|UFDC Home||myUFDC Home | Help|
WHY KNDA SINGS: NARRATIVE DISCOURSE AND THE MULTI FUNCTION ALITY OF BAKA SONG IN BAKA STORY By DANIEL FITZGERALD A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2011
2 2011 Daniel Fitzgerald
3 To Yves Lonard: colleague, confidant, friend, and brother
4 ACKNOWLEDGMENTS This study was made possible by so many people for whom I am grateful. Fir st and foremost, I am grateful for the constant and lively encourgement of my wife Debra my four children Emma, Aaron, Mary, and Jonathan, and my father and mother in law, Douglas and Suzanne Burton. In Cameroon, I became e specially indebted to Yves Le! onard, to whom this dissertation is dedicated. Without his foundational research of and gifted intuitions about Baka oral narrative, I would never had attempted this kind of interdisciplinary study. I have built upon his work because it has always seemed to carry the promise of making black ink on white paper serve real personal exchange, something I aspire to myself. My initial exploration into the Baka community would not have proceeded so smoothly without the humble and wise counsel of Pe" re Paul Cuypers. His many years of experience serving the Baka people eventually inspired me to do the same. The anthropological, linguistic, and missionary work of Kathleen Higgens was equally inspiring, though I have never met Kath face to face. Without the gracious gift of her original audio field recordings of traditional Baka narratives, my study would lack little historical depth. I would never have continued my research without the warm and patient welcome and engagement of so many Baka friends. In the Baka camp of Ndjibot, Konji was my first mentor and confidant. The fire in his drumming, the solace of his evening harp, and the animation in his storytelling hooked me. The masterful and comic narrative d elivery of Nguna (now deceased) eventually convinced me to focus my
5 research on Baka li! ka! n"! li!ka! n"! song, in particular. In the camp of Nomedjo, my interest in li! ka! n"! song deepened, as did my friendships. Nj#ma Mai And" and Dieudonne$ continu ally shepherded me in their language and music traditions. Their exemplary cultural leadership provided me with a rich and rare entry into Baka oral art. Without their expert and generally transparent point of view, this study would have been immeasurabl y impoverished, perhaps undoable. The broad support of my many colleagues in SIL Cameroon and SIL International provided me with helps too numerous to mention. I would not have navigated those early days of cross cultural communication in Cameroon without the collective experience of so many dedicated colleagues. I am further grateful to Cameroon's Ministe! re des Recherches Scientifique et Techniques (MINREST) for permission to carry out my research among the Baka people in Cameroon's southeastern provinc e. The c onfidence to actually sit down and write this dissertation was decisively built into me early on by my dissertation committee. I am particularly grateful to my advisor, Dr. Larry Crook, for his steady oversight. Finally, what a relief it was to h ave Dr. Robert Wright read through and comment on much of the final stages of my writing before I submitted it to my committee for final review Any errors in my research or writing are, of course, still my own
6 TABLE OF CONTENTS P age ACKNOWLEDGEMENTS 4 LIST OF FIGURES 9 LIST OF OBJECTS ... 11 ABSTRACT 12 CHAPTER 1 I NTRODUCTIONS TO THE STUDY OF BAK A LI! KA! N% SONG.. ... 14 1 1 Research Subject, Purposes, Theory, Models, and Motivatio ns.... 14 1.2 Research Literature Review 18 1. 3 Description of Chapters .. 27 1. 4 The Baka People ... 30 1. 4.1 Designations ... 30 1. 4.2 Demographic, Geographic, and H istoric G eneralities 30 1. 5 Baka Language ...... 31 1. 5.1 Language Classification ... 31 1. 5.2 Linguistic Structures ..... 32 1. 5.2.1 Phonology ....... 32 1. 5.2.2 Morphology and syntax .... 33 1. 5.3 Semantics .... 34 1. 5.4 Sociolinguistics .. 35 1.5 .5 Section S ummary .. 37 1. 6 Research Data Sample 37 2 BAKA VERBAL PERFORMA NCE IN SONG, STORY, AND STORYS SONG 46 2. 1 Baka Verbal Performance 48 2. 2 Poetic Verbal Performanc e : Song ... 51 2. 2.1 Baka Songs Rhythmic Generalities ... 53 2. 2.2 Baka Songs Melic Generalities ... 57 2. 2.3 Baka Songs Voice Register Generalities ... 64 2. 3 Narrative Verbal Performance: Story 73 2. 4 Sung Baka Narrative Performance: Storys Song 80 2. 5 S ummary 85 3 BAKA SONG TEXT POETICS: AN INV ENTORY OF POETIC DEV ICES .... 93
7 3. 1 Introduction: Sung versus Unsung Verbal Discourse .. 93 3. 2 Inventory of Baka Song Text Poetic Devices .... 99 3.3 Rhythmic Poetic Devices ..... 100 3.3.1 Line .. 100 3. 3. 2 Pulse .... 101 3. 3.3 Tempo .. 104 3. 3.4 Meter ... 105 3. 3.5 Summary of Rhythmic Poetic Dev ices ... 107 3. 4 Description of the Poetic Devices of One Baka Song .. 108 3.4 .1 Ethnographic Background .... 108 3. 4.2 Rhythmic Poetic Devices .. 110 3. 4.3 Syntactic Poetic Devices ... 111 3 .4 .3.1 Available grammatical units .... 112 3. 4.3.2 Available poetic syntactic units ... 112 3. 4.3.3 Complementary devices of syntactic development .. 113 3. 4.3.4 Frequency and distribution of syntactic poetic unit 114 3. 4.3.5 Syntac tic development and form ..... 114 3. 4.4 Lexical Poetic Devices .. 114 3. 4.5 Phonological Poetic Devices 116 3. 4.6 Semantic Level P oetic D evices 117 3. 5 Conclusion 120 4 BAKA LI! KA! N% SONG AS A DISCOURSE FEATURE OF BAKA STOR Y .. 135 4. 1 Performance Relationship of Baka Story and Song 135 4. 1.1 Story and Song in African Narrative Performance .. 135 4. 1.2 Story and Song in Baka Narrative Performance .. 138 4. 2. Cohesion of li! ka! n "! Story and Song .... 144 4. 2.1 Generative Cohesion (Stories Engender Songs) 144 4. 2.2 Performance Cohesion of li! ka!n "! Story and Song .. 147 4. 2.3 Textual Cohesion of Story and S ong .. 148 4. 2.3.1 Grammatical cohesion ... 148 Grammatical cohesion through lexical repetition 148 Grammatical cohesion through lexical substitution 151 Grammatical cohesion through common point of view ... 151 4. 2.3.2 Semantic cohe sion 154 Synonymy 154 Meronymy ... 156 4. 3 Narrative Discourse Development through Song 157 4. 3.1 Contextual N arrative D evelopment through Song ... 162 4. 3.1.1 Participant informati on through song .... 163 4. 3.1.2. Setting information through song .. 164 4. 3.1.3 Explanatory information through song ...... 165 4. 3.1.4 Collateral information through song ... 167 4. 3.1.5 Evaluative information through song 169
8 Evaluative information through propositions ... 170 Evaluative information through grammar .... 171 4. 3.2 Climactic Narrative Development through Song : Story of Lo! ndo! .. 181 4. 3.2.1 Climactic development through rhetorical underlining .. 186 Syntactic repetition 187 Rhythmic repetition 187 Melic (melodic) repetition 188 4. 3.2.2 Climactic development through heighten ed vividness 188 Through poetically organized di scourse .... 189 Through a shift in t ense ... 193 Through a change of pace 194 T hrough a shift in the incidence of ideophones 196 4. 4 Summary 19 7 5 WHY KU !NDA SINGS: SONG IN THE STORY OF SU$ A! T & KU! NDA .. 211 5. 1 One Storytelling Event : Fourteen Stories and Sixteen Songs 211 5. 1.1 Ethnographic Synopsis of the Storytelling E vent 211 5. 1.2 General Narrative Features .. 21 3 5. 1.3 General li! ka!n" Song F eatures .. 214 5. 1.3.1 Performance distribution of songs in stories 214 5. 1.3.2 Discourse functions of songs in stories ... 215 5. 1.3.3 Stylistic features of storys songs .... 216 Musicall y rhythmic features 217 Poetic features of song texts 218 5. 2 Songs Discourse Function s in the Narrative Performance Su$a! t# Ku! nda 220 5. 3 Conclusion: Why Ku! nda Sings .... 228 APPENDIX A TEXT TRANSCRIPTI ON OF SU$ A! T & KU! NDA ..... 244 B MUSICAL TRANSCRIPTIO N OF SU$ A! T & KU! NDA .. 251 REFERENCES .. 262 BIOGRAPHICAL SKETCH.. 280
9 LIST OF FIGURES Figure Page 1 1 Map of central African hunter gatherers .. 40 1 2 Baka camp of Nomedjo 40 1 3 Comparison of Baka and Aka language classifications 41 1 4 Table of Baka consonants ... .. 4 2 1 5 Baka vow el system 42 1 6 Maps : (1) Africa, (2) Cameroon, (3) SE Cameroon Rainforest, Baka Region (4) Abong Mbang Lomi Road .. ... 43 2 1 List of Baka song field recordings .. 87 2 2 Lexical sets of comm on Baka speech act terms 88 2 3 Fundamental categories and features of discourse genres 88 2 4 Composite transcription of li! ka!n" rhythms 89 3 1 Table of common Baka speech act terms ... 122 3 2 Inven tory of Baka song text poetic devices .. 122 3 3 Table of song tempi .. 123 3 4 Table of song meter types .. .. 123 3 5 Textual transcription of a Baka lament ... 124 3 6 Photo of Aye singing song and playing ng"!mbi na! pe! ke ... 127 3 7 Photo of Baka harp zither ( ng"!mbi na! pe! ke ) 127 3 8 Composite of available syntactic verse types .. 128 3 9 Frequency of available verse types 128 3 10 Chart of syntactic development 129
10 3 11 A djacent verses and diagram of allowable verse progressions ...... 13 0 3 12 Diagram of the poetic patterns of Aye$ s Lament 13 1 4 1 Axis graph of the potential perf ormance distribution of speech, song, and music in African oral monologue narrative discourse .... 200 4 2 Outline of discourse analysis of li! ka! n" song in li! ka!n" narrative 200 4 3 Grammatical devices of textual cohesion .. 201 4 4 Semantic devices of textual cohesion .. .. 201 4 5 Types of sung narrative development 201 4 6 Subtypes of sung contextual development 201 4 7 Higgens transcription and translation of Lo! ndo! ( Otter ) ... 202 4 8 Framework of the narrative performance development of Lo! ndo! .. 20 3 4 9 Composite melody of Lo! ndo refrain ... 204 4 10 Summary outline of the function s of li! ka! n "! song in li! ka! n" narrative .. .. 205 5 1 Diagram of the narrative development of the story of Su$a! t# Ku! nda (Leopard and Turtle)..... 237 5 2 Melodic motive of Ku! nda s Song 238 5 3 Motive of Ku! nda s Song restated in melodic sequence betwe en two voice parts .. 239 5 4 Motive of Ku! nda s Song strictly restated rhythmically, but with intervallic substitutions. 240 5 5 Motive of Ku! nda s Song restated with rhythmic contraction .. 241
11 LIST OF OBJECTS: AUDIO RECORDINGS Object: Audio Recording Page 2 1 Audio file of li! ka!n "! formula (clip) .. ... ..... 78 2 2 Audio file of rarer li! ka!n "! rhythm (clip) ......... 82 2 3 Audio file of typical li! ka! n "! rhythm (clip) .......... 82 3 1 Audio file of elephant hunting song (clip) .. 102 3 2 Audio file of song of a young boy playing with his parents (clip) .............. 102 3 3 Audio file of solo song of lament (clip) .. 103 3 4 Audio file of solo song of lament (with clapping) (clip) 103 3 5 Audio file of male initiation song (without accompaniment) (clip) 104 3 6 Audio file of male initiation song (with accompaniment) (clip) ... 10 4 3 7 Audio file of dance song (return form gorilla hunt) (clip) .. 106 3 8 Audio file of song associated with the spirit Jengi (clip) 106 3 9 Audio file of solo lament with harp zither accompaniment ... 108 3 9 Audio file of solo lament with harp z ither accompaniment ... 110 4 1 Audio file of li! ka!n "! story Lo! ndo! ... 184 4 1 Audio file of li! ka!n "! story Lo! ndo! ... 187 4 1 Audio file of li! ka!n "! story Lo! ndo! ... 20 4 5 1 Audio file of rarer li! ka!n "! rhythm (clip) .. 2 17 5 2 Audio file of typical li! ka! n "! rhythm (clip) ... 217 5 3 Audio file of li! ka!n "! story Su$a! t# Ku! nda ... 220
12 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfil lment of the Requirements for the Degree of Doctor of Philosophy WHY KU! NDA SINGS: NARRATIVE DISCOURSE AND THE MULTIFUNCTIONALITY OF BAKA SONG IN BAKA STORY By D aniel Fitzgerald May 2011 Chair: Larry Crook Major: Music The subject of this dissertatio n is 'e! na li! ka!n"! : a genre of song sung in the course of telling a traditional Baka narrative. The Baka people of Cameroon are one of several preliterate people groups in Equatorial Africa. Their culture is primarily oral. Most of what is socially s ignificant is typically marked by oral art, by song in particular. Contrary to the cumulative trope of both the scholarly and popular literature on the music of "pygmy peoples," the Baka do not simply, nor principally, yodel "wordless" songs. They most o ften sing songs with words. As one genre of Baka song, l i! ka! n"! song is both a specimen of the Baka song tradition and the Baka narrative discourse tradition. No other verbal performance phenomenon produces such an entanglement of Baka music and language as does li! ka! n"! song. My dissertation disentangles t he complexes of signs comprising li! ka! n"! song. Its aim is to describe the multiple function s of Baka song in Baka story through an analysis of the semiotic forces that order the complex textual, musical, and contextual interactions of li!ka! n"! music an d language. As a primary thesis of my dissertation, I argue that li! ka!n"! song most relevantly and inherently functions as a discourse feature of li! ka! n"! oral narrative. In particular, I
13 demonstrate that li! ka! n"! song effects both narrative cohesion and narrative development. Narrative cohesion is achieved generatively, performatively, and inter textually N arrative development is contextual and climactic in nature. As social acts li! ka!n"! songs encode multiple socio cultural functions: among them are social cohesion, collective expression, social critique, shared memory, and shared experience. Thus, when the well known narrative character Ku! nda (Turtle) sings, layers of communicative intent are potentially served. Minimally, Ku! nda gives voice to the development of the story's action and conflict. His voice, however, is narrated, that is, it is embodied by the choir of Baka voices present at and participating in the li! ka! n"! 's performance. His song is actually their song. He sings because th ey want to sing. And their reasons for singing Ku! nda s song, in turn, are intimately related to what is sung, how it is sung, when and where it is sung, and who sings it.
14 CHAPTER 1 INTRODUCTIONS TO THE STUDY OF #AKA LI! KA! N%! SONG: 1.1 Research Subje ct, Purposes, Theory, Models, and Motivations The subject of this dissertation is 'e! na li! ka!n "! : a genre of song sung in the course of telling a traditional Baka narrative. The Baka people of Cameroon are one of several semi nomadic, hunter gatherer pe ople groups in Equatorial Africa, commonly called pygmies. Although the Baka are now beginning to learn to farm, their culture is still primarily oral M ost of what is socially significant is typically marked by oral art, by song in particular. (e! na li! ka!n"! literally, "story's song," and glossed as / song / of / story / can denote song sung with any story. 1 More often, however, it connotes song sung during the telling of a traditional Baka narrative, that is, one with traditional characters, setti ngs, and themes. In order to maintain its traditional connotation, for the remainder of the dissertation I will most often refer to 'e! na li! ka! n"! as li! ka!n"! song," not as "story's song." My study of Baka li! ka!n"! song is an extension of my longstan ding investigation of Baka song in general. I first encountered li! ka!n"! song in 1997 during an all night wake in the Baka encampment of Ndjibot. Since that time, I have found that of all the Baka's song traditions, no other genre requires so much focus ed research on the nexus of music and language as does li! ka! n"! song. Indeed, no other verbal performance produces such an entanglement of Baka music and language as does li! ka! n"! song. Fox and Feld (1994) frame the historical trajectories of thought about language and music in terms of four major predications: music as language, language in music, music in language, and language about music ( 26). Any such research approach would be equally
15 applicable in the study of li! ka! n"! song. My purpose, howe ver, is to discover the functions of Baka song in Baka story through an analysis of the syntactic, semantic, and pragmatic "rules" that order the complex textual, musical, and contextual interactions of li! ka!n"! music and language. The theoretical framew ork of this study is based upon the assumption that li! ka! n"! song, if nothing else, is fundamentally a communicative act, a complex human sign. For this reason, I have chosen to approach my subject semiotically, with the specific inductive goal of identi fying the chains of musical, textual, and contextual sign forms, references, and effects that are commonly at play in li! ka! n"! song performance. The explanatory power of semiotics lies in the inherent interdisciplinary capacity of its conceptual terminol ogy to identify and explain the integration of the multivalent and multifunctional interactions of seemingly separate domains of communicative phenomena, whether musical, verbal, or contextual. The interdisciplinary evidence of the fruitful application of semiotics to music, anthropology, and linguistics was first proposed to me in t he scholarly works of Turino (1999 2000 ), Agawu (1991 2009 ), Nattiez (1990), Geertz (1973) and Turner (1967) and Jakobson (1973, 1960) Behind the theoretical strengths of se miotics lies "a happy methodological fault," in that the capacity of semiotics to interpret so many domains of semiosis also runs the risk of generating an unwieldy body of sign vehicles, that is, too much research data for a single dissertation. For this reason, I have chosen to focus this dissertation on the semiotic processes of li! ka! n"! song's verbal signs (as distinct, though not separate, from its visual, tactile, meta physical, or even non verbal auditory signs). This seemingly narrower field of data, however, does not then result in a "thin description" of the
16 research topi c. The discourse centered studies of Bauman (1977), Urban (1982), Sherzer (1987) and Sherzer and Urban (1988) are examples to the contrary. Moreover, to "focus on verbal signs" does not necessarily exclude a thorough investigation of pertinent musical o r contextual signs either. "Thick descriptions" of verbal phenomena abound in many full blown ethnographies of musical performances, as evidenced in the scholarship of Qureshi ( [1987 ], 2006 ) Coplan (1988), Stone (1988), and Waterman (1990) As ethnograp hic models, many of these studies suggest enumerable ways to focus on verbal phenomena without reducing them to "mere words." Instead, potential poetic, musical, and contextual signs are teased out from the performance fabric of diverse speech acts and co mbed for their significance. In the process of disentangling the complexes of signs comprising li! ka! n"! song, it has become apparent that, generically li! ka!n"! song is not simply sung verbal performance, but rather, sung narrative discourse performance. This song, then, is to be taken as integral to narrative, and in no way incidental. Indeed, as a primary thesis of this dissertation, I argue that li! ka! n"! song as a complex sign most relevantly functions as a discourse feature of li! ka! n"! narrative, that is, as one, albeit unique, discourse feature among many. T o deduce and recognize li! ka! n"! song's potential narrative discourse functions, I have most often consulted the analytical models of discourse analysts Robert Longacre (1996), Joel Sherzer (1986), Stephen Levinsohn and Robert Dooley (2001). The methodologies of Longacre, Dooley, and Le vinsohn are he l d to produce analytical descriptions that approximate how discourse is actually produced and understood (Levinsohn and Dooley iii). In short, their research findings are aimed at "application," and thus, are intended to facilitate the com petent communication of actual social acts. 2
17 Kathleen Higgens (1981) and Yves Le! onard (2003), as examples, have based their complementary text linguistic analyses of Baka li! ka!n"! on the analytical models of Longacre (1996) and Levinsohn (2001). They d id so with the expectation that these methodologies would facilitate more apt translations of Baka narrative discourse (Le! onard 1997). Le! onard's forthcoming publication of Baka biblical narratives will offer evidence of the aptness of his analytical mod els. Having adopted and expanded upon these same models of discourse analysis, I hope to complement Le! onard's analysis by innovatively modeling a suitable research method for perceiving, recognizing, and describing the sung narrative performance feature s of Baka li! ka! n"! And like Le! onard, my goal is not merely to objectively describe Baka narrative but also to subjectively participate in it, if only obliquely. That is why I have adopted a semiotic research approach. For as Clifford Geertz has state d, The whole point to a semiotic approach is to aid us in gaining access to the conceptual world in which our subjects live so that we can, in some extended sense of the term, converse with them (1973:24 as cited in Schrag 2005:41). To this end, I have proposed my analysis of li! ka!n"! song to Yves Le! onard as an indicator of how song might function in new Baka narrative translations (forthcoming). While the greater part of this dissertation progressively explains how li! ka! n"! song informs a Baka narrative performance, its pra gmatic (social) functions and effects are of no less importance. I imply as much in the dissertation's title: "Why Ku! nda (Turtle) Sings ." Ku! nda is a well known Baka li! ka!n"! character. In the story of Su$ a t# Ku! nda ("Leopard and Turtle") as na rrated in Chapter 5 Ku! nda sings. In doing so, he gives voice to some of the story's action and conflict. But his voice is narrated; that is, it is
18 embodied by the choir of Baka voices present at and participating in the li! ka!n"! 's performance. His s ong is actually their song. He sings because they want to sing. And as it turns out, their reasons for singing the li! ka! n"! song are intimately related as I hope to show to who actually sings, what they sing, how they sing, and even when and where they sing. 1.2 Research Literature Review Baka language and narrative My own fieldwork encounters with li! ka! n"! song have been consistently informed (and re informed) by the treasure of printed collections of traditional Baka stories transcribed and edited by Robert Brisson (1999), Daniel Boursier (1994), Kathleen Higgens (1981), Christa Kilian Hatz (1989), and Yves Le! onard (2003) (Sec. 4.1). To the study of these bodies of texts, I have also added the study of Brisson's Baka Franais dictionary (1979/1999 2002), Kilian Hatz's study of Baka grammar (1995), Kathleen Philips's Baka phonology (1981), and Kathleen Higgen's preliminary discourse analysis and anthropological studies (1981, 1984). Le! onard's unpublished discourse analysis became available to me in 2004. To all these, I then added my own Baka language acquisition routine, and most importantly, the audio and audiovisual documentation of Baka li! ka! n"! and li! ka!n"! song performances. Of all the documented li! ka! n"! performances mentioned above, i nitially only my personal collection provided both print and non print representations of li! ka!n"! performances in social settings. Arom and Renaud's 1977/1990 recordings of Baka li! ka! n"! only provided thematic synopses of narrative recordings; Brisson, Boursier, and Kilian Hatz generally appeared to record li! ka! n"! in their "natural" settings, but only published written
19 transcriptions of their recordings; and while Le! onard made both his recordings and transcriptions available to me, the "tellings" he elicited were only done so during one on one interviews, that is, without a participating audience. In 2005 Higgens graciously sent me the original audio recordings from which she produced her 1981 text transcriptions (e.g., Sec. 4.3.2). Cameroonian musi c research When I arrived in Cameroon in 1995 as a researcher with the Christian faith based non governmental organization "SIL," I did not initially know what language development project I would serve. As a result, my pre field bibliographic research did not specifically focus on Baka culture, though I was certainly familiar with certain aspects. By 1999, as chronicled in my essay A Profile of Cameroonian Music Research ," I was thoroughly familiar with the specifically ethnomusicological research ca rried out in that region of Africa (1999). The scholarship of four Cameroonian researchers, that is, Samuel Martin Eno Belinga (1965, 1970), Francis Bebey ( 1975, 1982 1995), Abbe! Pie Claude Ngumu ( 1964, 1966, 1972a, 1972b 1976 ), and Fr. Jean Marie Bodo (1992, 1996), was of particular interest to me. Their writings, as well as their personal musical creativity, tellingly signaled a collective interest during their l ifetimes for maintaining their most valued cultural identities, while still fruitfully engaging with the post colonial world. My friendship with Fr. Bodo sealed my own commitment to these same interests. Music of African hunter gatherers My engagement wi th the scholarly literature on the music of African hunter gatherers has generally tended to focus on six to eight authors. I will review them here by summarizing the writings of each and then
20 characterizing them as a whole I will conclude by assess ing a few of the l acunae in this domain of research. I do not consider the writings of Colin Turnbull as music research, strictly speaking. However his 1961 The Forest People, his 1965 Wayward Servants and his 1983 The Mbuti Pygmies: Change and Adaptation h ave served subsequent researchers with much helpful information. His earlier 1961 field recordings, Music of the Rainforest Pygmies are, of course, of infinite historical value but neither he (in his liner notes) nor any others that have followed have offered any substantive interpretive response to th ose recordings, other than perhaps, the oft i dealized trope s about pygmy egalitarianism and naturalness ." 3 Nicholas England, in his brief 1967 article "Bushmen Counterpoint," presented what appears to be the first (albeit brief) music centered study of African hunter gatherers. England presented excerpt ed transcriptions and a brief analysis of five songs of the Z$lu Khuisan people of South West Africa and Bechuanaland. His was the first analysis to posit evidence of contrapuntal techniques in Bushmen vocal music, "the distinctive mark of Bushmen communal music." Ten ye ars later, Alan Lomax's Folk Song Style and Culture (1968) and Cantometrics: An Approach to the Anthropology of Music (1976) also briefly included two other overtly "analytical" approaches to pygmy music, but eventually faded in explanatory power for lack of detailed, sustained ethnography. In 1971, Charlotte Frisbie published the article "Anthropological and Ethnomusicological Implications of a Comparative Analysis of Bushmen and African Pygmy Music." Three descriptive categories profile the music of Afri can Bushmen and
21 Pygmies: instruments, vocal music, and musical styles and structures. Frisbie's comparison of these three profiles suggests a high degree of similarity between Bushmen and Pygmy music, and are allegedly confirmed in other studies by Lomax, Grimaud and Rouget, Grauer, and Merriam. In turn, these similar profiles are said to support anthropologist G.P. Murdock's claim that the two pygmy groups have a common historical region of origin. The first sustained fieldwork on the music of the Aka pe ople of C.A.R. was undertaken by French Israeli ethnomusicologist, Simha Arom. Arom's publications peaked in the early 1990s. The subsequent publications of those whom he mentored at the French C.R.N.S. research institute are still accumulating, though w ith less regularity. The works of Susanne F rniss and Emmanuel Olivier are most notable. F rniss has most recently made research trips among the Baka in Cameroon, though to my knowledge, no print publications have as yet resulted. The bulk of the researc h carried out by Arom and his group at C.R.N.S. focuses on music analysis, that is, analyses of scales, modes, polyphonies, polyrhythm s "hocketing," yodeling, etc.. Contextual studies however, are not altogether neglected as illustrated in Ar om's African Polyphony and Polyrhythm (1991), and Olivier's "The Ju | hoansi Bushmen's Conception of their Musical World" (1998). Still, ( cognitive ) musico structural analysis is clearly their research emphasis. The following three paragraphs briefly ind icate the scope and method of their research. Arom's 1978 audio recordings, Centrafrique: Anthologie de la Musique des Pygmees Aka document thirty two audio music events of the Aka Pygmies in southwestern Central African Republic. The recordings are acco mpanied by a thirty two page
22 booklet of his commentary on Aka life and music, and more than half of that commentary is given to ethnographic description of the recordings. Arom's article "The Use of Play Back Techniques in the Study of Oral Polyphonies" ( 1976) posits that all music traditions are coherent systems of cognition, and in order to analyze the constituent elements of a music system, the transcription of music is held as necessary to the analytical process. To facilitate the analysis, especially of multi part textures, his innovative "play back" recording technique is explained and applied. Similarly, i n his article "Modlisation et Modles dans les Musiques de Tradition Orales" (1991), Arom posits that the scientific method can and has demonstr ated the existence of coherent systems of music cognition in oral societies, and that these, therefore, can be theorized, generalized, and taught. Arom's African Polyphony and Polyrhythm: Musical Structure and Methodology (1991) is his most well known stu dy. It is not, however, a study of pygmy music exclusively, but of the polyphony and polyrhythm found in much of central Africa. This book length work is primarily a synchronic study aimed (again) at the discovery of the underlying principles of musical f orms. Broader contextual elements are intermittently addressed as well. Arom's research theory, methodology, and analysis are explained along with 450 music transcriptions as supporting evidence. The resulting analysis claims sufficient findings to be a ble to generate "new" versions of central African music, which "traditional musicians are likely to find acceptable." Susanne Frniss's "Recherches Scalaires chez les Pygmes Aka" (1991 a ) gives an account of the research process that Frniss undertook in d iscovering the Aka Pygmies's use of "une chelle pentatonique anhmitonique bas sur l'agencement vertical des intervalles en certains points prdetermins." Frniss's article "La
23 Technique du Jodel chez les Pygmes Aka (Centrafrique): tude Phontique et Acoustique" (1991 b ) presents a phonetic and acoustic study of the vocal technique commonly called "yodeling" as performed in the familiar polyphonic vocal music of the Aka Pygmies. Her acoustic and phonetic analysis of Aka song recordings, complete wit h charts, music transcriptions, and "sonagrammes," argues that the most striking characteristics of yodel execution and timbre result from two alternating articulatory mechanisms, sung with specific vowels, on particular constitutive degrees of a scale sys tem. Later, i n "Rigueur et Libert: La Polyphonie Vocale des Pygmes Aka" (1993) Frniss explains how process es of thematic variation generate variations on the governing melodic pattern of one polyphonic Aka pygmy song, illustrating "the phenomenon that Aka music cognition is essentially polyphonic." Her 2006 article, "Aka Polyphony: Music, Theory, Back, and Forth," takes much the same approach. In 1996, Emmanuelle Olivier and Suzanne Frniss revisted the question of the relationship of pygmy music and Bushmen music in their article "Musique Pygme/Musique Bochiman: Nouveaux lments de Comparaison." The y conclu de that both music traditions are "multi voiced," though only pygmy music is strictly polyphonic, while Bushmen 's is monophonic. Yet a s simila r as these two traditions appear, they are said to be conceived of differently by their practitioners. Two years later, i n 1998, Olivier published a more contex t oriented study in the article "Ju | hoansi Bushmen's Conception of their Musical World." In i t, Olivier profiles the local meanings assigned to the interrelated musical and social entities of the Ju | hoansi Bushmen music Two broad genres of song are then posited: "healing songs and non healing songs
24 In 1995, around the same time that the C. R.N.S. group was publishing so prodigiously Louis Sarno published both Bayaka: The Extraordinary Music of the BaBenzele Pygmys and the Sounds of Their Forest and Song from the forest: my life among the Ba Benjell Pygmies In both works, Sarno chronicles his sustained experience among the Bayaka/Ba Benzelle Both accounts are more journalistic than scholarly in style Bayaka in particular, is part diary, part ethnography, part photo essay, social commentary, and multimedia presentation. T he oft ideali zed narrative of pygmy life is prevalent. I n sharp contrast with both the musico structural research orientation of the C.N.R.S. research group and the idealized accounts of Sarno, Arom, Lomax, and others, Michelle Kisliuk in her 1991 dissertation "Confron ting the Quintessential: Singing, Dancing, and Everyday life among the Biaka Pygmies" challenges "the utopian myth" of African egalitarian societies Kisliuk's reflexive ethnographic analysis strikingly shifts the scholarly research focus from decidedly m usical questions to socio musical questions. Her ethnography highlights the social negotiation of gender tensions as performed in a particularly popular Biaka dance, a dance controlled by Biaka women. M usic transcription and analysis is secondary to Kisl iuk's socio cultural emphasis A similar ethnographic approach is found in many of her subsequent publications (1997, 1998a, 1998b, and 2000). Her book Seize the Dance (a reworking of her dissertation material) is regularly cited in ethnomusicological li terature. Kisliuk's ethnographies regularly address issues of gender, identity, performance, and reflexivity. Her 1997 essay "(Un)doing Fieldwork: Sharing Lives, Sharing Songs" further underscores her research journey' as she narrates her concern f or the identity
25 of the field researcher the challenge of writing about field experience, and the problem of interpreting the relevance of those experiences. It is said by some that Kisliuk's reflexive writing style runs the risk of self indulgence, that readers learn more about her than her research "object" (Agawu 2003: 214 218). My perspective on Kisliuk's reflexive description is not the same as most readers. Having lived among the Baka in Cameroon for a similar period of time, I am not as much at a loss as a typical reader when it comes to imagining what it feels like to dance, sing, play, talk, get angry, cry, and laugh with African pygmy peoples. Thus, t he initial impact and attraction of her personal narrative for me is blunted. However, the fundamenta l point of her thick sociological representation is certainly not lost: that is, Aka or Baka or Mbuti music does not take place in a vacuum or on audio recordings, or in musical transcriptions but in Aka Baka or Mbuti communities. And to the degree th at anyone might want to participate relevantly in such music making, is the degree to which they would need to gain an adequate understanding of the social codes and cues that permeate these musical performances. Among all the scholars representing the mus ic of African pygmy peoples, the C.N.R.S group and Michelle Kisliuk are the only authors who have published book length studies. One publication privileges "the music itself," while the other privileges the music's c ontext; and b oth focus on a single gen re M y field recordings suggest that there are still many more genres and contexts yet to be described fruitfully (Sec. 1.6 and Fig. 2 1). This dissertation, therefore, will describe yet another, though not another study from the Aka music tradition. T o my knowledge, this will be the first book length study of a Baka music tradition.
26 Poetic analyses There are other lacunae as well. One of the most conspicuo u s lacunae in the research literature related to this region of Africa and of hunter gatherer peoples in general is in the area of poetic analysis of song. P oetic analysis is all too often reserved as the property of literary criticism, not linguistics and certainly not ethnomusicology But if my analysis of li!ka! n"! song is to begin to penetrate the "interpenetration" of sound and sense in music and language, an analysis of song text poetics is obligatory. The primary guide in my research of song text poetics is loosely modeled after linguist Roman Jakobso n s well known semiotic principle of equivalenc e" (1960, 1966, 1973; Sec. 3.1). To this principle, I have added Banti and Giannattasio's notion of "poetically organized discourse" (2004; Sec. 3.1 and Sec. 126.96.36.199). Numerous "lesser" poetic concepts have also found there way into my analysis, some from structural linguistics and literary criticism ; others from ethnopoetics, folkloristics, orality studies the ethnography of speaking and ethnomusicology Of the scholarship from these areas that have marke d me the most, the writings of Frisbie (1980) Feld (1982), Coplan (1988) Kindell (1996), Finnegan (1977, 1992, 2007), and Feld, Fox, Porcello, and Samuels (2004 ) have been most helpful. Oral narrative research The study of traditional oral narratives w orldwide and in Africa in particular has steadily increased over the last seventy five years (e.g., Malinowski 1922; Parry 1928; Lord 2000; Eno Belinga 1965; Levi Strauss 1969; Biebuyck and Mateene 1969; Dorson 1972; Abrahams 1972 1983 ; Scheub 1977, 1996, 1998, 2002; Opekw h o 1979, 1983, 1992; Foley 1987, 1988, 2002; Johnson, Hale, and Belchner 1997 ; Finnegan 1967, 1977, 1992, 2002, 2007 ). Similarly, though only
27 intermittently and in widely varying degrees of ethnographic detail the last twenty five year s of ethnomusicology's fruitful interdisciplinary dialogue with linguistics, anthropology, sociology, and literary studies have progressively drawn attention to the role of music and song in the narrative discourse of oral traditions (e.g., Ben Amos 1975; Coplan 1987, 1988, 1994; Stone 1988; Yung 1989; Kisliuk 1991 ; Opland 1983; Agawu 1995; Charry 2000; Seeger  2004; Stock 2006). My dissertation on Baka song in Baka story seeks to build on the momentum and collective understanding of these earlier studies, though it particular ly strives to describe, analyze, and interpret in greater detail the nexus of the interplay of a narrative 's sung multivalent multifunctional musical, textual, and contextual signs. 1.3 Description of Chapters The formal dev elopment of my dissertation is organized in five chapters. I began the present chapter (Chapter 1) by introducing my research subject, purposes, and motivations. Concomitantly, I have described the constellation of theoretical and methodological approach es that have most stimulated, and constrained, my research. The interdisciplinary nature of this study was reflected in the preceding review of the scholarly literature pertinent to Baka language and narrative, Cameroonian music research, the music of Afr ican hunter gatherers, poetic analyses and oral narrative research In the remainder of this chapter I will shift from introducing my research approach to introducing my research setting In Sections 1.4 and 1.5, I will describe select demographic, geog raphic, historic, linguistic, social, and cultural generalities about the Baka people themselves. And in Section 1.6, I will sketch the setting of my
28 research interactions with the Baka, and conclude with an outline of the scope of the research data. In C hapter 2, entitled "Baka Verbal Performance in Song, Story, and Story's Song," I begin my argument that li! ka! n"! song is not so much "musical" performance as it is "verbal" performance, or rather, verbal performance that is particularly poetically, musica lly, narratively, and socially "marked." Thus, as a type of verbal performance, li! ka!n"! song not only shares characteristic marks of both sung and narrated verbal performance types, but also bears distinctive song and story marks of its own. In Section s 2.2 and 2.3 under the headings of "Poetic Verbal Performance" and "Narrative Verbal Performance" I frame Baka song and story as particular verbal performance types as well. Subsequently, in Section 2.4 entitled "Sung Narrative Verbal Performance" I spec ify li! ka!n"! song as a unique sub type of both story and song. In Chapter 3, entitled "Baka Song text Poetics," I expand on Section 2.2 of Chapter 2 ("Verbal Poetic Performance") by setting out a detailed inventory of Baka songs' poetic devices. Section 3.3 covers Rhythmic Poetic Devices; Section 3.4.3 introduces Syntactic Poetic Devices; Section 3.4.4: Lexical Devices; Section 3.4.5: Phonological Devices; and Section 3.4.6: Semantic Devices. Section 3.5 concludes by introducing the potential effects of the cumulative patterning of these poetic devices. Chapter 4, entitled "Baka li! ka! n"! song as a Discourse Feature of Baka Story," expands on Sections 2. 3 and 2. 4 of Chapter 2 ( i.e., "Narrative Verbal P erformance" and "Sung Narrative Verbal Performance ," respectively). I do so by tackling two fundamental questions regarding li! ka !n"! song: that is, "Why do Baka stories include song?" and "What is the relationship of Baka songs to Baka stories?" The introductory
29 section (Section 4.1, "The Performance Relationship of Baka Story and Song") explores the potential "performance integri ty" of song and story, not just in Baka narrative performance, but throughout Africa. Are songs incidental to story, or intrinsic? In what ways and to what degree? In Sections 4.2 and 4.3, I progressively argue that li! ka!n"! song is intrinsic to li! ka! n"! story, both as a signifying force of narrative cohesion and narrative development In Section 4.2, I explain that li!ka! n"! song effects narrative discourse cohesion generatively, performatively, and textually Then in Section 4.3, I demonstrate how li! ka! n"! song signifies both contextual development and climactic development. In the final chapter (Chapter 5, entitled "Why Ku! nda Sings") I finally examine the complex of symbolic, iconic, and indexical signs of song musical signs, textual signs, and c ontextual signs in a single li! ka!n"! narrative performance: Su$ a t# Ku! nda ("Leopard and Turtle"). Section 5.1 of Chapter 5 situates the performance of Su$ a t# Ku! nda as one narrative performance among thirteen performed one evening in the Baka cam p of Ndjibot. In Section 5.2, I narrate the performance of that performance, that is, I describe the particular musical, textual, and contextual song features and their narrative functions that progressively unite, develop and distinguish the story of Su$ a t# Ku! nda I then highlight the fact that songs in Su$ a t# Ku! nda are sung by the narrative character Ku! nda (Turtle). Ku! nda 's voice, however, is actually performed (enacted, embodied) by all in attendance that evening. As a result, layers of co mmunicative intent and effects are simultaneously signified. And "why Ku! nda sings" is set out in the conclusion.
30 1.4 The Baka People 1.4.1 Designations Pygmy peoples are often spoken of as if they were culturally homogen e ous. This, of course, is not true; and the failure to recognize the distinctives of each group is an unfortunate hindr a nce to an adequate understanding of the unique identities and experiences of each group. Traditionally, t he Baka are one of several semi nomadic, hunter gatherer peop le groups in Equatorial Africa, commonly called "p ygmies ." 4 The generic designation pygmy (apart from how it may be politically construed) recognizes certain phenotypical and cultural similarities among these people groups. 5 Figure 1 1 plots approxim ate locations of these groups under seven commonly held ethnonyms : the Mbuti, Twa, Aka, Bongo, Baka, Tikar, and Bagieli Pygmy identities, however, are further distinguished by more than twenty five designations, each reflecting the perceptions of any numb er of local or academic communities. 6 The Baka themselves are variously refered to as Bangombe, Bebayaka, Babinga, Bibaya, and Baka. Baka is the designation most widely used. 7 The people in the Baka camp s of Ndjibot and Nomedjo where I have lived and wo rked call themselves and their language, Baka (Fig. 1 2) 1.4.2 Demographic, Geographic, and Historic Generalities There are an estimated 250,000 to 350,000 African pygmies 30,000 to 50,000 of which are Baka (Survival 1998:1, 3) The Baka are spread t hroughout the adjoining rainforests of Cameroon, Gabon, and Congo, the largest concentration living in Cameroon. The Baka form the largest of Cameroon's three pygmy groups the Tikar
31 Bagieli and Baka (Fig. 1 1 ). The Baka encampment of Ndjibot has an ave rage population of three hundred to four hundred people; the camp of Nomedjo is nearly twice as populated A synthesis of current studies tends to link the Baka with the Aka 8 and Mbuti pygmies, biologically, linguistically, and historically. 9 The emergin g hypothesis suggests that approximately 1000 AD a common ancestry of these three groups migrated west from the Ituri Rainforest (currently located in eastern Democratic Republic of Congo). Those that remained in the Ituri became known as the Mbuti people ; those that settled in western Central African Republic became the Aka"; and those that finally reached southeastern Cameroon emerged as the Baka (Duke 2001:8 13; Ba h uchet and Thomas 1986:90; Thomas 1991:19; Ba h uchet 1993b:49). 1.5 Baka Language 1.5 .1 Language Classification For all that may point to how similar the Baka, Aka, and Mbuti people groups are, it must be remembered that they do not speak the same language The Baka and Aka, for example, may live in relatively close proximity and even sha re a significant number of forest terms 10 but the Baka speak a Ubangian language, while the Aka speak a Bantu language. Thus, while both languages spring from the Niger Congo language family, they are still removed from each other by at least five languag e family stratums (Fig. 1 3 ) Baka and Aka, then, are as far removed from each other as, say, Italian and Russian 11 The linguistic bases of Baka and Aka, then, are critically different and therefore do not have the same phonological, grammatical, or proso dic resources to poetically
32 exploit. So as tempting and common as it is to cast the Baka and Aka in the same mold, their songs (sung in two different languages) are not mutually intelligible. Even if by some miracle their texts suddenly became mutually intellig i ble (propositionally), the poetic and rhetorical effect of those texts would still be of no effect as neither party would be familiar with the patterns of sounds, structures, and senses poetically at play; and the loss of a lyric's poetic effect is no lyric at all. 1.5.2 Linguistic Structures The poetic function rests upon the poetic exploitation of fundamental linguistic elements of a received language tradition. Poetic devices depend upon the perception of distinctive linguistic features for their poetic effect. Perception of the following basic linguistic features of the Baka language is prerequisite to the perception of Baka poetic devices. 188.8.131.52 Phonology P honemes: consonants, vowels, tones, and syllables The raw materials of phonolo gical level poetic devices ( e.g., alliteration, consonance, assonance, homophony, and meter as in Sec. 3.2 ) are taken from a language's distinct phonemic and phonological features. Baka has a basic inventory of thirty two phonemes : twenty five phonemic c onsonants and seven phonemic vowels. Figure 1 4 presents a table of the Baka consonants. Baka's twenty one phonemic vocalic sounds are derived from seven phonemic short vowels as shown in Figure 1 5 Each vowel may take three tone heights 12
33 Baka is a t one language with a register tone system of three level tones: high, mid, and low. Examples 1 3 illustrate the effects of these three phonemic tones on a single morphemic segment ( la ). Example 1: a high tone (/ a! /) on / la! / is glossed as catfish.' Ex ample 2: an (unmarked) mid tone (/ /) on / la / is glossed as sleep.' Example 3: a low tone (/ a" /) on / la" / is glossed as who.' There are two Baka syllable types; both carry one obligatory vowel and one tone. The smallest dynamic syllable consist s of a single vowel (V), while the second type carries a consonant onset (CV). As a result, with the exception of ideophones, some proper names, and a few Bantu loan words, no Baka word may contain a consonant cluster and all Baka morphemes end in a vowel (Phillips 1981:19; Killian Hatz  2004:8 9). 184.108.40.206 Morphology and syntax Syntactic poetic units are composed of grammatical phrases (Pinsky 1998:28, 34). Baka grammar, as outlined by Killian Hatz, employs four basic grammatical structures. From smallest to largest, they are ( 1) the nominal phrase, ( 2) non verbal clause types, ( 3) the simple sentence, and ( 4) complex sentences. A Baka simple sentence has an SVO structure with the direct object competing with an indirect object for the slot follow ing the verb (Kilian Hatz 9 14). In addition to these four basic structures, Baka grammar is marked by three striking linguistic devices: vowel lengthening ( e.g., Example 4; see also Sec. 3.2 ), reduplication ( e.g., Example 5; see also Sec. 3.2) and ideop hones ( e.g., Example 6; see also Sec. 3.2 ) (Kilian Hatz 37 40)
34 Example 4: vowel lengthening: of the verb jo ( to eat' ) : i nnnnnn + _d" !!!` # !!!!!in !!o`#l#! eat 3.SG.H IMPV eat.PRES wild boar He's taking a long time to eat the wild boar (Kil ian Hatz 40) Example 5: reduplication: of the verb f ( to go' ) : _d" !!!` # !!!! f f f !!!f !!!!!!f# 3.SG.H IMPV go.PRES go.VN go.VN go.VN go.VN He walks on and on..... (for a long time) (ibid.) Example 6: ideophone ln n # i#&> jo#g 2.SG PF V hear.PRES.ITR kp!"!h! Can you hear? _jo!g& (= a twig breaks)! (ibid.) These fundamental grammatical structures and devices, then, also become the building blocks of syntactic level poetic constructions (e.g., strophes, stanzas, phrase repetitions, par allelisms, enjambment ( see also Sec. 3. 2). 1.5.3 Semantics Semantically, Baka is marked by "extreme polysemy." Morphemes with ten or more functions are by no means rare, but rather characteristic of the language (Kilian Hatz 1, 170). 13 The particle / t # / +" enq" dw`lokd+" is defined as follows in a lexicon entry by Brisson & Boursier: preposition. With 1. co ordination, 2. means, 3. reference to the antecedent (relative), 4. benefactive, 5. locative, 6. factitive, 7. pronominal form. (Brisson & Boursier 1979: 453; Kilian Hatz 69) Kilian Hatz suggests as many as nineteen functions for the grammaticalization of / s / as the comitative preposition "with" (70). She concludes, This multiple synonymy does not follow necessity, but rather has a different motivatio n the creativity of the speaker who wishes to achieve stylistic variation (emphasis added, 170)
35 Kilian Hatz's diachronic analysis further proposes that Baka language change is not so much driven by phonological convenience, as by a progression of seman tic variations, [a] creative act which gives rise to a polyfunctional chain [that] takes place in tiny steps by way of metaphoric metonymic shifts in meaning through expansion of context. (emphasis added, 3) Stylistic variation, through metaphor and me tonomy, are, of course, common elements of poetic expression. To find them in elemental grammatical strategies like polysemy suggests that the Baka language tends by nature to ward poetic expression. Thus, metaphor and metonymy function not only diachron ically, as fundamental grammatical devices at work in Baka language change, but synchronically, as semantic level poetic devices in Baka song texts (Sec. 3. 2). Such stylistic strategies intensify, however, when perceptions move from non poetic to poetic structures. 1.5.4 Sociolinguistics: Language Change Baka language, then, like any living language, is a language in process. In addition to internal causes of language change, as demonstrated in the dynamics of polysemy, certain external forces are also at work effecting change. The most obvious external dynamics of Baka language change are Baka Bantu relations, modernization, sedentarization, and deforestation. Their effects on songtexts are most readily seen in new lexical and thematic material. Muc h lexical change is born of Baka Bantu relations. Baka encampments are spread throughout the vast adjoining rainforests of Cameroon, Gabon, and Congo. This loose web of Baka speaking communities is in contact with more than twenty five
36 Bantu language gro ups. 14 Each camp may be in vital everyday contact with as many as one to three neighboring Bantu communities. Baka camps commonly borrow Bantu words from their nearest Bantu neighbors. Some borrowed words find only limited local use, while others become more widely and commonly used. 15 Some then, enter song texts, and thus the Baka lexicon changes. The Baka camps represented in this study lie within three to five kilometers of a Bantu speaking village: the Baka speakers in Ndjibot for example, are fla nked by Makaa and Mpoumpoum villages; and the Baka in Mbalam and Menzo are neighbors to Badwe'e speaking villages (see maps in Fig. 1 6 ). Many of my text transcriptions reflect this borrowing Further lexical change is brought on by the spread of the nati onal language that is, French French is slowly coming into use among Baka speakers through modern educational, political, commercial, spiritual, medical, and technological institutions. Daily speech and songtexts reflect an ever increasing use of Frenc h loan words, not to mention more subtle changes in Baka phonology, syntax, and semantics. Through these "modern" institutions, not only does Baka language change, but also Baka experience As a result, new thematic materials gradually enter Baka songtex ts. 16 17 While sedentarization and deforestation may not so much effect grammatical change, they do, like modernization, effect changes in cultural themes. The gradual shift from a lifestyle of hunting and gathering to farming, from a life in the forest to a life in a village is a fundamental change of experience. The themes of everyday conversation, as well as the themes of newly composed songtexts, reflect this experience. 18 But s ongtexts in any culture are characteristically marked by what has
37 already be en or what may very well become the most highly valued common experiences of a community. 1.5.5 Section Summary In summary, Baka song texts are fundamentally informed by numerous anthropological, linguistic, and sociological forces some of which may be summarized as follows: A nthropologically, Baka song texts are thematically marked by the Baka people's particular experience as semi nomadic, pygmy ," hunter gatherer forest dwellers of equatorial Africa. Linguist i cally, Baka song texts are deeply marke d by the typological features of Ubangian languages as well as the distinctive features of the Baka language. For instance, p honologically, Baka is distinguished by a phonemic inventory of t h irty two phonemes (twenty five consonants and seven vowels), a register tone system of three level tones (high, mid, and low), and two syllable types (V and CV). And s yntactically, Baka is primarily characterized by four basic grammatical units: the nominal syntagm, non verbal clause types, the simple sentence (with an SVO structure), and complex sentences and secondarily characterized by three other striking linguistic devices: vowel lengthening, reduplication, and ideophones. S emantically, Baka is marked by extreme polysemy. S ociolinguistically, the Baka language is undergoing obvious lexical change (through lexical borrowing, in particular) due to dynamics of Baka Bantu relations and modernization. Literary themes are changing as well, due to modernization, sedentarization, and deforestation. 1.6 Research Data Sample Most of t he ethnographic data on which this dissertation is based e.g., audio and audio visual recordings, text transcriptions, music transcriptions, photographs and
38 fieldnotes is derived from seven years of sustained fieldwork among Baka speakers in the southeastern rainforest of Cameroon. During my initial period of fieldwork, f rom November 1996 to May 1998 I documented more than four hundred field recordings of Baka song performances These performances took place i n five Baka camps scattered along 175 kilometers of the Abong Mbang Lomie Road in the Eastern Province of Cameroon (Fig. 1 6, Map 4 ). During my fieldwork my family and I first lived in the Baka camp of Ndjibot (Fig. 1 6, Map 4, B). The research data cited in Chapter 5 is exclusively drawn from music performances in Ndjibot (Sec. 5.1). The data for Chapters 2 and 3 was documented from performance events in the camps of Ndjibot, Mayos, Mbalam, Malen/Menzo, and "le Bosquet" (Fig. 1 6, Map 4; see also, Sec. 3.1). Two years later, my fa mily and I relocated to the Baka camp of Nomedjo 100 kilometers south of Ndjibot and resided there from August 2001 to June 2004, and (intermittently) from August 2005 to June 2008 (maps in Figure 1 2 and Fig. 1 6, Map 4, F). 19 During our sojourn in Nomedj o I added and documented numerous additional field recordings and ethnographic metadata. The research data for Chapter 4 is drawn from my fieldwork in Ndjibot and Nomedjo, though greatly augmented by extensive research documentation from Pe" res Robert Brisson (1999) and Daniel Boursier (1994), and linguists Yves Le! onard (2005), Kathleen Higgens ( 1981), and Christa Kilian Hatz (1989). Collectively, the recordings, transcriptions, and analyses of these five researchers represent sung narrative performances of Baka li! ka! n"! that span some thirty years (i.e., 1973 2005) and take place in regions of the equatorial rainforest as near as 100 kilometers in proximity, or
39 as far as 400 kilometers (Map 3 in Fig. 1 6). This sample, then, represents the most geographically, historically, and socio linguistically diverse data sample possible from known existi ng documentation of Baka li!ka! n"! performance.
