2 1 HC1 lio 2 1 HV1 L.HV1.L (1) 2015 sphaer des, is, adj. 3 4 2 L.HV1.L m esferoide 4 2 HV 1.L esferoide 4 2 HV1.L esferide 4 2 HV1.L L.HV1.X (72) 2016 acrit do, -inis 3 4 2 L.HV1.L f acritud 3 1 L. HC1 acritud 3 1 L.HC1 acritude 4 2 L.L1.L 2017 *actit do 3 4 2 L.HV1.L f actitud 3 1 L.HC1 actitud 452 3 1 L.HC1 atitude 4 2 L.L1.L 2018 alt t do d nis 3 4 2 L.HV1.L f altitud 3 1 L.HC1 altitud 3 1 L.HC1 altitude 4 2 L.L1.L 2019 ampl t do, nis, f. 3 4 2 L.HV1.L f amplitud 3 1 L. HC1 amplitud 3 1 L.HC1 amplitude 4 2 L.L1.L 2020 an mal, lis 3 4 2 L.HV1.L m animal 3 1 L.HC1 animal 3 1 L.HC1 animal 3 1 L.HC1 2021 anul ris 3 4 2 L.HV1.L m anular 3 1 L.HC1 anular 3 1 L.HC1 anular 3 1 L.HC1 2022 aptit do 3 4 2 L.HV1.L f aptitud 3 1 L.HC1 aptitud 3 1 L.HC1 aptitude 4 2 L.L1.L 2023 bib tor, ris 3 4 2 L.HV1.L m bevedor 3 1 L.HC1 bebedor 3 1 L.HC1 bebedor 3 1 L.HC1 2024 br v tas tis, f. 3 4 2 L.HV1.L f brevetat 3 1 L.HC1 brevedad 3 1 L.HC1 brevidade 4 2 L.L1.L 2025 card n lis e 3 4 2 L.HV1.L m cardenal 3 1 L. HC1 cardenal 3 1 HC.L.HC1 cardeal 3 1 HC.L.HC1 2026 c r tas tis, f. 3 4 2 L.HV1.L f caritat 3 1 L. HC1 caridad 3 1 L.HC1 caridade 4 2 L.L1.L 2027 cast tas tis, f. 3 4 2 L.HV1.L f castedat 3 1 L.HC1 castidad 3 1 L.HC1 castidade 4 2 L.L1.L 2028 cav tas, tis 3 4 2 L.HV1.L f cavitat 3 1 L.HC1 cavidad 3 1 L.HC1 cavidade 4 2 L.L1.L 2029 c r lis, e 3 4 2 L.HV1.L m cereal 3 1 L.HC1 cereal 3 1 L.HC1 cereal 3 1 L.HC1 2030 circ l ris e 3 4 2 L.HV1.L f circular 3 1 L. HC1 circular 3 1 L.HC circular 3 1 L.HC 2031 cl r tas tis, f. 3 4 2 L.HV1.L f claredat 3 1 L. HC1 claridad 3 1 L.HC1 claridade 4 2 L.L1.L 2032 c r lis e 3 4 2 L.HV1.L m curial 3 1 L.HC1 curial 3 1 HC1 curial 3 1 HC1 2033 d tas tis, f. 3 4 2 L.HV1.L f detat 3 1 L. HC1 deidad 3 1 L.L.HC1 deidade 4 2 L.L1.L 2034 d g t lis e, adj 3 4 2 L.HV1.L m digital 3 1 L.HC1 digital 3 1 L.HC1 digital 3 1 L.HC1

PAGE 453

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 2035 aequ tas tis, f. 3 4 2 L.HV1.L f equitat 31L.HC1 equidad 31L.HC1 equidade 42L.L1.L 2036 fun r lis, e 3 4 2 L.HV1.L m funeral 32L.HC1 funeral 32L.HC1 funeral 32L.HC1 2037 g n r lis, e, adj. 3 4 2 L.HV1.L m general 3 1 L.HC1 general 3 1 L.L1.L geral 3 1 L.L1.L 2038 germ n lis, e, adj. 3 4 2 L.HV1.L m germinal 3 1 L.HC1 germinal 3 1 L.HC germinal 3 1 L.HC 2039 gr v tas, tis, f. 3 4 2 L.HV1.L f gravetat 3 1 L. HC1 gravedad 3 1 L.HC1 gravidade 4 2 L.L1.L 2040 hosp t lis, is, m. 3 4 2 L.HV1.L m hospital 3 1 L.HC1 hospital 3 1 L.HC1 hospital 3 1 L.HC1 2041 inf d lis, e, adj. 3 4 2 L.HV1. L m/f infidel 3 1 L.HC1 infiel 2 1 HC1 infiel 2 1 HC1 204 2 j v n lis (j v n l), e, adj. 3 4 2 L.HV1.L m/f juvenil 3 1 L.HC1 juvenil 3 1 L.HC1 juvenil 3 1 L.HC1 2043 lass t do, nis, f. 3 4 2 L.HV1.L f lassitud 3 1 L.HC1 lasitud 3 1 L.HC1 lassitude 4 2 L.L1.L 2044 l t r lis, e, adj. 3 4 2 L.HV1.L m lateral 3 1 L.HC1 lateral 3 1 L.HC1 lateral 3 1 L.HC1 2045 l t t do, nis, f. 3 4 2 L.HV1.L f latitud 3 1 L. HC1 latitud 3 1 L.HC1 latitude 4 2 L.L1.L 2046 l n tas, tis, f. 3 4 2 L.HV1.L f lenitat 3 1 L. HC1 lenidad 3 1 L.HC1 lenidade 4 2 L.L1.L 2047 l t r lis, e 3 4 2 L.HV1.L m litoral 3 1 L.HC1 litoral 3 1 L.HC1 litoral 3 1 L.HC1 2048 l c tor (l qu tor), ris, m. 3 4 2 L.HV1.L m locutor 3 1 L. HC1 locutor 3 1 L.HC1 locutor 3 1 L.HC1 2049 long t do, nis, f. 3 4 2 L.HV1.L f longitud 3 1 L.HC1 longitud 3 1 L.HC1 longitude 4 2 L.L1.L 2050 magn t do, nis, f. 3 4 2 L.HV1.L f magnitud 3 1 L. HC1 magnitud 3 1 L.HC1 magnitude 4 2 L.L1.L 2051 m n le, is, n. 3 4 2 L.HV1.L m manual 3 1 L.HC1 manual 2 1 HC1 manual 2 1 HC1 2052 m lit ris e 3 4 2 L.HV1.L m militar 3 1 L. HC1 militar 453 3 1 L.HC1 militar 3 1 L.HC1 2053 m n tor, ris, m. 3 4 2 L.HV1.L m monitor 3 1 L.HC1 monitor 3 1 L.HC1 monitor 3 1 L.HC1 2054 mult t do, nis, f. 3 4 2 L.HV1.L f multitud 3 1 L.HC1 multitud 3 1 L.HC1 multitude 4 4 L.L1.L 2055 n t r lis, e 3 4 2 L.HV1.L m natural 3 1 L.HC1 natural 3 1 L.HC1 natural 3 1 L.HC1 2056 n v tas tis, f. 3 4 2 L.HV1.L f novetat 3 1 L.HC1 novedad 3 1 L.HC1 novidade 4 2 L.L1.L 2057 ord n lis, e, adj. 3 4 2 L.HV1.L m ordinal 3 1 L.HC1 ordinal 3 1 L.HC1 ordinal 3 1 L.HC1 2058 p r tas, tis, f. 3 4 2 L.HV1.L f paritat 3 1 L. HC1 paridad 3 1 L.HC1 paridade 4 2 L.L1.L 2059 parv tas, tis, f. 3 4 2 L.HV1.L f parvitat 3 1 L.HC1 parvedad 3 1 L.HC1 parvidade 4 2 L.L1.L 2060 pect r lis, e, adj. 3 4 2 L.HV1.L m pectoral 3 1 L.HC1 pectoral 3 1 L.HC1 pectoral 3 1 L.HC1 2061 pisc tor, ris, m. 3 4 2 L.HV1.L m pescador 3 1 L. HC1 pescador 3 1 L.HC1 pescador 3 1 L.HC1 2062 p tas, tis, f. 3 4 2 L.HV1.L f pietat 3 1 L. HC1 piedad 2 1 L.HC1 piedade 3 2 L.L1.L 2063 pl n t do, nis, f. 3 4 2 L.HV1.L f plenitud 3 1 L.HC1 plenitud 3 1 L.HC1 plenitude 4 2 L.L1.L 2064 princ p lis, e 3 4 2 L.HV1.L m principal 3 1 L.HC1 principal 3 1 L.HC1 principal 3 1 L.HC1 2065 pr b tas, tis, f. 3 4 2 L.HV1.L f probitat 3 1 L.HC1 probidad 3 1 L.HC1 probidade 4 2 L.L1.L 2066 pulchr t do (pulcr-), nis, f. 3 4 2 L.HV1.L f pulcritud 3 1 L.HC1 pulcritud 3 1 L.HC1 pulcritude 4 2 L.L1.L 2067 p r tas, tis, f. 3 4 2 L.HV1.L f puritat 3 1 L. HC1 puridad 3 1 L.HC1 puridade 4 2 L.L1.L 2068 qual tas, tis 3 4 2 L.HV1.L f qualitat 3 1 L.HC1 calidad 3 1 L.HC1 qualidade 4 2 L.L1.L 2069 qu l tas tis, f. 3 4 2 L.HV1.L f qualitat 3 1 L.HC1 cualidad 3 1 L.HC1 qualidade 4 2 L.L1.L

PAGE 454

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 2070 quant tas tis, f. 3 4 2 L.HV1.L f quantitat 31 L.HC1 cantidad 31L.HC1 quantidade 42L.L1.L 2071 rect t do, nis, f. 3 4 2 L.HV1.L f rectitud 31L. HC1 rectitud 31HC1 rectitude 42L.L1.L 2072 r t lis, e 3 4 2 L.HV1.L m ritual 3 1 L.HC1 ritual 2 1 HC1 ritual 2 1 HC1 2073 s n tas, tis, f. 3 4 2 L.HV1.L f sanitat 3 1 L. HC1 sanidad 3 1 L.HC1 sanidade 4 2 L.L1.L 2074 sanct tas, tis, f. 3 4 2 L.HV1.L f santedat 3 1 L. HC1 santidad 3 1 L.HC1 santidade 4 2 L.L1.L 2075 serv t s, tis, f. 3 4 2 L.HV1.L f servitud 3 1 L.HC1 servitud 3 1 L.HC1 servitude 4 2 L.L1.L 2076 sing l ris, e 3 4 2 L.HV1.L m singular 3 1 L.HC1 singular 3 1 L.HC1 singular 3 1 L.HC1 207 7 s l tas, tis, f. 3 4 2 L.HV1.L f soledat 3 1 L.HC1 soledad 3 1 L.HC1 saudade 3 2 L1.L 2078 summ tas tis, f. 3 4 2 L.HV1.L f summitat 3 1 L.HC1 sumidad 3 1 L.HC1 sumidade 4 2 L.L1.L 2079 temp r lis e, adj. 3 4 2 L.HV1.L m temporal 3 1 L.HC1 temporal 3 1 L.HC1 temporal 3 1 L.HC1 2080 term n lis e, adj. 3 4 2 L.HV1.L m terminal 3 1 L.HC1 terminal 3 1 L.HC1 terminal 3 1 L.HC1 2081 tr d tor ris, m. 3 4 2 L.HV1.L m trador 3 1 L.HC1 traidor 2 1 HC1 traidor 2 1 HC1 2082 trin tas, tis 3 4 2 L.HV1.L f trinitat 3 1 L.HC1 trinidad 3 1 L.HC1 trindade 4 2 L.L1.L 2083 un tas, tis 3 4 2 L.HV1.L f unitat 3 1 L.HC1 unidad 3 1 L.HC1 unidade 4 2 L.L1.L 2084 van tas, tis 3 4 2 L.HV1.L f vanitat 3 1 L.HC1 vanidad 3 1 L.HC1 vanidade 4 2 L.L1.L 2085 ver tas, tis 3 4 2 L.HV1.L f veritat 3 1 L.HC1 verdad 2 1 HC1.HC verdade 3 2 L1.L 2086 vertic lis 3 4 2 L.HV1.L f vertical 3 1 L.HC1 vertical 3 1 L.HC1 vertical 3 1 L.HC1 2087 visu lis 3 4 2 L.HV1.L m visual 3 1 L.HC1 visual 2 1 HC1 visual 2 1 HC1 L.HV1.X (49) 2088 act o, nis 3 4 2 L.HV1.L f acci 3 1 L.HV1 accin 454 2 1 HC1 aco 2 1 HV1 2089 caut o, nis, f. 3 4 2 L.HV1.L f cauci 3 1 L.HV1 caucin 2 1 HC1 cauo 2 1 HV1 2090 cessio nis, f. 3 4 2 L.HV1.L f cessi 3 1 L.HV1 cesin 2 1 HC1 cesso 2 1 HV1 2091 coct o, nis 3 4 2 L.HV1.L f cocci 3 1 L.HV1 coccin 2 1 HC1 coco 2 1 HV1 2092 c l phon, nis, m. 3 4 2 L.HV1.L m colof 3 1 L.HV1 colofn 3 1 L.HC1 colofo 3 1 L.HV1 2093 dict o nis, f. 3 4 2 L.HV1.L f dicci 3 1 L.HV1 diccin 2 1 HC.HC1 dico 2 1 HV1 2094 scorp o, nis, m. 3 4 2 L.HV1.L m escorpi 4 1 L.HV1 escorpin 3 1 HC1 escorpio 3 1 HV1 2095 fact o, nis, f. 3 4 2 L.HV1.L f facci 3 1 L.HV1 faccin 2 1 HC1 faco 2 1 HV1 2096 fict o, nis, f. 3 4 2 L.HV1.L f ficci 3 2 L.HV1 ficcin 2 1 HC1 fico 2 1 HV1 2097 fiss o, nis, f. 3 4 2 L.HV1.L f fissi 3 1 L.HV1 fisin 2 1 HC1 fisso 2 1 HV1 2098 flex o, nis, f. 3 4 2 L.HV1.L f flexi 3 1 L.HV1 flexin 2 1 HC1 flexo 2 1 HV1 2099 fract o, nis, f. 3 4 2 L.HV1.L f fracci 3 2 L. HV1 fraccin 2 1 HC1 fraco 2 1 HV1 2100 frict o, nis, f. 3 4 2 L.HV1.L f fricci 3 1 L.HV1 friccin 2 1 HC1 frico 2 1 HV1 2101 funct o, nis, f. 3 4 2 L.HV1.L f funci 3 1 L.HV1 funcin 2 1 HC1 funo 2 1 HV1 2102 f s o, nis, f. 3 4 2 L.HV1.L f fusi 3 1 L.HV1 fusin 2 1 HC1 fuso 2 1 HV1 2103 gest o, nis, f. 3 4 2 L.HV1.L f gesti 3 1 L.HV1 gestin 2 1 HC1 gesto 2 1 HV1 2104 histrio, nis, m. 3 4 2 L.HV1.L m histri 3 1 L.HV 1 histrin 2 1 HC1 histrio 3 1 L.HV1

PAGE 455

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 2105 l g o, nis, f. 3 4 2 L.HV1.L f legi 31L.HV1 legin 21HC1 legio 21HV1 2106 laes o, nis, f. 3 4 2 L.HV1.L f lesi 31L.HV1 lesin 21HC1 leso 21HV1 2107 l t o, nis, f. 3 4 2 L.HV1.L f loci 31L. HV1 locin 21L.HC1 loo 21HV1 2108 l d o, nis, m. 3 4 2 L.HV1.L m ludi 31 L.HV1 ludin 21HC1 ludio 21HV1 2109 mans o, nis, f. 3 4 2 L.HV1.L f mansi 31 L.HV1 mansin 21HC1 manso 21HV1 2110 ment o, nis, f. 3 4 2 L.HV1.L f menci 31 L.HV1 mencin 21HC1 meno 22HV1 2111 minct o, nis, f. 3 4 2 L.HV1.L f micci 31L.HV1 miccin 21HC1 mico 21L.HV1 211 2 miss o, nis, f. 3 4 2 L.HV1.L f missi 3 1 L.HV1 misin 2 1 HC1 misso 2 1 HV1 2113 m t o, nis, f. 3 4 2 L.HV1.L f moci 3 1 L.HV1 mocin 2 1 HC1 moo 2 1 HV1 2114 n t o, nis, f. 3 4 2 L.HV1.L f naci 3 1 L.HV1 nacin 2 2 L.HC1 nao 2 1 HV1 2115 n t o nis, f. 3 4 2 L.HV1.L f noci 3 1 L.HV1 nocin 2 1 HC1 noo 2 1 HV1 2116 opt o, nis, f. 3 4 2 L.HV1.L f opci 3 1 L.HV1 opcin 2 1 HC.HC1 opo 2 1 HV1 2117 pass o, nis, f. 3 4 2 L.HV1.L f passi 3 1 L.HV1 pasin 2 1 HC1 paixo 2 1 HV1 2118 pens o, nis, f. 3 4 2 L.HV1.L f pensi 3 1 L.HV1 pensin 2 1 HC1 penso 2 1 HV1 2119 p t o, nis, f. 3 4 2 L.HV1.L f poci 3 1 L.HV1 pocin 2 1 HC1 poo 2 1 HV1 2120 port o, nis, f. 3 4 2 L.HV1.L f porci 3 1 L.HV1 porcin 2 1 HC1 poro 2 1 HC1 2121 press o, nis, f. 3 4 2 L.HV1.L f pressi 3 1 L.HV1 presin 2 1 HC1 presso 2 1 HV1 2122 punct o, nis, f. 3 4 2 L.HV1.L f punci 3 1 L.HV1 puncin 2 1 HC1 puno 3 1 HV1 2123 quaest o nis, f. 3 4 2 L.HV1.L f qesti 3 1 L.HV1 cuestin 455 2 1 HC1 questo 2 1 HV1 2124 rat o, nis 3 4 2 L.HV1.L f raci 3 1 L.HV1 racin 2 1 HC1 rao 2 1 HV1 2125 r g o, nis, f. 3 4 2 L.HV1.L f regi 3 1 L.HV1 regin 2 1 HC1 regio 3 1 L.HV1 2126 sanct o, nis, f. 3 4 2 L.HV1.L f sanci 3 1 L.HV1 sancin 2 2 HC1 sano 2 1 HV1 2127 sect o, nis, f. 3 4 2 L.HV1.L f secci 3 2 L.HV1 seccin 2 1 HC1 seco 2 1 HV1 2128 sess o, nis, f. 3 4 2 L.HV1.L f sessi 3 1 L.HV1 sesin 2 1 HC1 sesso 2 1 HV1 2129 t l o nis, f. 3 4 2 L.HV1.L m tali 3 1 L.HV1 talin 2 1 HC1 talio 3 1 HV1 2130 t m rix cis, f., 3 4 2 L.HV1.L m tamariu 3 1 L.HV1 tamariz 3 1 L.HC1 tamariz 3 1 L.HC1 2131 tens o nis, f. 3 4 2 L.HV1.L f tensi 3 1 L.HV1 tensin 2 1 HC1 tenso 2 1 HV1 2132 tors o nis, f. 3 4 2 L.HV1.L f torsi 3 1 L.HV1 torsin 2 1 HC1 toro 2 1 HV1 2133 tract o, nis 3 4 2 L.HV1.L f tracci 3 1 L.HV 1 traccin 2 1 HC.HC1 traco 2 1 HC.HV1 2134 unct o, nis 3 4 2 L.HV1.L f unci 3 1 L. HV1 uncin 2 1 HC1 uno 2 1 HV1 2135 n o nis, f. & m. 3 4 2 L.HV1.L f uni 3 1 L.HV1 unin 2 1 HC1 unio 2 1 HV1 2136 v s o nis, f. 3 4 2 L.HV1.L f visi 3 1 L.HV1 visin 2 1 HC1 viso 2 1 HV1 L.HV1.X (1) 2137 emphyteusis, os, f., = 3 4 2 L.HV1.L f emfiteusi 4 2 L.HV1.L enfiteusis 4 2 L. HV1.HC enfiteus e 4 2 L.HV1.L

