Citation
acquisition of English by Choctaw speaking children /

Material Information

Title:
acquisition of English by Choctaw speaking children /
Creator:
Kwachka, Patricia Butler, 1942-
Copyright Date:
1981
Language:
English

Subjects

Subjects / Keywords:
Grade levels ( jstor )
Grammatical tenses ( jstor )
Language ( jstor )
Learning ( jstor )
Nouns ( jstor )
Observational learning ( jstor )
Past tense ( jstor )
Pronouns ( jstor )
Sentences ( jstor )
Verbs ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
07883119 ( OCLC )
ABS1735 ( LTUF )
0028130538 ( ALEPH )

Downloads

This item has the following downloads:


Full Text

















THE


ACQUISITION


OF ENGLISH


BY CHOCTAW


SPEAKING


CHILDREN


PATRICIA


BUTLER


KWACHKA


A DISSERTATION PRESENTED T
THE UNIVERSITY
IN PARTIAL FULFILLMENT OF
DEGREE OF DOCTOR


0


T
0


THE GRADUATE C
OF FLORIDA
HE REQUIREMENTS
F PHILOSOPHY


COUNCIL


FOR


































Copyright 1981

by


Patricia Butler


Kwachka
















ACKNOWLEDGMENTS


am indebted


to a great


many


people


their


support


during


like


course


thank


this


study.


Mississippi


Particularly,


Band


of Choctaw


would


Indians


their


cooperation,


especially


Abbie


Gibson


Kennith


York,


who


were


very


patient


with


me.


Secondly,


am ex-


tremely


grateful


to Jerry


Matthews,


without


whom


lan-


guage


data


would


have


a certain


death


bowels


Univac.


would


also


like


to acknowledge


Mississippi


State


University


grant


enabling


computer


analysis


data.


Finally,


would


like


thank


my chairperson,


Hardman,


continuing


support


extremely


helpful


criticism


this


study


draft.

















TABLE


OF CONTENTS


Page


ACKNOWLEDGMENTS


LIST


S.... .. .. .. iii


OF TABLES.


LIST


OF FIGURES


S. . 0 5 5 S


ABSTRACT


. . . . Xll


CHAPTER


ONE.


INTRODUCTION


TO THE


RESEARCH.


Introdu


warning


action


Languag


Language
Case .


Indian


Edu


cation


tion,


tiv


, and


1.6.


CHAPTER


TWO.


LANGUAGE


IN THE


CHOCTAW


COMMUNITY


Intr
Lang
The
The
The
Lang
Summ
Note


duction


and
nd
f A
f S
hoc
usi


Co:
Eng
ge
ex
taw
on


mmun
lish
and
and
Con


tyles
ing.
ition
ition


*


CHAPTER


THREE.


MORPHOLOGY.


odu
an


n .
erb In
- 1_ -i











Table


Contents,


continued.


Page


CHAPTER


FOUR.


SOME


OBSERVATIONS


ON SYNTAX.


. . 241


Introdu


action


Sentence
Sentence


Types
Errors


Notes


Success


S . 242


S . . 249
* 269


CHAPTER


FIVE.


BEYOND


DATA.


. .. . . 270


Introdu


action.


. 270


Evaluation


. Some


Implic


Research


nations


Methodology
,e Children'


. 270


Deve


lopment


. 277


Notes


. . 2


APPENDICES


App
App
App
App


endix
endix
endix
endix


One
Two


Three
Four.


SWCEL


COMPUTER
CODING
CHOCTAW


TEST


RESPONSE


SYSTEM


FORM


. 2


a . . 2


ORTHOGRAPHY


S a a a 2


BIBLIOGRAPHY


. . 301


BIOGRAPHICAL


SKETCH.


* . 314


. . . 241
















LIST


OF TABLES


Table


Page


Communities


Ranked


core


A Compari
Dominance


son


Score


Cho


ctaw


Language


Ranking


Appropriate


Ranking


of Communities


ness


sponse


". 6 2


of Communities


Response"


Ranking


Communi t


"One


Word


Respon


Ranking

Ranking


of Communities

of Communities


Sentence


Imitative


Production

Ability .


Grade


Level


Score


Plural


Suffix


The


Plural


Suffix,


Child.


The


Plural


Suffix,


Imitative


Success.


Possessive


Suffix


Poss


essive


Suffix,


Imitative


Success.


Article


Pref


ixes.


The

The


Article

Article


Prefixes,

Prefixes,


Child


Imitative


Success


Arti


Omission


Errors,


Imitative


Responses.


. -S a a a


. . . . . .*


A










List


of Tables,


continued.


Table


Page


Third


Person Singular


Present


Tense


{-s}


Third Person Singular,


The Progressive

The Progressive
Success .


Imitative


Inflection

Inflection


-ing

-ing,


Success


Imitative


The


Past


Tense


Past


Tense,


Imitative Success.


Pronoun Subjects,


Non-Third Person Singular.


Third Person Singular Subject Pronouns


Pronoun
Success.


Subjects,


Non-Third Person,


Imitative


Third Person Singular


Subj


Pronouns


Imitative


Success.


Subject Pronouns,


Object


Summary.


Pronouns.


Object


Pronouns,


Imitative


Success


Third Person Singular


Possessive


Non-Third Person Singular


Pronouns.


Possessive


Pronouns.


Third Person Singular


Imitative


Possessive Pronouns,


Success


Non-Third
Imitative


Person
Success


Singular


Possess


Pronouns


Possessive


Pronouns,


Summary


All Auxiliary Verbs.










List


of Tables,


continued.


Table


Page


"Does"


"Doesn't"


A' uw]


S. a .


"DO"


"Does"


Imitative


Success.


"Doesn't"


, Imitative


Success.


Copular


Auxiliaries


Copular


Auxiliaries,


Imitative


Success.


Third Person Singular Copular


Auxiliary


Non-Third Person


Copular


Auxiliary


Contracted


Copular


Auxiliary,


Persons


Non-Contracted


Copular


Auxiliary,


All


Persons


Third Person


Singular,


Imitative


Success.


Non-Third Person,


Imitative


Success


Contracted,


All


Persons,


Imitative


Success.


Non-Contracted,


All


Persons,


Imitative


Success.


The Auxiliary


"Will"


The Auxiliary


"Will"


Imitative


Success


Percentage


Point


Production between


Change


Grades


Successful


Spontaneous


K and


Grades


Activity


3 Spontaneous
Indices .


success


Order


English Li


and L2


Acqui


sition


Sequences


"Do"










List of Tables,


continued.


Table


Page


Sentence Types
Variables .


Ranked by Three


Response


Sentence
Success.


Sentence


Types


Success,


, Spontaneous and


Grades


Imitative


2/3.


Transitive Sentences.


Interrogatives


Copular


Sentences


Intransitive Sentences,
Negative . .


Affirmative


and


Subject

Pronoun


Omission

Subject


Summary


Omission.


* . .

. . .


Noun Subject


Omission


Subject


Omission,


Imitative


Rate.


Objects


in Transitive


Sentences


Objects
Success


Transitive


Sentneces,


Imitative


Copular


Verbs,


Omission


Rate.


Copular


Verbs,


Imitative Omission Rate.


Adjectives.


Adjectives,


Imitative


Success


Locative


Prepositions


Locative


Prepositions,


Imitative


Success.

















LIST


OF FIGURES


Figure


Page


Plural


Imitative


Suffix
Success


, Spontaneous


. . a a a a a a


The


ssessive


Suffi


, Spontaneous


Imitative


cess


. I a a a a a .


The


Article


Imitative


Pre


success


xes


, Spontaneous


and


. a a a a a a a a a a


Third


Person


Singular,


pontaneous


Imitative


success


The Progr
Imitative


essive


Infl


success


section,


pontan


eous


. a . a a a a .


Past


Tense


Spontan


eous


Imitative


success


. a a a a a S S S a a a


Subj


Pronoun


, Spontan


eous


and


Imitative


success


. a S a a a a S S a a S S a a a a a


Third


Person


Singular


Subj


ect


Pronouns


Spontan


eous


Imitative


Success


Pre-Verb


Pronouns,


Summary,


Spontan


eous


and


Imitative


success


Pronouns


, Spontaneous


success


Imitative


Possessive


Imitati


Pronouns
ve Succe,


, Summary


, Spontaneous


. a a a . a a a


Auxiliary


Suc


cess


Verbs


, Spontan


eous


Imitative


a a a a a a a S S a a a a a a S


Auxiliary


, Spontaneous


LLA &.


Imitative


"do "










List


Figures


, continued.


Figure


Page


Copular


Auxiliaries


, Spontaneous


Imitative


success


pontaneous


Spontaneous
Inflections


pontan


eous


Success


, Summary.


Imitative


Dec


lining


Imitative


success


Grade


ucces


Inflections


that


Rise


Minimall


or Plateau


in Grade


Imitative


success,


Summary


Grade


K Imitative


Ability


and


Grades


1 and


Spontan


eous


Ability


Grade


Imitative


Ability


Grades


Spontan


eous


Ability










Abstract


of a Di


sse


rotation
the Univ


Pres


ented


ersity


Graduate


Council


of Florida


Partial


Fulfillment


Degree


Doctor


he Requirements
of Philosophy


ACQUISITION


OF ENGLISH


BY CHOCTAW


SPEAKING


CHILDREN


Patricia


Butler


Kwachka


March,


1981


Chairperson


Major


Martha


Department:


Hardman-de-Bautista


Linguistics


This


study


presents


results


four-year,


cross-


sectional


examination


of the


acquisition


of English


second


language


Choctaw


speaking


children.


The


primary


purpose


research


was


to di


cover


what


features


English

school,


the

when


which


children

specific


features


acquire


features


success


are


their

appear,


fully


first

and


four

the e


produced.


years


xtent


The


secondary


purpose


was


to explore


children


s learning


strategies


in an attempt


to establish


influence


of the


first


lan-


guage


to determine


role


of non-behavioralistic


cognitive


processes.


first


cussion


chapter


learning


contains


theories


a brief


applicable


general


to language


dis-


acqui-


sition,


followed


an examination


of English


language


_







Xlll


specifically:


language


variables


sex


domains;


, age,


communicative


community


styles


residence.


third


chapter


presents


extensive


cription


ana-


lyses


development


noun


verb


inflections,


pronoun


morphology


, and


auxiliaries.


Accompanying


developmental


data


each


inflection


are


: a comparison


with


imitative


development;


a contrastive


analysis


Choctaw


English


with


regard


feature


under


dis-


cussion;


an analysis


second


language


errors.


The


chapter


concludes


with


comparisons


these


data


with


other


studies


a second


of children


language.


learning


fourth


Engli


both


chapter


as a first


examines


the


and


devel-


opment


four


sentence


types,


transitive,


intransitive,


popular,


interrogative.


The


final


chapter


includes


brief


evaluation


research


methodology


some


impli-


cations


study


English


language


acqui


sition


future.


important


finding


this


study


that


over


a four-


year


period,


there


tremendous


variability


success


rates


individual


language


features.


This


variability


so great


from


that


data.


no regular


Secondly,


sequence


study


acqui


found


sition


that


emerges


none


inflections


was


produced


criterion


level


acquis


C a*,~-


--


I r *J1 i r \f -I .- 1 l l


learninac.


Itis


1 Tan rmar~ =


+* o Q-1/4 r


4-C sy ^4


I 1







XlV


major


variables


affecting


acquisition


of English


as a second


language.
















CHAPTER


INTRODUCTION


TO THE


RESEARCH


Introduction


Despite


fact


that


everyone


learns


a language


many


learn


or more


our


knowledge


of the


cesses


which


language


is acquired


more


speculative


than


concrete.

itself tc


It is not


experimental


an area


investigation


manipulation,


and,


which


because


lends


acqui-


sition


a lengthy


process


patient


observation


recording


analysis


large


bodies


either


longi-


tudinal


or cross-sectional


data


are


necessary.


This


study


based


the


latter


approach.


The


data


base


consists


of English


language


samples


from


Choctaw


children,


aged


five


through


ten,


who


speak


Choctaw,


a Muskogean


language,


their


first


language


(LI).


discover


important


general


report


English


research


objective


relative


morphological


descriptive


sequence


syntactic


emergence


structures


in the


development


children


second


language


(L2).


mhnhro


Tn s 1r ra +-


cf 00 Fl Q 7


4-^-i c


-M QQ/^


-I rrr^


IC:


T r ~-


r i


1 n










issue


of language


fundamental


their


resolution.


Section


1.3.


and


Chapter


further


discussion


ides


their


theoretical


interest,


studi


such


are


essential


teaching


material


assessment,


, and


the


development


evaluation


programs;


s the


only


extensive


study


writer


aware


speakers


a Native


American


language,


although


there


are


numerous


studi


English


acquisition


speakers


more


prominent


languages.


Finally,


rese


arch


this


nature


important


applications


development


of a coher


ent,


convincing,


theoretical


mode


second


language


acquisition.


Issues


in this


area


will


discussed


next


section,


efly,


they


may


summarized


question


underlying


acqui


sition


rese


arch:


what


minimal


conditions--


environmental,


phy


biological,


and


psychological--must


obtain


successful


language


development


take


place?


question


will


be addr


ess


ed in


next


section,


1.2.


which


sition.


The


compares c

following


current


section,


theory


1.3.,


language

scusses


acqui


English


education


of Native


Ameri


can,


and


specifically,


Choctaw


children.


Section


1.4.


cribes


data


base


method


analysis


Section


1.5.


format


. S


- -I-- - .1-. J --; -~ -.~ -1 -I- r I- - I -- e


13,


I










1.2.


Theories


Learning


Language


Acquisition


The


like


development


development


of a scientific


language


discipline


itself.


much


historical


curve


is marked


vertical


rushes


discovery,


uneven


periods


lateral


ensuing


periods


revolution,


assessment


followed


quiescent


incorporation


new


material.


Occasionally


, hiatuses


are


created


when


pre-


viously


accepted


material


must


be rejected


at least,


re-evaluated


face


new


data.


It is


just


this


situation


that


lingui


sts


the


past


decade


found


them-


selves


they


argued,


evaluated,


and


extended


their


work


in the

theory.


light

The


then


impact


newl


fiel


introduced tr

d of language


ansformational

learning


which,


heretofore


had


been


more


subdiscipline


psy-


chology


than


linguistics,


was


extreme,


Chomsky


proposed


objective


linguistic


theory


precise


language


specification


e" (Chomsky


. 1964


. a model


acquisition


:61) .


In fact,


is difficult


to separate


roles


language


and


learning


human


behavior,


but


develop-


ment


theory


cognition


even


encompassing


more


both


demanding.


language


task


structure


complicated


fact


that


we must


use


language


in our


attempts










through

is the


which


h


language
*


that


unconscious

is considered


we understand


summation


human;


ourselves


a great


one


deal

the


language


that


defining


characteristics


our


species.


Within


species,


serves


as a primary


mechanism


differentiation


intra-group


identification.


may


that


language


preeminently


responsible


through


language


our


that


survival


transmit


as a


our


species,

cumula-


tive


understanding


of ourselves.


Language


is both


complex


mysterious.


com-


plexity


attested


the


fact


that


satisfactorily


complete


grammar


exists


any


language,


yet,


the


other


talk


upright


hand,


children


they


posture,


learn


we can


inevitably


to walk.


account


effortlessly


case


their


learn


walking


development


terms


intersection


species


specific


anatomical


features


and


ineluctable


influence


genetic


matura-


tional


schema.


But


can


we account


language


develop-


ment


the


same


way?


The


various


current


models


acquisition


take


these


factors


into


account


variably.


There


fact,


embarrassment


riches


the


area


of language


acquisition


theory.


following


discussion


will


single


out


-S


.


S- S -- 9


- f l- a 4 W a f la la V-Y lf lf f lf lf lrf


-


I I__










characterized as


a behavioral model


and


second as


cognitive model.


In their


extreme


versions,


neither


theory


is adequate,


but both


contribute


to our under-


standing


acquisition


phenomena.


1.2.1.


Two


theories


of language


learning


The


first


theory of


language


learning,


the behavioral


model,


derives


from


traditional


learning


theory


and


conse-


quently views


the development


language


result


stimulus-response-reinforcement


processes.


Important


this model


role


parent


source of


primary


input


data


and


the reinforcer of


appropriate


re-


sponses.


The


learner's


ability to


imitate


the moti-


vation


provided by


rewarding


correct


imitations


are


also


important


factors.


essence,


this


theory maintains


that


language


is a set


successfully


learned habits.


(For


further


discussion,


see


Crothers


and Suppes


1967;


Palermo


1971;


Skinner


1957;


Staats


1971,


1974.


implications


learning


are


that


learner


will


transfer


L1 habits


form


of phonological,


morphological,


syntactic


structures of


patterns)


languages


are


the L2.


similar,


Where


these


habits will


facilitate


language


learning;


where they


are










examples


see


DiPietro


1971;


Ferguson


1965;


Ohannessian and


Gage


1969)


From


are


pedagogical


that a new


point of view,


of habits must be


implications


established


learners,


their


congruent


patterns


reinforced,


their


con-


flicting


habits extinguished


accepted method


the L2


accomplishing


this


context.


is reinforcement


through repetitive oral


language drills,


an approach known


the direct


or audio-lingual


language


teaching method.


(For further


discussion,


see


Dillard


1978)


The


second


theory


of language


learning


not been


systematically


developed.


It has,


moreover,


become


quite


controversial


as a result


extreme


claims made by


some


proponents.


4Despite
Despite


the multiplicity of viewpoints


held


this


theory'


adherents,


various versions


share a


common


thread:


that


the ability


to learn


a language


innate and dependent


the operation


of inherent,


uni-


versal


cognitive


processes.


The extent


of the


innate capacity


attributed


child


divides


adherents.


Chomsky's


own


position


is open


to interpretation and has


been


subject of much


specu-


lation,


e.g. ,


Braine


1971:182-186.


One


interpretation of


his position


would ascribe


human


infant


an actual


I 1


I _










(Chomsky


1965:59).


Whatever


Chomsky's own


position,


implications


First,


this


language development


theory may be


roughly


independent


delineated.


of such


external


factors


input


quality,


reinforcement,


and motivation.


presence of


language


in a normal,


communication


setting


sufficient


to activate children's


inherent


acquisition


capacities;


they will begin


actively


exploring this


language


by testing


hypotheses


until


they


eventually


arrive


of substantiated hypotheses,


other words,


a grammar


of the


language


(Katz


1966) .


extension,


implication


learning


is that


acquisition


will


proceed in much


the


same manner,


i.e.,


a process


of matching


hypotheses


input data.


It is


important


to note


that


this model


language


learning


claims


that


the role of


learner's


L1 i


s unimportant;


that

tue o


the determining factor


f


underlying relationships


itself which,


it embodies,


by vir-


dictates


the acquisition


process.


Therefore


, the


pedagogical


implications


this model


are nebulous.


This


discussion has not argued


either theory


but


merely presented


them in


simplified


outline


orientation.


For more detailed


discussion and


debate,


the reader


referred


to Beilin


1975;


Brown


1973


Carterette










1.2.2.


The evidence


of research


examination


research provides


inconclusive evi-


dence


evaluating


validity


either theory.


Len-


neberg


(1964,


1973)


finds


support


the cognitive model


in biological


language at


data.


He cites


a relatively


fixed


universal


point


appearance of


in the developmental


process,


similar


rates of


acquisition,


the attainment


proficiency


at comparable developmental


stages.


"Per-


chronological


phenomena,

argument.


culturally,

motivating


commensurability,


could not be


And,


coincidental,


considering the


input


factors,


data,

as well


as he refers

according to


variability,


reinforcement


as gross


to these

his


cross-


practices,


variation


in indi-


vidual


intelligence,


this commensurability


indeed


difficult to account


for within


learning theory model.


Observations


learning


process


itself


provide


evidence which also


tends


to support


the cognitive


theory


language


acquisition.


First,


with


regard


input


data:


a child's


traditional

ability to


learning theory


produce grammatical


account


sentences,


for

then


that


which


the child


listens


(the model)


must itself be


grammatical.


However,


there


some real


question about


whether


or not


are in


fact


consistently


grammatical


fect


- I I I










we are grammatical


our


utterances


6 (Labov


1972:203) .


More recently,


however,


Hat'ch


(1978b:64)


finds


in her work


that


casual narratives


are


replete with errors


inconsistencies.


Jakobovits


(1970:266)


also


finds


that


fluent,


adult


speakers


are


only


"semi-grammatical"


However,


there


the


possibility that


learners


dis-


regard any


data


they


cannot


process,


that


degree of


grammaticality


the general


quality


input


data


may not be


question


fact


relevant

obviously


that


issues.

v called


comprises


Further


for,


investigation


especially


an important


view of


component


the


behavioral model.


Moreover,


is one of


aspects


language


learning which


lends


itself


to experimental


manipulation and


Other questions


is certainly


relating


readily


observable.


the role of


input


data have been


examined.


Drach


(1969)


and Snow


(1972)


have observed


that adults


regulate


their


speech


with


chil-


dren by


simplifying


syntax


shortening utterances.


There


some


evidence


that


such


telegraphic


speech


(speech


with functors omitted)


may


aid


the comprehension


month


children


Shipley


et al.


1969) .


analysis


speech


the mothers


in Brown's


longitudinal


study


of child


speech


(1970,


1973)


Pfuderer


(1969)


demonstrated


++ a4-


r'1n -. n 7 rr


nh; l ra


Sn ,-.rd n c ci


of this


f CT- r -


n -


~i n


r


zn









Another


input


factor


which


might


influence


acquisi-


tion


frequency


adult


usage


example,


specific

Bellugi


morphological

(1970:138-152)


inflections.


concluded,


Brown,


after


Cazden


examining


and

this


question


thesis


that


that


there


was


expansions


no evidence


(adult


res


to support


ponses


hypo-


telegraphic


utterances


whi


repeat


children


s productions


supply


missing


functors


acqui


sition.


fact


modeling


parent


supplying


a grammatical


response


child


s utterance


was


demonstrated


to be


more


eff


ec-


tive


both


in an experimental


situation


in the


longi-


tudinal


study.


In the


latter,


they


report


"For


three


children,

system is


order


more


emergence


strongly


within


related


the

the


child


frequency


language

with


which


inflection


mod


eled


parent


than


proportion


frequency


expansion"


1970:148).


In onl


one


child,


however,


relationship


achieve


statistical


significance.


A caveat


should


observed


here


, since


frequency


is not


only


relevant


factor.


Many


high


frequency


items,


e.g.,


"the


", do not


emerge


early


the L1


acquit


sition


sequence


that


frequency


alone


cannot


be considered


determining.


must


assumed


that


there


ncnoee 7


an interaction


rxr 7 n


n. I


between


an ~- I- -


semantic


- -


syntactic


- - --


yh rj/ Tit^ il sl\


- ^









been


established.


a variable


which


deserves


further


investigation


and


which,


moreover,


lends


itself


experimental


manipulation.


final


point


with


regard


to input


data


must


mentioned


since


poses


particular


problems


any


theory


ascribing


preeminent


importance


roles


grammatical


modeling


imitation.


every


parent


knows


, children,


from


time


they


become


inte


lligibly


verbal,


produce


maintain


substantial


periods


of time


utterances


which


deviate


so markedly


from


dialect


of the


language


that


there


could


have


been


no existing


model,


alone


a rein-


forcing


one


(Brown


and


Bellugi


1970


:90;


McNeill


1970


:105


Menyuk


1969


:124) .


This


widespread


phenomenon


must


certainly


accounted


acquis


ition


model.


In concluding


this


discussion


role


input


data,


need


further


investigation


must


stressed.


results


are


especially


pertinent


to L2


acquisition


pedagogical


"unnatural"


language


learning


contexts.


Many

guage


current


data


teaching


materials


to provide


attempt


frequent,


sequence


reinforcing


lan-


exercises


reviews


assumption


that


this


method


presen-


station


facilitates


frequently


sons


have


concerninac


learning.


little


Yet


information


curriculum


Si ran


materials

on which


develop

to base


. nntimal


ers

deci-

a n i ncr


. P -


I


UJ










One promising


area


study whose results have been


overlooked


in acquisition research


is pidginization.


The


pidginization


in compressed


process,


time


and


periods


subsequent


thus


creolization,


offer


occur


an excellent


minilab


observing rapid


language change and development


(Ferguson


1971;


Valdman


and Phillips


1975).


A coordination


these


research


findings


with acquisition


findings


should


be fruitful.


A final


factor


to be considered,


role of


imitation,


important


There


the behavioral model


little agreement among


language


researchers


learning.


concerning the


relationship


between


spontaneous


productive


capacity


imitative


capabilities.


Many


researchers


find


that


imita-


tive


ability


exceeds


productive ability.


Brown and his


colleagues


(1970)


discovered


that


children


could


imitate


sentences which


were


longer


than


their


own


sponta-


neous


productions and


that


"production


sense of


imitation


proves


to be more advanced


three


year


olds


(1970:54).


Menyuk also


found,


in a comparison


of normal


deviant-speaking


children,


that normal-speaking


chil-


dren

the


"exceed, i

grammatical


n some


instances,


competence


displayed


their repetitions,

in their utterances"


(1969:141).










a mother patiently repeats


eight


times


the grammatically


correct

patient


version of


child


son's


imitates,


utterance while


incorrectly,


after


the equally


each modeling;


child'


final


imitation maintains


own


error


and


incorporates


his mother'


correction


as well.


Age may


be a


factor


ability


imitate.


Ervin-


Tripp


(1971:197)


suggests


that


the early


stages of


language development


children


not


imitate


sentences more


successfully than


they produce them spontaneously


(with


exception


imitations


that are


expansions


their


own


utterances),


the age,


that


the role


or developmental


stage,


of imitation


of the


changes with


learner.


Not


only


age but


also


individual


selection


com-


patible


learning


strategies may


be a factor.


Bloom and


colleagues


(1974)


report


that not all


children


imitate,


but


those who do,


imitated


structure appears


their


production


shortly


after


period


imitation


There


little


information


role


imitation


situation,


despite


fact


that many


current


language


textbooks


stress


importance of


repetitive


drills.


Jakobovits


sugg


ests


that,


while


practice and


repetition


should


assist


in gaining


sensory


and motor


integration,


it would be


interesting


to determine whether


. 1 1 .1 -A I


I


I


I It










regarding


the relationship


between age and


imitation


learners.


summary,


the role


of external


factors,


particularly


that


input


data,


not


clear.


Research neither


clearly


supports


nor


rejects either


language


learning


model.


Naturalistic L1


studies,


such as


that of Brown and


colleagues,


provide


us with


our most


complete


infor-


mation,


and


yet


very nature of


setting


often


pre-


cludes


possibility


assigning


significance


to any


single variable


either


child


s linguistic or non-


linguistic


environment.


Investigators


learning


gen-


erally


conclude


that


role of


external


factors


less


interesting


dren


less


themselves who


important


seem to


than


be guided by


role of


as yet


the chil-


unspecified,


innate capacities


to create


hypotheses


test


them,


gradually


decreasing their generalizability


in order to


account


rence.


Dr increasing

(For further


restricted


discussion,


patterns of


see


occur-


Brown and Bellugi


1970:75-79:


Cazden


1972;


Dulay


and Burt


1974b:36;


Langen-


doen


1970:1-6;


McNeil


1970:104;


Moskowitz


1978.


Turning now to a


closer


consideration


learner,


it is


reasonable


to assume


that


role of


external


and non-maturational


factors


(e.g.,


aural


dis-


--I 1


I C


1


*










Lenneberg


suggests


that


a negative


effect


language


learning


abilities


produced


completion


of brain


lateralization


during


teen


years


which


inhibits


cog-


nitive


flexibility


(Lenneberg


1973


see


also


Luria


1969;


Seliger


1978).


The


nonbiological


concomitants


to be


consi-


dered


are


both


numerous


and


nebulous.


Of preeminent


impor-


tance


fact


that


older


learners,


even


though


they


may


still


be children,


will


have


already


learned


a culture,


or a large


part


and


with


this


culture


they


will


have


acquired


a set


of attitudes


, beliefs,


values


, prac-


tices,


behaviors


whi


ch will


affect


their


attitudes


their


relationships


with,


world


outside


theri


culture.


Therefore,


learning


language


which


part


of that


outside


world


may


be affected


many


non-


linguistic


factors,


including


historical


relationship


learners'


culture


culture


language


they


are


acquiring


(Fishman


1972


:21;


Lambert


et al.


1963


learners'


attitude


to the L2


culture


which


usually


, but


not


entirely


or necessarily,


a result


cultures


historical


relationship


Oller


et al.


1977)


exposure,


opportunity


, motivation,


personality


a host


even


ess


tangible


factors,


e.g.,


learners


' cultural


beliefs


rnrr^orn ne-v


1I2 nryi /r yr


2nA


Sa r,, n Cr


T^Tt2 r.t, 1


I r,


4 1


t- n


I


3









that


children


who


begin


learning


an L2


several


years


after


they


have


begun


learning


their


L1 do


not


have


to learn


what


"language"


Even


they


are,


indeed,


born


with


metatheoretical


understanding


, they


have,


time


acquisition,


gained


experiential


knowledge


of this


con-


cept.

extent


Moreover,


may


a more


sophi


their


cognitive


separated


sticated


from


level


eve


their


than


lopment,

language


that


to whatever

development,


of L1


learners.


tageous


It is


or not.


clear


Ravem


whether


concludes


these


(1974a


facts


:132


are


from


advan-


the


study


of his


Norwegian


son


acquiring


Engli


sh that


normal


six-year-old


child


at all


levels


language


greatly


facilitated


linguistic


competence


sic


already


possessed


through


first


language.


second


environment


A fully


bilingual


factor


to be considered


manner


environment


exposure

where s


learning


language.


simultaneous


acquisi-


tion


two


languages


occurs


offers


substantially


different


circumstances


a partially


learner


bilingual


than


environment


sequential


(Leopold


acqui


1937;


sition


Vildo-


mec


1963) .