40 Figure 1 1. Map of central African h unter g atherers 20 Figure 1 2. Map of Baka camp of Nomedjo 21 Nomedjo
41 Figure 1 3 Comparison of Baka and Aka language classifications 22
42 bilabial dental alveolar palatal velar labio velar phary ngeal glottal nasal m n # plosive b t d k g kp gb fricative w s j h 23 affricative dz prenasalized mb nd ndz ng !gb lateral l trill (r) 24 implosive $ % Figure 1 4. Table of Baka cons onants (Kilian Hatz  2004:6; Brisson and Boursier 1979:VI; Le! onard 2009:1) Note: Transcriptions of Baka texts (depending upon the transcriber) variously represent [ % ] as [ny], [ & ] as [ng], [ ] as [p] or [f], [j] as [y], [dz] as [j], [ & ] as [ng], [ & gb ] as [ngb], [ndz] as [nj], [ ] as [ '] or  or ['] Front Central Back Closed i u Half closed e o Half open & Open a Figure 1 5. Baka vowel system (Brisson 1979:V)
43 Figure 1 6. Maps: [Map 1] Africa, [Map 2] Cameroon, [Map 3] SE Ca meroon Rainforest, Baka Region: (B) Abong Mbang, (C) Ndjibot, (D) Mayos, (E) Mbalam, (F) Malen/Menzo, (G) Nomedjo, (H) Lomi, (I) le Bosquet, (J) Yokadouma, (K) Salapoumb, (L) Moloundou, [Map 4] Abong Mbang Lomi Road: (A) Abong Mbang, (B) Ndjibot, (C) Ma yos, (D) Mbalam, (E) Malen/Menzo, (F) Nomedjo, (G) Lomi, (H) le Bosquet
44 1 Grammatically, na (of') is more a genitive marker than a benefactive or possessive marker (see Kilian Hatz 1995 translated in 2004 into an unpublished English version by Yves Le! onard 2 Longacre, Levinsohn, Dooley, Higgens, and Lonard have been or are curently members of SIL. Examples of SIL's research applications can be accessed through their Website at http://www.sil.org/ 3 Much of the same could be said of Grimaud and Rouget's earlier 195 7 recordings. 3 The Baka are primarily hunter gatherers, gradually learning to farm. 4 For further explanation and discussion of the term "pygmy" see Bahuchet (1993c), Kisliuk (1998:6), and Duke (2001:5). 5 See "The Index of Pygmy' Groups" in Forest Foragers of T ropical Africa by Survival for tribal peoples as compiled by Virginia Luling and Justin Kenrick, 1998. 6 C f. Grimes, Barbara F. 2000. Ethnologue: Languages of the World 14 t h Edition. Dallas: SIL International, p. 30 and "The Index of Pygmy' Groups" in Forest Foragers of T ropical Africa by Survival for tribal peoples as compiled by Virginia Luling and Justin Kenrick, 1998. 8 Variously referred to as Bayaka, Baaka, Ba Aka, B iaka, Ba MbenzeleBa Banjelle, Bambenga, Babenga, Benga (Duke 2001:10) 9 (For easier access to sources, please reread as) A synthesis of current studies tends to link the Baka with the Aka and Mbuti pygmies (Duke 2001:8 13) biologically (Cavalli Sforza 198 6), linguistically (Ba h uchet 1989:54, 1993b:39, 48; Thomas 1991:19 20f; Kilian Hatz [ 1995 ] 2004 :4 5 (in English translation); Ruhlen 1987) and historically (Ba h uchet and Thomas 1986:90; Ba h uchet 1993b:49; Thomas 1991:19f). 10 See Bahuchet 1993b:39 and Duke 2001:45. 11 Further study might be given to describing how Baka song texts do (or do not) reflect and exploit the distinctive features of Ubangian languages. 01 See Phillips 1981:23; Kilian Hatz [ 1995 ] 2004 :6 8. 13 See / pe / "back" and / na / "of" for other exemplary polysemic morphemes (see Kilian Hatz 2004: Sec. 3.8.1 and Sec. 3.10.3 na'). 03 Grimes, Barbara F. 2000. Ethnologue: Languages of the world 14 th Edition. Dallas: SIL International, pp. 17, 18, and 20.
45 04 See also, any of Brisson's lexicon e ntries marked (K), i.e., Kak"! (non Baka') (1979). 05 On the other hand, old themes may be re contextualized or disappear all together. 06 For example, my field recoridngs #A082097I II (Catholic Church songs); #A113097A (plucked ideophone songs, one about the governmental sedentari zation program); #A102497 (children's game songs wherein school is mentioned); #A042197 (aita accomp. Lomie Catholic church song); #A040397 (youth dance song : one about village rivalries). 07 Ibid. 19 In Nomedjo I acquired a level 3 ILR (FSI) Baka lang uage proficiency. 1/ Map adapted from Lisa Silcock's Baka: People of the Rainforest Chanel 4 Television, London, 1988:25. 10 Ibid.. 22 I have synthesized this classification from a number of sources, most notably Kilian Hatz 2004, Duke 2001, Grimes 2000, an d Bender Samuels 1989. 12 Kilian Hatz and Le! onard add the pharyngeal fricative [h] which occurs mainly in ideophones and borrowed words, also marginally as an initial consonant. 13 Kilian Hatz adds the alveolar trill [r] which occurs in ideophones and loan words which have become part of the lexicon of Baka.
46 CHAPTER 2 BAKA VERBAL PERFORMA NCE IN SONG, STORY, AND STORY'S SONG "The relationship of music to language is an enormously broad area of research" (Feld and Fox 1994:26). Feld and F ox reference 379 research works in their 1994 review of the English literature on interdisciplinary studies of the relationship of music and language. "Ethnomusicological surveysand substantial musical, linguistic, and literary dissertationsindicate how this vast interdisciplinary literature links research in musicology, acoustics, linguistics, literary studies, philosophy, psychology, and anthropology, and continues to inspire conferences, symposia, and research across these disciplines" (ibid.) T he b ibliography of my dissertation moreover, suggests that the subject of music and language is of no less interest to researchers today than it was sixteen years ago. Two years after Feld and Fox's review, I began my fieldwork among the Baka people of Camero on. A t that time, however, I had no explicit intention to focus my research on the nexus of Baka music and language. My interests then generally lay in musico structural phenomena, melody in particular. Eventually, however, as my collection of early fi eld recordings grew, a surprisingly verbal dimension of Baka music became apparent. My awareness of this dimension of Baka music was first triggered by the elemental discovery that no matter how many field recordings I collected most recorded songs remai ned a n enigma to me insofar as I remained ignorant of what semantically speaking the Baka were singing about. F or to my surprise, the Baka never seemed to just make music"; they sang songs songs with words. Contrary to what I had earlier and passively c ome to expect from the cumulative trope of both the
47 scholarly and popular literature on the music of pygmy peoples the Baka did not simply nor principally yodel wordless songs. To be sure, t he most striking features of Pygmy music include the ofte n wordless yodelling that results in disjunct melodies, usually with descending contours and densely textured multi part singing (Cooke and Kisliuk 2008:1) Yet, yodeled song styles are but one among many other styles. And more importantly, for the Baka music tradition in particular, there are numerous other stylistic aspects of their music tradition that are no less relevant to them depending upon the performance context they serve. I n as little time as the first three months of my fieldwork I also bec ame aware of the enormous variety of Baka music styles and settings The breadth of song styles represented in Simha Arom and Patrick Renauds 19 77 recordings of Baka music should have adequately prepared me for this variety (1990). B ut the prevailing W orld Music trope of the yodel like, hocket style choral polyphony sung by pygmy women as well as my own initial fascination with that vocal style predisposed me to expect otherwise. Eventually, however, my own field recordings revealed a broader repert oire Figure 2 1 lists all of the song performance styles that I have thus far documented with audio or audio visual media. The maps in Chapter 1, Sec. 1.6 indicate the geographical locations of these recordings; Chapter 3, Sec. 3. 1 summarizes their att ending cultural themes, composition histories, and performance practices. 1 I do not intend to imply any kind of song taxonomy in this list (i.e., Fig. 2 1) However, t he Baka terms for these groupings of song performances do generally seem to identify groups of song compositions according to similar constellations of stylistic and/or contextual features. That said, the list is still only intended to provide a n
48 introductory impression of the stylist i c breadth of the whole of their song tradition. More pertinently, the listing also introduces li! ka! n"! song, the subject of this dissertation, as one among many other song performance types. Li! ka! n"! song is listed third in the table. As indicated, the social context of li! ka!n"! song is most often a fun eral or a hunt. Occasionally, it occurs during a divination ritual, or even during evening recreation. What is not explicitly indicated in the table is that the typical verbal discourse context of li! ka! n"! song is oral narrative discourse (see also, Se c. 4.1.2) Li! ka! n"! song, then, is both a specimen of the Baka song tradition and the Baka narrative discourse tradition. As such, it exhibits certain characteristic verbal performance elements of both song and story, while simultaneously distinguishing itself from other Baka song genres and narrative discourse features. The aim of the rest of this chapter is to introduce some of the more salient generalities of all three verbal performance types songs, stories, and stories' songs in order to contextual ize the more detailed descriptions of particular Baka song, story, and story song features presented in Chapters 3, 4, and 5. 2.1 Baka Verbal Performance To reiterate, Baka song, story, and story's song are types of Baka verbal performance (Bauman 1992). Ethnographies of verbal performance commonly comprise a complex of research approaches, drawing most often on research practices in performance theory, the ethnography of speaking, discourse analysis, and ethnopoetics (Finnegan 1992: 29 52). From a macro level research orientation, performance theory recognizes verbal performance as but one cultural performance
49 type within a continuum of performance types (cf. Herndon and Brunyate 1975; McLeod and Herndon 1980). As a mid level research orientation, the e thnography of speaking specifies the particularities of distinctive speech acts. The most fundamental distinction is between everyday speech patterns (the default pattern) and performed speech patterns. S uch ethnographies claim that the patterns o f everyday speech when altered, are distinguished from performed speech by the execution of culture specific strategic devices that mark frame, key, cue, and codify particular verbal acts for particular social purposes (Bauman 1977, 1986, 1992, 1998 ; Sherzer 1987; Dooley 2001 ). Baka speakers identify a variety of speech acts. The Baka lexicon records a range of at least nine types of speech behaviors. Figure 2 2 displays a table of lexical entries of the most prevalent Baka speech act ter ms, along with their typical collocating verbs. From these sets of lexical entries, I observe a repertoire of nine speech acts and their typical collocating verbs. Thus, in the Baka language someone may be said to (1) speak a word or have a discussion ( m fn l`" ); (2) make bnmudqr`shnm"' klt" (: (3) explain formal counsel "' j` kn" (: (4) give personal advice ( kd vt" (: (5) relate an account ( ln rhln" (: (6) tell, recount, or narrate a story, legend, or fable ( k # j` m" (: (7) recount an allegory, oqnudqa+" nq" chbs tl" ( fa mf mf" (: (8) strike a comparison or speak a parable ( mb ` kh (: (9) sing ( d (" a song ( d" ). ( Brisson 2002; Kilian Hatz 1989 ) The two terms ngo! ma! (discussion) and l mu! ( conversation) implicitly identify an ordinary, unmarked speech style, that is, a Baka speakers everyday linguistic
50 code." The other seven terms ka! lo! le! wu! mo! simo! li! ka! n"! gb"!ng"!ng"! mba! li and 'e! however, identify a "special," "variant," "marked [speech] code" (cf. Herndon 1975; Urban 1982:18 22; Levinsohn 2001:38 48; Alvarez Ca! ccamo 2001:23 26; Banti and Giannattasio 294). And while the Baka do not as yet formulate abstractions about what actu ally constitutes distinctions between "marked" and "unmarked," "ordinary" and "special" speech acts, I have empirically abstracted several of the distinctive features that constitute the two variant speech acts that the Baka identify as 'e! (song') and li !ka! n"! (story'). The Baka also specify a third, variant speech act which occurs when song is sung in the course of telling a traditional story. They identify that speech act as 'e! na li! ka! n"! that is, story's song.' Thus, in this linguistic const ruction, that is, language about music, Baka speakers imply that li! ka!n"! song not only shares characteristic marks of both sung and narrated verbal performance types, but also bears distinctive song and story marks of its own. General descriptions of th ese marks will be set out in the sections that follow, that is, Sections 2.2, 2.3, and 2.4. In those sections, I will introduce the generalities of Baka song, story, and story's song under the headings of "Poetic Verbal Performance," "Narrative Verbal Per formance," and "Sung Narrative Verbal Performance," respectively. I will take up song ( 'e! ) first, but reframe it in the words of Banti and Giannattasio (2004) as "poetically organized discourse" (as opposed to "musically framed discourse"). I choose to frame song in this way because discourse about the nexus of music and language is more aptly discussed in poetic terms, especially when shared performance phenomena like rhythm and intonation are under discussion. Furthermore, the performance frame of li! ka! n"! song is generically
51 verbal, as opposed to musical, that is, song is sung in the context of "telling," not vice versa 2.2 Poetic Verbal Performance: Song E thnographic research has shown that "poetic" forms and behaviors are almost universally widespread" ( Banti and Giannattasio 2004: 290 ; Feld and Fox 1994:30, citing Lomax 1968 ) M oreover, "poetry is performed more commonly as sung rather than spoken discourse in all oral traditions" (2004:297). Such is the case in Baka verbal communication. It should be added, however, that the potential for poetically organized verbal performanc e sung or spoken is always present in any kind/genre of Baka discourse, even during everyday speech. For as with most cultural performance forms verbal or non verbal poetic procedures are "variable qualities, relatively more or less salient among the mult iple functions served by a communicative act" (Bauman 1992:44). Thus, for example, from within the nine speech acts previously displayed in Figure 2 2, if I were to sample the relative presence of a single poetic procedure such as repetition, we would gen erally observe less repetitive phenomena in "conversation," "discussion," and "personal advice," more repetition in "formal counsel," and personal "accounts," and still more in "proverbs," "parables," and "stories." Tautologies of the greatest variety an d frequency, however, would be reserved for "song." Similarly, along this same continuum of speech acts we could expect to find the same relative increase in the presence of other poetic procedures, although song, again, would exhibit more "poetic" speech than all other Baka speech acts.
52 In Chapter 3, entitled, "Baka Song text Poetics," I provide a detailed inventory of those poetic procedures that generally distinguish sung Baka discourse ( 'e! ') from everyday Baka discourse. That account, as its ti tle indicates, tends to emphasize song text poetics, though the poetic procedures that it describes extend beyond mere "textual" concerns. Theoretically, the poetic formalization of Baka song texts is characterized by what Roman Jakobson (1960:358) descri bes as "the [semiotic] projection of the poetic function into a verbal sequence to the point that the principle of equivalence is p r omoted as the constitutive device of the sequence ( brackets and emphases added; see also Sec 3.1 ) Formally, then, I iden tify five particular domains of p oetic e quivalen c e i n Baka song: rhythmic, syntactic, lexical, phonological, and semantic equivalence. Each domain, in turn, exhibits numerous poetic devices of its kind. For example, equivalent r hythmic forms emerge as pu lse and meter; syntactic similarities may include repeated verses and verse segments; l exical tautologies m ay be manifested as repeated words or as homonyms; phonological ly poetic phenomena may include assonances and ideophones; and semantic equivalences c an be recognized in metaphor or parallelism As is suggested by these examples, the boundaries of syntactic, lexical, and semantic domains tend to most often include strictly "textual" phenomena. Rhythmic and phonological domains, however, more often cons titute what many would consider to be "musical" phenomena, or, to put it in another way, "the music in language" (Feld and Fox 26). 2 This interpretation is more likely when the scope of the "phonological" patterns, that is, sonic patterns, of verbal perfo rmance is thought to include melodic or timbric qualities. When speech sounds are interpreted in this way, melic, timbric, and
53 rhythmic phenomena are recognized as shared qualities of both language and music. The formalization and re formalization of the se three dimensions of speech is a hallmark of "poetically organized discourse." In their 2004 essay on poetry and "poetically organized discourse, Banti and Giannattasio claim that the main procedures for formalizing speech sounds beyond those used in n ormal conversation seem to be no more than three: 1. altering voice register' ... 2. altering melodic contour... 3. segmenting utterances rhythmically ... (295). Banti and Giannattasio's three "main formalizing procedures" roughly correspond to the disti nctive timbric, melodic, and rhythmic qualities that typically infuse sung Baka discourse. Indeed, the most distinctive formal generalities that distinguish Baka singing from Baka speaking concern rhythmic segmentation, voice register (which is primarily constituted timbrically), and melodic contour. A more detailed discussion of these generalities will be set out in the immediately following sub sections. Section 2.2.1 will introduce Baka song's basic rhythmic distinctives, Section 2.2.2 will treat melo dic generalities, and Section 2.2.3 will discuss song's special voice register. 2.2.1 Baka Song's Rhythmic Generalities The predominant differentiation between everyday Baka discourse and poetic Baka discourse is rhythmic, periodic rhythm in particular. More generally, Jean Molino claims that "poetry cannot be confused with language or with any of its functions: it is the outcome of imposing upon language a structure that has very strong links with music and dance" (Molino 2002:31 cited in Banti and Giann attasio 2004:292) Vida Chenoweth frames the nexus of music and language ( in song ) in a similar fashion:
54 It is musical grammar superimposed upon verbal grammar which makes song transcend speech. This musical grammar has a life' of its own. It is the nature of song to enhance words. (italics added; 1972:102) In the case of Baka song, the most pertinent generality regarding the relationship of musical grammar to verbal grammar concerns rhythm. This characterization is so, because the essential music al mark of sung Baka discourse is that it is rhythmically organized Melodic organization may occas i onally be absent, as in vocal compositions employing sprechstimme like utterances ( i.e., rhythmic speech'), but rhythmic organization may not. Thus, the musical syntax of periodic phrase level rhythmic units defines the bounds of a song text's poetic line, transforming verbal sequences into lyrical lines (Sec. 3.3.4 3.3.5) Such a process is not unusual in African literary forms. "In fact the occurrence o f music or of a sung mode of expression has sometimes been taken as one of the main differentiating marks between [African] prose and verse" (Finnegan 1970:75). Yet, the underlying dynamic and character of the musical organization of verse any verse g oes beyond cultural particularities. Barbara Herrnstein Smith, in her book on poetic closure, posits that as soon as we perceive that verbal sequence has a sustained rhythm that it is formally structured according to a continuously operating principle of organization we know that we are in the presence of poetry and we respond to it accordingly expecting certain effects from it and not others, granting certain conventions to it and not others. (emphasis added; Herrnstein Smith 1968, cited in Alter 1985: 6) Thus, a s concerns Baka poetics in particular, the primary distinctive poetic feature of sung Baka discourse is the musical ly meter ed line to which it is set.
55 It must be emphasized, that the poetic nature of a Baka line is not so much verbal, as it is mu sical. More precisely, a line is a rhythmical unit (an organized unit of time), not a syntactical unit (Pinsky 1998:28, 34). However, a Baka song line does not express rhythm formed of patterned prosodic or semantic units (as in Greco Roman and Japanese meter, or Semitic parallelism), but periodic rhythm minimally formed from a continuously recurring sequence of pulses More specifically, Baka song line meter types are organized in units of either 16, 12, 8, or 4 p ulses per line Thus, the average meter is 1 2 p ulses per line though the predomina n t meter is 8 pulses per line (Chap. 3, Sec. 3.3.4). I set out a more detailed description of Baka rhythmic devices in Chapter 3, Section 3.3. For the purposes of this present introductory section on Baka song' s rhythmic generalities, it is enough to simply state that in all sung Baka discourse f our rhythmic devices converge to form what Herrnstein Smith would call the continuously operating principle of [poetic] organization ": they are, pulse, tempo, meter an d line. The recognition of a hierarchically organized "musically metered line" as the essential poetic mark of Baka song is an empirical though I trust accurate abstraction. The Baka, however, do not consciously articulate any such notion. Their conscio us conceptions of rhythm in song begin with an explicit identification of particular rhythmic motifs ( kole ') that they (the Baka) typically associate with this or that' song. For example, the traditional constellation of percussive rhythmic patterns ass ociated with li ka!n"! song as transcribed in Figure 2 4, Section 2.4 is simply identified by Baka speakers as kole na 'e! na li! ka! n"! ," that is, the li! ka! n"! song rhythm.' Even when this composite rhythmic motif (or simply one or two of its component patterns) is p ercussively performed without vocal performance, the rhythm is still
56 indexically recognized as the "right one" (" e! e na$$ k#! e jo$ ko! n#! ") for that particular song genre. Most Baka speakers can identify numerous song styles in this manner, especially da nce songs. Every traditional rhythm is correlated with the musically metered line of its respective song style. Motif and line are found to correlate in one of three ways: (1) the boundaries of certain rhythmic motifs may also coincide with the boundaries of a song's lyrical line, while (2) the boundaries of other rhythmic motifs are iterated twice before coinciding with song line boundaries; and (3) still other composite motifs may span the duration of two complete song lines, depending upon the relative measure of pulses per rhythmic motif and pulses per poetic line. Not only are specific rhythmic figures correlated with specific musical meters, grammatical constructions are correlated as well. Larger syntactic constructions, like verses, are set to pa rticular metric lines. Similarly, smaller constitutive verbal constructions, like syllables, are subsequently synchronized with the component segments of the larger rhythmic motif. A more detailed analysis of Baka songs' rhythmic tradition is beyond the scope of this present section. Detailed descriptions of traditional rhythms would only serve to distinguish one song style from another song style. It is enough, then, to close this present section on the rhythmic generalities of Baka song and reiterate that all sung verbal utterances as opposed to spoken verbal utterances are always segmented according to a musically metered line.
57 2.2.2 Baka Song's Melic Generalities In addition to special formalizations of rhythm Baka song as opposed to Baka speech is linguistically and musically marked by distinctive intonational and melic formalizations. The raw material of intonational formalizations is constituted by the phenomenon of pitch. Verbal intonation is a variation in pitch sequences in speech (Loos 2004). An intoned melody, by extension, is a continuum of culturally regularized sequences of emically discrete tones (Chenoweth 1979:125). Pitches in Baka culture, however, are not culturally regularized; intervals between pitches are. Thus, the minimal unit of melodic organization is a melodic interval a sequence of two tones (ibid.). To be sure, there are many other emic principles that subsequently order the succession of melodic intervals, but my goal here is not to posit all the patterns of intonationa l (or melodic) organization, only those that minimally distinguish the movement of pitch in Baka song from that in Baka speech. Alexander Reed begins his 2005 dissertation, The Musical Semiotics of Timbre in the Human Voice ," by distinguishing the semiot ic potential of the multiple individual frequencies that normally constitute the complex spectrum of all periodic sound. At a most elementary level, fundamental frequencies are shown to constitute signs of pitch while partial frequencies collectively sig nal timbre. In the process of discriminating these bands of frequencies, Reed formulates the simple yet important generalization that "o ne chief difference between speech and singing is that there is much greater movement of and stability between movement s of the fundamental frequency in a singing voice" ( emphasis added; Reed 2005:5 )
58 Generalizations like Reed's are normally posited with the aid of spectrograph analysis. Yet even without the confirmation of spectrograph measurements, the kinds of variatio ns in fundamental frequencies of which Reed describes are plainly evident when comparing Baka speaking and singing. To my knowledge, however, the Baka make no such abstractions about alterations in fundamental frequency patterns. They do not, for instanc e, have an equivalent term for "pitch" in their lexicon. For the Baka, as already noted in Section 2.2.1, a speech act is identified as "song" ( 'e! ) when verbal segments are rhythmically metered, that is, when verbal segments are performed in particular p eriodic rhythmic patterns ( kole '). However, even in the absence of explicit verbalizations about pitch, I have empirically observed that whenever such periodic segmentation is performed, "greater movement of and stability between the movement of the fund amental frequency" is always concomitant, though in varying degrees. Lesser increases in the degrees of alteration in fundamental pitches are exhibited in occasional Sprechstimme like song styles. Most often, however, song is indexically (and iconically) signified by still greater degrees of movement of and stability between movements of fundamental pitches. When regularized, these alterations generate patterned melodic contours. The empirical and emic measures of these movements have yet to be adequat ely determined, but certain basic characteristics can be preliminarily proposed even now. But before doing so, a word about my pre ethnographic research in this regard seems in order. My initial research of Baka melodic formalizations began with an orien tation to t he fruitful and seminal analyses of Aka melodic and polyphonic music systems by French ethnomusicologists at the Centre National de la Recherche Scientific (e.g., Arom
59 1991 ; Frniss 1991, 2006; Frniss and Olivier 1996). Susanne Frniss briefly summarizes Aka melody and polyphony as being based upon an anhemitonic pentatonic system a five tone scale in which neighboring degrees are separated by intervals that can vary between about a major second and a minor third. [Aka] Musicians, however, are more concerned with correct relationships between parts than they are with a rigid idea of interval sizes. (2006:177) With greater specificity, Frniss further explains that Aka music issues from a system of multiple levels of significance: the princip al parameter concerns the order of succession of scale degrees in a melodic pattern derived from a pentatonic scale in which the only constant is the absence of an interval of a half stepthe second significant parameter, which informs the realization of p olyphonic songs, is based on a vertical arrangement of intervals at certain predetermined points. The unfolding of melodic lines, with all the freedom of development it allows, is nevertheless dependant upon the anchor points of the polyphony (my transla tion; 1993:136, 137) 3 Given the common narrative concerning African pygmy peoples, one might presume that the melodic system of the Baka in Cameroon is the same as that of the Aka in Central Africa. However, the linguistic, geographical, and historical di fferences between the two groups as briefly described in Chapter 1, Sections 1.4 and 1.5.1 should suggest caution before making such an assumption. Moreover, even if the music cultures were identical, the application of another theoretical model i.e., one other than that of the C.R.N.S. group might yield a slightly different, though presumably complementary, description and analysis. To this end, I have begun my own analysis.
60 My emic analysis of the Baka melodic system is not yet complete, though a few f undamental characteristics are already emerging. The theoretical model for an emic analytical approach was first developed by Vida Chenoweth in the 1970s (Chenoweth  2006, 1979). I first applied Chenoweth's model during my 1989 Master's thesis' ana lysis of the melodic system of the Tewa Pueblo Indians of New Mexico. 4 Chenoweth's model is based upon a slightly, though critically, different generative premise than that of the C.N.R.S. group, a premise that presumes in short that as a phoneme is to la nguage, so a melodic interval is to music. Thus, from the start Chenoweth's model theoretically assumes that the "paramtre principal" of melodic organization is first constituted by a regularized quality of movement from one pitch to another within a sin gle given melodic interval ; and thus, presumably, melody could not be principally constituted as suggested by Frniss's hypotheses by the quality of movement between degrees of pitch within a given scale I do not intend to posit a full description of th e Baka melodic system in this dissertation, let alone constructively critique the analyses of other (presumably) similar melodic systems. My present focus is only on those qualities that minimally distinguish the patterns of fundamental frequencies of Bak a speaking from those of Baka singing. My intent, then, is simply to account for yet one more distinctive feature of Baka verbal performance, that is, one verbalized sign among many signs that both indexically and iconically marks singing as an extra ordi nary Baka speech act. Thus, I have found it necessary to introduce, if only briefly, some of the rationale for establishing the melodic interval as the essential, initial measure of intonational and melodic variation.
61 Having done so, I quickly resume my introductory description of the essential melic marks of Baka song. First, like the Aka of Central Africa, anhemitonic intervallic progressions predominate in Baka song. Thus, the melodic contours of Baka singing are predominantly informed by etic interva ls akin to unisons, major seconds, minor thirds, perfect fourths, and perfect fifths. 5 However, unlike Aka intervallic sequences, minor seconds and major thirds are not uncommon in the Baka repertoire. One probable reason for this discrepancy is undoubte dly due to the fact that unlike the corpus of songs analyzed by Arom and Frniss, I have included a number of song styles in my data that most likely have (relatively recently) been borrowed from neighboring music traditions. I have chosen to include some of these borrowed styles only when there seems to be adequate evidence that the Baka themselves appear to be generating and sharing (with some regularity) new compositions in these styles. Whenever this appears to be the case, I consider such composition s to be an emerging style in the regional Baka repertoire, just as it seems the Baka do. Not surprisingly, most of these song styles exhibiting the less common "hemitonic intervallic progressions" tend to emerge from performance contexts commonly associat ed with certain youth dance songs, particularly the basu$ ka and m#!ba! si! As is readily deduced from the work of Frniss et al ., there are several other complementary levels of Baka melodic organization that organically extend well beyond isolated abstrac tions of "minimal units of movement." One such formalization involves the typical distribution of the available melodic intervals. Simply put, that particular formalization stems from the fact that no single voice part in the traditional
62 Baka repertoire coordinates the intervallic boundaries of melodic contours with less than two or more than five distinct emic pitches per song 6 Some analysts would identify this parameter as a pentatonic scale. I prefer, however, to identify it as an inventory of fiv e pitches, not fixed pitches, but rather, an inventory of five tones dynamically relative to a tonal center. To clarify, only individual voice parts may not employ more than five pitches. The c omposite inventory of distinct pitches in songs with multipl e voice parts however, may exhibit more than five pitches. From a sample of eighty six diverse songs, slightly more than half exhibit five or fewer pitches; slightly less than half appear to use more than five, that is, six, or possibly seven, distinct p itches. Exceptions to the typical five tone inventory generally appear to occur in polyphonic songs. To reiterate, these five pitches are not fixed, but are only identified relative to a tonal center. And further, even though melody is essentially a con tinuum, any notion that these inventories of tones are somehow organized along a scalar continuum is not yet warranted in the light of my preliminary analysis of Baka song (Chenoweth 1979:125). In analyzing Baka melodic contours, the notion of pentatonic scales as organizing principles of melically inflected verbal discourse might arise from the potentially prescriptive effects of tuned instruments. In Baka instrumental performance, for example, no trad i tionally tuned chordophone, regardless of how many strings it may have, exhibits more than five discrete pitch classes But instrumental melodic practice does not necessarily prescribe sung melodic practice. For example, I know of no melodic sequence in the traditional Baka song repertoire even when acco mpanied by a tuned instrument that plainly exhibits an uninterrupted, descending or ascending
63 "scalar" succession of all five available melodic pitches. Such a continuum may be physically, and even theoretically possible, but within my corpus of recording s no such melodic figure is ever performed by a singing voice. Before drawing this section to a close, I note that the concept of "scale" is not only problematic in discussions of Baka melodic phenomena, but in other contexts as well. Arnold Whittall, i n his entry on melody in the Oxford Companion to Music argues that It is important to the consideration not only of ancient music but also of the Western tradition to realize that the concept of scale is based on practice, rather than being the basis of p ractice. The scale determines the kind of music made only in the sense that it becomes fixed in instrumental music. Thus the existence of five string lyres some three thousand years ago, as well as of wind instruments with certain possible fingerings and transpositional relationships to other instruments, suggests the fixed pitch content of ancient melody on a pentatonic basis ( Whittall 2010) 7 In an analogous article, Alexander L. Ringer writes, Leaving aside the admittedly important issue of fixed tuning however, scalar considerations can hardly be said to place more than very general constraints on melodic activity, if only because the scales themselves are derived from existing melodic practices. The image of the scalar tail wagging the melodic dog wou ld seem grotesque, were it not for the implied reminder of the extent to which musical notation, with all its blessings, has narrowed Western man's understanding of a cultural phenomenon that is always aural in essence and rarely if ever graphic (Ringer 2 010) 8 A much more in depth argument is called for in defense of my initial suppositions concerning Baka melodic perceptions. The preceding introductory comments are only
64 meant to clarify as much as seems necessary the bases for the measures I have chosen to characterize the kinds of variations of intonational patterns that distinguish sung from spoken verbal discourse. To summarize, then: The increased movement of and stability between movements of fundamental frequencies that typically distinguish Baka singing from Baka speaking is minimally signaled (1) through anhemitonic intervallic sequences which are most often expressed through an available intervallic inventory of unisons, major seconds, minor thirds, perfect fourths, and perfect fifths, respectiv ely, and (2) through two to five discrete tones (intervallically relative to a tonal center) that mark and correlate the boundaries of the available inventory of emic intervals. 2.2.3 Baka Song's Voice Register Generalities Not only are sung verbal perfor mances marked by special rhythmic and intonational formalizations special voice registers are often employed as well. At this point in my study of Baka song voice registers, I am only able to propose two distinctive sets of binary features: first, two co ntrastive sets of "more dense versus less dense" spectral fields of high register harmonics, and second, two contrastive sets of "closed vowel formants versus non closed vowel formants." These proposals are deduced in large part from earlier phonetic and acoustic studies by Susanne Frniss (1991b) of the singing techniques of Aka yodelers in Central Africa. 9 Before describing any fineries of Aka or Baka yodeling, it bears repeating that Baka yodeling ( yeyi ') according to the Baka is singing ( na 'e $ 'e! ). In Baka talk' about Baka song, Baka speakers often identify yodeling as "song" ( 'e! '), but more often they identify it as yeyi! (yodel'). Non yodeling songs, however, are never identified as
65 ye yi! '. I find it particularly telling that in regards to the specific act of yodeling ( yeyi! '), s ingers ( wa 'e $ 'e! ) are either said "to sing" (" na 'e" ) when yodeling, or more often in reference to female vocalists to whistle" (" na u! ) when yodelin g as in Wo$ s#!o, wo$ a u! ye$ yi (The women are whistling a yeyi! song'). This recognition by the Baka of a whistle like component in yodeling, as I will discuss near the end of this section, corroborates with Frniss's spectral analysis of the so call ed "head' voice sound in Aka yodeling. But b efore proceeding with a discussion of the head' voice chest' voice phenomenon associated with voice register, the basic term register needs some preliminary clarification Often, confusion ensues when empl oying the term "register" in interdiscpilinary studies involving both musicology and linguistics especially when the linguistic components include studies of phonetic s, phonolog y, and linguistic anthropology. Music theorists, for example, most often use register to denote a particular region of a vocal or instrumental frequency range In addition, musicians may even use the term to refer to a set, or the control of a set, of organ pipes. Anthropological linguistics (or, sociolinguistics), on the other h and, employs the term register to indicate a linguistic repertoire that is associated, culture internally, with particular social practices and with persons who engage in such practices.  Formally, registers differ in the type of repertoire involved ( e.g., lexemes, prosody, sentence collocations), and many registers involve repertoires of more than one kind. (Agha 2001:212) In yet another research context, phoneticians apply the term register in reference to "the VOICE QUALITY produced by a specific physiological constitution of the LARYNX ( Crystal 1997:327). For example, "variations in length, thickness, and tension of the
66 vocal cords combine to produce (in singing) the differences between, soprano, contralto, tenor, and bass etc. voices, and also (within one person) such differences as between head' ( falsetto ') and chest' voice" (ibid.). Of all the various applications of the term "register," this acoustic phonetic concept is most pertinent to my present discussion of Baka song's voice register generalities. While my present aims focus more on sonic phenomena, I do not want to imply that sociolinguistic notions of register are in any way irrelevant. The unique repertoire of physiological formations that produce such contrasting vibratory pheno mena as the alternating chest' and head' voice in Baka yodeling most certainly indexes particular Baka social events and practices. Indeed, such sonic signs are perpetuated precisely for culture specific events. But my aims are much more limited at the moment. And so, for those readers interested in the sociolinguistic implications of Baka or Aka voice register, I remind them of the various ethnographic descriptions by Turnbull (1961, 1965), Frisbie (1971), Arom (1991), Sarno (1995), Frniss and Olivie r (1996), Kisliuk (1998), or Frniss (2006) that have already drawn attention to the numerous ritual, spiritual, and sociological contexts so potently indexed by the unique sonic signs of yodeling. With that said, I return to my general description of th e physiological and acoustic formalizations of Baka song voice registers. Linguist Giorgio Banti and ethnomusicologist Francesco Giannattasio as cited earlier, claim that "altering voice register" is one of the three main procedures for formalizing speec h sounds. More specifically, "a ltering intentionlly and/or by convention one's voice register, is to alter its frequency range, timbre, and intensity (emphasis added; ibid.). Voice register s then, are particular bundle s (i.e.,
67 ensemble s") of subtle voice qualities generated by particular vocal tract conditions (Frniss 1991b:168). To better appreciate a voice register's acoustic phenomena, some understanding of its physiological formation is helpful. The initial physiological site of vocal producti on is the larynx. In both speaking and singing, the vibrating vocal folds of the larynx produce a complex spectrum of frequencies. The particular spectral formation of these frequencies, as alluded to earlier, is initially formed by particular combinator y variations in the length, thickness, and tension of the vocal cords. No rmally, the lowest and most prominent frequency generated by the vocal cords is the fundamental (Reed 2005:2) Th e perception of a fundamental frequency, however, has more to do wit h the sonic dimension of pitch, than with the dimension of timbre (ibid.) The specifically t imbral qualities of voice register are primarily signified by particular configurations of the multiple non fundamental spectral frequencies of a sound, that is, its partials or harmonic overtones 10 As these partials progress through the vocal tract, some (either naturally or through manipulation) resonate with and cluster around particularly resonant vocal tract cavities. This resonance reinforces these frequen cies. The loudest and most prominent clusters of frequencies are called formants "The number, intensity, and position of a sound's formants is the most important characteristic" of the timbral dimension of a sound (ibid.). Particular formant frequency resonances of a "lower" spectral tessitura (below F2, i.e., below 2000 3600 Hz) characterize phonemic vowel formations while particular partials of a higher spectral tessitura (above F2, i.e., above 2000 3600 Hz) signify consonants and many other acousti c qualities like "sibilance, breathiness, and brightness'" (2005:4). Further, and more importantly, from within
68 these higher levels of the spectral frequency range "we [also] receive a great deal of information regarding the location, clarity, and richne ss of a sound" (ibid.). Thus, while lower level formants critically signal the vocalic, phonemic basis of a people's language, 11 particular formations of a voice's higher partials signal who is speaking, or singing. These higher partials not only identify a person's gender, or likely age group, but who they are as unique individuals (Sundberg 1987:2; Howie and Delattre 1962:6 9, cited in Reed 2005:4). Given the special semiotic potential of sonic signs formed from a voice's higher partials, it is tempting, at this point, to speculate about the collective voice register identity of whole people groups like the Aka and Baka, people who share such unique physiological characteristics. One might ask, for example, "How significant is the relationship between th e collective sizes and shapes of Baka female vocal tracts to the seemingly unique acoustic quality of their choral yodeling technique?" "What might a comparative spectral analyses with other people groups reveal about the uniqueness of Baka voice?" These and other similar questions, however, must be left for another study. For I should digress no longer, but directly progress to a description of the distinctive voice register and formant qualities produced by the vocal tract of Baka singers. A "chest" v oice, produced by shorter, thicker, denser vocal cords, is used in ordinary speaking and ordinary singing (Frniss 1991b:169). 12 Frniss's acoustic measurements of the Aka head' voice partials reveal a dense field of harmonics, much more dense than that produced by a head' voice technique (176 183). When a "special," that is, a contrastive, voice register density is called for, as in the case of
69 yodeling, longer, thinner, and more attenuated vocal cords are called on to produce the head' voice sensatio n. Frniss's spectral measurements of the acoustic effects of this "mechanism" typically reveal a far less concentrated field of harmonics, affecting a "less dense," "lighter" acoustic quality (ibid. 182). The effect of this regularized alternation betwe en "more dense" and "less dense" voice qualities (timbres) acts to promote the timbre of voice register as an increasingly constitutive principle and special mark of this speech act (185). But as distinctive as this device may be, it is not the only devic e being timbrally foregrounded; special vowel formant patterns are correlated as well. The Baka and Aka, interestingly, share the same set of seven short phonemic vowels: /i/, /u/, /e/, /o/, /(/, /a/, and /)/ 13 The formants that characterize the timbre of both vowel systems, as in most languages, are constituted among lower formant frequencies. I assume that the formant formation of the seven phonemic vowels of the Baka language is phonemically the same in Baka singing as it is in Baka speaking. If this were not so, the resulting phonemic confusion would render texted voice parts unintelligble. The o ne significant exception to this pattern of vowel formalization takes place as one might guess, during the performance of yodeled song styles. In the Bak a song tradition, sung verbal performance is normally executed by non yodeling voices, or voice parts. Yodeling voices may vocalize but they do not simultaneously verbalize Not all the world's yodeling traditions exclude lexical performance so strictly For example, a number of American country music styles combine in the same verse yodeled vocables with lexical utterances (Wise 2007). I
70 cannot as yet propose the exact reasons as to why the Baka typically choose to exclude verbal performance from most of their yodeled voice parts. However, Frniss's study of Aka yodeling again, seems to s uggest a few potential forces, among which the formation of vowel formants is most pertinent. I n the cou r se of Aka yodeling, for instance, the contrastive alternatio n of head voice and chest' voice register is also correlate d with alterations between a specific set of the Aka's closed vowels that is, the phonemic /i/ and /u/, and the phonetic [y] and [ u ] with those of an opposing set of their non close d vowels, t hat is, the phonemic /(/, /a/, /)/, and /o/ and the phonetic [oe] and [ ] respectively (175). I can confirm that nearly the same correlation takes place in Baka performance, though I have not had opportunity as yet to verify the subtle presence of the f our additional, conditional phonetic vowels, that is, [y] and [ u ] [oe] and [ ]. Regardless, the cumulative effect of these correlated and homologous timbral "densities" i.e., closed vowels with less dense voice registers, and open vowels with more dense voice registers seems evident in their mutual reinforcement as emphatically contrastive timbres, and thus as marks of a "special" kind of verbal performance. These correlated alternations of "higher" and "lower" registers, along with their well known "dis junct intervallic motion," thus constitute the fundamental timbral character of the yodeling technique (Baumann 2011). As a result, whenever the contrastive timbral patterns of the yodel technique are promoted as the constitutive principle of a sung vocal sequence, the distinctive phonemic timbres of normal lexical utterances are at best, hard to realize, and more often, abandoned altogether (Frniss 1991b:175). 14
71 Frniss implicitly corroborates certain aspects of her empirical phonetic and acoustic analys is of Aka yodeling with ethnographic analysis. For example, she reports that within the typical four part performance textures of Aka song, only one voice part utters words, that is, mo! ta $ ngo! le! ; 15 and mo! ta $ ngo! le! does not yodel. Yodeling is reserved for one of the other voice parts, that is, di! ye! i$ 16 Correlationally, mo! ta $ ngo! le! (literally, one who counts') exclusively sings with the (lower) "chest" voice register, while di! ye! i$ (yodel') uses (in alternation) both a "chest" voice register and a (higher) "head" voice register (2006:175). 17 The Baka differentiate voice parts, as well, but not nearly so finely. In multi part voice textures, two categories apply: wa! nja! mba (the one who begins') and wa! tu$ ko! (the one who "pours"'). 18 While the re is only one wa! nja! mba voice part though more than one person may sing it, depending upon the song style there is often more than one distinct wa! tu$ ko! voice part. So, while any "chest voice register" voice part may utter either words, or non lexical vocables, yodeled voice parts with their alternating "chest" and "head" voice registers predominantly only utter non lexical vocables (especially when the song style is explicitly identified as y e y i or yeli! ). In general, then, yodeled parts are most o ften identified as wa! tu$ ko! voice parts, and lexical utterances are most often identified with the wa! nja! mba part. As noted earlier in this section (Sec. 2.2.3), the Baka recognize a "whistling" ( na u! ) u! ') quality in female y e y i singing. For the mo st part, both scholarly and popular studies identify yodeling's head' voice sound as falsetto ," or fausset ," (i.e., "artificial") (Frniss 1991b, Baumann 2011). Still, a number of acoustic and vocology studies further specify a difference between male and female "head and chest" voices. Johan Sundberg reports that
72 while both men and women have a head voice, these registers are apparently not the same thing: for when women yodel, it is not between their chest voice' and their head voice', but rather between their head voice and whistle register. (emphasis added; Wise 2007:7, citing Sundberg 1987:50) While some confusion still surrounds academic definitions of a whistle register the implications of the basic concept are clear enough in that the dimi nutive acoustic density of the higher range partials produced by a flute is acoustically similar to that produced by the vibrating outer layers of the vocal folds of a female Baka yodeler (Miller and Schutte 1993). 19 The special timbral status of the higher level voice register of y e y i is further reinforced by other socio cultural claims and traditions. From his many years of ethnographic work among the Baka, Robert Brisson reports that not only is the sound of y e y i likened to a whistle, or flute, some Ba ka have even said that yeli! song originates from the resonant sound of God's harp zither" ("le chant du "yeli" viendrait de la re! sonance de la guitare de Komba") (English translation, mine; 2002:626). 20 Baka talk' about y e y i sound is further contextu alized in the ethnographic work of Fr. Daniel Boursier and Robert Brisson. Both Boursier and Brisson document the ways in which the y e y i song tradition is socio culturally set apart. For example, the y e y i! tradition is set apart sociologically by perpet uating and framing it in a special rite of initiation for Baka girls. An insightful account of this tradition is documented in Pli: Mmoires d'une Femme Pygme: Tmoignage Auto Biographique d'une Femme Pygme Baka (Sud Est Cameroun ) an autobiography of a Baka women, Pli as told to Fr. Daniel Boursier (1996).