PAGE 456

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp L.HV1.X (Gk 1st decl)(1) 2138 tymp n tes, ae, m., = 3 4 2 L.HV1.L f timpanitis 42L.L1.HC ti mpanitis 42L.L1.HC timpanite 42L.L1.L L.HV1.X (1) 2139 m d cris, e 3 4 2 L.HV1.L m mediocre 43L. L1.L mediocre 32L1.L medocre 43L1.L.L HC1.L.X (1) 2140 c p t lis e 3 4 2 L.HV1.L m/fcapital 31 L1.HC capital 31L1.HC capital 31L1.HC HC1.L.X (3) 2141 interpr s, tis, m.& f. 3 4 3 HC1.L.L m/fintrpret 32HC1. HC intrprete 43HC1.L.L intrprete 43HC1.L.L 2142 s tell s, tis, m.& f. 3 4 3 HC1.L.L m satllit 32HC1.HC satlite 43L1. L.L satlite 43L1.L.L 2143 pr consul, is, m. 3 4 3 HC1.L.L m procnsol 32 HC1.HC procnsul 3 2 HC1.L procnsul 3 2 HC1.L HC1.L.X (3) 2144 penth sis, is, f., del gr. 3 4 3 HC1.L.L f epntesi 4 3 HC1.L.L epntesis 4 3 HC1.L.HC epntese 4 3 HC1.L.L HC1.L.X (2) 2145 pr boscis, dis, f., = 3 4 3 HC1.L.L f probscide 4 3 L1.L.L probscide 4 3 HC1.L.L probscide 4 3 L1.L.L HV1.L.X (1) 2146 r b go (r b-), nis, f. 3 4 3 HV1.L.L m rov 2 1 HV1 robn 2 1 HC1 rubim 2 1 HV1 HV1.L.X (4) 2147 imp bes, ris, adj. 3 4 3 HV1.L.L m impber 3 2 L1.HC impber 456 3 2 L1. HC impbere 4 3 L1.L.L 2148 n t lis, e, adj. 3 4 3 HV1.L.L m intil 3 2 L1.HC intil 3 2 L1.HC intil 3 2 L1.HC 2149 r go, nis, f. 3 4 3 HV1.L.L m origen 3 2 L1. HC origen 3 2 L1.HC1 origem 3 2 L1.HV1 2150 imb cillis (inb-), e, adj. 3 4 3 HV.L.L m/f imbc il 3 2 L1.HC imbcil 3 2 L1.HC imbecil 3 2 L1.HC HV1.L.X (4) 2151 m go, nis, f. 3 4 3 HV1.L.L f imatge 3 2 L1. L imagen 3 2 L1.HC imagem 3 2 L1.HV 2152 imm b lis, e, adj. 3 4 3 HV1.L.L m immoble 3 2 L1.L inmueble 3 2 L1.L imvel 3 2 L1.HC 2153 n t b lis e, adj. 3 4 3 HV1.L.L m notable 3 2 L1.L notable 3 2 L1.L notvel 3 2 L1.HC 2154 praef t o, nis, f. 3 4 3 HV1.L.L m prefaci 3 2 L1. L prefacio 3 2 L1.L prefcio 3 2 L1.L HV1.L.X (3) 2155 d c sis, is, f. = 3 4 3 HV1.L.L f dicesi 3 3 L1.L.L di cesis 3 3 L1.L.HC diocese 3 2 L.L1.L 215 6 d aer sis, is, f., = 3 4 3 HV1.L.L m diresi 4 3 L1.L.L diresis 3 3 L1.L.HC direse 4 3 L1.L.L 2157 synaer sis
PAGE 457

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp L1.L.X (1) 2159 ar es, tis 3 4 3 L1.L.L m ariet 3 1 L.HC1 ariete 3 2 L1.L arete 4 3 L1.L.L L1.L.X (3) 2160 art fex, f cis, m. 3 4 3 L1.L.L m artfex 3 2 L1. HC artfice 4 3 L1.L.L artfice 4 3 L1.L.L 2161 part ceps, pis, m. 3 4 3 L1.L.L m partcip 3 2 L1.HC partcipe 4 3 L1.L. L partcipe 4 3 L1.L.L 2162 pont fex (pont -), f cis, m. 3 4 3 L1.L.L m pontfex 3 2 L1. HC pontfice 4 3 L1.L.L pontfice 4 3 L1.L.L L1.L.X (4) 2163 cinn b ris is, f. (cinn b ri is, n. en mss.), del gr. 3 4 3 L1.L.L m cinabri 3 2 L1.L ci nabrio 3 2 L1.L cinbrio 3 2 L1.L 2164 flex b lis, e, adj. 3 4 3 L1.L.L m flexible 3 2 L1.L flexible 3 2 L1.L flexvel 3 2 L1.HC 2165 *fusib lis 3 4 3 L1.L.L m fusible 3 2 L1. L fusible 3 2 L1.L fusvel 3 2 L1.L 2166 zing b ri or zimp b ri, indecl. n., = 3 4 3 L1.L.L m gingebre 3 2 L1.L je ngibre 3 2 L1.L gengibre 3 2 L1.L L1.L.X (15) 2167 N s, dis, f., = 3 4 3 L1.L.L f niada 3 3 L1.L. L nyade 3 3 L1.L.L niada 3 3 L1.L.L 2168 l p s, dis, adj. 3 4 3 L1.L.L m/ f alpede 4 3 L1.L.L alpede 4 3 L1.L.L alpede 4 3 L1.L.L 2169 ant phr sis, is, f, del gr. 3 4 3 L1.L.L f antfrasi 4 3 L1.L.L antfrasis 4 3 L1. L.HC antfrase 4 3 L1.L.L 2170 ant th sis is, f., del gr. 3 4 3 L1.L.L f anttesi 4 3 L1.L.L anttesis 457 4 3 L1.L.HC anttese 4 3 L1.L.L 2171 p d sis is, f. y este del gr. 3 4 3 L1.L.L f apdosi 4 3 L1.L.L ap dosis 4 3 L1.L.HC apdose 4 3 L1.L.L 2172 c n t es, em, (other cases not in use), f. 3 4 3 L1.L.L f cancie 4 3 L1.L.L canicie 3 2 L1.L cancie 4 3 L1.L 2173 canth ris, dis, f., y este del gr. 3 4 3 L1.L.L f cantrida 4 3 L1.L.L cantrida 4 3 L1. L.L cantrida 4 3 L1.L.L 2174 cong ner, ris 3 4 3 L1.L.L m/f congn ere 4 3 L1.L.L congnere 4 3 L1.L.L congnere 4 3 L1.L.L 2175 m t th sis, is, f., = 3 4 3 L1.L.L f mettesi 4 3 L1.L.L mettesis 4 3 L1.L.HC mettese 4 3 L1.L.L 2176 m tr p lis, is, f., = 3 4 3 L1.L.L f metrpoli 4 3 L1.L.L me trpoli 4 3 L1.L.L metrpole 4 3 L1.L.L 2177 palm pes, p dis, adj. 3 4 3 L1.L.L m palmpede 4 3 L1. L.L palmpedo 4 3 L1.L.L palmpede 4 3 L1.L.L 2178 p r phr sis, is, f. 3 4 3 L1.L.L f parfrasi 4 3 L1.L.L parfrasis 4 3 L1.L. HC parfrase 4 3 L1.L.L 217 9 p r lysis, is, f. 3 4 3 L1.L.L f parlisi 4 3 L1.L.L parlisis 4 3 L1. L.HC parlise 4 3 L1.L.L 2180 p riphr sis, is, f. 3 4 3 L1.L.L f perfrasi 4 3 L1. L.L perfrasis 4 3 L1. L.HC perfrase 4 3 L1.L.L 2181 pyr mis, dis, f., = 3 4 3 L1.L.L f pirmide 4 3 L1.L.L pirmide 4 3 L1.L.L pirmide 4 3 L1.L.L HC.HV1.X (5) 2182 r pr hensor, ris, m. 3 5 2 HC.HV1.Lm reprensor 31HC. HC1 reprensor 31HC.HC1 repreensor 31HC.HC1 2183 agrimensor 3 5 2 HC.HV1.L m agrimensor 4 1 HC.HC1 agrimensor 4 1 HC.HC1 agrimensor 4 1 HC.HC1

PAGE 458

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 2184 benefactor, ris 3 5 2 HC.HV1.Lm benefactor 41HC.HC1 benefactor 41HC.HC1 benfeitor 31HV.HC1 2185 difficultas, tis 3 5 2 HC.HV1.Lf dificultat 41HC.HC1 dificultad 31HC.HC1 dificuldade 52HC.L1.L 2186 interventor, ris, m. 3 5 2 HC.HV1.Lm interventor 41HC.HC1 interventor 41HC.HC1 interventor 41HC.HC1 HC.HV1.X (1) 2187 m t morph sis, is, f., = 3 5 2 HC.HV1.Lf metamorfosi 52HC.L1.L meta morfosis 52HC.L1.HC metamorfose 52HC.L1.L HC.HV1.X (2) 2188 antecessor, ris, m. 3 5 2 HC.HV1.Lm antecessor 41L. HC1 antecesor 41L.HC1 antecessor 41L.HC1 2189 praed cessor, ris, m. 3 5 2 HC.HV1.Lm predecessor 41L.HC1 predecesor 41L.HC1 predecessor 41L.HC1 HV.HC1.LX (1) 2190 antec dens, -entis 3 5 2 HV.HC1.L m antecedent 4 1 L.HC1 antecedente 5 2 L.HC1.L antecedente 5 2 L.HC1.L HV.HV1.X (11) 2191 appar tor, ris 3 5 2 HV.HV1.L m aparador 4 2 L.HC1 aparador 4 1 L.HC1 aparador 4 1 L.HC1 2192 calc l tor ris, m. 3 5 2 HV.HV1.L m calculador 4 1 L. HC1 calculador 4 1 L.HC1 calculador 4 1 L.HC1 2193 c adj tor ris, m. 3 5 2 HV.HV1.L m coadjutor 4 1 L.HC1 coadjutor 4 1 L.HC1 coadjutor 4 1 L.HC1 2194 d g n lis e, adj. 3 5 2 HV.HV1.L f diagonal 4 1 L.HC1 diagonal 3 1 L.HC1 diagonal 3 1 L.HC1 2195 imper tor, ris 3 5 2 HV.HV1.L m emperador 4 1 L.HC1 emperador 4 1 L.HC1 imperador 4 1 L.HC1 2196 gl d tor, ris, m. 3 5 2 HV.HV1.L m gladiador 4 1 L. HC1 gladiador 3 1 L.HC1 gladiador 3 1 L.HC1 219 7 l gisl tor, ris, m. 3 5 2 HV.HV1.L m legislador 4 1 L.HC1 legislador 458 4 1 L.HC1 legislador 4 1 L.HC1 2198 labor tor, ris 3 5 2 HV.HV1.L m llaurador 3 1 L.HC1 labrador 3 1 L.HC1 lavrador 3 1 L.HC1 2199 n m r tor, ris, m. 3 5 2 HV.HV1.L m numerador 4 1 L. HC1 numerador 4 1 L.HC1 numerador 4 1 L.HC1 2200 r t n lis, e, adj. 3 5 2 HV.HV1.L m racional 4 1 L.HC1 racional 3 1 L.HC1 racional 3 1 L.HC1 2201 ventil tor, ris 3 5 2 HV.HV1.L m ventilador 4 1 L.HC1 ventilador 4 1 L.HC1 ventilador 4 1 L.HC1 HV.HV1.X (2) 2202 imper trix, cis 3 5 2 HV.HV1.L f emperadriu 4 1 L.HV1 emperatriz 4 1 L.HC1 imperatriz 4 1 L.HC1 2203 g n r trix, cis, f. 3 5 2 HV.HV1.L f generatriu 4 1 L.HV1 generatriz 4 1 L.HC1 generatriz 4 1 L.HC1 L.HC1.X (8) 2204 allic ens, -entis, m. 3 5 2 L.HC1.L m allicient 4 1 L.HC1 aliciente 4 2 HC1.L aliciente 4 2 HC1.L 2205 defic ens, -entis 3 5 2 L.HC1.L m deficient 3 1 L. HC1 deficiente 4 2 HC1.L deficiente 4 2 HC1.L 2206 aequ v l ns, -entis 3 5 2 L.HC1.L m equ ivalent 4 1 L.HC1 equivalente 5 2 L.HC1.L equivalente 5 2 L.HC1.L 2207 exp d ens, -entis 3 5 2 L.HC1.L m expedient 4 1 L.HC1 expediente 4 2 HC1.L expediente 4 2 HC1.L 2208 b. lat. extrav gans, antis 3 5 2 L.HC1.L m extravagant 4 1 L.HC1 extravagante 5 2 L.HC1.L extravagante 5 2 L.HC1.L 2209 impaen tens, entis, adj. 3 5 2 L.HC1.L m impenitent 4 1 L. HC1 impenitente 5 2 L.HC1.L impenitente 5 2 L.HC1.L 2210 ingred ens, -entis 3 5 2 L.HC1.L m ingredient 4 1 L.HC1 ingrediente 4 2 HC1.L ingrediente 4 3 HC1.L 2211 r c p ens, -entis 3 5 2 L.HC1.L m recipient 3 1 L. HC1 recipiente 4 2 HC1.L recipiente 4 2 HC1.L

PAGE 459

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp L.HC1.X (1) 2212 m n ensis, is, m. 3 5 2 L.HC1.L m/f amanuense 5 2 L.HC1.L amanuense 4 2 HC1.L amanuense 4 2 HC1.L L.HV1.X (2) 2213 *amic tas, tis, f. 3 5 2 L.HV1.L f amistat 3 1 HC1 amistad 3 1 HC1 amizade 4 2 L1.L 2214 h r d tas, tis, f. 3 5 2 L.HV1.L f heretat 3 1 HC1 heredad 3 1 HC1 herdade 3 2 L1.L L.HV1.X (2) 2215 infant o, nis 3 5 2 L.HV1.L m infan 3 1 HV1 infanzn 3 1 HC1 infano 3 1 HV1 2216 *acule ne(m), m. 3 5 2 L.HV1.L m agull 3 2 HV1 aguijn 3 1 HC1 agulho 3 2 HV1 L.HV1.X (78) 2217 activ tas, tis 3 5 2 L.HV1.L f activitat 4 1 L.HC1 actividad 4 1 L.HC1 actividade 5 2 L.L1.L 2218 admon tor, ris, m. 3 5 2 L.HV1.L m admonitor 4 1 L. HC1 admonitor 4 1 L.HC1 admonitor 4 1 L.HC1 2219 affin tas, tis 3 5 2 L.HV1.L f afin itat 4 1 L.HC1 afinidad 4 1 L.HC1 afinidade 5 2 L.L1.L 2220 alacr tas, tis 3 5 2 L.HV1.L f alacritat 4 1 L.HC1 alacridad 4 1 L.HC1 alacridade 5 2 L.L1.L 2221 anxi tas, tis 3 5 2 L.HV1.L f ansietat 4 1 L.HC1 ansiedad 3 1 L.HC1 ansiedade 4 2 L.L1.L 2222 ant qu tas tis, f. 3 5 2 L.HV1.L f antiguitat 4 1 L.HC1 antigedad 4 1 L.HC1 antiguidade 5 2 L.L1.L 2223 asp r tas tis, f. 3 5 2 L.HV1.L f asperitat 4 1 L. HC1 asperidad 4 1 L.HC1 asperidade 5 2 L.L1.L 2224 atroc tas, tis 3 5 2 L.HV1.L f atrocitat 4 1 L.HC1 atrocidad 4 1 L.HC1 atrocidade 5 2 L.L1.L 222 5 auctor tas, tis 3 5 2 L.HV1.L f autorit at 4 1 L.HC1 autoridad 459 4 1 L.HC1 autoridade 5 2 L.L1.L 2226 aux l ris, e 3 5 2 L.HV1.L m auxiliar 4 1 L. HC1 auxiliar 3 1 HC1 auxiliar 3 1 HC1 2227 b t t do nis, f. 3 5 2 L.HV1.L f beatitud 4 1 L.HC1 beatitud 4 1 L. HC1 beatitude 5 2 L.L1.L 2228 b nign tas tis, f. 3 5 2 L.HV1.L f benignitat 4 1 L. HC1 benignidad 4 1 L.HC1 benignidade 5 2 L.L1.L 2229 calam tas, tis 3 5 2 L.HV1.L f calamitat 4 1 L.HC1 calamidad 4 1 L.HC1 calamidade 5 2 L.L1.L 2230 callos tas, tis 3 5 2 L.HV1.L f callositat 4 1 L.HC1 callosidad 4 1 L.HC1 callosidade 5 2 L.L1.L 2231 capac tas, tis 3 5 2 L.HV1.L f capacitat 4 1 L.HC1 capacidad 4 1 L.HC1 capacidade 5 2 L.L1.L 223 2 carn l tas tis, f. 3 5 2 L.HV1.L f carnalitat 4 1 L. HC1 carnalidad 4 1 L.HC1 carnalidade 5 2 L.L1.L 2233 carnos tas, tis 3 5 2 L.HV1.L f carnositat 4 1 L.HC1 carnosidad 4 1 L.HC1 carnosidade 5 2 L.L1.L 2234 celer tas, tis 3 5 2 L.HV1.L f celeritat 4 1 L.HC1 celeridad 4 1 L.HC1 celeridade 5 2 L.L1.L 2235 c v l tas tis, f. 3 5 2 L.HV1.L f civilitat 4 1 L.HC1 civilidad 4 1 L.HC1 civilidade 5 2 L.L1.L 2236 comp s tor (conp), ris, m. 3 5 2 L.HV1.L m compositor 4 1 L.HC1 co mpositor 4 1 L.HC1 compositor 4 1 L.HC1 2237 comm n tas tis, f. 3 5 2 L.HV1.L f comunitat 4 1 L. HC1 comunidad 4 1 L.HC1 comunidade 5 2 L.L1.L 2238 concuss o nis, f. 3 5 2 L.HV1.L f concussi 4 1 L.HC1 concusin 3 1 HC1 concusso 3 1 HC1 2239 conform tas, tis, f. 3 5 2 L.HV1.L f conformitat 4 1 L.HC1 conformidad 4 1 L.HC1 conformidade 5 2 L.L1.L 2240 cont s o nis, f. 3 5 2 L.HV1.L f contusi 4 1 L.HC1 contusin 3 1 HC1 contuso 3 1 HC1 2241 convect o nis, f. 3 5 2 L.HV1.L f convecci 4 1 L. HC1 conveccin 3 1 HC1 conveco 3 1 HC1 2242 convent o nis, f. 3 5 2 L.HV1.L f convenci 4 1 L.HC1 convencin 3 1 HC1 conveno 3 1 HC1