Furthermore,


the


case


sequential


acqui-


sition,


one


would


expect


there


to be


significant


differences


between


exposure


in a natural


communicative


c1t-r i nr-


~II*I Ca a


i--i ni


rYn a


^ rll !


nl^p /" t^ V/^ ^


Y


I


_ n










Burt


Kiparsky


1972; Dulay


Burt


1974a).


Increased


attention


being


given


the


nature


interlanguage,


or the


systematic


approximation


target


language,


velop


ed by


learner


Selinker


1974;


Corder


1978


analy


zing


errors


erlanguage


, researchers


working


with


comparative


data


from


different


language


groups


acquiring


same


L2 hope


to determine


trans-


habit


development,


cognitive


analysis


Another


area


investigation


comparative


acqui


sitional

Natalicio


sequences


(Dulay


Natalicio


and


1971;


Burt

Hatch


1974b;

1978a)


Ravem

. Here


1974b;

, re-


searchers


compare


morphological


order


syntactic


emergence


forms


exhibited


specific


speakers


from


diff


erent


language


families


learning


same


If the


nitive


sequences


model


are


similar,


language


is argued,


learning


then


supported,


cog-


since


such


data


would


process


suggest


, i.e.


that


defines


etermines


set


sible


learning


hypotheses


If the


sequences


are


significantly


diff


erent,


then


may


argued


that


differences


are


caused


learners


bringing


L1 habits


assumptions


to bear


learning


process


The


same


arguments


apply


comparison


L1 and


acqui


sition


sequen


ces


same


language,


another


area


-r -










generalized


in order to maximize


comparison.


Other models


have been


proposed


(Braine


1971;


Richards


1978)


but,


the discussion has


pointed


out,


our


knowledge


our data


are


nowhere nearly


sufficient


to provide


us with


infor-


mation necessary


for conclusive


deci


sions.


are


parti-


cularly


constrained by


our


ignorance


the brain'


neuro-


chemical mechanisms.


Although


such information


will


cer-


tainly not answer


our


questions


concerning


language


learning,


it will more definitely


delimit


possible


boundaries


of our theoretical models.


Secondly,


past research has


focused


acquisition


information-communication


functions


language.


Language


and all


play,


the many


art,


other


evasion,


functions


status


controlled,


regulation,


somehow


acquired by the mature


speaker


are


largely uncharted areas


investigation.


(See,


however,


Peters


1977;


Scollon


1976.)


Granted


that


we have


to understand


acqul-


sition


former


function,


we must not


lose


sight


fact


that


our models must


also


allow


for the acquisi-


tion


latter,


sociolinguistic


functions.


Finally,


with reference


there are


to research.


further points


As mentioned


to be discussed


previously,


experimental manipulation


is difficult.


Not


only


are we


-- 1 -2 1


- 11 -- -- 1 l- - _


t. a -


1


I I 1 +










processes


reference


hibit


that


difficult


to language,


or distort


permanently


role


any


to define


experimentation


subjects'


damaging.


input


Thus,


factors


language


example


withholding


either


which


capacities


, we cannot


from


without


might


could


examine


an experi-


mental


group


access


question


transformation


serving

however,


if they


test


develop

; and t


one


his


their


own.


brings


our


What


sec


we can


point.


Language


testing,


as it


is currently


practiced,


still


fairly


primiti


ve.


reasons


this


that


we are


neither


certain


of what


we are


looking


nor


of how


interpret


what


we obtain.


Many


measures


in use


are


either


very


subjective


their


scoring


procedures


(e.g. ,


the


Bilingual


Syntax


Measure


, Burt


et al.


1973


very


arbitrary


in both


their


selection


tested


struc-


tures


et al.


scoring


1976).


procedures


(e.g.,


is to be hoped


SWCEL,


that


definitive


Silverman


descrip-


tions


actual


language


acquisition


sequences


and


proces-


ses


will,


future,


contribute


more


reali


stic


and


efficient


language


assessment


instruments.


The


preceding


discuss


outlined


models


research


slon


language


presents


acquis


a specific


ition.


case


The


following


acqui


sition


discus-


empha-


sizina


snn i nr~nl +r irnl


rni *rrrncsl-anrc


r.ih4 r' .1 t9


tjjh i rh'


LIIC:









1.3.


The


English


Language


Indian


Education,


the


Choctaw


case


an introduction


and


orientation


Choctaw


case


, this


section


begins


with


a general


discussion


English


language


Indian


education.


The


discussion


pro-


vides


context


presentation


the


Choctaw


case


and


also


serves


underline


the


fact


that


situation


faced


Choctaws


is not


isolated


circumstance;


many


tribes


this


country


and


Canada


are


faced


with


similar


issues


and


considerations


with


regard


the


acquisition


of English


as a


second


language


ESL).


1.3.1.


English


language


Indian


education


The


language


philosophy


country


has


been


extension


domestic,


sociopolitical


philosophy;


that


foreign


elements


should


dissolve


"melting


pot"


thereby


transmogrified


into


single


homo-


generous


Engli


sh speaking


identity.


From


Indian


point


view,


blatantly


revi


sionist


definitions


"foreign"


"native"


are


characteristic


the


entire


history


Anglo-Indian


relationships.


Despite


country


s defi-


nations


and


philosophy,


many


Indian


groups


have


resisted


loss


their


native


languages


much


more


successfully


than


they


resisted


loss


their


lands.










some


these


languages


are


dying


others


are


actually


growing


fact


that


Native


Americans


are


one


fast


est


growing


minority


groups


country


(Osborn

their 1


1970


:229)


anguages


Moreover,


will


lost


many

have


groups,

begun t


concerned


teachingg


that


their


children


Although


Native


there


language


are


second


no comprehensive


language.


or reliable


statistics


to refer


the


majority


of Native


language


L1 speakers


in this


country


also


speak


English


degrees


ranging


from


"limited"


fluent


bilingualism.


ex-


planation


this


lies


the


fact


that


most


adult


dians


attended


Bureau


Indian


Affairs


BIA)


schools


where


language


instruction


been


adamantly


English.


In fact


, one


of the fundamental


goals


these


schools


was


that


students'


"barbarous


dialects


would


blotted


English


language


substituted"


quoted


Leibowitz


1971


Use


Native


language


BIA


schools


was


until


recently


severely


punish


The


intent


BIA


was


overtly


assimilationist,


and


loss


Native


language


was,


quite


correctly,


seen


lever


in achieving


this


goal.


Considering


fact


that


students


is not


not


surprising


under


that


stand


high


language


dropout


instruction,


rates


U'* I- I= r aI a 4fi A on 4-


fir^ion


Tn^ i


CrTTTAQ


c4-n/"1b OTI+


L J I I


I


I \


Il I r r









low self-concepts


history

1970; U


have all


education


.S.C.C.R.


1975:38) .


played a


the American


Failure


part


Indian


to learn


the dismal


(Williams


fluent English13


certainly


contributed


overall


failure


and rea-


sons


for the


language


failure


are not difficult


find.


First,


relatively


reservation


little


Indians


incentive to


especially,


learn English.


there has


One


been


can


conduct


the majority


of one's


daily


life


the Native


language,


language of


the home.


English,


the


langugae


schools,


viewed


as an alien


imposition and


therefore resented


(Cornejo


1974:1)


Furthermore,


mas-


tering


English


in no way


guarantees


access


improved


economic


benefits,


since


prejudice


and


isolation


combine


to keep


Indian


lowest rungs


career


ladders.


Finally,


as Plumer


(1970:275)


observes:


If they
anyway,
will be


doing


see
then


themselves


locked


their motivation


understandably


they risk


low,


espec


cutting them


of society
learn English
ially if in


elves


from


associations


they


already


have,


namely their


peers


families.


In conclusion,


it has been


the overt


policy of


U.S.


government,


through


the BIA,


assimilate Native


Americans


depriving them of


their


languages


they will


be unable


to carry


their


cultures.


Ironically,


per-


haps


same


prejudice which


inspired


this


policy


has,









see


factors mentioned


the


preceding


dis-


cussion,


1.3.2.


and more,


operating


The Mississippi


Band


their


of Choctaw


situation.


Indians


English education


Today


there are


approximately


4,000


Choctaws


state of


Mississippi


(Spencer


et al.


1975;


histories,


see


Debo


1934


and DeRosier


1970).


Choctaw


is very much a


living


language.


families with


elementary


school-


aged


children,


83.4%


speak


Choctaw more


than


time


(York


Scott


1976).


Slightly


fewer than


7% of


households


1975) .


use English more


The remainder


than


report


Choctaw


equal


(Spencer


use of Choctaw


and English.


Language dominance


testing


children


entering


dominant


kindergarten


the Choctaw


1976


rated


language


87.6%


(Lewis


students


et al.


1977) .


Unlike many


other


Native American


groups,


the Choc-


taws


have


a long tradition


literacy


their


own


lan-


guage.


Missionaries


introduced


an orthography


the


early


19th


century


operated bilingual


schools.


1837,


there were


some


576,000


pages of works,


mostly


a religious


nature,


the Choctaw


language and


some


adults


were


literate


Choctaw


(Debo


1934:62) .


New Testament


the Bible


a hymnal


were


translated










Oklahoma,


a standardized


variant of


the missionary ortho-


graphy


been adopted by the Choctaw Bilingual Program


Nicklas


1972;


Jacob et


1977)


in Mississippi


"school"


orthography


has been


adapted for teaching purposes.


Although Choctaw


very


definitely


their


dominant


language,


most adult


Choctaws do


speak English.


Again,


this


six BIA


is more a


schools


result


exposure


the reservation


language


than


any need or


sire.


The


schools,


until recently,


operated on


the


stan-


dard BIA model:


an Anglo,


monolingual


teacher whose


struction was


translated


to the


students


by Choctaw,


bilingual


aides


classroom.


This


form of


instruction


persisted


until


fourth


or fifth


grade when


the chil-


dren'


English skills


were deemed adequate


classroom


purposes.


At no


point


, it


should be noted,


were


students


provided with English


language


instruction


predictable


results of


this


approach


to education


are


indicated by the


following


facts.


More


than


adult


Choctaws


have not


completed high school,


while


proximately


have had


less


than


three


years of


schooling


(Spencer


et al.


1975).


Those


students who


remain


in school


through


12th


grade


score


grade


level


national


achievement


tests


(Lewis


et al.


1977) .


Testing


\n7llrnnr r


rh' l rlrn


nr~ r


rU'v-S r-r'trL -F


. -


A3 QT / l Cn a r 3 4- Q (


\ / *- *


1 J I


--*


.,.4r g


_1










While


is obvious


that


a lack


of fluent


English


contributed


this


educational


shambles,


reasons


this


lack


are


entirely


obvious.


Some


are


, such


fact


that


was


never


taught


as a subject.


Students


were


expected


"pick


" Not


only


picking


up a language


inefficient


way


to learn,


also


requires


a highly


motivated


learner


who


immersed


communicative


opportunities.


These


requi


sites


simply


obtain


the


Choctaw


context.


Despite


urging


parents,


children


not


learn


English


quickly,


easily,


or well.


In fact,


many


develop


a strong


dislike


language.


basis


this


dislike


is entangled


a complex


matrix


of sociocultural


factors,


no one


factor


can


identify


as a primary


or independent


variable


affecting


English

is the


language

Choctaws'


acquit


sition.


attitude


One


toward


important


their


own


consideration


language.


them,


language


is extremely


significant


to be Choctaw


to speak


Choctaw.


eaking


language


poss


ibly,


single


most


important


aspect


one


s Choctaw


identity.


This


absolute


identification


language


social


group


membership


direct


implications


second


language


learning


can


argued


that


there


an unconscious










desirable.


One


comes


too


closely


identified


with


Anglos


Anglo


culture.


(For


discussion


other


Native


American


beliefs


regarding


language,


see


Ohannessian


1967


10-11.)


The


language

to force


Choctaw


must

the


attitude


only

childr


the


be exacerbated

en to speak En


dominant


culture


schools


glish.


Not


attempts


only


students


resist


the


language


itself,


their


culture


regards


form


overt


, direct


pressure


as highly


unacceptable.


culture


tolerates


extreme


deviance


from


expected


beha-


vior


patt


erns.


The


major


pressure


to conform


is the


viating


individual


s' expectation


collective,


community


disapproval


which,


though


rarely


expressed


directly


them,


circulated


effective


through


mechanism


considered


critical


of social


absolute


gossip

control


right


It is

Noneth


individual


a very

eless, it

s to dis-


cover


decide


themselves,


no matter


how


long


this


may


require


nor


how


flagrantly


they


may


violate


standards


interim.


grossly


improper


to directly


tell


a person


how


to behave.


Thus,


to place


individuals


the


position


of having


to do something,


schools


have


done,


creates


a great


deal


tension


and


students


respond


withdrawing


from


situation.


1~~ -


r-


- 1 *I I


* r -


__


-I 1


ql










improper

table:


their


answering


own


culture,


questions


whe


make

n the


them

v are


feel


uncomfor-


unsure


of the


response


have


not


been


given


time


to consider


themselves

allowed to


having


observe;


questions


being


single


rather


out


than


to perform


being


front


of others;


not


sharing


work


ass


ignments,


but


completing


them


individually;


being


expected


to maintain


steady


eye


contact;


high


decibel


and


movement


level


of Anglo


communicative


tribute


style;


to their


these,


unease.


Since


many

the s


other


factors,


school


con-


children


sole


exposure


to Engli


on an interactive


basis,


their


attitude


language


becomes


synonymous


with


their


school


experience;


that


, negative.


(For


comparable


findings


with


other


Native


American


students,


see


Cazden


John


1968;


Dumont


1972;


John


1972;


Ohannessian


1967


:26;


Phillips


1972


Wax


et al.


1964.


Ultimately,


incongruence


between


Anglo


teaching


styles


ctaw


learning


styles


Gibson


and


Kwachka


1978)


when


combined


with


a lack


opportunity


use


English


outside


classroom,


lack


necessity


res


ulting


from


economic


geographic


isolation;


cultural


about

Engli


language;

sh-speaking


a negative


people;


historical


produce


experience


an extremely


with


unfavor-


- 1 1


^ *


*










oral English,


it required


a Supreme Court


decision


(Lau


vs.


Nichols,


1974)


to make


possible


implementation of


1968


Bilingual


Education Act


(Title


VII)


1965


Elementary


and Secondary


Education Act


(U.S.C.C.R.


1975:180)


This


decision


declared


that


children with


limited English are


thereby


denied access


to equal


educa-


tion and


that


school


districts must


provide


for their


education.


a result


of this


decision,


monies


became


available


to minority


language groups


to institute bilin-


gual


education.


The Choctaws


began


receiving Title


funds


The Bilingual Education


for Choctaws


Mississippi


BECOME )


program was


established


reser-


vation and,


following


year,


a complementary program,


the Bilingual-Bicultural


Teacher


Education Project,


was


funded at nearby Mississippi


State University to


certify


Choctaw


students


in bilingual,


elementary


education.


programs


have


experienced mixed


success.


Chief


obstacles


implementation have been


teachers


entrenched


previous


system who are not academically prepared


teach


in bilingual


settings;


and


parents who are


afraid


that


languages


schools will


"confuse"


children


that


instructional


use of


Choctaw will hamper the


children


s acquisition


of English.


Nonetheless,


initial


-A-- .- i- - -


-C I 1- -


-1 .


Y I_


~____ ___~__


* '


,1 I









future


viability


larger


society,


a society


which


seems


to have


become,


lately,


a bit


more


accepting


cultural


pluralism.


Turning


now


testing,


next


section


will


describe


instrument


used


measure


the


linguistic


parameters


of elementary


school


children


s spoken


Engli


1.4.


The


data


base


Each


test


year


oral


November,


English


BECOME


proficiency


program


to all


administers


students


grades


Kindergarten


through


Three


test


tape-recorded


later


scoring.


It is


the


language


re-


corded


during


1977


admini


station


test


which


provides


data


this


study.


Language


at best,


testing,


a delicate


especially


proposition.


of oral


the


production


test


, is,


highly


structured,


i.e.,


there


are


specific


responses


or struc-


tures


to be elicited,


testers


cannot


certain


that


they


have


acc


urately


sampled


respondents'


productive


competence


Silverman


et al.


1976) .


On the


other


hand,


test


is more


flexible


and


allows


children


respond


freel


, thus


displaying


their


productive


range,


then


it is


difficult


to develop


and


apply


uniform


proce-


dures


score


ing.


The


test


elected


BECOME


a 7


Droaram


I


The


___


v


IJ









1.4.1.


SWCEL


test


The


tive


SWCEL


was


Educational


developed


Laboratory


Southwestern


accompany


an ESL


Coopera-


teaching


program


(Livero


n.d.) .


The


test


been


judged


appropriate,


both


technically


Chinese


, Navaho,


culturally,


Spanish


such


as well


diverse


as by


groups


Choctaws


(Locks


et al.


1978) .


test


in its


entirety


is presented


Appendix


SWCEL


composed


three


subtests.


The


first


deals


with


vocabulary,


second


with


pronunciation,


third


with


morphology


and


syntax.


The


total


instru-


ment


individually


administered,


requiring


about


fifteen


twenty


minutes


per


child.


vocabulary


subtest,


children


are


require


identify


three-dimensional


objects


labels


these


items


serve


stimuli


pronunciation


subtest.


third


subtest,


mor-


phology


syntax,


is compos


series


of pi


ctures


games,


control


to elic


a restricted


structural


responses.


The


vocabulary-pronunciation


sub-


tests


have


total


stimulus


items,


syntactic


subtest


total


items.


purp


oses


this


study


, the


phonological-lexical


portions


test


were


subjected


to analysis.


First,


playback


quality


was


freauentlyv


- -


Door


to discriminate


-










discrete


mechanical


lexical


aspect


items


perhaps,


language


acqui


a peripheral,


sition


even


uninter-


testing


1.4.


Administering


the


SWCEL


test


students


was


Grades


administered


through


in November,


the


1977,


Choctaw


to all


elemen-


tary


schools.


Native


English


speakers


the


BECOME


staff


gave


test.


Testing


took


place


schools


but,


because ac

recording,


:curate


scoring


areas


depends


isolated


quality


from


general


activities


schools.


test


Occasionally


lunchrooms


was


necessary


or auditoriums


to adminis-


and,


ese


cases


, tape


quality


is generally


rather


poor


but


sufficient


score


majority


of structures.


Because


children


s grammatical


proficiency


scored


basis


their


responses,


is extremely


important


to promote


a relaxed


atmosphere


and


encourage


children


talk


freely.


The


testers


made


every


effort


to do


those


who


were


shy,


the


obligatory


"warmup"


period


was


extended


those


with


little


English,


testers


allowed


children


take


long


they


wished


to respond.


Most


the


children


enjoyed


.8-.: n1r-.-


A- _1_ a*


n1, -,.,n


UA~l~~n~


J[ -


In ~









1.4.3.


Scoring


SWCEL


The


standard method for


first


section below


scoring the


, produces


SWCEL,


a single,


described


composite


score which does


linguistic or


allow


an evaluator to examine


sociolinguistic variables.


For this


specific


reason,


a different method,


was developed


described


for purposes


second


this


section below,


study.


1.4.3


standard


scoring method


The


standard


scoring method reflects


three measures:


the completeness of


response;


the grammaticality


the

mum


response;


points


spontaneity


is possible on


the response.

syntax portion


A maxi-


test


using


this method


scoring.


Each response


scored as


follows:


a complete,


spontaneous


response


receives


points;


a minimal


(not


complete


, spontaneous


response receives


points;


prompted


(imitated


, complete


response


also


receives


points;


a prompted


, minimal


response receives


1 point;


and no


response,


or a


response which


does not


include


desired


structure,


receives


no points


Southwestern


Cooperative Laboratory,


. 1971:4)


Prompting


takes


place


the children:


do not


res


pond


verbally,


1 a


. .. urn tin


_


heads


. Or ooint:


thir T


cshta










or d)


if they


give an


ungrammatical


response not containing


the element


in question.


The method


of prompting requires of


the children a


bit more


than


sheer


imitation


some questions.


For


example,


a child fails


to answer the question


"What do


you like?"


tester prompts


saying


"Tell me that


you


like


soup.


In order


to produce an appropriate,


grammati-


statement,


the child must make


the necessary reference


change


task is


pronouns:


primarily


like


imitative and


soup.


Nonetheless,


principal


problem with


standard scoring method is


that it does


not differen-


tiate between


such imitative


responses


spontaneous re-


sponses.


therefore


conceivable


that children


could


score 112


points on


syntax portion of


the test without


having produced a


single,


spontaneous


utterance.


Since


this


score


falls within


the range


expected for


competent


English speakers


(Lewis


et al.


1977) ,


assigning


children


to this


category who


have


only


imitated


interviewer


would


seem to contradict


assess


the children's


purpose of


ability


testing:


to generate


that is,


language.


Therefore,


second method


deriving


a score was de-


veloped


for this


research.


A -. a -


mm


n










grammaticality.


this


method,


a grammatically


correct,


spontaneous


response


any


length


receives


points


response


is a complete


sentence,


a point


added;


res


ponse


is a single


word,


a point


subtracted;


if the


response


appropriate


whether


or not


grammatical,


receives


a final


point.


maximum


pos-


sible


score


spontaneous


responses,


this


method,


points.


Imitative,


or prompted,


responses


are


scored


separately


similarly.


Because


standard


scoring


method


employed


BECOME


program


scores


are


reported


schools


assessing


student


progress,


one


first


questions


this


study


addresses


is whether


or not


method


actually

neously


indicative

generate gr


ammatica


students'

1 English


ability


to sponta-


as measured


the


scoring


method


employed


this


study.


In order


to do


this,


tional


sets


relationship


scores


. It


was


were


found


examined


that


a correla-


correlation


relatively


high


.83)


although


they


measure


dif-


ferent


variables,


both


scoresT


can


be considered


to reflect


children


Data


s linguistic


recording


competence.


analysis


nrTr., A1 n e


K2


P~nr*


1 M -P/NV


*t I-


ThC~l rJ^\ -l- -


^^n r- -


-^ v <


f~\


LJ


1









analysis.


The


size of


the data base,


some


12,000


utter-


ances


ranging


length from


single words


to paragraphs


of discourse,


necessitated


the development


special


techniques


recording


and retrieving


specific


types


acquisition


information


under


investigation.


Analyzing


such massive quantities


of data


not


typical

relying


of most


on a


linguistic


few major


studies,


consultants


traditional


linguists


transformational


linguists


on individual


(frequently their


own)


competence.


However,


in recent years both sociolinguists and


psycho-


linguists


have dealt with


large


corpuses.


Because


this


required


innovative approaches


analysis


not


pre-


sented


in most


linguistics


programs,


it is


rather


sur-


uprising


that


there


are


relatively


few published accounts


concerning


special


problems,


methodologies,


pro-


cedures


involved.


(For


exceptions,


see


old 1972


:31;


Shuy


et al.


1968.


The method


of recording


and retrieving the data


this


study


is reported here


detail


for the


following


reasons.


First


fact


that


the method


of recording


and coding


determines what


is available


analysis


therefore defines


inherent


limitations


that


analysis.


Second


usefulness


lack


L. ,


such accounts


interested researcher.


their potential


Thirdly, the










1.4.4.1.


Recording the data


Each


tape was


listened


to by theinvestigator


information


every


response recorded


exactly


the


child produced


Standard English orthography was


used


except


in those


cases


where


the


phonetic realization


pro-


vided information about the


phemic development.


stages of


facilitate


syntactic or mor-


recording process,


a response


form for


each


question was designed.


The


form


the expected,


of the page,


correct


followed by


response


a list


printed across


of the more


top


frequent


errors.


What


the child


said,


or did not


say,


was


recorded


entering


a mark beneath


appropriate word,


morpheme,


or error


column.


A system of


checks


and numbers


dis-


tinguished spontaneous


correct


from incorrect


responses,


these


from prompted


correct


incorrect answers.


Additional responses


unclassified


utterances were


recorded


in a


space


provided at


the right hand margin.


These were also


A sample response


identified as


form with


spontaneous or


explanations may


imitative.


found


Appendix


Each response,


furthermore,


was


preceded by


a number


code which


identified


student,


grade,


school,


test


question number.


The


identification


code was


followed by


i nfnrnan-4 nn Fnr


Aar4 74 nn


T f 1-a


cr'nnA 4-thnA n-F


t-h o rr ir










a spontaneous


response


was


appropriate;


whether


response


was


single


word


only;


whether


or not


child


imitated


correctly


prompted;


whether


or not


a complete


sentence


was


generated,


spontaneous


and


imitative


sentences


differentiated


this


category


use


of numbers.


Ultimately,


information


each


response


was


individually


punched


on a card,


transferred


tape,


stored


in a computer


mechanical


laborious


later

method


analysis.


of recording


This

was


highly

arrived


at after


considerable


trial


error


accompanied


increasing


appreciation


limitations


computers


instruments


recording


processing


lingui


stic


information.


Coding


data


ingenuity


which


the children


applied


lan-


guage le

creative


warning


task resulted


productions


both


suC


an immense


grammatical


varle


ungrammatical


that

or,


was


impossible


consequently,


to predict


to provide


the majority


categories


them


response


forms


which


would


allow


system


checks


to record


them.


It was t

rrnr 1i 71i flr


therefore

soir'h^ tnr


necessary


n cvntor*0a


to formulate

TO r^/M C C" ( r


a method


at"ri rTl orl


I1[ I


cate-


5.a










To accomplish


this,


general


categories


(verbs,


pro-


nouns,


etc.


were


selected


the basis of


their


relative


frequency


of inclusion


or exclusion


children


utterances.


These categories were each assigned a


dif-


ferent


pair


of numbers,


one of


pair


signifying


a cor-


rect


response,


the other


incorrect


one.


Verbs,


example,


were assigned


the general


the numbers


categories were


Several


further


specified by


pending


series of


subtopics,


or qualifiers,


which were


identified by


alphabetical


labels.


the case of verbs,


subtopics


are


present


tense,


past


tense,


future


tense,


third


person


singular present


tense,


auxiliary,


copula,


omission,


contraction.


Appendix


presents


the complete


coding


system.


By using


one or


the other


the number pair,


correct


responses


could be distinguished


from


incorrect


responses.


example,


and again with reference


to verbs,


if a child


said


present


= verb


singular;


hand,


"He big"


, the omission


tense copula

, incorrect;


= copula;


child


would be


= prese


and H


produces


third


coded as

nt tense;


person


follows:


their


= omission.


big"


singular,


1AEGH;

d person

the other


, the grammatical


response would be


indicated by the code:


2AEG;


= verb,


ohr'.- r ,~ 4 lr .r


-Fnr 4


--^ y\


f4 f" /"\










description,


there was


"other"


category for


those


pro-


ductions which defied


description.


They were


retrieved and


examined


separately.


1.5.


Data Presentation,


Objectives,


and Justification


The amount of


discrete data available from


tech-


niques


just described


the numerical


is tremendous.


information,


a great many tables and


In order to


succeeding


charts.


These are,


chapters


in all


summarize


contain


cases,


accompanied by prose


explanation and discussion.


The actual


presentation


format


follows necessarily


from


the objectives,


presented below.


1.5.1.


Research objective


general


goal


this


research is


to provide base-


line,


or descriptive data


in an area


increasing


research


interest,


second


language acqui


sition.


the basis of


the


data


collected


to sati


sfy this primary


descriptive goal,


several


corollary questions of


theoretical


interest may


addressed.


These


follow the


first,


descriptive objective


below.


The objectizes


are:


To determine


morphological


the extent


syntactic


acquisition


structures


selected


achieved by Choctaw










III.


examine


relationships


between


imitative


spontaneous


success


examine


language


learning


strategies


em-


played


children.


compare


acquisition


of English


as an L2


with


acqui


sition


of English


as an Ll.


The


first


objective


discussed


Chapters


Three


and


Four.


Chapter


Three


is concerned


primarily


with


morphology


auxiliaries,


objective


Chapter


is pursued


Four


Chapter


with


Two.


syntax.


The


The


variables


second


dis-


cus


that


chapter


are


language


domains,


communicative


style,


dence.


and


The


children


final


s sex


three


, age


ectives


and


require


community


some


of resi-


intro-


ductory


comments


explaining


their


relationship


the


descriptive


data


on which


they


are


based.


With


regard


to Objective


III,


it is


important


to note


that


imitation


an important


component


behavioral


model


language


to a certain


extent,


learning.


with


Thus


Objective


, Objective


However,


overlaps


one


problems


dealing


with


imitative


data


dual


interpretations


attributed


the


ability


to successfully


imitate.


Some,


particularly


behaviorists,


see


imitation


as a


tool


creating


underlying


competence,


1.e.,


as a


1 Parn i nc


strats av


which


assi


development


habits.


... .--


V.










production.


beyond


scope


this


study


to deter-


mine


however,


imitation

we can d


being


determine


employed


there


as a learning strategy;

is a relationship be-


tween


successful


imitation


successful


spontaneous


pro-


duction.