73 There are other aspects of the dimension of voice register that deserve further attention but can only be briefly mentioned at present. For instance, I have occasionally heard and documented Baka singers performing a speech act that some western academic traditions call sprechstimme, or speech voice.' The voice register of this speech act comprised of lower formant frequencies and an irregular movement of and between fundamental frequencies li es somewhere between speaking and singing. Baka speakers, nonetheless, simply refer to this verbal formalization as song' ( 'e! '), with no apparent regard for the etic distinctions that I perceive in the acoustic character of its fundamental frequency and formants (see also Sec. 2.1 and 2.2). To conclude this section, then: the Baka implicitly recognize three basic voice regist ers, those of speaking, singing, and yodelling. The voice registers of "ordinary" singing and yodelling, however, are often differentiated still further. Empirically, for example, they are found in contrast in two binary sets of timbres: first, as either "more dense" or "less dense" voice registers, and second as well as correlationally as either "closed vowels" or "non closed vowels." Ethnographically, the "more dense" and "closed vowel" registers may even be identified as whistle like, or possibly e ven harp like. 2.3 Narrative Verbal Performance: S tory Many scholars of discourse analysis explicitly claim, or even tacitly assume, that narrative discourse is a universal social act (e.g., Gulich and Quasthoff 1985:169; Dooley and Levinsohn 2001; Renkem a 2004). Implicit in these claims is the notion that there are universal c haracterisitics of narrative discourse, that is, general features
74 that distinguish narrative discourse from other types of verbal performance. To the degree that this is true, then any attempt to describe the distinctive features of a culture specific narrative for example, a traditional Baka li! ka!n"! must first establish the theoretical parameters of any such universal narrative features In this present section, then, I will fi rst outline the theoretical assumptions on which my description rests. Based on these assumptions, I will follow a decidedly functional and cognitive research approach, drawing particular attention to grammatical features, both formal and pragmatic. Once I have introduced the essentials of my adopted theoretical model, I will proceed to decribe the generalities of Baka narrative in terms of that model. In surveying the theoretical terrain of various approaches to discourse analysis, Mann and Thompson (19 92) claim that no matter what one 's research orientation may be, whether it is a "semantic orientation, a "speech act orientation or "social act orientation (or a combination of all three), the study of text relations practically (and historically) begi ns with semantic considerations. all three orientations are said to largely agree on 1) the centrality to discourse organization of semantic or discourse functional relations among the parts of the text, specifically between clause like units 2) predi cations' and propositions' are the basic units of text, 3) clauses' are prominent subsets of predications and propositions, and "minimal working units," 4) textual units are relational, e.g., there are interclausal relations,' and relations between pre dictions' and larger aggregated units.  Thus, discourse analysis minimally begins with the study of text relations' who's clauses (liberally defined) are used to build discourse structure in different waysfor a variety of ends (emphasis added; Mann and Thompson 1992:19, 21)
75 For the discourse analyst, then, grammar clause level grammar in particular acts as the basic strategic resource for "framing," "keying," and "codifying" particular discourse performances. Narrative discourse performance, as on e among several discourse genres, is marked by its own pattern of discourse features and pragmatic goals. Discourse analysis of narrative d iscourse There are a variety of theoretical models applied to the analysis of narrative discourse (Van Dijk 1997; Be augrande 1997; Renkema 2004). I have adopted an eclectic model, one proposed by Robert Dooley and Stephen Levinsohn as "a good approximation of how discourse is actually produced and understood" (2000:iii). 21 In their research approach, Dooley and Levinso hn draw on the scholarship of a number of discourse analysts, the work of Robert Longacre, in particular. Longacre places the investigation of grammatical structures at the heart of his approach, but only after boldly typifying his understanding of t he un iversal features of narrative discourse. Indeed, d iscourse typology is a major emphasis of Longacre, who argues that typology is an essential step in any linguistic analysis of a discourse. Characteristics of individual discourses can be neither describe d, predicted, nor analyzed without resort to a classification of discourse types. It is pointless to look in a discourse for a feature which is not characteristic of the type to which that discourse belongs. So determinative of detail is the general desi gn of a discourse type that the linguist who ignores discourse typology can only come to grief. (Longacre 1996:7) Longacre's classification of broad discourse genres is primarily based upon predicative, propositional content as distin c t from form. 22 He cl assifies the content of a
76 discourse any discourse, not just narrative discourse on the basis of four pairs of parameters: agent orientation, contingent temporal succession, projection (i.e. future orientation), and tension. Within this schema, Longacre's approach pote n tially yield s si x teen discourse types (Clendenon 1989; Longacre 1996). Dooley and Levinsohn interpret the primary features of Longacre's categorization schema as follows: Longacre's broad categorization makes use of plus and minus values for a set of four features. Two of these features contingent temporal succession and agent orientation can be taken as primary, and serve to identify the four broadest categories. C ONTINGENT TEMPORAL SUCCESSION refers to a framework "in which some (often most ) of the [discourse] events or doings are contingent on previous events or doings" (p. 9)The second primary feature, AGENT ORIENTATION refers to whether the discourse type deals with "events or doings" which are controlled by an agent (one who performs an action), "with at least a partial identity of agent reference running through the discourse." (emphasis added; Dooley and Levinsohn 2001) 23 The diagram in Figure 2 3 presents the four categories of discourse genre generated from these two primary featur es: From this diagram, then, I further note that t he content of narrative discourse is most broadly and universally marked by the dual values of agent orientation and contingent succession 24 Longacre further posits that the expression of these two primar y generic qualities of narrative discourse, that is, of agency in contingent temporal succession, typically reflects the following template for narrative development : stage, inciting incident, mounting tension, climax of tension, and release of tension (1996). Strategies for the actualization of agency and contingent temporal succession, and of "mounting tensions,
77 climaxes, and denouements" are of course, culture specific, and legion. An outline of the culture specific "strategies" of Baka narratives follow. Baka Narrative Discourse Features Both Kathleen Higgens (1981) and Yves Le! onard (2003) have identified and described the presence of narrative discourse among the wide variety of Baka verbal performance types. Higgens' description focuses on the more global concerns highlighted in Longacre's narrative (development) templat e "; Le! onard's description following Levinsohn's model gives a more detailed description of the grammatical devices of Baka narrative, though always with a view to explaining how such devices inform larger formal functional and pragmatic purposes Doole y and Levinsohn's model, as presented in Analyzing Discourse (2001) distills a number of Longacr e's basic theoretical concepts, synthesizes them with the work of other analysts, and reformulates them in a template to which analysts can refer for the disco very of some of the most common worldwide strategies in realizing narrative discourse, oral narrative in particular As stated earlier, Le! onard's analysis, for the most part, builds on Levinsohn's model. A minimal, clause level s ummar y of Le! onard's analysis is as follows. The principal grammatico pragmatic s trategies of traditional Baka oral narrative s are ( 1) variations in the default order of clause constituents, ( 2) devices that give prominence to focal clausal constituent s ( 3) devices that foreground or background narrative events, ( 4) devices that highlight climaxes, significant developments, or key assertions, ( 5) varieties of cla usal connectives or markers that coordinate sentences that describe the main events, describe successive events performed by the same subject, or describe new development or new material, ( 6) particular participant referencing, and
78 ( 7) distinct devices for the reporting of embedded narrative conversation (Le! onard 2003 and Dooley Levinsohn 2000). Each strategy may be realized through a ny number of a variety of grammatical, rhetorical, or poetic devices : for example, tense markers; aspect makers; complement izers; relative clause markers; cataphoric, exophoric, and anaphoric markers; repetition; ideophones; conjunctions; demonstratives; and developmental markers. Some of these devices may function in isolation, though more often they complement others. Thes e same devices may also function one way in one clause, and another way in another clause. Repetition, for example, "can bring emphasis, it can mark the progression of a story, or can slow down the story before a climax ( Le! onard 2005:12). Each narrative device, then, c ontributes in distinct and multiple ways to the realization of those common narrative qualities that typically focus on agency and contingent temporal succession expressed and developed in mounting tensi ons, climaxes, and denouemen ts." Before concluding this section on Baka narrative generalities, three other narrative performance devices should be briefly highlighted: first, opening and closing formulas; second, narrative aspect;' and third, song. First : Baka storytellers most often, though not always, indexically frame the beginning and/or ending of a traditional li ka!n"! or one of its episodes with all or part of a traditional call and response exchange with the participating audience. The teller begins with li! ka! n"! p"!ngu "; the audience responds with p"!ngu "; then, the teller utters, e sasa ," and the audienc e responds again with sa! a (Le! onard 2005:4). (An example of this formula can be heard in the first accompanying audio file (Object 2 1. Audio
79 clip of li! ka!n"! formula (.mp3 24 KB)) and its gloss can be read in Chapter 5, Section 5.2.) Second: O ne of t he most essential grammatical features of Baka narrative is the narrative aspect marker / a / The narrative tense is decisive in marking and recognizing a li!ka! n"! 's storyline (i.e., its "contingent temporal succession"). Its essential role in the nar rative cohesion and development of li! ka!n"! is discussed in more detail in Chapter 4, Sections 4.2 and 4.3. Third: The principal thesis of Chapter 4 is the claim that li! ka! n song functions as a narrative discourse feature of Baka li! ka! n When Baka song is sung in the course of telling a Baka story, two kinds of verbal performance phenomena converge, wherein song is pragmatically subsumed in story. In this verbal performance context, story and song are not simply two "kinds" of speech acts, but one, wherein story, is the genre of verbal discourse performed, and song (along with speech) is one manner of producing that one verbal discourse type. 25 The semantic, grammatical, syntactic, rhythmic, melic, and timbric peculiarities of this discourse type ar e encoded, configured, intertwined, and produced to achieve the especially inter subjective effects of narrativized communication. "Any old" song, however, will not do. Li! ka! n songs may bear the generic marks of all Baka songs, as generalized in Secti ons 2.2.1, 2.2.2, and 2.2.3, but they also bear marks of their own. In the remainder of this chapter, then, I will briefly outline those features that set li! ka! n song apart from other Baka song forms, song features that (for the Baka) are somehow more referentially, iconically, and indexically suited for narrative performance. For li! ka! n songs are typically found to
80 exhibit particular discursive functions, poetic lines, musical rhythms, melodic textures, and voice registers. 2.4 Sung Baka Narrativ e Performance: Story's Song The particular discourse functions of li!ka! n songs will be described at length in Chapter 4. However, a summary preview of those functions is presented here in order to introduce and emphasize the integrity of the multiple c omplementry signs with which li! ka! n songs are imbued. Fundamentally, Baka li! ka! n"! songs are found to function in two essential signifying domains of narrative discourse: first, as an agent of narrative discourse cohesion (Chap. 4, Sec. 4. 2 ), and se cond, as an agent of narrative discourse development (ibid., Sec. 4. 3). Li! ka! n"! songs are found to effect discourse cohesion generatively performatively and textually (ibid., Sec. 4. 2). Intertextual cohesion of spoken and sung li! ka!n"! texts most oft en emerges as grammatical and semantic in nature. Li! ka! n"! songs also effect discourse development both contextually and climactically (ibid., Sec. 4. 3). Contexualizing devices typically constitute five contextualizing information type s" (ibid, Sec. 4. 3.1), while climactic devices are recognized according to five common marks of climactic development (ibid., Sec. 4.3.2) While it is likely that songs are largely experienced as undifferentiated wholes, they are, nevertheless, complex (semiotic) wholes (see also Turino 1999:237 245, 2000:174 177). Thus, when describing the multiple potential discourse functions of li! ka!n"! song in li! ka!n"! story, it should be remembered that li! ka! n"! songs are actually
81 ensembles of signs whose effects might only be signified by certain components of the song. Recognition of this componential condition goes a long way toward expl aining li! ka!n"! songs' multiple discourse functions (Chapter 4). Li! ka! n"! song is not only typified by common textual discourse functions, but also by common poetic and musical characteristics. The most fundamental poetic feature of li! ka!n"! song is i ts poetic line, which according to my data sample tends to be limited to either eight or sixteen pulses per line. (Twelve or four pulse lines are absent; see Sec. 2.2.1). Other characteristics of li! ka! n"! song's poetic line include an average of 11 wor ds per line (ranging from 7 to 16 words per line), with an average 21 syllables per line (ranging from 20 to 22 syllables per line). The most common syntactic, phonological, and lexical poetic devices include iterative verses and verse segments, grammatic al substitution, vocables, and interlinear lexical repetition. There may be as few as one type of vocable per song or as many as seven. All such poetic devices are commonly derived from an average inventory of 14 words per song, some songs with as few as 4 words, others with as many as 24. Chapter 3 presents a detailed inventory of all the poetic devices thus far identified in a representative sample of my field recordings. For a better understanding of the distinctives of li! ka!n"! song text poetics, a comparative reading of Chapter 3 is recommended. For now, however, I move on to describe briefly the specific polyrhythmic patterns most often performed and associated with li! ka! n"! song. Not all sonic rhythms performed in li! ka! n"! song are vocally pro duced. Some are manually produced. The voices of people ( li bo ) are commonly accompanied by the sounds, or "voices" of drums ( li ndu! mu! ) and other percussive instruments. Specifically,
82 two percussive polyrhythms indexically key li! ka! n"! song performan ce. Excerpts of each one can be heard in the following two audio clips: Object 2 2 and Object 2 3, respectively (Object 2 2. Audio file of rarer li! ka! n"! rhythm (.mp3 705 KB) and Object 2 3. Audio file of typical li! ka! n"! rhythm (.mp3 1.1 MB)). The (pr imarily) simple meter pattern heard in the audio clip linked to Object 2 2, is rare according to my field recordings and recollecton and is not exclusively associated with li! ka!n"! song, but may be used in other song styles as well. For these reasons, I will forego its description for a later study. By far, the li! ka! n"! song rhythm ( kole na 'e! na li! ka! n"! ) that is performed most often is that which is heard in the audio clip linked to Object 2 3 and transcribed in Figure 2 4. The transcription in F igure 2 4 represents an archetypal composite of this traditional poly rhythm It does not account for any of its common variations, for my present intentions are only to highlight the most typifying features. The entire rhythm's traditional cycle of eight pulses is executed once or twice in the course of one verse (or song line) depending on whether the song line 's meter' is a sequence of eight pulses, or, as is more common, sixteen pulses. The rhythmic hierarchy of this pattern is built upon a pulse ran ging from 118 MM to 148 MM with a mean tempo of 125 MM. The tempo for each song not to mention the variations in the rhythm's execution significantly informs not only the character of each song, but the rhythm, pace, tone, and development of each narrativ e. A pulse is most commonly subdivided into some form of triplet rhythmic figure, though it is not uncommon especially in melodic rhythms to subdivide pulses into duple figures.
83 The li! ka! n"! song rhythm does not need to be actuated. Song performances may be executed "without the voice/sound' of drums" ( nde$ li ndu! mu! '). 26 A minimal rhythmic framework, however, is often actuated by the clapping of ( gb" $ gb"! kole t# kpa$ o$ ') at l east one of two complementary rhythmic patterns: kole a l#$l#! t# kpa$ o$ ("the smaller (infant) clapping rhythm") (as shown in the first staff of Figure 2 4), or kole a ny#$ny#! t# kpa$ o$ ("the mother clapping rhythm") (as shown in the second staff). Kole a l#$l#! t# kpa$ o$ is most predominant, and quite often occurs alone. 27 By it, li! ka! n"! song is indexically "keyed." When a song performance includes two drums, the kole a l#$l#! t# kpa$ o$ is further reinforced (as seen in the third staff of Figure 2 4) by the homologous dynamic accents of kole a l#$l#! t# ku'u! that is, "the smaller rhythm of the (smaller) ku'u! (hand drum)." Still more emphasis occurs during larger performance gatherings, when the basic framework of the smaller drum rhythm is mim icked in the mba! nda a rhythmic action performed by striking a large shaft of raffia bamboo which is laying on the ground with one or two smaller sticks of (raffia) wood. Yet in complementary contrast with the emphatic l#$l#! rhythm, Baka hand drummers a dditionally index li!ka! n"! song performance with kole na ny#$ny#! t# mo! kinda that is, the grander mother' rhythm of the larger mo! kinda hand drum' (as transcribed on the fourth staff of Figure 2 4). The signal of this seemingly "single" drum rhythm e xecuted as it is by one hand drummer on a large, single headed cylindrical wooden drum is actually distinguished by (1) a pattern of two contrastive levels of dynamics (e.g., "accented" and "unaccented") patterning with (2) a pattern of three contrastive t imbres, that is, a low compact and sustained tone, a high compact and sustained tone, and a high diffuse and clipped tone (again, see the fourth staff in Figure 2 4). No other sonic structure associated with
84 li! ka!n"! song so potently indexes li! ka! n"! song experience as does the li! ka! n"! song rhythm ( kole na 'e! na li! ka! n"! ) whether that rhythm is performed by a polyrhythmic percussion ensemble or a quiet, lone hand clapper. Before closing this section and chapter, one generality concerning li! ka!n"! song melody should be highlighted. I am only able to describe certain macro level melodic characteristics because as stated in Section 2.2.2 of this chapter my analysis of the systemic nature of critical micro level structures, such as melodic intervals is not yet complete. Thus, most generalities concerning particular song genres would be premature. Nonetheless, it is still possible to address the question of who typically sings what elements of li! ka! n"! so ng melodies. As in most but not all Baka song genres, li! ka! n"! song melodies are never performed solo, but by two or more voices, that is, by a choir of voices. Furthermore, li! ka!n"! song as with most genres always exhibits two or more melodic voice parts. Thus, two or more voice parts are typically distributed among two or more singers' voices. As set out earlier, in multi part voice textures, two categories apply: wa! nja! mba a nd wa! tu$ ko! And, while there is only one wa! nja! mba voice part thoug h more than one person may sing it, depending upon the song style there is often more than one distinct wa! tu$ko! melodic voice part. In the case of li! ka! n"! song, however, the wa! nja! mba voice part is normally performed by the storyteller singer, thou gh others may occasionally join. The wa! tu$ ko! voice parts, then, are performed by the participating audience choir (e.g., Chap. 5, Sec. 5.2). Li! ka! n"! song melodies may be segmented and organized as alternating, overlapping, interlocking, and/or polyph onic sequences, and then performed by responsorial, antiphonal and/or polyphonic choirs of voices. The
85 choirs of many other Baka song genres are commonly homogenous choirs of either males or females, young or old; but li! ka! n"! song exhibits no such gend ered or generational homogeneity. All attending voices, regardless of gender or age, join the choir. 2.5 Summary In summary, "everyday" speech and conversation constitute the "default" formalizations of Baka verbal discourse. Alterations in these ordin ary patterns of verbal discourse whether rhythmic, syntactic, lexical, phonological, or semantic alterations signal extra ordinary verbal discourse, effecting extra ordinary purposes. Baka song, story, and story's song, each in its own way, signal specia l verbal performance. As a complex sign, sung verbal performance ( 'e! ), regardless of genre or style, is poetically organized, and thus, typically constituted by a distinctive set of rhythmic, syntactic, and intonational formalizations. Extra ordinary p honological, lexical, and semantic formalizations are often employed as well. The content of narrative verbal discourse, on the other hand, is typified by the dual primary features of agent orientation and contingent temporal succession. In Baka li! k a n"!, agency and contingent succesion are actualized, unified, and developed by a variety of extra ordinary grammatical, rhetorical, and poetic devices, song being one such poetic device. When song ( 'e! ) is sung in the course of telling a traditional Bak a narrative ( li! ka! n"! ) Baka speakers specify it as 'e! na li! ka!n"! ( i.e., li! ka! n"! song ") As a type verbal performance, li! ka! n"! song not only shares characteristic marks of both sung and narrated verbal performance types but also bears distinctiv e song and narrative
86 features of its own. The most conspicuous rhythmic features are (1) the typical polyrhythmic percussion accompaniment kole na 'e! na li! ka! n"! (the li! ka! n"! song rhythm) and (2) metered poetic lines of eight or sixteen pulses. This li! ka! n"! rhythm and its corresponding line, in turn, inform the rhythmic segmentation of a li!ka! n"! song's lyric. Iterative, non contrastive verses and verse segments predominate. Simultaneously, m elodic formalizations of li! ka!n"! lyrics are ordinar ily distributed among two or more voice parts which in turn are (1) organized as alternating, overlapping, interlocking, and/or polyphonic sequences, and (2) performed by responsorial, antiphonal, and/or polyphonic choirs of voices. But unlike many Baka s ong styles and genres, the vocal timbres of li! ka! n"! song choirs are "mixed," that is, anyone in attendance may sing, regardless of gender or age So finally, complexes of rhythmic, syntactic and melodic signs that constitute and characterize li! ka! n"! song do not simply encode sung verbal perform ance, but also and more significantly serve narrative verbal performance. Li! ka! n"! song, then, is inherently a discourse feature of li! ka! n"! narrative In particular, li! ka!n"! song effects narrative cohesion and narrative development. Narrative cohes ion through song is achieved generatively, performatively, and inter textually. Narrative developmental, on the other hand, is contextual and climactic in nature. Elaborations on all of these generalities are set out in the chapters that follow.
87 Bak a Term(s) Etic Description Typical Contextual Associations e" na basu# ka youth recreational dance songs evening recreation e" na m $"ba" si" youth recreational dance songs evening recreation be" na li" ka" n %" traditional story's songs hunting/funeral e" na Mbo" a" mbo" a" dance songs w/ Mbo" a" mbo"a "" funeral e" na nga nga" divination songs divination / healing ritual e" na mbomba divination fire songs divination/healing ritual e" na J $ ngi" dance songs assoc. w / J $ ngi (spirit) spirit initiation e" na & e s %" nj%" dance songs to dispel malevolent spiri ts sorcery (Brisson 1999:84) e" na di" nda spiritual/ceremonial dance songs to find one lost in forest ( Brisson 81) e" na & e di" o" dance songs assoc. w/ sorcery /h ealing sorcery; healing; funeral e" na mange# le# bo" ritual song(s) to dispel ancestral spirits wake, funeral (81) e" na mo" ko" ndi" (dance) songs assoc w / socerer spirit healing (ibid.) e" na ebu" ma" dance songs w/ ebu" ma" spirit sorcery; funeral e" na mo" ling $" trapping songs trapping e" na ma" ka/na ndando hunting songs (s ung by women) for a successful hunt e" na ye # yi women's ritual association songs hunting; healing; benediction; sorcery e" na ma" ka /na s% women's hunting songs ( men's return ) great hunt celebration be" na ma" ka na gee" ya" mens elephant hunting son gs (return ) hunting celebration be" na ma" ka na gee" e "! o !o men's gorilla hunting songs hunting be" na ma" ka na gee" se" ko" men's chimpanzee hunting songs hunting be" na ma" ka na gee" pa" m $" men's wild pig hunting songs hunting e" na ngu# ma women's dam fishing song fishing e" na e" ka" circumcision ceremony dance song male circumcision e" na bo di# ndo# lullaby night /sickness e" na mu# ko# dance songs celebrating birth of twins birth celebration e" na s %"l % children's game song recreation e" na ku# lu children's vine swinging song recreation e" na nda a Komba church song Catholic Church celebration e" na a" si# k$ song s accomp. by bow harp recreation e" na aita song s accomp. by frame harp recreation/personal expression/narrative e" na ng %"mbi na" lo song s accomp b y plucked idiophone recreation/personal expression e" na ng %"mbi na" pe" ke song s accomp by harp zither recreation / personal expression / narrative e" na li# ngbidi song s accomp double string bow harp recreation/personal ex pression/narrative e" a ngu# ma/ba" le" song s accomp b y water drumming fishing/recreation/wash ing e" na p %"ki" honey gathering song s honey gathering e" na mo# ngulu song s while constructing leaf hut habitat construction Figure 2 1 List of Baka song field recordings
88 Nominal Term Collocating Verbs A`j`! Mntm Fknrr! A`j`! Udqa Fknrr mfn l`! discussion, word l! mfnl`! make, speak, have klt! conversation, discussion l! make, (be in) '_d (j` kn! (formal) counsel from elders l`m`! explain, give kd vt! (personal evening) advice s" 'od( give ln rhln! proverb, recitation, account, narrative, monologue (w/o song) jo c" s relate, recount, bring fa mfnmfn allegory, proverb, saying, dictum kp recount, tell la` kh comparison, parable gb mfn l`! strike, speak to, speak in k # j` m! story, legend, fable, history, chantefable jo recount, narrate, tell d! Song d sing Figure 2 2 Lexical s ets of c ommon Baka s peech a ct t erms Agent orientation + + narrati ve procedural Contingent temporal succession behavioral expository Figure 2 3. Fundamental categories and features of discourse genres
89 Figure 2 4 Composite t ranscription of l i ka!n"! rhythms (my transcription) ; Pulse = 125 MM 1 Reference to Robert Brisson's and Daniel Boursier's work especially the Petit Dictionnaire Baka (1979) could be consulted to expand or modify this list Brisson has told me in personal conversation ( 2006) that plans are in the works for Susanne Frniss to supplement his 2002 version of Petit Dictionnaire Baka with additional musical entries. In either case, corresponding audio recordings would be needed to corroborate any conclusions that might be dr awn from those lexical entires. 2 Fox and Feld initially frame their 1994 survey of the literature on the relationship of music and language "in terms of four major predications: music as language, language in music, music in language, and language about m usic." Further, they discuss "attempts to integrate these four themes through semiotics and sociomusicology" [prominent in the 1980s] and through "the ethnographic approach to intersections of language and music." i.e. "a refigured anthropology of the spe aking and singing voice to key issues in contemporary [ circa. 1994] social theory" (26). 3 Frniss' text reads: Pour la musique aka, il se dgage un systme plusiers niveuax de pertinence: le paramtre principal est l'ordre de succession des degrs dans un patron mlodique donn au sein d'une chelle pentatonique quelconque dont la seule constante est l'absence de l'intervalle de demi tonLe second niveau de pertinence, qui concerne la ralisation des chants polyphonic, est fond sur l'agencement vertical des intervalles en certains points prdtermins. Le droulement des lignes mlodiques, par del la license de ralisation qu'il admet, est nanmoins tributaire des points d'ancrage de la polyphonie" ( 1993:136, 1 37 ).
90 4 Fitzgerald, Dan. 1989. An Emic A nalysis of the Tewa Peublo Indian Melodic System. (unpublished). I plan to reformat and publish this study in a forthcoming website. 5 Terms such as major or minor seconds or thirds, or perfect fourths or fifths do not, of course, refer to the fixed conventions of western music theory, but simply indicate etic approximations of acoustic distance. The culture specific identities of such symbols is the subject and first step of emic analysis. Further, the melodic intervals of unyodeled song styles are normally articulate d solely through a head voice laryngeal mechanism. The boundaries of melodic intervals in yodeled song styles, however, are marked by alterations between chest voice and head voice tones. Still, the predominanc e of major seconds, minor thirds, and perfect fourths and fifths applies to both unyodeled and yodeled styles. However, head voice tones most commonly terminate intervals of a perfect fifth, or, not uncommonly, perfect fourths or minor sixths. The emic status of these four intervallica l ly sim i lar tones is yet to be determined. 6 This observation assumes that each voice sings only one melodic part per song (cf. Frniss 2006 ). 7 Arnold Whittall Whittall, Arnold. "melody." The Oxford Companion to Music Ed Alison Latham. Oxford Music Online 14 Oct. 2010
91 12 An excerpt of Frniss's definition of the yodel technique: "Yodel [( di! ye! i$ in Aka)] is sung above all the other parts by women. It is deter mined by the yodel technique, a constant alternation between laryngeal mechanisms 1 and 2, which is commonly called "falsetto," or head voice". It consists of melodies of mainly wide intervals and uses specific vowels correlated to the two yodel registers : low yodel register mechanism 1 [with] open vowels as [e, a, o]; high yodel register mechanism 2 closed vowels as [i, y, u]" (Frniss 1991, cited in Frniss 2005:11). 13 These shared vowel systems might suggest similar vocalic constraints in song pe rformance. The implications of such an assumption should not be taken to far given the fact that the Baka and Aka do not share the same set of phonemic consonants or tones, not to mention grammatical systems. The Aka, for example, have two phonemic tones while the Baka have three. 14 In the case of the Aka, the articulatory effect of the yodel technique can often even condtionally create its own set of vowels (see Frniss 1991:175). 15 Literally, "the one who counts" (Frniss 2006:175). 16 Literally, "yodel (ibid.). 17 Aka u se of the "head" voice does not oblige the singer to sing with words. For example, other voice parts, like the ngu$ e$ wa! le$ mbo! and o! s s most often articulate wordless syllables exclusively with head' voice. 18 Just as Frniss point s out that the Aka rarely "talk" about distinctive voice parts and their functions (see Frniss 2006:176), neither do the Baka. In fact, I hesitate to claim that all Baka speakers share the same lexical expressions when speaking of voice part performance. It is highly likely that the word wa! nja! mba is widespread (see Brisson 1992:465), but use of the word wa! tu$ ko! would still need further verification. 19 See Miller DG, Schutte HK. "Physical definition of the the flageolet register ." J Voice. 1993 Sep t. ; 7(3):206 12. School of Music, S yracuse University, New York. 20 Brisson claims that the etymology of the word yeyi! is formed from the lexemes ye (to like') + li (voice of,' or sound of'), meaning to like the voice sound of,' as in e$ a! ye li ng"mbi ( "He likes the voice sound of his guitar" ) ( see 19 9 9 :626 ; 2002 ). 21 I was initially motivated to adopt Dooley and Levinsohn's approach because SIL colleagues Yves Le! onard (2005) and Kathleen Higgens (1981) had both, in one way or another, at one time or another, based their complementa ry discourse analyses of Baka traditional narratives on the models of Longacre, Dooley, and Levinsohn 22 More specific genres often involve other textual properties. Drama, for example, may be a narrative according to broad genre, but one that is presen ted in the form of
92 dialogue (Chapter 1) and is typically written (Chapter 4) for live presentation. Letters are written discourses and may be of any one of several genres. Jokes are typically oral narratives with a particular goal (humor) and a specific r egister of speech, and so forth (Dooley and Levinsohn 20 01) 23 Besides these two primary features, Longacre discusses two further ones: projection and tension. P ROJECTION in its + value "has to do with a situation or action which is contemplated, enjoine d, or anticipated, but not realized"; prophecy is + projection narrative, stories are projection, and so forth. T ENSION "has to do with whether a discourse reflects a struggle or polarization of some sort". Narrative can be + or tension, as can scient ific articles (depending on how polemic they are), etc. (Dooley and Levinsohn 2001: 9, 10). 24 Procedural discourse ("how to do it, how it was done, how it takes place") is marked by contingent succession, but not agent orientation (since "attention is on what is done or made, not on who does it); behavioral discourse (exhortation, eulogy, some speeches of political candidates, etc.) clearly exhibits agent orientation, but not contingent succession (since "it deals with how people did or should behave"; an d expository discourse (budgets, scientific articles, etc.) is marked by neither feature (cf. ibid). 25 For an introductory discussion of a discourse's "manner of production," refer to Dooley and Levinsohn 2000/2001:1 6. 26 The dependent lexeme li can be translated here as voice of,' or sound of.' In some contexts, it commonly translates as language of' (cf. Brisson 2002:259). 27 A variation of kole a l#$l#! t# kpa$ o$ one that simply displaces the actuation of the entire pattern by two pulses is often peformed complementarily.
93 CHAPTER 3 BAKA SONG TEXT POETICS: AN INV ENTORY OF POETIC DEV ICES 3.1 Introduction: Sung versus Unsung Verbal Discourse In Section 2.1 of Chapter 2, I reported that Baka speakers identify a variety of v erbal discourse types The t able in Figure 3 1 a simplified version of the table in Figure 2 2 d isplays a set of lexical entries of nine of the most prevalent Baka speech act terms. As I explained in Section 2.1, the two terms ngo! ma! (discussion') and l"mu! ( conversation') identify "ordinary" verbal acts. The other seven terms ka! lo! le! wu! mo! simo! gb"!ngongo mba! li li!ka! n"! and 'e! identify "extra ordinary" verbal acts. Still, the Baka do not formulate abstractions about "ordinary" versus "e xtra ordinary" speech acts. In this chapter, however, I have empirically abstracted several of the distinctive features that constitute the particular (extra ordinary) verbal act that the Baka identify as 'e! that is, "song." My analysis is inductively developed from the notion that there is one underlying function in all of the features that distinguish song from non song, that being a "poetic function." What follows, then, is (1) a brief description of what generally constitutes poetically organized d iscourse and (2) an inventory of the poetic devices thus far discovered in Baka song texts. In the opening statement of their essay "Poetry" in Alessandro Duranti's A Companion to Linguistic Anthropology (2004) Giorgio Banti and Francesco Giannattasio asse rt that "ethnographic research has shown that "poetic" forms and behaviors are almost universally widespread" (Banti and Giannattasio 2004:290). Further, they elaborate on the notion of "poetry" in this way.
94 It should be clear at this point that two diff erent levels of definition for poetry have to be taken into consideration: [The first level is ] a wider level of poetic procedures whose general features broadly correspond to what [Roman] Jakobson [(1960)] said about his poetic function of language. Th ey not only characterize poetry proper, but also include magical spells, prayers, and ritual discourse, in proverbs and children's games, and so on, right up to today's advertising jingles and political slogans, and of course, in the different forms of cha nt and song. They all share the fact of being special, not ordinary speech (293 294) [The second level is ] a narrower level of poetry in the strict sense that is automonously defined by each culture in the course of its history on the basis of its own choice of genres, specific contents, ways of production, functions, occasions of performance, and aesthetic and social values. (ibid.) The scope of the analysis presented in this chapter is limited to Banti and Giannattasio's "wider level of poetic proced ures ," that is, that level which is roughly modeled after Roman Jakobson's notion of the "poetic function." In the field of twentieth century semiotics a field first engaged in the 1960s by musicologists and ethnomusicologists alike 1 the writings of Roman Jakobson (1896 1982) bridged the two major streams of semiotics that initially flowed from the pioneering works of Ferdinand de Saussure (1857 1913) and Charles Sanders Peirce (1839 1914) (Cobley and Janz 1997:144). 2 In treating song as sign, there are two kindred articles from Jakobson to which I have gravitated: one, because it gives specific attention to poetics as a semiotic process ; the other, because it addresses music as a semiotic process.
95 In his oft cited essay "Linguistics and Poetics" Jakob son justified and modeled a linguistic treatment of poetics, hitherto reserved as the property of the literary studies. He wrote, This separation of the two fields from each other is based on a current but erroneous interpretation of the contrast between the structure of poetry and other types of verbal structure. Many poetic features belong not only to the science of language but to the whole theory of signs, that is, to general semiotics. This statement is valid not only for verbal art but also for all varieties of language since language shares many properties with some other systems of signs or even with all of them (pansemiotic features). (1960) 3 In 1973, in his article "Le language en relation avec les autres systmes de communication," Jakobson lat er characterize d music as a semiotic system, one in which "introversive semiosis" (that is, the reference of each sonic element to the other elements to come) predominates over the "extroversive semiosis" (i.e., the referential link with the exterior world ) (1973:99 100). 4 Common to Jakobson's semiotic notions of both poetics and music is an emphasis on sound "figures of sound" and "sound symbolism." Yet even while Jakobson championed a call to continually and more deeply give sound (as sign) its due, he s imultaneously urged the investigation of yet another, more fundamental dynamic at work that is, "the principle of equivalence." Fully stated, it reads, The poetic function projects the principle of equivalence from the axis of selection into the axis of combination (1960:358) Restated in 1968, Jakobson wrote, "i n poetry similarity is superimposed on contiguity, and hence "equivalence is promoted to the constitutive
96 device of the sequence" (1966:602). 5 Literary scholar, Jonathan Culler tried to reduc e Jakobson's formulation this way: In other words, the poetic use of language involves placing together in a sequence items which are phonologically or grammatically related. Patterns formed by the repetition of similar items will be both more common and more noticeable in poetry than in other kinds of language. (1975:56) 6 Jakob son, however, would not agree to limiting his poetic principle to traditional literary categories of phonology and grammar. No doubt, verse is primarily a recurrent "figure of sound ." Primarily, always, but never uniquely. Any attempts to confine such poetry conventions as meter, alliteration, or rhyme to the sound level are speculative reasonings without any empirical justifications. The projection of the equational principle int o the sequence has a much deeper and wider significance. Valery's view of poetry as hesitation between the sound and the sense' is much more realistic and scientific than any bias of phonetic isolationism. 7 Russian folk poetry (for example) can be fruitf ully analyzed on all linguistic levels phonological, morphological, syntactic, and lexical: we learn what elements are conceived as equivalent and how likeness on certain levels is tempered with conspicuous difference on other ones. (1960:369) The inventor y of Baka song text poetic devices identified in this study are recognized within and across five musico linguistic poetic domains: rhythmic, syntactic, phonological, lexical, and semantic domains. In addition, the recognition of the poetic function in th ese five domains is guided not only by an abstract search for Jakobson's "principle of equivalence," but also, more specifically, by Peirce's "second
97 trichotomy" of semiotic concepts (Turino , 1999:226 229). In it, Peirce posits that there are "thre e ways that a "sign and [its] object' are related in perceivers" (emphasis added; ibid.). First, a sign and object may be related "iconically," whereby sign and object are said to be related through "resemblance." Second, sign and object may be related "indexically," wherein sign and object may be related through "co occurrence," or "association." Third, a sign and object may be related referentially, or "symbolically" as Peirce defines it through language (i.e., words) (ibid.). As it relates to poetic analysis, foreknowledge of these three distinct ways of relating sign and object aids poetic analyses in that it lends greater specificity to how Jakobson's "principle of equivalence" might be potentially constituted. Equivalencies among song signs that is, the ways that signs and objects might "relate" are not expected to be of one order, but of many, though iconic, indexical, and symbolic in nature. In addition to semiotic methodologies, the identification of the constitutive poetic elements whether rh ythmic, syntactic, phonological, lexical, semantic, or other is further aided by familiarity with (1) conventional analytical methods of music theory, linguistics, poetics, and literary criticism, and (2) pertinent local terminologies. Indeed, though t he theory and methodology employed in this analysis is rooted in semiotics, semiotic terminology will not be prevalent in the sections that follow. Rather, descriptive terms of poetic constituents are primarily drawn from the conventions of l inguistics, poet ics, ethnopoetics, literary criticism, and music theory. I have chosen to employ conventional terms simply because they are more widely understood than semiotic terms.
98 With this framework in mind, the plan for the rest of this ch a pter is first, to introdu ce the scope and sources of the Baka song texts under study; second, to present an overview of all Baka song text poetic devices found to date; third, to examine in detail the rhythmic devices at play in nine select Baka songs (Sec. 3.3); fourth, to descri be all of the poetic devices found at play in one particular Baka song (Sec. 3.4) ; and last, to summarize the cumulative effect of that one song's multiple, interconnected poetic signs (Sec. 3.5) Song text sources. In Section 1.6, I reported the scope an d location of the ethnographic data that I collected during my initial period of fieldwork f rom November 1996 through May 1998 During that period I collected ov er f our hundred field recordings of Baka song in five Baka camps in the Eastern Province of C ameroon ( maps in Fig. 1 6). From this corpus, preliminary textual transcriptions were made of approximately seventy five song texts representing a cross section of potentially thirty to forty song types (Fig. 2 1) From these seventy five song texts, nin e contextually divergent texts were selected for more detailed analysis (Fitzgerald 2003) These nine detailed song text transcriptions represent songs with a variety of cultural themes, composition histories, and performance practices. Compositely summ arized, they are songs both ancient and contemporary composed by individuals and groups of Baka people from all age groups and genders. Furthermore, they represent songs performed by individuals and groups (homogenous or mixed) of all genders and age grou ps, with and without instrumental accompaniment, with and without dance, antiphonally and polyphonically, staged and naturally, at night and in day, casually and formally, in solitude or in community, as sacred or profane.
99 3.2 Inventory of Baka Song Text Poetic Devices Figure 3 2 presents an inventory of all of the poetic devices thus far identified in my analysis of the nine selected song texts. The inventory is guided and organized by the etic perception of signifying patterns in five distinct but inte rrelated poetic domains, that is, rhythmic, syntactic, phonological, lexical, and semantic domains With some elaboration particularly in regard to the systemic distribution of these poetic devices the entire inventory may be read as follows: First, i n the domain of rhythmic poetic devices, all Baka song texts have the poetic devices of pulse, tempo, meter, and line. Second, i n the domain of syntactic poetic devices, all Baka song texts include the device conventionally termed verse ; most also have ve rse segments and some have stanzas All verses, verse segments, and stanzas are at some point repeated and most become v aried C ontrastive verses and verse segments are most commonly exhibited; derivative verse types occur less so E njambment both syn tactic and rhythmic in nature is not an uncommon syntactic poetic device. Third, i n the domain of lexical poetic devices, it is quite common to find intra and inter linear repetition reiteration or reduplication of phonemes, ideophones and words. Ho monyms occur less frequently, but certainly are not rare. Fourth, t he phonological functioning of poetic devices is too diverse to give much elaboration here. A simple listing of these devices is adequate for our present purposes. The phonological devic es, then, include the use of vocables, vocable repetition, vocable substitution, ideophones, assonance, consonance, alliteration, borrowed words, vowel lengthening, and sprechstimme ( sometimes referred to as rhythmic speech ). Fifth, and last, is t he fun ctioning of poetic devices on the semantic level
100 This also is too diverse to offer much elaboration at present. Simply listed, semantic level devices include the use of ideophones, reiterated ideophones, reduplicated ideophones, similes, metaphors, meto nyms, lexical reiteration of metonyms, anthropomorphism, meronymy, hyponyms, semantic parallelism, complimentary syntactic semantic parallelism, syntagmatic paradigmatic grammatical substitution, loan word synonymy, and ellipsis The preceding plain read ing of the entire inventory is only intended to present a quick profile and overall impression of the kinds of poetic devices performed in Baka song texts. I have, however, documented a description of each device listed in the inventory, but the presentat ion of that lengthy study lies outside the scope of this chapter What follows, then, is first, a description of all the rhythmic poetic devices found in the selected corpus of nine songs, and then second, a description of all of the poetic devices recogn ized thus far in one particular Baka song. 3.3 Rhythmic Poetic Devices 3.3.1 Line T he primary distinctive poetic feature of sung Baka discourse is its musically metered line. 8 It must be emphasized, that the poetic nature of a line is not so much verbal, as it is musical. More precisely, a line is a rhythmical unit (an organized unit of time), not a syntactical unit. 9 A Baka song line, however, does not express rhythm formed of patterned prosodic or semantic units (as in Greco Roman and Japanese meter, or Semitic parallelism), but rhythm minimally formed from a continuously recurring sequence of pulses Restated then, the primary distinctive poetic feature of sung Baka discourse is the musical meter to which it is set.