PAGE 460

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 2243 d form tas tis, f. 3 5 2 L.HV1.L f deformitat 41L.HC1 deformidad 41L.HC1 deformidade 52L.L1.L 2244 d vers tas tis, f. 3 5 2 L.HV1.L f diversitat 41L.HC1 diversidad 41L.HC1 diversidade 52L.L1.L 2245 d v n tas tis, f. 3 5 2 L.HV1.L f divinitat 41L. HC1 divinidad 41L.HC1 divindade 42L1.L 2246 excelsit do 3 5 2 L.HV1.L f excelsitud 41L.HC1 excelsitud 41L.HC1 excelsitude 52L.L1.L 2247 ext r or, ris 3 5 2 L.HV1.L m exterior 41L.HC1 exterior 31HC1 exterior 31HC1 2248 f m l ris, e 3 5 2 L.HV1.L m familiar 41L.HC1 familiar 31HC1 familiar 31HC1 2249 f t l tas, tis, f. 3 5 2 L.HV1.L f fatalitat 41L. HC1 fatalidad 41L.HC1 fatalidade 52L.L1.L 225 0 f l c tas, tis, f. 3 5 2 L.HV1.L f felicitat 4 1 L. HC1 felicidad 4 1 L.HC1 felicidade 5 2 L.L1.L 2251 f d l tas, tis, f. 3 5 2 L.HV1.L f fidelitat 4 1 L.HC1 fidelidad 4 1 L.HC1 fidelidade 4 1 L.HC1 2252 f n l tas, tis, f. 3 5 2 L.HV1.L f finalitat 4 1 L.HC1 finalidad 4 1 L.HC1 finalidade 4 1 L.HC1 2253 h l r tas, tis, f. 3 5 2 L.HV1.L f hilaritat 4 1 L. HC1 hilaridad 4 1 L.HC1 hilaridade 5 2 L.L1.L 2254 h m n tas, tis, f. 3 5 2 L.HV1.L f humanitat 4 1 L.HC1 humanidad 4 1 L.HC1 humanidade 5 2 L.L1.L 2255 humid tas, tis 3 5 2 L.HV1.L f humiditat 4 1 L. HC1 humedad 3 1 HC1 humidade 4 2 L1.L 2256 h m l tas, tis, f. 3 5 2 L.HV1.L f humilitat 4 1 L.HC1 humildad 3 1 HC1 humildade 4 2 L1.L 2257 ident tas, tis 3 5 2 L.HV1.L f identitat 4 1 L.HC1 identidad 4 1 L.HC1 identidade 5 2 L.L1.L 2258 inf r or, ius, adj. 3 5 2 L.HV1.L m inferior 4 1 L.HC1 inferior 3 1 HC1 inferior 3 1 HC1 2259 inf n tas, tis, f. 3 5 2 L.HV1.L f infinitat 4 1 L.HC1 infinidad 4 1 L.HC1 infinidade 5 2 L.L1.L 2260 ingr t t do, nis, f. 3 5 2 L.HV1.L f ingratitud 4 1 L. HC1 ingratitud 4 1 L.HC1 ingratido 4 1 HV1 2261 int r or, us, gen. ris 3 5 2 L.HV1.L m interior 4 1 L.HC1 interior 460 3 1 L.HC1 interior 3 1 L.HC1 2262 m m r le, is, n. 3 5 2 L.HV1.L m memorial 4 1 L.HC1 memorial 3 2 HC1 memorial 3 2 HC1 2263 mend c tas, tis, f. 3 5 2 L.HV1.L f mendicitat 4 1 L.HC1 mendicidad 4 1 L.HC1 mendicidade 5 2 L.L1.L 2264 b. lat. minoritas, tatis 3 5 2 L.HV1.L f minorit at 4 1 L.HC1 minoridad 4 1 L.HC1 menoridade 5 2 L.L1.L 2265 n t v tas, tis, f. 3 5 2 L.HV1.L f nativitat 4 1 L. HC1 natividad 4 1 L.HC1 natividade 5 2 L.L1.L 2266 n cess tas, tis, f. 3 5 2 L.HV1.L f necessitat 4 1 L.HC1 necesidad 4 1 L.HC1 necessidade 5 2 L.L1.L 2267 off c lis, is, m. 3 5 2 L.HV1.L m oficial 4 1 L.HC1 oficial 3 1 HC1 oficial 3 1 HC1 2268 orb c l ris, e, adj. 3 5 2 L.HV1.L m orbicular 4 1 L.HC1 orbicular 4 1 L.HC1 orbicular 4 1 L.HC1 2269 r g n lis, e, adj. 3 5 2 L.HV1.L m original 4 1 L. HC1 original 4 1 L.HC1 original 4 1 L.HC1 2270 part c l ris, e, adj. 3 5 2 L.HV1.L m particular 4 1 L. HC1 particular 4 1 L.HC1 particular 4 1 L.HC1 2271 p tern tas, tis, f. 3 5 2 L.HV1.L f paternitat 4 1 L. HC1 paternidad 4 1 L.HC1 paternidade 5 2 L.L1.L 2272 pl r l tas, tis, f. 3 5 2 L.HV1.L f pluralitat 4 1 L. HC1 pluralidad 4 1 L.HC1 pluralidade 5 2 L.L1.L 2273 post r tas, tis, f. 3 5 2 L.HV1.L f posteritat 4 1 L.HC1 posteridad 4 1 L.HC1 posteridade 5 2 L.L1.L 2274 praecaut o, nis 3 5 2 L.HV1.L f precauci 4 1 L.HC1 precaucin 3 1 HC1 precauo 3 1 HV1 2275 pr cl v tas, tis, f. 3 5 2 L.HV1.L f proclivitat 4 1 L. HC1 proclividad 4 1 L.HC1 proclividade 5 2 L.L1.L 2276 pr g n tor, ris, m. 3 5 2 L.HV1.L m progenitor 4 1 L.HC1 progenitor 4 1 HC1 progenitor 4 1 HC1 2277 pr pr tas, tis, f. 3 5 2 L.HV1.L f propietat 4 1 L.HC1 propiedad 3 1 L.HC1 propriedade 4 2 L.L1.L 2278 prosp r tas, tis, f. 3 5 2 L.HV1.L f prosperitat 4 1 L.HC1 prosp eridad 4 1 L.HC1 prosperidade 5 2 L.L1.L

PAGE 461

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 2279 rh n c ros, tis, m., = 3 5 2 L.HV1.L m rinoceront 41L.HC1 ri noceronte 52L.HC1.L rinoceronte 52L.HC1.L 2280 s t tas, tis, f. 3 5 2 L.HV1.L f sacietat 41L. HC1 saciedad 31L.HC1 saciedade 42L.L1.L 2281 s m v c l s, e, 3 5 2 L.HV1.L f semivocal 41L.HC1 semivocal 41L.HC1 semivogal 41L.HC1 2282 s m l t do, nis, f. 3 5 2 L.HV1.L f similitud 41L.HC1 similitud 41L.HC1 similitude 52L.L1.L 2283 s c tas, tis, f. 3 5 2 L.HV1.L f societat 41L. HC1 sociedad 31L.HC1 sociedade 42L1.L 2284 subt l tas, tis, f. 3 5 2 L.HV1.L f subtilitat 41L. HC1 sutilidad 41L.HC1 subtilidade 42L.HC1 2285 s p r or us 3 5 2 L.HV1.L m superior 41L.HC1 superior 31HC1 superior 31HC1 228 6 triplic tas, tis 3 5 2 L.HV1.L f triplicitat 4 1 L.HC1 triplicidad 4 1 L.HC1 triplicidade 5 2 L.L1.L 2287 unic tas, tis 3 5 2 L.HV1.L f unicitat 4 1 L.HC1 unicidad 4 1 L.HC1 unicidade 4 2 L.L1.L 2288 urb n tas tis, f. 3 5 2 L.HV1.L f urbanitat 4 1 L.HC1 urbanidad 4 1 L.HC1 urbanidade 4 2 L.HC1 2289 vari tas, tis 3 5 2 L.HV1.L f varietat 4 1 L.HC1 variedad 3 1 L.HC1 variedade 4 2 L.L1.L 2290 v l c tas tis, f. 3 5 2 L.HV1.L f velocitat 4 1 L.HC1 velocidad 4 1 L.HC1 velocidade 4 2 L.HC1 2291 v n l tas tis, f. 3 5 2 L.HV1.L f venalitat 4 1 L. HC1 venalidad 4 1 L.HC1 venalidade 5 2 L.L1.L 2292 vent s tas tis, f. 3 5 2 L.HV1.L f v entositat 4 1 L.HC1 ventosidad 4 1 L.HC1 ventosidade 5 2 L.L1.L 2293 v ciss t do nis, f. 3 5 2 L.HV1.L f vicissitud 4 1 L.HC1 vicisitud 4 1 L.HC1 vicissitude 4 1 L.HC1 2294 virgin tas, tis 3 5 2 L.HV1.L f virginitat 4 1 L.HC1 virginidad 4 1 L.HC1 virgindade 5 2 L.L1.L L.HV1.X (4th decl) (2 ) 2295 p sc p tus, s, m. 4 5 2 L.HV1.L m episcopat 4 1 L. HC1 episcopado 461 5 2 L.L1.L episcopado 5 2 L.L1.L 2296 pont f c tus, s, m. 4 5 2 L.HV1.L m pontif icat 4 1 L.HC1 pontificado 5 2 L.L1.L pontificad o 5 2 L.L1.L L.HV1.X (193) 2297 abduct o, nis, f. 3 5 2 L.HV1.L f abducci 4 1 L.HV1 abduccin 3 1 HC1 abduo 3 1 HC1 2298 abject o, nis, f. 3 5 2 L.HV1.L f abjecci 4 1 L.HV1 abyeccin 3 1 HC1 abjeco 3 1 HV1 2299 ablat o, nis, f. 3 5 2 L.HV1.L f ablaci 4 1 L.HV1 abl acin 3 1 HC1 ablao 3 1 HV1 2300 ablut o, nis, f. 3 5 2 L.HV1.L f abluci 4 1 L.HV1 ablucin 3 1 HC1 abluo 3 1 HV1 2301 absorpt o, nis, f. 3 5 2 L.HV1.L f absorci 4 1 L.HV1 absorcin 3 1 HC1 absoro 3 1 HV1 2302 accept o, nis 3 5 2 L.HV1.L f accepci 4 1 L. HV1 acepcin 3 1 HC1 acepo 3 1 HV1 2303 access o, nis, f. 3 5 2 L.HV1.L f accessi 4 1 L.HV1 accesin 3 1 HC1 acesso 3 1 HV1 2304 addit o, nis 3 5 2 L.HV1.L f addici 4 1 L.HV1 adicin 3 1 HC1 adio 3 1 HV1 2305 adhaes o, nis, f. 3 5 2 L.HV1.L f adhesi 4 1 L.HV1 adhesin 3 1 HC1 adeso 3 1 HV1 2306 adiunct o, nis, f. 3 5 2 L.HV1.L f adjunci 4 1 L.HV1 adj uncin 3 1 HC1 adjuno 3 1 HV1 2307 admiss o, nis, f. 3 5 2 L.HV1.L f admissi 4 1 L.HV1 admisin 3 1 HC1 admisso 3 1 HV1 2308 adust o, nis 3 5 2 L.HV1.L f adusti 4 1 L.HV1 adustin 3 1 HC1 adusto 3 1 HC1 2309 affect o, nis 3 5 2 L.HV1.L f afecci 4 1 L. HV1 afeccin 3 1 HC1 afeco 3 1 HV1 2310 afflict o, nis 3 5 2 L.HV1.L f aflicci 4 1 L. HV1 afliccin 3 1 HC1 aflio 3 1 HV1 2311 agnat o, nis 3 5 2 L.HV1.L f agnaci 4 1 L.HV 1 agnacin 3 1 HC1 agnao 3 1 HV1

PAGE 462

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 2312 allus o, nis 3 5 2 L.HV1.L f allusi 41L.HV1 alusin 31HC1 aluso 31HV1 2313 alluv o, nis 3 5 2 L.HV1.L m alluvi 41 L.HV1 aluvin 31HC1 aluvio 31HV1 2314 ambit o, nis 3 5 2 L.HV1.L f ambici 41L. HV1 ambicin 31HC1 ambio 31HV1 2315 aspers o, nis 3 5 2 L.HV1.L f aspersi 41L. HV1 aspersin 31HC1 asperso 31HV1 2316 assert o, nis 3 5 2 L.HV1.L f asserci 41L.HV1 asercin 31HC1 assero 31HV1 2317 assumpt o, nis, f. 3 5 2 L.HV1.L assumpci 41 L.HV1 asuncin 31HC1 assuno 31HV1 2318 astrict o, nis 3 5 2 L.HV1.L f astricci 41L.HV 1 astriccin 31HC1 adstrio 31HV1 2319 attent o, nis 3 5 2 L.HV1.L f atenci 4 1 L. HV1 atencin 3 1 HC1 ateno 3 1 HV1 2320 attract o, nis 3 5 2 L.HV1.L f atracci 4 1 L.HV 1 atraccin 3 1 HC1 atraco 3 1 HV1 2321 attr t o nis, f. 3 5 2 L.HV1.L f atrici 4 1 L. HV1 atricin 3 1 HC1 atrio 3 1 HV1 2322 audit o, nis 3 5 2 L.HV1.L f audici 4 1 L. HV1 audicin 3 1 HC1 audio 3 1 HV1 2323 vers o nis, f. 3 5 2 L.HV1.L f aversi 4 1 L. HV1 aversin 3 1 HC1 averso 3 1 HV1 2324 ch mael on nis, m. y este del gr. 3 5 2 L.HV1.L m camale 4 1 L.HV1 camalen 4 1 L.HC1 camaleo 2325 cent r o, nis, m. 3 5 2 L.HV1.L m centuri 4 1 L.HV1 centurin 3 1 HC1 centurio 3 1 HV1 2326 c act o nis, f. 3 5 2 L.HV1.L f coacci 4 1 L.HV1 coaccin 3 1 HC1 coaco 3 1 HV1 2327 cogn t o nis, f. 3 5 2 L.HV1.L f cognici 4 1 L.HV1 cognicin 3 1 HC1 cognio 3 1 HV1 2328 coll t o (conl), nis, f. 3 5 2 L.HV1.L f collaci 4 1 L.HV1 colacin 3 1 HC1 colao 3 1 HV1 2329 collect o, nis, f. 3 5 2 L.HV1.L f collecci 4 1 L.HV1 coleccin 462 3 1 HC1 coleo 3 1 HV1 2330 coll s o (conl), nis, f. 3 5 2 L.HV1.L f collisi 4 1 L.HV1 colisin 3 1 HC1 coliso 3 1 HV1 2331 combust o nis, f. 3 5 2 L.HV1.L f combusti 4 1 L.HV1 combustin 3 1 HC1 combusto 3 1 HV1 2332 compass o nis, f. 3 5 2 L.HV1.L f compassi 4 1 L. HV1 compasin 3 1 HC1 compaixo 3 2 HV1 2333 complex o (conp), nis, f. 3 5 2 L.HV1.L f complexi 4 1 L.HV1 complexin 3 1 HC1 compleio 3 1 HV1 2334 comm n o nis, f. 3 5 2 L.HV1.L f comuni 4 1 L.HV1 comunin 3 1 HC1 comunho 3 1 HV1 2335 cond c o, nis, f. 3 5 2 L.HV1.L f condici 4 1 L. HV1 condicin 3 1 HC1 condio 3 1 HC1 2336 confect o nis, f. 3 5 2 L.HV1.L f confecci 4 1 L. HV1 confeccin 3 2 HC1 confeco 3 2 HC1 2337 conf s o nis, f. 3 5 2 L.HV1.L f confusi 4 1 L.HV1 confusin 3 1 HC1 confuso 3 1 HV1 2338 congest o, nis 3 5 2 L.HV1.L f congesti 4 1 L.HV 1 congestin 3 1 HC1 congesto 3 1 HV1 2339 conjunct o nis, f. 3 5 2 L.HV1.L f conjunci 4 1 L.HV1 conjuncin 3 1 HC1 conjuno 3 1 HV1 2340 c nex o (conn), nis, f. 3 5 2 L.HV1.L f connexi 4 1 L.HV1 conexin 3 1 HC1 conexo 3 1 HV1 2341 consumpt o, nis, f. 3 5 2 L.HV1.L f consumpci 4 1 L.HV1 consuncin 3 1 HC1 consumpo 3 1 HV1 2342 contors o, nis 3 5 2 L.HV1.L f contorsi 4 1 L.HV 1 contorsin 3 1 HC1 contoro 3 1 HV1 2343 contr t o nis, f. 3 5 2 L.HV1.L f contrici 4 1 L.HV1 contricin 3 1 HC1 contrio 3 1 HV1 2344 convict o nis, f. 3 5 2 L.HV1.L f convicci 4 1 L.HV1 conviccin 3 1 HC1 convico 3 1 HV1 2345 convuls o, nis, f. 3 5 2 L.HV1.L f convulsi 4 1 L. HV1 convulsin 3 1 HC1 convulso 3 1 HV1 2346 cr m t o nis, f. 3 5 2 L.HV1.L f cremaci 4 1 L.HV1 cremacin 3 1 HC1 cremao 3 1 HV1