In order


to accomplish


this,


conclusions


of Chap-


ter s


Three


Four


will


compare


developmental


success


rates


imitative


spontaneous


oral


productions.


These


comparisons


will


be interpreted


basis


fol-


lowing,


very


tentative


assumptions.


First,


we will


assume


that


ability


taneously


produce


imitate


an inflection,


excee


and


ability


if this


spon-


relationship


remains


constant


across


grade


levels


while


both


abilities


increase,


then


we can


conclude


that


imitative


ability


an indicator


of underlying


competence.


other


hand


, imitative


ability


consistently


exceeds


spontaneous


ability


latter


does


not


increase


over


time,


then


we can


conclude


either


that


two


abi-


lities


are


related,


that,


they


are,


imitation


not


successful


learning


strategy


structure


question.


Finally,


we might


also


reach


s sec


conclusion


spontaneous


ability


excee


imitative


ability,


any


1- I 1 t -z -


-^.


11*


* t *


*1 1


r










conclusive.


chief


objective


simply


examine


the


data


for


the


existence


a systematic


relationship


between


two


productive


modes


children


s L2.


In the


most


investigation


accessible


learning


means


strategies,


children


Objective


s emerging


can


be described


as a system,


though


one


under


stood


to be


only


heuristically


stasis,


consistencies


discontinuities


evaluated


evidence


the


strategies


organic


proach


fact,


zing


not


data,


principles


within


as we will


that


scope


see


produced


this


, do not


them.


research


permit


This


and,


of a uni-


fying


grammar


or grammars.


Alternatively,


the


emerging


can


regarded


approximation


target


language


, with


errors


pro-


duced


providing


evidence


the


strategies


being


employed


to close


the


stance


between


the


approx


imation


the


target.


These


errors


may


result


from


L1 interference,


established

or from ove


contrastive


generalizations


analysis


hypothes


the L1 an

es based


on learner


analysis


of L2 data.


The


two


types


error,


thus


, imply


different


learning


strategies


different


theories


cognitive


organization


operation.


Chapters


Three


and


Four


contain


sections


dealing


with


the


analysis


errors


- Ut I *. .-


1


J


A


II ,- .- .* _


-i










with


those of


learning


ESL.


speakers


from different


We would expect


language backgrounds


the rates or


sequences


similar


the L2


inherent cognitive


capacities


which determine

variety of diss


the L1


learning


strategies.


similar patterns


that dominates


to emer


in structuring


We woul

ge if,


expect


however,


acquisition


stra-


tegies.


Therefore,


acquisition sequences and success


rates


will be compared,


where data


exist,


the conclusions of


Chapters


Three and Four.


Finally,


with reference


to Objective


a comparison


of L1 and L2


acqui


sition


same


language allows


explore


the question


of whether or not L2


learning


differs


significantly


from L1


learning.


Again,


there are


implica-


tions


learning


theory,


they are exactly the


same


those outlined in


also be discussed in


preceding paragraph.


the chapter


These will


conclusions.


1.5.2.


Data


presentation


format


Summarizing the


sociolinguistic


above discussion,


factors


Three and Four present


Chapter Two


learning process;


spontaneous and


imitative


presents


Chapters


production


success


rates,


followed by


error


analyses and


their


impli-


cations


learning


strategies


and models,


concluding with


comparative acquisition


smenunces and


summaries:


and


ChaD-











1.5.3.


Justification


research


objectives


In view


interest


in and


importance


of under-


standing


process


descriptive


ses


studi


language


have


great


acquis


value


ition,


their


both


implications


theory

absence


pedagogy.


of data,


but


Neither

we have


ese


relatively


can


develop


little


available


that


which


we have


comes


, primarily,


from


ESL


learners


from


world


s major


languages.


Data


from


well-


known


languages


are


espe


cial


value


that


they


may


pro-


vide


answers


to old


questions,


may


even


make


possible


new


interesting


ones,


expose


previous


conclu-


sons


either


to validation


or to reevaluation.


Although


this


research


breaks


no new


theoretical


ground,


it presents


anomali


inexplicable


framework


any


current,


single


theory


suggesting


that


we are


still


rather


removed


from


even


level of


theoretical


adequacy,


descriptive


capacity.


language


pedagogy


is equally


inadequate.


The


generally


accepted


explanation


the


lack


success


classroom


learning


that


a second


language


largely


irrelevant


many


learner


yet


, the


same


charge


irrelevance


can


be leveled


at much


of what


success


fully


learned


classroom


, so we must


look


ess


sim-


- a


*


*


* J


^^


I


I


*1 I










1.6.


Notes


Carol


Chomsky


the general belief


their


tactic


are


lingui


structures


years


(1969) d
that by


stic


systems


demonstrates


that,


contrary to


time children reach


are


fully


may not be mastered
age or older.


developed,


until


the


school


some


syn-


children


It i
language,
acquired


s possible
e.g., som
through ex


that


some


e English
tended li


extremely


gapping
teracy.


elaborate


rules


, may on


levels of
ly be


descriptions,


Haas


1941,


1971;


see


Nicklas


Brinton


1972,


1870;


1975;


Byington


Powell


1915;


1966.


The acquisition


a variety of


(Ravem 1974a,


of English as


other


1974b),


licio and Natalicio


Hatch


(1978a)


an L2


language groups,


Spanish


1971)


(Dulay


, Japanese


has been
e.g., No


and Burt


(Milon


examined
rwegian


1973;


1972)


Nata-


see


summaries.


Dulay


posure


Burt


to L2 data


to guarantee


or less


(1973) ,


example,


in a naturalistic


learning


irrelevant


that


propose


setting


teaching


learning


that


ex-


sufficient


therefore,


more


situation.


Bricker


debates


own


and Bricker


"pseudo-issues


theoretical position


argument which is


rather


than


defines


that


(1974:437-441)


refer to


, and one need not accept


agree with
resultant f


intent of


riction


various


their


their


obscures


issues.


6. In fact,
that ordinary


Labov


speech


lays


part


the blame


is ungrammatical


the belief


on linguists whose


conclusions may


other


converse at


have been


reached after


conferences


Braine


stening to
(1971:170)


each


further


discussion.


Hardman


(personal


communication)


pointed


out


to me


that her


extensive


text


collection


"shows,


in casual


recordings,
errors, or


virtually


mind-changes


complete grammaticality with


clearly marked by


lapsus,


intonation or


other


linguistic/paralinguistic marks.


Brown and Hanlon


adult


usage may


(1970:205)


be reflected


note


that while


the child's


frequency
production,


V -.I- I' -4- 4-. --a ,- ~ 9 J 4 L -J-9 9 -~- ...


1


-L-'1.. 1..1I--


L


A


^*J


rk









adults,


example


however,


Ervin


-Tripp


this


may


1971:190-


not


case.


191) ;


Sachs


See,


(1976:146).


Martin


child


several


languages


aker


which


there


are


examp 1


Native


New


York


Ameri


can


state


, the


Language
(Mithun a


Menominee


Project
d Chafe


sons


1974)
1979)


sonss
Mohawk


Bauer


(1971


summarizes


policy.


exceptions


this


13. The
stigmati
Harvey n


Engl


zed,
.d.;


ish
nons


that


was


standard


Leachman


and


earned


dial


Hall


gene


rally


features


1955)


contained


(Dubin


thus


1970


maintained,


even


contributed


ents


who


employment
imilation t


so insistent


speak


only


negative


themselves


difficulties


o be


that


inevitable


their


Engli


ster


experience


and
e a


who,


re


children


them


eotyp


es.


d failure


were


extreme


in some


earn


The eff


English


school


taught


cases
they


that


on the children


more harmful
a limited sp
deviant gram
ment in Choc
their peers


L


guistic


the
mann
who,
have


than


eaker


matical


taw
who


competence


notion
1966;
as a


helpful;
is also


the English


limited


d and


constructions.


is likewi


hamp


, as children,


of the


cognitive


Bernstein


res


difficulty


mature


deficiti


1964) ,
the just


communic


eating


children


usually


The


ere


not
spe


encies


ere


mar


children


their
control


er.


" (Ber


still


cribe


through


model
the
while
iter


remain


d language


lingui


stic


ked by
s develop-
s being


full
dismi
and E


lin-
ssing
ngel-


individuals
environment


mediums


average


Language
freshman


Profi


score


cienc


on the Mi


one


chigan


group


was


Test


ctaw


Engli
collea


16.
the
tion


eral


majority


problem,


upper


ctaw
their


ne require
obviously


evel


freshman
avoided


coll


eve


stud


ents


degree


d English


since


courses


composition


long


as pos


were


found


requirements


courses


eir
was


to have


with


Ability


Engli


courses


certainly
SEngli


complete


the
was


suffi
suffic


was


excep-


not


cient


the
for
for


simply


sible


rson


who


rac


iall


Cho


ctaw


does


not


earn


from


not










poi
sha
of
fro
wit
for
adu


wneac
h su
Ang
its.


f view
, in f
ateria
on by
spicio
los bu


, this
act an
1, int
the c
n and
t also


tomizes
cumulati
actual, o
re. One
term is
admonish


Angl
on t
r sp
who
used
chi


o cha
o exc
iritu
is n
not
idren


ter.
(wh
natu
l1o
y as
d to


the
e)
sr
ap
cha


s
garded
jorative
tise


Amern
gener
howev
when
tion
and C


e term
n commu
purpose
, disti
hiding
hattak
ctaws.


"Anglo
cities
es: I
nguish
the pe
('men


" also
simple
ndians
four:
jorati
') and


des
y di
and
wh
ve) ;
los


erves
sting
non-
ites
blac
sa ('


comment
ish two
ndians.
whom th
s (tako
lack');


Many
groups
The Ch
y call
sa, a c
other I


Nativ
for
octaw
Anglo
ontra
ndian


19. The
identify
become i
ceptable
attempt
There is
allowed
attitude
estingly
frequent
got our
thing le


obverse


ca
de

by
o
to
t

ly
la
ft


tion
ntif
give
Ang
nly,
lea
owar
by a
exp
nd;
."


s
d
t
s
e
L,
t
t
s
i


implic
that
with
he Cho
to le
local
theC
heir 1
her Mu
sed in


ation
if An
the C
ctaw
arn C
Angli
hocta
angua
skoge
vari


get


our


the


s s
taw
itu
taw
ho
ang
(an
gro
s o


language


and


peak
s.
de t


s stro
s lear
ge. T
ttitud
, the
the st


language


we won


cul
hen
iis
OS,
isco
r be
tect


ged.


, Inter-
ee), is
"They


t have


any-


20. Only in the past
ministering physical
the classrooms or on


decade have the schools
punishment for speaking
the school grounds.


stopped
Choctaw


ad-
in


26 vocabulary


ater
sma
ith
eight
uage
o th


items
number
'cified
minim
quisit
writer


items


in any
might
semant
ally us
ion has
's know


could


ultu
suf
don
ul,
een
dge.


hardly


, altho
cient t
ns. A
though
either s


encompass
h if chos
show acqu
dified Sw
s relevan
acested no


relevant


ca
nta
esh
to
exp


fully,
e
ist
an-
red,


The writer, w
BECOM program,
previous year'


rho at the time
participated
s testing.


was the ESL Speciali
in the administration


23. One child, on seeing me
l __ C a"*-* -^ ._ l l 1 --


in the


reservation


store not


thn


I










Inappropriate


irrelevant and


hend the q
Choctaw ra
beled so t
the degree
and 1 chil
singly low
diglossic
Situations


uesti
their
hey c
of C
dren,


clea
on,
than
would
hoct
the


The exp
situation
involving


responses


ly indica
nd those,
English.
be examine
w language
rate of C
lanation
which obt
Anglos d


included both


ted the
correct
The lat
ed separ
e domina
hoctaw r
for this
ains on


child


or
ter
ate
nce
esp
pr
the


emand Engli


did not


incorr
were f
ly. Co
among
onse wa
obably
reserve
sh or s


compre-


t, i
their
ider
ades
surp
es i
ion.


la-
ing
K
ri-
n the


ilence.


mumbl
their
the r
sound
Indiv
this


Responses that were inaudible
ed or otherwise consciously a
utterances were also scored
response was inaudible due to
s, etc.) the response was omi
idual scores were adjusted to
circumstance.


because the
nd purposeful
as inappropri
other factors
tted from the
avoid penali


children
ly obscured
ate. If
(competing
corpus.
zation in


As noted by


Shuy


et al.


linguists have formulated
than group performance."


(1968 :v) ,


theory


from


"Historically,


individual


rather


The
sissip
cost
write
the Un
that o
Coord
time


computer-


rammel
i Stat
or use


Compute


niver
the
office
Grat
to Ge
Compu
ce in


U
of
O


diversity.
office and
inator of
and patient


ese


r


1


sity
compu
for
eful
rald
ter S
tran


enc
wer
ter
Res
ack
Mat
tud
sla


who programmed


e Center facilities at
e utilized. Funds to cover
were generously granted to
earch and Graduate Studies
nowledgment is extended
thews, Associate Professor
ies, who gave freely of
ting linguistics-ese to
this entire study.


28. One child, struggling with
could not remember "under" to d
the toy pig used in testing. S
grammatical honor by producing:


locative prepositions,
describe the location of
he extricated herself with
"My hand is over the pig.


those


that were
















CHAPTER


TWO


LANGUAGE


IN THE


CHOCTAW


COMMUNITY


2.1.


Introduction


This


chapter


examines


important


sociolinguistic


fac-


tors


Choctaw


discussion


emphasized


culture.


previous


significance


of the


chapter


Choctaw


language


to Choctaw


identity,


culture


s negative


atti-


tude


toward


Anglos


and,


because


synonymity


lan-


guage


tive

This


culture


attitude

chapter


factors


Choctaw


toward

begins


that


thought,


language


discu

rigid


ssing a

division


resultant


Anglos,


corollary


of language


nega-


English.


these

domains,


or diglossia,


Cho


ctaw


life;


included


this


discussion


an examination


manner


which


communicative


styles


reinforce


domain


distinctions.


following


sections


of the


chapter


examine


three


sociolingui


stic


variables


, community


res


idence,


age,


and


sex


the children,


order


to determine


there


relationship


between


these


Engli


sh L2


acquis


ition.










individual


cultural


groups


may


not


themselves


recognize


this


link.


Choctaw,


as we have


seen


, are


an exception.


They

the


are


intensely


reservation,


they


language-conscious


maintain


people,


an absolute


and,


separation


English


Choctaw


language


domains.


The di


chotomy


very

nates


rigid

, but


and r

there


arely

are


violated

specific


The


contexts


hoctaw la

in which


nguage


domi-


use


English


three


The


deemed


major


most


not


English


visible


only


appropriate


contexts


context


are


but


obligatory.


described


tribal


council


below.


meeting


These


are


conducted


along


formal,


parliamentary


lines


entirely


in English


This


despite


fact


that


members,


majority


those


attending,


speak


Choctaw


at least


probably


more


, profi


ciently


than


they


English.


Very


occasional


an older


person


will


address


council


Choctaw,


such


occurrences


are


exceed-


ingly


rare.


English


also


used


almost


exclusively


at school


functions


and


when


interacting


with


school


personnel.


Some


Anglo


teachers


who


have


spent


many


years


Choctaw


schools


are


known


to partially


understand


language


but,


despite


this


, they


are


never


addressed,


either


childr


en or adults, in Choctaw.


r ,









even


impossible,


address


learner


casually


Choctaw,


invariably


respond


English


the


learner


addresses


them


Choctaw.


final


context


which


English


always


used


athletic


events


, a major


reservation


activity.


Announ


ce-


ments


play


descriptions,


even


Choctaw


tradi-


tional

Players


sport,


stickball,


' disputes


with


are


always


referees


are


made

also


English.


, initially,


Engli


, but


altercation


becomes


involved,


are


continued


referee


Choctaw.


s decision


Nonetheless,


again,


announcement


English.


On examining


contexts


which


English


used


clear


that


they


exhibit


common


features


they


volve


speech


a public


nature,


and/or


they


are


imported


institutions.


this


There


generalization


is,

and


however


one


that


surprising

church. C


exception


hoctaw


ministers


conduct


services


in Choctaw,


hymns


are


sung


Choctaw,


Bible


read


that


language.


Even


cases


where


minister


is non-Choctaw,


hymns


prayers


are


offered


in the


congregation


s first


language.


explanation


this


may


manner


which


Chri


stianity


was


first


introduced


Choctaw.


ssionaries


developed


a Choctaw


orthography


conducted


bilingual


schools.


Later


, many


Choctaws


themselves


became


themselves


became










some


communities


as an integral


part


of their


culture


and


an alien


institution.


event,


the


Choctaw


lan-


guage


traditionally


currently,


appropriate


religious


contexts.


Other


cult


domains


to define.


contexts


play,


just


of Choctaw


Choctaw


describe


at work.


language


spoken


The


Because


use


are


everywhere


language


most


Choctaws


diffi-


but


is spoken


live


those


at home,


work


reservation,


therefore


case


that


occasions


adults


use


Engli


are


relatively


infrequent.


Thus,


inability


to speak


fluent


Engli


sh is


not


particularly


disadvantageous,


but


the


inability


to speak


Choctaw


extremely


limiting.


inability


to speak


Choctaw


threatens


not


only


individual


identity


as previously


noted


, but


community


membership


as well,


since


each


Choctaw


community


defining,


idiosyncratic


speech


patterns.


Community


member-


ship


station


important.


which


Each


functions


community


to orient


traditional


expectations


repu-


one


introduced


unknown


person.


5 Not


only


there


alle-


giance

commun i


to community,

ty's speech w


there


whichh


is also

regard


allegiance


as superior


that

to others


whose


forms


are,


more


less


good-naturedly,


criti


cized


ridicule


__


__











Choctaw


enjoy


a well


argued


verbal


presentation


whether


or not


they


agree


with


speaker


s position.


One


ability


use


language


well,


however


, does


result


that


person


s being


regarded


as either


an authority


an arbiter,


since


"experts "


any


variety


are


routinely


despised.


The


Choctaw


function


principle


that


single


individual


embodied s


totality


or range


lan-


guage


impli


cit


community


as a whole.


Thus,


community


, not


individual,


that


dominates


cul-


ture


s social


psychological


orientation.


Just


some


speech


admired,


some


speech


is de-


plored


high


, primarily


school


teenagers.


"Pearl

6 This


River


slang"


speech


spoken


marked


a great


deal


of contraction,


slang


expressions,


and


a high


percen-


tage


English


loan


words.


Criticism


this


speech


a quite


different


tenor


than


criticism


arls


from


traditional


community


dialect


rivalry.


Indeed,


there


some


concern


among


older


people


that


current


generation


teenagers


will


not


return


to adult


norms


"degenera-


tion"


language


will


occur.


Besides


being


aware


language


domains


levels


speech,


relate


Choctaws


to speech


also


events.


recognize


The


rules


social


"rules"


Choctaw


which


conversa-


I -- J _I


..


I I __---


.. II


r"


*,


I


L


L


I










compared


pause


Cs.


with


lengths


Teasing,


Anglo,


characterized


in conversational


joking,


turns,


sophisticated,


volume,


different


often


long

kine-


risque


allusions


mark


conversation.


Humor


is much


admired


and


appropriate


to almost


every


conversational


topic.


Although


intercourse,


conversation


a major


is not,


case


feature


with


social


Anglos,


necessary


concomitant.


For


example


siting,


a regular


social


shed


activity,


that


may


a person


be silent.


After


s arrival


has been


social


rather


estab-


than


personal


reasons,


visitor


may


quietly


or join


household


activities.


Visitors


not


expect


to be


"enter-


trained"


There


converts


no demand


action


placed


a cessation


on either


of normal


sitor


routines.


or the


with


single


exception


food.


Visitors


arriving


during


meals


are


expected


to eat.


Silence


has


many


communicative


levels.


That


a visi-


s silence


social


can


detected


kinesics.


social


silence


marked


relaxed


body


posture;


eye


move-


ments


follow


whatever


activities


may


ongoing.


thought-


ful,


cons


idering


silence,


e.g.,


while


formulating


a re-


sponse,


is communicated


through


a more


rigid


body


posture


eyes


focused


on some


point


the


surrounding


environ-


Tmrnft -


I n rr vr


r noTrn'lIc


c i 1Pn


Sc ; rT-rvmrran i or


d


f= I


I I


I I \


CI










Kinesics accompanied by


silence are also


used


to com-


municate directions;


turns


a road,


or the


location


an object are


indicated by


a slight movement of


the head


or hand with no verbal


elaboration.


very


different


characteristics of Anglo


social


communicative


patterns make Choctaws


uncomfortable.


first difference


fact


that an Anglo


social


context


is obligatorily marked by


that


conversation,


to Choctaw participants,


and by


rapid


conversation


to allow re-


sponses.


volume


level


is very


loud


the Choctaw ear


swiftness and


compulsive exchange of


conversational


turns contributes


to an overall


impression of


incessant


noise.


only


silence not allowed


in Anglo conversation,


interpreted by them as withdrawal


or unfriendliness.


Moreover,


use of


contact by


Anglos makes


the Choc-


taws


feel as


they


are being


stared


i.e.,


they


are


isolated as objects of


curiosity rather than included


as social


co-participants.


Even


polite,


conventional


opening


gambits of


Anglo conversation immediately violate several


Choctaw


practices:


one does


utter


one's own name;


nor


does


one question another


about matters which are both


personal


TF :M -1/ a fl tn4- na a nar' rm4 1 Tr 4- '


i rrv hvttlI nC "iy^^


+-/\ //" n-an


=a r^,,










individual


after


a period


verbal


discussion


alterna-


tives;


whereas


a group


of Choctaws


remains


a group


only


as long


as each


individual


voluntarily,


without


discussion,


choo


ses


to continue


interaction.


group


member


may


withdraw


tion.


without


There


announcing


is no coercian,


either


intention


especially


not


or explana-


verbal,


follow


the


departing


member.


See


Witherspoon


(1977


:83)


similar


patterns


among


Navaho.


Finally


when


engaged


in conversation


with


Anglos


further


discomfort


results


from


a general


awareness


that


Engli


spoken


reservation


nonstandard


and,


because


Choctaw


people


expect


criticism


nonstandard


behavior


their


own


community


, they


project


this


anti


cipation


Anglo


communities,


and


are


therefore


re-


luctant


talk.


sum,


Choctaws


participate


minimally


English


language


varying


conversations


levels


, and,


of discomfort.


when


The


they


divergences


experience


social


practices


treme.


related


They


to language


intensify


use


trans


are,


cend


cumulatively


simple,


ex-


sheerly


linguistic


problems


appropriate


language


selection


basic,


grammatical


communication.


-n r^Tn i i nu


Fnrr 4i


ch T,.


T rn I na


0 rh


rl I I I









having


a population


around


1,200


persons


(Spencer


1975) .


They


are


scattered


over


four northeastern Mis-


sissippi


counties,


the central


largest


community


located


in Neshoba


County.


All


the communities


consi-


dered in


this


study


are within


40 miles


of Pearl


River,


the central


community


and seat


tribal


government.


Each of


the communities


not only


own


particular


speech patterns


own history


and reputation,


and relationship with


as mentioned above,


surrounding


but also


Anglo com-


ture.


Since


is conceivable


that any


or all


these


might affect L2


acquisition,


community


residence has been


selected


as a variable


examination


this


study.


Unfortunately,


there


no precise data


on either


dialect


or historical


differences.


With reference


to dia=


lect


differences,


it is


believed


that


originally there were


three major


dialect


groupings


corresponding


to political


regions:


the Ahi Apat


the Okla


Okla


Falaya


or Potato


or Longtown


Eaters


dialect


dialect


the west;


east;


the Okla Hannali


or Sixtown


dialect


in the


south


(Jacob


et al.


1977:49).


These


three may


still


exist


in both


Okla-


homa and Mississippi


(Nicklas


1972


substantiation


this


point awaits


further


research.


There


is no question but


that


communities


exhibit


dialect


differences.


but


the differpncnp may


hb nnlv


Cllnnor-










than


others,


there


less


contraction


of suffixes;


and


some


communities


contract


or drop


suffixes


more


than


others.


Contraction


may


or may


be a superficial


phenome-


non.


In casual


speech,


a widespread


practice.


How-


ever,


there


evidence


that,


at least


some


speakers,


recovery


noncontracted,


or even


intermediate


forms


is not


possible,


suggesting


that


several


syntactic-


semantic


relationships


are


unavailable


these


speakers


grammar.


Whether


or not


such


a deep


structure


difference


might


present


different


learning


configurations


depends


degree


influence


attributable


the


learning.


With


no more


than


minimal


and


somewhat


impressionistic


data


on dialect


differences,


impossible


to state


that


language


one


community


s sufficiently


different


affect


acquisition,


assuming


influence.


question


examine


must


approached


L2 data


terms


indirectly;


that


community


res


we can


idence


determine


there


a consistent


interaction


between


community


and


English


acquisition.


Such


intera


action


might


also


be produced


torical


contemporary


relationship


a community


with


surrounding


Ana lo


culture -


Aarin -


we are


1 i mi -pt


1I.


I 1 \









maintenance


traditional


customs.


The


introduction


of churches


and


schools


was


strongly


resisted


community


members.


Tucker


community,


other


hand,


is regarded


as more


"Anglo"


than


other


communities.


People


believe


this


was


to be


established


there


fact


that


in 1883


strong


Catholic


remained.


mission


was


the


only


church


which


both


Anglos


and


Choctaws


attended


(al-


though


seating


was


divided)


was


staffed


Anglos


Peterson


1970


:183


-187).


Finall

various


Pearl


some


prominent


River,


communities

families,


example


are

and


are


recognized as th

others, Standing


linked


seats

Pine


generations


of kinship


ties.


sum,


there


are


a variety


factors


related


community


While


rank


data


residence


beyond


these


factors,


to discover


presence


which


might


scope


this


is within


affect


acquisition.


research


capac


of a relation


ship


isolate


of these


between


community


SWCEL.


of residence


results


and


this


several


variables


investigation


are


tested


presented


next


section.


acquisition


communities


communities


are


compared


basis


para-


4-










confidence.


results


these


analysis


are


reported


below.


Ranking


communities


SWCEL


test


performance


communities


are


ranked


Table


below


basis


revised


score


(Score


see


Section


method


of derivation).


The


actual


number


table


a mean


score


based


on all


children


aces


each


community.


This


provides


a general


measure


comparison


among


communities.


Table


Communities


Ranked


Score


Rank Score Community


1 56.30 Bogue Chitto

2 62.53 Pearl River

3 67.30 Conehatta

4 68.17 Standing Pine

5 69.34 Red Water

6 95.32 Tucker


Although


there


are


differences


among


coImmUu-


- ._ I _


. .. .. --


m- .1 --


* -- I 4 -


r- I n v 0 n y fl^ fl rnrn n r *v I* C L1 r if I i 4 I


* _










To a certain


community


Table


Table


extent


s dominance


, below,


Score


Choctaw


compares


A Comparison
Dominance


degree


language


, ranked


of Score


Choctaw


each


correlate.


Score


Language


Community Score I % L1 Dominance


Bogue Chitto 56.30 91

Pearl River 62.53 71

Conehatta 67.30 87

Standing Pine 68.17 91

Red Water 69.34 82

Tucker 95.32 73


However,


correlation


not


consistent


that


degree


of L1 dominance


cannot


employed


as a gauge


of L2 profi-


ciency.


.1.2.


Ranking


of communities


"Appropriate


Res


ponse"


This


measure


provides


indication


children


comprehension


of English


ignores


grammaticality


their


response.


other


words


, a measure


pas-









development


that


can


extrapolated


from


testing


data.


score


in Table


represents


mean


number


times


each


child,


community,


respond


ed appropriately.


maximum


number


times


possible


Table


Ranking
ness"


of Communities


"Response


Appropriate-


Rank Score Community


1 24.56 Bogue Chitto

2 26.38 Pearl River

3 27.50 Red Water

4 28.54 Standing Pine

5 28.81 Conehatta

6 35.97 Tucker


Tucker


community,


again,


significantly


different


from


others.


we compare


communities


bases


both


(Score


appropriateness


find


grammatical


similarities


oral


lowest


production


highest


extremes


ranks


Bogue


Chitto


and


Pearl


River


chil-


dren


neither


respond


very


grammatically


nor


comprehend


well;


Tucker


children,


other


hand,


perform


well


in both










Pine


children


are


more


less


average


in both


areas;


Water


children


not


comprehend


well


respond


gram-


matically


when


they


.1.3.


Ranking


of communities


Response"


Another


prehension


measure


whether


least


or not


partially


children


indicative


attempt


com-


to respond


test


item.


This


measure


is not


so reliable


the


previous


one,


since


a variety


factors


may


enter


into


children


s deci


sion,


including


shyness,


degree


previous


exposure


to Anglos,


illness,


forth.


Table


below,


did


reports


respond


mean


at all,


number


even


after


items


to which


prompting.


each


The


child


maximum


again,


Table


Ranking


of Communities


Response"


Rank Score Community


1 5.99 Tucker

2 10.80 Red Water

3 10.84 Standing Pine

4 11.37 Conehatta

5 13.32 Pearl River


'










communities


at either


scale,


Tucker


Bogue


Chitto,


are


significantly


different


from


the


others,


Tucker


having


lowest


number


ponse"


Bogue


Chitto


having


highest


number


child


per


test.


.1.4.