101 The poetic line in all nine song texts in this study is musically metered. 10 Musical meter however, is not always apparent in conventional text transcriptions using Roman script. Meter of this nature is better represented by way of a music notation system 11 In lieu of music notation i n all the examples that follow, it is adequate for my purposes here p urposes decidedly linguistic and literary in emphasis to ( 1) tacitly mark the bounds of a musically metered poetic line by the typographical convention of "stopping at the right margin an d returning to the left margin" 12 and ( 2) explicitly note both tempo and meter in the upper left margin of the first line. In addition to text examples, audio excerpts (clips) accompany most text examples in this section. 3.3.2 Pulse The recognition of pul se is fundamental to an authentic understanding of a song line's nature. A pulse (a periodic unit of time) is the minimal rhythmical unit of a line. However, as the pulse in any given song or song line is not always actuated, a cultural outsider may no t immediately experience its presence. To a cultural insider, however, its presence, actuated or not, is still experienced. As Vida Chenoweth writes: In any particular composition not all the pulses are necessarily actuated, but because of their regula rity, all of the pulses are anticipated and felt by the listener [i.e., a cultural insider]. (1972:99) This phenomenon of perception was corroborated in the transcription of the data under question. The pulse in Example 1, for instance, is clearly actuat ed by the simple clapping of hands and was readily perceived and transcribed by me a cultural outsider. I iconically represent it here with numbered quarter notes equidistantly plotted above the text, roughly synchronizing pulse and text. 13
102 Example 1: E lephant Hunting Song (Object 3 1. Audio file of elephant hunting song (mp3. 944 KB)) 8 pulses per line 0 2 3 4 5 5 7 8 q q q q q q q q '2`( 'g! _( x` !ln k!'mc( '2a( 'g! _( x` !ln k!x` !j+! mfnhjnmi`!' ( In another example, Example 2, the pulse was again readily perceived, but its actuation was less evident in the accompanimental rhythm of an aita a Baka seven string frame harp. (Again, numbered quarter notes are added to the original transcription to make the line's rhythmic character explicit.) 14 Example 2: S ong of a Young Boy Playing with His Parents (Object 3 2. S ong of a young boy playing with his parents (mp3. 932 KB)) 7"otkrdr"odq"khmd 0 1 2 3 4 5 6 7 q q q q q q q q '0( !!!!!!!_@l` + !!!!!v`!ahmi`#!kd#! L`l`m !!Hkr!ld!sqnlodms '1( !!!!! !!_@o`+! !!!!!v`!ahmi`#!kd#!jn+! mjn"ld Uq`hldms+!o`o` !!Hkr!ld!sqnlodms-!!'Ldld+ ( !lnh !! '2( @l ah`mbd !`#!atl` kd#!jn!' h d(!an' n(!vd# O`tuqd!cd!lnh !!'Dbntsdy(!snts!kd!lnmcd !!B&drs!uq`h!'pt&hkr!ld!sqnlodms!!ldld+!lnh ( "'3( !!!!!!!!mf``#!' d( !!!!!v`!ahmi`#!kd+ "'4( !!!!!!!!_`l`+! !!!!!!v`!ahmi`#!kd#+ In still another case, the presence of the pulse was not plainly evident, yet still suggested in the regularity of both the melodic rhythm and the accompanimental rhythm of the ng"!mbi na! pe! ke, a plucked harp zither. However, its emic timing was
103 initially unknown to me. As an outsider, I was not culturally competent to immediately respond to it. It was, however, eventually revealed to me when Baka friends began clapping and dancing to a recording playback of the song. The pulse is plotted below in Example 3. (An audio clip in Object 3 3 presents the original recording (Object 3 3. A udio file of solo song of lament (.mp3 952 KB)). The audio clip in Object 3 4 records a Baka speaker clapping to a playback of the original recording ( Object 3 4 A udio file of solo song of lament with clapping (.mp3 804 KB)). Example 3: Solo Song of Lament, with Harp Zither Accompaniment (ibid., i.e., Object 3 3 and Object 3 4 above) 1 2 3 4 1 2 3 4 q q q q q q q q '03/"LL( 'khmd"0( !!!!!! !!!!!Vt!vd!vt!vd' n(+ 'khmd"1( !!!!!Vt!vd!vt!vd' n(+ 'khmd"2( ! !!!!!k`#! !ln#!!` #!!rh`#: 'khmd"3( !!!!!ln# ln#!` #!mfnl a : 'khmd"4( !!!!!i ln#!!!` #!!!i#!!!!!a !! d!!k!!d-! Finally, there are those songs where the pulse is not actuated at all, where neither accompaniment nor dancing exist to realize it. Melodic rhythm alone is left to suggest it. Yet, should the melodic rhythm be characterized by syncopation, whi ch of course is not uncommon in sub Saharan Africa, the pulse may be all the more imperceptible to the uninitiated. This phenomenon was the case in song Example 4. It was not until the song was re performed months later in another context (with drumming and dancing)
104 that the pulse was made explicit. ( The audio clip in Object 3 5 presents an unaccompanied performance and the audio clip in Object 3 6 documents an accompanied version, with drums ( ndu! mu! '). ( Object 3 5. A udio file of male initiation song with out accompaniment (.mp3 1 MB) and Object 3 6. Audio file of male initiation song with accompaniment (.mp3 944 KB) respectively.) Example 4: Male Initiation Song (with the spirit, Jengi ) (ibid., i.e., Object 3 5 and 3 6 above) 16 pulses per line 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 '6`( '_h!x` ( 'v( d# !vd# mf`mf`!k`#!j ,,, ! !!!!!!!!!'_h!x` ,,,,, ( ..!`g !.!o`tuqd!cd!mntr!.!pth!.!bnlld!b`!. ! ..!Nm!u`!e`hqd!b nlldms+!`udb!Idmfh!bnlld!b`>.. '6a( ! r`#kd#r`kd' d( !!!!!!'_ (Lnla`' ( " ..!dmsqdy !. !!!Idmfh!'l`hsqd!ct!bntsd`t(!.. I have found, then, that the first step in the process of defining the rhythmic nature of a poetic line begins with emicall y distinguishing the minimal dynamic unit of rhythm, the pulse. Once competently determined, a pulse may then be described in terms of its tempo 3.3.3 Tempo Tempo is pulse rate. Detailed etic descriptions of song performance tempi are more properly the realm of musicological studies. Linguistic and literary conventions do not typically account for this element of a poetic line. I include tempo markings at the beginning of each of my linguistic transcriptions to underscore the extent of the musical cha racter of the Baka poetic line and to indicate, even if only slightly, the quality of movement and energy animating a song line. Data of performance tempi
105 might also become pertinent to linguistic description if correlations emerge between tempo and tende ncies concerning compositional word choice, word performance, or a lines text load. Plainly put, there are only so many verbal segments one can comfortably articulate in a given period of time. Thus, for example, poetic devices such as elision and ellip sis may tend to be found in songs of faster tempi. In my transcriptions, a lines tempo is described by way of the musical convention of metronome markings displayed next to the meter measure at the upper left margin of the first line. The t able in Figure 3 3 displays the various tempi in our song sample. The range of song line tempi in our sample varies fairly evenly between 105MM to 160MM, the mean tempo being 126MM. Personal recollection suggests that a more complete sample may extend these extremes, but not radically so. Even with a broader range, the mean tempo would not likely change significantly 3.3.4 Meter While pulse and tempo are fundamental dynamics in the temporal organization of a Baka song line, they do not define its boundaries. Pulse i s but a minimal rhythmic unit from which other larger units are constructed. 15 We recognize the organization of larger units in prevailing patterns of repetition that is, varyingly repeated segments of music, texts, and performance textures. In Baka song texts, patterns of syntax are patterned with patterns of rhythm. In most cases, syntax patterns coincide with line patterns. T his patterning defines a song lines boundaries. The number of pulses within its boundaries is termed its meter Meter then, is the length of a Baka song line, the measure of its primary organizing principle. 16
106 The t able in Figure 3 4 displays the distribution of the four meter types found in my data. The four meter types are 16, 12, 8, or 4 pulses per line. Each type is show n in Examples 5 8, respectively. Some example s excerpt two lines from a complete song text transcription in order to demonstrate coinciding patterns of line and syntax. Example 5: Dance Song (1) Returning from a Gorilla Hunt or, (2) during a Circumcision Rite ( Object 3 7. Audio file of dance song for return from gorilla hunt (.mp3 848 MB)) 16 pulses per line; pulse = 118MM 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 '02`( _@lt #! xdjd # xdjd #! sod rjd #! !!!!!!! j` # mfn #+ '02a( _@ xd j` # mfn #+ !!!!!! '_dd ( !!!!!!! j` # mfn #+! _` xd j` # mfn #! Example 6: Song Associated Jengi (Spirit of the Forest) (Object 3 8. Audio file of song associated with the spirit, Jengi (.mp3 976 KB) 12 pulses per line 105MM 1 2 3 4 5 6 7 8 9 10 11 12 '5`( '_d x( at k` a` # mf $' (! ! '_d x( '5a( '_d d d ! _d d _h !x n( '5b( ! _d cn # sn 's( o s $ l` #! cd' ( '_ ( v` st# cd' ( Example 7: (see Example 1) 8 pulse per line 125MM Example 8: (see Example 3) 4 pulse per line 140MM Within the sample, then, there are four songs with metered lines of 8 p.p.l. (pulses per line), three songs with 16 p.p.l., two songs with 12 p.p.l., and one song with 4 p.p.l.. Taken as a whole, the sa mple suggests the following preliminary summary of
107 Baka song line meter types: (1) lines are organized in metrical units ("lengths") of either 16, 12, 8, or 4 p.p.l.; 17 (2) the maximum line length allowable is 16 pulses; (3) the minimum line length allowabl e is 4 pulses; (4) half the line types have 12 or 16 p.p.l., while another half have 4 to 8 p.p.l.; (5) the average line length is 12 pulses; (6) the predominate (typical) song line length is 8 pulses. 3.3.5 Summary of Rhythmic D evices In summary, the rhyt hmic poetic devices of Baka song texts may be described as follows: The primary distinctive poetic feature of sung Baka discourse is its musically metered line. A Baka song line is essentially a rhythmical unit, an organized unit of time. The minimal rhy thmic unit of a Baka song line is a pulse. From a pulse other larger rhythmic units are constructed. In any given Baka song a pulse is not necessarily actuated, but can nonetheless be emically recognized because of its regularity. The measure of its re gularity is its tempo. Baka song tempos range from 105MM to 160MM, with a mean tempo of 126MM. A song line is minimally formed from a continuously recurring sequence of pulses. The organization of these sequences is recognized in patterns of repeated se gments of music, texts, and textures. In Baka songtexts, patterns of syntax are patterned with patterns of rhythm. In most cases, syntax patterns coincide with line patterns. T his patterning then, defines a song line's boundaries. The number of pulses within its boundaries is its meter, that is, its relative "length," the measure of its primary organizing principle. Baka song line meter types are organized in units of either 16, 12, 8, or 4 p.p.l..; the average meter being 12 p.p.l., the predomina n t meter 8 p.p.l..
108 3.4 Description of the Poetic Devices of One Baka Song Having presented an analysis of the rhythmic poetic devices of Baka songs, what follows is a description of all of the poetic devices rhythmic, syntactic, phonological, lexical, and semantic thus far analyzed in one particular Baka song a lament Figure 3 5 presents a textual transcription of the song. Object 3 9 documents the audio performance of the song (Object 3 9. Audio file of solo song of lament (.mp3 1.9 MB)). The analysis is applied to the entire song performance, not just one representative refrain as is so often the case in literary studies of partially improvised song texts. 18 Performance examples will be identified and presented for each device found in the transcripti on, but elaboration on their poetic function must be reserved for a future article. 3.4.1 Ethnographic Background The song selected the simplest of the nine select, transcribed songs was recorded in 1998 in the Baka encampment of Mbalam (Map 3 in Fig. 1 6) It was sung by a 30 to 45 year old Baka man named Aye (Figure 3 6). The composer is unknown. Aye$ sang and instrumentally accompanied himself on a plucked four metal string rafia frame harp zither ( ng mbi na! pe! ke or ng mbi na! kpokpo ') (Figure 3 7). The six components of the ng"!mbi na! pe! ke are constructed from five locally ava i lable materials, mostly forest materials. The bridge ( mbo$ o! njo! ) is whittled from the wood of the mo! sasa! a tree. The neck ( mbe$ nga = lance ) and the soundboard/resonator ( d#!d# ) are fabricated from various pa rts of a raffia palm tree ( pe! ke ). Traditionally, the four strings ( ku ') of the harp zither were made from the rattan kpopkpo vine; today, however, wire
109 strands ( wa$ ya$ = Eng. wire') from metal trapping cable are used. The strings are attached to two end pegs whittled from the wood of mo! sasa! a tree saplings, or are simply wound around three to five inch metal carpentry nails. Finally, the eight tuning cinches ( ki$ y" ) are made from thin strips of bark peeled from kpongbo vines. The f our strands of metal trapping cable are attached to and stretched along the straight, slender shaft of the raffia palm branch The four metal cords are then further stretched and set in four notches ( t #!t # ) carved in the wooden bridge which is perpendicularly attac hed to the raffia shaft. The intersection of bridge and cords effectively doubles the four cords, producing eight distinct pitches when plucked. The pitches of Aye 's zither where approximately tuned to F 3, Ab 3, Bb 3, C 4, Eb 4, F 4, Ab 4, Bb 4. 19 Tensi on in the cords, and thus tuning, is effected by sliding each of the eight thin bands of vine that encircle the raffia shaft and cord s toward or away from the bridge nodes. The zither is normally laid across a players lap and plucked with the index finger s. The song is sung solo, as are many Baka songs which are sung to the accompaniment of the ng mbi na! pe! ke The themes of songs sung with ng mbi accompaniment are often personal and even intimate, in nature. Aye$ sings a personal lament a song of w oe while communing with the forest for solace, a telling expression of the degree to which many Baka identify with their forest home. In brief, i n the opening verses that is, Verses 1 2 (Example 9), Aye$ utter s the "moan like" ideophone Wu we wu we o Wu we wu we o ."
110 Example 9: Song of Lament, Verses 1 5 '0(" Vt!vd!vt!vd n+ Vt!vd!vt!vd n-!'vt!vd!'vt!vd n(9!hcdnognmd!< !_ln`m& '1( Vt!vd!vt!vd n+ Vt!vd!vt!vd n! '2( K`# ln#!`#!rh`#+ Lx!dxdr!`qd!knnjhmf!sn '3( ln# ln#!`#!mfnl`#' d(+ Lx!lntsg!hr!rod`jhmf! sn '4( i# ln#!`#!i#!adkd! Lx!d`qr!`qd!khrsdmhmf!sn!sgd!enqdrs-" Then, in Verses 3 5, he directs his attention to the forest, singing, La! mo! 'a! sia ( m y eyes are looking to ), mo! mo! 'a ngoma! ( e) ( my mouth is speaking to ), [and] j#! mo! 'a j#! bele ( my ears are listening to the forest [for help for consolationfor provision] ') The accompanying audio file in Object 3 9 documents Aye$ 's entire performance (Object 3 9. Audio file of solo lament with harp zither accompaniment (.mp3 1.9 MB)). In the sections that follow, that is, Sections 3.4.2 through 3.4.6, I will describe the rhythmic (Sec. 3.4.2), syntactic (Sec. 3.4.3), phonological (Sec. 3.4.4), lexical (Sec. 3.4.5), and semantic (Sec. 3.4.6) poetic devices thus far discovered in Aye$ 's text. 3.4.2 Rhythmic Poetic Devices Pulse, tempo, meter, and line (Fig. 3 2) Four rhythmic devices converge in Baka poetic texts to form what Herrnstein Smith would call the "sustained rhythmthe continuously operating principle of organization" in Baka poetic discourse: pulse, tempo, meter, and line. And as explained in Section 3.2, Example 3 of this Chapter, there are four pulses pe r line in Aye 's lament (Example 10). T he presence of the pulse is not plainly evident but is nonetheless suggested in the regularity of both the song's
111 melodic rhythm and the zithers accompanimental rhythm In Example 10 (as in Example 3) t he pulse is placed above the transcription and is represented with numerals equidistantly plotted above the text, roughly synchronizing pulse and text. Example 10: 4 pulses per poetic line Pulses 1 2 3 4 1 2 3 4 '03/" LL( 'khmd 0( !!!!!! !!!!! Vt vd vt! vd' n(+ 'khmd 1( !!!!! Vt vd vt vd' n(+ 'khmd 2( ! !!!!! k` #! ln#!!` #!! rh`#: 'khmd 3( !!!!! ln# ln #!` #! mfnl`: 'khmd 4( !!!!!i ln#!!!` #!!! i #!!!!!a !! d !! k !!d ( L ine boundaries and text boundaries do not however, coinicide as described further i n Section 3.4.3 and Chap. 2, Sec. 2.2.1. ) The measure of a song line's pulse is its tempo that is, its pulse rate. This song normally averages 140 MM, an average tempo for a Baka song Four meter types are present in Baka song: 16, 12, 8, or 4 pulses per line. T he meter of this song is four pulses per line. 3.4.3 Syntactic Poetic Devices C ontrastive verses and verse variations ( Fig. 3 2). Sung Baka discourse is not only subject to particular patterns of rhythm, but to particular patterns of verbal synta x as well. Syntactic patterns, as opposed to rhythmic patterns, are composed of grammatical units, not chronological units; and every kind of grammatical unit (morpheme, word, phrase, sentence, etc.) is potentially subject to poetic treatment. Generally stated, the poetic function distinctively marks the verbal syntax of a Baka
112 song by the cumulative action of particular poetic devices acting on particular syntactic units particularly ordered 220.127.116.11 Available grammatical units As for the natural gram matical units available for poetic exploitation in this song, one solo voice part versifies an available grammatical inventory of 10 different words and 2 6 distinct vocables. For each verse of this poetic discourse, 0 5 words and 1 6 vocables are uttered in a span of 5 6 syllables and synchronized to poetic lines of 4 pulses. 18.104.22.168 Available poetic syntactic units From the text transcription (shown in its entirety in Figure 3 5) we discover that this song is composed of four available poetic syntactic units (represented here in Figure 3 8 as Verses A, B, B 1 and B 2 ). Example 11 presents the two basic contrasting verses (Verse A, e.g., l.1 and Verse B, l.3). In addition, one verse, (Verse B) is further subject to two kinds of variation (e.g., Verse B 1 l.4 and B 2 l.5). Example 11: Verses in Contrast and in Variation 'otkrd( 0 1 2 3 '0 1 2 3 dsb-( 'khmd"0(" ZUdqrd!@\ Vt!!vd!!vt!!vd n+ ..!vt!vd!vt!vd-' n(!.. Vt!vd!vt!vd n-!'vt!vd!'vt!vd n(9!hcdn-!< !fdlhq'eq-(.ln`m'dmf-(( 'khmd"1( Vt!!vd!!v t!!vd n+ ..!vt!vd!vt!vd-' n(!.. Vt!vd!vt!vd n! 'khmd"2( ZUdqrd!A\ K`# ln#!!!`#!!!rh`#+ ..!ndhk!cd!lnh!.!HLO!.!qdf`qcdq!.. Ldr!xdtw!qdf`qcdms:
113 'khmd"3( ZUdqrd!A 0 \ ln# ln#!!!`#!!!mfnl`#' d(+ ..!antbgd!cd!lnh!.!HLO!.!o`qkdq-' d(!.. l`! antbgd!o`qkd: 'khmd"4( ZUdqrd!A 1 \ i# ln#!!!`#!!!i#!!!adkd! ..!nqdhkkd!cd!lnh!.!HLO!.!dbntsdq!.!enqds!.. ldr!nqdhkkdr!dbntsdms!k`!enqds! 22.214.171.124 Complementary devices of syntactic development In this song, both verse variations (i.e., variatio ns B 1 and B 2 e.g., l.4 and l.5 respectively) are created by grammatical substitution that is, syntagmatic paradigmatic substitution One variation is also altered by textual expansion (i.e., B 2 e.g., l.5). To explore grammatical substitution further as it is quite common in Baka song texts we note that two syntagmatic 20 slots, that is, the subject slot and the predicate slot, in contrastive verse "B" are subject to paradigmatic substitution. Both slots may be improvisationally filled by one of three par adigmatic substitutes: first, the subject slot in all "B" verse types may be filled with one of three co hyponyms from a semantic set of what might be called the singer s communicative sensory organs, that is his "mouth," "eyes," and "ears" ; second, the predicate slot in all "B" verse types may be filled by one of three co hyponyms from a semantic set of what might be called functions performed by the singer s communicative sensory organs," that is, "speaking, "looking," and "listening." The textual ex pansion of the verse is accomplished by the addition of an indirect object (i.e., "forest.") as heard in "B 2 verse types. As briefly noted earlier, verse boundaries and line boundaries do not coincide in verse type "B 2 "; in this song, verse terminal bound aries overlap line initial boundaries.
114 Technically, this is a kind of enjambment though its regular ity tends to temper its typically tension producing effect. 126.96.36.199 Frequency and distribution of available syntactic poetic units The table in Figure 3 9 presents the frequency and distribution of the available verse types. Relative to the frequency of each verse type, the two basic verse types are distributed fairly evenly throughout. The initial contrastive verse (A) evenly pervades nearly two thirds of the discourse while the subsequent contrastive verse (B) and its variations (B 1 and B 2 ) together, fairly evenly share the remainder of the sung discourse. 188.8.131.52 Syntactic development and form Figure 3 10 charts the overall syntactic development of t he entire song text. The table and diagram in Figure 3 11 depict s the allowable pairs of adjacent verse types and verse progressions. From Figures 3 10 and 3 11, I induce that the syntactic development and form of the entire discourse generally tends to be both iterative and revertive in that most often a series of one of the two contrastive verses is followed by a series of the other contrastive verse (or one of its variations). However, while a series of immediately repeated verses seems to generally m ark the syntax as iterative, the alternation of the two contrastive sets of immediately repeated verses eventually produces a revertive effect as well. 21 These sets of contrastive verses always revert to A verses types. 3.4.4 Lexical Poetic Devices L exi cal repetition and homonym repetition (Fig. 3 2). In the domain of lexical poetic devices, this song exhibits two devices: simple lexical repetition and
115 homonym repetition Lexical repetition that is simple lexeme repetition, is intra linearly recurren t in the performance of Verse A (e.g., line 2 1 Example 12) Example 12: Intra linear Lexical Repetition '10( Vt vd vt vd n! It is also inter linearly recurrent in words sung in a series of "B" verse types ( e.g., lines 25 27 Example 13 ). Example 13: Inter linear Lexical Repetition '14( K`# ln# ` #!rh`#: '15( ln# ln# `# !mfnl`#' d(: '16( i# ln# `# !i#!adkd! Repetition in homonyms recurs intra linearly in most stanza final lines, as in line 31 and 32 of Example 14 ; with the homonyms mo! (the noun for "mouth") and mo! (the 1S possesive pronoun), as well as the homonyms j (the noun for "ear") and (na)j (the verb "to hear"). Example 14: Homonyms '20( ln# ln# !`#!mfnl`#' d(: .. lntsg !.! 0R onr .!HLOU!.!rod`j!.!unb!.. Lx lntsg !hr!rod` jhmf! '21( i# ln#!`#! i# !adkd! .. d`q !.!0R!onr!.!HLOU!.! khrsdm !.!enqdrs!.. Lx! d`qr !`qd! khrsdmhmf sn!sgd!enqdrs!
116 3.4.5 Phonological Poetic Devices A ssonance and alliteration (Fig. 3 2). In the domain of phonological poetic devices, there are two basic devices: assonance and alliteration Assonance occasionally recurs improvisationally in Verse B in word final, line final vocable repetition ( e.g., line 38 w ith line 39 Example 15) Example 15: Assonance (word final, line final vocable repetition) '27( K`# ln#!`#!rh`#' d (: '28( ln# ln#!`#!mfnl`#' d (: Assonance also intermittently recurs in Verse A in word initial, line initial vocable repetition ( e.g., lines 42, 43, and 46 in Example 16) Example 16: Assonance (word initial, line initial vocable repe tition) "'31( d (Vt!vd!vt!vd n! '32( d (Vt!vd!vt!vd n! '33( Vt!vd!vt!vd n! '34( '_(Vt!vd!vt!vd n! '35( d (Vt!vd!vt!vd n! Alliteration (of voiced, bilabial fricatives) is recurrent in Verse A, phoneme initially, in the reiterated phonemes of the ideop hone Wu We Wu We o ( e.g., line 22, Example 17) Example 17: Alliteration (phoneme initially) '11( V t v d v t v d&!
117 3.4.6 Semantic Level Poetic Devices I deophones, syntagmatic paradigmatic grammatical substitution, reiteration of co hyponyms, reiteration o f meronyms, complimentary syntactic semantic parallelism, and anthropomorphism (Fig. 3 2). Finally, in the domain of semantic level poetic devices, this song makes use of at least six semantic level poetic devices: ideophones, syntagmatic paradigmatic gra mmatical substitution, reiteration of co hyponyms reiteration of meronyms complimentary syntactic semantic parallelism and anthropomorphism Unfortunately, the scope of this chapter only allows for a presentation of examples; detailed description of th eir poetic function must be left for a future study The ideophone 22 wu we wu we o (an ideophone that stands for "moaning") serially recurs in all occurrences of the contrastive Verse A (e.g., lines 1 and 2 Example 18) Example 18: Ideophone '0( Vt!vd!vt!v d n! Vt!vd!vt!vd n-!'vt!vd!'vt!vd n(9! hcdn-!
118 The first semantic device is grammatical substitution, syntagmatic p aradigmatic substitution in particular. It is a commonly constituted when, as in Example 19, two syntagmatic 23 slots (e.g., the subject slot and the predicate slot in contrastive verse B ") are subject to paradigmatic substitution. Example 19: Syntagmatic paradigmatic Grammatical Substitution (three co hyponyms) '2/( K`# ln#!`#! rh` # d(: ..!dxd!.!0R!onr!.!HLOU!.!knnj!.!unb!.. Lx!dxdr!`qd!knnjhmf! '20( ln# ln#!`#! mfnl`# d(: ..!lntsg!.!0R!onr!.!HLOU!.!rod`j!.!unb!.. Lx!lntsg!hr!rod`jhmf! '21( i# ln#! `#! i# !adkd! ..!d`q!.!0R!onr!.!HLOU!.!khrsdm!.!enqdrs!.. Lx!d`qr!`qd!khrsdmhmfsn!sgd!enqdrs! Improvisationally, b oth slots may be filled by one of three paradigmatic substitutes: first, the subject slot in all B verse types may be filled with one of th ree co hyponyms from a semantic set of the singer s communicative sensory organs: h is mouth, eyes, and ears "; second, the predicate slot in all B verse types may be filled by one of three co hyponyms from a semantic set of functions performed by t he singer s communicative sensory organs ": speaking, looking, and listening. The s econd semantic device, also shown in Example 18 (above), is the reiteration of hyponyms 13 Throughout the song, the reiteration of hyponyms is inter linearly intermitte nt, though prevalent, in the same syntax slot co hyponyms la! mo! and j ("eyes, mouth, and ears", i.e., "kinds of the sensory organs") as well as the co hyponyms sia! ngoma! j ("see, talk, and listen", i.e., "kinds of sensory communication") as seen in lines 30 32.
119 The third device is l exical reiteratio n of meronyms 25 The reiteration of meronyms is inter linearly intermittent, though prevalent, in the same syntax slot meronyms la! mo! mo! mo! j mo! (i.e., "my eyes, my mouth, and my ears") in lines 30 32, Example 20. These organs are meronyms in tha t they each ultimately "stand for" the person of whom they are sensorial parts. Example 20: Meronymy '2/( K`# ln #!`#!rh`#' d(: '20( ln# ln# !`#!mfnl`# d(: '21( i# ln #!`#!i#!adkd! The fourth semantic device, complimentary syntactic semantic paralleli sm is inter linearly intermittent, though prevalent, where successive lines, like lines 30, 31, and 32 in Example 21 progressively develop and intensify similar semantic images in three adjacent verses containing what might be called "sensory communicati on modes." Example 21: Complimentary syntactic semantic parallelism '2/( K`# ln#!`#!rh`#' d(: ..!dxd!.!0R!onr!.!HLOU!.!knnj!.!unb!.. Lx dxdr `qd knnjhmf '20( ln# ln#!`#!mfnl`#' d(: ..!lntsg!.!0R!onr!.!HLOU!.!rod`j!.!unb!.. Lx lntsg hr rod`jhmf '21( i # ln#!`#!i#!adkd! ..!d`q!.!0R!onr!.!HLOU!.!khrsdm!.!unb!.. Lx d`qr `qd khrsdmhmf The fifth, and last, semantic device is anthropomorphism In the Aye $ 's lament, anthropomorphism is semantically implicit in the inter linear and intermittent occurrences of lyrics that giv e expression to the singer s personal engagement with the forest that is, in his "looking to," "speaking to," and "listening to" the f orest for solace,
120 sympathy, or help. They are performed in any recurrence of Verse B 2 ( e.g., lines 3 5 in Example 22) Example 22: Anthropomorphism '2( K` ln#!`#!rh`#' d(: ..!dxd!.!0R!onr!.!HLOU!.!knnj!.!unb!.. Lx!dxdr!`qd!knnjhmf! '3( ln# ln#!`#!mfnl` d(: ..!lntsg!.!0R!onr!.!HLOU!.!rod`j!.!unb!.. Lx!lntsg!hr!rod`jhmf! '4( i ln#!`#!i! adkd ..!d`q!.!0R!onr!.!HLOU!.!khrsdm!.! enqdrs !.. Lx!d`qr!`qd!khrsdmhmfsn! sgd!enqdrs 3.5 Conclusion Having discussed each of the principal poetic devices for this s ingle song one by one, I conclude by re presenting them together, as a complementary whole. I recapitulate in this manner because t he peculiar effects of poetic ally organized discourse are not ultimately experienced by considering poetic devices one by on e but in the cumulative perception and experience of the ebb and flow of multiple interconnecting poetic devices. In the diagram in Figure 3 12, I have depict ed the interconnectedness of most o f the aforementioned poetic devices in Aye $ s lament. Any one of the song text's ten words and two vocables might function in multiple poetic devices. The poetic "equivalencies" of the lyric's sixteen poetic devices signify multiple "chains of semiosis" (Peirce, cited in Turino 1999: 222 240). An y one of these sixteen poetic devices is potentially linked to another. For example, r hythmic poetic devices ( like pulse, tempo, meter, line ) are linked to syntactic poetic devices ( in verses repeated, verses contrasted, verses in variation, or verses in enjambment) These linked devices, in turn, are also "chained" to l exica l poetic devices ( like repeated words
121 and homonyms), or to phonological devices (signified in assonance and alliteration), or to semantic poetic devices ( like ideophones, hyponyms, me ronyms, semantic parallelism, and anthropomorphism) E ach device communicates its own qualities, yet all are unified in time. The most conspicuous semiotic effect of these semiotic chains is the interconnected patterns that they create. These p atterns, in and over time, create expectations and unify experience (Meyer 1956:22 42; Dooley and Levinsohn 2000:26; Chenoweth 2001:112) Variations on these patterns reshape expectations and thus engender a sense of drama as they unfold. In Baka song, as in a ll poetic traditions, the perception and performance of trad i tional poetic patterns like those in Aye$ 's lament is indispensable for any one who would competently participate in the communication of the unique and powerful communicative effect s of poetic all y organized discourse. Even t he generation of new poetic creation s will build upon received forms, even and especially if these new creations would generate new effects to serve new purposes. Expectation as an experiential effect of poetic patterning is b ut one, albeit fundamental, poetic function of Baka song texts. Other effects are of course created. Examples of such effects will be taken up in Chapters 4 (e.g., Sec. 4.3.2) and 5 (e.g., Sec. 5.2). The scope of those effects, however, will be limited to the particular poetic effects achieved through songs sung in the course and context of performing traditional Baka oral narratives.
122 A`j`"Mntm Fknrr" mfn"l`" discussion, word klt" conversation, discussion '_d"(j`"kn" (formal) counsel from elders kd"vt" (personal evening) advice ln"rhln" proverb, recitation, account, narrative, monologue w/o song fa"mfnmfn allegory, proverb, saying, dictum la`"kh comparison, parable k#j`"m" story, legend, fable, history, chantefable d" song Figure 3 1. Common Baka speech act terms F IVE P OETIC D OMAINS (1) R HYTHMIC P OETIC D EVICES pulse tempo meter line (musically metered) (2) S YNTACTIC P OETIC D EVICES repeated verses repeated verse segments repeated stanzas contrastive verses contrastive verse segments verse variations verse segment variations stanza variations verse derivations enjambment (3) P HONOLOGICAL P OETIC D EVICES vocables vocable repetition vocable substitution ideophones assonance consonance alliteration borrowed words vowel lengthening sprech stimme (speech song) (4) L EXICAL P OETIC D EVICES (intra and inter linearly) repeated phonemes repeated ideophones, repeated words homonyms reiterated homonyms reduplicated homonyms (rhythmically) (5) S EMANTIC P OETIC D EVICES ideophones reiteration of ide ophones reduplication of ideophones similes metaphors metonyms reiteration in metonyms anthropomorphism meronymy (co )hyponyms semantic parallelism (complimentary) syntactic semantic parallelism syntagmatic paradigmatic grammatical substitution rhetorical questions archaic language loan word synonymy ellipsis Figure 3 2. Inventory of Baka song text poetic devices
123 Song Tempo (pulse rate: MM =) V98032100338: story song with bow harp accompaniment ( lngbidi ) 160 @/70386HHH1409 c`mbdrnmf"hm"bdkdaq`shnm "ne"sgd"ahqsg"ne"svhmr 140 U87/1/3H/100/9 solo "lament" with zither accompaniment 140 C87/116H//3129 dkdog`ms"gtmshmf"rnmf 125 U87/1/3H//8//9 song with frame harp accompaniment of a young boy playing with his parents 125 @0/1886@1//9 dancesong (1) retu rning from a gorilla hunt or, '1("ctqhmf"`"bhqbtlbhrhnm"qhsd 118 @0/1886@1419 rnmf"`mmntmbhmf"sgd"qdstqm"eqnl"`"rtbbdrretk"dkdog`ms"gtms 115 @00/586@/119 male initiation song 'vhsg"sgd"rohqhs" Idmfh ( 105 @00/586@///9 rnmf"`rrnbh`sdc"vhsg" Idmfh +"`"rohqhs 105 Figure 3 3. Table of song tempi Song Meter (pulses per line) @00/586@/119 male initiation song 'vhsg"sgd"rohqhs" Idmfh ( 16 @0/1886@1//9 dancesong (1) returning from a gorilla hunt or, '1("ctqhmf"`"bhqbtlbhrhnm"qhsd 16 @0/1886@1419 rnmf"`mmntmbhm f"sgd"qdstqm"eqnl"`"rtbbdrretk"dkdog`ms"gtms 16 @00/586@///9 rnmf"`rrnbh`sdc"vhsg" Idmfh +"`"rohqhs 12 @/70386HHH1409 c`mbdrnmf"hm"bdkdaq`shnm"ne"sgd"ahqsg"ne"svhmr 12 / 8 C87/116H//3129 dkdog`ms"gtmshmf"rnmf 8 U87/1/3H//8//9 song with frame harp accompa niment of a young boy playing with his parents 8 V98032100338: story song with 2 string bow harp accompaniment ( lngbidi ) 8 U87/1/3H/100/9 solo song of woe with zither accompaniment 4 Figure 3 4. Table of song meter types
127 Figure 3 6. Photo of Aye$ singing song and playing ng"!mbi na! pe! ke Figure 3 7. Photo of a Baka harp zither ( ng mbi na! pe! ke / ng mbi na! kpokpo ')
128 'd-f-+"khmd"0( A 'd-f-+"khmd"2( A 'd-f-+"khmd"3( A 0 (e.g., line 5) B 2 Figure 3 8. Composite of available syntact ic verse types number of unit occurrences percentage of unit occurrences @"w"24 A"w"7 A 0 "w"5 B 2 x 7 @"< "51$ A"w"03$ A 0 "< "00$ B 2 = 13% Figure 3 9. Frequency of available verse types
129 Rxms`w (1) = line 1, etc. and A, B, = verses (1) A (2) A (3) B (4) B 1 (5) B 2 (6) A (7) A (8) B (9) B 1 (10) B 2 (11) A (12) B (13) B 1 (14) B 2 (15) A (16) A (17) A (18) B (19) B 1 (20) B 2 (21) A (22) A (23) A (24) A (25) B (26) B 1 (27) B 2 (28) A (29) A (30) B (31) B 1 (32) B 2 (33) A (34) A (35) A (36) A (37) A (38) B (39 ) B 1 (40) B 2 (41) A (42) A (43) A (44) A (45) A (46) A (47) (B) (48) A (49) B (50) B 2 (51) A (52) A (53) A (54) A (55) A (56) B (57) B 1 (58) A (59) A (60) A (61) B (62) B (63) A (64) A (65) B (66) B 2 (67) A (68) A (69) B (70) B 1 (71) B 2 (72) A (73) A (74 ) A (75) A (76) A (77) A (78) B (79) B 2 (80) B 1 (81) A (82) A (83) A (84) A (85) A (86) B (87) B 2 (88) A (89) A (90) A (91) A (92) A (93) A (94) B (95) B 1 (96) B 1 (97) B 2 (98) A (99) A (100) A (101) A (102) A Figure 3 10. Chart of syntactic develo pment
130 Figure 3 11. Adjacent verses and diagram of allowable verse progressions
131 Figure 3 12: Diagram of the poetic patterns of Aye 's Lament." (The color coded arrows, boxes, highlighting, numbers, letters, and words are intended to iconically and symbolically suggest the complementarity of the poetic devices previously set out in Sections 3.4.2 through 3.4.6. The performance of all ten verses is hypothetically patterned with ten metered lines of four pulses each, lasting approximately twenty secon ds.)
132 1 See Jean Jacques Nattiez, "The Contribution of Musical Semiotics to the Semiotic Discussion in General," The Garland L ibrary of Readings in Ethnomusicology Vol.2 (New York: Garland Publishing, 1990), 221 242. 2 Paul Cobley and Litza Janz, Introducing Semiotics, New York: Totem Books, 1997), 144. 3 Roman Jakobson, "Linguistics and Poetics," in Style in Language, edited by Thomas A. Sebeok (The Massachusetts Institute of Technology, 1960), 351. 4 Roman Jakobson, "Le langage en relation avec les autres systmes de communication," in Essais de linguistique gnrale II (Paris: Minuit, 1973), 99 100. 5 Roman Jakobson, Selected W ritings, vol.III Poetry of Grammar and Grammar of Poetry (Mouton: The Hague, 1966), 602. 6 Jonathan Culler, Structuralist Poetics (Ithaca: Cornell University Press, 1975), 56. 7 P. Valery, The Art of Poetry, Bollingen Series 45 (New York, 1958). 7 The met ered line is the primary poetic feature because it is the sole poetic device (thus far in this study and in my personal recollection of the whole of the Baka song tradition) which no Baka song may be without and still be called song ( 'e! ) While it is true that most sung Baka discourse employs highly formantal vocal techniques and highly repetitive syntactical structures, not all songs do. Thus, songs may exclusively employ sprechstimme ( spoken song ," vocal techniques with relativel y little formant atriculation), or (non denotative, non grammatical) vocables (as in the yodelling' song style called yeyi! ) and still be called song, primarily because it is marked by a musically metered poetic line. 8 Robert Pinsky. 1998. The Sounds o f Poetry New York: Farrar, Straus and Giroux. p.28, 34. 0/ Indeed, every song from a corpus of more than 400 field recordings of Baka songs has a musically metered poetic line. 00 In either case, technical verification of this auto segmental reality require s competency in music literacy. 01 There are cases that typographically necessitate using half lines. Such cases are indicated accordingly. The effect (real and representational) of superimposing a line upon the syntax of a verbal utterence is described later in Sec. 184.108.40.206 after line and syntax have been described as units in themselves. The relationship of line to verbal syntax is like wind is to leaves on a tree, that is, it is only perceived in the effect one
133 has on the other. This effect then, of the line on syntax, is symbolized in the formatting conventions of a writing system. 02 Other conventions, slashes for example, would be sufficient modifications in text transcriptions using Roman script 03 Another poetic effect of the line (a rhythmical p rinciple) on the verbal syntax (a grammatical principle) is indicated by the relative spacing of words and syllables as performed in time, that is in reference to the pulse. 04 And tempo is but an elementary quality of a pulse. 05 It is not yet known whethe r or not the meter of a particular Baka song or song genre is "principally" dictated by received musical forms, textual forms, or both. (Forthcoming music analysis will help to determine this.) As has been said, what is known from the data is that meter i n a Baka song line respects the prevailing patterns of verbal syntax. This relationship of line and syntax can be measured. It is doubtful that a lingua literary description would need a more detailed description of the musically rhythmic nature of a poe tic line than a description of its tempo and meter. 06 It will be pertinent to music analysis to then characterize these metrical units (even the lines of 12 p.p.l.) as formed from duple rhythms. Thus (so far), we see pulses organized into dulpe rhythms which in turn are organized into metrical lines divisible by two. 18 Composite representations of song refrains may serve many purposes, but they cannot bring into relief variational or contrastive performance features that may progressively signify any of the emerging structures of oral performance. Only a thick analysis of a full performance can suggest such structures. 19 Tunings vary with compositions. Another song, preformed by another musician, for example, uses the approximate tuning of F 3, [G 3], Bb 3, C 4, [D 4], F 4, [G 4], Bb 4. Careful measurements have been documented for many Baka instruments and will likely be accounted for in a forthcoming emic analysis of the Baka melodic system, currently underway. 20 This term refers to the ordering of sequences in grammatical structures. 21 Both iterative and revertive processes involve repetition, but an iterative process is one in which you repeat something again and again, using the results from an immediately preceding stage (e.g., A A A A A); wher eas revertive processes repeat results from a more distant stage (e.g., A B C B A) (Chenoweth 1972:94, citing Nettl 1956:66 71). 22 See Section 4. 3.2.2 for an explanation of ideophones.
134 23 Again, this term refers to the ordering of sequences in grammatical s tructures. 24 A hyponym is a word of more specific meaning than a general or superordinate term applicable to it. For example, spoon is a hyponym of cutlery ." From: "hyponym". Oxford Dictionaries. April 2010. Oxford Dictionaries. April 2010. Oxford Univer sity Press. 5 January 2011
135 CHAPTER 4 BAKA LI! KA! N%! SONG AS A DISCOURSE FEATURE OF BAKA STOR Y 4.1 Performance Relationship of Baka Story and Song Song ( 'e! ) sung in the course of telling a traditional Baka story ( li!ka! n"! ) is referred to by the Baka as 'e! na li! ka!n"! ( s tory's song '). Most Baka li! ka! n"! include song. Of 147 Baka li! ka! n"! documented in seven distinct collections of traditional Baka narratives, 109 (77%) include song (Higgens 1981; Brisson 1981, 1996, 1999 ; Kilian Hatz 1989; Boursier 1994 ; Le! onard 2003; Fitzgerald 2007) A dditional songs would have likely been included in these seven collections had all of the original a udio recordings been carried out in more "natural" circumstances, that is, in settings with both a storyteller and a participat ing audience, as opposed to a storyteller and an ethnographer in a "staged," one on one interview. Even so, it is plain enough that most Baka stories do include songs. To say so may state the obvious, but not all of the world's story telling tradition s do, in fact, include song. And so, it is reasonable to ask, "Why do most traditional Baka narratives include song? What is the relationship of Baka songs to Baka stories?" These two questions are the principal concerns of this chapter. 4.1.1 Story and Song in African Narrative Performance In Africa and throughout the world story and song (or, story and music ) co occur in countless varieties of social settings. African communities combine texts with music to create recitations of narrative songs, gene alogies, and epics. Aside from a study on the Kpelle Woi epic (Stone 1988) and phonograph recordings documenting sung stories, genealogies, and histories of the Zulu, the Shona, the Karanga, the Duma, the Soga, the Wolof, and their West African neighbors,
136 ethnomusicologists have given little attention to analyzing such music thoroughly. (Hampton 2000:107) This present analysis as it seeks to contribute to the body of research that addresses this lacuna begins with the observation that while story and song do indeed co occur in countless African communities and in a variety of social settings, it is equally, if not more, important to note that their performance integrity varies just as greatly A brief survey of Africanist scholarship linguistic, folklorist ic, anthropological, and musicological documenting African oral narrative performance types ranging from North and West African epics ( e$ pope$ es ), to Central African chantefables, to East African folktales, to South African sung narrative poetry, initially suggests three broad performance distinctions. At one extreme, story may, of course, be told without music (and vice versa ) (Dorson 1972; Abrahams 1983; Finnegan 1967; Kisliuk 1991:373 383). In other performance contexts, story and song may be performed "in alternation" with varying degress of cohesion. The music and songs of some of these narrative events are variously described as "intermissions," "interludes," or "intercalcated [songs] which punctuate the narrative" (L e! onard 2003:2; Agawu 1995:170 179; Johnson, Hale, and Belchner 1997: xvii, 255 ; see also Ben Amos 1975:24, 51; Finnegan 1977:109, 119 2007:51 ). The songs of other "alternating" performance practices, however, are viewed as significantly more integral t o the narrativ e' s structure and storyline ( Biebuyck and Mateene 1969; Stone 1988 :3, 6 ; Kisliuk 1991: 74 80, 364 372; Agawu ibid. ). The most extreme integration of story and song, of course, is achieved when certain narratives are entirely sung (Coplan 198 7, 1988, 1994; Opland 1983). And
137 whether exclusively spoken, or exclusively sung, or a combination of both, African monologue narrative discourse is also commonly performed with intermitent or continual instrumental or choral accompaniment, or both (Okpew ho 1979:59 66; Johnson, Hale, Belcher 1997:255). I have initially described these three broad characterizations of African monologue narrative performance in terms of a relative and variegated "promoting and demoting" of certain binary distinctions, compar ing the descriptors "story and song," "song and music," "speaking and singing," "instrumental and choral," "intermittent and continual." Cumulatively, this description generally and simultaneously suggests that (1) the semiotic relationship between story and song, that is, their performance integrity, inherently tends more toward synthesis than antithesis, (2) the nature of their tendency toward unity is predicated on their shared verbal properties, and (3) the nature of their boundaries is more complex th an simple. Ruth Finnegan expands on the relationship of verbal and non verbal modes in her book The Oral and Beyond: Doing Things with Words in Africa So we can see Africa as indeed the rich home of words but emphatically not of words separate from thei r multiplex multi modal settings or of some mysterious and distinctive "orality." (2007:223) getting rid of the over ambitious claims for "language" in fact allows a clearer perspective on humans' active use of words but words now seen, more modestly, as set in the context of, and intermingled with, the array of other communicative modes of which verbal language is only one. (ibid., 210) The boundaries of "the oral", of "performance," and of literature may come to dissolvebut at the same time such notion s, treated more
138 critically and imaginatively, can lead into a deeper understanding of people's creative and reflective ways with words, more sophisticated and more richly intermingled into intricately meshed modalities than we once realized. (224) In Figu re 4 1 I have abstracted the verbal and aural qualities of the aforementioned song and story performance practices and (re)configured them in a graph in order to (re)present the relationship of story and song in Africa as "more richly intermingled" and "i ntricately meshedthanonce realized." The axis graph abstractly posits that the performance distribution of speech song and music in African monologue narrative discourse results from the axis of a "speech song continuum" variously projected into the a xis of an "instrumental/choral accompaniment continuum." From the graph, we read that a long a simple speech song continuum, monologue narrative discourse performance may be ( 1) exclusively spoken with varying degrees of "everyday" and "heightened" speech r egisters, ( 2) alternately spoken or sung in varying manners and degrees of frequency and distribution, or ( 3) exclusively sung Simultaneously, e ach of these verbal performance practice s whether spoken, sung, or both is potentialy subject to varying degre es of instrumental and/or choral accompaniment. 4.1.2 Story and Song in Baka Narrative Performance To find examples of some of the basic permutations of African narrative performance practices, one need not comb a ll of the literature on all of the narrativ e
139 discourse traditions of Africa: a cross section of Baka narratives conveniently exhibits a good number of verbal performance types. The distinction between those Baka stories with song and those without was first made by French priests Robert Brisson a nd Daniel Boursier. Brisson and Boursier lived among the Baka, in different communities, for over twenty and thirty years respectively, and were the earliest to document Baka traditional stories (1979, 1984, 1994, 1996). In their respective commentaries on traditional Baka narratives a s well as their collaborative work on a dictionary of the Baka language they specifically assign the French term chantefable ( sung tale ) to describe those Baka li! ka! n"! that include singing. 1 However, both Brisson and Bo ursier and later, Higgens, Kilian Hatz and Le! onard in each of their collections, only modestly describe the characterisitics that distinguish sung narrative texts from spoken narrative texts. Further, the typographical conventions of the text transcripti ons of their audio field recordings could not (nor doubtless, were they ever intended to) adequately represent either the temporal distribution of song in a story, or the performance duration of such songs especially when speaking and singing overlap. In deed, when a song is first uttered or reintroduced in these recorded narrative performances, most song text transcriptions only present a single, typified version of the refrain. Neither subsequent lyric variation, nor the relative temporal performance re lationship of speaking and singing were germane to the initial purposes of these collections (and no one would necessarily expect otherwise). By drawing attention to the limitations of these conventional transcription practices, I only mean to aid the app reciation of the emerging processes
140 that have guided the understanding of oral narratives. This present analysis, to which I return, only benefits from these former representations regardless of their limitations. As reported in the opening paragraph of t his chapter, my tabulations of the narrative collections of Brisson and Boursier, as well as Higgens, Kilian Hatz, Le! onard and Fitzgerald, confirm that most Baka narratives include song. Yet plainly, those same tabulations indicate that it is not uncommo n to hear some Baka narratives exclusively spoken Roughly a quarter of the documented narrative collections have no singing. These exclusively spoken performances, however, vary greatly in their proportional usage of everyday versus heightened spee ch patterns, depending upon who is narrating and what they are recounting. For example, a sought after and practiced adult Baka teller telling a well known and highly formalized traditional story ( li! ka! n"! ) employs many more heightened speech devices than does a Baka child recounting (and re recounting) a memorable personal experience in the forest. 2 By far the most common Baka narrative performance styles and practices exhibit both spoken and sung performance sometimes in alternation, sometimes overlapping. Taken as a structural and sequential whole, the relative temporal distribution of speaking to singing in a Baka li! ka!n"! at least in my initial analys i s, seems to resist generalization: sung discourse may be distributed at the beginning, middle, or end of a li! ka!n"! ; it may occur one or multiple times, intermitently or constantly, briefly or protractedly In some cases, for instance, choral accompaniment may be performed during the entire na rrative performance while the teller singer weaves in and out of speaking and singing.