PAGE 463

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 2347 d ceptio nis, f. 3 5 2 L.HV1.L f decepci 41L.HV1 decepcin 31HC1 decepo 31HV 2348 d c r o nis, m. 3 5 2 L.HV1.L m decuri 41 L.HV1 decurin 31HC1 decurio 31HV1 2349 d fect o nis, f. 3 5 2 L.HV1.L f defecci 41 L.HV1 defeccin 31HC1 defeco 31HV1 2350 d ject o nis, f. 3 5 2 L.HV1.L f dejecci 4 1 L.HV1 deyeccin 3 1 HC1 dejeco 3 1 HV1 2351 d v t o nis, f. 3 5 2 L.HV1.L f devoci 4 1 L.HV1 devocin 3 1 HC1 devoo 3 1 HV1 2352 d gress o nis, f. 3 5 2 L.HV1.L f digressi 4 1 L.HV1 digresin 3 1 HC1 digresso 3 1 HV1 2353 d l t o nis, f. 3 5 2 L.HV1.L f dilaci 4 1 L.HV1 dilacin 3 1 HC1 dilao 3 1 HV1 235 4 d lect o nis, f. 3 5 2 L.HV1.L f dilecci 4 1 L.HV1 dileccin 3 1 HC1 dileco 3 1 HV1 2355 d mens o nis, f. 3 5 2 L.HV1.L f dimensi 4 1 L.HV1 dimensin 3 1 HC1 dimenso 3 1 HV1 2356 disjunct o, nis, f. 3 5 2 L.HV1.L f disjunci 4 1 L.HV1 disyuncin 3 1 HC1 disjuno 3 1 HV1 2357 dissens o nis, f. 3 5 2 L.HV1.L f dissensi 4 1 L.HV1 disensin 3 1 HC1 dissenso 3 1 HV1 2358 edit o, nis 3 5 2 L.HV1.L f edici 4 1 L.HV1 edicin 3 1 HC1 edio 3 1 HV1 2359 eff s o nis, f. 3 5 2 L.HV1.L f efusi 4 1 L.HV1 efusin 3 1 HC1 efuso 3 1 HV1 2360 elat o, nis 3 5 2 L.HV1.L f elaci 4 1 L. HV1 elaci n 3 1 HC1 elao 3 1 HV1 2361 lect o nis, f. 3 5 2 L.HV1.L f elecci 4 1 L.HV1 eleccin 3 1 HC1 eleio 3 2 HV1 2362 emot o, nis 3 5 2 L.HV1.L f emoci 4 1 L.HV 1 emocin 3 1 HC1 emoo 3 1 HV1 2363 aequ t o nis, f. 3 5 2 L.HV1.L f equaci 4 1 L.HV1 ecuacin 3 1 HC1 equao 3 1 HV1 2364 r s o, nis, f. 3 5 2 L.HV1.L f erosi 4 1 L. HV1 erosin 3 1 HC1 eroso 3 1 HV1 2365 rupt o, nis, f. 3 5 2 L.HV1.L f erupci 4 1 L.HV1 erupcin 463 3 1 HC1 erupo 3 1 HV1 2366 exact o, nis, f. 3 5 2 L.HV1.L f exacci 4 1 L.HV1 exaccin 3 1 HC1 exaco 3 1 HV1 2367 except o, nis, f. 3 5 2 L.HV1.L f excepci 4 1 L. HV1 excepcin 3 1 HC1 excepo 3 1 HV1 2368 excurs o, nis, f. 3 5 2 L.HV1.L f excursi 4 1 L. HV1 excursin 3 1 HC1 excurso 3 2 HV1 2369 exempt o, nis, f. 3 5 2 L.HV1.L f exempci 4 1 L.HV1 exencin 3 1 HC1 iseno 3 1 HV1 2370 expl s o nis, f. 3 5 2 L.HV1.L f explosi 4 1 L.HV1 explosin 3 1 HC1 exploso 3 1 HV1 2371 express o nis, f. 3 5 2 L.HV1.L f expressi 4 1 L. HV1 expresin 3 1 HC1 expresso 3 1 HV1 2372 extors o nis, f. 3 5 2 L.HV1.L f extorsi 4 1 L. HV1 extorsin 3 1 HC1 extorso 3 1 HV1 2373 fru t o, nis, f. 3 5 2 L.HV1.L f fruci 4 1 L. HV1 fruicin 2 1 HC1 fruio 2 1 HV1 2374 gest t o, nis, f. 3 5 2 L.HV1.L f gestaci 4 1 L.HV1 gestacin 3 1 HC1 gestao 3 1 HV1 2375 gr d t o, nis, f. 3 5 2 L.HV1.L f gradaci 4 1 L.HV1 gradacin 3 1 HC gradao 3 1 HV1 2376 ill t o (inl-), nis, f. 3 5 2 L.HV1.L f illaci 4 1 L.HV1 ilacin 3 1 HC1 ilao 3 1 HV1 2377 ill s o (inl-), nis, f. 3 5 2 L.HV1.L f illusi 4 1 L.HV1 ilusin 3 1 HC1 iluso 3 1 HV1 2378 inflex o, nis, f. 3 5 2 L.HV1.L f inflexi 4 1 L.HV1 inflexin 3 1 HC1 inflexo 3 1 HV1 2379 inf s o, nis, f. 3 5 2 L.HV1.L f infusi 4 1 L.HV1 infusin 3 1 HC1 infuso 3 1 HV1 2380 immers o (inm-), nis, f. 3 5 2 L.HV1.L f inmersi 4 1 L. HV1 inmersin 3 1 HC1 imerso 3 1 HV1 2381 inspect o, nis, f. 3 5 2 L.HV1.L f inspecci 4 1 L.HV1 inspeccin 3 1 HC1 inspeco 3 1 HV1 2382 intent o, nis, f. 3 5 2 L.HV1.L f intenci 4 1 L.HV1 intencin 3 1 HC1 inteno 3 1 HV1

PAGE 464

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 2383 irr s o, nis, f. 3 5 2 L.HV1.L f irrisi 41L.HV1 irrisin 31HC1 irriso 31HV1 2384 l c t o (l qu tio), nis, f. 3 5 2 L.HV1.L f locuci 41L.HV1 locucin 31HC1 locuo 31HV1 2385 lux t o, nis, f. 3 5 2 L.HV1.L f luxaci 41L.HV1 luxacin 31HC1 luxao 31HV1 2386 m gr t o nis, f. 3 5 2 L.HV1.L f migraci 41L. HV1 migracin 31HC1 migrao 31HV1 2387 m n t o, nis, f. 3 5 2 L.HV1.L f munici 41L. HV1 municin 31HC1 munio 31HV1 2388 m t t o, nis, f. 3 5 2 L.HV1.L f mutaci 41L.HV1 mutacin 31HC1 mutao 31HV1 2389 n t t o, nis, f. 3 5 2 L.HV1.L f nataci 41L. HV1 natacin 31HC1 natao 31HV1 239 0 n t t o nis, f. 3 5 2 L.HV1.L f notaci 4 1 L.HV1 notacin 3 1 HC1 notao 3 1 HV1 2391 object o, nis, f. 3 5 2 L.HV1.L f objecci 4 1 L.HV1 objecin 3 1 HC1 objeco 3 1 HV1 2392 obl t o, nis, f. 3 5 2 L.HV1.L f oblaci 4 1 L.HV1 oblacin 3 1 HC1 oblao 3 1 HV1 2393 obsess o, nis, f. 3 5 2 L.HV1.L f obsessi 4 1 L.HV1 obsesin 3 1 HC1 obsesso 3 1 HV1 2394 obvent o, nis, f. 3 5 2 L.HV1.L f obvenci 4 1 L.HV1 obve ncin 3 1 HC1 obveno 3 1 HV1 2395 occ s o, nis, f. 3 5 2 L.HV1.L f ocasi 4 1 L.HV1 ocasin 3 1 HC1 ocasio 3 1 HV1 2396 occl s o, nis 3 5 2 L.HV1.L f oclusi 4 1 L.HV 1 oclusin 3 1 HC1 ocluso 3 1 HV1 2397 miss o, nis, f. 3 5 2 L.HV1.L f omissi 4 1 L.HV1 omisin 3 1 HC1 omisso 3 1 HV1 2398 p n o, nis, f. 3 5 2 L.HV1.L f opini 4 1 L.HV1 opinin 3 1 HV1 opinio 3 1 HV1 2399 oppress o (obp-), nis, f. 3 5 2 L.HV1.L f opressi 4 1 L.HV1 opresin 3 1 HC1 opresso 3 1 HV1 2400 r t o, nis, f. 3 5 2 L.HV1.L f oraci 4 1 L.HV1 oracin 3 1 HC1 orao 3 1 HV1 2401 v t o, nis, f. 3 5 2 L.HV1.L f ovaci 4 1 L.HV1 ovacin 464 3 1 HC1 ovao 3 1 HV1 2402 percept o, nis, f. 3 5 2 L.HV1.L f percepci 4 1 L.HV1 percepcin 3 1 HC1 percepo 3 1 HV1 2403 percuss o, nis, f. 3 5 2 L.HV1.L f percussi 4 1 L.HV1 percusin 3 1 HC1 percusso 3 1 HV1 2404 perd t o, nis, f. 3 5 2 L.HV1.L f perdici 4 1 L.HV1 perdicin 3 1 HC1 perdio 3 1 HV1 2405 perfect o, nis, f. 3 5 2 L.HV1.L f perfecci 4 1 L.HV1 perfeccin 3 1 HC1 perfeio 3 1 HV1 2406 persu s o, nis, f. 3 5 2 L.HV1.L f persuasi 4 1 L.HV1 persuasin 3 1 HC1 persuaso 3 1 HV1 2407 pervers o, nis, f. 3 5 2 L.HV1.L f perversi 4 1 L.HV1 perversin 3 1 HC1 perverso 3 1 HV1 2408 p t t o, nis, f. 3 5 2 L.HV1.L f petici 4 1 L.HV1 peticin 3 1 HC1 petio 3 1 HV1 2409 plant t o, nis, f. 3 5 2 L.HV1.L f plantaci 4 1 L.HV1 plantacin 3 1 HC1 plantao 3 1 HV1 2410 poll t o, nis, f. 3 5 2 L.HV1.L f polluci 4 1 L.HV1 polucin 3 1 HC1 poluo 3 1 HV1 2411 p s t o, nis, f. 3 5 2 L.HV1.L f posici 4 1 L.HV1 posicin 3 1 HC1 posio 3 1 HV1 2412 possess o, nis, f. 3 5 2 L.HV1.L f possessi 4 1 L.HV1 posesin 3 1 HC1 possesso 3 1 HV1 2413 prostr t o, nis, f. 3 5 2 L.HV1.L f postraci 4 1 L. HV1 postracin 3 1 HC1 prostrao 3 1 HV1 2414 praecess o, nis, f. 3 5 2 L.HV1.L f precessi 4 1 L.HV1 precesin 3 1 HC1 precesso 3 1 HV1 2415 praec s o, nis, f. 3 5 2 L.HV1.L f precisi 4 1 L.HV1 precisin 3 1 HC1 preciso 3 1 HV1 2416 praedict o, nis, f. 3 5 2 L.HV1.L f predicci 4 1 L.HV1 prediccin 3 1 HC1 predio 3 1 HV1 2417 prael t o, nis, f. 3 5 2 L.HV1.L f prelaci 4 1 L.HV1 prelacin 3 1 HC1 prelao 3 1 HV1 2418 praesumpt o, nis, f. 3 5 2 L.HV1.L f presumpci 4 1 L.HV1 presuncin 3 1 HC1 presuno 3 1 HV1

PAGE 465

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 2419 praetens o, nis, f. 3 5 2 L.HV1.L f pretensi 41 L.HV1 pretensin 31HC1 pretenso 31HV1 2420 praevent o, nis, f. 3 5 2 L.HV1.L f prevenci 41L. HV1 prevencin 31HC1 preveno 31HV1 2421 praev s o, nis, f. 3 5 2 L.HV1.L f previsi 41 L.HV1 previsin 31HC1 previso 31HV1 2422 pr cess o, nis, f. 3 5 2 L.HV1.L f processi 41 L.HV1 procesin 31HC1 procisso 31HV1 2423 pr fess o, nis, f. 3 5 2 L.HV1.L f professi 41L. HV1 profesin 31HC1 profisso 31HV1 2424 pr f s o, nis, f. 3 5 2 L.HV1.L f profusi 41L. HV1 profusin 31HC1 profuso 31HV1 2425 pr miss o, nis f. 3 5 2 L.HV1.L f promissi 41L. HV1 promisin 31HC1 promisso 31HV1 242 6 pr m t o, nis, f. 3 5 2 L.HV1.L f promoci 4 1 L. HV1 promocin 3 1 HC1 promoo 3 1 HV1 2427 pr pens o, nis, f. 3 5 2 L.HV1.L f propensi 4 1 L.HV1 propensin 3 1 HC1 propenso 3 1 HV1 2428 pr port o, nis, f. 3 5 2 L.HV1.L f proporci 4 1 L. HV1 proporcin 3 1 HC1 proporo 3 1 HV1 2429 pr script o, nis, f. 3 5 2 L.HV1.L f proscripci 4 1 L.HV 1 proscripcin 3 1 HC1 proscrio 3 1 HV1 2430 prospect o, nis, f. 3 5 2 L.HV1.L f prospecci 4 1 L. HV1 prospeccin 3 1 HC1 prospeco 3 1 HV1 2431 pr v s o, nis, f. 3 5 2 L.HV1.L f provisi 4 1 L.HV1 provisin 3 1 HC1 proviso 3 1 HV1 2432 p n t o, nis, f. 3 5 2 L.HV1.L f punici 4 1 L.HV1 punicin 3 1 HC1 punio 3 1 HV1 2433 r bell o, nis, f. 3 5 2 L.HV1.L f rebelli 4 1 L.HV1 rebelin 3 1 HC1 rebelio 3 1 HV1 2434 r cens o, nis, f. 3 5 2 L.HV1.L f recensi 4 1 L.HV1 recensin 3 1 HC1 recenso 3 1 HV1 2435 r cept o, nis, f. 3 5 2 L.HV1.L f recepci 4 1 L.HV1 recepcin 3 1 HC1 recepo 3 1 HV1 2436 r cess o, nis, f. 3 5 2 L.HV1.L f recessi 4 1 L.HV1 recesin 3 1 HC1 recesso 3 1 HV1 2437 redact o, nis, f. 3 5 2 L.HV1.L f redacci 4 1 L.HV1 redaccin 465 3 1 HC1 redaco 3 1 HV1 2438 r dempt o, nis, f. 3 5 2 L.HV1.L f redempci 4 1 L.HV1 redencin 3 1 HC1 redeno 3 1 HV1 2439 r flex o, nis, f. 3 5 2 L.HV1.L f reflexi 4 1 L.HV1 reflexin 3 1 HC1 reflexo 3 1 HV1 2440 refract o, nis, f. 3 5 2 L.HV1.L f refracci 4 1 L.HV1 refraccin 3 1 HC1 refraco 3 1 HV1 2441 r l t o, nis, f. 3 5 2 L.HV1.L f relaci 4 1 L.HV1 relacin 3 1 HC1 relao 3 1 HV1 2442 r l g o, nis, f. 3 5 2 L.HV1.L f religi 4 1 L.HV1 religin 3 1 HC1 religio 4 1 L.HV1 2443 r m t o, nis, f. 3 5 2 L.HV1.L f remoci 4 1 L.HV1 remocin 3 1 HC1 remoo 3 1 HV1 2444 r press o, n is, f. 3 5 2 L.HV1.L f repressi 4 1 L. HV1 represin 3 1 HC1 represso 3 1 HV1 2445 repuls o, nis 3 5 2 L.HV1.L f repulsi 4 1 L. HV1 repulsin 3 1 HC1 repulso 3 1 HV1 2446 r c s o, nis, m. 3 5 2 L.HV1.L f rescissi 4 1 L.HV1 rescisin 3 1 HC1 resciso 3 1 HV1 2447 r sect o, nis, f. 3 5 2 L.HV1.L f resecci 4 1 L.HV1 reseccin 3 1 HC1 resseco 3 1 HV1 2448 r strict o, nis, f. 3 5 2 L.HV1.L f restricci 4 1 L.HV1 restriccin 3 1 HC1 restrio 3 1 HV1 2449 lat. med. retorsi ne, f. 3 5 2 L.HV1.L f retorsi 4 1 L.HV1 retorsin 3 1 HC1 retorso 3 1 HV1 2450 r tract o, nis, f. 3 5 2 L.HV1.L f retracci 4 1 L.HV1 retraccin 3 1 HC1 retraco 3 1 HV1 2451 r vers o, nis, f. 3 5 2 L.HV1.L f reversi 4 1 L.HV1 reversin 3 1 HC1 reverso 3 1 HV1 2452 r v s o, nis, f. 3 5 2 L.HV1.L f revisi 4 1 L.HV1 revisin 3 1 HC1 reviso 3 1 HV1 2453 r vuls o, nis, f. 3 5 2 L.HV1.L f revulsi 4 1 L.HV1 revulsin 3 1 HC1 revulso 3 1 HV1 2454 s cess o, nis, f. 3 5 2 L.HV1.L f secessi 4 1 L.HV1 secesin 3 1 HC1 secesso 3 1 HV1

PAGE 466

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 2455 s cr t o, nis, f. 3 5 2 L.HV1.L f secreci 4 1 L. HV1 secrecin 3 1 HC1 secreo 3 1 HV1 2456 s d t o, nis, f. 3 5 2 L.HV1.L f sedici 4 1 L.HV1 sedicin 3 1 HC1 sedio 3 1 HV1 2457 s lect o, nis, f. 3 5 2 L.HV1.L f selecci 4 1 L.HV1 seleccin 3 1 HC1 seleco 3 1 HV1 2458 sens t o, nis, f. 3 5 2 L.HV1.L f sensaci 4 1 L. HV1 sensacin 3 1 HC1 sensao 3 1 HV1 2459 septentr o, nis, m. 3 5 2 L.HV1.L m s eptentri 4 1 L.HV1 septentrin 3 1 HC1 setentrio 4 1 L.HV1 2460 s l t o, nis, f. 3 5 2 L.HV1.L f soluci 4 1 L.HV1 solucin 3 1 HC1 soluo 3 1 HV1 2461 subject o nis, f. 3 5 2 L.HV1.L f subjecci 4 1 L.HV1 sujecin 3 1 HC1 sujeio 3 1 HV1 246 2 summiss o (subm), nis, f. 3 5 2 L.HV1.L f submissi 4 1 L.HV1 sumisin 3 1 HC1 submisso 3 1 HV1 2463 subvent o, nis, f. 3 5 2 L.HV1.L f subvenci 4 1 L. HV1 subvencin 3 1 HC1 subveno 3 1 HV1 2464 subvers o nis, f. 3 5 2 L.HV1.L f subversi 4 1 L. HV1 subversin 3 1 HC1 subverso 3 1 HV1 2465 success o, nis, f. 3 5 2 L.HV1.L f successi 4 1 L.HV1 sucesin 3 1 HC1 sucesso 3 1 HV1 2466 s d t o, nis, f. 3 5 2 L.HV1.L f sudaci 4 1 L.HV1 sudacin 3 1 HC1 sudao 3 1 HV1 2467 suggest o, nis, f. 3 5 2 L.HV1.L f suggesti 4 1 L.HV1 sugestin 3 1 HC1 sugesto 3 1 HV1 2468 suppress o, nis, f. 3 5 2 L.HV1.L f supressi 4 1 L. HV1 supres in 3 1 HC1 supresso 3 1 HV1 2469 suspens o nis, f. 3 5 2 L.HV1.L f suspensi 4 1 L.HV1 suspensin 3 1 HC1 suspenso 3 1 HV1 2470 tempt ti nis, f. 3 5 2 L.HV1.L f temptaci 4 1 L.HV1 tentacin 3 1 HC1 tentao 3 1 HV1 2471 tr d t o nis, f. 3 5 2 L.HV1.L f traci 4 1 L.HV1 traicin 2 1 HC1 traio 2 2 HV1 2472 transact o nis, f. 3 5 2 L.HV1.L f transacci 4 1 L.HV1 transaccin 466 3 1 HC1 transaco 3 1 HV1 2473 transcript o, nis, f. 3 5 2 L.HV1.L f transcripci 4 1 L. HV1 transcripcin 3 1 HC1 transcrio 3 1 HV1 2474 b. lat. transfix o, nis 3 5 2 L.HV1.L f transfixi 4 1 L.HV 1 transfixin 3 1 HC1 transfixo 3 1 HV1 2475 transgress o nis, f. 3 5 2 L.HV1.L f transgressi 4 1 L.HV 1 transgresin 3 1 HC1 transgresso 3 1 HV1 2476 trans t o nis, f. 3 5 2 L.HV1.L m transici 4 1 L. HV1 transicin 3 2 HC1 transio 3 1 HV1 2477 transmiss o nis, f. 3 5 2 L.HV1.L f transmissi 4 1 L. HV1 transmisin 3 1 HC1 transmisso 3 1 HV1 2478 vacat o, nis 3 5 2 L.HV1.L f vacaci 4 1 L.HV1 vacacin 3 1 HC1 vacao 3 1 HV1 2479 vocat o, nis 3 5 2 L.HV1.L f vocaci 4 1 L.HV1 vocacin 3 1 HC1 vocao 3 1 HV1 L.HV1.X (3) 2480 anadipl sis, y este del gr. 3 5 2 L.HV1.L f anadiplosi 5 2 L.L1.L anadiplosis 5 2 L. L1.HC anadiplose 5 2 L.L1.L 2481 anastom sis, y este del gr. 3 5 2 L.HV1.L f anastomosi 5 2 L.L1.L anas tomosis 5 2 L.L1.HC anastomose 5 2 L.L1.L 2482 m tempsych sis, is, f., = 3 5 2 L.HV1.L f metempsicosi 5 2 L.L1.L mete mpsicosis 5 2 L.L1.HC metempsicose 5 2 L.L1.L HV1.L.X (1) 2483 v n r b lis e 3 5 3 HV1.L.L m venerable 4 2 L1.L venerable 4 2 L1.L venervel 4 2 L1.HC L1.L.X (1) 2484 haemorrh is, dis, f. 3 5 3 L1.L.L f hemorroide 4 2 HV1.L hemorroide 4 2 HV1.L hem orrides 4 2 HV1.HC