Ranking


of communities


"One


Word


Response"


Children


may


res


pond


with


single


words


several


reasons.


minimal


some


(viz.


cases


single


Question


word


secondly,


adequate,


their


ability


use


English


may


sufficient


res


pond


more


fully;


they


may


be uncomfortable


using


language.


There-


fore,


it is


impossible


to determine


exactly


what


this


mea-


sure


in the


reflects.


children


It is,


s L2.


nonetheless,


score


an index


Table


"volubility"


below,


mean


number


time s


"One


Word


Response"


was


given


child,


community.


Table


Ranking


of Communities


"One


Word


Response"


Rank Score Community


1 15.07 Bogue Chitto

2 16.12 Pearl River










Here


we find


that


Bogue


Chitto


children


are


most


voluble


Standing


Pine


children


most


reticent.


Since


Bogue


Chitto


children


are


also


least


likely


respond


at all,


least


likely


to respond


either


appropriately


or grammatically


when


they


do respond,


this


is a rather


been


surprising


significantly


finding.


different


Tucker

other m


children


measures


who


are


have


more


less


average


this


one.


There


is no obvious


explanation


these


findings,


but


apparently


not


case,


one


might


assume,


that


more


proficient


speakers


talk


more.


.1.5.


Ranking


of communities


sentence


production


Responses


beyond


single


word


utterances


may


or may


be complete


sentences.


In order


to find


out


to what


extent


children


spontaneous


generate


complete,


grammatical


sentences,


a separate


index


was


prepared,


Table


, below.


tabulation


includes


ellipses,


e.g.,


and


the


score


reflects


average


number


times


each


child


produced


grammatical


sentences


during


test.


Tucker


significantly


different


from


the


first


three


communities,


Pearl


River,


Bogue


Chitto,


and


Cone-


hatta,


generation


of grammatical


sentences


The










Table


Ranking


of Communities


Sentence


Production


Rank Score Community


1 1.51 Pearl River

2 2.00 Bogue Chitto

3 2.28 Conehatta

4 2.96 Red Water

5 3.10 Standing Pine

6 7.45 Tucker


.3.1


Ranking


of communities


imitative


ability


Finally,


because


imitation


factor


which


may


con-


tribute


to L2


acquisition


the


communities


were


ranked


according


to the children


s ability


imitate


correctly.


score


Table


, below,


reflects


number


times


each


child


imitated


correctly,


either


words,


phrases,


sentences,


during


test.


last


community


, Tucker,


significantly


different


from


first


communities,


Water


Bogue


Chitto.


Comparing


imitative


ability


with


overall


performance


(Score


we find


highest
._


rank


results


n both


to be equivocal.


measures


Tucker


so we might


has

the


the

con-


. -


.I










Table


Ranking


of Communities


Imitative


Ability


Rank


Score


Community


4.28


Red


Water


5.45


Bogue


Pearl


Chitto

River


Standing


6.98


Pine


Conehatta


Tucker


though


these


are


relatively


crude


measures


comparison,


they


suggest


that


relationship


between


spontaneous


imitative


proficiency


may


be neither


straightforward


nor


simple.


This


question


will


explored


at length


following


chapters.


2.3.


. Language
Conclusions


Communities


: Summary


Two


communities,


Tucker


Bogue


Chitto,


are


distin-


guished


on the


basis


several


methods


analysis.


Tucker


s differences


are


significant


statistical


sense


indicate


that


children


this


community


acquire


Engli


sh much


more


successfully


than


children











contrast,


Bogue


Chitto


presents


opposite


SOCIo-


historical


experience:


community


remained


isolated


from


Anglo


influence


res


isted


importation


Anglo


institutions.


stance


This


is reflected


research


in a redu


suggests


rate


that


of L2


Bogue


acquis


Chitto


ition.


Therefore


, it


the conclusion


this


research


that


expo-


sure


outs


classroom,


single


most


important


sociolinguistic


variable


acquisition


reflected


variables


isolated


from


test.


Correlation


Age


English


L2 Acquisition


large,


children


this


study


are


five


years


before


they


are


exposed


to English


on a regu-


interactive


basis.


Before


this,


their


opportunities


to hear


language


are


fairly


numerous


but


sporadic


a passive


nature.


Televi


sion


is present


in almost


every


home,


shopping


trips


to nearby


towns


off-reservation


stores


are


conduct


ed in


Engli


parents)


and


athletic


school


events


are


announced


that


language.


With


their


parents


peers


children


speak


Choctaw


and


hear


Choctaw.


Thus


, with


sole


exception


school,


most


chil-


dren


s opportunities


to communicate


Engli


sh do


not


t *


nI - L a 1-- I. --


*


X





4I


r ,,,,


L


n


L


L










As a result


these


circumstances,


any


changes


children


attributable


s English


their


proficiency


school


with


experience.


are


Until


directly


recently,


this


experience


has


been


primarily


in English,


but


currently


children


are


instructed


bilingually,


Anglo


teachers


teamed


with


Choctaw


aides.


latter


are


enrolled


Mississippi


State


University'


bilingual


teacher


education


program


and


will


eventually


graduate


certified


bilingual-


bicultural


education.


BECOME


program


Choctaw


schools


transi-


tional


one


whose


goal


move


children


from


content ins

instruction


truck


tion in

English


Choctaw

in Grade


Grade


with


100%


structured


content

English


instruction


providing


bridge


between


languages.


All


children


this


study


participated


the


BECOME


program,


thus


children


received


some


instruction


Engli


as a


subject


anc,


with


exception


of Grade


some


content


instruction


English.


Therefore,


the


scores


Table


, below,


represent


result


of Engli


sh instruction


schools,


and,


moreover,


we see


that


this


instruction


successful


extent


that


diff


erences


between


grade


levels


are


sig-


nificant


the


eve


of confidence.










Table


Grade


Level


Score


Grade Score I


K 45.85

1 68.85

2/3 105.65


children,


more


proficient


their


English.


This


crease


in proficiency


is primarily


a product


instruction


exposure


the


schools.


Correlation


of Sex


L2 Acquisition


Scores


results


two


were


sexes


negative;


were


there


isolated


compared.


significant


dif-


ference


between


male


and


female


children


this


study.


Both


sexes


are


equal


proficiency


reflected


Score


Language i
Conclusions


Choctaw


Community


: Summary


preceding


discussion


has


emphasized


limited


role


English


plays


everyday


lives


Choctaw


people.


use


appropriate


in a very


few,


highly


cir-









factors,


in conjunction


with


antipathy


to English-


speaking


people


noted


in Chapter


, result


in an extremely


unfavorable


environment


second


language


learning.


an attempt


to determine


any


the


six


commu-


niti


provides


significantly


different


learning


environ-


ment


from


others,


various


measures


were


analyzed,


with


result


that


one


community,


Tucker


was


found


favor s

unusual


second 1

nature


language

of its


development.

relationship


The

with


stinct


Engli


and

speaker


comparison


with


other


communities


was


judged


to be


stinguishing


ctor.


Finally,


was


found


that


age


correlates


positively


with


language


acquis


ition,


but


sex


does


not.


Note


The Choctaw


lands


in northcentral


S1SS


ippi


are


divided
tances


term


into


SIX


individual


17-40 miles


ese


communities


rvation"


coll


communities


of non-Indian


will


used


land


separate
s from


to ref


ed by
each


dis-
other


these


ectively.


There


ization
scious
suffici
writer


Later


is some


evid


of languages


deci
ent


ecid


Sshe


sion


switch


success


to addr


reported


ence


that


so deeply


h languages


For


ful.


ess


that she. a


extreme


rooted


is not


example


tribal


contextual


that


a mere


always


con-


either


a friend


counc


n el ahnra-t-


in Cho


ctaw


and aiftp









An Anglo attempting
only because it is
e of the attitude


by the
Kenneth


following
York who


ane
se r


g to learn Choctaw is extremely rare
discouraged by the Choctaws but be-
of the Anglo community, as illustrated


cdote,
elative


for
was


whi
th


h I am indebted
translator in


event.


During


coun
was
tran
the
whet
Coun
to k
The
judg
the
why


co
sco
ato
ans
r o
an
w w


hoctaw


was
the
ans
was
del
al
red
peak


hoct
if
udge
on't


nda
tfo
res
was


man


much
'nec
;ngli
the
felo
ansli
t di
Ssom
dent
the


appear
ut of
sary t
. Aft
dge be


local
n it


emp
se
r s


g resident
ted as "y
not spea
time, th
of the Co
defendant


Dy asK
of Nes
," dem
Englis
asked
ty. W


asked,


Choctaw?"


hen
"Then


4. Council
contact and
greatly in


meetings and
early contact


form


from


those


stickball
periods,


today


games occurred in
but they differed


(Swanton


pre-


1931)


5. Because there are sanctions against speaking one's
own name, they are not used in introductions and thus
family relationships, a possible means of orientation,
not immediately available.


are


6. The
and many
there.


only Cho
student


taw high
from the


school
more


is located in Pearl
distant communities


River
board


ire


"What


ct quest
ill-mann
only in
and rou
eyond th
esponsib
e, it is
immedia
ut, a fa
While
they st


bring


ions


concerning


or o
oung
11 i
over


personal
, request
Since th
assistance
idual is
is own ac
e but nec
s a probl
ce cannot
are less
question


matters


are


tions and
essary to
em. If th
be request
bound by t
visitors,


con-
nee
e
d
bso-
blems.

question
by the
con-
*.


Another community,
,,4-lkorn nfl r4- n-^ F 4-ha -\c


Bogue
4-4-a +


Homma, i
nA^ r-.4-4^11


s located


in the


"_ -


n









Swan ton,


in his


introduction


to Byington's dictionary,


states:


Anciently


only


one of


living


there were


these,


several


that


southern


Choctaw dialects,


the Sixtown


part


the old


but


Indians
Choctaw


country,


differed


standard,


Western
ference
certain
language


part o
seems
words,


to any


or Longtown,


f the Nation.
to have been


involving


a whole


considerablee
dialect sp


Moreover,


degree
oken in


this


from
the
dif-


confined mainly to


but


(Byington


very


slightly


1915:ix).


scores


other numerical


data have been rectified


interviewer


items


error;


interviewers


1.e.,
failed


children are not
to administer.


penalized


The reader


switching


is reminded


or pronominal


that


24 of


reference as


items


part of


require


imitative


task.
and


In almost all


cases,


pronouns


involved are


"you"


time of


this


study,


only two classroom


teachers


Grades


were Choctaw.


As can be expected with any new program,


the quality


and
from


degree of
teacher t


BECOM program implementation


O


teacher,


from school


some children received more or


less


varied


to school


greatly
; that


English instruction


than
many


others.


Anglo


There was,


teachers,


who


indeed,


felt


resistance


their


jobs


the part


threatened,


but


presence and


compensate
programs
materials


for t


everywhe


dedication of
he reluctant t
re, there were


with


teacher


the Choct


teacher.


aw aides usually
As with bilingual


problems with


training


curriculum


and retraining.


None-


theless, t
with which


program was,


the writer


in comparison


is familiar,


a model


with many


in its


others


rigorous


administration,


evaluation,


and actualization.


If Score


the difference


for Grades


significant


is analyzed
e .05 level


separately,


only;


and


when


standard


the difference


scores
is not


these


two grades


statistically


are


significant.


compared,


The Grade


score


in Table


is an average based


on a Grade


score of
in later


- -- -


92.98


a Grade


discussion,


1.-^ .... -


L2 growth


score of


appears


fl2 rr..~


118.32.


As mentioned


level
/- -


some


\ ln 1 fl Tf n rn11 fl 1 rr1 l I f 1 fl r n1/ lli [f rf --ll... / l f r r















CHAPTER THREE

MORPHOLOGY


3.1.


Introduction


This


chapter


begins


presentation,


analysis,


and


discussion


language


data.


There


are


three areas


focus:


noun


verb


inflections;


pronominal


case


gender;


and auxiliaries


and


copulas.


format


dis-


cussion will be as


follows.


Each


structure


is presented


individually


and independently


the quantitative data


are


the others.


described.


For


These data


each,


summarize


the number


times


structure


in question


was


obliga-


torily required


the children's


utterances and


compare


this


figure with


the number of


times


the


structure was


actually


correctly


supplied.


These


figures combine


to produce a


basis


"percentage


success"


comparing the degree


figure which


rate


acquisition


across


grade


levels.


success


rate


generally


been


accepted


"acquired"


will


criterion

therefore


classifying


be adopted


for this


item

study


(Brown


1973:258;


Cazden


1972


:33;


Dulay


and Burt


1974a).










Following


quantitative


data


a discussion


types


errors


produced


children.


This,


turn,


followed


a discus


sion


learning


strate-


gies


thought


to affe


ct second


language


acquisition


:the


transfer


of Ll patterns


active


analysis


These


will


evaluated


with


reference


their


relevance


acqui


sition


the


structure


under


discussion.


The


chapter


concludes


with


a comparative


and


cumula-


tive


summary


structures


sented,


a discussion


acquit


sition


pattern


portrayed


data,


and


a com-


prison


this


pattern


with


those


found


other


acqui-


sition


studies,


both


English


L1 and


Noun


Verb


Inflections


Although


English


language


noted


relatively


rigid


word


order


a dependence


on this


order


indi-


cate


role


relationships


within


sentence,


inflectional


morphology


nonetheless


indicates


many


important


relation-


ships.


inflections


are


examined


following


sec


tions.


noun


inflections


are


plural


suffix


{-S};


poss


essive


suffix


definite


and


indefinite


pref


ixes


{a-}


All


these


affixes


are


bound.


S.-% -


I


Y-l----










Noun


inflections


semantic


notions


expressed


three


inflec-


tions


discussed


this


section


are


plurality,


possession,


definiteness.


first


inflections,


although


semantically


cally


very


exactly


distinct,


same.


are


Because


phonetically


of this


and


homophony


phonologi-


we can


determine


basis


eir


emergence


relative


to each


other


whether


children


s mastery


a result


phonetic-phonemic


considerations


or of


syntactic-semantic


ones.


The


emergence


two


relatively


simultaneously


would

more


suggest

dependent


acquisition


to be,


superficial


this


aspe


case


language


least,

compe-


tence,


1.e.,


mast


of phonetic


and


phonemic


systems,


with


present.


implication


2 Independent


that


semantic


emergence


would


notions


suggest


are


acquis


eady


ition


to be dependent


derations,


on more


their


complex,


order


syntacti


emergence


c-semantic


interesting


consi-


as an


indication


of which


concept,


plurality


or possession,


more


obvious,


"easier"


or more


compatible


with


precedents.


Unlike


plural


and


possessive


inflections


, those


indicating


definiteness


are


prefixed


their


use


governed


complicated


co-occurrence


restrictions


both


syntactic


- --


discourse


levels


language.


Briefly,


v










from occurring with plural


count


nouns


(plural


indefinite-


ness


being marked


either


by non-insertion


an article


by the


inclusion


with mass


an indefinite quantifier,


nouns.


Moreover,


e.g.,


both indefinite


"some")


defi-


nite articles


are


usually restricted


from occurring with


proper


nouns


and nouns


referring


to body parts.


These restrictions


are


exceedingly


complex and,


effect,


require the


learner to differentiate among


several


class


ses


of English nouns


in order


to express


plurality


definiteness


appropriately


(see Brown


(1973:340-356)


more


extensive


discussion).


following


three


sections


consider


each


inflections


order


discussed


above.


3.2.1.1.


The


plural


suffix


There


are


six opportunities


included


the SWCEL


test


for producing the


plural


inflection


(Questions


73-75).


The responses


these questions


form


the basic


data


pool.


Added


to these


every


occurrence


in casual


conversation


during


this morpheme.


testing


Table


an obligatory


summarizes


the number


context


obli-


gatory


number


contexts


(opportunities),


successful


attempts,


grade


level,


expresses


the children's


a a 4- 1 a ~ -b% a n 1 a, ~ 4 a- % n A-- 1- n -


1


r*


-c ,FL j_


-L-l -.-: ^- J- -


/-^^-V ** -^/- 4


C


?n


n











Table


The


Plural


Suffix


Grade # Opportunities # Successful % Successful



K 157 59 38%

1 280 116 41%

2/3 234 129 55%





inflection although it has not yet been mastered to the

criterion point of acquisition. Table 10, below, suggests

that age is an important variable in acquiring this inflec-

tion.




Table 10. The Plural Suffix, per Child




Grade # Opportunities # Successful


From


this


breakdown,


we can


see


that


rate


ac-


quisition


increases


dramatically


with


age.


The


number


-1-' 2 a -I -- -s- I I--**.- --


/^--. 3 / -


I


^I,-^ -


- 1 "I e 1


I


r. ulr










sum,


children


produce


plural


inflection


spontaneously


successfully


slightly


more


than


half


time


rate


of successful


production,


relative


opportunities,


makes


a large


increase


between


Grades


and


2/3.


Imitative


abilities


differ


markedly


from


spontaneous


abilities.


Table


reports


children


s imitative


success


and


Figure


compares


this


with


their


spontaneous


success.


Table


Plural


Suffix,


Imitative


Success


Grade


Opportunities


Successful


Successful


As Table


indicates


, the


children


do not


attempt


imitate


this


inflection


very


frequently.


It must


re-


membered


that


they


were


requested


to imitate


every


time


they


produced


response,


an incorrect


column


an incomplete


headed


Opportunities"


(non-sentence)


represents


r~~~~~~~~ Ti 0 I ~nlrr L S. nnr 4-n 4-


4i mae


n I rn or


#--


^*l- I I~ rhn^ "^


1..


T- r i C-A


f


rT-


h














100%

90

80

70

60

50


-I.


Figure


Plural


Suffix,


Spontaneous


and


Imitative


cess


a per


child


basi


changes


little


between


Grades


K and


imitations.


Furthermore,


while


there


a marked


increase


children


s ability


imitate


between


Grades


K and


this










inflection


spontaneously


in Grade


exceeds


their


tative ability.


We must


therefore conclude


that


abilities


are not


consistently related


the acquisition


plural


inflection.


3.2.


Discussion


errors


With


exceptions


noted below,


type of


error


found


for this


inflection


simple omission


suffix


obligatory


to Question


contexts.


"What do


you


For


see


example,


the box?"


in response


(three cats


are


view),


the most


frequent response


"cat"


and even,


occasionally,


cat"


The other


errors are discussed by


grade


level.


In Grade K,


is an


there


extension


are only two,


plural


non-omission


to a mass noun:


errors.


soups"


(3K70138).


The other


the application


plural


a singular noun:


"books"


(5K62035;


only


one book


depicted).


In Grade K,


there are no


cases


spontaneous


error


other than


omission;


however,


classification


responses


to Question


is unclear.


The


children


are


guessing what


inside


a box


possible


they


are


changing their minds about


the number


items


contained:











These


responses


were


not


included


numerical


summary


are


noted


here


because


they


illustrate


a general


char-


acteristic


the


data,


that


variability.


In Grade


there


are


no errors


other


than


omission.


In Grade


, there


are


only


three


non-omission


errors.


one


case,


plural


suffixed


a mass


noun:


"That


a bowl


of soups"


(4369170)


In another


case,


a child


extends


plural


to a sin-


gular


noun.


this


test


item,


child


holds


marbles,


one


"Ask


in each


me which


hand.


one


response


want"


stimulus


child


question:


replies:


"Which


ones,


which


marble s


you


want?"


(3345183)


seems


likely


that


problem


here


partitive


"which"


with


requirement


a singular,


rather


than


a plural,


referent.


A third


response


in Grade


defies


classification:


re clappings"


(4373171)


Although


child


speculation


attempting


inconclusive,


to capture


may


repetitive


be that


nature


verb


pluralizing


this


vein


- a -


, it


1 _


_~


m -- m


.I .










In conclusion,


principal


error at


every


grade


level


the omission


of plural


suffixes.


3.2.1.1.2.


Learning


strategies


following two sections


consider,


first,


role


L1 in


the acquisition


process,


and


secondly,


role of


cognitive


analysis.


Choctaw orthography


ex-


plained


in Appendix


3.2.1.1.2.1.


role of


In order to determine


there


interference or


transfer


from the


children's


we must


first


understand


how Choctaw treats


plurality.


Consider the


following sen-


tences.


Hattakat


itime


pisatok.


'A man saw that


tree.


Hattak


pisa


tok


Man


SCM5


tree


that


see


Hattakat itimg


'Two men saw that


pisatoklotok.


tree.


Hattak


pisa


toklo +


Man


tree


that


see


p.d.


a -


s-s


I 1 --I .. -- A-- --Ll .


1.1 C. ,4-


r I I


I










These


examples


demonstrate


that


verbs,


rather


than


nouns,


are


inflected


number


Choctaw.


Nouns


may


modified


specificity


use


of adjectives6


quantifiers


but


are


not


mark


number.


Therefore


there


is no precedent


children


s LI


plural


flection


on nouns;


other


mechanisms


may


optionally


com-


municate


this


information.


Before


concluding


that


lack


of precedent


children


s L1


accounts


fact


that


they


supply


plural


inflections


only


of the


time


after


four


years


Eng-


instruction


, we must


first


compare


their


performance


with


that


of ESL


learners


from


other


language


backgrounds.


If the


children


this


study


are


similar


to others


would


reject


assigning


overwhelming


influence


Such


a comparison


will


be made


in the


final


section


of this


chapter.


3.2.1.1.


role


cognitive


analyst


specific


learning


strategy


impli


ed by


term


"cognitive


analyst


the


active


organization,


learners,


language


data


they


are


exposed


formulation


of hypotheses


which


regularize


data;


and


realization


these


hypotheses


the children


s L2


*


* *


- *


*


C ~


m


*


I










being


utilized


is the


occurrence


overgeneralizations


children


s L2,


e.g.,


heated "


"ate"


It should


noted


that


classification


errors


as overgeneralizations


not


always


straightforward


case


"eated"


clappingg"


error


in the


pre-


vious


section


illustrates


one


difficulties.


It is


clearly


possible


that


this


transfer


error.


Thus


dis-


cussion


will


be limited


those


errors


which


are


undeniably


overgeneralizations.


second


form


of evidence


that


cognitive


analysis


is actively


employed


learners


occurrence


similar


acquisition


sequences


among


English


learners


from


will


different


L1 backgrounds.


discussed


cognitive


Overgeneralization


analysis


errors


sections


this


chapter;


ESL


acquis


ition


sequences


will


be compared


chapter


s conclusion.


With


regard


error


data


plural


suffix,


there


are


only


cases


of overgeneralization,


occurring


Grades


K and


pluralization


mass


nouns.


therefore


conclude


acquisition


that


process,


these c

to not


children,

employ


this


cognitive


point


analysis


as a general


learning


strategy










being


acquired


but


not


point


that


it is


firmly


part


of the


children


s grammatical


competence.


Imitatively,


children


s productions


are


more


suc-


cessful


than


their


spontaneous


productions


Grade


but


Grades


K and


reverse


case.


relation-


ship


between


modes


production


this


inflection


therefore


not


consistent.


major


error,


every


grade


level,


is omission


inflection.


Cognitive


analy


does


not


appear


to play


an active


role


learning


process


during


these


first


four


years


data,


this


acquis


point,


ition.


we can


absence


tentatively


more


conclude


specific


that


prevalence


omission


errors


influenced


nonex-


istence

This co


noun


inclusion


plural


will


inflections

re-examined


the

the


children

final


s LI.


section


this


chapter.


.1.2.


possessive


suffix


SWCEL


test


includes


only


one


item


requiring


use


possessive


suffix


(Question


As a result,


there


are


fewer


productions


this


morpheme


than


there


are


plural.


occurrences


appear


summary

s below


the distribution


these


Table


e I S I dl---------------------- 2 -SI-- -


I


I .


11 11 *


1


' _




Full Text

PAGE 1

THE ACQUISITION OF ENGLISH BY CHOCTAW SPEAKING CHILDREN By PATRICIA BUTLER KWACHKA A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE U N IVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1981

PAGE 2

Copyright 1981 by Patricia Butler Kwachka

PAGE 3

ACKNOWLEDGMENTS I am indebted to a great many people for their support during the course of this study. Particularly, I would like to thank the Mississippi Band of Choctaw Indians for their cooperation, and especially Abbie Gibson and Kennith York, who were very patient with me. Secondly, I am ex tremely grateful to Jerry Matthews, without whom the lan guage data would have died a certain death in the bowels of Univac. I would also like to acknowledge Mississippi State University for the grant enabling computer analysis of the data. Finally, I would like to thank my chairperson, Dr. Hardman, for her continuing support and her extremely helpful criticism of this study in draft. iii

PAGE 4

TABLE OF CONTENTS Page ACKNOWLEDGMENTS LIST OF TABLES. LIST OF FIGURES ABSTRACT . . iii vi X xii CHAPTER ONE. INTRODUCTION TO THE RESEARCH .. 1 1 3 1.1. 1. 2. 1. 3. 1. 4. 1.5. 1. 6. Introduction . . . . Theories of Learning and Language Acquisition. . . ....... The English Language in Indian Education, The Choctaw Case ............ The Data Base .............. Data Presentation, Objectives, and Justification. .......... Notes . . . . . . 20 29 39 45 CHAPTER TWO. LANGUAGE IN THE CHOCTAW COMMUNITY. 49 2 .1. 2 2 2. 3. 2. 4. 2 5 2. 6. 2 7 Introduction. ........ Language Domains and Communicative Styles. The Communities and English L2 Learning .. The Correlation of Age and L2 Acquisition. The Correlation of Sex and L2 Acquisition. Language in the Choctaw Community: Summary and ConclusioR . . . Notes . . . . . . . 49 49 56 68 70 70 71 CHAPTER THREE. MORPHOLOGY .. 74 3 .1. 3. 2. 3. 3. 3. 4. 3 5 3 6 Introduction . . . Noun and Verb Inflections .. Pronoun Morphology ... Auxiliary Verbs ..... Morphology: Summary and Conclusio n Notes . . . . . . 74 75 136 169 202 237

PAGE 5

Table of Contents, continued. CHAPTER FOUR. SOME OBSERVATIONS ON SYNTAX .. Page 241 4 .1. 4 2 4 3 4. 4. Introduction ... Sentence Types and Sentence Errors. . 2 41 Success . ..... 242 . . . 249 Notes ...... 2 6 9 CHAPTER FIVE. BEYOND DATA. 270 5.1. Introduction ................ 270 5.2. Evaluation of Research Methodology. . 270 5.3. Some Implications of the Children's L2 Development. . . . . 277 5.4. Notes ................... 284 APPENDICES Appendix One. THE SWCEL TEST .... Appendix Two. COMPUTER RESPONSE FORM. Appendix Three. CODING SYSTEM .. Appendix Four. CHOCTAW ORTHOGRAPHY .. BIBLIOGRAPHY BIOGRAPHICAL SKETCH. V 2 8 6 292 296 298 . 301 314

PAGE 6

LIST OF TABLES Table 1. Communities Ranked by Score I ..... 2. A Comparison Score I and Choctaw Language Page 60 Dominance . . . . . . . . 61 3. Ranking of Communities by "Response Appropriateness". . . . . Ranking of Communities by "No Response" 4. 5. 6. 7. 8. 9. 10. Ranking of Communities by "One Word Response" Ranking of Communities by Sentence Production Ranking of Communities by Grade Le~el and Score I The Plural Suffix. Imitative The Plural Suffix, per Child .. Ability 11. The Plural Suffix, Imitative Success. 12. 13. 14. 15. The Possessive Suffix ........ The Possessive Suffix, Imitative Success .. The Article Prefixes. The Article Prefixes, per Child .. 16. The Article Prefixes, Imitative Success . 62 63 64 66 67 70 78 78 79 87 88 95 95 96 17. Article Omission Errors, Imitative Responses. 98 18. Article Omission Errors 19. Types of Article Errors, Excluding Omission 20. Examples of Article Errors. vi 99 100 101

PAGE 7

List of Tables, continued. Table 21. Third Person Singular Present Tense {-s} Page 112 22. Third Person Singular, Imitative Success. 113 23. The Progressive Inflection -ing ..... 123 24. The Progressive Inflection -ing, Imitative Success . . . . . . . . . 124 25. The Past Tense. 26. The Past Tense, Imitative Success. 129 130 27. Pronoun Subjects, Non-Third Person Singular. 138 28. Third Person Singular Subject Pronouns . 139 29. Pronoun Subjects, Non-Third Person, Imitative Success. . . . . . . . 140 30. Third Person Singular Subject Pronouns, Imitative Success .......... 31. Subject Pronouns, Summary .. 32. Object Pronouns. 142 149 152 33. Object Pronouns, Imitative Success 152 34. Third Person Singular Possessive Pronouns. 154 35. Non-Third Person Singular Possessive Pronouns .. 155 36. Third Person Singular Possessive Pronouns, Imitative Success. . . . . .. 37. Non-Third Person Singular Possessive Pronouns, 156 Imitative Success. . . . 15 7 38. Possessive Pronouns, Summary 39. All Auxiliary Verbs .. 40 All Auxiliary Verbs, Imitative Success 41. The Auxiliary "do", Overall Success. 158 170 171 173 42. The Auxiliary "do", Overall Imitative Success .. 173 vii