141 The only verbal performance style not exhibited in Baka narrative performance is that in which the entire narrative discourse strictly speaking is sung. In fact, as th e bulk of the rest of this chapter will point out, "mainline" narrative material is rarely produced in sung mode. So while occasionally an entire narrative performance may be permeated with singing, as highlighted above, the main storyline is rarely produ ced in sung mode. Instead, the storyline is reserved for speech mode. However, I do not mean to imply that sung discourse material does not intrinsically and critically inform the "mainline" material of a li! ka! n"! To the contrary, whether mainline or supportive, foreground or background, the pragmatics of singing in Baka narrative discourse indicate that sung discourse mode can be decisive in achieving the intra and supra pragmatic goals of a narrative performance This particular role of song is progressively set out in greater detail in Sections 4.2 4.3 of this chapter, and Sections 5.2 5.3 in Chapter 5. Having introduced in some detail the breadth of variation in the relative co occurrence of speaking, singing, and instrumental accompanying in Baka narrative (or any similar narrative tradition), I now move from a description to an analysis of these variations in actual performance practice. For my interest is not so much in discovering how much or how little is spoken or sung in a traditional Baka narrative but in understanding the nature of such a variety of choices. For the choosing of one configuration of speaking to singing to musiking over another implies communicative intent. I borrow this conclusion from functional linguistics whic h claims according to Levinsohn, that
142 one basic principle of a functional approach to text linguistics is that choice implies meaning In virtually every sentence, authors have the option of expressing themselves in more than one way. Are these ways sim ply stylistic variations? Text linguistics answers, No! Because there is a choice of ways, the ways differ in significance; there are reason s for the variations. ( 2004:0.2.1) 3 T o move on then from quantitative and comparative descriptions of surface stru ctures of speaking and singing and musiking, to interpreting the semiotic significance of the principal aural signs of Baka song in Baka story, what follows is a discourse analysis of the intertextual nature of e! na li! ka! n "! (Baka story s song ) in l i! ka! n "! (Baka story). The aim of the analysis will be to understand better why the Baka sing as opposed to simply speak when they tell a traditional narrative. A discourse analysis is applied because the nature of d iscourse analysis, regardless of its application whether sociological, political, linguisitic, or other, is essentially ordered to asking why this or that discourse is executed in this or that way. 4 The data for the analysis is based upon the documentation of 12 select Baka li! ka! n "! t wo narratives from each of the six collections of traditional Baka narratives cited at the beginning of this chapter. This data sample represents the most geographically, historically, and socio linguistically diverse sample possible from existing documen tation. First, because the areas from which these field recordings originate represent the most culturally diverse regional concentrations of Baka speakers in the southeast rainforest of Cameroon (Map 3 in Fig. 1 6). And second, because the recorded perf ormances are not limited to my fourteen years of field experience, but span nearly forty years, the earliest recorded in 1973. A comparison of these diverse
143 documentation projects is made quite feasible by the fact that most, if not all, of the recordists and ethnographers where quite competent in the Baka language and followed a fairly similar and consistent transcription methodology and technique. The original field recording for every documented narrative is still not available for study. Thus, while all twelve text transcriptions are available, only the recordings of Higgens, Le! onard, and Fitzgerald provide audio documentation. For this reason, the text linguistics findings of the following textual analyses are more fruitful than the more performan ce centered analyses. Once the audio recordings for all of the narrative performances are gathered into one place, additional performance centered studies can be performed. The basic format of the analysis is presented in Figure 4 2. The prose sections t hat follow (Sections 4.2 through 4.3) mirror the order of the outline. Fundamentally, Baka li! ka! n"! songs are found to function in two essential signifying domains of narrative discourse: first as an agent of narrative discourse cohesion (presented in Section 4. 2), and second as an agent of narrative discourse development (presented in Section 4. 3). In Section 4. 2, my analysis reveals that li! ka! n"! songs effect discourse cohesion generatively performatively and textually 5 Descriptions and examples of each of these four sub domains of cohesion are documented in their respective sub section s, though textual analyses predominate. Within this domain, (inter)textual cohesion of spoken and sung li! ka!n"! texts emerges as chiefly grammatical and semantic in nature. In the sub sections of Sections 4. 2.3.1 and 4. 2.3.2, examples of typical grammat ical and
144 semantic devices are cited from the texts of our aforementioned sample of 12 traditional narratives (Sec. 4. 1 .2 ). In Section 4. 3, the analysis further reveals that li! ka! n"! songs also effect discourse development both contextually and climacti cally The sub sections of Section 4. 3.1 then present typical examples of contexualizing devices for five "contextualizing information types." Most are textual devices. Finally, the sub sections of Section 4. 3.2 cite examples of devices that frequently contribute to climactic development. These devices are recognized according to five common marks of climactic development. And while climactic devices like cohesive devices and contexulualizing devices are often signified textually, many other climactic signs are formalized musico poetically, as we shall see. 4.2 Cohesion of li# ka# n!# Story and Song 4.2.1 Generative Cohesion (Stories Engender Songs) Baka songs are not merely "included" in most stories; stories engender songs. My longtime friend Tombomb o once explained that, "All [Baka] songs come from storiesnot just [ li! ka! n"! ] story songsbut even hunting songs, lullabies, divination or dance songs [come from "narratives"] ( Tombombo 2007) He further claimed that "stories might be forgotten, but [ stories ] songs are not" ( ibid. ). 6 Some songs, then, as they become disassociated with the story from which they arose ( for any number of reason s ), may gradually and eventually become appropriated for use in social contexts other than the narrative conte xts in which they originated, as in the case of dance li! ka!n"! songs at Baka funerals. In such cases, the temporal cohesion of story and song normally implied by their mere co occurrence is absent.
145 During Baka funerals li! ka! n"! songs may be sung outside the temporal context of a storytelling event. At su ch times, li! ka! n"! songs are performed, though no story is told, no story teller telling. This does not mean that stories are not or cannot be told during funerals, for they often are; but they need not be. A story's song can stand on its own. I obser ved this practice during my first few years of fieldwork among the Baka in the encampement of Ndjibot (1996 1997). In Ndjibot I recorded and participated in that is, attended, mourned, danced, and contributed monetarily to a number of funerals. The aud io recordings were mainly of the dance music performed at funerals. Initially, I assumed that all or most of the dance songs performed during a funeral were exclusive for that context. However, as I continued to gather audio visual recordings, collect et hnographic information, and progress in Baka language competency, I gradually learned that four or five distinct dance song genres were typically performed in the course of an all night funeral celebration. These included dance song genres invoking the f orest spirits Jeng i E! bu! ma! E! s"!nj"! and Mbo! a mbo! a as well as li! ka! n"! dance songs. Most of these dance songs for spirits, however, were also performed in social and ritual contexts other than funerals. In addition to these dance songs for spirits, however, li! ka! n"! songs were also sung, and "danced," though without narrative Initially, with the absence of a storyteller, I did not assume that the term li! ka! n"! (dance) and li!ka! n"! (story) necessarily referred to the same thing. None of the literature, Bris son, Boursier, et al. made explicit mention of dancing taking place during storytelling. So I assumed at first that perhaps these songs shared the same term, but not the same song texts. As each song was sung and danced, Baka friends would simply seem t o describe the import of the dance song as somehow ( ipso facto )
146 referring to the character of a well known forest animal (fieldnotes: dry season, 1997). What was not made explicit though plainly assumed was that song texts were not abstractly composed and associated with some forest animal, or legendary character, but rather, generated in and thereby associated with a particular Baka narrative, whose protagonist (or antagonist) was the subject of the dance song at hand. Thus, songs index stories. For the Baka, the integral link between dance, song, and story is assumed. They eventually made this connection explicit to me in their claim that, "We clearly see the story as we sing the story s song and dance its dance" (emphasis added; Nomedjo group intervie w, March 2008). It is unlikely, then, that the Baka generate many songs, if any, "for song's sake." In all my collection of several hundred field recordings, I know of no song ( whether li! ka!n"! song or non li! ka! n"! song ) that is clearly not somehow th e related experience of some individual or group, fictitious or non fictitious Over time the intertextual cohesion of a Baka song any Baka song and the experience that generated it may become obscure indeed. Yet, what is suggested by the process that ge nerates li! ka! n"! song is that Baka song in Baka story is no mere interlude or intermission ," but an instrinsic feature of the discourse. For even when li! ka! n"! songs occur outside the passing context of a storytelling event, the cohesive effects of other coh esive phenomena still indexically, iconically, and symbollically effect a connection in the minds of the Baka.
147 4.2.2 Performa nce Cohesion of li! ka! n"! Story and Song The cohesive nature of Baka story and story song is further confirmed not only in what is learned from how they are generated (i.e., from a common experience ) but in how they are performed Baka narrative cohesion is embodied in the constellation of spoken and sung speech acts performed by a Baka community at each storytelling event. Story a nd song cohere all the more because they are performed by a socially homogenous choir of voices, a Baka speaking community. At the hub of this constellation of voices stands the storyteller singer ( wa kpo ). The story unites with the song through the voic e of one narrator who intermittently tells the narrative, intones the song, and leads others to join in the refrain (see, for example, the narrative of Lo! ndo!, Sec. 4. 3.2, Fig. 4 8, or Su$a! t# Ku! nda Chap. 5). Through the performance of this refrai n the participants voices are joined to that of the storyteller singer, and through his (or her) voice all are joined to the story. If the storyteller is not competent to both tell the story and lead in singing, the story with song ( chantefable ) is not normally performed. A Baka story is not typically told by one person and sung by others, as is the case in some African cultures (Agawu 1995:165 166). 7 The common performance thread, then, is the voice of the singer storyteller, who simultaneously te lls, sings, and joins: joining story to song, people to song, people to story, and people to people. Having considered the cohesive relationship of li! ka!n "! and e! na li! ka!n "!, both generatively and performatively, I now move on to consider how they als o unite textually
148 4.2.3 Textual Cohesion of Story and Song The most salient phenomena of (verbal) discourse is the fact that sentences or utterances are linked together (Renkema 2004:103). Texts are considered discourse if they are interpreted as conne cted to each other somehow. Two criteria of this connectedness, according to de Beaugrande (1981), are intertextuality and cohesion (Renkema ibid.:49 51). In discourse analysis, connections which result when the interpretation of a textual element is d ependent on another element in the text is called cohesion (ibid.). In the spoken and sung texts of Baka li ka!n "! textual cohesion is manifested grammatically and semantically ( Figure s 4 3 and 4 4 ). 220.127.116.11 Grammatical cohesion Three grammatical devic es may effect inter textual cohesion between spoken and sung li ka!n "! texts: lexical repetition, lexical substitution (anaphora, specifically), and common point of view (Figure 4 3). 8 Grammatical cohesion through lexical repetition. Intertextual cohesion is compounded when the narratives spoken and sung texts not only share a common voice (embodied in the storyteller singer), but also share common words and phrases. Sung texts often simply reiterate, verbatim or near verbatim, previously spoken texts. I n the li! ka!n "! Ka$la$ for example, the entire refrain of the li! ka!n "! song (Example 1, Verse 8b) is a verbatim repetition of the opening spoken line (Example 1, Verse 1) of the story. Example 1: Brisson (1996) Ka# la# v 1 and 8b (verbatim sung reiteration of previously spoken phrases) (Verse 1, the character, Komba (God), speaks )
149 ''Ma a! ng"ng"!, ng#! sia! ke! : bele a bo '' ''I look all around, only to see this forest of people'' ''Ma a! ng"ng"!, ng#! sia! ke! : yo! mbo! a bo '' ''I look all around, only to see this world of people'' (Verse 8b, the character, Komba (God), sings ) ''Ma a! ng"ng"!, ng#! sia! ke! : bele a bo '' ''I look all around, only to see this forest of people'' ''Ma a! ng"ng"!, ng#! sia! ke! : yo! mbo! a bo '' ''I look all around, only to see this world of people'' Many song texts are near verbatim repetitions of previously spoken texts, as exemplified in the following excerpt from the story D("ngb(" (Example 2). Example 2: Kilian Hatz (1989), D$"ngb$" ," v. 9 and 10 (near verbatim sung reiteration of previously spoken phrases) (Verse 9, the narrator speaks ) li ndu!mu! a! Su$ a! e$ %a! m""! 'kp"ng! kp"ng! The sound of the Leopard's drum goes, 'Kp$ng! Kp$ng!' (Verse 10, the narrator sings ) ya$ ya bi$ na$ siki si! ki! Friend! Move! Move! ya bi$ na$ e de & ? li ndu!mu! a D"!ngb"!, e$ %a! m""! si! 'ki Friend don't you agree? The sound of Leopard's drum moves [us to dance]. No less common is a song's simple echo of the spoken name of a story's principal character, as in the case of the name Kpa! ngbala in the story Kpa! ngbala (Example 3). Example 3: Le' onard (2003) Kpa# ngbala v 6, 8, 10 and 38 (sung reiteration of previously spoken words, e.g., proper names ) (Verse 6, 8, 10, the narrator intermittently speaks the name of the main ch aracter) (6) Wa$ k#t#! t" bit" Kpa$ ngbala nyi$ a$ o$ They go to live with their mother, Lizard. (8) Wa$ m""! pe Kpa$ ngbala bela na jo. They work for their mother, Lizard, preparing food.
150 (10) Ek#!, Kpa$ ngbala e$ pe, '' w# la! a ng"$o$ wo$ be p"$ wa! wode! However, Lizard says that her daughter in laws don't gather any firewood for her. (Verse 38, the character, Kpa'ngbala, sings ) k" li ng" t" k" li ng" t", ha yo ha yo k" li ng" t" e! (archaic, unknown words) ''e! e$ Kpa$ ngbala ma m""! pe?'' ''Mother Lizard, what am I to do?'' Lyric texts not only reiterate spoken texts; spoken texts may occasionally reiterate lyric texts. In the story of Se! ko the protagonist's central imperative ( Go away! ) is first expressed in song (Example 4, Verse 13), and then l ater reiterated in speech (Example 4, Verses 2, 24, 25). Example 4: Kilian Hatz (1989), Se" ko" ," verses 13, 22, 24, 25 (spoken reiteration of previously sung phrases) (Verses 13, the character, Chimpanzee, sings ) (choir) ''mbamba t#$ mbamba t#$ l"!ji k mbamba t#$ mbamba t#$ (meaning unknown) Go away! (meaning unknown) (storyteller) l"!ji l"!ji k"!; t" Ka! ka! ? t" Ba$ ka? mbamba t#$ T" pe m"? l"!ji k"!'' Go! Leave here Are you a villager? A Baka? (?) A spirit? Go away! (Verse 22, the ch aracter, Chimpanzee, speaks ) '''!#ngi na sika! le! ? Oka! L"!ji pee! '' Were you waiting for me? Come on! Go away '' (Verse 24, the character, Chimpanzee, speaks ) ''e mo! la! ? T" pe Ka! ka! ? T" pe Ba$ ka?'' ''Who are you? Are you are villager? Are yo u a Ba# ka?'' (Verse 25, the character, Chimpanzee, speaks ) ''T" pe m"? e mo! la! ? L"!ji pee! !'' ''Are you a spirit? Who are you? Get away from me!''
151 Grammatical cohesion through lexical substitution. As illustrated above, lexical reiteration signals inter textual cohesion both iconically and symbolically: reiterated words first spoken, then sung, or sung then spoken bring about inter textual connections because they sound the same, and mean the same. However, when particular sung words are interprete d as lexical substitutes for certain previously spoken words and vice versa inter textual cohesion is signaled symbolically; that is the words do not sound the same, but do have the same or similar meanings. 9 In Baka li ka! n"! anaphora is a common type of lexical substitution, therefore often functioning as a symbolic semitoic sign of textual cohesion. Through anaphora, back referential pronouns in li! ka!n"! song texts function intertexually as grammatical substitutes for anteceding spoken li! ka! n"! narrative subjects (Example 5). Example 5: Le' onard (2003), Mbo" a se" ka ," lines 11 12 (Line 11, the narrator speaks ) Pe e$ k#t#! mb#li t#$, e$ j"! e! mbo! a! se! kao When he arrives there, he hears Deerflies (Line 12a, the n arrator continues to speak ) wa $ tongoa! t" na& %e! t" 'e l"$. They start to sing his name (i.e., about him) (Line 12b, the characters "Mbo"a" se" ka" sing ) ''kpa Komba ko mbi! ndo! na$ '' ''God's hand is filthy.'' (Line 12c, the narrator sings about "them," i.e., Deerflies) E $ ng"ng"! lu! e $ ng"ng"! lu$ u $ u $ 0/ They (he/she) mock [ God ] .. They (he/she) mock [ God ] [Eng. trans., mine] Grammatical cohesion through common point of view. Every Baka li! ka! n"! song is connected to a li! ka!n"! story by way of a commo n point of view The point of view taken in a story's song is most often that of a story's principal character,
152 occasionally that of the narrator, or both (though the personal identity of the song's voice is not always explicit). In the course of the sto ry song of L o! ndo for example, the narrator occasionally tells us explicitly that it is the main character, Lo! ndo! (Otter), who is speak ing or s ing ing (Example 6) 11 Example 6: Higgens (1981), Lo" ndo" ," Verses 5f 6b, 9a b (Lines 5f 6a narrator narrates the acti on and speech of the main character, Otter) Lo! ndo! pe : ''Ma a! b" mb""! n"! mo! subu! .'' Na g# Lo! ndo! k#! yi$"! Otter says ''I will go a little way downstream.'' Otter goes (Line 6b: the main character, Otter, sings) ''Hi, hi lu! ka! ny"$"! ngo, A g# a lu! ka! ny"$"! ngo'' ''We are going to prospect at the great river, hi hi'' (Line 9a the narrator narrates) E be!" di! di$ li$ na be be! k#!p" Finally, he [Lo" ndo" (Otter)] sings sings the whole song (Line 9b the main character (Otter) sings) ''H i, hi lu! ka! ny"$"! ngo, A g# a lu! ka! ny"$"! ngo'' ''We are going to prospect at the great river, hi hi'' As a li! ka! n"! unfolds, the point of view in the spoken discourse may frequently change between that of the narrator and one of the principal cha racters; the voice of the narrative's sung discourse, however, with its smaller text load, typically expresses but one point of view, most often, that of a principal character. The following excerpt from the story Ka$ la$ (Example 7) briefly illustrates this tendency: the song refrain (Verse 8b) only expresses the main character s point of view ( Komba ), but the immediately preceding and ensuing spoken verses narrate two points of view, first through the voice of the main character (Verse 8a) and then thro ugh the voice of the narrator (9a).
153 Example 7: Brisson (1996), Ka# la# ," Lines 8 9a (Line 8a, the character, Komba (God), speaks) ''I w"!" na gee! le! ge$ e! !'' ''So that, then, was why you were looking for me.'' (Line 8b, the character, Komba (God), sings) ''Ma a! ng"ng"!, ng#! sia! ke! : bele a bo'' ''I look all aroun d, only to see this forest of people'' (Line 9a, the narrator narrates) E! e na$ a! m""! k#!, d""! E k"!, Ka$ la$ o$ wa$ wee! t" li! ba! njo! nyi! a! o$ %o yi$"! So then, the Crabs, they come and touch the forehead of their Father (God) While the point of view of a li! ka! n"! song is most often that of a principal character, occasionally the voice of the narrator sings, as in the story Gbanga ( Example 8, Verse 5). Example 8: Brisson (1999) Gbanga 4b 5 12 (Line 4b, the narrator narrates) na g# gbanga a! bu ngo k#! yi$"! the calebasse goes down the river (Line 5, the narrator voices the song) oka! Gbanga wa! ni! oo! a! mi! na! nku! e! a! mi! na! nku! e! gbanga ti$ ta! oo, gbanga wa! ni! (e! Let's go! [let's sing!] 13 Calebasse of God! Oh! Nothing bu t water! Nothing but water! Uncle's calebasse, Oh! God's calebasse. [Sing!] S ongs typically express only one point of view, yet two are apparently not unheard, as in the following example from the li! ka! n"! Gbanga in which half of the refrain is the v oice of the Deerfly characters and the other half is that of the narrator (Example 9, Verses 12b and 12c, respectively).
154 Example 9: Le' onard (2003), Mbo" a se" ka ," Verses 12a,b,c (V.12a, the narrator narrates) wa$ tongoa! t" na& %e! t" 'e l"$. They start to sing his name (i.e., about him) (V.12b, the characters "Mbo" a"se"ka" sing) ''kpa Komba ko mbi! ndo! na$ '' ''God's hand is filthy.'' [Eng. trans., mine] (V.12c, the voice of the narrator sings) E$ ( ) ng"ng"! lu! e$ ( ) ng"ng"! lu$ u $ u $ [null' symbo l added] They mock God.. They mock God. [Eng. trans., mine] Over the entire course of a li! ka! n"! then, both the narrator, or a narrative character, may speak or sing, or both. In either case, and most importantly both texts sung or spoken, poetic o r prosaic are united by a common point of view a common voice The grammatical effect is inter textual cohesion, whereby a common voice weaves together a narration of sung texts and speaking texts. Thus, in contrast to other narrative performance tradit ions that include singing, li!ka! n"! songs do not merely comment on, embellish, or "frame" a narrative through some anonymous, disembodied voice but speak as an integral voice from within the narrative (cf. Agawu 1995:165 179). 18.104.22.168 Semantic c ohesion In addition to the typical lexical devices of inter textual cohesion mentioned previously ( i.e., lexical repetition and anaphora in Sec. 22.214.171.124) Baka li! ka! n"! and their songs occasionally signal their relatedness semantically particularly through the devices of synonymy and mero nymy (Fig 4 4).
155 Synonymy. Inter textual lexical synonymy is not merely limited to synonymous words. In Example 10, the synonymous lexemes are the verb na subu! nga! ( to walk in water ) and the ideophone nja! nga! mu ( walking in the water sound ). Example 10: Boursier (1994), Mbi" lo ," Verses 30b, 35 (Verse 30b, the character, Mbi"lo, speaks) e k#! k"! n"!: na subu! nga! 14 e k#!, I g# t" j#!, ga! je pa$ sa$ it is this: to walk in (the) water ; (you) cross to the other side of the river. (Verse 35, the cha racter, Mbi"lo, sings) Mbi! lo, k# le! k#!, nja! nga! mu taa! ba! le! Ngba! nda! mo! p"!ng#, nja! nga! mu taa! ba! le! Ma mo! o! lo Mbi! lo de, nja! nga! mu taa! ba! le! Ngba! nda! mo! ka! ke, nja! nga! mu taa! ba! le! Mbi" lo, my husband, then [ walking through water idio ] truly in the river, I'm testifying to the truth [ walking through water idio .] truly in the river, I did not murder Mbi" lo [ walking through water idio .] truly in the river, I'm bearing witness to the truth [walking through water idio .] truly in the river, Neither of the two synonymous lexemes, however, are necessarily word s In the same li! ka! n"! excerpted above, the intertextual cohesion of the li! ka! n"! 's sung and spoken texts is further signaled by the utterance of two synonymous ideophones : tchubu and nja! nga! mu both meaning "walking in the water sound" (Example 11). Example 11: Boursier (1994), Mbi" lo ," (1994), Verses 12, 35 (Verse 12, 15 the character, Mbi"lo speaks) Tchubu tchubu n"! a! bu ngo [ Walking in the water sound (idio), Walki ng in the water sound (idio.)] here in the stream (Verse 35, the character, Mbi"lo, sings) Mbi! lo, k# le! k#!, nja! nga! mu taa! ba! le! Ngba! nda! mo! p"!ng#, nja! nga! mu taa! ba! le! Ma mo! o! lo Mbi! lo de, nja! nga! mu taa! ba! le! Ngba! nda! mo! ka! le, nja! nga! mu taa! ba! le!
156 Mbi" lo, my husband, then [ walking through water sound (idio .)] truly in the river, I'm testifying to the truth [ walking through water sound (idio .)] truly in the river, I did not murder Mbi" lo [ walking through water sound (idio. )] tru ly in the river, I'm bearing witness to the truth [walking through water sound (idio .)] truly in the river, Meronymy. The last inter textually cohesive device to note is meronymy In the story of Sa $ ngo! ngo ," the narrator tells us that Sa $ ngo! ngo b ird is off in search of yams, singing along the way. As Sa$ ngo! ngo bird comes upon each yam plant, his lyric does not explicitly indicate that he has found a yam, rather, he merony mi cally implies it by describing each kind of yam flower that he comes upon (Example 12). Example 12 Higgens (1981) Sa# ngo" ngo Verses 2a, 2c, & 9 (Verse 2a, the narrator narrates) Sa$ ngo# ngo k#! la%o! a! ga! je na! k#$n#! sapa! yi$"! de? That Sa# ngo"ngo bird goes off in one direction... [and] isn't he looking for yams ? (Ve rse 2c, the narrator narrates) Sa$ ngo! ngo a$ sia! "$k"! n##! mo! fi! ma Sa# ngo"ngo bird sees another yam (stem). (Verse 9, the character, Sa'ngo" ngo bird, says, then sings) Sa$ ngo! ngo a$ do" njamba di! di$ li$ Sa# ngo"ngo bird has started up singing, and goe s on and on. i", Ko$ ngofa$ lo i" i" i", Ko$ ngofa$ lo! i" i" i$, Ko# ngofa# lo" i$ i$ i$, Ko# ngofa#lo" i$ i$ e$ ga! je na ma! nju! mba He is by the yam flower e$ ga! je na be! le! bo He is by the yam flower e$ ga! je na ma! nju! mba ko He is by the yam flow er there
157 4.3 Narrative Discourse Development through Song Having described the intrinsic unity of li!ka! n"! song in li! ka!n"! narrative performance, that is, the cohesive relationship of li! ka! n"! 's sung mode and speech mode discourse, I now turn to cons ider the internal, purposive nature of that unity how li! ka!n"! song serves narrative discourse development As introduced earlier, li! ka! n"! is shown to effect two kinds of narrative development: contextual development and climactic development (Figure 4 5). The particulars of contextual development will be discussed in Section 4.3.1 and those of climactic development in Section 4.3.2. First, however, both of these dynamics of narrative development will first be distinguished from the more prominent and fundamental notion of storyline development, for such a distinction is critical to recognizing the distinctly narrative discourse functions of li! ka!n"! song in li!ka! n"! story. Storyline prominence. Among many discourse analysts, the most salient developmental dynamic of narrative discourse concerns the formation of the narrative's "storyline." "It is commonly recognized that viable discourse h as cohesion/coherence and prominence (emphasis added; Longacre 1996:33). Longacre refers to storyline prominence, that is the basic constituitive elements of the narrative framework, as "mainline" material; Dooley and Levinsohn use the term "foreground "; Grimes describes it as "event" information ( Longacre 21 24; Dooley and Levinsohn 41 44 ). Discourse analysts (particularly in textlinguisitics) not only single out a narrative's mainline/foreground/event material, but also distinguish it from the same narrative's supportive/background/non event material ( ibid. ). Grimes asserts that the first
158 distinction made in the analysis of discourse is between events and non events (emphasis added; Dooley Levinsohn 2001: Chap. 12:2, citing Grimes 1975: 35; see al so Longacre 21 for others). To distinguish event and non event, foreground and background, mainline and supportive material, Longacre and others maintain that such a search is most often signalled by "a characterisitic constellation of verb forms," or rat her, is marked by particular "uses of a given tense/aspect/mood form" (2003:39 48). Discourse grammarians are coming to recognize more and more that in the telling of a story in any language, one particular tense is favored as the carrier of the backbone o r storyline of the story while other tenses serve to present the background, supportive, and depictive material in the story (ibid, 59, 64) Thus, "f or any language, each type of text [i.e., genre of discourse] has a mainline development and contains other material which can be conceived of as encoding progressive degrees of departure from the mainline" (Lonagacre 1996:23, citing Longacre 1989a). Such is the case in Baka narrative. In his discourse analysis of Baka li! ka!n"! Le! onard modeling Dooley and L evinsohn's methodology reports that in Baka li! ka! n"! foreground events are carried out with the perfective/ narrative aspect marked by [ a ] which elipses with the pronoun in the 3rd person singular and plural. Perfective/Narrative aspect (verb to go') 1.SG ma a" g & 1.PL.(excl) nga a" g& 2.SG mo o" g & 1.PL.(excl) nga a" g& 3.SG.(human) a g & 1.PL.(incl) a g& ni" 3.SG.(neutral) a g & 1.PL(dual) a g& ni" 2.PL i a" g& 3.PL wa' g&
159 The nar rative [ aspect ] has two functions in discourse: It indicates completeness of the immediate action, and it indicates [strict] chronological sequence (emphasis added; 2005:13) Complementarily, "the background in Baka narrative allows the use of different t enses" (emphasis added; ibid. 14) Thus, the default grammatical form used to signal Baka storyline is the narrative aspect; background ( non event ) material is uniquely marked with other tense aspect forms; and nowhere is this binary distinction of foreg round versus backgound more apparent than in the case of li! ka! n"! song for the perfective narrative aspect is never observed in any of the song text transcriptions at hand. Moreover in all of the six collections of Baka traditional stories (noted earl ier), li! ka! n"! song rarely if ever, manifests constituitive elements of the basic narrative framework of a li! ka! n"! that is, its mainline narrative events. Instead, li! ka! n"! song communicates supportive elements Song text clauses with their non narr ative tense aspect status uniquely serve as contextual material, and are therefore marked by either the imperfective aspect (Ex amples 13 14), distant and recent past tense markers (Ex 15 1 6), general future tense (Example 17), or more often as verb forms with no tense aspect at all (Ex. 18 19). 16 Examples 13 and 14 (below) demonstrate performances of the imperfective aspect: Example 13: Kilian Hatz (1989) D$"ngb$" Verse 1 0 [Choir sings ] ya$ ya bi$ na$ siki si! ki! // ? ? / ami / la" / bouger / bouger // Friend! Move! Move!
160 [Soloist sings ] ya bi$ na$ e de & ? li ndu! mu! a D"!ngb"!, e$ %a! m""! si! 'ki! // ? ami / la" / n(est ce pas / son de tambour // // de / antilope / il / IMP / fait / bouger // Friend don't you agree? The sound of Leopard's drum moves [us to dance]. ( To avoid confusion, t he [ a! ] aspect in the Ma a! ng#ng# clause in Example 14 should not be mistaken for a perfective narrative aspect marker; it is, rather, a phonologically eroded variant of the imperfective marker 'a! (Kilian Hatz , translated by Le! onard 2007:29). The potential for this formal confusion is most apt to occur in the conjugation of 1.S, 1.PL.EXCL, 1.PL.INCL 1.PL.DUAL, and 2.PL structures. The discourse context of the verb phrase, however, clarifies the intended aspect; as is the case in Example 14 where the mere repetition of the verbal phrases alone indicates that the action of looking side to side is habitual', progressive' (ibid ). ) Example 14: Brisson (1996), Ka# la# ," Verses in 8b (Verse 8b, the voice of the character, Komba (God), singing) ''Ma a ng"ng"!, ng#! sia! ke! : bele a bo'' // 1S / IMP / look side to side / OBL / see // forest / POS / people // ''I look all around only to see this forest of people '' ''Ma a ng"ng"!, ng#! sia! ke! : yo! mbo! a bo'' // 1S / IMP / look side to side / OBL / see // world / POS / people // "I look all around only to see this world of people. Among the twelve transcriptions under analysis only one lyric (Example 15) employs the past tense; in this case the distant past tense. Example 15: Boursier (1994), Mbi" lo ," verse 35 [the character, Mbi" lo sings] Ma mo! o lo Mbi! lo de, nja! nga! mu taa! ba! le! // 1.S / murder / P3 / NEG / idio.walking through water / truly / ri ver // I did not murder Mbi" lo [ walking through water idio.] truly in the river,
161 Despite this solitary example from the twelve principal narratives under investigation it is not, however, uncommon to find other non mainline clauses within the larger c orpus of 109 documented narratives whose verb constructions also u se past tense s not to mention future tenses or no verbal aspect marker at all Example 16 illustrates singing of the recent past (glossed "P1") Example 16: Brisson (1999:162), Gi" li" m a ndo" ," verse 22b ma '#!" mo! w "!" 'e! e a le! a nda! e! /je/ ai laisse' /toi/ auxil. //chose/de/moi/a" /maison/he'/ [Brisson's French gloss] //1S/leave. P1 /2S/be. P1 /thing/POS/1S/POS/house/he' .ideo// [English gloss, mine] Je t avais laisse # tu e# tais mon bien, la" dans la maison! He# e#! Example 17 illustrates singing of with the general future tense (glossed "fut.") Example 17: Brisson (1999:206), KPINYA III ," verse 11 k#!mb" ma m""! la! pe // fut. /je/fais/comment// Qu'est ce que je vais faire? What am I go ing to do? [Eng. trans., mine] And finally from our primary corpus of twelve narratives E xample s 18 and 19 illustrate lyrics with no aspect marker in the verbal syntagm (see '; also, Lonard 2003: Sec. 5.2.2) Example 18: Brisson (1996), Ma# ng%, Verse 9 Sima longo (K.), ma ( ) kp" l#!k#! //sima/longo(K.)/1S/fear/unclean// [English gloss, mine] Miracle! Je n'aime pas la salete# (J'ai peur de la salete# !) (Miracle! I don't want (to be near)/ am afraid of this ritually unclean corpse). [Eng. trans., min e] Example 19: Le' onard (2003), Mbo" a se" ka Line 12c (the narrator sings about Deerflies)
162 E$ ( ) ng"ng"! lu! e$ ( ) ng"ng"! lu$ u $ u $ [null' symbol added] They mock God.. They mock God. [Eng. trans., mine] 4.3.1 Contextual Narrative Development through Song While it is informative to note quantitatively the relative use of the various non narrative tense aspect forms in li! ka! n"! lyrics, it is more pertinent to this study to move on from this elemental distinction and begin to more finely characterize s ong's multiple structural functional role s in Baka li! ka! n"! So, having posited that song does not ever constitute a li!ka! n"! 's basic narrative "skeleton," but rather, its supportive "sinew," I will now move on to describe the particular kinds of suppor tive material it provides. For as fundamental as it is to determine song's basic role vis a vis the "event/non event saliency scheme" of a Baka li! ka!n"! what song does n o t do is not nearly as perti n ent as what it does do, and how. Grimes distinguished event from non event and divided the latter into setting, background, evaluations, and collateral 17 Setting was posited as having to do with "where, when, and under what circumst a nces actions take place" (p.51). By background he meant secondary informati on that is used to clarify a narrative (p. 56)... In reference to evaulations Grimes stated: "Not only do speakers report the state of the world; they tell how they feel about it (p. 61). He mentions that evaluations can be directly on the part of the nar rator or that he may put such an evaluation into the mouth of one of the participants in the story. As to collateral information, Grimes comments: "Some information, instead of telling what did happen, tells us what did not happen. It ranges over possibl e events, and in so doing sets off what actually happens against what might have happened" (p. 64). Under this heading Grimes includes most instances of negation, adversatives, questions, and the content of quotations (especially denial,
163 question, or pred iction (pp. 64 70) ( emphasis added ; Lonagacre 1996:23 24, citing Grimes 1975) 18 Dooley and Levinsohn build on Grimes' categories and add that These categories [i.e., setting, background, evaluation, collateral,] are not always mutually exclusive; bits of information in a text can belong to more than one, having more than one discourse function. Quite often, different kinds of information are mixed together in a single utterance (2001: 43) Many, if not all of Grimes' non event information subtypes are demonstrated in li! ka!n"! song texts, and often occur "mixed together." Figure 4 6 lists five of Grimes' six subtypes evidenced i n the lyric texts of the twelve li! ka! n"! under investigation. Participant orientation, setting, explanatory, and collateral i nformation are common, but e valuative information is most predomi n ant P erformative information is the only contextualizing material absent from the data 19 Examples of each non event text type are cited and explained in the sections that follow, Section s 126.96.36.199 through 188.8.131.52. 184.108.40.206 P articipant i nformation through song Occasionally, song texts "introduce, reintroduce, or describe in greater detail" particular narrative participants, as in the stories of Su$ a t# Ku! nda (Example 20) and Ku! njenje (Example 21). In the lyric of Su$ a t# Ku! nda the text makes explicit, as well as poetically highlights by way of paradigmatic substitution (Sec. 220.127.116.11 and 3.4.6 ) the local knowledge that the two principa l characters, Su$ a (Leopard') and Ku! nda (Turtle'), are not merely two disinterested and unrelated characters, but Su$ a is Ku! nda 's uncle ( ti$ ta ) (Example 20) 20 21
164 Example 20, Fitzgerald ( Appendix A), Su#a" t$ Ku" nda Verse 8 L "! k la & mo o! l ti$t a! ( e)? La & mo o! l Su$a !( e)? Ku! nda! Ti$ ta !( o ), la & mo o!l ti$ta !( e)? Here, who killed Uncle ? Who killed Sua (Leopard)? Ku" nda (Turtle)! O Uncle who killed Uncle ? In the story of Ku! njenje (Example 21), local knowledge of the narrative setting and natural habitat of the protagonist, Ku! njenje is made explicit in the songs lyric. Ku! njenje is held to resemble both a duck and a chicken (personal conversation; Brisson 2002:213). Local knowledge of Ku! njenje s peculiar forest role as nu na ngo (wa ter bird) and nu na bale (river bird) is made explicit and reinforced in the songs text. Example 21, Fitzgerald, Ku" njenje Ku!njenje, Ku!njenje( o), Ku!njenje, nu na! ngo ... Ku!njenje( o), Ku!njenje, nu na! ba! le !... Ku" njenje! Ku" njenje( o)! Ku njen je... bird of the water ... Ku" njenje! Ku" njenje( o)! Ku njenje... bird of the river ..., 18.104.22.168 S etting i nformation through song Just as frequently as participant orientation information is included, setting background information is occassionally embe dded in a lyric clause. 22 In the song of the story Ka$ la$ (Example 22), Komba ( God) locates, both generally and specifically, the scene of the storys conflict, that is the (humanly) inhabited forest ( bele ), or the worldof people ( yo! mbo! ) 23 Example 2 2: Brisson (1996), Ka# la# Verse 8b (Verse 8b, the character, Komba (God), sings) ''Ma a! ng ng "!, ng #! sia! ke! : bele a bo '' ''I look all around, only to see this forest of people '' ''Ma a! ng ng "!, ng #! sia! ke! : yo! mbo! a bo '' ''I look all around, only to see this world of people ''
165 In the story of Sa$ ngo! ngo (Example 23), Sa$ ngo! ngo has wandered off in search of wild yams ( safa! /sapa! ), and details his surroundings as he goes describing the specific types of flowering yam plants that he comes upon along his path. 24 Example 23 : Higgens ( 1981 ) Sa# ngo" ngo ," Verse 9 (Verse 9, the character, Sa'ngo" ngo bird, sings) i" Ko$ ngofa$ lo # i" i" i" Ko$ ngofa$ lo# i" i" i(, Ko! ngofa! lo" i( i( i(, Ko! ngofa! lo" i( i( e$ ga! je na ma! nju! mba! He is by the (large) yam flower e $ ga# je na be# le# bo He is by the (small) yam flower 22.214.171.124 E xplanatory i nformation through song In addition to the "participant orientation" and "setting" information subtypes, explanatory information may also inform a lyric clau se and by comparison, does so more frequently In my initial analysis, I simply searched for explanatory material in adjectival or adverbial phrase constructions, but eventually found that the most pr evalent explanatory material s were communicated throug h ideophone s Explanatory information through sung ideophones Kilian Hatz's (1989, 2001) fruitful study of Baka ideophones provides a most helpful orientation to the signifying potential of Baka ideophones. This present section on explanatory backgrou nd information is but one domain of discourse in which ideophones are prominent. In an early, introductory description of Baka ideophones, Kilian Hatz writes, [Ideophones] serve the [Baka] story teller as a stylistic device allowing him to describe as re alistically as possible what happened and so give the
166 listener proof of the authenticity of the story or experience (Kilian Hatz 1989:37) 25 Le' onard later confirms the descriptive power of Baka ideophones in Baka narrative discourse. These short utterances provide greater precision in the description of an action that would otherwise require sentences to communicate. A single ideophone can d escribe with accuracy abstract concepts, such as sounds, feelings, odours, colours, actions, or even complete activities. (Lonard 2005:11) Baka ideophones are either nominal or verbal, though most often verbal, as is reflected in the li$ ka$ n"! songs of t he narratives Lo! ndo! (Ex. 24) Su$ a t# Ku! nda (Ex. 25) Mbo! a se! ka (Ex. 26) and Mbi! lo (Ex. 27). In the story of Lo! ndo! (Example 24) the reduplicated verbal ideophone hi (to go away quickly') symbolically signals a more realistic and inte nse image of the main character's "going" in the forest, that is the continuous, brisk nature of his walking and searching. 26 Example 24: Higgens (1981), Lo" ndo" ," Verse 6b (Line 6b: the main character, Otter, sings) A g # a lu! ka! ny"$"! ngo... '' hi, hi lu! ka! ny"$"! ngo... // 3S / go / INF / to prospect / mother / water / IDE. "to go away quickly / to prospect / mother / water // ''We are going to prospect at the great river, hi hi '' In another song, from the li! ka! n"! of Su$ a t# Ku! nda (Example 25 ) the verbal and reduplicated ideophone kp" (to chop with an ax') symbolically signals a more vivid representation of the actors busily "chop... chop... chopping" the "tree of calamity." Example 25, Fitzgerald (Appendix A), Su# a t$ Ku" nda ," Verse 39 Nga buu wa! na! wee! ( eeee)... Nga buu! wa! na! wee! ( e)... kp#$! kp#!, kp# We are cutting down the "tree of calamity"... Chop! Chop! Chop!
167 In the following excerpt from the story of Mbo! a se! ka (Example 26) the verbal ideophone lu$$ (rendered "l u! u! u! when subjected to vowel lengthening, to indicate continual action) more realisitically portrays the mocking gesture the Deerflies perform as they continually and activ ely "shake their heads and eyes" at Komba Example 26: Le' onard (2003), Mbo" a se ka ," Line 12c [the narrator sings about "them," i.e., Deerflies] E$ ng"ng"! lu! e$ ng"ng"! lu$ u $ u $ // 3S/agite/action de se tourner.[IDEO] / 3S/agite/ action de se tourner.[IDEO] // [ideophone marker added; Brisson 2002:285] They mock ingly shake their heads a t God. [Eng. trans., mine] And for a final example of the use of an ideophone as a device for conveying background information, I cite the narrative of Mbi! lo (Example 27) Here, the nominal ideophone nja! nga! mu symbollically encodes "the sound of walk ing in water," evoking a more realisitic sonic environment during the climax of the narrative's conflict, as the accused wades through the poisoned waters. Example 27: Boursier (1994), Mbi" lo ," verse 35 [the character, Mbi" lo, sings] Ma mo! o! lo Mbi! lo de, nja! nga! mu taa! ba! le! // 1.S / murder / P3 / NEG / idio.walking through water / really / river // I did not murder Mbi" lo [ walking through water idio .] really in the river, 126.96.36.199 C ollateral information through song: quoted questions and denials I n addition to the occasional presence of participant, setting, and explanatory information in the contextual discourse materials of li! ka!n"! song, collateral information is sometimes included. Some information, instead of telling what did happen, tells wh at did not happen. It ranges over possible events and in doing so sets off what actually does happen against what might have happened. Under this
168 heading Grimes includes most instances of negation, adversatives, questions, and the content of quotations (e specially denial, question, or prediction. (Longacre, 24; citing Grimes 1975:64 70) Collateral information, simply stated, relates non events to events. By providing a range of non events that might take place, it heightens the significance of the real ev ents. (Grimes 65) In li! ka! n"! song lyrics (the content of) quoted questions of the principal characters are the most typical form of expressing collateral information; quoted denials are less frequent, but not uncommon. The question voiced by the main c haracter in the song of Se! ko! (Example 28) for instance brings into relief Se! ko! 's concern over the unknown identity of a creature in the distance along his path, that is, whether or not it is "friend or foe The future "eventline," in turn, hinges on this non event collateral question. Example 28: Kilian Hatz (1989) Se" ko" verses 13 (Verses 13, the character, Chimpanzee, sings) l"!ji, l"!ji k"!; t" Ka! ka! ? t" Ba$ ka? mbamba t#$ T" pe m"? l"!ji k"!'' Go! Leave here! Are you a "villager" (Ba ntu)? Baka? (?) A spirit? Go away! In another example, from the li! ka! n"! song of Su$ a t# Ku! nda (Ex. 29), the protangonist, Ku!nda asks the question "W ho killed my uncle ( Su$ a ) ?" And like the function of the collateral question in Ex ample 28 ( imme diately above), Ku! nda 's (non event, collateral) question heightens the listener's expectation as to what will actually happen, that is, what significant event will take place, once the alledged tragedy is brought to light. Example 29, Fitzgerald (Append ix A), Su# a t$ Ku" nda ," Verse 8 L"! k", la&mo! o! l" ti$ ta! ( e) ? La&mo! o! l" Su$ a! ( e) ? Ku! nda! Ti$ ta! ( o "), la&mo! o! l" ti$ ta! ( e) ? Here, who killed Uncle ? Who killed Sua (Leopard) ? Ku" nda (Turtle)! O Uncle, who killed Uncle ?
169 In addition to quoted questio ns, quoted denials can communicate collateral information in song texts, though less frequently. In the narrative of Mbi! lo (Example 30), each of the accused denies murdering their husband ( Mbi! lo ) Their denials heighten the anticipation (in the liste ner) of a likely future event their potential demise from the swallowing of a poisoness truth serum that will potentially reveal the truth of their respective professions. Example 30: Boursier (1994), Mbi" lo Verse 35c [ a wife of Mbi" lo sings] Ma mo! o! l o Mbi! lo de nja! nga! mu taa! ba! le! // 1.S / murder / P3 / NEG / idio.walking through water / really / river // I did not murder Mbi" lo [walking through water idio.] really in the river, 188.8.131.52 E valuative i nformation through song Of all the contextu al information subtypes found in li!ka! n"! song texts, evaluative texts predominate. Through them "speakers not only report the state of the world," but how they feel about it ." Evaluative materials in li! ka! n"! song texts may be encoded propositionall y, lexically, grammatically, rhetorically, poetically, and musically. They may report the speakers' subjective assesment of the relative virtue of the state of affairs at hand; or as is more often the case, they may express the speakers' feelings or atti tudes toward another person or about circumstances in the past, present, or future The affective evaluations expressed through Baka li! ka!n"! songs are numerous; they encode feelings of anxiety, grief (or pseudo grief), annoyance, obligation, eagerness, anticipation, fear, self pity, dread, helplessness, victimization, offense, shock, scandal, isolation, frustration, unhappiness, lonli ness, antagonism, dislike, intoleration, confidence, self satisfaction, and delight.
170 The following sub section s of this sub section (Sec. 4.3 .1.5) highlight the most prevalent devices found in Baka li!ka! n"! song for the expression of evaluative disco urse: propositions, obligative mood, imperative mood, exclamatives, interjectives, ideophones, intensifiers, rhetorical questions, and vocatives. Evaluative information through propositions. Explicitly evaluative utterances are most plainly communicated propositionally In the song of Ma $ ng" (Example 31), the story's principal character expresses his fear of what might happen to him if he goes to his mother's burial. Example 31: Brisson (1996), Ma# ng%, Verse 9 [ the character, Ma# ng% sings] Sima longo (K.), ma kp" l#!k#! Miracle! Je n'aime pas la salete# ( J'ai peur de la salete# !) (Miracle! I don't want (to be near)/am afraid of this ritually unclean corpse). [Eng. trans., mine] And in another story (Example 32), the protagonist, Bo!ki! sa! (the solitary male gorilla), plainly lament s his isolation when he sings wa$ m#!l" (malheureux' Fr., poor me' Eng.) as he complains of his lonely predicatment. Example 32: Boursier (1994) Bo" ki" sa line 22 [the character, Bo"ki"sa, sings] 'a! l#!ng#! na mo! bi wa$ m"!l# 'e! / se' jour / de / bo uche buisson / malheureux / he' / Se# jour de l'entre# e du buisson Malheureux que je suis Opening day of entanglements poor me In the story song of Mbi! lo (Example 33) Mbi! lo 's wives flatly deny their culpability in Mbi! lo 's murder, and thus, assert the injustice of their trial.