PAGE 467

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp L1.L.X (2) 2485 aborig nes (pl.) 3 5 3 L1.L.L m aborigen 4 2 L1.HC aborigen 4 2 L1.HC aborgene 5 3 L1.L.L 2486 c tyl don, nis, f., = 3 5 3 HV1.L.L m cotildon 4 2 L1.HC co tiledn 4 1 L.HC1 cotildone 5 3 L1.L.L L1.L.X (1) ph m ris, dis, f., 3 5 3 L1.L.L f efemride 5 3 L1.L.L efemride 5 3 L1.L.L efemride 5 3 L1.L.L *cosuetumen < consuet do, nis, f. 3 6 3 HV1.L.L f costum 2 1 HC1 co stumbre 3 2 HC1.L costume 3 2 L1.L L.HV1.X (18) 2489 animal tas, tis 3 6 2 L.HV1.L f animalitat 5 1 L.HC1 animalidad 5 1 L.HC1 animalidade 6 2 L.L1.L 2490 animos tas, tis 3 6 2 L.HV1.L f animositat 5 1 L.HC1 animosidad 5 1 L.HC1 animosidade 6 2 L.L1.L 2491 consangu n tas tis, f. 3 6 2 L.HV1.L f consanguin itat 5 1 L.HC1 consanguinidad 5 1 L.HC1 consanguinidade 6 2 L.L1.L 2492 copios tas, tis 3 6 2 L.HV1.L f copiositat 5 1 L.HC1 copiosidad 4 1 L.HC1 copiosidade 6 2 L.L1.L 2493 Christ n tas tis, f. 3 6 2 L.HV1.L f cristiandat 3 1 HC1 cristiandad 3 1 HC1 cristandade 4 2 L1.L 2494 *inimic tas < inimicit a 3 6 2 L.HV1.L f enemistat 4 1 HC1 enemistad 4 1 HC1 inimizade 5 2 L1.L 2495 intellect lis, e, adj. 3 6 2 L.HV1.L m intellectual 4 1 L.HC1 intelectual 4 1 HC1 intelectual 4 1 HC1 2496 long n m tas, tis, f. 3 6 2 L.HV1.L f longan imitat 5 1 L.HC1 longanimidad 4675 1 L.HC1 longanimidade 6 2 L.L1.L 2497 mult pl c tas, t is, f. 3 6 2 L.HV1.L f multiplicitat 5 1 L.HC1 multiplicidad 5 1 L.HC1 multiplicidade 6 2 L.L1.L 2498 t s tas, tis, f. 3 6 2 L.HV1.L f ociositat 5 1 L.HC1 ociosidad 4 1 L.HC1 ociosidade 5 2 L.L1.L 2499 perpend c l ris, e, adj. 3 6 2 L.HV1.L m perpendicular 5 1 L. HC1 perpendicular 5 1 L.HC1 perpendicular 5 1 L.HC1 2500 p p l r tas, tis, f. 3 6 2 L.HV1.L f popularitat 5 1 L.HC1 popularidad 5 1 L.HC1 popularidade 6 1 L.HC1 2501 poss b l tas, tis, f. 3 6 2 L.HV1.L f possibilitat 5 1 L.HC1 posibilidad 5 1 L.HC1 possibilidade 6 2 L.L1.L 2502 pr b b l tas, tis, f. 3 6 2 L.HV1.L f prob abilitat 5 1 L.HC1 probabilidad 5 1 L.HC1 probabilidade 6 2 L.L1.L 2503 sens b l tas, tis, f. 3 6 2 L.HV1.L f sens ibilitat 5 1 L.HC1 sensibilidad 5 1 L.HC1 sensibilidade 6 2 L.L1.L 2504 superflu tas, tis, f. 3 6 2 L.HV1.L f superflutat 5 1 L.HC1 superfluidad 4 1 L.HC1 superfluidade 5 2 L1.L 250 5 univers tas, tis, f. 3 6 2 L.HV1.L f universitat 5 1 L.HC1 universidad 5 1 L.HC1 universidade 6 2 L.L1.L 2506 visual tas, tis, f. 3 6 2 L.HV1.L f visualitat 5 1 L.HC1 visualidad 4 1 L.HC1 visualidade 5 2 L.L1.L L.HV1.X (7) 2507 benedict o, nis, f. 3 6 2 L.HV1.L f benedicci 4 1 L.HV1 bendicin 3 1 HC1 bno 2 1 HV1 2508 r d t o, nis, f. 3 6 2 L.HV1.L f erudici 4 1 L.HV1 erudicin 3 1 HC1 erudio 3 1 HV1 2509 inv l t o, nis, f. 3 6 2 L.HV1.L f involuci 4 1 L.HV1 involucin 3 1 HC1 involuo 3 1 HV1 2510 p p l t o, nis, f. 3 6 2 L.HV1.L f poblaci 4 1 L.HV1 poblacin 3 1 HC1 populao 3 1 HV1 2511 pond r t o, nis, f. 3 6 2 L.HV1.L f ponderaci 4 1 L.HV1 ponderacin 3 1 HC1 ponderao 3 1 HV1

PAGE 468

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 2512 r pr hens o, nis, f. 3 6 2 L.HV1.L f reprensi 41L. HV1 reprensin 31HC1 repreenso 31HV1 2513 calefact o, nis 3 6 2 L.HV1.L f calefacci 41L.HV1 calefaccin 41HC1 calefaco 41HC1 L.HV1.X (58) 2514 abnegat o, nis, f. 3 6 2 L.HV1.L f abnegaci 51 L.HV1 abnegacin 41HC1 abnegao 41HV1 2515 adiurat o, nis, f. 3 6 2 L.HV1.L f adjuraci 51L. HV1 adjuracin 41HC1 adjurao 41HC1 2616 admonit o, nis, f. 3 6 2 L.HV1.L f admonici 51L. HV1 admonicin 41HC1 admonio 41HV1 2517 advocat o, nis, f. 3 6 2 L.HV1.L f advocaci 51 L.HV1 advocacin 41HC1 avocao 41HV1 2518 allocut o, nis, f. 3 6 2 L.HV1.L f allocuci 51L.HV1 alocucin 41 HC1 alocuo 4 1 HV1 2519 apparit o, nis 3 6 2 L.HV1.L f aparici 5 1 L. HV1 aparicin 4 1 HC1 apario 4 1 HV1 2520 apposit o, nis 3 6 2 L.HV1.L f aposici 5 1 L. HV1 aposicin 4 1 HC1 aposio 4 1 HV1 2521 bifurcat o, nis 3 6 2 L.HV1.L f bifurcaci 5 1 L. HV1 bifurcacin 4 1 HC1 bifurcao 4 1 HV1 2522 circumspect o nis, f. 3 6 2 L.HV1.L f circumspecci 5 1 L. HV1 circunspeccin 4 1 HC1 circunspeco 4 1 HV1 2523 b. ll. coalitio, -onis 3 6 2 L.HV1.L f c oalici 5 1 L.HV1 coalicin 4 1 HC1 coalizo 4 1 HC1 2524 coarctat o, nis, f. 3 6 2 L.HV1.L f coartici 5 1 L. HV1 coartacin 4 1 HC1 coarctao 4 1 HV1 2525 confl gr t o nis, f. 3 6 2 L.HV1.L f conflagraci 5 1 L.HV 1 conflagracin 4 1 HC1 conflagrao 4 1 HV1 2526 conform t o nis, f. 3 6 2 L.HV1.L f conformaci 5 1 L. HV1 conformacin 4 1 HC1 conformao 4 1 HV1 2527 constellat o, nis, f. 3 6 2 L.HV1.L f constellaci 5 1 L. HV1 constelacin 4 1 HC1 constelao 4 1 HV1 2528 d lect t o nis, f. 3 6 2 L.HV1.L f delectaci 5 1 L.HV1 delectacin 468 4 1 HC1 deleitao 4 1 HV1 2529 d n t t o nis, f. 3 6 2 L.HV1.L f denotaci 5 1 L. HV1 denotacin 4 1 HC1 denotao 4 1 HV1 2530 d p s t o nis, f. 3 6 2 L.HV1.L f deposici 5 1 L. HV1 deposicin 4 1 HC1 deposio 4 1 HV1 2531 d pr c t o nis, f. 3 6 2 L.HV1.L f deprecaci 5 1 L. HV1 deprecacin 4 1 HC1 deprecao 4 1 HV1 2532 disqu s t o nis, f. 3 6 2 L.HV1.L f disquisici 5 1 L.HV1 disquisicin 4 1 HC1 disquisio 4 1 HV1 2533 bull t o nis, f. 3 6 2 L.HV1.L f ebullici 5 1 L. HV1 ebullicin 3 1 HC1 ebulio 3 1 HV1 2534 d c t o nis, f. 3 6 2 L.HV1.L f educaci 5 1 L.HV1 educacin 4 1 HC1 educao 4 1 HV1 2535 l c t o nis, f. 3 6 2 L.HV1.L f elocuci 5 1 L.HV 1 elocucin 4 1 L.HC1 elocuo 4 1 L.HV1.L 2536 qu t t o nis, f. 3 6 2 L.HV1.L f equitaci 5 1 L. HV1 equitacin 4 1 HC1 equitao 4 1 HV1 2537 v l t o, nis, f. 3 6 2 L.HV1.L f evoluci 5 1 L. HV1 evolucin 4 1 HC1 evoluo 4 1 HV1 2538 exh l t o, nis, f. 3 6 2 L.HV1.L f exhalaci 5 1 L.HV1 exhalacin 4 1 HC1 exalao 4 1 HV1 2539 expect t o, nis, f. 3 6 2 L.HV1.L f expectaci 5 1 L. HV1 expectacin 4 1 HC1 expectao 4 1 HV1 2540 n n t o, nis, f. 3 6 2 L.HV1.L f inanici 5 1 L. HV1 inanicin 4 1 HC1 inanio 4 1 HV1 2541 ind gest o, nis, f. 3 6 2 L.HV1.L f indigesti 5 1 L.HV1 indigestin 4 1 HC1 indigesto 4 1 HV1 2542 indign t o, nis, f. 3 6 2 L.HV1.L f indignaci 5 1 L.HV 1 indignacin 4 1 HC1 indignao 4 1 HV1 2543 indurat o, nis, f. 3 6 2 L.HV1.L f induraci 5 1 L.HV1 induracin 4 1 HC1 indurao 4 1 HV1 2544 insurrect o, nis, f. 3 6 2 L.HV1.L f insurrecci 5 1 L.HV1 insurreccin 4 1 HC1 insurreio 4 1 HV1 2545 intellect o, nis, f. 3 6 2 L.HV1.L f intellecci 5 1 L.HV1 inteleccin 4 1 HC1 inteleco 4 1 HV1 2546 interdict o, nis, f. 3 6 2 L.HV1.L f interdicci 5 1 L.HV1 interdiccin 4 1 HC1 interdio 4 1 HV1

PAGE 469

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 2547 interject o, nis, f. 3 6 2 L.HV1.L f interjecci 51L. HV1 interjeccin 41HC1 interjeio 41HV1 2548 intersect o, nis, f. 3 6 2 L.HV1.L f intersecci 51L. HV1 interseccin 41HC1 interseco 41HV1 2549 iurisdict o, nis, f. 3 6 2 L.HV1.L f jurisdicci 51L.HV1 jurisdiccin 41HC1 jurisdio 41HV1 2550 m l dict o, nis, f. 3 6 2 L.HV1.L f maledicci 51L.HV1 maldicin 31HC1 maldio 31HV1 2551 m n miss o, nis, f. 3 6 2 L.HV1.L f manumissi 51 L.HV1 manumisin 41HC1 manumisso 41HV1 2552 occaec t o, nis, f. 3 6 2 L.HV1.L f obcecaci 51 L.HV1 obcecacin 41HC1 obcecao 41HV1 2553 opp s t o, nis, f. 3 6 2 L.HV1.L f oposici 51L.HV1 oposicin 41HC1 oposio 41HV1 255 4 p r r t o, nis, f. 3 6 2 L.HV1.L f peroraci 5 1 L.HV1 peroracin 4 1 HC1 perorao 4 1 HV1 2555 pers c t o, nis, f. 3 6 2 L.HV1.L f persecuci 5 1 L.HV1 persecucin 4 1 HC1 persecuo 4 1 HV1 2556 praecogn t o, nis, f. 3 6 2 L.HV1.L f precognici 5 1 L.HV 1 precognicin 4 1 HC1 precognio 4 1 HV1 2557 praep s t o, nis, f. 3 6 2 L.HV1.L f preposici 5 1 L. HV1 preposicin 4 1 HC1 preposio 4 1 HV1 2558 pr s c t o, nis, f. 3 6 2 L.HV1.L f prossecuci 5 1 L. HV1 prosecucin 4 1 HC1 prossecuo 4 1 HV1 2559 p tr factio, nis, f. 3 6 2 L.HV1.L f putrefacci 5 1 L.HV1 putrefaccin 4 1 HC1 putrefaco 4 1 HV1 2560 r d t o, nis, f. 3 6 2 L.HV1.L f radiaci 5 1 L. HV1 radiacin 3 1 HC1 radiao 3 1 HV1 2561 r percuss o, nis, f. 3 6 2 L.HV1.L f repercussi 5 1 L. HV1 repercusin 4 1 HC1 repercusso 4 1 HV1 2562 r p s t o, nis, f. 3 6 2 L.HV1.L f reposici 5 1 L.HV1 reposicin 3 1 HC1 reposio 3 1 HV1 2563 r p t t o, nis, f. 3 6 2 L.HV1.L f reputaci 5 1 L.HV1 reputacin 4 1 HC1 reputao 4 1 HV1 2564 r s l t o, nis, f. 3 6 2 L.HV1.L f resoluci 5 1 L.HV1 resolucin 4 1 HC1 resoluo 4 1 HV1 2565 r surrect o, nis, f. 3 6 2 L.HV1.L f resurrecci 5 1 L.HV1 resurreccin 469 4 1 HC1 ressurreio 4 1 HV1 2566 r v l t o, nis, f. 3 6 2 L.HV1.L f revoluci 5 1 L.HV1 revolucin 4 1 HC1 revoluo 4 1 HV1 2567 s l t t o nis, f. 3 6 2 L.HV1.L f salutaci 5 1 L.HV1 salutacin 4 1 HC1 saudao 4 1 HV1 2568 s tisfactio, nis, f. 3 6 2 L.HV1.L f satisfacci 5 1 L.HV 1 satisfaccin 4 1 HC1 satisfao 4 1 HV1 2569 superstit o, nis, f. 3 6 2 L.HV1.L f superstici 5 1 L.HV1 supersticin 4 1 HC1 superstio 4 1 HV1 2570 s p n t o nis, f. 3 6 2 L.HV1.L f supinaci 5 1 L.HV1 supinacin 4 1 HC1 supinao 4 1 HV1 2571 supp s t o, nis, f. 3 6 2 L.HV1.L f suposici 5 1 L.HV1 suposicin 4 1 HC1 suposio 4 1 HV1 L.HV1.X (6) 2572 bipartit o, nis 3 6 2 L.HV1.L f bipartici 5 2 L.HV 1 biparticin 4 1 HC1 bipartio 4 1 HV1 2573 constern t o nis, f. 3 6 2 L.HV1.L f consternaci 5 2 L.HV 1 consternacin 4 1 HC1 consternao 4 1 HV1 2574 crucifix o, nis 3 6 2 L.HV1.L f crucifixi 5 2 L.HV 1 crucifixin 4 1 HC1 crucifixo 4 1 HV1 2575 d praed t o nis, f. 3 6 2 L.HV1.L f depredaci 5 2 L.HV1 depredacin 4 1 HC1 depredao 4 1 HV1 2576 praet r t o, nis, f. 3 6 2 L.HV1.L f preterici 5 2 L.HV1 pretericin 4 1 HC1 preterio 4 1 HV1 2577 tr b l t o nis, f. 3 6 2 L.HV1.L f tribulaci 5 2 L.HV1 tribulacin 4 1 HC1 tribulao 4 1 HV1 L.HV1.X (2) 2578 alienat o, nis, f. 3 6 2 L.HV1.L f alienaci 6 1 L.HV1 ajenacin 4 1 HC1 alienao 4 1 HV1 2579 st p l t o, nis, f. 3 6 2 L.HV1.L f estipulaci 6 1 L.HV1 estipulacin 5 1 HC1 estipulao 5 1 HV1