PAGE 8

List of Tables, continued. Table 43. 44. 4 5. 46. 47. 48. "Do" and "Does" "Doesn't" "Do" and "Does", Imitative Success. "Doesn't", Imitative Success .. Copular Auxiliaries .... 49. Copular Auxiliaries, Imitative Success . so. 51 52. Third Person Singular Copular Auxiliary .. Non-Third Person Copular Auxiliary ... Contracted Copular Auxiliary, All Persons Page 175 175 175 176 177 187 188 189 190 190 53. Non-Contracted Copular Auxiliary, All Persons 191 54. Third Person Singular, Imitative Success. 192 55. Non-Third Person, Imitative Success .... 192 56. Contracted, All Persons, Imitative Success. 192 57. Non-Contracted, All Persons, Imitative Success .. 193 58. The Auxiliary "Will" .... 198 59. The Auxiliary "Will", Imitative Success ..... 198 60. Percentage Point Change in Successful Spontaneous Production between Grades Kand 2/3 . ... 204 61. 62. 63. Grades 2 / 3 Spontaneous Success Order and Activity Indices ......... English Ll and L2 Acquisition Sequences. English Ll and L2 Acquisition .. 64. Comparative English L2 Sequences .. 65. Three English L2 Acquisition Sequences. 66. Ll Precedent for L2 Inflections viii 205 222 226 227 228 233

PAGE 9

List of Tables, continued. Table 67. Sentence Types Ranked by Three Response Variables. . . . . . 68. Sentence Types, Spontaneous and Imitative Success 69. Sentence Success, Grades K 2/3. 70. 71. Transitive Sentences .. Interrogatives ... 72. Copular Sentences .... 73. Intransitive Sentences, Affirmative and Page 243 245 246 247 247 248 Negative. . . . . . . . 249 74. 75. 76. Subject Omission, Summary. Pronoun Subject Omission. Noun Subject Omission. 77. Subject Omission, Imitative Rate. 78. Objects in Transitive Sentences .. 79. Objects in Transitive Sentneces, Imitative 80. 81. Success . . . . . . . Copular Verbs, Omission Rate. Copular Verbs, Imitative Omission Rate .. 82. Adjectives. 83. Adjectives, Imitative Success 84. Locative Prepositions 85 Locati v e Prepositions, Imitative Success ix 250 251 251 252 256 256 260 261 263 263 266 266

PAGE 10

LIST OF FIGURES Figure 1. 2 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. The Plural Suffix, Spontaneous and Imitative Success ......... The Possessive Suffix, Spontaneous and Imitative Success . . . .. The Article Prefixes, Spontaneous and Imitative Success . .. Third Person Singular, Spontaneous and Imitative Success .. The Progressive Inflection, Spontaneous and Imitative Success . . . ..... The Past Tense, Spontaneous and Imitative Success . . . . . Subject Pronouns, Spontaneous and Imitative Success. . . .. Third Person Singular Subject Pronouns, Spontaneous and Imitative Success. Pre-Verb Pronouns, Summary, Spontaneous and Imitative Success . . Object Pronouns, Spontaneous and Imitative Success . . . . Possessive Pronouns, Summary, Spontaneous and Imitative Success . . Auxiliary Verbs, Spontaneous and Imitative Success . . . . . . . Auxiliary "do", Spontaneous and Imitative Success . . . . Forms of "do" Spontaneous and Imitative Success . . . . . X . . . . Page 80 88 97 114 125 130 140 142 . 150 . 153 158 . 171 . 174 . 178

PAGE 11

List of Figures, continued. Figure 15. Copular Auxiliaries, Spontaneous and Imitative Success. . 16. Spontaneous Success, Summary. 17. Spontaneous and Imitative Success of Inflections Declining in Grade 1. 18. Spontaneous and Imitative Success of Inflections that Rise Minimally or Plateau in Grade 1 . . . . 19. Imitative Success, Summary . . . 20. Grade K Imitative Ability and Grades 1 and Spontaneous Ability . . . . 21. Grade 1 Imitative Ability and Grades 2/3 Spontaneous Ability . . . xi . 2/3 . . . . Page 188 207 210 211 214 216 217

PAGE 12

Abstract of a Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE ACQUISITION OF ENGLISH BY CHOCTAW SPEAKING CHILDREN By Patricia Butler Kwachka March, 1981 Chairperson: Martha Hardman-de-Bautista Major Department: Linguistics This study presents the results of a four-year, cross sectional examination o f the acquisition of English as a second language by Choctaw speaking children. The primary purpose of the research was to discover what features of English the children acquire in their first four years of school, when specific features appear, and the extent to which the features are successfully produced. The secondary purpose was to explore the children's learning strategies in an attempt to establish the influence of the first lan guage and to determine the role of non-behavioralistic cognitive processes. The first chapter contains a brief and general dis cussion of learning theories applicable to language acqui sition, followed by an examination of English language learning in the Native American context, in general, and the Choctaw case, in particular. The second chapter explores sociocultural concomitants of language acquisition, xii

PAGE 13

xiii specifically: language domains; communicative styles; and the variables of sex, age, and community of residence. The third chapter presents extensive description and ana lyses of the development of noun and verb inflections, of pronoun morphology, and of auxiliaries. Accompanying the developmental data for each inflection are: a comparison with imitative development; a contrastive analysis of Choctaw and English with regard to the feature under dis cussion; and an analysis of second language errors. The chapter concludes with comparisons of these data with other studies of children learning English both as a first and as a second language. The fourth chapter examines the devel opment of four sentence types, transitive, intransitive, copular, and interrogative. The final chapter includes a brief evaluation of the research methodology and some impli cations of the study for English language acquisition in the future. An important finding of this study is that over a four year period, there is tremendous variability in the success rates of individual language features. This variability is so great that no regular sequence of acquisition emerges from the data. Secondly, the study found that none of the inflections was produced at the criterion level of acquisi tion at the end of four years of language learning. It is suggested that the students' limited contact with and need for the language, in conjunction with the socio-historical facts of their relationship with speakers of English, : : are

PAGE 14

xiv the major variables affecting the acquisition of English as a second language.

PAGE 15

CHAPTER ONE INTRODUCTION TO THE RESEARCH 1.1 Introduction Despite the fact that everyone learns a language and many learn two or more, our knowledge of the processes by which language is acquired is more speculative than it is concrete. It is not an area of investigation which lends itself to experimental manipulation, and, because acqui sition is a lengthy process, 1 patient observation, recording, and analysis of large bodies of either longi tudinal or cross-sectional data are necessary. This study is based on the latter approach. The data base consists of English language samples from 206 Choctaw children, aged five through ten, who speak 2 Choctaw, a Muskogean language, as their first language (Ll). The general research objective is descriptive: to discover and report the relative sequence of emergence of important English morphological and syntactic structures in the development of the children's second language (L2). There is an urgent need for studies of this type in minority language communities, particularly Native American communities. The educational problems besetting Native Americans in this country are complex, but the 1

PAGE 16

2 issue of language is fundamental to their resolution. (See Section 1.3. and Chapter Two for further discussion.) Besides their theoretical interest, studies such as this are essential for assessment, for the development of teaching materials, and for the evaluation of programs; yet this is the only extensive study the writer is aware of for speakers of a Native American language, although there are numerous studies of English L2 acquisition by speakers of more prominent languages. 3 Finally, research of this nature has important applications in the development of a coherent, convincing, theoretical model of second language acquisition. Issues in this area will be discussed in the next section, but, briefly, they may be summarized in the question underlying all acquisition research: what minimal conditionsenvironmental, physiological, and psychological--must obtain for successful language development to take place? This question will be addressed in the next section, 1.2., which compares current theories of language acqui sition. The following section, 1.3., discusses English and the education of Native American, and specifically, Choctaw children. Section 1.4. describes the data base and method of analysis, and Section 1.5. the format for data presentation and the research objectives.

PAGE 17

1.2. Theories of Learning and Language Acquisition The development of a scientific discipline is much like the development of language itself. The historical curve is marked by vertical rushes of dlscovery, uneven periods of ensuing revolution, followed by quiescent lateral periods of assessment and incorporation of new material. Occasionally, hiatuses are created when pre viously accepted material must be rejected or, at least, re-evaluated in the face of new data. It is in just this situation that linguists of the past decade found them selves as they argued, evaluated, and extended their work in the light of the then newly introduced transformational theory. The impact on the field of language learning which, heretofore, had been more a subdiscipline of psy chology than of linguistics, was extreme, for Chomsky proposed "an objective for linguistic theory (to be) the precise specification of ... a model for acquisition of language" (Chomsky, N. 1964:61). In fact, it is difficult to separate the roles of language and learning in human behavior, but the develop ment of a theory encompassing both language structure and cognition is even more demanding. The task is complicated by the fact that we must use language in our attempts to understand language; language is at once the means and the end of our research. Moreover, not onl y must we try to understand language itself in this fashion, but it is 3

PAGE 18

through language that we understand ourselves: language is the unconscious summation of a great deal of that which is considered human; it is one of the defining characteristics of our species. Within the species, it serves as a primary mechanism of differentiation and of intra-group identification. It may be that language is preeminently responsible for our survival as a species, for it is through language that we transmit our cumula tive understanding of ourselves. 4 Language is both complex and mysterious. Its com plexity is attested by the fact that no satisfactorily complete grammar exists for any language, yet, on the other hand, children as inevitably and effortlessly learn to talk as they learn to walk. In the case of walking and upright posture, we can account for their development in terms of the intersection of species specific anatomical features and the ineluctable influence of genetic matura tional schema. But can we account for language develop ment in the same way? The various current models of acquisition take these factors into account variably. There is, in fact, an embarrassment of riches in the area of language acquisition theory. The following discussion will single out two of the more prominent models, one which dominated pre transformational linguistics and whose influence is still felt; and one which arose from the implications of transformational theory. The first may be broadly

PAGE 19

characterized as a behavioral model and the second as a cognitive model. In their extreme versions, neither theory is adequate, but both contribute to our under standing of acquisition phenomena. 1.2.1. Two theories of language learning The first theory of language learning, the behavioral model, derives from traditional learning theory and conse quently views the development of language as the result of stimulus-response-reinforcement processes. Important to this model is the role of the parent as the source of the primary input data and the reinforcer of appropriate re sponses. The learner's ability to imitate and the moti vation provided by rewarding correct imitations are also important factors. In essence, this theory maintains that language is a set of successfully learned habits. (For further discussion, see Crothers and Suppes 1967; Palermo 1971; Skinner 1957; Staats 1971, 1974.) The implications for L2 learning are that the learner will transfer Ll habits (in the form of phonological, morphological, and syntactic patterns) to the L2. Where the structures of the two languages are similar, these habits will facilitate language learning; where they are different, they will be irrelevant to the learning process. L2 learning errors are explained as the transfer of inappropriate and / or conflicting Ll habits. (Fo~ 5

PAGE 20

6 examples see DiPietro 1971; Ferguson 1965; Ohannessian and Gage 1969) From the pedagogical point of view, the implications are that a new set of habits must be established in the learners, their congruent patterns reinforced, and their con flicting habits extinguished in the L2 context. The accepted method for accomplishing this is reinforcement through repetitive oral language drills, an approach known as the direct or audio-lingual language teaching method. (For further discussion, see Dillard 1978) The second theory of language learning has not been so systematically developed. It has, moreover, become quite controversial as a result of extreme claims made by some 4 proponents. Despite the multiplicity of viewpoints held by this theory's adherents, the various versions share a conu~on thread: that the ability to learn a language is innate and dependent on the operation of inherent, uni versal cognitive processes. The extent of the innate capacity attributed to the child divides adherents. Chomsky's own position is open to interpretation and has been the subject of much specu lation, e.g., Braine 1971:182-186. One interpretation of his position would ascribe to the human infant an actual metatheor y of language itself: "some delimitation of a class of p oss i ble h y potheses ab o ut lan g ua g e structure (Chomsk y N. 19 6 5:39). A st r on g er interpretation w o uld endow the infant with "innate ideas and p rinciples" II

PAGE 21

7 (Chomsky 1965:59). Whatever Chomsky's own position, 5 the implications of this theory may be roughly delineated. First, language development is independent of such external factors as input quality, reinforcement, and motivation. The presence of language in a normal, communication setting is sufficient to activate children's inherent acquisition capacities; they will begin actively exploring this language by testing hypotheses until they eventually arrive at a set of substantiated hypotheses, in other words, a grammar of the language (Katz 1966). By extension, the implication for L2 learning is that acquisition will proceed in much the same manner, i.e., by a process of matching hypotheses to input data. It is important to note that this model of language learning claims that the role of the learner's Ll is unimportant; that the determining factor is the L2 itself which, by vir tue of the underlying relationships it embodies, dictates the acquisition process. Therefore, the pedagogical implications of this model are nebulous. This discussion has not argued either theory but merely presented them in simplified outline for orientation. For more detailed discussion and debate, the reader is referred to Beilin 1975; Brown 1973; Carterette and Friedman 1976; Fodor and Katz 1964; Hatch 1978a; Moore 1973; Richards 1974a, 1978; Schiefelbusch and Llo y d 1974; Slobin 1971.

PAGE 22

1.2.2. The evidence of research An examination of research provides inconclusive evi dence for evaluating the validity of either theory. Len neberg (1964, 1973) finds support for the cognitive model in biological data. He cites the universal appearance of language at a relatively fixed point in the developmental process, similar rates of acquisition, and the attainment of proficiency at comparable developmental stages. "Per8 feet chronological commensurability," as he refers to these phenomena, could not be coincidental, according to his argument. And, considering the variability, cross culturally, of input data, reinforcement practices, and motivating factors, as well as gross variation in indi vidual intelligence, this commensurability is indeed difficult to account for within the learning theory model. Observations of the learning process itself provide evidence which also tends to support the cognitive theory of language acquisition. First, with regard to the input data: if t~aditional learning theory is to account for a child's ability to produce grammatical sentences, then that which the child listens to (the model) must itself be grammatical. However, there is some real question about whether or not we are in fact consistently grammatical in our e v er y day, unmonitored speech. Choms ky claims w e are not; t h at the input d ata are of a deg e n er a te q ualit y and narrowl y limited" (1 9 6 5 : 58) O n the o ther hand, sociolinguists, whose resear cn domai n is casual speech, claim

PAGE 23

we are grammatical in 75% of our utterances 6 (Labcv 1972:203). More recently, however, Hatch (1978b:64) finds in her work that casual narratives are replete with errors and inconsistencies. Jakobovits (1970:266) also finds that fluent, adult speakers are only "semi-grammatical". However, there is the possibility that learners dis regard any data they cannot process, so that degree of grammaticality and the general quality of the input data may not be relevant issues. Further investigation of this question is obviously called for, especially in view of the fact that it comprises an important component of the behavioral model. Moreover, it is one of the few aspects of language learning which lends itself to experimental manipulation and is certainly readily observable. Other questions relating to the role of the input data have been examined. Drach (1969) and Snow (1972) have observed that adults regulate their speech with chil dren by simplifying syntax and shortening utterances. 9 There is some evidence that such telegraphic speech (speech with functors omitted) may aid the comprehension of 18 month old children (Shipley et al. 1969). An analysis of the speech of the mothers in Brown's longitudinal study of child speech (1970, 1973) by Pfuderer (1969) demonstrated that as the children's speech increased in complexity so did the mothers'. Ervin-Tripp (1971:192) suggests that this may be a mechanism for maintaining a consistent rela tion between the children's ability to comprehend and the input they are required to handle.

PAGE 24

10 Another input factor which might influence acquisi tion is the frequency of adult usage of, for example, specific morphological inflections. Brown, Cazden and Bellugi (1970:138-152) concluded, after examining this question, that there was no evidence to support the hypo thesis that expansions (adult responses to telegraphic utterances which repeat the children's productions and supply the missing functors) aid acquisition. In fact, modeling (the parent supplying a grammatical response to the child's utterance) was demonstrated to be more effec tive both in an experimental situation and in the longi tudinal study. In the latter, they report: "For all three children, order of emergence within the child's language system is more strongly related to the frequency with which the inflection is modeled by the parent than it is to the proportion of frequency of expansion" (1970:148). In only one child, however, did the relationship achieve . 1 . f. 7 statistica signi icance. A caveat should be observed here, since frequency is not the only relevant factor. Many high frequency items, e.g., "the", do not emerge early in the Ll acquisition sequence so that frequency alone cannot be considered determining. It must be assumed that there is an interaction between semantic and/or syntactic necessit y and redundancy, on the one hand, and frequency, on the other, which affects the order of acquisition. In sum, a definite relationship between input fre quencies and the appearance of structures in Ll production

PAGE 25

11 has not been established. It is a variable which deserves further investigation and which, moreover, lends itself to experimental manipulation. One final point with regard to input data must be mentioned since it poses particular problems for any theory ascribing preeminent importance to the roles of grammatical modeling and imitation. As every parent knows, children, from the time they become intelligibly verbal, produce and maintain for substantial periods of time utterances which deviate so markedly from any dialect of the language that there could have been no existing model, let alone a rein forcing one (Brown and Bellugi 1970:90; McNeill 1970:105; Menyuk 1969:124). This widespread phenomenon must certainly be accounted for by any acquisition model. In concluding this discussion of the role of input data, the need for further investigation must be stressed. The results are especially pertinent to L2 acquisition in pedagogical or "unnatural" language learning contexts. Many current teaching materials attempt to sequence lan guage data and to provide frequent, reinforcing exercises and reviews on the assumption that this method of presen tation facilitates learning. Yet materials developers frequently have little information on which to base deci sions concerning curriculum design, e.g., optimal learning sequences (which may or may not parallel the Ll sequence); or optimum amounts and timing of reinforcement e x ercises.

PAGE 26

12 One promising area of study whose results have been overlooked in acquisition research is pidginization The pidginization process, and subsequent creolization, occur in compressed time periods and thus offer an excellent minilab for observing rapid language change and development (Ferguson 1971; Valdman and Phillips 1975). A coordination of these research findings with acquisition findings should be fruitful. A final factor to be considered, the role of imitation, is important to the behavioral model of language learning. There is little agreement among researchers concerning the relationship between spontaneous productive capacity and imitative capabilities. Many researchers find that imita tive ability exceeds productive ability. Brown and his colleagues (1970) discovered that children could imitate sentences which were longer than any of their own sponta neous productions and that "production in the sense of imitation proves to be more advanced in three year olds (1970:54). Menyuk also found, in a comparison of normal and deviant-speaking children, that normal-speaking chil dren "exceed, in some instances, in their repetitions, the grammatical competence displayed in their utterances" (1969:141). On the other hand, McNeill (1966, 1970) maintains that children cannot imitate utterances which are not alread y within their productive capacities, and reports an experience probably familiar to all linguist-parents: II

PAGE 27

a mother patiently repeats eight times the grammatically correct version of her son's utterance while the equally patient child imitates, incorrectly, after each modeling; the child's final imitation maintains his own error and incorporates his mother's correction as well. 13 Age may be a factor in the ability to imitate. Ervin Tripp (1971:197) suggests that in the early stages of language development children do not imitate sentences more successfully than they produce them spontaneously (with the exception of imitations that are expansions of their own utterances), and that the role of imitation changes with the age, or developmental stage, of the learner. Not only age but also individual selection of com patible learning strategies may be a factor. Bloom and colleagues (1974) report that not all children imitate, but for those who do, the imitated structure appears in their production shortly after the period of imitation. There is little information on the role of imitation in the L2 situation, despite the fact that many current language textbooks stress the importance of repetitive drills. Jakobovits suggests that, while practice and repetition should assist in gaining sensory and motor integration, it would be interesting to determine whether or not there is a point in the L2 learner's development when "automaticity practice is likely to be valuable (1970:264). This speculation seems deserving of investi gation and is consonant with Ervin-Tripp's conclusion II

PAGE 28

regarding the relationship between age and imitation in Ll learners. 14 In summary, the role of external factors, particularly that of the input data, is not clear. Research neither clearly supports nor rejects either language learning model. Naturalistic Ll studies, such as that of Brown and his colleagues, provide us with our most complete infor mation, and yet the very nature of the setting often pre cludes the possibility of assigning significance to any single variable either in the child's linguistic or non linguistic environment. Investigators of Ll learning gen erally conclude that the role of external factors is less interesting and less important than the role of the chil dren themselves who seem to be guided by as yet unspecified, innate capacities to create hypotheses and test them, gradually decreasing their generalizability in order to account for increasingly restricted patterns of occur8 rence. (For further discussion, see Brown and Bellugi 1970:75-79: Cazden 1972; Dulay and Burt 1974b:36; Langen doen 1970:1-6; McNeil 1970:104; Moskowitz 1978.) Turning now to a closer consideration of the L2 learner, it is reasonable to assume that the role of external and non-maturational factors (e.g., aural dis crimination) would increase in importance as the age of the learner increases. The biological and social attri butes of older individuals differ from those of individuals who are neither physically nor culturally adults.

PAGE 29

Lenneberg suggests that a negative effect on language learning abilities is produced by completion of brain lateralization during the teen years which inhibits cog nitive flexibility (Lenneberg 1973:96; see also Luria 1969; Seliger 1978). 15 The nonbiological concomitants of age to be consi dered are both numerous and nebulous. Of preeminent impor tance is the fact that older L2 learners, even though they may still be children, will have already learned a culture, or a large part of it; and with this culture they will have acquired a set of attitudes, beliefs, values, prac tices, and behaviors which will affect their attitudes to, and their relationships with, the world outside theri culture. Therefore, the learning of a language which is a part of that outside world may be affected by many non linguistic factors, including the historical relationship of the learners' culture to the culture of the language they are acquiring (Fishman 1972:21; Lambert et al. 1963); the learners' attitude to the L2 culture which is usually, but not entirely or necessarily, a result of the cultures' historical relationship (Oller et al. 1977); exposure, opportunity, motivation, personality, and a host of even less tangible factors, e.g., the learners' cultural beliefs concerning language and learning. We will return to a discussion of these factors in the next section. At this point, there are two further aspects of L2 learning which require consideration. First is the fact

PAGE 30

16 that children who begin learning an L2 several years after they have begun learning their Ll do not have to learn what "language" is. Even if they are, indeed, born with a metatheoretical understanding, they have, by the time of L2 acquisition, gained experiential knowledge of this con cept. Moreover, their cognitive development, to whatever extent it may be separated from their language development, is at a more sophisticated level than is that of Ll learners. It is not clear whether these facts are advan tageous or not. Ravem concludes (1974a:132) from the study of his Norwegian Ll son acquiring English that "a normal six-year-old child at all levels of language is greatly facilitated by the linguistic competence he (sic] already possessed through his first language." 9 The second factor to be considered in the L2 learning environment is the manner of exposure to the language. A fully bilingual environment where simultaneous acquisi tion of two languages occurs offers substantially different circumstances to the learner than sequential acquisition in a partially bilingual environment (Leopold 1937; Vildomec 1963). Furthermore, in the case of sequential acqui sition, one would expect there to be significant differences between exposure to the L2 in a natural communicative setting and an artificial one, e.g., a classroom. L2 research dealing with these questions has been highl y data-oriented. Error anal y sis is an important methodological tool in the field (Richards 1974a, 1974b;

PAGE 31

17 Burt and Kiparsky 1972; Dulay and Burt 1974a). Increased attention is being given to the nature of the interlanguage, or the systematic approximation to the target language, de veloped by the learner (Selinker 1974; Corder 1978). By analyzing errors and interlanguage, researchers working with comparative data from different language groups acquiring the same L2 hope to determine the roles of trans fer and habit development, and of cognitive analysis. Another area of investigation is comparative acqui sitional sequences (Dulay and Burt 1974b; Ravem 1974b; Natalicio and Natalicio 1971; Hatch 1978a). Here, re searchers compare the order of emergence of specific morphological and syntactic forms exhibited by speakers from different language families learning the same L2. If the sequences are similar, it is argued, then the cog nitive model of language learning is supported, since such data would suggest that the L2 determines the learning process, i.e., defines the set of possible hypotheses. If the sequences are significantly different, then it may be argued that the differences are caused by the learners bringing Ll habits and assumptions to bear on the L2 learning process. The same arguments apply to comparison of Ll and L2 acquisition sequences of the same language, another area of investigation. In concluding this discussion, several points should be emphasized. First, the above references to curre~t theoretical models have been somewhat exaggerated and

PAGE 32

18 generalized in order to maximize comparison. Other models have been proposed (Braine 1971; Richards 1978) but, as the discussion has pointed out, our knowledge and our data are nowhere nearly sufficient to provide us with the infor mation necessary for conclusive decisions. We are parti cularly constrained by our ignorance of the brain's neuro chemical mechanisms. Although such information will cer tainly not answer all our questions concerning language learning, it will more definitely delimit the possible boundaries of our theoretical models. Secondly, past research has focused on the acquisition of the information-communication functions of language. Language as play, as art, as evasion, as status regulation, and all the many other functions controlled, and somehow acquired by the mature speaker are largely uncharted areas for investigation. (See, however, Peters 1977; Scollon 1976.) Granted that we have yet to understand the acquisition of the former function, we must not lose sight of the fact that our models must also allow for the acquisi tion of the latter, sociolinguistic functions. Finally, there are two further points to be discussed with reference to research. As mentioned previously, experimental manipulation is difficult. Not only are we subject to the usual restrictions which apply when humans are involved; but we must also be sensitive to the fact that, because the object of our research is language, an attribute so integral to our personalities and thought -____________________ ____.

PAGE 33

19 processes that it is difficult to define either without reference to language, any experimentation which might in hibit or distort the subjects' language capacities could be permanently damaging. Thus, for example, we cannot examine the role of input factors by withholding from an experi mental group access to the question transformation and ob serving if they develop one on their own. What we can do, however, is test; and this brings us to our second point. Language testing, as it is currently practiced, is still fairly primitive. One of the reasons for this is that we are neither certain of what we are looking for nor of how to interpret what we obtain. Many measures in use are either very subjective in their scoring procedures (e.g., the Bilingual Syntax Measure, Burt et al. 1973) or very arbitrary in both their selection of tested struc tures and scoring procedures (e.g., the SWCEL, Silverman et al. 1976). It is to be hoped that definitive descrip tions of actual language acquisition sequences and proces ses will, in the future, contribute to more realistic and efficient language assessment instruments. The preceding discussion has outlined models and research in language acquisition. The following discus sion presents a specific case of L2 acquisition empha sizing the sociocultural circumstances which obtain in the learning environment.

PAGE 34

1.3. The English Language in Indian Education, the Choctaw case 20 As an introduction and orientation to the Choctaw case, this section begins with a general discussion of the English language in Indian education. The discussion pro vides context for the presentation of the Choctaw case and also serves to underline the fact that the situation faced by the Choctaws is not an isolated circumstance; many tribes in this country and in Canada are faced with similar issues and considerations with regard to the acquisition of English as a second language (ESL). 1.3.1. The English language in Indian education The language philosophy of this country has been an extension of its domestic, sociopolitical philosophy; that is, all foreign elements should dissolve in a "melting pot" and be thereby transmogrified into a single, homo geneous English speaking identity. From the Indian point of view, the blatantly revisionist definitions of "foreign'' and "native" are characteristic of the entire history of Anglo-Indian relationships. Despite the country's defi nitions and philosophy, many Indian groups have resisted the loss of their native languages much more successfully than they resisted the loss of their lands. Today there are around 500,000 speakers of Native American and Alaskan Native languages in the U.S. and Canada (Martin 1975; Spolsky 1972). These speakers re present approximately 200 different languages. Although

PAGE 35

21 some of these languages are dying, 10 others are actually growing due to the fact that Native Americans are one of the fastest growing minority groups in the country (Osborn 1970:229). Moreover, many groups, concerned that their languages will be lost, have begun teaching their children the Native language as a second language. 11 Although there are no comprehensive or reliable statistics to refer to, the majority of Native language Ll speakers in this country also speak English in degrees ranging from "limited" to fluent bilingualism. One ex planation for this lies in the fact that most adult In dians attended Bureau of Indian Affairs (BIA) schools where the language of instruction has been adamantly English. In fact, one of the fundamental goals of these schools was that the students' "barbarous dialects would be blotted out and the English language substituted" (quoted in Leibowitz 1971:2). Use of the Native language in BIA schools was, until recently, severely punished. The intent of the BIA was overtly assimilationist, and loss of the Native language was, quite correctly, seen as a lever in achieving this goai. 12 Considering the fact that the students did not understand the language of instruction, it is not surprising that high dropout rates and low proficienc y scores characterized the Indian student ( Osborn 1970; U.S.C.C.R. 1975:18-19). Negative expectations, the self-fulfilling prophec y effect of the dominant culture's stereot y pe of the minorit y

PAGE 36

22 and low self-concepts have all played a part in the dismal history of the education of the American Indian (Williams 1970; U.S.C.C.R. 1975:38). Failure to learn fluent English 13 has certainly contributed to the overall failure and rea sons for the language failure are not difficult to find. First, for reservation Indians especially, there has been relatively little incentive to learn English. One can conduct the majority of one's daily life in the Native language, the language of the home. English, the langugae of the schools, is viewed as an alien imposition and therefore. resented (Cornejo 1974:1). Furthermore, mastering English in no way guarantees access to improved economic benefits, since prejudice and isolation combine to keep the Indian on the lowest rungs of career ladders. Finally, as Plumer (1970:275) observes: If they see themselves locked out of society anyway, then their motivation to learn English will be understandably low, especially if in doing so they risk cutting themselves off from associations they already have, namely their peers and families. In conclusion, it has been the overt policy of the U.S. government, through the BIA, to assimilate Native Americans by depriving them of their languages so they will be unable to carry on their cultures. Ironically, per haps the same prejudice which inspired this policy has, in conjunction with physical isolation, kept the Indian apart from the dominant Anglo society and their culture alive. Turning now to the Choctaws of Mississippi, we can

PAGE 37

see all of the factors mentioned in the preceding dis cussion, and more, operating in their situation. 1.3.2. The Mississippi Band of Choctaw Indians and English education Today there are approximately 4,000 Choctaws in the state of Mississippi (Spencer et al. 1975; for histories, see Debo 1934 and DeRosier 1970). Choctaw is very much a living language. Of the families with elementary school aged children, 83.4% speak Choctaw more than 90% of the time (York and Scott 1976). Slightly fewer than 7% of all households use English more than Choctaw (Spencer et al. 1975). The remainder report equal use of Choctaw and English. Language dominance testing of children entering kindergarten in 1976 rated 87.6% of the students dominant in the Choctaw language (Lewis et al. 1977). Unlike many other Native American groups, the Choc taws have a long tradition of literacy in their own lan guage. Missionaries introduced an orthography in the early 19th century and operated bilingual schools. By 1837, there were some 576,000 pages of works, mostly of a religious nature, in the Choctaw language and some 80% of all adults were literate in Choctaw (Debo 1934:62). 23 The New Testament of the Bible and a hymnal were translated using the missionary orthography and older people today read and write Choctaw as a result of their strong asso ciation with church and religious activities. In

PAGE 38

24 Oklahoma, a standardized variant of the missionary ortho graphy has been adopted by the Choctaw Bilingual Program (Nicklas 1972; Jacob et al. 1977) and in Mississippi a "school" orthography has been adapted for teaching purposes. Although Choctaw is very definitely their dominant language, most adult Choctaws do speak English. Again, this is more a result of exposure to the language in the six BIA schools on the reservation than of any need or de sire. The schools, until recently, operated on the stan dard BIA model: an Anglo, monolingual teacher whose in struction was translated to the students by Choctaw, bilingual aides in the classroom. This form of instruction persisted until the fourth or fifth grade when the chil dren's English skills were deemed adequate for classroom purposes. At no point, it should be noted, were the students provided with English language instruction per se. The predictable results of this approach to education are indicated by the following facts. More than 77% of adult Choctaws have not completed high school, while ap proximately 27 % have had less than three years of schooling (Spencer et al. 1975). Those students who remain in school through the 12th grade score at the 8th grade level on national achievement tests (Lewis et al. 1977). Testing of y ounger children over a period of y ears demonstrates a cumulati v e deficit effect; the longer the y remain in school, the farther behind the y score in terms of national norms.