171 Example 33: Boursier (1994), Mbi" lo ," Verse 35c [ a wife of Mbi" lo sings] Ma mo! o! lo Mbi! lo de nja! nga! mu taa! ba! le! // 1.S / murder / P3 / NEG / idio.walking through water / really / river // I did not murder Mbi" lo [wa lking through water idio.] really in the river, In the song of M bo! a se! ka (Ex ample 34 ), Deerflies express their opposition to God as he approaches; they mock him in song, singing that his "hand is filthy," that is that what he is likely to do is unj ust unfair immoral arbitrary. Example 34: Le' onard (2003), Mbo" a se" ka ," Line 12b [the characters Mbo" a"se" ka sing] ''kpa Komba ko mbi! ndo na$ '' ''God's hand is filthy .'' [Eng. trans., mine] Evaluative information through grammar. The import of m ost evaluative strategies, however, is not so explicit ly propositional. Instead, valuations are more often implied through the inherently heightened evaluative and expressive functions of certain grammatical forms. In song texts, the most common evalua tive grammatical forms are obligative and imperative moods, exclamations and interjections, vocatives, ideophones, intensifiers, and rhetorical questions. Evaluative information through obligative mood m arker In the lyrics of two story songs, Bo! ki! sa! (Ex. 35) and Ka$ la$ (Ex. 36) the principal characters utter the obligative mood marker ng"! (must') in order to express the obligation they feel to act on the intolerable situation in which they find themselves. In the story of Bo! ki! sa! (Ex 35), Bo! ki! sa! feels compelled to do something about his intolerable isolation and lon e liness. And i n the li! ka! n"! Ka$ la$ (Example 36 ), Komba (God') feels so annoyed with
172 the distracting sound of someone's ( Ka$ la$ 's) noisy singing that he is dr iven to set out to find the source of the noise and put and end to it. Example 35 : Boursier (1994), Bo" ki" sa "" line 22 [the character, Bo" ki"sa "" sings] o ng#! m""! taa! 'e! e, ng#! m""! taa! 'e! e, / oh / OBL 27 / faire.IMP / vrai / chose / OBL / faire.IM P / vrai /chose / "Oh! fais vraiment quelque chose, Oh! fais vraiment quelque chose," "Oh! Really... (you must ) do something! Really...(you must ) do something! [Eng trans mine] Example 36: Brisson (1996), Ka# la ," line 8b [ Komba sings] Ma a! ng"ng"!, ng#! sia! ke! : bele a bo Je regarde de tous les cotes, mais regarde donc! C'est la foret des gens! "I look all around, (but you should ) see this a forest of people!" [English translation, mine] Evaluative information through imperative mood marker In addition to the obligative mood, the imperative marker also encodes heightened feelings, and attitudes regarding the action and hand. In the story song of Se! ko! (Ex ample 37 ), the cowardly protagonist is inordinately leary of the unknown object in his path ahead an object which in the end simply turns out to be a stump In the face of a potential threat, Se! ko! call s out with false bravado and orders "it" to "Go (away)!", to "Leave!" ( l#!ji ) and let him pass. Example 37: Kilian Hatz, (1989) Se" k o verse 13 [ Se" ko" sings] l"!ji, l"!ji k" ; t" Ka! ka! ? t" Ba$ ka? mbamba t#$ T" pe m"? l"!ji k"! '' "Go! Leave here! Are you a villager' (Bantu)? Baka? (?) A spirit? Go away!"
173 And again in the song of Bo! ki! sa! (Example 38), Bo! ki! sa! compelle d by his growing sense of an obligation to take action, commands himself (at least interiorly) to do something about his intolerable isolation. Example 38 : Boursier (1994) Bo" ki" sa"" l ine 22 [ Bo" ki"sa" sings] o ng#! m ""! taa! 'e! e ng# m "" taa! 'e! e / oh / OBL / faire.IMP / vrai / chose / OBL / faire.IMP / vrai / chose / "Oh! fais vraiment quelque chose Oh! fais vraiment quelque chose ," (French translation, Boursier) "Oh! Really... (you must) do something Really...(you must) do something [E nglish trans., mine] Evaluative i nformation through e xclamative and interjective markers Other grammatico rhetorical devices may co occur with and augment the expressive effect of devices like exclamatives and interjectives; they may even function as th e sole textually heightening formalization of a song lyric. I n the song of Bo! ki! sa! (Ex ample 39 ), in addition to the obligative marker ng"! and the imperative m##!...e! e$ the exclamative O! and the intensifier taa! are added to the text to further enhance the lyric's expressively heightening effect. 28 29 Example 39 : Bou rsier (1994), Bo" ki" sa" l ine 22 [the character, Bo" ki"sa" sings] o ng#! m""! taa! 'e! e ng#! m""! taa! 'e! e / oh / OBL 30 / faire.IMP / vrai / chose / OBL / faire.IMP / vrai / chose / "Oh! fais vraiment quelque chose Oh! fais vraiment quelque chose ," "Oh! Really ... (you must ) do something Really ...(you must ) do something [Eng trans ., mine] Not all songs, however, bundle so many expressive signs in its lyric as does that of Bo! ki! sa! In the song of Ma$ ngo (Ex ample 40 ), for example, the exc lamatory phrase sima longo (miracle!' ) borrowed from the neighboring Djem language, is the sole grammatical formalization used to indicate the strength of the emotional impact of the
174 protagonist's experience, that is of his horror of coming in contact with his mother's dead body. Example 40, Brisson (1996), Ma # ng% ," Verse 9 Sima longo (K.) ma kp" l#!k#! Miracle! Je n'aime pas la salete# (J'ai peur de la salete# !) Miracle! I don't want [to be near ] ( am afraid of) this ritually unclean corpse. [Eng. trans., mine] Evaluative information through intensifier mar kers In another example, in M bo! a se! ka (Ex ample 41 ), the lone grammatical intensifier ko is chosen to heighten the already affectively pregnant poetic import of the mockery expressed in the lyric of the Deerflies. Example 41: Le' onard (2003), Mbo" a s e" ka ," line 12 [Deerflies sing] ''kpa Komba ko mbi! ndo! na$ '' // hand God / INT / dirty / DEM // "God's hand...(it)is really filthy." [Eng. trans., mine] Evaluative information through ideophones. Along with obligatives, imperatives, exclamatives, and in tensifiers, ideophones may occasionally, connotatively contribute to a lyric's evaluatory signal. The ideophone hi (to go quickly', as note d earlier in Ex. 24 from the story song of Lo! ndo! ) not only explicitly details the quality of the action at ha nd, but also implies the eager (perhaps carefree, though ultimately self centered) attitude with which the actor is carrying it out. Evaluative information through rhetorical questions Rhetorical questions may also constitute evaluative material, for whi le their direct illocutionary force is a question, "[they are] not generally used with the expectation of an answer, but with some different, indirect force, such as a command, or an evaluation (brackets and
175 emphasis added; see "illocutionary force" in Lo os 2004). In the story song of D#!ngb#! (Example 42), the effectiveness of the sound of D#!ngb#! 's drum to "move one to dance" is judged to be so self evident that the speaker confidently elicits that same evaluation from the others in the form of the rhe torical question "e de)?" ( n'est pas?' ). Example 42 : Kilian Hatz (1989) D$"ngb$" line 10 [Choir] ya$ ya bi$ na$ siki si! ki! // ? ? / ami / la" / bouger / bouger // Friend! Move! Move! [Soloist] ya bi$ na $ e de & ? li ndu! mu! a D"!ngb"!, e$ %a! m""! si! 'ki! // ? ami / la" / n(est ce pas / son de tambour // // de / antilope / il / IMP / fait / bouger // Friend D on't you agree)? The sound of Leopard's drum moves [us to dance]. [Eng. trans, nine] Another rhetorical question commonly uttered as a n evaluative expression is the interrogative phrase ma m##! pe?" (What am I to do?'). As in Example 43, i t expresses fear frustration, or more commonly, a cry of self pity. Example 43 : Le' onard (2003) Kpa# ngbala ," line 38 (the character, Kpa' ngbala, sings) k" li ng" t" k" li ng" t", ha yo ha yo k" li ng" t" e! (archaic, unknown words) ''e! e$ Kpa$ ngbala ma m""! pe? '' // thing / Lizard / 1S / do / ITR.what // "Mother Lizard, what am I to do? Evaluative information through vocatives Finally, t he mos t common grammatical device connoting evaluative and expressive illocutionary force is a vocative that is a noun whose referent is being addressed" ( Loos 2004 ). Seven of the
176 twelve stories sampled contain vocatives in their songs' texts; some songs may even employ more than one. According to McCormick and Richardson "t he functions of vocatives is a topic that is highly debated amongst linguists (2006 ). Vocatives are said to "express attitude, politeness, formality, status, intimacy, or a role relatio nship, and most of them mark the speaker, characterizing him or her to the addressee" (Zwicky, 1974) ... Elizabeth Axelson (2003) references research that identifies vocatives as "markers of power and solidarity (Hook, 1984), in group status (Wood & Kroger 1991, Brown & Levinson, 1978), or pseudo intimacy (McCarthy & O'Keeffe, 2001), equality (Troemel Ploetz, 1994) or condescension (Wood & Kroger, 1991) and as redressive action for facethreatening acts (Ostermann, 2000, Brown & Levinson, 1978)" the Lon gman Grammar of Spoken and Written English (1999) states that there are three main functions: Summoning Attention Addressee Identification Establishing and maintaining social relationships. (McCormick and Richardson 2006) Paul Portner adds to the debat e by suggest ing that the meaning of vocatives be formulated as expressive content ( emphasis added; Portner 2004 :5 ). 31 In Baka song texts, vocatives are most often proper names, though not always. Singers/speakers u tter vocatives to address other narrativ e characters and/or themselves. The illocutionary point of a sung vocative, that is, a speaker's basic purpose in uttering (or singing ) it, is at once directive and expressive ( Loos 2004 ). It is directive in that li! ka! n"! vocatives may summon attention elicit pity, solicit solidarity, plead for sympathy, call for help, and/or direct (one's) attention; it is expressive, and thus evaluative, in that vocatives may index (co n note ) feelings of anxiety, fear, self pity,
177 dread, anticipation, expectation, isol ation, vulnerability, self absorbtion, self satisfaction, obligation, lonliness, unhappiness, frustration, discontent, and/or anger. The referent of a li! ka! n"! song vocative is most often a proper name (Ex amples 44 49 ), though not always (Ex. 50 ). Further, the addressee of the vocative is most often the addresser. Thus, sung vocatives are often self referential (Ex amples 44 49 ) Through these vocativ es the speaker is primarily "talking to him/herself," revealing his/her inner thoughts, attitudes, and feelings about the narrative circumstance in focus. Indeed, as is evidenced by the pattern of the contexts in which these vocatives are uttered in our s ample, li! ka! n"! song vocatives are most often contextually associated with tense emotional circumstances, and thus, are indices of strong emotional experience. Expressive and directive purposes of vocatives need not be mutually exclusive; but rather, can be co mplementary, wherein strong emotional states act as perlocutionary forces that incite the addressee to act as a responsible agent in the face of a problem at hand. 32 In the songs of Kpa$ ngbala, Bo! ki! sa! and Su$ a t# Ku! nda for example, each of the stories' respective protagonists address themselves in order to incite themselves to action. For her part, Kpa$ ngbala (Example 44) frustrated and even angered with the undutiful behaviour of her daughter in laws, calls out her own name in song and index es her frustration in the question, "What am I to do?" Example 44: Le' onard (2003) Kpa# ngbala line 38 (the character, Kpa' ngbala, sings) ''e! e$ Kpa$ ngbala ma m""! pe?'' // thing / VOC.Lizard / 1S / do / ITR.what // Mother Lizard what am I to do?"
178 The self referential vocative uttered by the li! ka!n "! character Bo! ki! sa! in Example 45 however, does not so much collocate with an interogative and its accompanying feelings of frustration, but rather, with an inciting directive to Do something! given Bo !ki! sa! s vulnerable, and thus, anxious state of isolation (as a solitary male gorilla). Example 45: Boursier (1994) Bo" ki" sa" line 22 [the character, Bo ki sa, sings] O ng #! m ""! taa! ) e! e! ng #! m ""! taa! ) e! e! // oh/ asp. / faire / vrai / chose / asp. / faire / vrai /chose // Oh! fais vraiment quelque chose, Oh! fais vraiment quelque chose! O! You must do something, really! You must do something, really! [Eng. trans., mine] Bo! ki! sa! 'e 'a! l #!ng #! wa! m "$"! // VOC Bo" ki"sa" / he / se' jour / celui faire // Bo" ki" sa" he# se # jourcelui qui fait 'a! l #!ng #! na mo! bi wa$ m "!l # 'e! / se' jour / de / bouche buisson / malheureux / he/ Se# jour de l'entre# e du buisson Malheureux que je suis This negative, affective evaluation of solitary living is not, of course, simply the self expression of a single abstract narrative character such as Bo! ki! sa! Through the group performance of the narrative voice of Bo! ki! sa! the Baka people strongly and socially express in song their collective valuation of communit y over individuality In a slightly different example, the pragmatics of the self referential vocative of the character Ku! nda, in the story Su$a! t# Ku! nda (Example 46) indexes neither frustration nor anxiety, but grief and self pity though ingenuinely as Ku! nda pretentiously laments his uncle Su$a! s alledged death. Example 46: Fitzgerald ( Appendix A), Su#a" t$ Ku" nda Verse 8 [the character Ku" nda, sings; performed by storyteller singer and choir] L "! k la & mo o! l ti$ ta!( e)? La & mo o! l Su$ a!( e)? Ku! nda! Ti$ ta!( o ") la & mo o!l ti$ta!( e)? Here, who killed Uncle? Who killed Sua (Leopard)? Ku nda (Turtle) O Uncle who killed Uncle? [my translation]
179 Not all li! ka! n"! song vocatives are self referential. In the li! ka! n"! Ku! njenj e (Example 47) as Ku! njenje (Water fowl') floats down the rising river he utters his disgust ( ma a! kpo' ) with what he sees along the river bank. He does not, however, utter his disgust to himself Instead, he first draws attention to himself, identifying himself to others as the one speaking/singing the self identified "moral authority He then summon s those along the river bank to listen to him, Ku! njenje ," and stop all their ostentatious behaviour ( m"ng"m"!ng"! ). 33 And as I noted in the story of Bo! ki! sa! (E x. 45), the collective sung social performance of such evaluative strategies, through the voice of narrative character s like Ku! njenje not only develops the narrative context, but as a social act potentially signifies and reaffirms traditional Baka social values. Example 47: Fitzgerald (2008), Ku" njenje [Ku"njenje sings] Ku! njenje, Ku! njenje( o), Ku! njenje, nu na! ngo... Ku! njenje( o), Ku! njenje, nu na! ba! le! ... Ku" njenje! Ku" njenje( o)! Ku"njenje...bird of the water... Ku" njenje! Ku" njenje( o)! Ku"nje nje...bird of the river... [Eng. trans. mine] Similarly, the carefree and self centered Ko$ ngofa$ lo! (Ex ample 48 ) sings his own name not to address himself, but to make himself the narrative subject of his own "happy go lucky," self satisfied song. Exampl e 48: Higgens (1981), Sa# ngo" ngo ," Verse 9 [Verse 9, the character, Sa'ngo" ngo bird, sings] i", Ko$ ngofa$ lo i" i" i", Ko$ ngofa$ lo i" i" // go quickly.ideo. / Ko' ngofa'lo / go quickly.ideo. / Ko' ngofa'lo / go quickly.ideo. // The referent of a sung li ka! n"! vocative is not, however, always the addresser, but may occasionally be another character in the narrative. In Example 49 the wives of
180 Mbi! lo call out the ( proper ) name of their deceased husband, in fear and in search of pity during their present trial. The illocutionary point of their vocative is "to maintain and reinforce their [familial] relationship" and thereby gain an advocate in their crisis (Biber 1999:1108, cited in McCormick and Richardson 2006 ) Example 49: Boursier (1994), Mbi" lo ," v erse 35 (Verse 35, the character, Mbi"lo, sings) Mbi! lo, k# le k#!, nja! nga! mu taa! ba! le! Ngba! nda! mo! p"!ng#, nja! nga! mu taa! ba! le! Ma mo! o! lo Mbi! lo de, nja! nga! mu taa! ba! le! Ngba! nda! mo! ka! ke, nja! nga! mu taa! ba! le! Mbi" lo, my husband then [w alking through water idio.] truly in the river, I'm testifying to the truth [walking through water idio.] truly in the river, I did not murder Mbi" lo [walking through water idio.] truly in the river, I'm bearing witness to the truth [walking throu gh water idio.] truly in the river, Similarly, i n Example 50, the addresser calls out and refers to the addressee an anonymous narrative character, or a performing participant as "friend" ( bi ) in an attempt to elicit his solidarity of opinion regardin g the exceptional sound and power of the drum to move one to dance. Example 50: Kilian Hatz (1989) D$"ngb$" line 10 [Choir] ya$ ya bi $ na$ siki si! ki! // ? ? / ami / la" / bouger / bouger // Friend Move! Move! [Eng. trans, mine] [Soloist] ya bi$ na$ e de & ? li ndu! mu! a D"!ngb"!, e$ %a! m""! si! ki! // ? ami / la" / n(est ce pas / son de tam bour // // de / antilope / il / IMP / fait / bouger // Friend don't you agree? The sound of Leopard's drum moves [us to dance]. [Eng. trans, mine]
181 4.3.2 Climactic N arrative Development through Song: the Story of Lo! ndo! Having dwelt at some length on a description of the unifying and contextualizing/developmental functions of Baka song in Baka story, a brief recapitulation of the preceding sections may be helpful before moving on to the final section of my discourse analysis. From the point of view o f the broadest textual concerns of narrative discourse analysis, it has been necessary to account for how (if at all) li! ka! n"! song serves the fundamental narrative dynamics of these performances, with the essential dynamics, thus far, being narrative cohesion and prominence (or saliency ) (Sections 4. 2 and 4.3 ). Within the dynamic of prominence, I have shown that li! ka! n"! song texts rarely, if ever, signal foreground storyline content. Rather, a song text typically communicates background content, that is, it does not extend the basic narrative framework but primarily "aids in [the] internal and external contextualizat ion" of the narrative's storyline (Dooley and Levinsohn 2000:12). And further, of all the subtypes of background information, song texts most often signify evaluative information, that is expressive content. Yet song texts serve narrative development in other ways. The formal and functional development of any narrative not only forms a unity while simultaneously signifying peculiar supportive qualities more or less salient to the emerging storyline framework; but a narrative also progresses Longacre pu ts it plainly: While a discourse has cohesion/coherence and prominence, it just as necessarily involves progress a well formed discourse is going somewhere. The [macrostructure] progress of a discourse typically issues in some sort of climactic developme nt (or developments). (1996:33)
182 Of the various and potential signs of narrative "progress climactic development is typical. Longacre underscores the analytical importance of the recognition of narrative climax. It is especially crucial that we be able t o recognize the peak... in the surface structure. We can then identify prepeak episodes and postpeak episodes and can consequently better articulate the [entire] surface structure of the narrative (emphasis added; 1996:38) 34 Longacre characterize s clima ctic development (or "peak", as he prefers to call it) [as] essentially...a zone of turbulence in regards to the flow of the discourse in its preceding and following parts. Routine features of the storyline may be distorted or phased out at the peak. T hus, the characteristic storyline tense/aspect may be substituted for by another tense/aspect. (emphasis added; 1996:38) Such a substitution, of course, is true of the default grammar of Baka li! ka! n"! song ( as mentioned earlier in the "Storyline Prominen ce paragraphs of Section 4.3) There, we noted that the perfective narrative aspect that so normally constitutes the foreground of Baka traditional narrative is never heard in a li! ka! n"! song text (at least not in the data sample at hand). Thus, the p erformance of li! ka! n"! song and its attending non storyline verb forms has the innate potential to signal climactic development. Such a single "distorted" feature, however, is but one mark of "turbulence." Other "positive" features, according to Longacr e, commonly contribute to the evidence that development is afoot. Thus, in addition to the conditions of "minus [verb form] features and distortion," "peak has features peculiar to itself and the marking of such
183 features takes precedence over the marking of the mainline (1996:38). 35 And such generalized features are observed in sung li! ka! n "! discourse. Longacre posits six common marks/keys of narrative structure peaks: ( 1 ) rhetorical underlining, (2 ) concentration of participants, (3) heightened vividness, ( 4 ) change of pace, ( 5 ) change in vantage point, and ( 6 ) incidence of particle s and onomatopoeia (ibid., 39 48). Of these six signs, four are exhibited in the corpus of li! ka! n "! songs analyzed in this chapter. They are ( 1) rhetorical underlining, ( 2) heightened vividness, ( 3) change of pace, and ( 4) a shift of incidence of onomat opoeia (or as we shall specify later, a shift of incidence of ideophones ) Several, or even all four, of these four qualities commonly coales c e in song to form the zone of turbulence that typifies climactic development. Climactic development in the stor y and s ong of Lo! ndo! (Otter) In the following description of the climactic development of one traditional Baka story, Lo! ndo! all of the four common marks of sung climatic development (i.e., rhetorical underlining, heightened vividness, change of pace, and shif t of incidence of ideophone ) are exhibited. As each of the four key marks is presented, it will also be elaborated upon. By choosing to exemplify each of the four features from a single story as opposed to presenting select examples from all twelve narra tives not only will the particularities of each distinct mark be considered but also the ir cumulative complementarity should be made more apparent as well With some modification, Figure 4 7 (re)presents Kathleen Higgens transcription and English transla tion of the traditional Baka narrative Lo! ndo! ( Otter ) (1981:35 42). Higgens abbreviates her transcription of the sung refrain by presenting in italics only
184 its first and last iterations (6b and 9d). I have added gray highlighting to the representati on to indicate the continual repetition of the refrain as it simultaneously overlaps with the spoken narration. The relative sonic prominence and timing of the overlapping telling and singing is best appreciated, at least initially, by listening to the ac companying audio excerpt of the original field recording (Object 4 1. Audio file of li ka!n"! story Lo! ndo! (.mp3 3.4 MB)) (Higgens 1981). Figure 4 8, then, reframes Higgen's transcription in order to further disclose the particularly developmental structures and features of this narrative performance. What follows is a brief "reading" o f th e diagram in Figure 4 8. The reading summarizes the temporal representation of the plot and the relative distribution of singing and speaking to the plot's peak About a quarter of the way into the narrative [i.e., at :50 sec.], the principal charac ter ( Lo! ndo! ) shifts from speaking to singing (mode) at which time he sings the words we are going prospecting, hi hi, at the great river ." As he sings on his way to the river, Komba (God) hears Lo! ndo! 's song and becomes increasingly annoyed with his singing ( Lo! ndo! 's singing is "disturbing the peace of the forest" and hindering Komba from hearing the humming of the bees, the very sound that, according to local knowledge, invariably leads Komba to his prized forest honey.) The resulting two minute and fif ty second syntactic versification of the single clause "we are going prospecting at the great river issues in a refrain (of modest musical variation) that continues throughout the last three quarters of the story's performance [i.e., from :50 3:40 sec.] a nd terminates suddenly at the narrative's peak [i.e., at 3:40 sec.], that is,
185 the point at which Komba touches the forehead of Lo! ndo! "the person," and thus transforms him ( kwa, kwa, kwa ) into Lo! ndo! "the otter." The initial two thirds of the refrain's performance [:50 2:37] are sung in a simple two part polyphonic texture: the narrator sings one part (the wa! nja! mba part) while the small audience choir sings the other (the wa! tu$ ko! part) (Example 51 ). Example 51: Higgens (1981), Sung Refrain from Lo" ndo" (Otter'). [The following is a modified (re)presentation of Higgens' transcription: format and gloss are added; numbers 1 4 indicate the pulses of a metered line; wa" tu# ko" = choir, wa" nja"mba = storyteller singer] 1 2 3 4 [ wa" tu# ko : accompanying voices ] a g! a lu# ka# ny"$"# ngo... / prospect / mother river // // 3.SG.N / go / LOC We're going prospecting at the great river. [ wa" nja" mba : storyteller singer ] hi( i) hi... hi( i) hi... (bal" mb" o) During the last third of the performance [2:37 3:40) the narrator leaves off singing and takes up the spoken narrative once more while the choir continues singin g the ( wa! tu$ ko! ) refrain "in the background." Having presented the basic narrative framework of Lo! ndo! 's story and the distribution of song in it I move on to describe the most common marks of sung narrative development in Baka li! ka! n"! again, as exe mplified in the narrative of Lo! ndo! (Otter'). Figures 4 2 and 4 10 outline these five marks of sung climactic development
186 alongside the five marks of sung contextual development to serve as a reminder that these signs are not signified in isolation, b ut in concert potentially with many other distinct, though complementary signs. 184.108.40.206 Climactic development through r hetorical u nderlining : repetition Of all the marks of li! ka! n" climactic development in the story Lo! ndo or any other Baka li! ka! n" none is more ubiquitous than that of rhetorical underlining. The importance of rhetorical underlining must not be underestimated. It is one of the simplest and most universal devices for marking the important point not only of a narration but of other sorts of discourse as well. (1996:39) And of all the potential types of emphatic rhetorical devices repetition is most common. The narrator does not want you to miss the important point of the story so he employs extra words at that point. He may emplo y parallelism, paraphrase, and tautologies of various sorts to be sure you don't miss it. (emphasis added; ibid.) In the song of Lo" ndo" for instance, the most prominent form of rhetorical underlining observed in the versification of the clause we are going prospecting at the great river is repetition. Indeed, Baka song of any kind, li" ka"n%" song or otherwise, typically employ s not one or two, but a variety of tautologies. As demonstrated in previous chapters treating li" ka" n%" song musical structures and text poetics (Chapters 2 and 3, respectively ), bundle s of syntactic, lexical, phonological, rhythmic, and melodic redundanc ies are always converging in the musical and poetic formalizations of any genre of Baka song. The plainly redundant nature of Baka song thus imbues it with the
187 requisite emphatic characteristics, that is, the rhetorical underlining, so typical of li" ka" n %" climatic development. Syntactic repetition. Within the initial bundle of repeated textual constituents in Lo! ndo! s song in particular, syntax level repetition is most apparent. Two concurrent (overlapping and/or interlocking) syntactic units the pro positional clause we are going prospecting at the great river and the ideophonic phrase hi hi (to go away quickly), as shown previously in Example 51 are repeated in one varied form, or another, approximately sixty four times [from :50 3:40 sec.] bef ore culminating at the narratives peak [at 3:40 sec.]. Both the clause and the phrase are initially intoned in their entirety by the singer storyteller [from :50 54]; however, the clause and phrase are subsequently and varyingly repeated and distributed between the singer storytellers voice part and the audience choirs voice part [from :54 3:40] (again, see Example 51, and listen to accompanying audio file in Object 4 1 (Object 4 1. Audio file of li! ka! n "! story Lo! ndo! (.mp3 3.4 MB)). Rhythmic repeti tion. The texts syntactic repetition [from :50 3:40] coincides with units of rhythmic repetition, as the boundaries of these two concurrent, repeated textual units synchronize, transforming sixty four (merely) verbal sequences into sixty four poetic/lyri c lines (Example 51; see also Chapter 3, Sec tions 3.3 .4, 3.3.5 and 3. 4.3). And furthermore, within this musically metered line, still other redundancies emerge, converge and thus multiply the refrains rhetorical effect when the syllabic units of these v erses are also rhythmically segmented and iconically patterned with li! ka!n "! songs typical eight pulse poetic meter and underlying (though latent) generic polyrhythmic percussion pattern (Sec. 2.4 and Figure 2 4).
188 Melic (melodic) repetition. In addition the redundancies of Lo! ndo! 's song are neither limited to syntactic nor to rhythmic elements. The rhythmic formalizations of speaking are also intonationally formalized into melic patterns. Sixty four repetitions of complementary syntactic and rhythmic patterns are also patterned with sixty four repetitions of the single, basic melodic refrain shown in Figure 4 9. If the scope of this chapter allowed, a detailed description of the compositional devices that develop this simple musical refrain w ould furt her reveal many more musical redundancies. Such a description, however, is reserved for the narrative of Su$ a t# Ku! nda in the following chapter (Chapter 5), and even there, only modestly so. In this present section, our point is simply to give adequate evidence of the basic types of syntactic, rhythmic, and melic repetitive 36 devices that typically coalesc e in the service of "rhetorical underlining," or rather, in the service of climatic narrative development. Having done so, we move on for as was said earlier, the "zone of turbulence" enroute to a narrative's peak consists of more than rhetorical underli ning. "Heightened vividness" as Longacre refers to it is typically effected as well. 37 220.127.116.11 Climactic development through h eightened vividness" In his chapter "Plot and Peak" Longacre (1996) does not define "heightened vividness," per se but rather, deduces it: "Heightened vividness [as a mark of climatic development] may be obtained in a story by  a shift in the nominal verbal balance,  by a tense shift,  by a shift to a more specific person, or  by a shift along the narrative drama param eter" (40 43). In addition to Longacre's four parameters of "heightened vividness," I would propose a fifth: that of a shift to a more poetic
189 organization of discourse ." I base my proposal to expand Longacre s notion of "heightened vividness" on the cum ulative implications of a number of factors. A brief summary of these factors seems in order before continuing with our description of th e parameters of heightened vividness found in li! ka! n"! song. Heightened vividness through poetically organized discou rse. First, the decriptive term "heightened vividness" is vague, and presumably intentionally so. It connotes certain semiotic phenomena more than denotes them, and in doing so, already allows for the perception of a broader range of related signs (which is what naturally happens when a more interdisciplinary approach is taken to discourse analysis). Second, poetic procedures are inherent in nearly all of Longacre's higher level inventory of prototypical features of narrative peak (e.g., rhetorical under lining, heightened vividness, change of pace, and use of onomatopoiea). Third, Longacre's own exemplary analysis of the Hebraic Genesis narrative claims that poetic discourse lies at the heart of both episodic and global narrative peaks of the Genesis nar rative (2003:21,39,300). And last it is striking how many interdisciplinary studies of the relationship of music and language share the kindred terms "heightened text," "heightened speech, and "heightened voice" in their discussions of a number of express ive and poetic speech acts. For example, in their survey of (English) interdisciplinary studies investigating "the relationship of music and language," Feld and Fox, citing the studies of Lomax (1967), Hinton (1984), Basso (1985), and Feld (1990), posit t hat "music's formal redundancy and auto referentiality heighten poetic texts (emphasis added; 1994:27). In other studies, List (1963), Finnegan ( 1977, 1988), Banti and Gian n attasio
190 (2004), and Feld, Fox, Porcello, and Samuels (2004) similarly use the kin dred terms of "heightened voice," or "heightened speech" to develop the notion of "different levels of formalization of speech by means of timbric, rhythmic, and/or melodic procedures that heighten and specialize its symbolic effect (emphasis added; Banti and Gian n attasio 295). Linguistic anthropologist Giorgio Banti and ethnomusicologist Francesco Giannattasio then recast and re term these "levels of formalization of speech" as "poetically organized discourse (POD). Poetically organized discourse (POD) and, in general, poetic procedures may be regarded as a special way of formalizing speech by means of a number of constraints on how the text is organized such as meter, rhythm, morphosyntactic parallelism, assonance, and other procedures [procedures] that effect speech sounds in the voice register, in the melodic and accentual contour, and, especially, in their recurrence through time, on the basis of an equivalence between vocallic or syllabic units, stresses, and/or several recurring text or sound units, so as to frame one's speech in a cyclic time, often relying on a periodic measure unit or beat. Such constraints, that often occur together, heighten and specialize the symbolic impact of an utterance. (emphasis added; 315) The marks of "poetically organ ized discourse," then, are numerous. In the case of Baka song in general, many such marks have already been illustrated in our previous chapter on Baka poetic texts (Chapter 3). In this present chapter and section, however, these poetic procedures (and m ore) are not simply abstracted and enumerated, but framed and interpreted in a particular context, that is, that of narrative discourse, and more specifically, within the particular activities of climactic development in narrative discourse In this genre of Baka discourse, as we shall see, the "zone of turbulence" so often associated with climactic development often bears many marks of poetically
191 organized discourse; and these marks are not only perceived in the timbric, rhythmic, and melodic formalizatio ns of speech that effect heightened vividness, but may also be present in (1) the redundant procedures of rhetorical underlining, (2) the rhythmic nature of a narrative "change of pace," and (3) the semantic and phonological effects of onomatopoeia. Having expanded and reconfigured the notion of "heightened vividness" to include the parameter of "a shift to a more poetic organization of discourse," we resume our identification of poetically organized discourse in the climatic development of the particular Baka narrative of Lo! ndo! "P.O.D." through voice register, intonation, and rhythm To reiterate, the main procedures for formalizing speech sounds beyond those used in normal conversation seem to be no more than three: 1. altering voice register' ... 2. altering melodic contour... 3. segmenting utterances rhythmically ... ( Banti and Giannattasio 295). As for the particular repertoire of formalized speech sounds in the subtle domain of voice register "a ltering intentionlly and/or by convention one's voice register, is to alter its frequency range, timbre, and intensity ( emphasis added; ibid.). Shifts in these three vocal qualities alone (i.e., pitch, timbre, and intensity) can by themselves signal heightened vividness (Sec. 2.2.3) Yet, even when bundled with other shifts in poetic intonational contours and rhythmic segmentation (Sec. 18.104.22.168) "any of these procedures can work as an indexical framing device for picking out an utterance as special in some way (emphasis mine; 295 ; Bauman 1977:15 25 ) Many such poetic procedures are at play in the climactic development signaled by the performance of 'e! na li! ka! n"! The Baka, as we have already seen in Sec. 2.1 "pick
192 out a number of distinct speech acts (e.g., conversations, discussions, accounts, counsels, proverbs, parables, stories, and songs). Within the particular performance context of trad itional story ( li! ka! n "! ), the Baka explicitly distinguish speech mode from sung mode discourse. Their conceptual distinctions (i.e., what they say about it) can be correlated initially with alterations in vocal pitch range, timbre, and intensity, and eve ntually with patterned alterations in melodic contour and rhythmic segmentation During the performance of the story of Lo! ndo! for example, the singer storyteller first tells the story ( na kp li! ka!n "! ) [from 0 to :50 sec.], then intones the son g ( na t" nja!mba ) 38 [at :50 sec.], and then sings the song ( na # e # e! ) [from :50 to 2:37 sec.]. The first signs of the shift are signified within a second or two [from :50 to : 52 sec.] by an alteration in the sound ( li ) of the intoners voice ( li na wa! nja!mba ), that is, a conventional alteration of the teller turning singers vocal frequency range, timbre, and intensity. This vocal shift at and in the nja!mba (the intoning of a song) instantly indexes the possibility that something speci al as Banti and Gianna t tasio put it is in the offing. This special ( i.e., extra ordinary, more than everyday ) speech sign is quickly formed as the teller turning intoner turning singer reflexively changes the size and shape of his vocal tract in or der to intone the phrase a g a lu! ka! He changes from the default vocal register associated with a telling voice to the special register of the nja! mba voice, a register marked by a prolonged low range fundamental frequency a distinctive cluster of overtones (formants), an d increased intensity. 39 In just a few seconds more [ Fig. 4 8, at :52 sec. ] intoning quickly becomes singing as the inflected voice utters the words a g" a lu! ka! in new and protracted melic and rhythmic patterns At this point in the narratives devel opment a more poetic organization of
193 the verbal discourse predominates. This shift, then, perlocutionarily evokes qualities of heightened vividness qualities that so often index climactic development in Baka li! ka!n"! "Heightened vividness" through a shif t in tense. At the beginning of this subsection (Sec. 22.214.171.124 ), I noted that "a shift in tense," according to Longacre, is also a common parameter of heightened vividness. And of all of the four parameters of heightened vividness originally observed by L ongacre, only "a shift in tense" is exhibited in all twelve of the li! ka! n"! songs under investigation. This shift of course, is consistent with our earlier discussion of "s toryline prominence in Section 4.3 in that song text clauses rarely, if ever, employ the conventional narrative tense aspect marker reserved for th e foreground storyline, but rather use either the imperfective aspect, the distant and recent past tense markers, the general future tense marker, or verb forms with no tense aspect markers at all (e.g., Examples 13 19) 40 These are the tenses and aspects conventionally associated with supportive background material. The song of Lo! ndo! then, is no exception. The verbal construction a g" a lu! ka ..." (we are going prospecting'), especially in its prolonged and redundant song form, conspicuously contrasts with the strict chronological sequencing that typifies much of the previ ous narrative foreground. The construction does not signal a sequence of actions but indicates an action in progress, one that is ongoing and oriented to the future. Le! onard describes such verbal constructions in this way: The two auxiliary verbs g % go ing' and d % coming' play a crucial role in setting a point of orientation in Baka stories. G % is most often used, and its main function is to mark that the story is going towards a point of
194 interest. By doing this, it indicates that an action is already in progress and is taking time to fulfill (emphasis mine; 2005:36) Heightened vividness through change of pace. As stated earlier, Longacres third peculiar feature of climactic development is change of pace. In most li! ka!n "! songs change of pac e can be signified in more than one manner. In Lo! ndo! s song in particular, a change of pace is signaled through ( 1) variation in the regularity of narrative action, as indicated in the relative pattern of foreground versus background verb constructio ns, ( 2) variation in the relative sizes of grammatical constructions (e.g., clauses, paragraphs, or embedded discourses), and ( 3) variation in the relative rhythmic regularity of spoken utter a nce s versus sung utterances. Variation in the regularity of narr ative action The foreground action that precedes Lo! ndo! s song is fairly regular and indicative of the stricter sequential flow of mainline li! ka!n "! discourse (cf. Le! onard citation early in Sec. 4. 3). But Lo! ndo! s song, like all li! ka!n "! song, is not a foreground verbal construction, but rather, ba ckground. In Lo! ndo! s refrain the verbal construction a g" as previously discussed in Sec. 4. 3.2.2 is understood as meaning taking time to fulfill, that is, an action in progress. This prolonged action contrasts with the previous narrative action (i.e., in Fig. 4 8: lines 1 6a, 0 :50 sec.) which changes fairly regularly, as recognized in the foregrounded sequence Otter says... Otter sets off... Otter goes... He arrives... He says... Otter spears... he looks... Otter says... Otter goes. Variation in the relative sizes of grammatical constructions Beyond these shifts in verb constructions and regularity of action, there are variat i ons in the sizes of discourse constructions. The redundancies of poetic constructions, like songs, lengthen
195 grammatic al units; and the mere length of time it takes to perform the sixty four repetitions of Lo! ndo! 's refrain plainly "slows down" discourse development. Le! onard observes three or four other (spoken) discourse features that "slow down" a story in order to "bu ild the climax" : they include "use of non narative aspect," "variation in the order of clause constituents," "tail head linkage," and "repetition," (2003:11, 13, 15 ). The principle of repetition, in fact, is at play in most of these "pace changing" device s, and illustrates the multifunctionality of the syntactic, rhythmic, and melodic repetition treated earlier in our analysis of Lo! ndo! 's song, and other li! ka! n"! songs. Variation in the relative regularity of spoken utterances to sung utterances The last kind of pace changing device to highlight is that which is observed in the variation that takes place between the relative rh ythmic regularity of spoken utterances versus sung utterances. In sung discourse the "ordinary" rhythms of everyday speech (or even those of oratory) are subjected to "extra ordinary" rhythms The most striking quality of these extra ordinary rhythms is their periodicity a fairly strict periodicity at that, in the case of li! ka! n"! song. T he utterances of sung li! ka!n"! discourse are always segmented and rhythmically patterned with one or two particular traditional polyrhythmic percussion patterns : kole na! 'e! na li! ka! n"! (as shown in Sec. 2.4 ). I n turn, the boundaries of kole na! 'e! na li! ka!n"! are patterned with a fairly strict pulse which, in turn, is metered into groups of 8, or 16 pulses (Se c. 2.4 ). The combined and complementary periodicities of pulse, meter, and percussion pattern may be further affected by var iations in tempo; and to the degree that the segmented utterances of narrative discourse are (re)patterned with the periodic patterns of kole na!
196 'e! na li! ka!n"! a "change of pace" is perceived in the development of a li! ka! n"! and a narrative climax may be anticipated. Climactic development through a shift in incidence of ideophones. The last of four features exhibited in li! ka! n"! song that commonly contributes to the "zone of turbulence" that so often marks climactic narrative development is a shift in incidence of ideophones. Longacre initially chooses to describes it as "a shift in incidence of onomatopoeia ," but later implies that the concept of onomatopoeia' is more appropriately termed ideophone when speaking of many African languages (48). F or the Baka language in particular, the different formal features of ideophones... [semantically] represent two points on a scale leading from onomatopes (i.e., ideophones representing pure audible sensations) to non onomatopes (i.e., ideophones represe nting all other kinds of sensations). (Kilian Hatz, 2001:163) [Thus,]... the various semantic concepts [of Baka ideophones] don't illustrate only audible events but... can generally point out every detail of every lexeme that [a Baka speaker] wants to use ." (emphasis added; ibid., 162) Le onard observes that "[spoken] ideophones often appear at the climax of a story" (5.3.4). 41 And from the data at hand I observe that sung ideophones are occasionally exhibited in li! ka!n"! songs as well. In fact, as has already been illustrated in Sec. 3. 2, ideophones are common in the general Baka song repertoire. Of the twelve li! ka!n"! songs sampled for this present chapter, two employ ideophones ( Su$ a t# Ku! nda (2 nd song) and Lo! ndo! 's song). Both participate in climactic development.
197 The ideo phonic phrase in Lo! ndo! 's refrain, as shown earlier in Example 24, is hi hi (to go away quickly'). Minimally, it grammatically functions to modify the verbal phrase a g" a lu! ka ..." (we are going prospecting'). But Kilian Hatz claims that the pr agmatics of Baka ideophones exceed mere grammatical categories. Ideophones... are direct speech, a verbalized imitation of extra linguistic events or situations. By the use of ideophones the speaker simulates or raises the illusion that the verbalized e vent happens simultaneously the moment of its production/pronunciation (ibid, 155) [It] is a vivid re presentation or re creation of an event in sound. (emphasis added; Kilian Hatz, 155, citing Fortune 1962:6) [Its] function is to dramatize a narration.. (157) A shift in a narrative's incidence of ideophones," then, not only has the potential to indexically signify narrative climactic development, but can also iconically and symbolically signify numerous affective qualities of "heightened vividness" wh ich, as presented earlier, are also so often associated with climatic development. This bi functionality of ideophones is indicative of a number of verbal discourse devices and conveniently segues to the generalization that few narrative discourse devices in a li! ka!n"! song act as isolated, unifunctional signs. (This generalization will be made more evident in the following chapter, Chapter 5). 4.4 Summary In summary, li! ka! n"! song functions as a discourse feature of li! ka! n"! narrative in at least two signifying domains of narrative discourse: first as an agent of narrative discourse cohesion and second as an agent of narrative discourse development. In the immediately preceding sub section (Sec. 4.3.2), I described li! ka! n"! song's effect on
198 climact ic development; in the prior sub section (Sec. 4.3.1), I discussed its effects on contextual development; and in Section 4.2, I identified song's unifying effect on the entire narrative performance. Figure 4 10 recaps the numerous specific ways in which l i! ka!n"! song can potentially function as a feature of li! ka! n"! narrative discourse. I summar ize the outline as follows L i! ka! n"! song effects discourse cohesion generatively performatively and (inter ) textually The fundamental unity of story and song is inherent in that stories engender songs and that both are generated from a common experience. This cohesion is r einforced performatively throug h the embodiment of the discourse in the speaking and singing voice of the storyteller singer who in tur n is joined by the singing voices of the storytelling 's particpants. The intertextual cohesion of the ir spoken and sung texts is further m anifested both gramm a tically and semantically Grammatical devices may include lexical repetition lexical substitut ion (e.g., anaphora ), and/or a shared point of view Semantic devices often include synonymy and meronymy Narrative texts do not simply cohere, however, but also develop. Li! ka! n"! song effects narrative development both contextually and climactically Song lyrics commonly contextualize the main storyline by providing supportive participant information, setting information, explanatory information, collateral information, and evaluative information. Any of these information types can be communicated directly through propositions, but more often they are expressed indirectly through devices like ideophones, obligative mood markers, imperative mood markers, exclamations, interjections, intensifiers, rhetorical questions and vocatives. Finally, l i!ka! n"! song may develop a narrative's climax through either special verb forms,
199 rhetorical underlining, heightened vividness, a change of pace, or a shift in the incidence of ideophones or a combination thereof S igns of rhetorical underlining include s yntactic, rhythmic, and melodic repetition. S igns of heightened vividness may include a shift in tense or a shift to a more poetic organization of discourse The potential devices of poetic organization are particularly encoded in special formalizations of verbal synta x, phonolog y, semantic s rhythm, intonation, and timbre. The changes of pace that so commonly mark climactic development are typically signaled in the s hifts in the kind s and size s of a song's verbal constructions or the shift s in the regu larity of action suggested by a song's verb forms. Finally, li! ka! n"! song texts also contribute to the "zone of turbulence" that so often marks climactic development by employing a shift in the incidence of ideophone s. Li! ka! n"! song, then, functions as a discourse feature of Baka narrative discourse in multiple, complementary ways, ways that both unify and develop narrative discourse performance. Yet even as I close this section of my analysis I am still becoming aware of other related functions. I sus pect, for instance, that the unifying function of song does not simply unite sung and spoken "texts," but more specifically unites particular narrative episodes, or sequences of actions These potential functions, and others, however, must be left for a f uture study. Still, in Chapter 5, I will gradually extend my description of li! ka!n"! song's multifunctionality beyond the domain of discourse and into the more pragmatic domains of social and spiritual experience.