PAGE 470

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp L1.L.X (1) 2580 l phant sis, is, f., del gr. 3 6 3 L1.L.L f elefantiasi 62L.L1.L elef antiasis 52L1.H elefantase 63L1.L.L L.HV1.X (2) 2581 f m l r tas, tis, f. 3 7 2 L.HV1.L f familiaritat 61L.HC1 familiaridad 51L.HC1 familiaridade 62L.L1.L 2582 particular tas, tis, f. 3 7 2 L.HV1.L f particularitat 61L.HC1 particularidad 61L.HC1 particularidade 72L.L1.L L.HV1.X (11) 2583 animadvers o, nis, f. 3 7 2 f animadversi 61L.HV1 animadversin 51HC1 animadverso 51HC1 2584 annuntiat o, nis, f. 3 7 2 L.HV1.L f anunciaci 61L. HV1 anunciacin 41HC1 anunciao 41HV1 2585 capitulat o, nis 3 7 2 L.HV1.L f capitulaci 61L.HV1 capitulacin 51HC1 capitulao 51HV1 2586 circuml c t o nis, f. 3 7 2 L.HV1.L f circumlocuci 6 1 L. HV1 circunlocucin 5 1 HC1 circunlocuo 5 1 HV1 2587 comm s r t o nis, f. 3 7 2 L.HV1.L f commiseraci 6 1 L. HV1 conmiseracin 5 1 HC1 comiserao 5 2 HV1 2588 interl c t o, nis, f. 3 7 2 L.HV1.L f interlocuci 6 1 L. HV1 interlocucin 5 1 HC1 interlocuo 4 1 HV1 2589 m r g r t o, nis, f. 3 7 2 L.HV1.L f morigeraci 6 1 L. HV1 morigeracin 5 1 HC1 morigerao 5 1 HV1 2590 p c f c t o, nis, f. 3 7 2 L.HV1.L f pacificaci 6 1 L.HV1 pacificacin 5 1 HC1 pacificao 5 1 HV1 2591 praedest n t o, nis, f. 3 7 2 L.HV1.L f predestinaci 6 1 L.HV 1 predestinacin 5 1 HC1 predestinao 5 1 HV1 2592 r susc t t o, nis, f. 3 7 2 L.HV1.L f ressuscitaci 6 1 L.HV1 resucitacin 5 1 HC1 ressuscitao 5 1 HV1 2593 subordinat o, nis, f. 3 7 2 L.HV1.L f subordinaci 6 1 L. HV1 subordinacin 470 5 1 HC1 subordinao 5 2 HV1 2594 transf g r t o nis, f. 3 7 2 L.HV1.L f transfiguraci 6 1 L. HV1 transfiguracin 5 1 HC1 transfigurao 5 1 HV1 HC1.X (121) 2595 aphthae, rum, f., = 1 2 2 HC1.L f afta 2 2 HC1.L afta 2 2 HC1.L afta 2 2 HC1.L 2596 alba, ae, f. 1 2 2 HC1.L f alba 2 2 HC1.L alba 2 2 HC1.L alba 2 2 HC1.L 2597 alga, ae, f. 1 2 2 HC1.L f alga 2 2 HC1.L alga 2 2 HC1.L alga 2 2 HC1.L 2598 arca, ae, f. 1 2 2 HC1.L f arca 2 2 HC1.L arca 2 2 HC1.L arca 2 2 HC1.L 2599 ansa ae, f. 1 2 2 HC1.L f ansa 2 2 HC1.L asa 2 2 L1.L asa 2 2 L1.L 260 0 asthma, y este del gr. 1 2 2 HC1.L f asma 2 2 HC1.L asma 2 2 HC1.L asma 2 2 HC1.L 2601 barba, ae 1 2 2 HC1.L f barba 2 2 HC1.L barba 2 2 HC1.L barba 2 2 HC1.L 2602 barca ae, f. 1 2 2 HC1.L f barca 2 2 HC1.L barca 2 2 HC1.L barca 2 2 HC1.L 2603 bucca, ae, f., voz de or. Celta 1 2 2 HC1.L f boca 2 2 L1.L boca 2 2 L1.L boca 2 2 L1.L 2604 bursa 1 2 2 HC1.L f bossa 2 2 L1.L bolsa 2 2 HC1.L bolsa 2 2 HC1.L 2605 burra ae, f. 1 2 2 HC1.L f borra 2 2 L1.L borra 2 2 L1.L borra 2 2 L1.L 2606 br chus, brocchus broccus, a, um, 1 2 2 HC1.L f broca 2 2 L1.L broca 2 2 L1.L broca 2 2 L1.L 2607 bulla ae, f. 1 2 2 HC1.L f butlla 2 2 HC1.L bula 2 2 L1.L bula 2 2 L1.L 2608 capsa ae, f. 1 2 2 HC1.L f caixa 2 2 HV1.L caja 2 2 L1.L caixa 2 2 L1.L

PAGE 471

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 2609 c l da (calda ), ae, f. 1 2 2 HC1.L f calda 22HC1.L calda 22HC1.L calda 22HC1.L 2610 caltha ae, f. 1 2 2 HC1.L f calta 22HC1.L calta 22HC1.L calta 22HC1.L 2611 calva ae, f. 1 2 2 HC1.L f calba 22HC1.L calva 22HC1.L calva 22HC1.L 2612 canna ae, f. 1 2 2 HC1.L f canya 22L1.L caa 22L1.L canha 22L1.L 2613 charta ae, f. 1 2 2 HC1.L f carta 22HC1.L carta 22HC1.L carta 22HC1.L 2614 cella ae, f. 1 2 2 HC1.L f cella 22HC1.L celda 22HC1.L cela 22L1.L 2615 comma tis, n., del gr. 1 2 2 HC1.L f coma 22L1.L coma 22L1.L coma 22L1.L 261 6 cuppa 1 2 2 HC1.L f copa 2 2 L1.L copa 2 2 L1.L copa 2 2 L1.L 2617 crusta ae, f. 1 2 2 HC1.L f crosta 2 2 HC1.L costra 2 2 HC1.L crosta 2 2 HC1.L 2618 crista ae, f. 1 2 2 HC1.L f cresta 2 2 HC1.L cresta 2 2 HC1.L cresta 2 2 HC1.L 2619 crypta ae, f. 1 2 2 HC1.L f cripta 2 2 HC1.L cripta 2 2 HC1.L cripta 2 2 HC1.L 2620 concha ae, f. 1 2 2 HC1.L f conca 2 2 HC1.L cuenca 2 2 HC1.L conca 2 2 HC1.L 2621 chorda ae, f. 1 2 2 HC1.L f corda 2 2 HC1.L cuerda 2 2 HC1.L corda 2 2 HC1.L 2622 costa ae, f. 1 2 2 HC1.L f costa 2 2 HC1.L cuesta 2 2 HC1.L costa 2 2 HC1.L 2623 culpa ae, f. 1 2 2 HC1.L f culpa 2 2 HC1.L culpa 2 2 HC1.L culpa 2 2 HC1.L 2624 lat. v. falla, ae, f. 1 2 2 HC1.L f falla 2 2 L1.L falla 2 2 L1.L falha 2 2 L1.L 2625 flamma, ae, f. 1 2 2 HC1.L f flama 2 2 L1.L flama 2 2 L1.L flama 2 2 L1.L 2626 forma, ae, f. 1 2 2 HC1.L f forma 2 2 HC1.L forma 2 2 HC1.L forma 2 2 HC1.L 2627 fossa, ae, f. 1 2 2 HC1.L f fossa 2 2 L1.L fosa 471 2 2 L1.L fossa 2 2 L1.L 2628 funda, ae, f. 1 2 2 HC1.L f funda 2 2 HC1.L funda 2 2 HC1.L funda 2 2 HC1.L 2629 gemma, ae, f. 1 2 2 HC1.L f gema 2 2 L1.L gema 2 2 L1.L gema 2 2 L1.L 2630 glossa, ae, f. 1 2 2 HC1.L f glossa 2 2 L1.L glosa 2 2 L1.L glossa 2 2 L1.L 2631 g mma
PAGE 472

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 2644 mamma, ae, f. 1 2 2 HC1.L f mama 22L1.L mama 22L1.L mama 22L1.L 2645 manna, ae, f., = 1 2 2 HC1.L f mann 21L.L1 man 21L.L1 man 21L.L1 2646 mappa, ae, f. 1 2 2 HC1.L m mapa 22L1.L mapa 22L1.L mapa 22L1.L 2647 b. lat. marca, y este del germ. 1 2 2 HC1.L f marca 22HC1.L marca 22HC1.L marca 22HC1.L 2648 marga, ae, f., de or. celta 1 2 2 HC1.L f marga 22HC1.L marga 22HC1.L marga 22HC1.L 2649 massa, ae, f., = 1 2 2 HC1.L f massa 22L1.L masa 22L1.L massa 22L1.L 2650 matta, ae, f. 1 2 2 HC1.L f mata 22L1.L mata 22L1.L mata 22L1.L 265 1 menta (mentha), ae, f., = 1 2 2 HC1.L f menta 2 2 HC1.L menta 2 2 HC1.L menta 2 2 HC1.L 2652 mensa, ae, f. 1 2 2 HC1.L f mesa 2 2 L1.L mesa 2 2 L1.L mesa 2 2 L1.L 2653 merda, ae, f. 1 2 2 HC1.L f merda 2 2 HC1.L mierda 2 2 HC1.L merda 2 2 HC1.L 2654 murra (alt. myrrha), ae, f., = 1 2 2 HC1.L f mirra 2 2 L1.L mirra 2 2 L1.L mirra 2 2 L1.L 2655 missa, ae, f. 1 2 2 HC1.L f missa 2 2 L1.L misa 2 2 L1.L missa 2 2 L1.L 2656 musca, ae, f. 1 2 2 HC1.L f mosca 2 2 HC1.L mosca 2 2 HC1.L mosca 2 2 HC1.L 2657 multa (mulcta), ae, f. 1 2 2 HC1.L f multa 2 2 HC1.L multa 2 2 HC1.L multa 2 2 HC1.L 2658 murta, ae, f. 1 2 2 HC1.L f murt a 2 2 HC1.L murta 2 2 HC1.L murta 2 2 HC1.L 2659 naphtha, ae, f., 1 2 2 HC1.L f nafta 2 2 HC1.L nafta 2 2 HC1.L nafta 2 2 HC1.L 2660 nassa o naxa, ae, f. 1 2 2 HC1.L f nassa 2 2 L1.L nasa 472 2 2 L1.L nassa 2 2 L1.L 2661 *natta, var. del b. lat. matta 1 2 2 HC1.L f nata 2 2 L1. L nata 2 2 L1.L nata 2 2 L1.L 2662 nympha ae, f., del gr. 1 2 2 HC1.L f nimfa 2 2 HC1.L ninfa 2 2 HC1.L ninfa 2 2 HC1.L 2663 n rma ae, f. 1 2 2 HC1.L f norma 2 2 HC1 .L norma 2 2 HC1.L norma 2 2 HC1.L 2664 olla, ae, f. 1 2 2 HC1.L f olla 2 2 L1.L olla 2 2 L1.L olha 2 2 L1.L 2665 unda, ae, f. 1 2 2 HC1.L f onda 2 2 HC1.L onda 2 2 HC1.L onda 2 2 HC1.L 2666 orca, ae, f. 1 2 2 HC1.L f orca 2 2 HC1.L orca 2 2 HC1.L orca 2 2 HC1.L 2667 palma, ae, f. 1 2 2 HC1.L f palma 2 2 HC1.L palma 2 2 HC1.L palma 2 2 HC1.L 2668 [uva] passa; passus, a, um
PAGE 473

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 2677 prensus, a, um 1 2 2 HC1.L f pre ssa 22L1.L presa 22L1.L pressa 22L1.L 2678 VL. posta< p s tus, a, um 1 2 2 HC1.L f posta 22HC1.L puesta 22HC1.L posta 22HC1.L 2679 pugna, ae, f. 1 2 2 HC1.L f pugna 22HC1.L pugna 22HC1.L pugna 22HC1.L 2680 pulpa, ae, f. 1 2 2 HC1.L f polpa 22HC1.L pulpa 22HC1.L polpa 22HC1.L 2681 puncta, ae, f. 1 2 2 HC1.L f punta 22HC1.L punta 22HC1.L ponta 22HC1.L 2682 ruptus, a, um 1 2 2 HC1.L f ro ta 22L1.L rota 22L1.L rota 22L1.L 2683 salsus, a, um 1 2 2 HC1.L f salsa 22HC1.L salsa 22HC1.L salsa 22HC1.L 268 4 sanctus, a, um 1 2 2 HC1.L f sant a 2 2 HC1.L santa 2 2 HC1.L santa 2 2 HC1.L 2685 sectus, a, um 1 2 2 HC1.L f secta 2 2 HC1.L secta 2 2 HC1.L secta 2 2 HC1.L 2686 silva, ae, f. 1 2 2 HC1.L f selva 2 2 HC1.L selva 2 2 HC1.L selva 2 2 HC1.L 2687 sextus, a, um 1 2 2 HC1.L f sexta 2 2 HC1.L sexta 2 2 HC1.L sexta 2 2 HC1.L 2688 serra, ae, f. 1 2 2 HC1.L f serra 2 2 HC1.L sierra 2 2 HC1.L serra 2 2 HC1.L 2689 serva, ae, f. 1 2 2 HC1.L f serva 2 2 HC1.L sierva 2 2 HC1.L serva 2 2 HC1.L 2690 sexta [hora]
PAGE 474

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp HC1.L (prothesis) 2712 sponsus, a, um 1 2 2 HC1.L f espsa 3 2 L1.L esposa 3 2 L1.L esposa 3 2 L1.L 2713 sporta, ae, f. 1 2 2 HC1.L f esporta 3 2 HC1.L espuerta 3 2 HC1.L esporta 3 2 HC1.L 2714 stuppa, ae, f., = 1 2 2 HC1.L f estopa 3 2 L1.L estopa 3 2 L1.L estopa 3 2 L1.L 2715 stella, ae, f. 1 2 2 HC1.L f estrella 3 2 L1.L estrella 3 2 L1.L estrela 3 2 L1.L HV1.X (16) 2716 aula ae, f. 1 2 2 HV1.L f aula 2 2 HV1.L aula 2 2 HV1.L aula 2 2 HV1.L 2717 aura ae, y este del gr. 1 2 2 HV 1.L f aura 2 2 HV1.L aura 2 2 HV1.L aura 2 2 HV1.L 2718 br ma ae, f. 1 2 2 HV1.L f broma 2 2 L1.L bruma 2 2 L1.L bruma 2 2 L1.L 2719 causa, ae, f. 1 2 2 HV1.L f causa 2 2 HV1.L causa 2 2 HV 1.L causa 2 2 HV1.L 2720 c na, ae, f. 1 2 2 HV1.L f cena 2 2 L1.L cena 2 2 L1.L ceia 2 2 HV1.L 2721 c ra ae, f. 1 2 2 HV1.L f cera 2 2 L1.L cera 2 2 L1.L cera 2 2 L1.L 2722 cl va ae, f. 1 2 2 HV1.L f clava 2 2 L1.L clava 2 2 L1.L clava 2 2 L1.L 2723 cauda (c da), ae, f. 1 2 2 HV1.L f c oda 2 2 L1.L coda 2 2 L1.L coda 2 2 L1.L 2724 c da, ae, del lat. cauda, lat. vulg. 1 2 2 HV1.L f cua 2 2 L1.L cola 2 2 L1.L cola 2 2 L1.L 2725 causa, ae, f. 1 2 2 HV1.L f cosa 2 2 L1.L cosa 2 2 L1.L coisa 2 2 HV1.L 2726 cr ta ae, f. 1 2 2 HV1.L f creta 2 2 L1.L creta 2 2 L1.L creta 2 2 L1.L 2727 qu dra ae, f. lat. tardo 1 2 2 HV1.L f quadra 2 2 L1.L cuadra 474 2 2 L1.L quadra 2 2 L1.L 2728 c ra ae, f. 1 2 2 HV1.L f cura 2 2 L1.L cura 2 2 L1.L cura 2 2 L1.L 2729 d pla ae, f. 1 2 2 HV1.L f dobla 2 2 L1.L dobla 2 2 L1.L dobra 2 2 L1.L 2730 n na ae, f. 1 2 2 HV1.L f nana 2 2 L1.L enana 3 3 L.L1.L an 2 2 L.HV1 2731 aera ae, f. 1 2 2 HV1.L f era 2 2 L1.L era 2 2 L1.L era 2 2 L1.L HV1.L (prothesis) (12) 2732 sc la, ae, f. 1 2 2 HV1.L f escala 3 2 L1.L escala 3 2 L1.L escala 3 2 L1.L 2733 squ ma, ae, f. 1 2 2 HV1.L f escama 3 2 L1.L escama 3 2 L1.L escama 3 2 L1.L 2734 scaena, ae (sc na), f. 1 2 2 HV1.L f escena 3 2 L1.L escena 3 2 L1.L cena 2 2 L1.L 2735 scr ba, ae, m. 1 2 2 HV1.L m escriba 3 2 L1.L escriba 3 2 L1.L escriba 3 2 L1.L 2736 sphaera, ae, f. 1 2 2 HV1.L f esfera 3 2 L1.L esfera 3 2 L1.L esfera 3 2 L1.L 2737 sp ca, ae, f. 1 2 2 HV1.L f espiga 3 2 L1.L espiga 3 2 L1.L espiga 3 2 L1.L 2738 sp na, ae, f. 1 2 2 HV1.L f espina 3 2 L1.L espina 3 2 L1.L espinha 3 2 L1.L 2739 sp ra, ae, f. 1 2 2 HV1.L f espira 3 2 L1.L espira 3 2 L1.L espira 3 2 L1.L 3740 sp ma, ae, f. 1 2 2 HV1.L f espuma 3 2 L1.L espuma 3 2 L1.L espuma 3 2 L1.L 2741 sch ma, ae, f. 1 2 2 HV1.L f esquema 3 2 L1.L esquema 3 2 L1.L esquema 3 2 L1.L 2742 st la, ae, f., = 1 2 2 HV1.L f estela 3 2 L1.L estela 3 2 L1.L estela 3 2 L1.L 2743 *st va, dialect.de st va, ae, f. 1 2 2 HV1.L f esteva 3 2 L1.L esteva 3 2 L1.L esteva 3 2 L1.L