PAGE 39

25 While it is obvious that a lack of fluent English has contributed to this educational shambles, the reasons for this lack are not entirely obvious. Some are, such as the fact that the L2 was never taught as a subject. Students were expected to "pick it up." Not only is picking up a language an inefficient way to learn, it also requires a highly motivated learner who is immersed in communicative opportunities. These requisites simply do not obtain in the Choctaw context. Despite the urging of parents, 14 children do not learn English quickly, easily, 15 or well. In fact, many develop a strong dislike for the language. 16 The basis for this dislike is entangled in a complex matrix of sociocultural factors, and no one factor can be identified as a primary or independent variable affecting English language acquisition. One important consideration is the Choctaws' attitude toward their own language. To them, language is extremely significant: to be Choctaw is to speak Choctaw. Speaking the language is, possibly, the single most important aspect of one's Choctaw identity. 17 This absolute identification of language and social group membership has direct implications for second language learning: it can be argued that there is an unconscious extension of this belief so that, if one speaks English, one has become, in some fundamental sense, nahollo or 'An 1 I 18 g O Because the Choctaws have little liking for Anglos, speaking English, though sometimes exigent, is not

PAGE 40

26 desirable. One becomes too closely identified with Anglos 19 and Anglo culture. (For discussion of other Native American beliefs regarding language, see Ohannessian 1967: 10-11.) The Choctaw attitude to the dominant culture and its language must only be exacerbated by the schools' attempts t f h h ld k 1 h 20 N 1 d o orce t e c i ren to spea Eng is ot on y o students resist the language itself, their culture regards any form of overt, direct pressure as highly unacceptable. The culture tolerates extreme deviance from expected beha vior patterns. The major pressure to conform is the de viating individuals' expectation of collective, community disapproval which, though rarely expressed directly to them, is circulated through critical gossip. It is a very effective mechanism of social control. Nonetheless, it is considered the absolute right of individuals to dis cover and decide for themselves, no matter how long this may require nor how flagrantly they may violate standards in the interim. It is grossly improper to directly tell a person how to behave. Thus, to place individuals in the position of having to do something, as the schools have done, creates a great deal of tension and the students respond by withdrawing from the situation. In fact, withdrawal fairly well typifies the Choctaw children's total response to the school environment. Anglo teaching styles appear coercive to them and require them to behave in ways that, because they are considered

PAGE 41

27 improper in their own culture, make them feel uncomfor table: answering questions when they are unsure of the response and have not been given time to consider it for themselves; having to ask questions rather than being allowed to observe; being singled out to perform in front of others; not sharing work assignments, but completing them individually; being expected to maintain steady eye contact; the high decibel and movement level of Anglo communicative style; these, and many other factors, con tribute to their unease. Since the school is the children's sole exposure to English on an interactive basis, their attitude to the language becomes synonymous with their school experience; that is, negative. (For comparable findings with other Native American students, see Cazden and John 1968; Dumont 1972; John 1972; Ohannessian 1967:26; Phillips 1972; Wax et al. 1964.) Ultimately, the incongruence between Anglo teaching styles and Choctaw learning styles (Gibson and Kwachka 1978) when combined with a lack of opportunity to use English outside the classroom, a lack of necessity resulting from economic and geographic isolation; cultural beliefs about language; and a negative historical experience with English-speaking people; all produce an extremely unfavor able learning environment. While the obvious remed y for the conditions described above, at least in terms of academic improvement, is to teach content, or subject areas in Choctaw and structured,

PAGE 42

28 oral English, it required a Supreme Court decision (Lau vs. Nichols, 1974) to make possible the implementation of the 1968 Bilingual Education Act (Title VII) of the 1965 Elementary and Secondary Education Act (U.S.C.C.R. 1975:180). This decision declared that children with limited English are thereby denied access to equal educa tion and that school districts must provide for their education. As a result of this decision, monies became available to minority language groups to institute bilin gual education. The Choctaws began receiving Title VII funds in 1975. The Bilingual Education for Choctaws of Mississippi (BECOM) program was established on the reser vation and, in the following year, a complementary program, the Bilingual-Bicultural Teacher Education Project, was funded at nearby Mississippi State University to certify Choctaw students in bilingual, elementary education. The programs have experienced mixed success. Chief obstacles to implementation have been teachers entrenched in the previous system who are not academically prepared to teach in bilingual settings; and parents who are afraid that two languages in the schools will ''confuse" children and that the instructional use of Choctaw will hamper the children's acquisition of English. Nonetheless, initial testing is promising. If the programs continue, education ma y gain a more integral and positi v e p ositio n in Choctaw culture with ultimatel y salutar y e f fects on the culture's

PAGE 43

future viability in the larger society, a society which seems to have become, lately, a bit more accepting of cultural pluralism. 29 Turning now to testing, the next section will describe the instrument used to measure the linguistic parameters of elementary school children's spoken English. 1.4. The data base Each year in November, the BECOM program administers a test of oral English proficiency to all students in grades Kindergarten (K) through Three (3). The test is tape-recorded for later scoring. It is the language re corded during the 1977 administration of the test which provides the data for this study. Language testing, especially of oral production, is, at best, a delicate proposition. If the test is highly structured, i.e., there are specific responses or struc tures to be elicited, the testers cannot be certain that they have accurately sampled the respondents' productive competence (Silverman et al. 1976). On the other hand, if the test is more flexible and allows the children to respond freel y thus displaying their productive range, then it is difficult to develop and apply uniform proce dures for scoring. The test selected b y the BECOM pro g ram suffers from the first problem since it attempts to elicit v er y specific mor p hological and syntactic items.

PAGE 44

30 1.4.1. The SWCEL test The SWCEL was developed by the Southwestern Coopera tive Educational Laboratory to accompany an ESL teaching program (Livero n.d.). The test has been judged appropriate, both technically and culturally, by such diverse groups as Chinese, Navaho, and Spanish as well as by the Choctaws (Locks et al. 1978). The test in its entirety is presented in Appendix A. The SWCEL is composed of three subtests. The first deals with vocabulary, the second with pronunciation, and the third with morphology and syntax. The total instru ment is individually administered, requiring about fifteen to twenty minutes per child. In the vocabulary subtest, children are required to identify 24 three-dimensional objects; the labels for these items serve as the stimuli for the pronunciation subtest. The third subtest, mor phology and syntax, is composed of a series of pictures and games, controlled to elicit a restricted set of structural responses. The vocabulary-pronunciation sub tests have a total of 26 stimulus items, and the syntactic subtest a total of 57 items. For purposes of this study, the phonological-lexical portions of the test were not subjected to analysis. First, playback quality was frequently too poor to discriminate absolutel y between phonemes, e.g., / 9 / and / 1 / and, in any event, acquisition of phonology was not crucial to the analysis of morphology and syntax. Secondly, mastery of

PAGE 45

discrete lexical items is, perhaps, a peripheral, even mechanical aspect of language acquisition and uninter 21 est1.ng. 1.4.2. Administering the SWCEL The test was administered in November, 1977, to all students in Grades K through 3 in the six Choctaw elemen tary schools. Native English speakers on the BECOM staff gave the test. 22 Testing took place in the schools but, because accurate scoring depends on the quality of the recording, in areas isolated from the general activities 31 of the schools. Occasionally it was necessary to adminis ter the test in lunchrooms or auditoriums and, in these cases, tape quality is generally rather poor but sufficient to score the majority of structures. Because the children's grammatical proficiency is scored on the basis of their responses, it is extremely important to promote a relaxed atmosphere and to encourage the children to talk freely. The testers made every effort to do so; for those who were shy, the obligatory "warrnup" period was extended; for those with little English, the testers allowed the children to take as long as they wished to respond. Most of the children enjoyed taking the test and regarded it as a "play" situation. 23

PAGE 46

32 1.4.3. Scoring the SWCEL The standard method for scoring the SWCEL, described in the first section below, produces a single, composite score which does not allow an evaluator to examine specific linguistic or sociolinguistic variables. For this reason, a different method, described in the second section below, was developed for purposes of this study. 1.4.3.1. The standard scoring method The standard scoring method reflects three measures: the completeness of the response; the grammaticality of the response; and the spontaneity of the response. A maxi mum of 171 points is possible on the syntax portion of the test using this method of scoring. Each response is scored as follows: 24 a complete, spontaneous response receives 3 points; a minimal (not complete), spontaneous response receives 2 points; a prompted (imitated), complete response also receives 2 points; a prompted, minimal response receives 1 point; and no response, or a response which does not include the desired structure, receives no points (Southwestern Cooperative Laboratory, Inc. 1971:4) Prompting takes place if the children: a) do not respond v erbally, i.e., nod or shake their heads, or point; b) if the y gi v e a one-word response where a more complete response is possible; c) if the y gi v e a response which does not contain the grammatical element being evaluated;

PAGE 47

33 or d) if they give an ungrammatical response not containing the element in question. The method of prompting requires of the children a bit more than sheer imitation for some questions. For example, if a child fails to answer the question "What do you like?" the tester prompts by saying "Tell me that you like soup." In order to produce an appropriate, grammati cal statement, the child must make the necessary reference change in the pronouns: "I like soup." Nonetheless, the task is primarily imitative and the principal problem with the standard scoring method is that it does not differen tiate between such imitative responses and spontaneous re sponses. It is therefore conceivable that children could score 112 points on the syntax portion of the test without having produced a single, spontaneous utterance. Since this score falls within the range expected for competent English speakers (Lewis et al. 1977), assigning children to this category who have only imitated the interviewer would seem to contradict the purpose of testing: that is, to assess the children's ability to generate language. Therefore, a second method for deriving a score was de veloped for this research. 1.4.3.2. The revised scoring method This method allows the relationship between sponta neous and imitative ability to be examined and also pro vides a means for evaluating comprehension apart from

PAGE 48

grammaticality. By this method, a grammatically correct, spontaneous response of any length receives 2 points; if the response is a complete sentence, a point is added; if the response is a single word, a point is subtracted; and f h . 25 h h 1 t e response is appropriate, wet er or not it is grammatical, it receives a final point. The maximum pos sible score for spontaneous responses, by this method, is 228 points. Imitative, or prompted, responses are scored separately and similarly. 34 Because the standard scoring method is employed by the BECOM program and the scores are reported to the schools for assessing student progress, one of the first questions this study addresses is whether or not the method is actually indicative of the students' ability to sponta neously generate grammatical English as measured by the scoring method employed for this study. In order to do this, the two sets of scores were examined for a correlational relationship. It was found that the correlation is relatively high (.83) so, although they measure dif ferent variables, both scores can be considered to reflect the children's linguistic competence. 1.4.4. Data recording and analysis Because neither method of scoring provides any infor mation about what the children actually say, whether correct or incorrect, spontaneous or imitated, it was necessary to transcribe and code everything the children said for

PAGE 49

analysis. The size of the data base, some 12,000 utter ances ranging in length from single words to paragraphs of discourse, necessitated the development of special techniques for recording and retrieving the specific types of acquisition information under investigation. Analyzing such massive quantities of data is not typical of most linguistic studies, traditional linguists relying on a few major consultants and transformational linguists on individual (frequently their own) competence. 26 However, in recent years both sociolinguists and psycho_ linguists have dealt with large corpuses. Because this has required innovative approaches to analysis not pre sented in most linguistics programs, it is rather sur prising that there are relatively few published accounts concerning the special problems, methodologies, and pro cedures involved. (For exceptions, see Fasold 1972:31; Shuy et al. 1968.) The method of recording and retrieving the data for this study is reported here in detail for the following reasons. First is the fact that the method of recording and coding determines what is available for analysis and therefore defines the inherent limitations of that analysis. Second is the lack of such accounts and their potential usefulness to the interested researcher. Thirdl y the method developed has proven to be v iable and productive for future research. 35

PAGE 50

36 1.4.4.1. Recording the data Each tape was listened to by theinvestigator and the information for every response recorded exactly as the child produced it. Standard English orthography was used except in those cases where the phonetic realization pro vided information about the stages of syntactic or mor phemic development. To facilitate the recording process, a response form for each question was designed. The form has the expected, correct response printed across the top of the page, followed by a list of the more frequent errors. What the child said, or did not say, was recorded by entering a mark beneath the appropriate word, morpheme, or error column. A system of checks and numbers dis tinguished spontaneous correct from incorrect responses, and these from prompted correct and incorrect answers. Additional responses and unclassified utterances were recorded in a space provided at the right hand margin. These were also identified as spontaneous or imitative. A sample response form with explanations may be found in Appendix II. Each response, furthermore, was preceded by a number code which identified the student, grade, school, and test question number. The identification code was followed by the basic information for deriving the second method of scoring described above. This consisted of: 1) whether or not the child responded; 2) whether or not a sponta neous response was grammatically correct; 3) whether or

PAGE 51

37 not a spontaneous response was appropriate; 4) whether or not the response was a single word only; 5) whether or not the child imitated correctly if prompted; and 6) whether or not a complete sentence was generated, spontaneous and imitative sentences being differentiated in this category by the use of numbers. Ultimately, the information for each response was individually punched on a card, transfered to tape, and d f 1 1 27 store in a computer or ater ana ysis. This highly mechanical and laborious method of recording was arrived at after considerable trial and error accompanied by an increasing appreciation for the limitations of computers as instruments for recording and processing linguistic information. 1.4.4.2. Coding the data The ingenuity which the children applied to the lan guage learning task resulted in such an immense variety of d 1 28 creative productions, both grammatical an ungramrnatica, that it was impossible to predict the majority of them or, consequently, to provide categories on the response forms which would allow a system of checks to record them. It was therefore necessary to formulate a method of cate gorizing such non-expected responses (recorded in the right margin of the form, as previously described) so that the linguistic information could be separately ex tracted for study.

PAGE 52

To accomplish this, general categories (verbs, pro nouns, etc.) were selected on the basis of their relative frequency of inclusion or exclusion in the children's utterances. These categories were each assigned a dif ferent pair of numbers, one of the pair signifying a cor rect response, the other an incorrect one. Verbs, for example, were assigned the numbers "l" and "2". Several of the general categories were further specified by ap pending a series of subtopics, or qualifiers, which were identified by alphabetical labels. In the case of verbs, the subtopics are present tense, past tense, future tense, third person singular present tense, auxiliary, 38 copula, omission, and contraction. the complete coding system.) (Appendix III presents By using one or the o~her of the number pair, correct responses could be distinguished from incorrect responses. For example, and again with reference to verbs, if a child said "He big", the omission of the third person singular, present tense copula would be coded as follows: lAEGH; 1 = verb, incorrect; A= present tense; E = third person singular; G = copula; and H = omission. If, on the other hand, the child produces "He is big", the grammatical response would be indicated by the code: 2AEG; 2 = verb, correct; and so forth. In this fashion, it was possible to categorize the ma j orit y of utterances and subsequentl y retrie v e them for analysis. However, since language alwa y s eludes complete

PAGE 53

39 description, there was an ''other" category for those pro ductions which defied description. They were retrieved and examined separately. 1.5. Data Presentation, Objectives, and Justification The amount of discrete data available from the tech niques just described is tremendous. In order to summarize the numerical information, the succeeding chapters contain a great many tables and charts. These are, in all cases, accompanied by prose explanation and discussion. The actual presentation format follows necessarily from the objectives, presented below. 1.5.1. Research objectives The general goal of this research is to provide base line, or descriptive data in an area of increasing research interest, second language acquisition. On the basis of the data collected to satisfy this primary descriptive goal, several corollary questions of theoretical interest may be addressed. These follow the first, descriptive objective below. The objectizes are: I. To determine the extent of acquisition of selected morphological and syntactic structures achieved by Choctaw children over a period of four years. II. To delineate sociolinguistic factors in the chil~ dren's environment affecting L2 acquisition.

PAGE 54

40 III. To examine the relationships between imitative and spontaneous success in the L2. IV. To examine the language learning strategies em ployed by the children. V. To compare the acquisition of English as an L2 with the acquisition of English as an Ll. The first objective is discussed in Chapters Three and Four. Chapter Three is concerned primarily with morphology and auxiliaries, and Chapter Four with syntax. The second objective is pursued in Chapter Two. The variables dis cussed in that chapter are language domains, communicative style, and the children's sex, age, and community of resi dence. The final three objectives require some intro ductory comments explaining their relationship to the descriptive data on which they are based. With regard to Objective III, it is important to note that imitation is an important component of the behavioral model of language learning. Thus, Objective III overlaps, to a certain extent, with Objective IV. However, one of the problems in dealing with imitative data is the dual interpretations attributed to the ability to successfully imitate. Some, particularly behaviorists, see imitation as a tool for creating underlying competence, i.e., as a learning strategy which assists in the development of habits. Others see the ability to successfully imitate as a sign of underlying competence, whether or not the imitated structure has surfaced in the form of spontaneous oral

PAGE 55

41 production. It is beyond the scope of this study to determine if imitation is being employed as a learning strategy; however, we can determine if there is a relationship be tween successful imitation and successful spontaneous pro duction. In order to accomplish this, the conclusions of Chap ters Three and Four will compare developmental success rates of imitative and spontaneous oral productions. These comparisons will be interpreted on the basis of the fol lowing, very tentative assumptions. First, we will assume that if the ability to imitate exceeds the ability to spon taneously produce an inflection, and if this relationship remains constant across grade levels while both abilities increase, then we can conclude that imitative ability is an indicator of underlying competence. If, on the other hand, imitative ability consistently exceeds spontaneous ability and the latter does not increase over time, then we can conclude either that the two abi lities are not related, or that, if they are, imitation is not a successful learning strategy for the structure in question. Finally, we might also reach this second conclusion if spontaneous ability exceeds imitative ability, at any point in time; or if the two do not display a systematic relationship over a period of time. These assumptions will serve merely as interpretive guidelines and are certainly neither inclusive nor

PAGE 56

42 conclusive. The chief objective is simply to examine the data for the existence of a systematic relationship between the two productive modes in the children's L2. In the investigation of learning strategies, Objective IV, the most accessible means is the children's emerging L2. It can be described as a system, though one understood to be only heuristically in stasis, and its consistencies and discontinuities evaluated for evidence of the strategies and organizing principles that produced them. This ap proach is not within the scope of this research and, in fact, the data, as we will see, do not permit of a uni fying grammar or grammars. Alternatively, the emerging L2 can be regarded as an approximation to the target language, with the errors pro duced providing evidence for the strategies being employed to close the distance between the approximation and the target. These errors may result from Ll interference, established by contrastive analysis of the Ll and the L2; or from overgeneralizations of hypotheses based on learner analysis of L2 data. The two types of error, thus, imply different learning strategies and different theories of cognitive organization and operation. Chapters Three and Four contain sections dealing with the analysis of errors and the implications of these analyses for the behavioral and cognitive models of acquisition. Another approach to the evaluation of learning theories is to compare success rates of these children

PAGE 57

43 with those of speakers from different language backgrounds learning ESL. We would expect the rates or sequences to be similar if it is the L2 or inherent cognitive capacities which determine learning strategies. We would expect a variety of dissimilar patterns to emerge if, however, it is the Ll that dominates in structuring acquisition stra tegies. Therefore, acquisition sequences and success rates will be compared, where data exist, in the conclusions of Chapters Three and Four. Finally, with reference to Objective v, a comparison of Ll and L2 acquisition of the same language allows us to explore the question of whether or not L2 learning differs significantly from Ll learning. Again, there are implica tions for learning theory, and they are exactly the same as those outlined in the preceding paragraph. These will also be discussed in the chapter conclusions. 1.5.2. Data presentation format Summarizing the above discussion, Chapter Two presents sociolinguistic factors in the learning process; Chapters Three and Four present spontaneous and imitative production success rates, followed by error analyses and their impli cations for learning strategies and models, concluding with comparative acquisition sequences and summaries; and Chap ter Five reviews the methodology and discusses implications and applications of the research findings.

PAGE 58

44 1.5.3. Justification for research objectives In view of the interest in and importance of under standing processes of language acquisition, both Ll and L2, descriptive studies have great value in their implications for theory and pedagogy. Neither of these can develop in the absence of data, but we have relatively little available and that which we have comes, primarily, from ESL learners from the world's major languages. Data from less well known languages are of especial value in that they may pro vide answers to old questions, may even make it possible to ask new and interesting ones, and expose previous conclu sions either to validation or to reevaluation. Although this research breaks no new theoretical ground, it presents anomalies inexplicable in the framework of any current, single theory suggesting that we are still rather removed from even the first level of theoretical adequacy, the descriptive capacity. Our language pedagogy is equally inadequate. The generally accepted explanation for the lack of successful classroom L2 learning is that a second language is largely irrelevant for many learners; yet, the same charge of irrelevance can be leveled at much of what is successfully learned in the classroom, so we must look for less sim plistic explanations. In order to do this, we must begin with a clear delineation of what is and what is not learned, and an understanding of the factors related to the suc cesses and failures. That beginning is the purpose of this study.

PAGE 59

45 1. 6. Notes 1. Carol Chomsky (1969) demonstrates that, contrary to the general belief that by the time children reach school age their linguistic systems are fully developed, some syn tactic structures may not be mastered until the children are ten years of age or older. It is possible that some extremely elaborate levels of language, e.g., some English gapping rules, may only be acquired through extended literacy. 2. For descriptions, see Brinton 1870; Byington 1915; Haas 1941, 1971; Nicklas 1972, 1975; Powell 1966. 3. The acquisition of English as an L2 has been examined for a variety of other language groups, e.g., Norwegian (Ravem 1974a, 1974b), Spanish (Dulay and Burt 1973; Nata licio and Natalicio 1971), Japanese (Milon 1972); see Hatch (1978a) for summaries. 4. Dulay and Burt (1973), for example, propose that ex posure to L2 data in a naturalistic setting is sufficient to guarantee learning and that teaching is, therefore, more or less irrelevant to the learning situation. 5. Bricker and Bricker (1974:437-441) refer to the various debates as "pseudo-issues", and one need not accept their own theoretical position to agree with the intent of their argument which is that the resultant friction obscures rather than defines issues. 6. In fact, Labov lays part of the blame for the belief that ordinary speech is ungrammatical on linguists whose conclusions may have been reached after listening to each other converse at conferences See Braine (1971:170) for further discussion. M. J. Hardman (personal communication) has pointed out to me that her extensive text collection "shows, in casual recordings, virtually complete grammaticality with lapsus, errors, or mind-changes clearly marked by intonation or other linguistic/paralinguistic marks." 7. Brown and Hanlon (1970:205) note that while frequency of adult usage may be reflected in the child's production, where the structure is beyond the child's developmental stage, it will "become lodged in his [sic] speech as an unassimilated fragment." 8. It is possibly not without significance that this is exactly the method followed by field linguists in eliciting and analyzing an unrecorded language.

PAGE 60

46 9. For adults, however, this may not be the case. See, for example, Ervin-Tripp (1971:190-191); Sachs (1976:146). 10. Martin lists several languages for which there are no child speakers. 11. For example, the Menominee of Wisconsin (Wisconsin Native American Language Project 1974) and the Mohawk of New York state (Mithun and Chafe 1979). 12. Bauer (1971) summarizes the few exceptions to this policy. 13. The English that was learned generally contained stigmatized, nonstandard dialect features (Dubin 1970; Harvey n.d.; Leachman and Hall 1955) and thus maintained, and even contributed to negative stereotypes. 14. Parents who themselves experienced failure in school and employment difficulties and who, also, were taught assimilation to be inevitable are, in some extreme cases, so insistent that their children learn English that they speak only English to them. The effect on the children is more harmful than helpful; the English children learn from a limited speaker is also limited and usually marked by deviant grammatical constructions. The children's develop ment in Choctaw is likewise hampered, their modeLs being their peers who, as children, do not control the full lin guistic competence of the mature speaker. While dismissing the notion of cognitive "deficiencies'' (Bereiter and Engel mann 1966; Bernstein 1964), there still remain individuals who, as a result of the just described language environment, have difficulty communicating through linguistic mediums. 15. The average score on the Michigan Test of English Language Proficiency for one group of Choctaw college freshman was 70.4. 16. Several Choctaw students were found to have completed the majority of their degree requirements with the excep tion of the required English courses. Ability was not the problem, obviously, since their English, if sufficient for upper level college courses, was certainly sufficient for freshman level composition courses. English was simply avoided for as long as possible. 17. A person who is racially Choctaw but who does not speak the language is regarded as an anomaly and, of neces sity, alienated from normal, social interaction since this interaction is almost exclusively conducted in Choctaw. 18. The word nahollo translates literally as 'one who is greedy, or grasping for material things'. From the Choctaw

PAGE 61

47 point of view, this epitomizes Anglo character. Not sharing, in fact any accumulation to excess (whether it be of a material, intellectual, or spiritual nature) is frowned on by the culture. One who is nahollo is regarded with suspicion and the term is used not only as a pejorative for Anglos but also to admonish children and to chastise adults. The term "Anglo" also deserves comment. Many Native American communities simply distinguish two groups for general purposes: Indians and non-Indians. The Choctaws, however, distinguish four: whites (whom they call Anglos when avoiding the pejorative); blacks (takossa, a contrac tion of hattak ('men') and lossa ('black'); other Indians; and Choctaws. 19. The obverse implication of the language and culture identification is that if Anglos speak Choctaw then they become identified with the Choctaws. Since this is unac ceptable, given the Choctaw attitude toward Anglos, any attempt by Anglos to learn Choctaw is strongly discouraged. There is only one local Anglo who has learned, or been allowed to learn, the Choctaw language. The protective attitude toward their language (an attitude shared, inter estingly, by another Muskogean group, the Miccosukee), is frequently expressed in variants of the statement: "They got our land; if they get our language we won't have any thing left." 20. Only in the past decade have the schools stopped ad ministering physical punishment for speaking Choctaw in the classrooms or on the school grounds. 21. 26 vocabulary items could hardly encompass relevant material items in any culture, although if chosen carefully, a small number might be sufficient to show acquaintance with specified semantic domains. A modified Swadesh list might be minimally useful, although its relevance to lan guage acquisition has been neither suggested nor explored, to this writer's knowledge. 22. The writer, who at the time was the ESL Specialist for the BECOM program, participated in the administration of the previous year's testing. 23. One child, on seeing me in the reservation store not long after I had tested her, pointed me out to her parents as "the lady I played games with at school." 24. In most cases, "complete" indicates a sentence re sponse. In some cases, the response is a cloze task requiring the child simply to supply one or two words; these were also scored "complete".

PAGE 62

48 25. Inappropriate responses included both those that were irrelevant and clearly indicated the child did not compre hend the question, and those, correct or incorrect, in Choctaw rather than English. The latter were further la beled so they could be examined separately. Considering the degree of Choctaw language dominance among Grades K and 1 children, the rate of Choctaw response was surpri singly low. The explanation for this probably lies in the diglossic situation which obtains on the reservation. Situations involving Anglos demand English or silence. Responses that were inaudible because the children mumbled or otherwise consciously and purposefully obscured their utterances were also scored as inappropriate. If the response was inaudible due to other factors (competing sounds, etc.) the response was omitted from the corpus. Individual scores were adjusted to avoid penalization in this circumstance. 26. As noted by Shuy et al. (1968:v), "Historically, linguists have formulated theory from individual rather than group performance." 27. The Trammel Computer Science Center facilities at Mississippi State University were utilized. Funds to cover the cost for use of the computer were generously granted to the writer by the Office for Research and Graduate Studies at the University. Grateful acknowledgment is extended to that office and to Gerald Matthews, Associate Professor and Coordinator of Computer Studies, who gave freely of his time and patience in translating linguistics-ese to computer-ese and who programmed this entire study. 28. One child, struggling with locative prepositions, could not remember "under" to describe the location of the toy pig used in testing. She extricated herself with grammatical honor by producing: "My hand is over the pig."