200 Figure 4 1. Axis graph of the potential performance distribution of speech, song, and music in African oral monologue narrative discourse L I # KA # N % # S ONG AS A D ISCOURSE F EATURE OF L I # KA # N % # N ARRATIV E I. I N N ARRATIVE D ISCOURSE C OHESION ( Sec.2) 1. G ENERATIVE C OHESION ( Sec.2.1) 2. P ERFORMANCE C OHESION ( Sec.2.2) 3. T EXTUAL C OHESION ( Sec.2.3) Grammatically ( Sec.2.3.1) & Semantically ( Sec.2.3.2) II. I N N ARRATIVE D ISCOURSE D EVELOPMENT (Sec.3) S UNG T EXTS # 1 C ONTEXTUAL D EVELOPMENT (Sec.3.1) th rough a. Participant Information (Sec.3.1.1) b. Setting Information (Sec.3.1.2) c. Explanatory Information (Sec.3.1.3) d. Collateral Information (Sec.3.1.4) e. Evaluative Information (Sec.3.1.5) 2 C LIMACTIC D EVELOPMENT (Sec.3.2) through a. Particular Verb Forms (Sec.3.2.1) b. Rhetorical Underlining (Sec.3.2.2) c. Heightened Vividness (Sec.3.2.3) d. Change of Pace (Sec.3.2.4) e. Incidence of Ideophones (Sec.3.2.5) S POKEN T EXTS $ Figure 4 2. Outline of discourse analysis of li! ka! n"! song in li!ka! n"! narrative
201 3. T EXTUAL C OHESION Grammatically Semantically L EXICAL R EPETITION L EXICAL S UBSTITUTION P OINT OF V IEW S YNONYMY M ERONYMY Figure 4 3. Grammatical devices of textual cohesion 3. T EXTUAL C OHESION Grammatically Semantically L EXICAL R EPETITION L EXICAL S UBSTITUTION P OINT OF V IEW S YNONYMY M ERONYMY Figure 4 4. Semantic devices of textual cohesion L I # KA # N % # S ONG AS A D ISCOURSE F EATURE OF L I # KA # N % # N ARRATIV E II. I N N ARRATIVE D ISCOURSE D EVELOPMENT 1 C ONTEXTUAL D EVELOPMENT 2 C LIMACTIC D EVELOPMENT Figure 4 5. Types of sung narrative development II. I N N ARRATIVE D ISCOURSE D EVELOPMENT 1 1 C O NTEXTUAL D EVELOPMENT through a. Participant Information b. Setting Information c. Explanatory Information d. Collateral Information e. Evaluative Information Figure 4 6. Subtypes of sung contextual development
202 Figure 4 7. Transcription and translati on of Lo ndo! (Otter') (Higgens 1981:35 42) (also, Object 4 1. Audio file of li! ka! n"! story Lo! ndo! (.mp3 3.4 MB))
203 Figure 4 8. Framework of the narrative performance development of Lo! ndo (Otter') (also, see Object 4 1. Audio file of li! ka! n"! story Lo! ndo! (.mp3 3.4 MB))
204 Figure 4 9. Composite melody of Lo! ndo! refrain (Object 4 1. Audio file of li! ka! n"! story Lo! ndo! (.mp3 3.4 MB)). (Two voice parts are performed: that of the storyteller singer (Voice 1) and of the storytelling partici pants (Voice 2). The basic melodic figures for Voice 1 and 2 are framed in dashed rectangles; improvised melodic substitutes are then aligned with their respective slots below each of the two basic figures. Pulse % 94 MM. The refrain is repeated sixty four times during the performance.) 42
205 L I # KA # N % # S ONG AS A D ISCOURSE F EATURE OF L I # KA # N % # N ARRATIV E I. F OR N ARRATIVE D ISCOURSE C OHESION 1. G ENERATIVELY 2. P ERFORMATIVELY 3. T EXTUALLY Grammatically Semantically L EXICAL R EPETITION L EXICAL S UBSTITUTIO N P OINT OF V IEW S YNONYMY M ERONYMY II. F OR N ARRATIVE D ISCOURSE D EVELOPMENT C ONTEXTUAL D EVELOPMENT C LIMACTIC D EVELOPMENT 1. Participant Information 1. Verb Forms ( ) 2. Setting Information 2. Rhetorical Underlining 3. Explanatory Informa tion R EPETITION I DEOPHONES S YNTACTIC 4. Collateral Information R HYTHMIC Q UOTED Q UESTIONS /D ENIALS M ELODIC 5. Evaluative Information 3. Heightened Vividness (Explicit Evaluation) V ERB FORMS (+) P ROPOSITIONS P OETICALLY O RGANIZED D ISCOURSE (Implicit Grammatical Evaluation) V OICE REGISTER O BLIGATIVE M OOD P ITCH I MPERATIVE MOOD T IMBRE E XCLAMATIONS I NTENSITY I NTERJECTIONS I NTONATION I DEO PHONES R HYTHM I NTENSIFIERS R HETORICAL Q UESTIONS V OCATIVES (proper names ) o ther and self referential directive and expressive inciting action validating experience 4. Change of Pace V ARIATION in the regularity of narrative action (verb forms) V ARIATION in the relative sizes of grammatical constructions V ARIATION in the relative regularity of spoken utterances to su ng utterances S POKEN T EXTS # 5. Incid ence of Ideophones S UNG T EXTS $ Figure 4 10. Summary outline of the functions of l i! ka! n"! song in li! ka!n"! narrative
206 1 The term chantefa ble was first coined to describe a particular French medieval narrative, part recited prose and part sung verse (see "Chante fable." In Grove Music Online Oxford Music Online http://www.oxfordmusiconline.com.lp.hscl.ufl.edu/subscriber/article/grove/musi c/05 418 (accessed February 1, 2011). Finnegan refers to chantefable as "prosimetric" form in which prose alternates with verse (1977:109, 119). 2 To become familiar with stories told by Baka youths, consult Brisson's Contes des Pygmes baka du Sud Camerou n Vol. 1 4. Douala: Collge Libermann, 1981. 200 p. 3 This quote from Stephen Levinsohn is taken from the introductory chapter of his lecture notes for a course in the "Analysis of Narrative Texts" given Februrary 3 14, 2003 at Horsleys's Green, UK. A s imilar course was presented in Yaounde! Cameroon in 2004. 4 Despite its common association with postmodern thinkers (e.g., Derrida, Foucault, Kristeva, Jameson, etc.), discourse analysis is not the exclusive methodological property of deconstructive or cr itical theory. As mentioned above, all discourse analysis, regardless of its application, asks why "this or that" discourse is executed in "this or that" way. However, answers to such questions need not soley be applied to deconstructing or "subverting" the hidden motivations, aims, or agendas of social, political, or philosophical discourse (cf. Beard and Gloag 2005:38). In translation studies, for example, discourse analysis (or "text linguisitics") is often appropriated for use in addressing the pragm atic problems of "relevency" in translation. In the task of translation, for example, an informed choice between two similar linguistic forms (whether lexical, syntactic, semantic, or poetic) is greatly aided by a more explicit understanding of which of t wo or more optional forms are most relevent to the purposes of the discourse at hand. Likewise, in the case of Baka narrative discourse, it is hoped that discourse analysis will help determine the relevency of choosing the peculiar verbal forms of sung di scourse over those of similar, yet distinct, spoken discourse when performing a traditional story. 5 Temporal cohesion that is, the mere fact that story and song occur together "in time" during discrete social events, is implicit in Sections 4. 1, 4. 1.1, a nd 4. 1.2. It is also presented more explicitly in Section 4. 2.1. Otherwise, there seems little advantage in treating temporal cohesion more fully in this analsysis. 6 Having made this claim, Tombombo would not, however, deny that certain elements of a s ong (especially the lyrics) would be forgetton, modified, or become archaic. Often, when he and other Baka speakers would assist me in transcribing song texts, the meaning of various words or phrases could not be identified, and would often be explained as archaic, even by elders.
207 7 Yves Le! onard has told me that some Baka speakers in the camp of Ndjibot claim that a li! ka!n"! 's song may occasionally be initiated by someone other than the story 's teller (2010, personal conversation). 8 My search for cohesive devices is built on the work of Halliday and Hassan (1976, cited in Renkema 2004:103 108) who distinguish five types of cohesion: substitution, ellipsis, reference, conjunction, lexical cohesion (e.g., repetition, synonymy, hyponymy, meronymy, antonymy). 9 Possible exceptions may include certain ideophones, grammatical reduplication, or onomatopoiea. 10 Le! onard explains this ideophonic gesture this way: "Les bourdons s'agitent en faisant des gestes de provocation pour se moquer de Komba. Ils bougent leurs yeux et leurs ttes d'un ct l'autre" (2003:80). 11 Earlier in the narrative, the narrator does no t explicitly indicate that it is the voice of Otter performing the first entrance of sung discourse. 12 This example is not taken from the twelve representative songs in focus in this chapter because it is not clear that any of the twelve indisputably prese nts the point of view of the narrator in the song (though it is very likely that the song in Higgen's D e! ngbe! (verse 10) also functions as the grammatical voice/point of view of the narrator. Example 8, then, is from the story of Gbanga a Ti$ ta on pages 152 153 in Volume I of Robert Brisson's Mythologie des Pygme$ es Baka 1999. 13 The accuracy of Briss on's translation and transcription is in question here. I have bracketed those words and phrases that I have reinterpreted. oka and e are not, then, properly part of the narrative, but the narrator's instructions to the gathered community to "begin singing!" and "sing strong!" respectively. Therefore, it is not likely that these phrases are repeated in the refrain. 14 "Na subu! nga! is also uttered in verses 31 and 32. 15 Also uttered in verses 13, 27, 28, and 36. 16 Le! onard outlines similar, as wel l as additional, coorelations of verb forms with narrative information types, though only non lyric/poetic narrative clauses are cited (2005:14 17). 17 Dooley and Levinsohn (and Longacre) call "collateral inf o rmation" "irrealis" (cf. 2001:43; 1996). 18 Doole y and Levinsohn cite two additional "Grimes like" non event types'': participant orientation and performative information (ibid ).
208 19 According to Dooley and Levinsohn, P ERFORMATIVE INFORMATION (Grimes, Chapter 5) deals with aspects of the situation under which the text is produced, especially the speaker hearer axis. This comes out when the speaker speaks in first person to the hearer in second person. Also included in this category are morals, conclusions, and applications to the audience, which also ov erlap with evaluation" (2000:43). It is to be expected, then, that performative information is typically absent from "traditional" song texts, g iven that sung lyrics, among the Baka, are pre composed and remain relatively fixed. 20 According to local kno wledge, Su$ a is Ku! nda 's notorious uncle. 21 It can also be said that this information is not only particpant orientation background information but explanatory information as well. 22 In addition to the following two examples, the story of Lo! ndo! could also be cited as an example of a lyric with setting/orientation information. 23 It is common (local) knowledge that this stereotypical scene, that is, the forest...with noisy people in it, is likely to annoy Komba 24 It might also be said that this informa tion is both setting and background material. 25 Also from Kilian Hatz, "Ideophones in Baka have the function of communicating reality. Their proper use is "Proof that one knows the world in which the story is taking place" (Kpokpo Maurice, personal commun ication)... ( 2004:37). 26 Reduplication, in Baka, indicates that the meaning of the lexeme is somehow augmented, either in duration, iteration, speed, or quantity (see Kilian Hatz 39). 27 Boursier glosses ng"! as a verbal aspect (asp.'), but linguist Kilian Hatz more accurately identifies ng"! as an "agent oriented obligative modality marker" (OBL') (see p. 30). 28 See Le! onard regarding the grammatico pragmatic function of the lexeme taa! (2003). 29 See taa! in Mbi! lo! also. 30 Again, Boursier glosses ng"! as a verbal aspect (asp.'), but linguist Kilian Hatz more accurately identifies ng"! as an "agent oriented obligative modality marker" (OBL') ( 2004:30). 31 For further clarification, Portner goes on to write that the meaning of vocatives be formulated as expressive content in the sense discussed recently by Potts (2003a, 2003b, 2003c, 2004)
209 32 Cf. Austin, J.L. How to Do Things with Words. Harvard Univeristy Press, Cambridge, 1962. 33 While the refrain of Ku! njenje does make Ku! njenje 's disg ust/disapproval explicit, the improvised lyric of the storyteller singer also does, as indicated in the excerpted phrases ma a! kpo (I refuse/disaprove/hate') and m"ng"m"!ng"! (ostentatious strutting'). 34 I too have found in a previous musical analysis t hat the identification of macro level climatic structures seems to be an analytical prerequisite to an apt description of many lesser level structures and developmental strategies. In my comparative analysis of Monteverdi's Italian opera Orfeo and Yang's Chinese opera The Prostitute's Tears for example, "the larger scale formal designs of both Orfeo and Tears greatly inform smaller scale forms: larger scale narrative development gives rise to larger scale musical development; "lesser scale" musical forms and devices serve the greater, though the subtlest devices (e.g., ornaments, register, voice timbre, and silence) seem most critical to the realization of a narratives heightened dramatic affect. No musical "sign," then, no matter how nuanced, should go u nidentified, if one would understand why any sung narrative "sounds the way it does" (Fitzgerald 2009, unpublished). 35 By "minus features" Longacre means those verb forms that are not typically used in the main storyline material. 36 Therefore, repetition a s a quintessential iconic and indexical sign of rhetorical underlining is plainly signaled as the narrative's climax builds. 37 In his Storyline Concerns and Word Order Typology in East and West Africa Longacre uses the term "immediacy" instead of "heighte ned vividness" (1990:9). 38 It may also be expressed as na k p"$"! nja! mba or na tongo! a 'e! (begins the song') (cf. Brisson 2002:465). Further, I surmise that the term nja! mba (intonation) potentially signifies any combination of four aspects of that speech act: for example, (1) the solo beginning of a song, (2) the "lead" song part, (3) the essential and composite elements of the song (text and tune) from which other singers are enabled to "pick up" or recall their complementary "part" in the song, and (4) the essential sonic features associated with the typ ical vocal register of Baka singing, that is its frequency range, timbre, and intensity. 39 A fuller description of these voice register qualities, that is frequency range, formant, and intensity, is beyond the scope of this chapter, but intended for futu re studies. Vocal formant qualities is of particular interest. Pages 332 336 of Feld, Fox, Porcello, and Samuel's "Vocal Anthropolgy: From the Music of Language to the Language of Song" (2004) seem to indicate a fruitful approach.
210 40 Le! onard outlines sim ilar, as well as additional, coorelations of verb forms with narrative information types, though only non lyric/poetic narrative clauses are cited (cf. Le! onard 2005:14 17). 41 Le! onard focuses on the texts of spoken di s course, not the texts of sung discour se. 42 This composite version of the refrain of Lo! ndo! is an etic transcription, and not very detailed at that. It is merely intended to bring into reliecf some of the more obvious repetitive melodic figures that overlap with, intermingle with and enha nce the other redundant non melic, non rhythmic elements at play. A fuller emic analysis of the Baka melodic system has already begun and is intended for publication at a later date.
211 C HAPTER 5 WHY KU! NDA SINGS: SONG IN THE STORY OF SU$ A! T& KU! NDA 5.1 One Storytelling Event: Fourteen Stories and Sixteen Songs From my original data sample of four hundred recorded songs, more than forty are li! ka! n"! songs (Sec. 1.6). Sixteen of those forty li! ka! n"! songs were recorded during a single storytelling event one evening in August 1997. What immediately follows is a brief description of that event : the setting, the people, and a few generalities conce rning that evening 's fourteen stories and sixteen songs. This general description will then segue into and provide the context for a more detailed description of one particular story, Su$ a t# Ku! nda ("Leopard and Turtle"), and its embedded songs. My d escription of this single li! ka! n"! song will serve as the preamble to my conclusions regarding (1) how a li! ka! n"! song functions in a traditional Baka story and in Baka experience in general, and (2) why Ku! nda (Turtle) in the story of Su$ a t# Ku! nda s ings, rather than speaks, when expressing himself. 5.1.1 Ethnographic Synopsis of the Storytelling Event One clement evening (August 14, 1997) in the short dry season of the southeast rainforest of Cameroon in the Baka encampment of Ndjibot, fourteen trad itional Baka stories ( li! ka! n"! ) were performed over a period of approximately one and one half hours. The principal organizer of the storytelling event, as well as its principal storyteller, was a Baka man named Nguna an elder ( kobo ) of about 60 years of age. A few days pri or to the event, Nguna had asked me for some money. ( I had been living in Ndjibot about three months at this point.) I gave him what he asked for, and in return I asked to hear
212 him tell a few traditional Baka stories (as weeks prior I had heard that he w as a very good storyteller (" wa kp" $ kp"! e$ ko kp#$k# ). 1 On that evening Nguna assembled a group of fifteen to th i rty people for a story telling under a 4 x 5 m eter rafia shelter at the entrance of my home in Ndjibot. The shelter was illuminated by two or three kerosene lamps. Th e gathered group was fairly evenly mixed in gender and age, from nursing infants, children, and adolescents, to young adults, older adults, and elders of approximately seventy years of age. Almost all were members of one of three or four clans ( ye! e ). A ll were Baka speakers; no Bantu neighbors from nearby villages were present. The number of participants fluctuated slightly during the evening, as a few came and went at will. Most sat around in a small circular arrangment, some on wooden benches and som e on the ground. A few others, like Nguna stood. Nguna was posit i oned well inside the northern rim of the circle. On the opposite rim of the circle were two adolescent male drummers, each seated upon and straddling an approximately 40 cm. by 14 cm., sin gle headed, cylindrical wooden drum ( ndu! mu! ) that laid horizontally on the ground. With the drumheads straddled between their legs, the drummers struck the deer skinned heads with their hands. Nguna narrated and led the singing in thirteen of the fourtee n stories. A younger man, Konji recounted one other story The sonic ambience for the entire evening's event was one of continual low level chatter, light banter, intermittent giggles, and occasional harmless reprimands ( as certain sporadic behavior by t he children during the performance seemed ill regarded ). A few utterances from within this continual verbal stream converged with
213 the formal narrative, as storyteller singer and participating audience choir intermittently exchanged fleeting comments and rhetorical questions. These exchanges also included interjections of surprise and disapproval, playful mockery, and occasional questions regarding the story's characters or the story's action. 5.1.2 General Narrative Features The average duration of a s ingle story 's narration, including any pre and post narrative discussion, was six and one half minutes, though the story proper averaged only four minutes. The shortest story lasted two and one half minutes; the longest nearly nine. Each story segued into another. The settings, characters, and plots of all fourteen stories were typical of the Baka li! ka!n"! repertory, in that (1) the setting is always in the forest at a hunting camp, a forest camp, a river's edge, or some spot along any number of fore st paths between well known areas for hunting, fishing, gathering food, or collecting honey; (2) the characters are most commonly drawn from a familiar cast of animals, people, and spirits, such as Ku! nda (Turtle), Su$ a (Leopard), Se! ko! (Chimpanzee), Li! mbe! nu! (an ancient heroine), Komba (God), J#l"! (God's first daughter), and m# (spirits); other less predominant, though no less familiar characters are Mo! njombe (the great leaping deer), Bi$ li! (Mouse), Kpa$ ngbala (Lizard), Kpi! nya! (Anteater), mo$ k"$s# (an ordinary man), and wo$ s#! (an ordinary wom a n); (3) plots ordinarily revolve around well known, ongoing and inherent conflicts between certain stereotypical characters (like the dangerous Su$ a the clever Ku! nda or the attractive J#l"! ). M any other plots however, are built around the familiar challenges, hazards and dangers of life lived in the forest, ranging from lesser problems posed by such things as prickly thorns and stinging bees, to more
214 earnest problems experienced in childbirth, child rea ring, marriage, and death. Still other plots are etylogical, told in order to explain, or at least imply, the causes, nature, or origins of certain animals, peoples, or spirits. 2 5.1.3 General Li! ka! n"! Song Features 126.96.36.199 Performance distribution of songs in stories As for the general patterns of the occurrence of song during the evening's tellings, t welve of the fourteen stories included at least one song; two stories included two songs. 3 Each song occurred only once (as opposed to a few stories in the greater repertory that intermittently reiterate a song throughout). Songs ranged from forty seconds to five minutes in length, the average song lasting about two and one half minutes. The re lative duration of a li! ka!n"! to its song as well as the chronological distribution of the song within a story both bear significantly on a story's rhythm and pace, and by extension, upon its dramatic development and poetic effect. It is therefore noteworthy that on the ave rage, well over half the duration (60%) of a story involves singing (regardless of song length); a few stories were nearly completely performed with s o ng. 4 All songs occurred within the main body of a story, with the main body normally preceeded, at the very least, by a spoken opening formula and some type of introductory section ( S ome stories however, included an episode or two in the main body of the narrative before introducing the song). 5
215 188.8.131.52 Discourse functions of songs in stories As for the narrative discourse function of all the story's songs, little can be said about their general character that has not already been outlined in Chapter 4. That is, every song in every story that evening, in a variety of ways, contributed to both the cohesi on and development of its respective narrative. Each song was story specific; each one connected and developed intra narrative material; and none functioned inter narratively, that is, as if it were a mere interlude between contiguous narrative performanc es In regards to narrative cohesion (Sec. 4.2), it is worth reiterating that in all fourteen of the evening's li! ka! n"! the voice of the storyteller singer embodied the cohesive locus of both verbal discourse and narrative performance: his voice told and sang the narrative, and his voice led and enjoined the choir. Occasionally, songs were performed during the telling of a short series of contingent narrative actions, or during the transition from one narrative episode to another. Such an occurrence of so ng seemed to reinforce the connection between these contiguous discourse units. 6 As generally regards the narrative development of all fourteen stories, all song texts, in various ways and in varying degrees, provided supportive, contextual information for each story. Song texts often contained information about the narrative's characters or settings, commonly provided explanatory or collateral information, and always communicated evaluative information wherein and whereby the thoughts, emotions, and attit udes of narrative characters and performers were
216 explicitly or implicitly expressed. In addition to contextual development, songs often indexically signaled ("foreshadowed") climactic development (Sec. 4.3.2). 184.108.40.206 Stylistic features of story's songs All fourteen songs were performed by an adult male soloist and mixed vocal choir composed of all in attendence male and female, young and old. D rummi n g (on two drums) and hand clapping accompanied nearly all the songs ( Sec. 2.4 ). It is presum ed that dru mming and clapping would have accompanied all the songs had the drums and drummers been in place earlier. The choral performance textures of all fourteen li ka!n"! song performances generally reflect the typified descriptions presented in earlier chapters (Chapters 2 and 4). To briefly reiterate and elaborate: The Baka refer to two fundamental voice parts in their performance of li! ka! n"! song: the voice par t of wa! nja! mba and the voice part of wa tu$ ko! T he storyteller is primarily responsible for singing the wa! nja! mba part ; the audience choir sings the w a tu$ ko! parts. Often, w a tu$ ko! is subdivided into two or three other constituent parts. 7 The term wa! nja! mba also, and more fundamentally refers to "one who" sings the opening nja! mba part N ja! mba denotes the action whereby t he song leader prompts the audience choir to begin singing their respective wa! tu$ ko! voice parts. The prompt is signaled when the storyteller singer intones an aggregate song line from segments of b oth the wa! nja! mba and wa! tu$ ko! voice parts. 8 Once the nja! mba is repeated once or twice members of the particpating audience choir immediately take up the ir respective wa! tu$ k o part (s) 9 Soon after, the storyteller singer in this case, Nguna begins to gradually leave off t hose figures and phrases of the composite song line that are strictly proper to the wa! tu$ ko! part in favor of those figures and phrases
2 17 normally associated with the wa! nja! mba part. The storyteller singer then tends to only sing the wa! nja! mba part, though intermittently may perform wa! tu$ ko! phrases, freely varying either part, though within the unspoken bounds of Baka improvisation. Once all voice parts have entered, from that point forward their composite performance relationship may generally be described as an overlapping responsorial choral style. Occasionally, voice parts in some songs, or song sections, may overlap and interlock to the extent that the choral texture can no longer be considered homophonic, but polyphonic. Thus, the relative choral textures and contours of the li! ka!n"! song voice parts performed during the evening 's fourteen story tellings ranged from irregularly alternating unison and/or parallel chordal textures, to modestly polyphonic textures. Musically rhythmic features. All but one of the sixteen li! ka! n"! songs performed that evening employed the accompanimental rhythm most often associated with li! ka! n"! song that is, kole na 'e! na li! ka! n"! The audio clip for Object 5 1 presents one song s performance of the predominant rhythm pattern ; Object 5 2 documents the only song featuring the single alternative rhythm (see Object 5 2 audio file of "rarer" li! ka! n"! rhythm (.mp3 705 KB) and Object 5 1 audio file of "typical" li! ka! n"! rhythm (.mp3 1.1 MB)). Both rhythms, especially the most common one, have already been described in more detail in Section 2. 4. Instrumental accompaniment did not begin promptly with the first li! ka! n"! telling. However, by the time the second li! ka! n"! was being recounted, both drummers had arrived with drums. From that ti me on, all songs were accompanied by the drumming duo and audience hand clapping. Pulse tempi ranged from 118 MM to 148 MM. Often
218 enough, the tempo of a song would accelerate as the song progressed. Frequently a song s tempo would accelerate once the ch oir had refamiliarized themselves with their respective voice parts. Other times, the drums and singers accelerated the tempo at the direct command of the storyteller to "Get going!" (" oka !"), or "[Sing] stronger!" (" t # kp#$k#$ !"). At such times, dynamics often increased as well. As already described in Chapter 2, the percussion ensemble most commonly subdivides the pulse into some form of triplet figure, though it was not uncommon in th e songs performed that evening for the pulse of melodic rhythms to be subdivided into duple figures as well. The entire rhythm's traditional cycle of eight pulses was executed either once or twice in the course of a poetic line, depending on whether or no t a song line 's meter was constituted by a sequence of eight pulses, or, as was more common, a sequence of sixteen pulses. Poetic features of song texts The most characteristic poetic features of the evening's sixteen song texts can be adequately general ized from the texts of just four select songs. It does not suit my present purposes to account for the poetics of all sixteen song texts. For a more systemic discussion of songtext poetics, the reader is directed to Chapter 3. The texts of all sixteen l i ka! n"! songs are set to recurring poetic lines of either eight or sixteen pulses. Settings are primarily syllabic. Q uasi neumatic figures are common enough though most are performed as vocables in word final, phrase final, or line final positions. In addition to lexical units (words), a lyric almost always employs vocables (non lexical phonemes). Average song lines carry a text load ranging from 5 to 16
219 words per line, and 2 to 11 vocables per line A line 's syllabic count ranged from 20 to 22 syll ables. Beyond the phonological level of syllable and pulse, the nature of the relationship of text and line is most obviously expressed syntactically. In particular, what is most conspicuous is that in most of the sixteen l i! ka! n"! songs, as in any Baka song (regardless of genre), verse boundaries typically coincide with line boundaries. Enjambment is occasionally employed, but co terminal boundaries of text and tune predominate. Whether verse boundaries coincided with line boundaries or are enjambed, there tends to be but one basic recurring verse in any song conventionally represented as Verse "A" From the songs sampled, no verse was found in contrast with another textually or musically; that is, there was no Verse B or C (cf. Sec. 3.4.3) A basic li! ka!n"! song verse can, however, undergo variation as is often represented with the convention s Verse "A 1 or A 2 or A 3 ". Verse (or verse segment ) variations were most often due to some form of textual or musical expansion, diminution, or substitution V arious paradigmatic ( textual ) substitutions were particularly prevalent. The syntactic development of the se li! ka! n"! songs may be characterized as iterative given there are no contrasti ve verse types. Even so, strict repetition is avoided and synta ctic development is still achieved through the more subtle complementary devices of verse variation and derivation as implemented through various improvisatory musical and textual expansions, diminutions, and substitutions. The aforementioned generalities are culled from a brief survey of the fourteen stories and sixteen songs of a single performance event. Many more generalizations could be induced from this event, its stories, and its stories songs. However, my
220 purpose in this section is not to be exha ustive, but rather, to prepare an adequate context for a more detailed description of one particular li! ka!n "! performance that evening the li! ka!n "! of Su$a! t# Ku! nda (Leopard and Turtle) In the following section (Sec. 5. 2.) I will detail the par ticular song features and their narrative functions that unite, differentiate, and develop Su$a! t# Ku! nda 5.2 Songs Discourse Functions in the Narrative Performance of Su$ a# t" Ku# nda The li! ka!n "! Su$a! te Ku! nda ( Leopard and T urtle ) was the t enth traditional story told that evening. The elder male, Nguna was the storyteller songleader. The setting, characters, plot, and dialogue are archetypal. The audio file in Object 5 3 documents the performance (Object 5 3. Audio file of li! ka!n "! stor y Su$a! t# Ku! nda (.mp3 4.4 MB), and a text transcription is presented in Appendix A. A synopsis of the story is as follows: Su#a" (Leopard) feigns his death in the hope that all the other animals of the rainforest will gather for his funeral, at which time Su#a" plans to suddenly arise, spring on them, and kill them. Meanwhile, from another part of the forest, Ku" nda (Turtle, Su#a" s nephew) hears the rumor, but is naturally skeptical, given Su#a" s dubious and dangerous reputation. Ku" nda then sets o ut for the funeral with his own scheme to outwit Su#a" and thwart his likely plans for treachery. Along the way, Ku" nda gathers every manner of itchy, prickly plant and insect in the forest. Ku" nda arrives at the funeral where the other animals are gathe red around Su#a" s seemingly lifeless body. Then, by stealth, Ku" nda takes a notori o usly poisonous caterpillar from his pouch and slips it next to Su#a" s inert body right on Su#a" s testicles! Su#a" can pretend no longer; and his body, ever so slightly begins to twitch. Ku" nda his suspicions confirmed, quickly organizes a few other animals (behind the scenes) to fell a nearby tree in the hope that it might fall on the pretender Su#a" and crush him before he rouses in
221 fury to carry out his plan to att ack and kill them all. But just as the tree begins to fall, Su# a jumps up in pain, and in a craze, flees into the forest and the schemes of both rivals, once again, are foiled. The performance of Su$ a t# Ku! nda is a multi layered progression of inter woven signs: words, rhythms, tunes, timbres and gestures The diagram in Figure 5 1 progressively represents the simultaneous and contingent interrelationship of a number the narrative 's more salient layers of signifying sonic phenomena Special attenti on is given to the interrelationship of the dynamic structures of the song s' texts and tunes, and how they, in turn, relate to the entire story's development. An amplified and expanded reading of and commentary on the performance represented in the diagr am in Figure 5 1 follows. Reference to performance chronology is parenthetically inserted when indicated. The entire performance of "Su$ a te Ku! nda" lasts a little more than four and one half minutes (:00 4:50) slightly longer than the average duration of the other thirteen li! ka!n"! performed that eve n ing. It was one of two li! ka!n"! performances that included two li! ka!n"! songs. The first song of Su$ a te Ku! nda as is typical, lasted two and a half minutes; the second song, however, only lasted forty seconds. 10 And though bits of spoken narration are interjected in the course of both songs, we note that the temporal ratio of song to story is unusually high, that is, three quarters of the story is taken up in song. The basic structure of Su$ a t # Ku! nda is typical of any li!ka! n"! : a traditional opening formula is stated (:00 :05) followed by an introductory exposition of the circumstances (:05 :19) ; introductory material then transitions into the "inciting moment" and main
222 body of the story (i n this case, the main body consists of two episodes (:20 2:50 and 2:50 4:30 ); the story finally climaxes and soon closes with the traditional closing formula (4:30 4:49) (Higgens 19 81:1 34; Longacre 1996:34 38 ). T he introduct ory material of Su$a! t# Ku! nd a ( :0 0 :20 ) is entirely spoken. At the beginning of the introduction Nguna initiates the first responsorial formula and his audience, as expected, responds (Appendix A, line 1). (call) e sa $ sa (response) saa # (call) li# ka# n!# p!#ngu (response) p "!ngu This opening formula is a p otent index of the li! ka! n "! performance experience. Most traditional Baka narratives are generically framed this way, although many of the participants freely admit that they do not know what this expression lexically denotes. Minimally at this point in the narrative the formula is explicitly held to signify the beginning ( a! na to$ ngo!a! ). Implicitly, such performative speech given its repeated, shared group use over time also typically indexes a host of exp ectations as to what kind of discourse is likely to follow, and what kind of experiences it might likely invoke (Behague! 1992:174 177; Gumperz 1995:395; Kahlberg 1998:5, 9 10; Turino 2000:174 176; Briggs and Bauman 2009:226 in Duranti 2009). The formulai c opening, then, is the first generic, indexical mark of a shift from ordinary verbal interaction, to a specific extra ordinary verbal interaction (cf. Sec. 2.1). Once the formula is uttered, the general circumstances of the narrative as so often happ ens are briefly explained (:05 :20, see also Appendix A, lines 2 5). Then
223 immediately, with the scene set, Nguna shift s from speaking mode to singing mode and transitions to the first episode (at :20) Together, the alteration of his vocal intonation, th at is, "the increase in movement of and stability between the movement of the fundamental frequency," and his shift to a periodic segmentation of his speech (both iconically and indexically) signal a "heightened" orality, and with it the potential of narra tive "progress" (cf. Sec. 2.2; Sec. 4.3.2). The pace of the narrative has changed. The introductory nja! mba has been sounded and the audience choir is implicitly being cued to participate (cf. Sec. 2.4). Concurrently, the sung narrative voice and lyric of Ku! nda (Turtle) (Appendix A, line 8), embodied through Nguna's voice, reveals critical additional co ntextual information: K # j+!k ` !!ln#n#k!s"s`#' d(>!!K ` !!ln#n#k!s"s`#' d(>!!Jt#mc`! Vgn!sgdm>!!Vgn!g`r!jhkkdc!Tmbkd>!!Vgn!g`r!jhkkdc!Tmbkd>!!Ng!Jtmc` S"s`#' n (-!!K ` !!ln#n#k!s"s`#' d(> Tmbkd !!Vgn!g`r!jhkkdc!Tmbkd> Through it we learn that t he allegedly deceased, Su$ a (Leopard), is not just any old animal in the forest, but is Ku! nda 's own (albeit, infamous) uncle ( ti$ ta ). In addition, Ku! nda 's collateral, value laden question "Who killed Uncle?" forcefully expresses and highlights Ku! nda 's distress and apprehen sion regarding the circumstances surrounding his uncle's alleged demise, not to mention the potential difficulties that lie ahead at the wake. Still more evaluative context is added when the intensity of Ku! nda 's ambivalent feelings are reinforced through the exclamatory vocatives Ku! nda !" and Ti$ ta ( o # )!" Through these vocatives, he reveals and expresses his familial solidarity with his
224 deceased uncle, and his subsequent self pity, distress, and anxiety or so we are lead to believe (see also Sec. 4.3 .1.5). After Nguna has intoned these emotionally expressions two and one half times, other participants begin to join him (and Ku! nda ?) i n singing (:40) Soon after, Ku! nda s questioning lyric is segmented ( & :45): Nguna sings one segment in the wa! nja!mb a voice part; the audience and choir sings the other in a wa!tu$ ko! voice part. At this point in the narrative, each one in attendance is now participating in the story and its telling. The multiple voice parts of the solo male singer and mixed choir soo n begin to overlap and interlock the lyric s complementary verse segments whereby a new collective vocal timbre emerges Lexemes, words, verses, lines, and motifs both rhythmic and melodic are repeated. By the end of the fourth verse ( & :50) the clipped timbres of drums and clapping have also emerged as the unmistakable l i! ka!n "! song rhythm begins to permeate the performance. More voices are added; the volume grows a little louder, the tempo a little faster, and even the choirs pitch begins to go sharp (1:00 and 2:30). Clearly, t he patterns of ordinary verbal discour se are being reconfigured, and a unmistakable zone of turbulence is developing (Chap.4, Sec. 3.2). Suddenly (though not unexpectedly) about halfway through the song (1:18) Nguna begins to narrate not sing how Ku! nda on the path to Su$a! s funeral, i s slowly gathering all manner of poisonous plants and insects as perfect tools for his devilish scheme ( Appendix A, line 9) The choir continues the refrain. (Everyone assembled teller and choir, already knows what Ku! nda will do with these itchy, prick ly things, so expectations seem to mount all the more. O ne or two participants momentarily abandon the singing and blurt out feigned cries of incredulity at Ku! nda s dubious
225 intentions, exclaiming e! e $ ? (Whats this?), or a yi! ( No way! ) (e.g., 2 :26, line 12b; 2:33, line 15b). By the last four verses of the singing, the choir is full throated, the drums and clapping are strong, and Ku! nda has gathered all the poisons he needs The song performance and storyline seem to be peaking (2:36) But su ddenly, Nguna abruptly and sharply cues the choir to stop singing and the drummers to stop drumming ( he he he! (2:42)), and without pause again initiates the formulaic episodic call and response (2:44): (call) e sa # sa # (response) saa $ (call) li! ka! n"! p"!ngu (response) p "!ngu The performance formula marks the end of the first episode. But Ku! nda still does not know who has killed his uncle, and we, or rather, the Baka have yet to hear about or perhaps, participate in? Ku! nda s purposes for col lecting all those harmful insects and plants. Narrative resolution is delayed. We are told that Ku! nda simply paddles upstream a little further and arrives at the funeral site where the rest of the animals are gathered around Su$a! s lifeless looking bod y the scene of a second episode (2:47; see also Appendix A, line 18) Episode 2 is shorter than Episode 1. Most of the first h alf of Episode 2 is spoken (2:50 3:50) ; most of the second half is sung (3:50 4:30) As outlined in the storys synopsis, at th e funeral site we are told that Ku! nda s suspicions are quickly revealed and confirmed when the narrator recounts how Su$a! can no longer lie lifeless, moments after Ku! nda places the prickly caterpillar on Su$a! s body (2:51 3:13; Appendix A, lines 20 28).
226 Having incited Su$ a to acti on, Ku! nda immediately counteracts. Nguna quickly narrates Ku! nda 's covert and frantic effort to organize the other animals to fell a tree and crush Su$ a beneath it (3:13 3:38; Appendix A, lines 28 36). The conflict now develops quickly. A second disco urse shift, that is, another "zone of turbulence," is conspicuously signaled in the narrative as Nguna and the entire audience and choir give voice in song to the entire choir of narrative characters singing and narrating their frenzied attempt to fell the tree. Nguna as expected, begins this second song (3:40). First, he intones the opening verse of the wa! tu$ ko! voice part: Nga buu! wa! na! wee! ( e )" ("We are cutting down this tree of calamity!"). The lyric contextualizes the action: the narrative's participants are intensely concentrating, the chopping is continuous, and the scene is calamitous. Ngun a then switches to his wa! nja! mba part. His ideophone kpo! kpo! kpo! (chop chop chop') heightens and vivifies the sound and movement of the tree chopping action, and its repetition reinforces it. His words are sung, not spoken; poetically organized, no t ordinarily organized. His vocal sounds are altered; intonational contours and rhythmic segmentations are re formalized. The intonational contours are not as overtly melic as in the narrative's earlier song; the intervals of the fundamental frequencies are not as melodically formalized; they are more sprechstimme like" though the Baka still recognize such verbal performance as 'e! (song) (Chap. 2, Sec. 2.3). The sung clauses are repeated, and thus become verses; and these same verses then overlap and pattern (in enjambment) with sixteen pulse lines. By the second repetition of the these lines, Nguna 's wa! nja! mba line and voice part is regularly alternating with the choir's line and wa! tu$ ko part. And by now, the extra linguistic sonic signs of clapping and drumming are re sounding again. The drummers,
227 as anticipated, index the li! ka!n "! song experience with the familiar li!k a!n "! song rhythm (Chap. 2, Sec. 4). Climactic development resurges. Expectations are again building. The tempo even increases slightly, as does the overall amplitude. But a third and fourth verse delay the narrative climax when Nguna as the voice of K u! nda lingers and shouts out two quick commands in an effort to coordinate the groups desperate efforts to fell the tree (4:00, 4:08). He even improvises with additional, though brief, melic variation (Figures 5 3, 5 4, 5 5). But finally, in the fifth v erse, with everyone participating singing, clapping, drumming, performing Nguna suddenly stops singing and abruptly shifts to speaking the penultimate closing formula, e sa # sa # (4:16). And all the others immediately respond, saa # Their singing clapping and drumming stop too. And we are immediately told from within the relative silence that Su#a! unexpectedly springs up, and flees away, into the forest in pain (Appendix A, line 42). So, Ku! nda thwarted by Su#a! s escape, blurts out the narra tives final words: Youre lucky! If I had cut the tree down myself, youd be dead! (Appendix A, lines 44 45). To this last ditch threat and expression of frustration, many of the performers respond in laughter; and a few (once again) pretend increduli ty. The closing formula is once more performed and thereby marks the end of that li! ka!n "! s performance (4:33). In general, t he pace of Episode 2 moves faster than that of Episode 1; its length is also shorter almost a third as long. The duration of its song is also a third as long as that of the first episode. B ut the development of Episode 2 in many ways still mirrors that of Episode 1 rhythmically, syntactically, dynamically, and dramatically. As the entire li! ka!n "! unfolds as the intensifying red hue of the diagram in Figure 5 1 indicate s each section of this oral narrative and every element within each section,
228 builds upon another to bring about the collectively desired dramatic effect. The Baka 's song is particularly effective, especially in its ability to both unify and develop their narrative. 5.3 Conclusion: Why Ku# n da Sings At the heart of this performance stands the storyteller singer, Nguna ; and at the heart of his li! ka! n"! stands the character Ku! nda (Turtle). Nguna narrates Ku! nda 's actions, but embodies Ku! nda 's voice. When Ku! nda sings Nguna sings, but Nguna does not sing alone; he is joined by others. Everyone present, then, embodies Ku! nda's singing voice. And it is that voice which both unifies and develops the evening's verbal discourse in so many distinct and complementary ways: rhythmically, syntactic ally, grammatically, phonologically, melodically, timbrically, and semantically. To be sure, not all Baka li! ka!n"! exhibit singing. Discourse cohesion and development are not, therefore, dependent upon song (Lonard 2003:44 155). But most li! ka! n"! do have song, presumably because there are unique advantages to sung narrative discourse. Sung verbal discourse encodes narrative cohesion and development in ways that speech, even heightened speech, does not. As I have demonstrated in Ku! nda's song and nume rous other songs cited in Chapters 3 and 4, Baka song's peculiar rhythmic, syntactic, melic, timbric, and semantic formalizations far outnumber those of most Baka speech acts. Yet the advantages of li! ka! n"! song are not simply discursive, but social, as well. A s one kind of song among many, li! ka!n"! song is a mode of collective social communication that provides a type of communal expression unlike any other in Baka
229 culture. As I noted earlier, virtually no one present at this storytelling event did not actively participate in singing. Except for the two drummers and a few infants, everyone in attendance that evening sang the li! ka!n"! songs male and female, young and old. This inclusive participation is not the case with most Baka song types. Certain kinds of song s tend to be sung exclusively by women, or girls; others are only sung by men, or teenage boys; still others are typically reserved for the youth, or just for solo singing in a quiet corner of the camp. Not so with li! ka! n"! song. Nearly ev eryone gathered sings. When Ku! nda or any li! ka! n"! character sings, everyone present at the storytelling event is afforded the opportunity to lend their voice to his and join in singing. As such, the opportunity that li! ka!n"! song provides for social cohesion is unique among Baka performance genres. This unity is not, however, exclusively formed through shared vocal performance, but also achieved through shared expressive performance. In the first song of the li! ka! n"! of Su$ a t# Ku! nda ," Ku! nda exp resses his incredulity, self pity, and distress over the news of his uncle Su$ a 's alleged death. Later, in the second song, he and all the other animals present, express their anxious anticipation in felling the "calamitous" tree. Many li! ka!n"! songs, maybe even most, give voice to intensified emotional experience, grief and delight, loneliness and confidence, eagerness and dread (Sec. 220.127.116.11; Boursier 1994:8). It is not likely that such performed expressions are simply the imagined experiences of na rrative characters like Ku! nda, but rather, are opportunities for li! ka! n"! singers to collectively express and participate in those emotional experiences common to any Baka community. Fr. Daniel Boursier, having lived among the Baka for many years, and witnessed numerous li! ka! n"!
230 performances in their natural settings, frames the collective expressive function of li! ka!n"! song in a very similar way. Borrowing from French sociolinguist Ferdinand Agblemagnon, Boursier writes: These refrains contribute to creating an atmosphere and to rendering the hearers invested parties: "These songscarry an emotional charge which again encourages and emphasizes collective public participation, such that the storyteller and his audience form but one and the same person, one and the same body. To a large degree, the story is a veritible socio drama, not only involving participation in the play of one or several privileged personalities (actors), but those of the entire assembly (my translation; Agblemagnon 1969:220 cited in Boursier 1994:8) 11 Thus, when Ku! nda sings, he sings in order that the Baka themselves may sing and give "free expression to genuine social conflict" and experience (ibid. 10). Beneath the mask of personalities who distract us, arises our admiration or our indignation, o ften hiding an exacerbated irony or social critique (ibid.) 12 To be sure, this social critique is not always veiled behind "the mask" of a traditional narrative character like Ku! nda Whether sung or not, Elders may use storytelling to [explicitly] addr ess particular problems among the people, and thus resolve conflicts within an atmosphere of relaxation, humour, and entertainment. The storyteller can end a story with an explanation, a comment, or a moral. If he or she is an authority figure, such as a n elder, he/she can direct an application to a situation, such as a rebuke or an admonition (Le' onard 2003:4) Li! ka! n"! song's collective expression of shared experience and shared values also reflects and reinforces a shared memory At many, maybe even most Baka social events involving song or dance, everyone participates in one way or anothe r, whether
231 dancing, singing, drumming, clapping, speaking, or otherwise. But in li! ka! n"! song nearly everyone unites around and utters the same words important words for the Baka, words that form and re form traditional narratives long associated with B aka notions of who they are, where they came from, and what is most important about life in the rainforest (Boursier 1994, Brisson 1999, Higgens 1981, and Le! onard 1997 ) As I have reported in Section 4. 2.1, even when a li! ka!n"! song is performed in iso lation from its li! ka! n"! narrative (e.g., a s a dance song at a funeral ) the Baka nonetheless "see" (i.e., remember ) the basic story with which it is associated and from which it was generated. Thus, it is assumed that when the Baka sing Ku! nda 's song in the story of Su$ a t# Ku! nda ," that story through song is somehow m ade more memorable. Plainly, li! ka! n"! song plays an important role in the Baka 's collective memory The mnemonic capacity of li! ka!n"! song is not surprising, for the power of music t o trigger associations in the memory is legendary" ( Dunsby 2011, referencing Nattiez 1989). Indeed, i t is well documented that song is an effective cognitive aid to memory (Yalch 1991, Wallace 1994 Rainey & Larsen 2002 ). 13 Songs performed in many traditi onal African stories or in any primarily oral society are generally held to be powerful mnemonic aids in the collective memory of that community (Finnegan, 1970). The implications of this phenomenon merit a closer look at some of the more potent mnemonic devices at work in li! ka! n"! song. A fundamental dynamic of mnemonics i s, of course, repetition; and li! ka! n"! song, as has so often and in so many ways been pointed out in this study, is replete with repetition: repeated vowel sounds, repeated phonemes, r epeated lexemes, words, phrases, syntaxes, meters, lines, verses, verse segments, semantic categories, rhythmic
232 figures, melodic figures, tempi, dynamics, and timbres. As demonstrated in Chapter 4, m ost of these tautologies build cohesion, others effect d evelopment, and many signal both Nearly all such patterning also make a song and its story more memorable But no pattern marks a li! ka!n "! song so uniquely as does melodic patterning. All li! ka!n "! songs may share certain typical metrical patterns, syntactic patterns, and rhythmic patterns (Sec. 2.2), but no two li! ka!n "! songs bear the same melodic pattern. A songs melody is li! ka! n "! specific. Yet, while no two song lyrics are set to the same melody, nearly all melodic movement in Baka song is constituted by at least two basic melodic units: repeated melodic refrains and repeated melodic motives. The motives constitute the refr ains while the refrains constitute the songs greater open, iterative structure. The choice to repeat or vary a motive is acted on improvisatorially. Within any li! ka! n "! songs choral texture, melodic motives may be repeated or developed, that is, repea ted with variation. Their repetition and development may take place intra linearly and/or inter linearly, for example, within a single voice part (like the wa! nja!mba or the wa! tu$ ko! ) or between multiple voice parts. Thus, the refrain of Ku! nda s first song, for instance, is iconically constituted by numerous variations on a single melodic motive: a short short long triplet formation of an intervallic sequence etically comprised of a descending perfect fourth/major third and a unison, as shown in Figur e 5 2. 14 The motive regularly begins on the first, fifth, ninth, and thirteenth pulses of Ku! nda s sixteen pulse refrain. Most often the motive is restated as a melodic sequence in a single voice, as shown in measures 9, 13, 17, and 21 of Figure 5 2. Melodic sequence between two voices, as
233 between measures 121 of the lead voice and 125 of Vo ice 1 (in Figure 5 3), is also common. Often enough, the motive's rhythm is strictly repeated, but intervallic substitutions are improvised and added. This kind of variation is exemplified in measures 81 and 85 of Voice 1 and measures 49 and 53 of the lea d voice, shown in Figure 5 4. At other times, however, the motive's rhythm is contracted (Fig. 5 5, measures 149, 157, and 165, V1 & V2). In general, subtly improvised combinations of sequences, substitutions, and contractions permeate the forty iteratio ns of Ku! nda 's melodic refrain. Such repeated refrains and their constitutive motives act as the distinguishing m n em on ic devices of each unique li!ka! n"! song. As suggested by a number of popular music studies, melodic motives, in particular, function a s songs principal m n em on ic hooks (Adorno 1976:34 37 Burns 1987, Kronengold 2005:381 397) 15 Informal evidence of the mnemonic effectiveness of these melodic hooks has often been observed whenever I have elicited the identity of a particular Baka song by simply humming a small portion of its melody to a Baka friend. The moment my Baka friends recognize a melodic motive the song's words soon follow; and if the song is a li! ka! n"! song, a synopsis of the story soon follows as well, maybe even an entire story telling. Nearly every li! ka! n"! song verse has at least one hook' of some kind be it a repeated verse segment, a repeated rhythmic figure, a repeated melodic motif, or some combination thereof. P rocesses of repetition do not aid memory on their own, for they are inextricably linked with that which is repeated. The particular form of each motive is the unique index of each story and song ; its multiple reiterations increasingly ensure its
234 recognition. It is not likely that all melodic motives are equally effective hooks, but to the degree that the melodic motives of li! ka! n"! songs are emically "well formed," songs (and their stories) are more likely made more memorable. Li! ka! n"! songs and stories are not, however, simply "things to be remembered." As Chapter 4 suggests, li! ka! n"! song's developmental function also makes the performance experience more memorable In the performance of Ku! nda 's song, for example, as Ku! nda gather s his poisonous plants and insects melodic and syntactic variations incrementally unfold, the choral performance intensifies the p itch ris es, the tempo increases and drums and voices crescendo. Layers of symbolic, iconic, and indexical signs coalesce and induce a "semantic snowballing effect" (Turino 2000:175 176). Thus, as the performance develops, not only is the narrative marked by li! ka! n"! song, the Baka are "marked" as well. For "signs calling forth densely layered meanings often initially create complex effects which we' experience as feeling (italics added; ibid.). Thus, the affective experiences associated with li! ka! n"! song performance further complicate the semiotic potential of li! ka! n"! song and so index the Baka experience of the narrative that the memory of both its performance and storyline are rendered all the more memorable. As social acts, then, li! ka! n"! songs imp licitly encode multiple socio cultural functions: among them are social cohesion, collective expression, social critique, shared memory and shared experience. But there is more, according to the Baka. In t he ethnographic accounts of Boursier, Brisson, Hi ggens, and L e! onard many more explicit functions of li!ka! n"! song and story are reported (Boursier 1994:8 10 and Higgens 1981:5 7, cited in Le! onard 2003:3 5 ). For example, i n semi permanent forest camps,
235 songs and stories are traditionally held to "weaken the heart of the animals" so that they fall into Baka traps (ibid.). Similarly, at camps on hunting trips, songs and stories "weaken the forest" so that hunters' "chances" for a successful hunt are ensured ( Higgens 1985:101, ibid. ). These are n o small purposes fo r a people so dependent on the forest for their food. In addition to the basic struggle for food, elders are said to tell stories during wakes "to please the spirit of the deceased person in order to maintain peace in the community (Lonard 2003:5). Yet the purposes of stories and their songs are not always so explicitly pragmatic. As explained above, li! ka! n"! song and story are memorably marked. But they are not only memorable in themselves, but also for what they explicitly and implicitly "pass on" a bout everyday life in the forest, beliefs about the origins of things, and traditional customs, mores, and spirituality (Boursier 1991:26 31, 16 1994:10 17, 26 30; Brisson 1999:8 103). Strikingly, such consequential themes are not only sung and spoken in ea rnest, but are plainly and often simultaneously performed to amuse and recreate ( n a m##! soba/s"!l" ) as well. As Boursier describes it, In an atmosphere of play and relaxation, an impressive knowledge is transmited to the newer generations regarding all that touches the forest, from animals and their habits, to the world of the sacred. There emerges a very clear overall impression: this very unified vision of the world of the Baka where the realm of the supernatural, the world of men and of nature communicate and exchange incessantly The daily life and conduct of the Baka only serves to reinforce this impression (my translation; emphasis added; 1994:10) 17 Therefore, when K u! nda sings layers of communicative intent are potentially served. The greater part of this dissertation has decidedly focused on li!ka! n"! song s
236 discursive purposes, that is, how it serves the narrative discourse of which it is a part. Yet, it is like ly that the Baka take such abstract functions quite for granted. Their purposes, as I have just explained, are much more social ly physical ly and sp i ritual ly pragmatic Yet, to say that the Baka would take abstractions like discourse functions for gra nted, does not mean that they would be indifferent to how a li ka! n"! song is composed or performed. For the discernable patterns in their traditional performances strongly suggest that the relationship of form and function are not arbitrary, but ardently intentional. So, for example, Ku! nda does not merely speak, but also sings. And when he sings, he sings in earnest. And we know he is in earnest because of what he sings, and when and how he sings it.