PAGE 475

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp HV1.L (43) 2744 Fauna, ae, f. 1 2 2 HV1.L f fauna 22HV1.L fauna 22HV 1.L fauna 22HV1.L 2745 f bra, ae, f. 1 2 2 HV1.L f fibra 2 2 L1.L fibra 2 2 L1.L fibra 2 2 L1.L 2746 Fl ra, ae, f. 1 2 2 HV1.L f flora 2 2 L1.L flora 2 2 L1.L flora 2 2 L1.L 2747 ph c ae, f. 1 2 2 HV1.L f foca 2 2 L1.L foca 2 2 L1.L foca 2 2 L1.L 2748 cr ta, ae, f. 1 2 2 HV1.L f greda 2 2 L1.L greda 2 2 L1.L greda 2 2 L1.L 2749 ra, ae, f. 1 2 2 HV1.L f ira 2 2 L1.L ira 2 2 L1.L ira 2 2 L1.L 2750 l pra, ae, f. 1 2 2 HV1.L f lepra 2 2 L1.L lepra 2 2 L1.L lepra 2 2 L1.L 2751 l bra, ae, f. 1 2 2 HV1.L f lliura 2 2 HV1.L libra 2 2 L1.L libra 2 2 L1.L 2752 l ma, ae, f. 1 2 2 HV1.L f llima 2 2 L1.L lima 2 2 L1.L lima 2 2 L1.L 2753 l ra, ae, f. 1 2 2 HV1.L f lira 2 2 L1.L lira 2 2 L1.L lira 2 2 L1.L 2754 pl ga, ae, f. 1 2 2 HV1.L f plaga 2 2 L1.L llaga 2 2 L1.L chaga 2 2 L1.L 2755 l na, ae, f. 1 2 2 HV1.L f lluna 2 2 L1.L luna 2 2 L1.L lua 2 2 L1.L 2756 m ta, ae, f. 1 2 2 HV1.L f meta 2 2 L1.L meta 2 2 L1.L meta 2 2 L1.L 2757 m ca, ae, f. 1 2 2 HV1.L f mica 2 2 L1.L mica 2 2 L1.L mica 2 2 L1.L 2758 m tra, ae, f., = 1 2 2 HV1.L f mitra 2 2 L1.L mitra 2 2 L1.L mitra 2 2 L1.L 2759 M sa, ae, f., = 1 2 2 HV1.L f musa 2 2 L1.L musa 2 2 L1.L musa 2 2 L1.L 2760 nauta, ae, m. 1 2 2 HV1.L f nauta 2 2 HV1.L nauta 2 2 HV1.L nauta 2 2 HV1.L 2761 pausa, ae, f. 1 2 2 HV1.L f pausa 2 2 HV1.L pausa 475 2 2 HV1.L pausa 2 2 HV1.L 2762 poena, ae, f. 1 2 2 HV1.L f pena 2 2 L1.L pena 2 2 L1.L pena 2 2 L1.L 2763 p la, ae, f. 1 2 2 HV1.L f pila 2 2 L1.L pila 2 2 L1.L pilha 2 2 L1.L 2764 pl ga, ae, f. 1 2 2 HV1.L f plaga 2 2 L1.L plaga 2 2 L1.L plaga 2 2 L1.L 2765 pl na, ae, f. 1 2 2 HV1.L f plana 2 2 L1.L plana 2 2 L1.L plana 2 2 L1.L 2766 pl ma, ae, f. 1 2 2 HV1.L f ploma 2 2 L1.L pluma 2 2 L1.L pluma 2 2 L1.L 2767 pr mus, a, um 1 2 2 HV1.L f prim a 2 2 L1.L prima 2 2 L1.L prima 2 2 L1.L 2768 pr sa, ae, f. 1 2 2 HV1.L f prosa 2 2 L1.L prosa 2 2 L1.L prosa 2 2 L1.L 2769 r ga, ae, f. 1 2 2 HV1.L f rua 2 2 L1.L ra 2 2 L1.L rua 2 2 L1.L 2770 r ta, ae, f., = 1 2 2 HV1.L f ruda 2 2 L1.L ruda 2 2 L1.L ruda 2 2 L1.L 2771 saeta (s ta), ae, f. 1 2 2 HV1.L f s eda 2 2 L1.L seda 2 2 L1.L seda 2 2 L1.L 2772 t la ae, f. 1 2 2 HV1.L f tela 2 2 L1.L tela 2 2 L1.L tela 2 2 L1.L.L 2773 tina 1 2 2 HV1.L f tina 2 2 L1.L tina 2 2 L1.L tina 2 2 L1.L 2774 v ra ae, f. 1 2 2 HV1.L f vara 2 2 L1.L vara 2 2 L1.L vara 2 2 L1.L 2775 v na ae, f. 1 2 2 HV1.L f vena 2 2 L1.L vena 2 2 L1.L veia 2 2 L1.L 2776 v ta ae 1 2 2 HV1.L f vida 2 2 L1.L vida 2 2 L1.L vida 2 2 L1.L 2777 ra ae, f. 1 2 2 HV1.L f ar a 2 2 L1.L ara 2 2 L1.L ara 2 2 L1.L 2778 d va (dia ), ae, f. 1 2 2 HV1.L f diva 2 2 L1.L diva 2 2 L1.L diva 2 2 L1.L

PAGE 476

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 2779 f ma, ae, f. 1 2 2 HV1.L f fama 22L1.L fama 22L1.L fama 22L1.L 2780 gl ma, ae, f. 1 2 2 HV1.L f gluma 22L1.L gluma 22L1.L gluma 22L1.L 2781 fata, f. (< f tum, i, n.) 1 2 2 HV1.L f fada 2 2 L1.L hada 2 2 L1.L fada 2 2 L1.L 2782 h ra, ae, f. 1 2 2 HV1.L f hora 2 2 L1.L hora 2 2 L1.L hora 2 2 L1.L 2783 chra, ae, f., del gr. 1 2 2 HV1.L f ocre 2 2 L1.L ocre 2 2 L1.L ocre 2 2 L1.L 2784 da, ae, f., del gr. 1 2 2 HV1. L f oda 2 2 L1.L oda 2 2 L1.L oda 2 2 L1.L 2785 p la, ae, f. 1 2 2 HV1.L f pala 2 2 L1.L pala 2 2 L1.L p 1 1 HV1 278 6 l na, ae, f. 1 2 2 HV1.L f llana 2 2 L1.L lana 2 2 L1.L l 1 1 HV1 L1.X (40) 2787 ala, ae, f. 1 2 2 L1.L f al a 2 2 L1.L ala 2 2 L1.L ala 2 2 L1.L 2788 b a, ae, f. 1 2 2 L1.L f boa 2 2 L1.L boa 2 2 L1.L boa 2 2 L1.L 2789 capra ae, f. 1 2 2 L1.L f cabra 2 2 L1.L cabra 2 2 L1.L cabra 2 2 L1.L 2790 Der. regres. del lat. cac re 1 2 2 L1.L f caca 2 2 L1.L caca 2 2 L1.L caca 2 2 L1.L 2791 c sa ae, f. 1 2 2 L1.L f casa 2 2 L1.L casa 2 2 L1.L casa 2 2 L1.L 2792 *cova 1 2 2 L1.L f cova 2 2 L1.L cueva 2 2 L1.L cova 2 2 L1.L 2793 lat. tardo [charta] data 1 2 2 L1.L f data 2 2 L1.L dat a 2 2 L1.L data 2 2 L1.L 2794 sch la (sc la), ae, f. 1 2 2 L1.L f escola 3 2 HC.L1. L escuela 3 2 HC.L1.L escola 3 2 HC.L1.L 2795 sp tha, ae, f. 1 2 2 L1.L f espasa 3 2 HC. L1.L espada 476 3 2 HC.L1.L espada 3 2 HC.L1.L 2796 sp tha, ae, f., = 1 2 2 L1.L f espata 3 2 HC.L1.L es pata 3 2 HC.L1.L espata 3 2 HC.L1.L 2797 st la, ae, f., = 1 2 2 L1.L f estola 3 2 HC.L1.L es tola 3 2 HC.L1.L estola 3 2 HC.L1.L 2798 str pha, ae, f., = 1 2 2 L1.L f estrofa 3 2 HC.L1.L estrofa 3 2 HC.L1.L estrofe 3 2 HC.L1.L 2799 f ra, ae, f. 1 2 2 L1.L f fera 2 2 L1.L fiera 2 2 L1.L fera 2 2 L1.L 2800 f ga, ae, f. 1 2 2 L1.L f fuga 2 2 L1.L fuga 2 2 L1.L fuga 2 2 L1.L 2801 g la, ae, f. 1 2 2 L1.L f gola 2 2 L1.L gola 2 2 L1.L gola 2 2 L1.L 2802 g la, ae, f. 1 2 2 L1.L f gola 2 2 L1.L gula 2 2 L1.L gula 2 2 L1.L 2803 f ba, ae, f. 1 2 2 L1.L f fava 2 2 L1.L haba 2 2 L1.L fava 2 2 L1.L 2804 f bra, ae, f. 1 2 2 L1.L f fibr a 2 2 L1.L hebra 2 2 L1.L fibra 2 2 L1.L 2805 Hydra, ae, f., = 1 2 2 L1.L f hidra 2 2 L1.L hidra 2 2 L1.L hidra 2 2 L1.L 2806 l pa ae, f. 1 2 2 L1.L f lloba 2 2 L1.L loba 2 2 L1.L loba 2 2 L1.L 2807 m na, ae, f. 1 2 2 L1.L f mina 2 2 L1.L mina 2 2 L1.L mina 2 2 L1.L 2808 m la, ae, f. 1 2 2 L1.L f mola 2 2 L1.L mola 2 2 L1.L mola 2 2 L1.L 2809 m ra, ae, f. 1 2 2 L1.L f mora 2 2 L1.L mora 2 2 L1.L mora 2 2 L1.L 2810 n ta ae, f. 1 2 2 L1.L f nota 2 2 L1.L nota 2 2 L1.L nota 2 2 L1.L 2811 papas (pappas), ae & tis, m. 1 2 2 L1.L m papa 2 2 L1.L papa 2 2 L1.L papa 2 2 L1.L

PAGE 477

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 2812 p tra, ae, f., = 1 2 2 L1.L f pedra 2 2 L1.L piedra 2 2 L1.L pedra 2 2 L1.L 2813 pyra, ae, f., = 1 2 2 L1.L f pira 2 2 L1.L pira 2 2 L1.L pira 2 2 L1.L 2814 r sa, ae, f 1 2 2 L1.L f rosa 2 2 L1.L rosa 2 2 L1.L rosa 2 2 L1.L 2815 r ta, ae, f. 1 2 2 L1.L f roda 2 2 L1.L rueda 2 2 L1.L roda 2 2 L1.L 2816 lat. vulg. socra, ae, f.
PAGE 478

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp HC.HV1.X (54) 2845 alb ra 1 3 2 HC.HV1.Lf albura 32HC.L1. L albura 32HC.L1.L alvura 32HC.L1.L 2846 ang na 1 3 2 HC.HV1.Lf angina 32HC.L1. L angina 32HC.L1.L angina 32HC.L1.L 2847 ann na 1 3 2 HC.HV1.Lf annona 32L. L1.L anona 32L.L1. L anona 32L.L1.L 2848 arm ta, f. de arm tus 1 3 2 HC.HV1.Lf armada 32HC.L1. L armada 32HC.L1.L armada 32HC.L1.L 2849 Harpyia (trisyl.), ae, f y este del gr. 1 3 2 HC.HV1.Lf harpia 32HC.L1.L arpa 32HC.L1.L harpia 32HC.L1.L 2850 b. lat. asc ta, y este del gr. 1 3 2 HC.HV1.Lm/fasceta 32HC.L1.L asceta 32HC.L1.L asceta 32HC.L1.L 2851 thl ta ae, com. 1 3 2 HC.HV1.Lm/fatleta 32HC.L1.L atleta 32 HC.L1.L atleta 3 2 HC.L1.L 2852 cens ra ae, f. 1 3 2 HC.HV1.L f censura 3 2 HC.L1.L censura 3 2 HC.L1.L censura 3 2 HC.L1.L 2853 cinct ra ae, f. 1 3 2 HC.HV1.L f cintura 3 2 HC.L1.L cintura 3 2 L.L1. L cintura 3 2 L.L1.L 2854 coll ga (conl), ae, m. 1 3 2 HC.HV1.L m co llega 3 2 HC.L1.L colega 3 2 L.L1.L colega 3 2 L.L1.L 2855 cort na ae, f. 1 3 2 HC.HV1.L f cortina 3 2 HC. L1.L cortina 3 2 HC.L1.L cortina 3 2 HC.L1.L 2856 cult ra ae, f. 1 3 2 HC.HV1.L f cultura 3 2 HC.L1.L cultura 3 2 HC.L1.L cultura 3 2 HC.L1.L 2857 cogn ta ae, f. 1 3 2 HC.HV1.L e cunyada 3 2 L.L1.L cuada 3 2 L.L1. L cunhada 3 2 L.L1.L 2858 discr tus, a, um 1 3 2 HC.HV1.L f discreta 3 2 HC.L1.L discreta 3 2 HC.L1. L discreta 3 2 HC.L1.L 2859 dyspnoea ae, f., del gr. 1 3 2 HC.HV1.L f dispnea 3 2 HC.L1.L disnea 3 2 HC.L1.L dispneia 3 2 HC.L1.L 2860 doctr na ae, f. 1 3 2 HC.HV1.L f doctrina 3 2 HC. L1.L doctrina 478 3 2 HC.L1.L doutrina 3 2 HV.L1.L 2861 g ng va ae, f. 1 3 2 HC.HV1.L f geniva 3 2 L.L1.L enca 3 2 HC.L1. L gengiva 3 2 HC.L1.L 2862 gangraena, ae, f. 1 3 2 HC.HV1.L f gangrena 3 2 HC.L1.L gangrena 3 2 HC.L1.L gangrena 3 2 HC.L1.L 2863 script ra, ae, f. 1 3 2 HC.HV1.L f escriptura 4 2 HC.L1.L escritura 4 2 L.L1. L escritura 4 2 L.L1.L 2864 sculpt ra, ae, f. 1 3 2 HC.HV1.L f escultura 4 2 HC. L1.L escultura 4 2 HC.L1.L escultura 4 2 HC.L1.L 2865 struct ra, ae, f. 1 3 2 HC.HV1.L f estructura 4 2 HC. L1.L estructura 4 2 HC. L1.L estrutura 4 2 L.L1.L 2866 fact ra, ae, f. 1 3 2 HC.HV1.L f factura 3 2 HC. L1.L factura 3 2 HC.L1. L factura 3 2 HC.L1.L 2867 fiss ra, ae, f. 1 3 2 HC.HV1.L f fissura 3 2 L. L1.L fisura 3 2 L.L1.L fissura 3 2 L.L1.L 2868 font na, ae, f. 1 3 2 HC.HV1.L f fontana 3 2 HC.L1.L fontana 3 2 HC.L1.L fontana 3 2 HC.L1.L 2869 fort na, ae, f. 1 3 2 HC.HV1.L f fortuna 3 2 HC. L1.L fortuna 3 2 HC.L1.L fortuna 3 2 HC.L1.L 2870 fract ra, ae, f. 1 3 2 HC.HV1.L f fractura 3 2 HC. L1.L fractura 3 2 HC.L1.L fractura 3 2 HC.L1.L 2871 germ na, ae, f., 1 3 2 HC.HV1.L f germana 3 2 HC.L1.L hermana 3 2 HC.L1. L irm 2 1 HC.HV1 2872 form ca, ae, f. 1 3 2 HC.HV1.L f formiga 3 2 L. L1.L hormiga 3 2 L.L1.L formiga 3 2 L.L1.L 2873 junct ra ae, f. 1 3 2 HC.HV1.L f juntura 3 2 HC.L1.L juntura 3 2 HC.L1. L juntura 3 2 HC.L1.L 2874 lampr da 1 3 2 HC.HV1.L f llamprea 3 2 HC.L1.L lamprea 3 2 HC.L1.L lampreia 3 2 HC.L1.L 2875 lact ca, ae, f. 1 3 2 HC.HV1.L f lletuga 3 2 L.L1.L lechuga 3 2 L.L1. L leituga 3 2 L.L1.L 2876 lect ra, ae, f. 1 3 2 HC.HV1.L f lectura 3 2 HC. L1.L lectura 3 2 HC.L1. L leitura 3 2 HC.L1.L 2877 mant sa (mantissa), ae, f. 1 3 2 HC.HV1.L f mantissa 3 2 HC.L1.L mantisa 3 2 HC. L1.L mantissa 3 2 HC.L1.L 2878 membr na, ae, f. 1 3 2 HC.HV1.L f membrana 3 2 HC.L1.L membrana 3 2 HC.L1.L membrana 3 2 HC.L1.L

PAGE 479

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 2879 mens ra. ae, f. 1 3 2 HC.HV1.Lf mesura 32L. L1.L mesura 32L.L1.L mesura 32L.L1.L 2880 mist ra (mix-), ae, f. 1 3 2 HC.HV1.Lf mixtura 32HC.L1.L mixtura 32HC. L1.L mistura 32HC.L1.L 2881 obl tus, a, um 1 3 2 HC.HV1.Lf oblata 32 L.L1.L oblata 32HC.L1. L oblata 32HC.L1.L 2882 oct va, ae, f. 1 3 2 HC.HV1.Lf octava 32HC. L1.L octava 32HC.L1.L oitava 32HV.L1.L 2883 urt ca, ae, f. 1 3 2 HC.HV1.Lf ortiga 32 HC.L1.L ortiga 32HC.L1.L urtiga 3 2 HC.L1.L 2884 panth ra, ae, f. 1 3 2 HC.HV1.L f pantera 3 2 HC.L1.L pantera 3 2 HC.L1. L pantera 3 2 HC.L1.L 2885 pers na, ae, f. 1 3 2 HC.HV1.L f persona 3 2 HC.L1.L persona 3 2 HC. L1.L pessoa 3 2 HC.L1.L 288 6 pisc na, ae, f. 1 3 2 HC.HV1.L f piscina 3 2 HC. L1.L piscina 3 2 HC.L1.L piscina 3 2 HC.L1.L 2887 rupt ra, ae,f. 1 3 2 HC.HV1.L f ruptura 3 2 HC.L1. L ruptura 3 2 HC.L1. L ruptura 3 2 HC.L1.L 2888 sard na, ae, f. 1 3 2 HC.HV1.L f sardina 3 2 HC. L1.L sardina 3 2 HC.L1.L sardinha 3 2 HC.L1.L 2889 *subt na, de subtus, adv. 1 3 2 HC.HV 1.L f sotana 3 2 L.L1.L sotana 3 2 L.L1.L sotana 3 2 L.L1.L 2890 text ra ae, f. 1 3 2 HC.HV1.L f textura 3 2 HC.L1.L textura 3 2 HC.L1. L textura 3 2 HC.L1.L 2891 tinct ra 1 3 2 HC.HV1.L f tintura 3 2 HC.L1. L tintura 3 2 HC.L1.L tintura 3 2 HC.L1.L 2892 tons ra ae, f. 1 3 2 HC.HV1.L f tonsura 3 2 HC.L1.L tonsura 3 2 HC.L1.L tonsura 3 2 HC.L1.L 2893 tort ra ae, f. 1 3 2 HC.HV1.L f tortura 3 2 HC.L1.L tortura 3 2 HC.L1.L tortura 3 2 HC.L1.L 2894 unct ra ae, f. 1 3 2 HC.HV1.L f untura 3 2 HC.L1.L untura 3 2 HC.L1.L untura 3 2 HC.L1.L 2895 lat. tardo vent sa, ae 1 3 2 HC.HV1.L f ventosa 3 2 HC.L1. L ventosa 3 2 HC.L1.L ventosa 3 2 HC.L1.L 2896 verb na ae, f. 1 3 2 HC.HV1.L f berbena 3 2 HC. L1.L verbena 3 2 HC.L1.L verbena 3 2 HC.L1.L 2897 verr ca ae, f. 1 3 2 HC.HV1.L m berruga 3 2 L. L1.L verruga 479 3 2 L.L1.L verruga 3 2 L.L1.L 2898 vert bra ae, f. 1 3 2 HC.HV1.L f vrtebra 3 3 HC1 .L.L vrtebra 3 3 HC1.L.L vrtebra 3 3 HC1.L.L HC.L1.X (1) 2899 esch ra, ae, f., = 1 3 2 HC.L1.L f escara 3 2 HC.L1.L e scara 3 2 HC.L1.L escara 3 2 HC.L1.L L.HC1.L (7) 2900 d fensa ae, f. 1 3 2 HV.HC1.L f defensa 3 2 L.HC1.L defensa 3 2 L. HC1.L defesa 3 2 L.L1.L 2901 d fensa ae, f. 1 3 2 HV.HC1.L f devesa 3 2 L.L1.L dehesa 3 2 L. L1.L devesa 3 2 L.L1.L 2902 Syringa ae, f., 1 3 2 HV.HC1.L f xeringa 3 2 L.HC1.L jeringa 3 2 L.HC1.L seringa 3 2 L.HC1.L 2903 l bertus, a, um 1 3 2 HV.HC1.L f lliberta 3 2 L. HC1.L liberta 3 2 L.HC1.L liberta 3 2 L.HC1.L 2904 praebenda, ae, f. 1 3 2 HV.HC1.L f prebenda 3 2 L.HC1.L prebenda 3 2 L.HC1.L prebenda 3 2 L.HC1.L 2905 t trarches, ae, m., del gr. 1 3 2 HV.HC1.L m tetrarca 3 2 L.HC1.L tetrarca 3 2 L.HC1.L tetrarca 3 2 L.HC1.L 2906 p pilla, ae, f, 1 3 2 HV.HC1.L f pupilla 3 2 L.HC1.L pup ila 3 2 L.L1.L pupila 3 2 L.L1.L HV.HV1.L (28) 2907 aur ga ae 1 3 2 HV.HV1.L m auriga 3 2 HV.L1. L auriga 3 2 HV.L1. L auriga 3 2 HV.L1.L 2908 aur ra ae, f. 1 3 2 HV.HV1.L f aurora 3 2 HV .L1.L aurora 3 2 HV.L1. L aurora 3 2 HV.L1.L 2909 b laena (b l na ), ae, f., y este del gr. 1 3 2 HV.HV1.L f balena 3 2 L.L1.L bal lena 3 2 L.L1.L baleia 3 2 L.L1.L 2910 caut la ae, f. 1 3 2 HV.HV1.L f cautela 3 2 HV .L1.L cautela 3 2 HV.L1. L cautela 3 2 HV.L1.L