PAGE 63

CHAPTER TWO LANGUAGE IN THE CHOCTAW COMMUNITY 2.1. Introduction This chapter examines important sociolinguistic factors in the Choctaw culture. In the previous chapter, discussion emphasized the significance of the Choctaw language to Choctaw identity, the culture's negative atti tude toward Anglos and, because of the synonymity of lan guage and culture in Choctaw thought, the resultant nega tive attitude toward the language of Anglos, English. This chapter begins by discussing a corollary of these factors; that is the rigid division of language domains, or diglossia, in Choctaw life; included in this discussion is an examination of the manner in which communicative styles reinforce domain distinctions. The following sections of the chapter examine three sociolinguistic variables, community of residence, age, and sex of the children, in order to determine if there is a relationship between these and English L2 acquisition. 2.2. Language Domains and Communicative Styles Anthropologists and linguists take for granted the inextricable link between language and culture; however, 49

PAGE 64

50 individual cultural groups may not themselves recognize this link. The Choctaw, as we have seen, are an exception. They are an intensely language-conscious people, and, on the reservation, 1 they maintain an absolute separation of English and Choctaw language domains. The dichotomy is very rigid and rarely violated. The Choctaw language domi nates, but there are specific contexts in which the use of English is deemed not only appropriate but obligatory. The three major English contexts are described below. The most visible context is tribal council meetings. These are conducted along formal, parliamentary lines and entirely in English. This is despite the fact that all members, and the majority of those attending, speak Choctaw at least as, and probably more, proficiently than they do English. Very occasionally an older person will address the council in Choctaw, but such occurrences are exceed. 1 2 ing y rare. English is also used almost exclusively at school functions and when interacting with school personnel. Some Anglo teachers who have spent many years in the Choctaw schools are known to partially understand the language, but, despite this, they are never addressed, either by children or adults, in Choctaw. Because the schools are a major context for Choctaw Anglo interaction, English language use has extended to any Choctaw-Anglo interaction. Even in the unusual event of an Anglo being taught Choctaw, people find it very difficult,

PAGE 65

51 even impossible, to address the learner casually in Choctaw, and invariably respond in English if the learner addresses them in Choctaw. 3 The final context in which English is always used is athletic events, a major reservation activity. Announce ments and play descriptions, even for the Choctaw tradi tional sport, stickball, are always made in English Players' disputes with referees are also, initially, in English, but if the altercation becomes involved, are continued in Choctaw. Nonetheless, the announcement of the referee's decision is, again, in English. On examining the contexts in which English is used, it is clear that they exhibit two common features: they in volve speech of a public nature, and/or they are imported institutions. 4 There is, however, one surprising exception to this generalization and that is the church. Choctaw ministers conduct services in Choctaw, hymns are sung in Choctaw, and the Bible read in that language. Even in cases where the minister is non-Choctaw, hymns and prayers are offered in the congregation's first language. The explanation for this may lie in the manner in which Christianity was first introduced to the Choctaw. Missionaries developed a Choctaw orthography and conducted bilingual schools. Later, man y Choctaws themsel v es became missionaries and ministers and pla y ed an acti v e role d uring the period of removals (Debo 1934; M orrison 1975; Peter son 1970). Perhaps, then, church came to be regarded by

PAGE 66

some communities as an integral part of their culture and not an alien institution. In any event, the Choctaw language is, traditionally and currently, appropriate in religious contexts. 52 Other domains of Choctaw language use are not diffi cult to define. Choctaw is spoken everywhere but in those contexts just described. The language is spoken at home, at play, and at work. Because most Choctaws live and work on the reservation, it is therefore the case that occasions for adults to use English are relatively infrequent. Thus, the inability to speak fluent English is not particularly disadvantageous, but the inability to speak Choctaw is extremely limiting. The inability to speak Choctaw threatens not only individual identity, as previously noted, but community membership as well, since each Choctaw community has defining, idiosyncratic speech patterns. Community member ship is important. Each community has a traditional repu tation which functions to orient expectations if one is introduced to an unknown person. 5 Not only is there alle giance to community, there is also allegiance to that community's speech which is regarded as superior to others whose forms are, more or less good-naturedly, criticized and ridiculed. Individual, as well as community speech differences are recognized. Although "showing off" is strongly dis approved of, good speech-makers and debaters are admired.

PAGE 67

53 The Choctaw enjoy a well argued verbal presentation whether or not they agree with the speaker's position. One's ability to use the language well, however, does not result in that person's being regarded as either an authority or an arbiter, since "experts" of any variety are routinely despised. The Choctaw function on the principle that no single individual embodies the totality or range of lan guage implicit in the community as a whole. Thus, it is the community, not the individual, that dominates the cul ture's social and psychological orientation. Just as some speech is admired, some speech is de plored, primarily the "Pearl River slang" spoken by the high school teenagers. 6 This speech is marked by a great deal of contraction, slang expressions, and a high percen tage of English loan words. Criticism of this speech is of a quite different tenor than the criticism arising from traditional community dialect rivalry. Indeed, there is some concern among older people that the current generation of teenagers will not return to adult norms and "degenera tion" of the language will occur. Besides being aware of language domains and levels of speech, the Choctaws also recognize social "rules" which relate to speech events. The rules for Choctaw conversa tion in the context of social interaction are quite dif ferent from those of English and there is no question but that the differences reinforce and entrench language diglossia. Choctaw conversation in social situations, as

PAGE 68

compared with Anglo, is characterized by low volume, long pause lengths in conversational turns, and different kine sics. Teasing, joking, and sophisticated, often risque, allusions mark conversation. Humor is much admired and appropriate to almost every conversational topic. Although conversation is a major feature of social intercourse, it is not, as is the case with Anglos, a necessary concomitant. For example, visiting, a regular social activity, may be silent. After it has been estab lished that a person's arrival is for social rather than 7 personal reasons, the visitor may sit quietly or join in 54 household activities. Visitors do not expect to be "enter tained" by conversation and a cessation of normal routines. There is no demand placed on either the visitor or the host, with the single exception of food. Visitors arriving during meals are expected to eat. Silence has many communicative levels. That a visi tor's silence is social can be detected by kinesics. A social silence is marked by relaxed body posture; eye move ments follow whatever activities may be ongoing. A thought ful, considering silence, e.g., while formulating a re sponse, is communicated through a more rigid body posture, the eyes focused on some point in the surrounding environ ment. An angry or nervous silence is accompanied by lowered shoulders and head, the eyes directed briefly and alternately to and away from the source of displeasure.

PAGE 69

Kinesics accompanied by silence are also used to com municate directions; turns off a road, or the location of an object are indicated by a slight movement of the head or hand with no verbal elaboration. 55 The very different characteristics of Anglo social and communicative patterns make Choctaws uncomfortable. The first difference is the fact that an Anglo social context is obligatorily marked by conversation, and by conversation that is, to Choctaw participants, too rapid to allow re sponses. The volume level is very loud to the Choctaw ear and the swiftness and compulsive exchange of conversational turns contributes to an overall impression of incessant noise. Not only is silence not allowed in Anglo conversation, it is interpreted by them as withdrawal or unfriendliness. Moreover, the use of eye contact by Anglos makes the Choc taws feel as if they are being stared at, i.e., as if they are isolated as objects of curiosity rather than included as social co-participants. Even the polite, conventional opening gambits of an Anglo conversation immediately violate several Choctaw practices: one does not utter one's own name; nor does one question another about matters which are both personal and, from the Choctaw perspectiv e irrelevant to social izing, e.g., "What do you do?" A final social interaction difference is that a group of Anglos frequently follows the decision of a dominating

PAGE 70

56 individual after a period of verbal discussion of alterna tives; whereas a group of Choctaws remains a group only as long as each individual voluntarily, without discussion, chooses to continue the interaction. A group member may withdraw without announcing either intention or explana tion. There is no coercian, especially not verbal, to follow the departing member. (See Witherspoon (1977:83) for similar patterns among the Navaho.) Finally, when engaged in conversation with Anglos, further discomfort results from a general awareness that the English spoken on the reservation is nonstandard and, because Choctaw people expect criticism for nonstandard behavior in and by their own community, they project this anticipation to Anglo communities, and are therefore re luctant to talk. In sum, Choctaws participate minimally in English language conversations, and, when they do, they experience varying levels of discomfort. The divergences in social practices related to language use are, cumulatively, ex treme. They intensify and transcend the simple, sheerly linguistic problems of appropriate language selection and basic, grammatical communication. 2.3. The Communities and English L2 Learning As a result of shifting federal policies, the Choctaw are located in separate communities, rather than on a single reservation. The communities vary in size, the largest

PAGE 71

57 having a population of around 1,200 persons (Spencer et al. 1975). They are scattered over four northeastern Mis sissippi counties, the central and largest community located in Neshoba County. All of the communities consi dered in this study are within 40 miles of Pearl River, the central community and seat of tribal government. 8 Each of the communities not only has its own particular speech patterns and reputation, as mentioned above, but also its own history and relationship with surrounding Anglo com ture. Since it is conceivable that any or all of these might affect L2 acquisition, 9 community of residence has been selected as a variable for examination in this study. Unfortunately, there is no precise data on either dialect or historical differences. With reference to dia= lect differences, it is believed that originally there were three major dialect groupings corresponding to political regions: the Okla Falaya or Longtown dialect in the west; the Ahi Apat Okla or Potato Eaters dialect in the east; and the Okla Hannali or Sixtown dialect in the south (Jacob et al. 1977:49). These three may still exist in both Okla homa and Mississippi (Nicklas 1972:2) but substantiation of this point awaits further research. There is no question but that the communities exhibit dialect differences, but the differences ma y be onl y super ficial, i.e., le x ica1 10 and / or phonological. Bogue Chitto speakers, for example, drop word final / h / ; Cone Hatta people are thought to speak a more conservative dialect

PAGE 72

than others, i.e., there is less contraction of suffixes; and some communities contract or drop suffixes more than others. Contraction may or may not be a superficial phenome58 non. In casual speech, it is a widespread practice. However, there is evidence that, at least for some speakers, the recovery of noncontracted, or even intermediate forms is not possible, suggesting that several syntactic semantic relationships are unavailable in these speakers' grammar. Whether or not such a deep structure difference might present different L2 learning configurations depends on the degree of influence attributable to the Ll in L2 learning. With no more than minimal and somewhat impressionistic data on dialect differences, it is impossible to state that the language of any one community is sufficiently different to affect L2 acquisition, assuming Ll influence. The question must be approached indirectly; that is, we can examine the L2 data in terms of community of residence and determine if there is a consistent interaction between community and English acquisition. Such interaction might also be produced by the his torical and contemporary relationship of a community with the surrounding Anglo culture. Again, we are limited by incomplete data. Nonetheless, some communities have repu tations which reflect, to some extent, their histories. Bogue Chitto, for example, is recognized for its retention

PAGE 73

59 and maintenance of traditional customs. The introduction of churches and schools was strongly resisted by community members. Tucker community, on the other hand, is regarded as more "Anglo" than other communities. People believe this to be due to the fact that a strong Catholic mission was established there in 1883 and remained. It was the only church which both Anglos and Choctaws attended (al though seating was divided) and it was staffed by Anglos (Peterson 1970:183-187). Finally, some communities are recognized as the seats of various prominent families, and others, Standing Pine and Pearl River, for example, are linked by generations of kinship ties. In sum, there are a variety of factors related to community of residence which might affect L2 acquisition. While it is beyond the scope of this research to isolate and rank these factors, it is within the capacity of these data to discover the presence of a relationship between community of residence and several variables tested by the SWCEL. The results of this investigation are presented in the next section. 2.3.1. ESL acquisitio in the six communities The communities are compared on the basis of six para meters e x trapolated from the SWCEL test data and the dif ferences among the communities, on each parameter, tested to determine if they are significant at the .05 level

PAGE 74

of confidence. The results of these analysis are reported below. 2.3.1.1. Ranking of communities by SWCEL test performance The communities are ranked in Table 1, below, on the basis of the revised score (Score I; see Section 1.4.3.2 for method of derivation). The actual number in the table is a mean score based on all children of all ages in each community. This provides a general measure of comparison among the communities. Table 1. Communities Ranked by Score I Rank Score Community 1 56.30 Bogue Chitto 2 62.53 Pearl River 3 67.30 Conehatta 4 68.17 Standing Pine 5 69.34 Red Water 6 95.32 Tucker 60 Although there are differences among all the commu nities, only one community, Tucker, is significantly different from each and all the others. The Tucker children perform significantly better than the children of the other communities.

PAGE 75

To a certain extent, Score I and the degree of each community's dominance in the Choctaw language correlate. Table 2, below, compares the two, ranked by Score I. Table 2. A Comparison of Score I and Choctaw Language Dominance Community Score I % Ll Dominance Bogue Chitto 56.30 91 Pearl River 62.53 71 Conehatta 67.30 87 Standing Pine 68.17 91 Red Water 69.34 82 Tucker 95.32 73 61 However, correlation is not consistent so that degree of Ll dominance cannot be employed as a gauge of L2 profi ciency. 2.3.1.2. Ranking of communities by "Appropriate Response" This measure provides an indication of the children's comprehension of English and ignores the grammaticality of their response. It is, in other words, a measure of pas sive competence and, although it is not definitive (some children may choose not to respond even though they under stand), it is the only indicator of passive (versus oral)

PAGE 76

62 development that can be extrapolated from the testing data. The score in Table 3 represents the mean number of times each child, by community, responded appropriately. The maximum number of times possible is 57. Table 3. Ranking of Communities by "Response Appropriate ness" Rank Score Community 1 24.56 Bogue Chitto 2 26.38 Pearl River 3 27.50 Red Water 4 28.54 Standing Pine 5 28.81 Conehatta 6 35.97 Tucker Tucker community, again, is significantly different from all the others. If we compare the communities on the bases of both appropriateness and grammatical oral production (Score I), we find similarities at the lowest and highest extremes of the ranks: Bogue Chitto and Pearl River chil dren neither respond very grammatically nor comprehend well; Tucker children, on the other hand, perform well in both areas. However, in the middle ranks, there is some varia tion. Conehatta children have comprehensive skills which are superior to their grammatical productions; Standing

PAGE 77

63 Pine children are more or less average in both areas; and Red Water children do not comprehend well but respond gram matically when they do. 2. 3 .1. 3. Ranking of communities by "No Response" Another measure at least partially indicative of com prehension is whether or not the children attempt to respond to a test item. This measure is not so reliable as the previous one, since a variety of factors may enter into the children's decision, including shyness, degree of previous exposure to Anglos, illness, and so forth. Table 4, below, reports the mean number of items to which each child did not respond at a11, even after prompting. The maximum is, again, 57. Table 4. Ranking of Communities by "No Response" Rank Score Community 1 5.99 Tucker 2 10.80 Red Water 3 10.84 Standing Pine 4 11. 37 Conehatta 5 13.32 Pearl River 6 18.87 Bogue Chitto

PAGE 78

The communities at either end of the scale, Tucker and Bogue Chitto, are significantly different from the others, Tucker having the lowest number of "No Response" and Bogue Chitto having the highest number per child per test. 2.3.1.4. Ranking of communities by "One Word Response" Children may respond with single words for several reasons. In some cases, a single word is adequate, if minimal (viz. Question 27); secondly, their ability to 64 use English may not be sufficient to respond more fully; or, they may be uncomfortable using the language. There fore, it is impossible to determine e x actly what this measure reflects. It is, nonetheless, an index of "volubility" in the children's L2. The score in Table 5, below, is the mean number of times a "One Word Response" was given per child, by community. Table 5. Ranking of Communities by "One Word Response" Rank Score Community 1 15.07 Bogue Chitto 2 16.12 Pearl River 3 16.89 Red Water 4 17.98 Tucker 5 18.81 Conehatta 6 18.87 Standing Pine

PAGE 79

65 Here we find that Bogue Chitto children are the most voluble and Standing Pine children the most reticent. Since Bogue Chitto children are also the least likely to respond at all, and the least likely to respond either appropriately or grammatically when they do respond, this is a rather surprising finding. Tucker children who have been significantly different by other measures are more or less average on this one. There is no obvious explanation for these findings, but it is apparently not the case, as one might assume, that more proficient L2 speakers talk more. 2.3.1.5. Ranking of communities by sentence production Responses beyond single word utterances may or may not be complete sentences. In order to find out to what extent the children spontaneously generate complete, grammatical sentences, a separate index was prepared, Table 6, below. The tabulation includes ellipses, e.g., "It is'', and the score reflects the average number of times each child produced grammatical sentences during the test. Tucker is significantly different from the first three communities, Pearl River, Bogue Chitto, and Cone hatta, in the generation of grammatical sentences. The differences between Tucker and Red Water and Standing Pine are large but not statistically significant.

PAGE 80

66 Table 6. Ranking of Communities by Sentence Production Rank Score Community 1 1.51 Pearl River 2 2.00 Bogue Chitto 3 2.28 Conehatta 4 2.96 Red Water 5 3.10 Standing Pine 6 7.45 Tucker 2.3.1.6. Ranking of communities by imitative ability Finally, because imitation is a factor which may con tribute to L2 acquisition, the communities were ranked according to the children's abilit y to imitate correctly. The score in Table 7, below, reflects the number of times each child imitated correctly, either words, phrases, or d h 12 sentences, uring t e test. The last community, Tucker, is significantly different from the first two communities, Red Water and Bogue Chitto. Comparing imitative ability with overall performance (Score I), we find the results to be equivocal. Tucker has the highest rank in both measures so we might leap to the con clusion that overall proficiency and imitative ability are closely related; however, Red Water presents a somewhat different picture: a moderately high overall proficiency rating coupled with the least imitative ability. Even

PAGE 81

-------------------------------67 Table 7. Ranking of Communities by Imitative Ability Rank Score Community 1 4.28 Red Water 2 5.45 Bogue Chitto 3 6.46 Pearl River 4 6.92 Standing Pine 5 6.98 Conehatta 6 8.60 Tucker though these are relatively crude measures of comparison, they do suggest that the relationship between spontaneous and imitative proficiency may be neither straightforward nor simple. This question will be explored at length in following chapters. 2.3.2. Language in the Six Communities: Summary and Conclusions Two communities, Tucker and Bogue Chitto, are distin guished on the basis of the several methods of analysis. Tucker's differences are significant in the statistical sense and indicate that the children of this community acquire English much more successfully than the children of other communities. In the absence of other, discernible differences it is suggested that Tucker's access and exposure to English speakers, its distinguishing socio historical characteristic, is the responsible factor. In

PAGE 82

68 contrast, Bogue Chitto presents the opposite socio historical experience: the community has remained isolated from Anglo influence and resisted the importation of Anglo institutions. This research suggests that Bogue Chitto's stance is reflected in a reduced rate of L2 acquisition. Therefore, it is the conclusion of this research that expo sure to the L2, outside the classroom, is the single most important sociolinguistic variable in L2 acquisition as reflected by the variables isolated from the test. 2.4. The Correlation of Age and English L2 Acquisition By and large, the children of this study are five or six years old before they are exposed to English on a regu lar, interactive basis. Before this, their opportunities to hear the language are fairly numerous but sporadic and of a passive nature. Television is present in almost every home, shopping trips to nearby towns and off-reservation stores are conducted in English (by parents), and athletic and school events are announced in that language. With their parents and peers the children speak Choctaw and hear Choctaw. Thus, with the sole exception of school, most chil dren's opportunities to communicate in English do not change with age. Only when they enter the labor force or college may this picture change. However, while college necessarily requires English, work may not, since most people work on the reservation.

PAGE 83

69 As a result of these circumstances, any changes in the children's English ? roficiency with age are directly attributable to their s:hool experience. Until recently, this experience has be e~ primarily in English, but currently 13 the children are instr u :ted bilingually, Anglo teachers teamed with Choctaw ai c ~s. The latter are enrolled in Mississippi State Univ e =sity's bilingual teacher education program and will event u ~lly graduate certified in bilingual bicultural education. The BECOM program :n the Choctaw schools is a transi tional one whose goal is to move the children from 90 % content instruction in C hoctaw in Grade K to 100% content instruction in English : n Grade 4, with structured English instruction providing t i e bridge between languages. All the children in this study ; a rticipated in the BECOM program, and thus all children h cj received some instruction in English as a subject an c with the exception of Grade K, . 1 h 14 some content 1nstruct1 0~ in Eng is Therefore, the sco=es in Table 8, below, represent the result of English i ~ struction in the schools, and, moreover, we see that t~is instruction is successful to the e x tent that the differe ~c es between grade levels are sig nificant at the .01 le v ~l of confidence. 15 In sum, age, whic h in the Choctaw case correlates closely (Tucker and Bo g ~e Chitto, as we have seen, are the e x ceptions) with the a rn~u nt of L2 e x perience, is an indica tion of language devel o :ment and proficiency: the older

PAGE 84

Table 8. Grade Level and Score I Grade K 1 2/3 Score I 45.85 68.85 105.65 70 the children, the more proficient their English. This in crease in proficiency is primarily a product of instruction in and exposure to the L2 in the schools. 2.5. The Correlation of Sex and L2 Acquisition Scores for the two sexes were isolated and compared. The results were negative; there is no significant dif ference between male and female children in this study. Both sexes are equal in L2 proficiency as reflected by Score I. 2.6. Language in the Choctaw Community: Summary and Conclusions The preceding discussion has emphasized the limited role English plays in the everyday lives of the Choctaw people. Its use is appropriate in a very few, highly cir cumscribed contexts, and its concomitants, social and communicative styles of interaction, are not only different from but frequently repugnant to Choctaw customs. These r,

PAGE 85

71 factors, in conjunction with the antipathy to English speaking people noted in Chapter One, result in an extremely unfavorable environment for second language learning. In an attempt to determine if any of the six com.mu nities provides a significantly different learning environ ment from the others, various measures were analyzed, with the result that one community, Tucker, was found to favor second language development. The distinct and unusual nature of its relationship with English speakers in comparison with other communities was judged to be the distinguishing factor. Finally, it was found that age correlates positively with language acquisition, but sex does not. 2.7. Notes 1. The Choctaw lands in northcentral Mississippi are divided into six individual communities separated by dis tances of 17-40 miles of non-Indian lands from each other. The term "reservation'' will be used to refer to these communities collectively. 2. There is some evidence that the extreme contextual ization of languages is so deeply rooted that a mere con scious decision to switch languages is not always either sufficient or successful. For example, a friend of the writer decided to address the tribal council in Choctaw. Later, she reported that she, an elaborate and gifted speaker, had e x treme difficult y using the language; she had trouble finding the ri g ht words and hesitated in her speech.

PAGE 86

72 3. An Anglo attempting to learn Choctaw is extremely rare not only because it is discouraged by the Choctaws but be cause of the attitude of the Anglo community, as illustrated by the following anecdote, for which I am indebted to Kenneth York whose relative was the translator in the event. During the 1920s, a Choctaw man appeared in the local county court and put the judge much out of temper when it was discovered that it would be necessary to send for a translator; the man spoke no English. After some delay, the translator was secured and the judge began by asking whether or not the man was a lifelong resident of Neshoba County and, the answer being translated as "yes," demanded to know why it was the defendant did not speak English. The Choctaw man deliberated for some time, then asked the judge if he was a lifelong resident of the County. When the judge answered that he was, the defendant asked, "Then why don't you speak Choctaw?" 4. Council meetings and stickball games occurred in pre contact and early contact periods, but they differed greatly in form from those of today (Swanton 1931). 5. Because there are sanctions against speaking one's own name, they are not used in introductions and thus family relationships, a possible means of orientation, are not immediately available. 6. The only Choctaw high school is located in Pearl River and many students from the more distant communities board there. 7. Direct questions concerning personal matters are con sidered ill-mannered; moreover, requests for assistance are made only in dire necessity since the Choctaw have familial and routinized mutual assistance networks and since, beyond that, every individual is regarded as abso lutely responsible for her or his own actions and problems. Therefore, it is not only polite but necessary to ask visitors immediately if there is a problem. If the question is not put, a favor or assistance cannot be requested by the visitor. While younger people are less bound by this con vention, they still indirectly question visitors, e.g., "What brings you over here?" 8. Another community, Bogue Homma, is located in the southern part of the state, and still another group has settled around Ripley, Tennessee. 9. In addition, it was the impression of BECOM staff mem bers that the children of some communities spoke more pro ficient English, or were more willing to speak English, pro ficient or not, than the children of other communities.

PAGE 87

73 10. Swanton, in his introduction to Byington's dictionary, states: Anciently there were several Choctaw dialects, but only one of these, that of the Sixtown Indians living in the southern part of the old Choctaw country, differed to any considerable degree from the standard, or Longtown, dialect spoken in the Western part of the Nation. Moreover, this dif ference seems to have been confined mainly to certain words, involving but very slightly the language as a whole (Byington 1915:ix). 11. All scores and other numerical data have been rectified for interviewer error; i.e., children are not penalized for items interviewers failed to administer. 12. The reader is reminded that 24 of the 57 items require switching or pronominal reference as part of the imitative task. In almost all cases, the pronouns involved are "you" and "I". 13. At the time of this study, only two classroom teachers in Grades K-3 were Choctaw. 14. As can be expected with any new program, the quality and degree of BECOM program implementation varied greatly from teacher to teacher, and from school to school; that is, some children received more or less English instruction than others. There was, indeed, resistance on the part of many Anglo teachers, who felt their jobs threatened, but the presence and dedication of the Choctaw aides usually compensated for the reluctant teacher. As with bilingual programs everywhere, there were problems with curriculum materials, and with teacher training and retraining. None theless, the program was, in comparison with many others with which the writer is familiar, a model in its rigorous administration, evaluation, and actualization. 15. If Score I for Grades 2 and 3 is analyzed separately, the difference is significant at the .05 level only; and when standard scores for these two grades are compared, the difference is not statistically significant. The Grade 2/3 score in Table 8 is an average based on a Grade 2 score of 92.98 and a Grade 3 score of 118.32. As mentioned in later discussion, L2 growth appears to level at some point, and the minimal difference between Grades 2 and 3 may presage this phenomenon.

PAGE 88

3.1. Introduction CHAPTER THREE MORPHOLOGY This chapter begins the presentation, analysis, and discussion of the language data. There are three areas of focus: noun and verb inflections; pronominal case and gender; and auxiliaries and copulas. The format for dis cussion will be as follows. Each structure is presented individually and independently of the others. For each, the quantitative data are described. These data summarize the number of times the structure in question was obliga torily required in the children's utterances and compare this figure with the number of times the structure was actually and correctly supplied. These two figures combine to produce a "percentage of success" figure which is the basis for comparing the degree and rate of acquisition across grade levels. A success rate of 90% has generally been accepted as the criterion for classifying an item as "acquired" and will therefore be adopted for this study (Brown 1973:258; Cazden 1972:33; Dulay and Burt 1974a). Both spontaneous and imitati v e utterances are subjected to this anal y sis. 74

PAGE 89

Following the quantitative data is a discussion of the types of errors produced by the children. This, in turn, is followed by a discussion of two learning strate gies thought to affect second language acquisition: the transfer of Ll patterns; and active analysis of the L2. These will be evaluated with reference to their relevance to the acquisition of the structure under discussion. The chapter concludes with a comparative and cumula tive summary of the structures presented, a discussion of the acquisition pattern portrayed by the data, and a com parison of this pattern with those found in other acqui sition studies, both English Ll and L2. 3.2. Noun and Verb Inflections 75 Although the English language is noted for relatively rigid word order and a dependence on this order to indi cate role relationships within the sentence, inflectional morphology nonetheless indicates many important relation ships. Six inflections are examined in the following sections. The noun inflections are: the plural suffix {-S}; the possessive suffix {-S}; and the definite and indefinite prefixes {oa-} and {a-}. 1 All of these affixes are bound. The verb inflections examined are: the third person, present tense suffix {-s}; the progressive suffix {-ing}; and the simple past tense suffix {-D}.