237 Figure 5 1. Diagram of the narrative development of the story of Su$ a t Ku! nda (Leopard and Turtle')
238 Figure 5 2. Melodic motive of Ku! nda s Song: a short short long triplet formation of a descending perfect fourth/major third and a unison. (See also within the complete transcription in Appendix B, measures 9 24)
239 Figure 5 3. Motive of Ku! nda 's Song" restated in melodic sequence between two voice parts
240 Voice 1, Measures 81, 85 Lead Voice, Measures 49, 53 Figure 5 4. Motive of Ku! nda 's Song" strictly restated rhythmically, but with intervallic substitutions
241 Figure 5 5 Motive of Ku! nda 's Song" restated with rhythmic contractions 1 wa! kp" $ kp"! ko kp#$k# is more literally "a very strong story tell er." 2 For fuller accounts of the typical scenes, characters, plots, and functions of Baka li! ka!n"! I refer the reader, again, to the works of Higgens (1981), Brisson (1981, 1996, 1999), Kilian Hatz (1989), Boursier (1994), and Le! onard (1997). 3 Intere stingly, the second song in both of these narratives employed sprechstimme throughout. 4 This should not be taken to mean that the story still carried a typical text load; it did not. In fact, it is difficult to understand if there was any narration spoke n at all in such cases. Just how such a short lyric and such little spoken narration could constitutute a "complete" story will need further investigation. 5 More complete transcriptions are needed to verify this, and to better describe the distribution i n a story's structure. 6 This discourse function warrants more study, though requires better representation of the overlap of speaking and singing in performances. 7 In other contexts, for example other li! ka! n"! songs and other Baka song genres, these te rms would need further clarification. The clarification is needed in that while these terms may be the lexical extent to which the Baka distinguish voice parts, there may be other distinctives.
242 8 The nature of these composite versions of both parts vari es with any li ka!n"! song and is an interesting study in itself. As an example, the description of Su# a te Ku! nda presents some of the particulars as to the composite nature of its introductory wa! nja! mba 9 The wa! nja! mba part is normally only performed by the lead singer, but not always. Occasionally, once the lead singer has adequately stated the wa! nja! mba part, the wa! nja! mba part may be picked up by other voices as well (especially other adult male voices, as in the case of a few of these sixteen li! ka! n"! song s, for example Ku! nda t$ Jelo ). 10 When a traditional Baka story includes more than one occurrence of song in the course of a narrative's performance, songs typically become progressively shorter in duration. This pattern is consistent with the "change of pace" that Longacre claims is so commonly found in narrative discourse development, climactic development in particular. 11 Aggblemagnon, F. Sociologie des socie# te# s orales d 'Afrique noire, Mouton, Paris, 1969. Boursier's text reads: Ces refrains co ntribuent a! cre" er un climat et a! rendre les auditeurs partie prenante: "Ces chansonsapportent une charge emotive qui favorise et accentue encore la participation collective du public, de telle sorte que le conteur et son auditoire ne forment qu'une seu le et meme personne, qu'un seul meme corps. Dans une large mesure, le conte est un ve" ritable socio drame n'impliquant pas seulement participation au jeu d'un ou plusieurs personnages privile" ge" s (acteurs), mais celle de l'assemble" e entie! re" (Agblemagn on 1969:220 cited in Boursier 1994:8). 12 Ibid.: Sous le masque de personnages qui nous distraient, suscitent notre admiration ou notre indignation, se cache souvent une forme exacerbe" e d'ironie ou de critique sociale" (ibid.). 13 Wallace, W. (1994). "Memo ry for music: effect of melody on recall of text." Learning, Memory, and Cognition 20(6), 1471 1485; Rainey, David W. & Larsen, Janet D (2002) "The Effect of Familiar Melodies on Initial Learning and Long term Memory for Unconnected Text." Music Perceptio n Winter, Vol. 20, No. 2, 173 186; Yalch, Richard (1991) "Memory in a Jingle Jungle: Music as a Mnemonic Device in Communicating Advertising Slogans." Journal of Applied Psychology Vol.76, No. 2, 268 275. 14 My etic transcription includes quarter tones, i ndicated by a "+" over the notehead. Because of this, a number of the motive s descending intervals are neither /P4/ (perfect fourths), nor /M3/ (major thirds), but rather M3+ (e.g., see measure 13 ) 15 See Adorno 1976:34 37; Burns, Gary. 1987. "A Typolog y of Hooks' in Popular Records," Popular Music 6(1): 1 20; Middleton, Rich a rd 1990:51, 98; Shepherd, John, 2003. "Hooks" in Continuum Encyclopedia of Popular Music of the World London: Continuum, p. 563 564; Vernallis, Carol. Experiencing Music Video: A esthetics and
243 Cultural Context / Carol Vernallis. New York: Columbia University Press, 2004. Kronengold, Charles. 2005. "Accidents, hooks and theory." Popular Music 24, pp 381 397. 16 See Boursier, Daniel, Rflexions sur l'vanglisation des Baka Vivant U nivers n¡ 396, nov. dc. 1991, pp. 26 31. 17 Boursier's text reads: Dans une ambience de jeu and de de# tente un savoir impressionante est transmit aux nouvelles ge" ne" rations pour tout ce qui touche a! la fort, aux animaux et leur moeurs, au monde du sacre" Il s'en degage une impression globale tre! s nette: cette vision tre! s unifie" e du monde baka ou! l'au dela! le monde des hommes et la nature communique et e" changent sans cesse. Le ve" u et le comportement quotidien s des Baka ne font que renforcer cette impression" (italics added; 1994:10).
251 APPENDIX B ETIC MUSIC TRANSCRIP TION OF VOICE PARTS : 1 KU! NDA 'S SONG" FROM THE BA KA NARRATIVE SU# A! T% KU! NDA ( Object 5 3. Audio file of li! ka! n"! story Su# a t$ Ku! nda (.m p3 4.4 MB) ) Note: (1) This transcription only represents the melodic and rhythmic elements of the voice parts; percussion parts are represented in a composite transcription in Chapter 2, Figure 2 4; textual representation is only provided during the initia l entries of certain voice parts; (2) a plus symbol over a note head indicates that the pitch is a quarter tone sharp; a diagonal slash through a plus symbol however, indicates that a quarter tone alteration is cancelled ; (3) note stems without note heads rhythmically indicate a vocal presence, but only an approximate pitch; (4) notes with x note heads indicate indefinite pitches/tones. q = 140 MM.
261 1 In Section 2. 2.2, I implicitly question etic (i.e., outsider") notions of Baka perceptions of melodic intervals. Thus, in preparing for my forthcoming emic analysis of Baka intervals, I have included quarter tones in most of my transcriptions in the event that these melodic variations turn out to be emical ly significant to the Baka. I have left the representations of quarter tones in this present transcription of Kunda's song," but they have little relevance to my present concerns.
262 REFERENCES Abrahams, Roger D. 1972. "Folklore and Literature as Performance." In Journal o f the Folklore Institute 9 (2/3) (Aug. Dec.): 75 94. 1983. African Folktales: Traditional Stories of the Black World The Pantheon Fairy Tale and Folklore Library; variation: Pantheon Fairy Tale & Folklore Library. 1st ed. New York: Pantheon Books. A dorno, Theodor W. 1976. Introduction to the Sociology of Music New York: Seabury Press. Agawu, V. Kofi. 1991. Playing with Signs: A Semiotic Interpretation of Classic Music Princeton, N.J.: Princeton University Press. 1995. African Rhythm: A North ern Ewe Perspective Cambridge; New York: Cambridge University Press. 2003. Representing African Music: Postcolonial Notes, Queries, Positions New York: Routledge. 2009. Music as Discourse: Semiotic Adventures in Romantic Music Oxford Studies in Music Theory. Oxford; New York: Oxford University Press. Agblemagnon, F. NSougan. 1969. Sociologe des socits orales d'afrique noire; les eve du sud togo Le monde d'outre mer, pass et prsent; premire srie; etudes; 35. Paris: Mouton. Agha, Asif 2001. "Register." In Key Terms in Language and Culture edited by Alessandro Duranti, 212 215. Malden, Mass.: Blackwell. Alter, Robert. 1985. The Art of Biblical Poetry New York: Basic Books. Alvarez Cccamo, Celso. 2001. "Codes." In Key Terms in Lan guage and Culture edited by Alessandro Duranti, 23 26. Malden, Mass.: Blackwell. Arom, Simha. 1976. "The Use of Play back Techniques in the Study of Oral Polyphonies." Ethnomusicology 20 (3) (Sep.): 483 519. 1985. Polyphonies et polyrythmies instrum entales d'afrique centrale: Structure et mthodologie Ethnomusicologie; 1. Paris: SELAF. 1991. African Polyphony and Polyrhythm: Musical Structure and Methodology Cambridge; New York; Paris: Cambridge University Press; Editions de la maison des sci ences de l'homme.
263 1991. "Modlisation et Modles dans les Musiques de Tradition Orales." Analyse Musicale 1 er Trimestre (1991): 67 78. Arom, Simha, and Patrick Renaud. 1990. Cameroon Baka Pygmy Music UNESCO collection; musics and musicians of the world; musiques et musiciens du monde variation: [Cameroun: la musique des Pygmes Baka]. Reissue Auvidis ed. Vol. 1 sound disc : digital; 4 3/4 in. Ivry sur Seine: Auvidis. Arom, Simha, and Pierre Toureille. 1987. Centrafrique anthologie de la musique d es pygmes aka Vol. 2 sound discs (135 min.): digital, stereo; 4 3/4 in. + Includes: 1 pamphlet (36 p.). Paris: Ocora: Distribution, Harmonia Mundi France. Austin, J. L. 1962. How to Do Things with Words Cambridge: Harvard University Press. Banti, Gior gio and Francesco Giannattasio. 2004. "Poetry." In A Companion to Linguistic Anthropology edited by Alessandro Duranti, 290 319. Malden, MA: Blackwell Pub. http://www.netlibrary.com/urlapi.asp?action=summary&v=1&bookid=10 8704. Basso, Ellen B. 1985. A Mus ical View of the Universe: Kalapalo Myth and Ritual Performances University of Pennsylvania Publications in Conduct and Communication. Philadelphia: University of Pennsylvania Press. Ba h uchet, Serge and J.M.C. Thomas. 1986. "Linguistique et histoire des Pygmes de l'ouest du bassin Congolais." Sprache und geschichte in Afrika 7.2, p.73 103. Ba h uc h et, Serge. 1989. "Les Pygmes Aka et Baka: contribution de l'ethnolinguistique a l'histoire des populations forestieres d'Afrique Centrale." Thse de Docto rat d'Etat des Lettres et Sciences Humaines, Universite Rene Descartes Paris V, Paris. 1993b. "History of the inhabitants of the central African rain forest: perspectives from comparative linguistics. tropical forests, people and food." Biocultur al interactions and applications to development ed. by C.M. Hladik, 37 54. Paris: UNESCO/Parthenon. 1993c. "L'invention des Pygmes." Cahiers d'Etudes africaines 129, XXXIII I, p.153 181. Bauman, Richard.  1984. Verbal Art as Performance P rospect Heights, Ill: Waveland Press.
264 1986. Story, Performance, and Event: Contextual Studies of Oral Narrative Cambridge Studies in Oral and Literate Culture; 10. Cambridge; New York: Cambridge University Press. 1992. Folklore, Cultural Perf ormances, and Popular Entertainments: A Communications C entered Handbook New York: Oxford University Press. 1998. "The Ethnography of Performance in the Study of Oral Tradition." In Teaching Oral Traditions ed. John Miles Foley, 106 122. New York: Modern Language Association. Bauman, Richard, and Charles L. Briggs. 1990. "Poetics and Performance as Critical Perspectives on Language and Social Life." Annual Review of Anthropology 19: 59 88. Bauman, Richard, and Joel Sherzer. 1989. Explorations in t he Ethnography of Speaking Studies in the Social and Cultural Foundations of Language; no. 8. 2nd ed. Cambridge; New York: Cambridge University Press. Baumann, Max Peter. 2011. "Yodel." In Grove Music Online Oxford Music Online http://www.oxfordmusicon line.com.lp.hscl.ufl.edu/subscriber/article/grove /music/52555 (accessed February 1, 2011). Beard, David, and Kenneth Gloag. 2005. Musicology: The Key Concepts Routledge Key Guides. London; New York: Routledge. Bebey, Francis. 1969. Musique de l'Afrique Expressions. Paris: Horizons de France. 1975. African Music: A People's Art 1 U.S. ed. New York: Lawrence Hill & Company. 1982. "La merveilleuse musique des pygmes: Des choeurs dans la forte." Balafon (55, 2e trim.): 56 64. 1995. Nandolo with Love: Francis Bebey, Works, 1963 1994 Vol. 1 sound disc (ca. 75 min.): digital; 4 3/4 in. + Includes: 1 folded sheet. Tivoli, NY: Original Music. Belcher, Stephen Paterson, Thomas A. Hale, Thomas A. Hale, Thomas A. Hale, and John William Johnson. 1 997. Oral Epics from Africa : Vibrant voices from a vast continent African Epic Series. Bloomington: Indiana University Press. Ben Amos, Dan. 1975. Sweet Words: Storytelling Events in Benin Philadelphia: Institute for the Study of Human Issues.
265 Bendor Samuel, John, and Rhonda L. Hartell. 1989. The Niger Congo languages: a classification and description of Africa's largest language family. Lanham: University Press of America. Biber, Douglas. 1999. Longman Grammar of Spoken and Written English Harlow, En gland; New York: Longman. Biebuyck, Daniel P., and Mateene Kahombo. 1969. The Mwindo Epic from the Banyanga (Congo Republic) Berkeley: University of California Press. Bodo, Jean Marie. 1992. La Musique, Instrument Privilgi de la Liturgie pour la Second e vanglisation du Cameroun Paris: ditions ARM. 1996. Bienvenue sa Saintet le Pape Jean Paul II au Cameroun (audio recording). Paris: ARM 03. Boursier, Daniel. 1991. Rflexions sur l'vanglisation des Baka.'' Vivant Univers n¡ 396, nov. d c., pp. 26 31. 1994. Depuis ce jour l: Contes des pygmes baka du sud est cameroun: Contes bilingues, baka franais Collection "la lgende des mondes"; variation: Collection "la lgende des mondes." Paris: L'Harmattan. Brisson, Robert. 1981. Contes des pygmes aka du sud cameroun Vol. 1 4 Douala: Collge Libermann. 1996. Contes des pygmes baka Livret 1 3. Yaound: Imprimerie Saint Paul. 1999. Mythologie des pygmes baka (sud cameroun): Mythologie et contes Divers (socit d'tudes linguistiques et anthropologiques de france); 13. Leuven; Leuven; Paris: Peeters; Peeters; SELAF. 2002. Petit dictionnaire baka franais Limited edition of an unpublished revision of the 1979/1984 Brisson Boursier publication (below). and Dani el Boursier. 1979/1984. Petit dictionnaire baka franais College Libermann. Douala, Cameroun. Burns, Gary. 1987. "A Typology of Hooks' in Popular Records," Popular Music 6(1): 1 20. Cavalli Sforza, Luigi Luca (ed.). 1986. African Pygmies Orlando: Harcou rt Brace Jovanovich.
266 Charry, Eric S. 2000. Mande Music: Traditional and Modern Music of the Maninka and Mandinka of Western Africa Chicago Studies in Ethnomusicology. Chicago: University of Chicago Press. Chenoweth, Vida. 1966. "Song Structure of a New G uinea Highlands Tribe. Ethnomusicology 10 (3) (Sep.): 285 2 97. 1968. "Managalasi Mourning Songs." Ethnomusicology 12 (3) (Sep.): 415 41 8. 1972. Melodic Perception and Analysis: A Manual on Ethnic Melody Ukarumpa, Papua: Summer Institute of Li nguistics. 1979. The Usarufas and their Music Publication of SIL Museum of Anthropology 5. Dallas, Texas: SIL Museum of Anthropology. and Darlene Bee. 1971. "Comparative generative Models of a New Guinea Melodic Structure." American Anthropolog ist 73 (3) (Jun.): 773 82. Clendenen, E. Ray. 1989. "The Interpretation of Biblical Hebrew Hortatory Texts: A Textlinguistic Approach to the Book of Malachi," Ph.D. diss., University of Texas at Arlington. Cobley, Paul, Litza Jansz, and Richard Apignanes i.  1999. Introducing Semiotics Cambridge: Icon. Cogan, Robert. 1984. New Images of Musical Sound Cambridge, Mass.: Harvard University Press. Conference on Style, edited by Thomas A. Sebeok. 1960. Style in Language Cambridge: Technology Press of Massachusetts Institute of Technology. Coplan, David B. 1987. "The Power of Oral Poetry: Narrative Songs of the Basotho Migrants." Research in African Literatures 18 (1, Special Issue on Literature and Society) (Spring): 1 35. 1988. "Musical Underst anding: The Ethnoaesthetics of Migrant Workers' Poetic Song in Lesotho." Ethnomusicology 32 (3) (Fall; Fall 1988): 337. 1994. In the Time of Cannibals: The Word Music of South Africa's Basotho Migrants Chicago Studies in Ethnomusicology. Chicago: Uni versity of Chicago Press. Cooke, Peter and Michelle Kisliuk. "Pygmy Music." In Grove Music Online Oxford Music Online
267 http://www.oxfordmusiconline.com.lp.hscl.ufl.edu/subscriber/article/gro ve/music/22590 (accessed January 31, 2011). Crystal, David. [199 7] 2008. A Dictionary of Linguistics and Phonetics The Language Library, 6th ed. [Other editions: 1985, 1990, 1991, 1997, 2001, 2004.] Malden, MA; Oxford: Blackwell Pub.. Culler, Jonathan D. 1975. Structuralist Poetics: Structuralism, Linguistics, and the Study of Literature Cornell Paperbacks; variation: Cornell Paperbacks. Ithaca, N.Y.: Cornell University Press. De Beaugrande, Robert. 1997. New Foundations for a Science of Text and Discourse : Cognition, Communication, and the Freedom of Access to Know ledge and Society Advances in Discourse Processes; v. 61. Norwood, N.J: Ablex. De Beaugrande, Robert, and Wolfgang U. Dressler. 1981. Introduction to Text Linguistics Longman Linguistics Library; title no. 26; A Longman Paperback. London; New York: Long man. Dooley, Robert A., and Stephen H. Levinsohn. 2001. Analyzing Discourse: A Manual of Basic Concepts Dallas, Tex: SIL International. Dorson, Richard M. 1972. Folklore and Folklife, An Introduction Chicago: University of Chicago Press. 1975. Fo lktales Told around the World Chicago: University of Chicago Press. Duke, Daniel Joseph. 2001. Aka as a Contact Language: Sociolinguistic and Grammatical Evidence Masters Thesis: University of Texas at Arlington. Dunsby, Jonathan. 2011. "Memory, Me morizing." In Grove Music Online Oxford Music Online http://www.oxfordmusiconline.com.lp.hscl.ufl.edu/subscriber/article/grove /music/42568 (accessed February 1, 2011). Duranti, Alessandro. 2001. Key Terms in Language and Culture Malden, Mass.: Blackwell England, Nicholas. 1967. "Bushmen Counterpoint." Journal of the International Folk Music Council : 58 66. Eno Belinga, Samuel Martin. 1965. Littrature et musique populaires en afrique noire, Collection culture et coopration, 1. Paris: ditions Cujas.
268 1970. Dcouverte des chantefables beti bulu fang du cameroun. prf. de pierre alexandre: Langues et littratures de l'afrique noire, 7. Paris: Klincksieck. Feld, Steven.  1990. Sound and Sentiment: Birds, Weeping, Poetics, and Song in Kaluli Expr ession Publications of the American Folklore Society. New series (unnumbered). 2nd ed. Philadelphia: University of Pennsylvania Press Feld, Steven, and Aaron A. Fox. 1994. "Music and Language." Annual Review of Anthropology 23: 25 53. Feld, Steven, Aaron Fox, Thomas Porcello, and David Samuels. 2004. "Vocal Anthropology: From the Music of Language to the Language of Song." In A Companion to Linguistic Anthropology edited by Alessandro Duranti, 322 342. Malden, MA: Blackwell Pub. Finnegan, Ruth H. comp. 1 967. Limba Stories and Story telling compiled and translated by Finnegan Oxford Library of African Literature. Oxford: Clarendon P. 1977. Oral Poetry: Its Nature, Significance, and Social Context Cambridge; New York: Cambridge University Press. 1988. Literacy and Orality: Studies in the Technology of Communication Oxford, UK; New York, NY, USA: Blackwell. 1992. Oral Traditions and the Verbal Arts: A Guide to Research Practices ASA Research Methods in Social Anthropology (Routledge). Lond on; New York: Routledge. 2002. Communicating: The Multiple Modes of Human Interconnection London: Routledge. 2007. The Oral and Beyond: Doing Things with Words in Africa Oxford; Chicago; Pietermaritzburg: James Currey; University of Chicago Pr ess; University of KwaZulu Natal Press. Fitzgerald, Dan. 1999. "A Profile of Cameroonian Music Research." Unpublished research paper for SIL Cameroon. firstname.lastname@example.org 2003. "Baka Song Text Poe tics: A Preliminary Inventory of Poetic Devices in Nine Songs." Unpublished study partially submitted in 2003 as an SIL Cameroon research report for Cameroon's Ministere de la Recherche Scientifique et Technique. Yaounde, Cameroon.
269 2009. "What Can One Learn by Comparing European Opera and Chinese Opera?" Unpublished final paper for a musicology course (MUH 5634) at the University of Florida. email@example.com. Foley, John Miles. 1988. The Theory of Oral Composition: History and Methodology Fo lkloristics. Bloomington: Indiana University Press. 2002. How to Read an Oral Poem Urbana: University of Illinois Press. 1998. Teaching Oral Traditions Options for Teaching; 13. New York: Modern Language Association. and Milman Parry. 19 87. Comparative Research on Oral Traditions: A Memorial for Milman Parry Columbus, Ohio: Slavica. Fortune, G. 1962. Ideophones in Shona: An Inaugural Lecture Given in the University College of Rhodesia and Nyasaland on 28 April 1961 Oxford University Pr ess. Frisbie, Charlotte J. 1971. "Anthropological and Ethnomusicological Implications of a Comparative Analysis of Bushmen and African Pygmy Music." Ethnology 10 (3) (Jul.): 265 90. 1980. "Vocables in Navajo Ceremonial Music." Ethnomusicology 24 (3) (Sep.): 347 92. Frniss, Susanne. 1991a. "Recherches Scalaires chez les Pygmes Aka." Analyze Musicale 23, p.31 35. 1991b. "La Technique du Jodel chez les Pygmes Aka (Centrafrique): tude Phontique et Acoustique." Cahiers de Musiques Traditionelle s 4 (1991): 167 187. 1993. "Rigueur et Libert: La Polyphonie Vocale des Pygmes Aka." In Polyphonies de Tradition Orale: Histoire et Traditions Vivantes edited by Michel Huglo, Marcel Prs, and Christian Meyer, 101 131. Paris: ditions Craphis, 1993. 2006. "Aka Polyphony: Music, Theory, Back and Forth." In Analytical Studies in World Music, edited by Michael Tenzer. Oxford University Press, p.163 204. Geertz, Clifford. 1973. The Interpretation of Cultures: Selected Essays New York: Basic Bo oks.
270 Grimaud, Yvette, Gilbert Rouget, Jane Wenning, Muse de l'homme (Musum national d'histoire naturelle), and Dpartement d'ethnomusicologie. 1957. Notes sur la musique des bochiman compare celle des pygmes babinga, tablies Paris: Muso de l'Homm e Dpartement d'Ethnomusicologie. Grimes, Barbara F. 2000. Ethnologue: Languages of the World 14 th Edition. Dallas: SIL International. Grimes, Joseph Evans. 1975. The Thread of Discourse Janua Linguarum. Series Minor; 207. The Hague: Mouton. Gulich, Elisabeth, and Uta M Quasthoff. 1985. "Narrative Analysis," in Handbook of Discourse Analysis, Vol. 2, Dimensions of Discourse edited by Teun A. van Dijk, 169 197. London: Academic Press. Halliday, M. A. K. and Ruqaiya Hasan. 1976. Cohesion in English E nglish Language Series; no. 9. London: Longman. Hampton, Barabra. 1998. "Identities: Music and other African Arts." In The Garland Encyclopedia of World Music : Volume 1, Africa, edited by Ruth M. Stone, 102 123. New York: Garland Pub.. Herndon, Marcia, an d Roger Brunyate. 1975. Proceedings of a Symposium on Form in Performance: Hard core Ethnography Symposium on Form in Performance (1975 : University of Texas, Austin). Austin: Office of the College of Fine Arts. Herndon, Marcia, and Norma McLeod. 1979. M usic as Culture Norwood, Pa.: Norwood Editions. Higgens, Kathleen. 1981. Narrative Folklore Discourse of the Baka Language: Preliminary Analysis. SIL Academic Library,Yaounde, Cameroon. 1981. Analog Audio Field Recordings of Baka Narrative Performa nces The audio data for Higgen's 1981 Narrative Folklore Discourse of the Baka Language: Preliminary Analysis. (Unpublished. Provisionally archived in SIL Cameroon's audio visual archive, Yaound, Cameroon.) 1985. Ritual and Symbol in Baka Life His tory." Anthropology and Humanism Quarterly, December, 100 106. Hinton, Leanne. 1977. Havasupai Songs: A L inguistic Perspective Ph.D. diss., University of California, San Diego.
271 Howie, John and Pierre Delattre. 1970. "An Experimental Study of the Effect of Pitch on the Intelligibility of Vowels." The Bulletin of National Association of Teachers of Singing XVIII (May, 1962): 6 9. Quoted in William A. Hunt. "Spectrographic Analysis of the Acoustical Properties of Selected Vowels in Choral Sound." (Ph.D. diss. North Texas State University, p.2.) Jakobson, Roman. 1960. "Linguisitics and Poetics," In Style in Language, edited by Thomas A. Sebeok. The Massachusetts Institute of Technology. 1 966. Selected Writings (vol.III "Poetry of Grammar and Grammar of Poetry") Mouton: The Hague. 1973. Essais de linguistique gnrale: Rapports internes et externes du langage Arguments; 57. Paris: Les Editions de Minuit. Jakobson, Roman, C.G.M. Fant, and M. Halle. 1967. "Preliminaries to Speech Analysis." MIT Acoustic Laboratory Technical Report No. 13 (1952). Quoted in Andrew H. Harper Jr. "Spectrographic Comparison of Certain Vowels to Ascertain Differences Between Solo and Choral Singing, Reinforced by Aural Comparison." (Ph.D. diss., Indiana University. ) Ki lian Hatz, Christa, Robert Brisson and Daniel Boursier. 1989. Contes et proverbes des pygmes baka Collection paroles d'afrique. Paris: Agence de coopration culturelle et technique. Kilian Hatz, Christa. 1995 . Das Baka: Grundzge einer Grammatik aus der Grammatikalisierungsperspektive Kln: Institut fr Afrikanistik, Universitt zu Kln. [Unpublished English translation, translated in 2004 from German by Wycliffe Associates UK as Baka Grammar: A Grammaticalization Account edited by Yves Leonard. Held in SIL Academic Library, Yaound, Cameroon.] Kilian Hatz, Christa, and Erhard Friedrich Karl Voeltz. 2001. Ideophones Typological Studies in Language, 0167 7373; v. 44., ed. International Symposium on Ideophones. Amsterdam; Philadelphia: J. Benjami ns. Kindell, Gloria. 1996. "Ethnopoetics: Finding Poetry." Notes on Literature in Use and Language Programs no. 50. Kisliuk, Michelle Robin. 1991. Confronting the Quintessential: Singing, Dancing, and Everyday Life among Biaka Pygmies (Central African Republic) ." Ph.D. diss., N.Y.U..
272 1997. "(Un)doing Fieldwork: Sharing Lives, Sharing Songs." In Shadows in the Field edited by Gregory F. Barz and Timothy J. Cooley, 23 44. Oxford University Press. 1998. "Musical Life in the Central African Rep ublic." In The Garland Encyclopedia of World Music; Africa Vol. 1., edited by Ruth M. Stone, 681 697. New York: Garland Press. 1998. Seize the Dance!: BaAka Musical Life and the Ethnography of Performance Oxford; New York: Oxford University Press. 2000. "Performance and Modernity among BaAka Pygmies: A Closer Look at the Mystique of Egalitarian Forgers in the Rainforest." In Music and Gender edited by Pirkko Moisala and Beverley Diamond, 25 50. Urbana: University of Illinois Press. Kronengold Charles. 2005. "Accidents, hooks and theory." Popular Music 24, pp 381 397. Lonard, Yves. 1997. "The Baka: A People between Two Worlds." M.A., Providence Theological Seminary. 2003. Baka Oral Narratives. Yaounde: SIL, Cameroon. 2005. Prag matic Features of Baka Narrative Discourse An SIL Cameroon research report for Cameroon's Ministere de la Recherche Scientifique et Technique. Yaounde, Cameroon. [A revision of Baka Oral Narratives (2003).] 2009. Orthographe baka." Yaound, Camerou n: SIL. An SIL research report for Ministre de la Recherche Scientifique et de l'Innovation au Cameroun. Lvi Strauss, Claude. 1969. The Raw and the Cooked New York: Harper & Row. Levinsohn, Stephen. 2004. "Analysis of Narrative Texts." Lecture notes f or a course given from Februrary 3 14, 2003 at Horsleys's Green, UK. (A similar course was presente d in Yaounde" Cameroon in 2004.) List, George. 1963. "The Boundaries of Speech and Song." Ethnomusicology 7 (1) (Jan.): 1 16.
273 Lomax, Alan 1967. "Special Features of the Sung Communication." In Essays on the Verbal and Visual Arts, Proceedings of the 1 966 Annual Spring Meeting of the American Ethnological Society June Helm, ed., 109 127. Seattle: University of Washington Press. 1976. Cantometrics: An Approach to the Anthropology of Music Berkeley, CA.: distributed by University of California, Ex tension Media Center. and Edwin E. Erickson. 1968. Folk Song Style and Culture Washington: American Association for the Advancement of Science. Longacre, Robert E. 1989a. "Two Hypotheses Regarding Text Generation and Analysis." Discourse Processe s, 12 p.413 460. 1990. Storyline Concerns and Word Order Typology in East and West Africa .# Studies in African Linguistics, supplement, 10. Los Angeles: James S. Coleman African Studies Center and Department of Linguistics, University of California. viii, 181 p. 1996. The Grammar of Discourse 2 nd Edition. Plenum Press. 2003. Joseph: A Story of Divine Providence: A Text Theoretical and Textlinguistic Analysis of Genesis 37 and 39 48.# 2nd ed. Winona Lake, Indiana: Eisenbrauns, 59. 20 06. Discourse Structure, Verb Forms, and Archaism in Psalm 18. Journal of Translation Volume 2, Number 1, 17. Lord, Albert Bates, Stephen Mitchell, Stephen A. Mitchell, and Gregory Nagy. 2000. The Singer of Tales Harvard Studies in Comparative Literatu re; 24. 2nd ed. Cambridge, Mass.: Harvard University Press. Luling, Virginia, Justin Kenrick, and Survival International. 1998. Forest Foragers of Tropical Africa: A Dossier on the Present Condition of the Pygmy' Peoples ." London: Survival for Tribal Pe oples. Malinowski, Bronislaw.  1961. Argonauts of the Western Pacific: An Account of Native Enterprise and Adventure in the Archipelagoes of Melanisian New Guinea New York: Dutton. Mann, William C., Sandra A. Thompson. 1992. "Relational Discourse S tructure: A Comparison of Approaches to Structuring Text by Contrast." In Language in Context: Essays for Robert Longacre edited by Shin Ja J. Hwang and Robert R. Merrifield, 19 46. Arlington, TX: Summer Institute of Linguistics and the University of Texa s at Arlington, Publication 107.
274 McCormick, Jennifer, Sarah Richardson. 2007. "Vocatives in MICASE." http://micase.elicorpora.info/files/0000/0055/Kibb_12 Vocative.pdf (acc essed january 2011). Loos, Eugene, ed.. 2004. Glossary of Linguistic Terms In LinguaLinks Library, edited by Eugene Loos, with Susan Anderson, Dwight H. Day Jr., Paul C. Jordan, and J. Douglas Wingate. SIL International http://www.sil.org/linguistics/GlossaryOfLinguisticTerms/WhatIsIntonation .htm (accessed january 31, 2011). McLeod, Norma, and Marcia Herndon. 1980. The Ethnography of Musical Performance Norwood Pa: Norwood Editions. Middleton, Richard. 1990. Studying Popular Music Milton Keynes England; Philadelphia: Open University Press. Miller DG and HK Schutte. 1993. "Physical Definition of the "Flageolet Register"." J Voice 1993 Sep; 7(3):206 12. Schoo l of Music, Syracuse University, New York. Molino, Jean. 2002. "La poesia cantato." In Sul verso cantato [on sung verse] ed. M. and F. Giannattasio Agamennone, 17 33. Padoa: il Poligrafo. Molino, Jean and Jolle Tamine Gardes. 1982. Introduction l'anal yse linguistique de la posie Linguistique nouvelle. 1re d ed. Paris: Presses universitaires de France. Nattiez, Jean Jacques. 1973. "Linguistics: A New Approach for Musical Analysis?" International Review of the Aesthetics and Sociology of Music 4 (1) (Jun.): 51 68. 1974. "Smiologie musicale: L'tat de la question." Acta Musicologica 46 (2) (Jul. Dec.): 153 71. 1988. De la smiologie la musique Cahiers du dpartement d'tudes littraires; 10. 1re d ed. Montral, Quebec: Service des pub lications, Universit du Qubec Montral. 1989. Proust as Musician Cambridge; New York: Cambridge University Press. 1990. Music and Discourse: Toward a Semiology of Music Princeton, N.J.: Princeton University Press.
275 1990. "The Contri bution of Musical Semiotics to the Semiotic Discussion in General." In The Garland Library of Readings in Ethnomusicology Vol.2 : 221 242. New York: Garland Publishing. 1999. "Inuit Throat games and Siberian Throat Singing: A Comparative, Historical, a nd Semiological Approach." Ethnomusicology 43 (3) (Autumn): 399 418. Ngumu, Pie Claude. 1964. La Psalmodie Camerounaise (audio recording). Espagne : Edigsa CE 6. 1966. La Messe Ewondo (audio recording). Paris: R. Nagy Eau bonne CAM 02. 1971. Matrise des Chanteurs la Croix d'bene de Yaound Yaound: n.p.. 1976. Les Mendzan: des Chanteurs de Yaound: Histoire, Ouganologie, Fabrication, Systme de Transcription WEIN. (ACTA Ethnologistica et Linguistica, no. 34.) 1984. "Les Reche rchers Ethnomusicologiques en Afrique Centrale: le Cas du Cameroun." Recherche, Pdagogie, et Culture no. 65 66. (1984): 64 69. 1984. "L'ethnomusicologie et l'Agence de Coopration Culturelle et Technique Erpamao, EMAC." Recherche, Pdagogie, et Cu lture no. 65 66. (1984): 91 98. 1985. "Cameroon Cultural Identity and Musical Art." In Cameroon Cultural Identity Yaound: Ministry of Information and Culture, Dept. of Cultural Affairs. Nj$ma and Tombombo, Mai, And% and Awu. 2008. Recorded interv iew with authors, Nomedjo, Cameroun, March. (Baka musicians, singers, and story tellers) Okpewho, Isidore. 1979. The Epic in Africa: Toward a Poetics of the Oral Performance New York: Columbia University Press. 1992. African Oral Literature: Backgr ounds, Character, and Continuity Bloomington: Indiana University Press. 1983. Myth in Africa: A Study of its Aesthetic and Cultural Relevance Cambridge Cambridgeshire; New York: Cambridge University Press.
276 Olivier, Emmanuelle. 1998. "The Ju/ hoans i Bushmen's Conception of their Musical World." Lacito. (1998): 1 15. (Paris: CNRS.) a nd Susanne Frniss. 1996. "Musique Pygme/Musique Bochiman: Nouveaux lments de Comparaison." Lacito (1996): 1 19. (Paris: CNRS.) Opland, Jeff. 1983. Xhosa Oral Po etry: Aspects of a Black South African Tradition Cambridge Studies in Oral and Literate Culture; 7. Cambridge Cambridgeshire; New York: Cambridge University Press. Parry, Milman. 1928. Les formules de la mtrique d'homre Paris: Les Belles Lettres. Par ry, Milman, and Adam Parry. 1971. The Making of Homeric Verse: The collected Papers of Milman Parry Oxford: Clarendon Press. Philips, Kathleen. 1981. A Phonology of Baka Yaounde, Cameroon: SIL. Pinsky, Robert. 1998. The Sounds of Poetry: A Brief Guid e 1st ed. New York: Farrar, Straus and Giroux. Pli, and Daniel Boursier. 1996. Pli: Mmoire d'une femme pygme: Rcit de pli Paris: L'Harmattan. Portner, Paul. 2004. Vocatives, Topics, and Imperatives. Paper presented at IMS Workshop on Information Structure, Bad Teinach, Georgetown University, July 16, 2004. https://www9.georgetown.edu/faculty/portnerp/my_papers/Stuttgart_hand out.pdf (accessed january 30, 2011). Qureshi, Regula. 1987. "Musical Sound and Contextual Input: A Performance Model for Musical Analysis." Ethnomusicology 31 (1): 56 86. Rainey, David W. and Janet Larsen. 2002. "The Effect of Familiar Melodies on Initial Learning and Long term Memory for Unconnected Text." Music Perception Winter, Vol. 20, No. 2, 173 186. Reed, Smith Alexander. 2005. "The Musical Semiotics of Timbre in the Human Voice." Ph.D. diss., University of Pittsburgh. Renkema, J. 2004. Introduction to Discourse Studies Amste rdam; Philadelphia: John Benjamins Pub.
277 Ringer, Alexander L. 2010. "Melody." G rove Music Online Oxford Music Online http://www.oxfordmusiconline.com.lp. hscl.ufl.edu/subscriber/article/grove /music/18357 (accessed 8 Oct. 2010). Ruhlen, Merritt. 1987. A Guide to the World's Languages Stanford, Calif.: Stanford University Press. Sarno, Louis, and Bernard L. Krause. 1995. Bayaka: The Extraordinary Music of the Babenzele Pygmies and Sounds of their Forest Home Roslyn, N.Y.: Ellipsis Arts. Scheub, Harold. 1977. African Oral Narratives, Proverbs, Riddles, Poetry, and Song Bibliographies and Guides in African Studies. Boston: G. K. Hall. 1996. The Shadow s of Memory: Storytellers of South Africa Struggle with Apartheid Madison: University of Wisconsin Press. 1996. The Tongue is Fire: South African Storytellers and Apartheid Madison: University of Wisconsin Press. 1998. Story Madison, WI: Univ ersity of Wisconsin Press. 2002. The Poem in the Story: Music, Poetry, and Narrative Madison, Wis.: University of Wisconsin Press. Schrag, Brian Edward. 2005. "How Bamileke Music makers Create Culture in Cameroon." Ph.D. diss., University of Califo rnia at Los Angeles. Seeger, Anthony. 1986. "Oratory is Spoken, Myth is Told, and Song is Sung, But They Are All Music to My Ears," in Native South American Discourse edited by Joel Sherzer and Greg Urban, 59 82. New York: Mouton de Gruyter. [1987 ] 2004. Why Suy Sing: A Musical Anthropology of an Amazonian People Urbana: University of Illinois Press. Shepherd, John, 2003. "Hooks," in Continuum Encyclopedia of Popular Music of the World edited by John Shephard. London: Continuum, p. 563 564; Sher zer, Joel. 1987. A Discourse centered Approach to Language and Culture. American Anthropologist 89 (2) (Jun.): 295 309. Silcock, Lisa. 1988. Baka: People of the Rainforest Chanel 4 Television, London. Smith, Barbara Herrnstein. 1968. Poetic Closure: A Study of How Poems End Phoenix Books. Chicago: University of Chicago Press.
278 Stock, Jonathan P. J. 2003. Huju: Traditional Opera in Modern Shanghai Oxford; New York: Published for the British Academy by Oxford University Press. 2006. "Yang's Ei ght Pieces: Composing a Musical Set Piece in a Chinese Local Opera Tradition," in Analytical Studies in World Music, edited by Michael Tenzer, 275 302. Oxford University Press. Stone, Ruth M. 1988. Dried Millet Breaking: Time, Words, and Song in the W&i Ep ic of the Kpelle Bloomington: Indiana University Press. Sundberg, Johan. 1987. The Science of the Singing Voice DeKalb, Ill: Northern Illinois University Press. Thomas, Jacqueline M.C.. 1991. Encyclopedie des pygmes aka I fascicule 4: La Langue. Etu des Techniques et langage des chasseurs cuilleurs de la foret centrafricaine. Etudes Pygmees VI. Paris; SELAF. Tombombo, Dieudonn. 2007. Personal conversation/interview with author, Nomedjo, Cameroun (Baka cultural leader). Turino, Thomas. 1999. "Sig ns of Imagination, Identity, and Experience: A Peircian Semiotic Theory for Music." Ethnomusicology Vol. 43, No. 2 (Spring Summer, 1999): pp. 221 255. 2000. Nationalists, Cosmopolitans, and Popular Music in Zimbabwe Chicago Studies in Ethnomusicolo gy. Chicago: University of Chicago. Turnbull, Colin M. 1955. "Pygmy Music and Ceremonial." Man 55 (Feb.): 23 4. 1961. The Forest People New York: Simon and Schuster. and American Museum of Natural History. 1965. Wayward Servants: The Two Worl ds of the African Pygmies ed. by the Natural History Press. 1983. The Mbuti Pygmies: Change and Adaptation New York: Holt, Rinehart, and Winston. 1990. Music of the Rainforest Pygmies Vol. 1 sound disc (50 min.) : digital, stereo.; 4 3/4 in. New York, N.Y: Lyrichord. Turner, Victor Witter. 1967. The Forest of Symbols: Aspects of Ndembu Ritual Ithaca, N.Y.: Cornell University Press. Urban, Greg, and Joel Sherzer. 1988. "The Linguistic Anthropology of Native South America." Annual Review of Anthropology 17: 283 307.
279 Urban, Greg and Southwest Educational Development Laboratory. 1982. The Semiotics of Two Speech Styles in Shokleng ." Sociolinguistic working paper no. 103. Austin, Tex: Southwest Educational Development Laboratory. Valry, Paul T. S. Eliot, and Denise Bernard Folliot. 1958. The Art of Poetry Bollingen series 45: 7. New York: Pantheon Books. van Dijk, Teun A. 1997. Discourse Studies: A Multidisciplionary Introduction. Vol. 1, Discourse as Structure and Process London: Sage. Vernallis, Carol. 2004. Experiencing Music Video: Aesthetics and Cultural Context New York: Columbia University Press. Wallace, W. 1994. "Memory for Music: Effect of Melody on Recall of Text." Learning, Memory, and Cognition 20(6), 1471 1485. Waterman, Christopher Alan. 1990. Jj: A Social History and Ethnography of an African Popular Music Chicago Studies in Ethnomusicology. Chicago: University of Chicago Press. Whittall, Arnold. "Melody." The Oxford Companion to Music Edited by Alison Latham. Oxfo rd Music Online http://www.oxfordmusiconline.com.lp.hscl.ufl.edu/subscriber/article/opr/t 114/e4345 (accessed 14 Oct. 2010). Wise, Tim. 2007. "Yodel Specie s: A Typology of Falsetto Effects in Popular Music Vocal Styles," Radical Musicology 2 (2007):57 pars. 17 January 2008, http://www.radical musicology.org.uk (accessed january 30 2011). Yalch, Richard. 1991. "Memory in a Jingle Jungle: Music as a Mnemonic Device in Communicating Advertising Slogans." Journal of Applied Psychology Vol.76, No. 2, 268 275. Yung, Bell. 1989. Cantonese Opera: Performance as Creative Process Cambridge Studies in Ethnomusico logy. Cambridge Cambridgeshire; New York: Cambridge University Press.
280 BIOGRAPHICAL SKETCH Dan Fitzgerald was born in Auburn, New York in 1955. The youngest of five children, Dan grew up in western New York in the town of Lyons and graduated from Lyons C entral High School in 1973. He earned his M.M. in performance and literature in applied percussion from the Eastman School of Music in Rochester, New York in 1977 and his M.A. in missions/intercultural studies and ethnomusicology from Wheaton College, Whe aton, Illinois in 1989. During his ethnomusicology internship at Wheaton College, Dan carried out ethnographic field work among the Tewa Pueblo Indians in New Mexico, under the supervision of Dr. Vida Chenoweth. He pursued an M.A. in linguistics from the University of Texas at Arlington, Arlington, Texas from 1992 1993. Dan performed as a percussionist with the Rochester Philharmonic Orchestra from 1976 1977, with the Pasadena Symphony from 1978 1979, and as an ethnic percussionist with the Rwanda Lewis A fro Cuban Dance Company in Los Angeles in 1979. He also taught and performed percussion as a member of the faculties of music at the State University of New York at Geneseo and Brockport from 1979 1986 and 1981 1983, respectively. Dan began his ethnomusic ological field work in Cameroon, Africa, in 1995, as a member of SIL, an international Christian faith based language development NGO. From 1996 to 2008, Dan and his family intermittently lived in Cameroon's eastern rainforest among the Baka people as par t of SIL's Baka Language Development Project. Dan and his wife, Debra, and their four children, Emma, Aaron, Mary, and Jonathan, returned to the U.S. in June 2008. Currently, Dan is teaching at the
281 Graduate Institute of Applied Linguistics (GIAL) in Dal las, Texas. Upon completion of his Ph.D in May 2011, he will continue to pursue a career in teaching and ethnomusicological research.