PAGE 480

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 2911 cael ta 1 3 2 HV.HV1.Lf celada 32L.L1. L celada 32L.L1.L celada 32L.L1.L 2912 claus ra, ae, f. 1 3 2 HV.HV1.Lf clausura 32HV. L1.L clausura 32HV.L1.L clausura 3 2 HV.L1.L 2913 qu dr ga ae, f., 1 3 2 HV.HV1.L f quadriga 3 2 L.L1.L cuadriga 3 2 L.L1. L quadriga 3 2 L.L1.L 2914 quaest ra ae, f. 1 3 2 HV.HV1.L f qestura 3 2 HC. L1.L cuestura 3 2 HC. L1.L questura 3 2 HC.L1.L 2915 l tr na, ae, f. 1 3 2 HV.HV1.L f latrina 3 2 L. L1.L letrina 3 2 L.L1. L latrina 3 2 L.L1.L 2916 l r ca, ae, f. 1 3 2 HV.HV1.L f lloriga 3 2 L. L1.L loriga 3 2 L.L1.L loriga 3 2 L.L1.L 2917 m tr na, ae, f. 1 3 2 HV.HV1.L f matrona 3 2 L.L1.L matrona 3 2 L.L1.L matrona 3 2 L.L1.L 291 8 n t vus, a, um 1 3 2 HV.HV1.L f nativa 3 2 L.L1.L nativa 3 2 L. L1.L nativa 3 2 L.L1.L 2919 n t ra, ae, f. 1 3 2 HV.HV1.L f natura 3 2 L.L1.L natura 3 2 L.L1. L natura 3 2 L.L1.L 2920 r na, ae, f. 1 3 2 HV.HV1.L f orina 3 2 L.L1.L orina 3 2 L. L1.L urina 3 2 L.L1.L 2921 p g nus, a, um, adj. 1 3 2 HV.HV1.L f pag ana 3 2 L.L1.L pagana 3 2 L.L1.L pag 2 1 L.HV1 2922 p r ta, ae, m., = 1 3 2 HV.HV1.L m pirata 3 2 L.L1.L pirata 3 2 L.L1.L pirata 3 2 L.L1.L 2923 pl b us (jus), a, um, adj. 1 3 2 HV.HV1.L f plebea 3 2 L.L1.L pl ebeya 3 2 L.L1.L plebeia 3 2 L.L1.L 2924 praet ra, ae, f. 1 3 2 HV.HV1.L f pretura 3 2 L.L1.L pretura 3 2 L.L1.L pretura 3 2 L.L1.L 2925 r s ra, ae, f. 1 3 2 HV.HV1.L f rasura 3 2 L. L1.L rasura 3 2 L.L1.L rasura 3 2 L.L1.L 2926 r g na, ae, f. 1 3 2 HV.HV1.L f reina 3 3 L.L1.L reina 2 2 HV1. L rainha 3 2 L.L1.L 2927 r s na, ae, f. 1 3 2 HV.HV1.L f resina 3 2 L. L1.L resina 3 2 L.L1.L resina 3 2 L.L1.L 2928 s t ra ae, f. 1 3 2 HV.HV1.L f sutura 3 2 L. L1.L sutura 480 3 2 L.L1.L sutura 3 2 L.L1.L 2929 t t la ae, f. 1 3 2 HV.HV1.L f tutela 3 2 L.L1.L tutela 3 2 L.L1.L tutela 3 2 L.L1.L 2930 r thra ae, f., y este del gr. 1 3 2 HV.HV1.L f uretra 3 2 L.L1.L uretra 3 2 L.L1.L uretra 3 2 L.L1.L 2931 s ra ae, f. 1 3 2 HV.HV1.L f usura 3 2 L.L1.L usura 3 2 L. L1.L usura 3 2 L.L1.L 2932 v g na ae, f. 1 3 2 HV.HV1.L f vagina 3 2 L.L1.L vagina 3 2 L.L1. L vagina 3 2 L.L1.L 2933 v g na ae, f. 1 3 2 HV.HV1.L f beina 3 2 HV1.L vaina 2 2 HV1.L bainha 2 2 HV1.L 2934 v s ca, ae, f. (mss. vess ca) 1 3 2 HV.HV1.L f veixiga 3 2 L.L1. L vejiga 3 2 L.L1.L bexiga 3 2 L.L1.L L.HC1.L (29) 2935 r missa, ae, f. 1 3 2 L.HC1.L f remesa 3 2 L.L1.L remesa 3 2 L. L1.L remessa 3 2 L.L1.L 2936 agenda 1 3 2 L.HC1.L f agenda 3 2 L.HC1.L agenda 3 2 L.HC1.L agenda 3 2 L.HC1.L 2937 lumna ae 1 3 2 L.HC1.L f alumna 3 2 L.HC1.L alumna 3 2 L.HC1.L aluna 3 2 L.L1.L 2938 rista ae, f. 1 3 2 L.HC1.L f aresta 3 2 L.HC1.L arista 3 2 L.HC1.L aresta 3 2 L.HC1.L 2939 Kalendae (Cal-; often written K), rum, f. 1 3 2 L.HC1.L f calenda 3 2 L.HC1.L calenda 3 2 L.HC1.L calenda 3 2 L.HC1.L 2940 b. lat. *cappella, dim. de cappa, 1 3 2 L.HC1.L f capella 3 2 L.L1.L c apilla 3 2 L.L1.L capela 3 2 L.L1.L 2941 c terva ae, f. 1 3 2 L.HC1.L f caterva 3 2 L. HC1.L caterva 3 2 L.HC1.L caterva 3 2 L.HC1.L 2942 c verna ae, f. 1 3 2 L.HC1.L f caverna 3 2 L. HC1.L caverna 3 2 L.HC1.L caverna 3 2 L.HC1.L

PAGE 481

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 2943 c culla ae, f. 1 3 2 L.HC1.L f cogulla 32 L.L1.L cogulla 32L. L1.L cogula 32L.L1.L 2944 c lumna ae, f. 1 3 2 L.HC1.L f columna 32L. HC1.L columna 32L.HC1.L coluna 32L.L1.L 2945 c rolla ae, f. 1 3 2 L.HC1.L f corolla 32L.HC1.L corola 32L.L1.L corola 32L.L1.L 2946 g nesta o g nista, ae, f. 1 3 2 L.HC1.L f ginesta 32L. HC1.L ginesta 32L.HC1.L giesta 32L.HC1.L 2947 g nesta or g nista, ae, f. 1 3 2 L.HC1.L f ginesta 32L.HC1 .L hiniesta 32L.HC1.L giesta 32L.HC1.L 2948 m taxa, ae, f., = 1 3 2 L.HC1.L f madeixa 32L.L1.L madeja 32L.L1.L madeixa 32L.L1.L 2949 m dulla, ae, f. 1 3 2 L.HC1.L f medulla 32 L.HC1.L mdula 33L1. L.L medula 32L.L1.L 295 0 p laestra, ae, f. 1 3 2 L.HC1.L f palestra 3 2 L. HC1.L palestra 3 2 L.HC1.L palestra 3 2 L.HC1.L 2951 p pilla, ae, f. 1 3 2 L.HC1.L f papilla 3 2 L.HC1.L papila 3 2 L.L1.L papila 3 2 L.L1.L 2952 pl centa, ae, f. 1 3 2 L.HC1.L f placenta 3 2 L.HC1 .L placenta 3 2 L.HC1.L placenta 3 2 L.HC1.L 2953 p lenta, ae, f. 1 3 2 L.HC1.L f polenta 3 2 L.HC1.L polenta 3 2 L.HC1.L polenta 3 2 L.HC1.L 2954 qu r la o qu rella, ae, f. 1 3 2 L.HC1.L f querella 3 2 L.L1.L querella 3 2 L. L1.L querela 3 2 L.L1.L 2955 r pressus, a, um 1 3 2 L.HC1.L f represa 3 2 L.L1.L represa 3 2 L. L1.L represa 3 2 L.L1.L 2956 r pulsa, ae, f. 1 3 2 L.HC1.L f repulsa 3 2 L. HC1.L repulsa 3 2 L.HC1.L repulsa 3 2 L.HC1.L 2957 r tortus, a, um 1 3 2 L.HC1.L f retorta 3 2 L. HC1.L retorta 3 2 L.HC1.L retorta 3 2 L.HC1.L 2958 s gitta, ae, f. 1 3 2 L.HC1.L f sageta 3 2 L.L1.L saeta 3 2 L.L1.L seta 2 2 L1.L 2959 s cundus, a, um 1 3 2 L.HC1.L f segona 3 2 L. L1.L segunda 3 2 L.HC1.L segunda 3 2 L.HV1.L 2960 S bylla, ae, f.,= 1 3 2 L.HC1.L f sibilla 3 2 L.HC1 .L sibila 3 2 L.L1.L sibila 3 2 L.L1.L 2961 s phista, ae, m., = 1 3 2 L.HC1.L m/f sofista 3 2 L.HC1.L sofista 4813 2 L.HC1.L sofista 3 2 L.HC1.L 2962 t berna ae, f. 1 3 2 L.HC1.L f taverna 3 2 L. HC1.L taberna 3 2 L.HC1.L taverna 3 2 L.HC1.L 2963 *tragella, de trag l a 1 3 2 L.HC1.L f tragella 3 2 L. L1.L tralla 3 2 L. L1.L trela 2 2 L1.L 2964 m ca 1 3 2 L.HV1.L f amiga 3 2 L. L1.L amiga 3 2 L.L1. L amiga 3 2 L.L1.L 2965 ar na 1 3 2 L.HV1.L f arena 3 2 L.L1. L arena 3 2 L.L1.L arena 3 2 L.L1.L 2966 v na ae, f. 1 3 2 L.HV1.L f avena 3 2 L.L1.L avena 3 2 L. L1.L aveia 3 2 L.L1.L 2967 cat na, ae 1 3 2 L.HV1.L f cadena 3 2 L.L1. L cadena 3 2 L.L1.L cadeia 3 2 L.L1.L 2968 camp na, ae, f. 1 3 2 HC.HV1.L f campana 3 2 HC. L1.L campana 3 2 HC.L1.L campana 3 2 HV.L1.L 2969 cand la ae, f. 1 3 2 HC.HV1.L f candela 3 2 HC. L1.L candela 3 2 HC.L1. L candeia 3 2 HV.L1.L 2970 c th dra ae, f., del gr. 1 3 2 L.HV1.L f ctedr a 3 3 L1.L.L ctedra 3 3 L1.L.L ctedra 3 3 L1.L.L 2971 c c ta ae, f. 1 3 2 L.HV1.L f cicuta 3 2 L.L1.L cicuta 3 2 L.L1. L cicuta 3 2 L.L1.L 2972 Ch maera ae, f. 1 3 2 L.HV1.L f cimera 3 2 L.L1.L cimera 3 2 L.L1. L cimeira 3 2 L.HV1.L 2973 sciss ra ae, f. 1 3 2 L.HV1.L f cissura 3 2 L. L1.L cisura 3 2 L.L1.L cissura 3 2 L.L1.L 2974 clca ae, f. 1 3 2 L.HV1.L f cloaca 3 2 L.L1.L cloaca 3 2 L.L1. L cloaca 3 2 L.L1.L 2975 c qu na ae, f. 1 3 2 L.HV1.L f cuina 2 2 HV 1.L cocina 3 2 L.L1.L cozinha 3 2 L.L1.L

PAGE 482

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 2976 c m tes ae, m., del gr. 1 3 2 L.HV1.L m/f comet a 3 2 L.L1.L cometa 3 2 L.L1.L cometa 3 2 L.L1.L 2977 c r na, ae, f. 1 3 2 L.HC1.L f corona 3 2 L.L1.L corona 3 2 L.L1.L coroa 3 2 L.L1.L 2978 qu terni ae, a, adj. pl. 1 3 2 L.HV1.L f quaderna 3 2 L.HC1.L cuaderna 3 2 L.HC1.L caderna 3 2 L.HC1.L 2979 c l bra ae, f. 1 3 2 L.HV1.L f colobra 3 2 L.L1.L culebra 3 2 L.L1.L cobra 2 2 L1.L 2980 diaeta, ae, del gr. 1 3 2 L.HV1.L f dieta 3 2 L.L1.L dieta 3 2 L.L1.L dieta 3 2 L.L1.L 2981 st t ra, ae, f. 1 3 2 L.HV1.L f estatura 4 2 L.L1.L estatura 4 2 L.L1.L estatura 4 2 L.L1.L 2982 f g ra, ae, f. 1 3 2 L.HV1.L f figura 3 2 L.L1.L figura 3 2 L.L1.L figura 3 2 L.L1.L 298 3 g l na, ae, f. 1 3 2 L.HV1.L f galena 3 2 L.L1.L galena 3 2 L. L1.L galena 3 2 L.L1.L 2984 gall na, ae, f. 1 3 2 L.HV1.L f gallina 3 2 L.L1.L gallina 3 2 L.L1.L galinha 3 2 L.L1.L 2985 f r na, ae, f. 1 3 2 L.HV1.L f farina 3 2 L.L1.L harina 3 2 L.L1. L farinha 3 2 L.L1.L 2986 hyaena, ae, f., = 1 3 2 L.HV1.L f hiena 2 2 L1.L hiena 2 2 L1.L hiena 2 2 L1.L 2987 ta ae, f., del gr. 1 3 2 L.HV1.L f jota 2 2 L1.L jota 2 2 L1.L jota 2 2 L1.L 2988 l c na, ae, f. 1 3 2 L.HV1.L f llacuna 3 2 L. L1.L laguna 3 2 L.L1.L laguna 3 2 L.L1.L 2989 latr a, ae, f. 1 3 2 L.HV1.L f latria 3 2 L. L1.L latra 3 2 L.L1.L latria 3 2 L.L1.L 2990 m g a, ae, f., = 1 3 2 L.HV1.L f mgia 3 3 L1. L.L magia 2 2 L1.L magia 3 2 L.L1.L 2991 m r nus, a, um 1 3 2 L.HV1.L f marina 3 2 L.L1.L marina 3 2 L.L1. L marinha 3 2 L.L1.L 2992 M n ta, ae, f. 1 3 2 L.HV1.L f moneda 3 2 L. L1.L moneda 3 2 L.L1. L moeda 3 2 L.L1.L 2993 Murena (alt. Muraena), ae, f., = 1 3 2 L.HV1.L f morena 3 2 L.L1.L morena 3 2 L.L1.L moreia 3 2 L.L1.L 2994 n v nus a, um 1 3 2 L.HV1.L f novena 3 2 L.L1.L novena 482 3 2 L. L1.L novena 3 2 L.L1.L 2995 l va, ae, f. 1 3 2 L.HV1.L f oliva 3 2 L. L1.L oliva 3 2 L.L1.L oliva 3 2 L.L1.L 2996 pyr tes, ae, m., = 1 3 2 L.HV1.L f pirita 3 2 L.L1.L pirita 3 2 L.L1.L pirita 3 2 L.L1.L 2997 pl n ta, ae, m. 2 3 2 L.HV1.L m/f planeta 3 2 L.L1.L planeta 3 2 L.L1. L planeta 3 2 L.L1.L 2998 p ta, ae, m. 1 3 2 L.HV1.L m poeta 3 2 L.L1.L poeta 3 2 L. L1.L poeta 3 2 L.L1.L 2999 pr ph ta & pr ph tes, ae, m., = 1 3 2 L.HV1.L f profeta 3 2 L.L1.L profeta 3 2 L. L1.L profeta 3 2 L.L1.L 3000 pr p na, ae, f. 1 3 2 L.HV1.L f propina 3 2 L. L1.L propina 3 2 L.L1.L propina 3 2 L.L1.L 3001 qu tus a, um 1 3 2 L.HV1.L f queda 2 2 L1.L queda 2 2 L1.L queda 2 2 L1.L 3002 Ch maera, ae, f. 1 3 2 L.HV1.L f quimera 3 2 L.L1.L quimera 3 2 L.L1. L quimera 3 2 L.L1.L 3003 r s da, ae, f. 1 3 2 L.HV1.L f reseda 3 2 L.L1.L reseda 3 2 L.L1.L reseda 3 2 L.L1.L 3004 lat. med. ret na < rete 1 3 2 L.HV1.L f retina 3 2 L.L1.L retina 3 2 L.L1.L retina 3 2 L.L1.L 3005 r na, ae, f. 1 3 2 L.HV1.L f runa 3 2 L.L1.L ruina 2 2 L1.L runa 3 2 L1.L 3006 s l nae, rum, f. 1 3 2 L.HV1.L f salina 3 2 L. L1.L salina 3 2 L.L1.L salina 3 2 L.L1.L 3007 s l va, ae, f. 1 3 2 L.HV1.L f saliva 3 2 L. L1.L saliva 3 2 L.L1.L saliva 3 2 L.L1.L 3008 s qu la o s quella, ae, f. 1 3 2 L.HV1.L f seqela 3 2 L1.L secuela 3 2 L1.L sequela 3 2 L1.L 3009 sent na, ae, f. 1 3 2 HC.HV1.L f sentina 3 2 HC. L1.L sentina 3 2 HC.L1.L sentina 3 2 L.L1.L 3010 ti ra, y este del gr 1 3 2 L.HV1.L f ti ara 3 2 L.L1.L tiara 2 2 L1.L tiara 2 2 L1.L

PAGE 483

Etymon Dcl Ac Temp Gen CAT Ac Temp CAS Ac Temp POR Ac Temp 3011 tribuna