PAGE 90

76 3.2.1. Noun inflections The semantic notions expressed by the three inflec tions discussed in this section are plurality, possession, and definiteness. The first two inflections, although semantically very distinct, are phonetically and phonologi cally exactly the same. Because of this homophony, we can determine on the basis of their emergence relative to each other whether the children's mastery is a result of phonetic-phonemic considerations or of syntactic-semantic ones. The emergence of the two relatively simultaneously would suggest acquisition to be, in this case at least, more dependent on superficial aspects of language compe tence, i.e., the mastery of phonetic and phonemic systems, with the implication that the semantic notions are already 2 present. Independent emergence would suggest acquisition to be dependent on more complex, syntactic-semantic consi derations, and their order of emergence interesting as an indication of which concept, plurality or possession, is more obvious, or "easier", or more compatible with Ll precedents. Unlike the plural and possessive inflections, those indicating definiteness are prefixed and their use governed by complicated co-occurrence restrictions both at the syntactic and discourse levels of language. Briefly, the definite and indefinite articles are obligatorily prefixed to the class of English nouns commonly referred to as count nouns. The indefinite article, however, is restricted

PAGE 91

77 from occurring with plural count nouns (plural indefinite ness being marked either by non-insertion of an article or by the inclusion of an indefinite quantifier, e.g., "some") and with mass nouns. Moreover, both indefinite and defi nite articles are usually restricted from occurring with proper nouns and nouns referring to body parts. These restrictions are exceedingly complex and, in effect, require the learner to differentiate among several classes of English nouns in order to express plurality and definiteness appropriately (see Brown (1973:340-356) for more extensive discussion). The following three sections consider each of the inflections in the order discussed above. 3.2.1.1. The plural suffix There are six opportunities included in the SWCEL test for producing the plural inflection (Questions 41, 55, 56, 73-75). The responses to these questions form the basic data pool. Added to these is every occurrence in casual conversation during the testing of an obligatory context for this morpheme. Table 9 summarizes the number of obli3 gatory contexts (opportunities), by grade level, the number of successful attempts, and expresses the children's correct use of the inflection as a percentage of their total number of opportunities. There is a 17 percentage point gain between Grades K and 2 / 3, indicating that the children are acquiring the

PAGE 92

78 Table 9. The Plural Suffix Grade # Opportunities # Successful % Successful K 1 2/3 157 280 234 59 38% 116 41% 129 55% inflection although it has not yet been mastered to the criterion point of acquisition. Table 10, below, suggests that age is an important variable in acquiring this inflec tion. Table 10. The Plural Suffix, per Child Grade K 1 2/3 # Opportunities 2.4 4.3 4.7 # Successful 9 1.3 2.6 From this breakdown, we can see that the rate of ac quisition increases dramatically with age. The number of obligatory contexts increases only slightly between Grades 1 and 2/3 (.4 opportunities), but the number of successful productions increases by 1.3, or three times the increase in the number of opportunities.

PAGE 93

79 In sum, the children produce the plural inflection spontaneously and successfully slightly more than half the time and the rate of successful production, relative to opportunities, makes a large increase between Grades 1 and 2/3. Imitative abilities differ markedly from spontaneous abilities. Table 11 reports the children's imitative success and Figure 1 compares this with their spontaneous success. Table 11. The Plural Suffix, Imitative Success Grade K 1 2/3 # Opportunities 41 61 35 # Successful 10 31 14 % Successful 24% 51% 40% As Table 11 indicates, the children do not attempt to imitate this inflection very frequently. It must be re membered that they were requested to imitate every time they produced an incorrect or an incomplete (non-sentence) response, so the column headed"# Opportunities" represents the number of times the children actually attempted to imitate, not the number of times they were requested to do so. In fact, if the rate of imitation is computed on a

PAGE 94

100% 90 80 70 60 50 40 30 20 10 K 1 2/3 Figure 1. The Plural Suffix, Spontaneous and Imitative Success a per child basis, it changes little between Grades Kand 2/3: .6 to .7 imitations. Furthermore, while there is a marked increase in the 80 children's ability to imitate between Grades Kand 1, this ability declines between Grades 1 and 2/3 although, in Grade 2/3, it still exceeds their Grade K ability. There is no obvious explanation for this decline, particularly in view of the fact that their ability to produce the

PAGE 95

81 inflection spontaneously in Grade 2/3 exceeds their imi tative ability. We must therefore conclude that the two abilities are not consistently related in the acquisition of the plural inflection. 3.2.1.1.1. Discussion of errors With the few exceptions noted below, the type of error found for this inflection is simple omission of the suffix in obligatory contexts. For example, in response to Question 56, "What do you see in the box?" (three cats are in view), the most frequent response is "cat" and even, occasionally, "a cat". The other errors are discussed by grade level. In Grade K, there are only two, non-omission errors. One is an extension of the plural to a mass noun: "a soup~" (3K70138). The other is the application of the plural to a singular noun: "books" (5K62035; only one book is depicted). In Grade K, there are no cases of spontaneous error other than omission; however, the classification of two responses to Question 40 is unclear. The children are guessing what is inside a box and it is possible they are changing their minds about the number of items contained: "The marbles. I say the marble." (1K400641) "Toys. Toy." (5140126)

PAGE 96

82 These responses were not included in the numerical summary but are noted here because they illustrate a general char acteristic of all the data, and that is, its variability. In Grade 2, there are no errors other than omission. In Grade 3, there are only three non-omission errors. In one case, the plural is suffixed to a mass noun: "That is a bowl of soup__" (4369170) In another case, a child extends the plural to a sin gular noun. In this test item, the child holds two marbles, one in each hand. In response to the stimulus question: "Ask me which one I want", the child replies: "Which one__, which marbles do you want?" (3345183) It seems likely that the problem here is the partitive "which" with its requirement for a singular, rather than a plural, referent. A third response in Grade 3 defies classification: "We're clapping~" (4373171) Although speculation is inconclusive, it may be that the child is attempting to capture the repetitive nature of the verb by pluralizing it; and in this vein, it is in teresting to note that Choctaw has a verbal infix to indi cate repetitive action. On the other hand, this may simply be a stray, overgeneralized third person singular, present tense verb suffix, or the categorization of a verb as a gerundive noun.

PAGE 97

In conclusion, the principal error at every grade level is the omission of plural suffixes. 3.2.1.1.2. Learning strategies The following two sections consider, first, the role of the Ll in the acquisition process, and secondly, the role of cognitive analysis. Choctaw orthography is ex plained in Appendix IV. 3.2.1.1.2.1. The role of the Ll In order to determine if there is interference or transfer from the children's Ll, we must first understand how Choctaw treats plurality. Consider the following sen4 tences. 1. Hattakat itim't pisatok. 'A man saw that tree. I Hattak + at iti + . m~ pl.Sa + tok Man SCM 5 tree that see RPT 2. Hattakat itim~ pisatoklotok. 'Two men saw that tree. I Hattak + at iti + m~ pisa + toklo + tok Man SCM tree that see 3 p.d. RPT 3. Hattak lawat itim~ oklapisatok. 'Many men saw that tree.' Hattak lawa + (a)t iti + m~ okla + pisa + tok Man many SCM tree that 3 p.pl. see RPT 83

PAGE 98

These examples demonstrate that verbs, rather than nouns, are inflected for number in Choctaw. Nouns may be modified for specificity by the use of adjectives 6 or quantifiers but are not marked for number. Therefore, there is no precedent in the children's Ll for plural in flection on nouns; other mechanisms may optionally com municate this information. 84 Before concluding that the lack of precedent in the children's Ll accounts for the fact that they supply plural inflections only 55% of the time after four years of Eng lish instruction, we must first compare their performance with that of ESL learners from other language backgrounds. If the children of this study are similar to others we would reject assigning overwhelming influence to the Ll. Such a comparison will be made in the final section of this chapter. 3.2.1.1.2.2. The role of cognitive analysis The specific learning strategy implied by the term "cognitive analysis" is the active organization, by the learners, of the language data they are exposed to; the formulation of hypotheses which regularize the data; and the realization of these hypotheses in the children's L2 speech. It is frequently the case that the learners' initial hypotheses are too general, either because of incomplete data, or because of rule exceptions in the L2. Thus, one form of evidence that this learning strategy is

PAGE 99

85 being utilized is the occurrence of overgeneralizations in the children's L2, e.g., "eated" for "ate". It should be noted that the classification of ~rrors as overgeneralizations is not always so straightforward as in the case of "eated". The "clapping~" error in the previous section illustrates one of the difficulties. It is clearly possible that this is a transfer error. Thus dis cussion will be limited to those errors which are undeniably overgeneralizations. The second form of evidence that cognitive analysis is actively employed by L2 learners is the occurrence of similar acquisition sequences among English L2 learners from different Ll backgrounds. Overgeneralization errors will be discussed in the cognitive analysis sections of this chapter; ESL acquisition sequences will be compared in the chapter's conclusion. With regard to the error data on the plural suffix, there are only two cases of overgeneralization, occurring in Grades Kand 3, the pluralization of mass nouns. We therefore conclude that these children, at this point in the acquisition process, to not employ cognitive analysis as a general learning strategy. 7 3.2.1.1.3. The plural suffix: summary and conclusions The plural suffix is used successfully 38% of the time by children in Grade K, and 55% of the time by children in Grades 2/3. The suffix, therefore, is in the process of

PAGE 100

being acquired but not to the point that it is firmly a part of the children's grammatical competence. 86 Imitatively, the children's productions are more suc cessful than their spontaneous productions in Grade 2, but in Grades Kand 2/3 the reverse is the case. The relation ship between the two modes of production of this inflection is therefore not consistent. The major error, at every grade level, is omission of the inflection. Cognitive analysis does not appear to play an active role in the learning process during these first four years of acquisition. In the absence of more specific data, at this point, we can tentatively conclude that the prevalence of omission errors is influenced by the nonex istence of noun plural inflections in the children's Ll. This conclusion will be re-examined in the final section of this chapter. 3.2.1.2. The possessive suffix The SWCEL test includes only one item requiring the use of the possessive suffix (Question 66) 8 As a result, there are fewer productions of this morpheme than there are of the plural. A summary of the distribution of these occurrences appears below in Table 12. As illustrated by this table, there is a drastic drop in the successful production of the possessive suffix between Grades Kand 1. A twofold increase in opportunities (i.e., obligatory contexts) is accompanied by a 50%

PAGE 101

87 decrease in successful production. There is no immediately obvious explanation for this phenomenon. Perhaps it is simply a more vivid than usual illustration of the fact that language learning has not only quiescent but retro gressive periods. Table 12. The Possessive Suffix Grade K 1 2/3 # Opportunities 32 62 46 # Successful 7 7 14 % Successful 22% 11% 30% In any event, although the increase in success between Grades 1 and 2/3 is large, the possessive inflection is only marginally within the children's competence at the end of the period for which we have data. In fact, when we compare the plural and possessive inflections, we find that the latter is much less firmly controlled than the former by Grade 2/3. The success of the children's imitative responses is reported below in Table 13 and compared graphically with the success of their spontaneous responses in Figure 2. According to Table 13, the children's ability to successfully imitate the possessive inflection does not increase with age. In fact, of all the grades, Grade K

PAGE 102

100% 90 80 70 60 50 40 30 20 10 I s K \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ 1 2/3 Figure 2. The Possessive Suffix, Spontaneous and Imitative Success Table 13. The Possessive Suffix, Imitative Success Grade K 1 2/3 # Opportunities 26 28 28 # Successful 13 2 11 % Successful 50% 7% 39% 88

PAGE 103

89 demonstrates the most successful imitative ability. More over, the decrease in spontaneous success experienced by Grade 1 is replicated to an even greater degree in imita tive success. With the exception of Grade 1, imitative success ex ceeds spontaneous success, although the difference is much greater in Grade K than in Grades 2/3. Thus it appears that successful spontaneous production proceeds independently of successful imitative production. Without data from later years it is impossible to determine if the greater imita tive success indicates passive competence presaging later spontaneous success. 3.2.1.2.1. Discussion of errors The only error found for the possessive suffix is simple omission of the inflection, as illustrated below. Grade K: "A ball Tony hands, a marbles on his hands" (3K521372) Grade 1: Grade 2: Grade 3: "Tony daddy" (2166099) "Her friend book" (6261157) "His girl, her sister" understood) (4361170; "book" The children indicate possession by producing the passessing noun with no inflection, and by producing the correct word order, Possessor+ Possessed.

PAGE 104

90 3.2.1.2.2. Learning strategies The following two sections will present data concerning the learning of the possessive inflection. It is an inter esting question since, as two of the above examples point out, possessive pronoun morphology is already present in Grades 2/3 (although gender is not yet sorted out), but the seemingly less complex noun possessive inflection still has not been acquired. 3.2.1.2.2.1. The role of the Ll Choctaw expresses possession by the use of personal possessive prefixes which attach to the item possessed and agree in person and number with the possessor. For example: 1. ~cokka iyatok. 'S/he went to my house.' + cokka iya + tok My house go RPT 2. Ohoyohat Rose icokka iyatok. 'A woman went to Rose's house. I Ohoyoh + at Rose i + cokka + tok 1ya Woman SCM Rose her house go RPT 3. Ohoyohat oklicokka iyatok. 'A woman went to their house. I Ohoyoh + at okli + cokka iya + tok Woman SCM their house go RPT

PAGE 105

91 These examples demonstrate that the possessive pattern in Choctaw is somewhat different from the English pattern. In Choctaw, the possessing affix is attached to the item possessed and not, as it is in English, to the possessor. However, the languages are similar to the extent that, if the possessor is not mentioned, the affix precedes the pos sessed item, with the difference that the personal affixes are bound in Choctaw whereas, in the case of the masculine third person singular English pronoun "his" they may be independent, e.g., "That is his". Choctaw has a separate, independent pronoun paradigm which only partially resembles, morphologically, the dependent personal affixes. A final difference between the two languages, not illustrated by the examples, is the presence in Choctaw of noun classes which are inalienably possessed (body parts and some kinship terms). These affixes are, for the most part, non-nasal counterparts of the possessive inflections; and, in transitive sentences, these prefix to the verb to indicate patient relationships. In sum, although there are many resemblances between personal possessives in Choctaw and English, there is no counterpart in Choctaw to the English inflection which suffixes to the possessor. on the possessed item. Choctaw employs personal prefixes It seems very likely that the slow acquisition rate is influenced by the lack of precedent in the children's Ll. There is, however, no evidence of transfer of the entire Choctaw pattern, e.g. "Mary her house". No such phrases occur.

PAGE 106

92 3.2.1.2.2.2. The role of cognitive analysis Because the only error type is omission of the pos sessive suffix, there is no evidence, at this point, that the children are analyzing the L2 data. Cazden (1972:44) notes that children acquiring English as an Ll generalize the possessive inflection to independent pronouns ("mines") but this did not occur among these children, although one of the questions (79) provided the opportunity. 3.2.1.2.3. The possessive suffix: summary and conclusion The possessive suffix is produced spontaneously and correctly in 22% of the required contexts by children in Grade Kand in 30% of the required contexts by children in Grades 2/3. This increase in success is evidence that the inflection is being acquired, but the process is neither rapid nor regularly incremental, as demonstrated by the drop in Grade l's success rate. The reduced success in Grade 1 is reflected in the children's imitative produc tions as well. With the exception of this grade, they imitate possessive suffixes more successfully than they spontaneously produce them. Because the only error is that of omission it is dif ficult to determine what learning strategies the children employ. While transfer from the Choctaw pattern may in fluence omission of the suffix, this mechanism, if it is operating, does not do so consistently in view of the fact that the children do not transfer the English equivalent

PAGE 107

(which by Grades 2/3 they have already acquired) of the Choctaw inflection onto the noun that would be inflected 93 in that language. The absence of overgeneralization errors suggests that cognitive anal y sis is not, at this point, a relevant learning strategy. Having dismissed both learning strategies, we are somewhat at a loss to e x plain the acquisition of the possessive suffix. In terms of real gain over the four-year period, there is not a great deal to explain since the suffi x is only produced at a 30 % success rate by the oldest children in the sample. In comparison with the phonologically homophonous plural suffi x we find that the possessive suffix enters the children's grammar at a much slower rate. Moreover, the two suffi x es are very dissimilar in their patterns of spontaneous and imitative success so that we can conclude that other than phonological factors affect the acquisition of the two suffixes. Influence from the surrounding dialect of English ma y probably be dismissed as a factor. Wolfram (1971:146) found the possessive morpheme present in more than 89 % of obligatory contexts in the speech of white, Mississippi delta children. E x actly what these considerations are remains to be determined, but necessity and redundance should not be dismissed. English is adamant that number be specified on nouns and, with the exception of the third person singular, present tense verb suffix, au x iliaries and the copula, the language does not redundantly supply this information in

PAGE 108

simple, declarative sentences. Possession, on the other hand, is redundantly available through apposition. The relationship of these observations to the differential acquisition rates of the plural and possessive suffixes is certainly speculative, but it is not inconceivable 94 that semantic priorities and syntactic provisions for re dundance in the L2 are reflected in the learner's acquisi tion sequence. 3.2.1.3. The article prefixes In terms of frequency, articles are one of the most ubiquitous types of morphological affixes in the English language. While they are not so essential to communicative success as, for example, subject-verb word order, their absence is very noticeable and frequently distinguishes the non-native from the native speaker. The following dis cussion presents the children's learning of these inflec tions. Of the 57 questions in the SWCEL test, 29 require the presence of an article. Ten questions require the indefinite article, 16 require the definite, and 3 require either one or the other. This large number of opportunities results in a proportionately large sample for this item. The children's success for definite and indefinite articles combined is reported in Table 14, below.

PAGE 109

95 Table 14. The Article Prefixes Grade # Opportunities # Successful % Successful K 1 2/3 573 1301 954 231 544 449 40% 42% 47% On an individual basis (Table 15, below), we find that the number of opportunities in each child's speech increases with age, so that the success rate presented in Table 14 is based on an increasingly larger figure from grade level to grade level. Table 15. Articles, per Child Production Grade K 1 2/3 Total# Opportunities 16.9 24.1 33.4 Total# Successful 6.8 10.l 15.7 Tables 14 and 15 demonstrate an increase in article usage and success as the children increase in age. This increase is not large, however, and we cannot characterize articles as established in the children's grammars by Grades 2/3 inasmuch as they are correctly produced in

PAGE 110

slightly less than half the contexts in which they are required. 96 The children's imitative success is presented in Table 16 which includes both definite and indefinite articles in the numerical summary. Imitative success is compared with spontaneous success in Figure 3. Table 16. The Article Prefixes, Imitative Success Grade K 1 2/3 # Opportunities 528 896 715 # Successful 199 44i 417 % Successful 38% 49% 58% When we compare imitative success with spontaneous success (Figure 3), we find that, with the exception of Grade Kin which spontaneous responses are two percentage points more successful than imitative responses, imitative success exceeds spontaneous success and that the degree of excess increases with age; that is, in Grade 1, imitative responses are 7 percentage points more successful than spontaneous ones, and in Grades 2/3 imitative responses are 11 percentage points more successful. Moreover, success in imitation appears to result, at least in part, from the supplying of articles omitted in the children's spontaneous responses. Table 17, which

PAGE 111

97 100 % 90 80 70 60 50 40 s I 30 20 10 K 1 2/3 Figure 3. The Article Prefixes, Imitative and Spontaneous Success tabulates imitative errors resulting from omission, indi cates that this error is much less frequent in imitative responses than it is in spontaneous productions. Other types of imitative errors, though more frequent, are much the same as other types of spontaneous article errors, to be discussed below. In other words, the children's imitative productions do not radicall y depart from their spon taneous ones as far as articles are concerned.

PAGE 112

98 Table 17. Article Omission Errors, Imitative Responses Grade Total # Errors Total # Omission % Omission I* s K 329 109 33% 78% 1 455 169 37% 89% 2/3 298 118 49% 88% *I = Imitative; s = Spontaneous Turning to the question of the relationship between imitative and spontaneous production, we see that articles display the most consistent relationship of any structure thus far considered. Yet, the relationship is not entirely transparent. On the assumption that the ability to imitate a structure is indicative of this structure's development within the speaker's grammar and, although nascent, it will soon emerge productively, we would expect an imitative ex cess at one grade level to forecast, at the next, an in crease in spontaneous ability. To a certain extent that expectation is met, but the rate of increase in spontaneous success is much slower than the rate of increase in imita tive success, so that we cannot conclude with any degree of certainty that the children's increasing ability to imitate articles correctly is a reflection of an increasing incor poration of the structures into their passive competences.

PAGE 113

99 There is, in the analysis of errors, an indication that imitation is being utilized as a learning strategy, albeit not entirely successfully. For example, with errors involving prepositional structures, it would appear that some of the children imitate what they have heard, [a], but have, as a consequence of the reduction in casual speech of (av) ("of") to [a] and resulting homophony with "a", the two syntactically different morphemes are coalesced in the children's grammars, e.g., "soup a bowl" (1K69008) for "a bowl of soup", and "front a cow" (3333203) for "in front of the cow". However, there are too few examples of this type to describe the process as a general stage in the developmental sequence or, for that matter, as a general learning strategy. 9 3.2.1.3.1. Discussion of errors At every grade level, omission of the article is the 10 most frequent error encountered. The number of omission errors is shown as a percentage of the total number of errors for each grade level in Table 18 below. The Table 18. Article Omission Errors Grade K 1 2/3 Total# Errors 342 757 505 Total# Omission 267 674 443 % Omission 78% 89% 88%

PAGE 114

proportion of article errors due to omission, rather than decreasing with age, actually increases. 100 The frequency of other article errors is not numeri cally significant, but their forms illuminate other aspects of the children's developing grammars, providing us with clues about their contemporary syntactic systems. Table 19 enumerates the major types of article errors other than omission. Table 19. Types of Article Errors, excluding Omission Grade K (N=65*) 1 (N=9 l) 2/3 (N=50) Subcategorization Errors With proper noun 3 4 0 With mass noun 11 13 8 5 With plural noun 13 12 8 With body part noun 10 11 7 With kin term noun 1 0 1 With adjective or predicate adjective 5 9 4 With numerical quantifier 3 0 1 Definiteness "a" for "the" 5 10 7 "the" for II a II 1 7 5 Substitution of nasal syllabic 5 3 3 *Refers to number of children in the grade level In Table 20, below, the errors in Table 19 are illus trated. The examples are arranged in grade order, beginning with Grade K, for each error category.

PAGE 115

101 Table 20. Examples of Article Errors, excluding Omission Category Proper noun Mass noun Plural noun 12 Body part noun Kin term noun d . 13 A Jective Numerical quanti fier Definiteness "a" for "the" "the" for "a" Syllabic nasal Grade K 1 2/3 K 1 2/3 K 1 2/3 K 1 2/3 K 1 2/3 K 1 2/3 K 1 2/3 K 1 2/3 K 1 2/3 K 1 2/3 E x ample "A Tony play a fireman" (4K59025) "A Tony" (2157094) None "A ice cream" ( 6K70 039) "Tony got a water" (3160083) "She have a soup" (4369166) "a marbles in his hand" (3K521372) "Ton y have a books in that box ( 115 5 0 4 7 ) "I see a kittens" (225192) "Under the hand" (5K36036) "You have it in the hand" (1142055) "What Tony has in the hand" (6251158) "This a sister's" ( 2K6114 9) None "A father" (3257193) "Take the blue" (1K46011) "I give you the blue" (3146082) "It' s a b 1 ue" ( 2 2 4 4 2 0 0) "A one" (3K62132) None "A one marble" (4341177) "A ball" (2K29149) "Side of a cow" (2134094) "I would like a cow" (4338173) "The fireman" (2K59148) "The big one red" (1164064) "The big red dog" (3264197) "ij big dog" (1K64001) "ij bowl" ( 1160969) "That's ij big box" (3253191)

PAGE 116

102 According to Table 19, the most common error is the extension of article prefixes to inappropriate noun sub classes, and to adjectives and quantifiers. The syntactic implication of these article extensions is that the chil dren have not yet subcategorized nouns, and have provi sionally classed some adjectivals as nouns. The definiteness errors require no comment other than that the contexts clearly required the opposite of the form supplied by the children. Here, as elsewhere, all ambi guous cases were discarded. The syllabic nasal errors are interesting for reasons to be discussed in Section 3.2.1.3.2.1. below. There are other variations than those listed in Table 20, chiefly nasalized mid central and low back vowels, e.g. "[~] bowl" (1169069), and"[~] kitten, has a kitten" (3255197); and nasal consonants plus one of these vowels, e.g., "[na] dog" (3163084). Many other probable examples were discarded due to the impossibility of distinguishing them from re duced forms of the -ing verbal suffix and the preposition II in II Finally, a variety of other errors, too numerous and too original to be easily classified, were produced. Among these, errors with prepositions predominate. They are of two kinds: the substitution of an article for a preposi tion ("front a cow", 3333203); and the insertion of an article before a locative preposition ("pig the behind", 1K33019). Both of these types will be discussed in Section 3.2.1.3.2.1.

PAGE 117

103 There were also occasional instances of articles prefixed to verbs: "A stand (2173097); "A clap" (2183091); "The write-a" (1K50019). Several examples of what appear superficially to be articles suffixed to verbs ("He write-a the paper", 4150112; "She have-a the ball", 2128097) will be discussed in Section 3.2.1.3.2.1. Finally, at every grade level, variability of inser tion is evident, an indication that the use of articles is developing in the children's language despite the high rate of omission. The following examples illustrate this varia bility at every grade level, beginning with Grade K. "He got a book. Book. 11 (4K64019) "He's got big cat and a little kitten." (21550931) "And a boots and firehat and a firecoat. (52591531) 3.2.1.3.2. Learning strategies Although the children are not acquiring English arti cles rapidly, the variety of errors produced makes it pos sible to extrapolate information about their approaches to learning these prefixes. We have already noted that imita tion appears to be one strategy employed by the children. The following discussion suggests that both Ll transfer and cognitive analysis operate as well. 3.2.1.3.2.1. The role of the Ll Choctaw has a complex system of inflections which corresponds not only to English articles but also to the

PAGE 118

104 deictic functions realized, in English, by demonstrative pronouns. The Choctaw system, moreover, includes discourse functions, emphasis and focus, which in English are rea lized through paralinguistic and syntactic mechanisms, respectively. The Choctaw inflections are suffixed to nouns (or to adjectives, locatives, and quantifiers if the noun is modified), and are marked for subject and oblique case. The morphemes are bound and may be combined with each other to communicate increasingly detailed specification. The following sentences illustrate some properties of the Choctaw system. It should be noted that the Subject Case Marker is obligatory; the other inflections are dependent, in these examples, upon the speaker's intention. Unmarked: Faniyat ossak apatok. 'Squirrel(s) ate nut(s) .' Fani + (y)at ossak nut apa + tok Squirrel SCM eat Definite Article, Oblique Case only: Faniyat ossak~ apatok. 'Squirrel(s) ate the nut(s).' Indefinite Focus, Subject: Faniyos ossak apatok. RPT 'It was squirrel(s) that ate nut(s) .'

PAGE 119

Indefinite Focus, Oblique: Faniyat ossak~ apatok. 'It was nut(s) squirrels ate.' Emphasis, Subject: Faniyato ossak apatok. 'Squirrel(s) ate nut(s).' Emphasis, Oblique: Faniyat ossakano apatok. 'Squirrel(s) ate nut(s).' Deixis, Subject and Oblique: Fani~at ossak~ apatok. 'That squirrel(s) ate this nut(s).' Faniat ossak~ apatok. 'This squirrel(s) ate that nut(s).' Emphasis and Deixis, Subject: Fanimato ossak apatok. 'That squirrel(s) ate nut(s).' Emphasis and Deixis, Oblique: Faniyat ossakpano apatok. 'Squirrel(s) ate this nut(s) .' 105 As mentioned previously, if a noun is modified, the modifier follows the noun and the inflections are suffixed to the final modifier of the NP. An example with an adjective

PAGE 120

106 modifier is the following: Fani losamat ossak apatok. 'That black squirrel(s) ate nut(s)." Even a superficial examination of this system (which is more complex than here indicated in that other inflec tions, such as partitives, may combine with these suffixes) indicates there are many points of contrast between the two languages. The most obvious difference is the fact that the inflections are suffixed in Choctaw and prefixed in English. A second point of difference is that the inflec tions are bound and suffixed to the final modifier in the NP; whereas in English the modifiers, with the exception of articles, are free, and always precede the head noun in the NP. Finally, of the Choctaw inflectional system, only the definite article{-~} and the deictic inflections {-pa~-p~} and {-ma~-m~} have lexical or inflectional coun terparts in English. The Choctaw definite article, further more, is only partially similar to English "the" since its use is restricted to the oblique case. Thus, in Choctaw, nouns are inherently indefinite (i.e., there is no indefi nite morpheme), and only oblique nominals can be inflected for definiteness. Turning now to the question of whether or not the points of contrast and similarity are reflected in the children's acquisition of articles, we must first consider the fact that the major error is that of omission. From

PAGE 121

107 this, we might conclude that the children are, in the case of indefinite nominals, transferring the Choctaw pattern, i.e., no inflection, to English. However, if the Choctaw pattern were being transferred, we would also expect the definite article to be produced regularly in appropriate contexts, which it is not. In fact, articles, both definite and indefinite, are produced sporadically. The attractive hypothesis that first definite articles emerge (having a precedent in the Ll), then indefinite articles, is not supported by the data. We might also expect the place of affixation to be transferred from the Ll, but this, apparently, is not a problem for the learners; that is, there are no cases in the corpus of an article being suffixed rather than pre fixed.14 Even in Grade K when Choctaw lexical items appear more frequently than in other grades, the English article, if produced at all, appears in the proper position: A katos A sokka (2K6144; 'a cat') (4K30026; 'a pig') From the above discussion, we conclude that broad and obvious patterns are not transferred from the Ll. However, three types of errors suggest that some Ll patterns which are more specific do indeed appear in the L2. These are: 1) errors involving nasal and nasalized syllabics; 2) the extension of articles to adjeGtives; and 3) the extension of articles to prepositions.

PAGE 122

108 In the case of the nasal syllabics produced in English for articles (see Section 3.2.1.3.1. for examples), the similarity of these to the Choctaw definite article is sufficient to be convincing as a case of rather direct transfer and substitution of semantically equivalent and phonetically close morphemes: Choctaw{-~}; English {a-~an-}. In the case of adjectives and prepositions, the chil dren's prefixation of articles may not be a result of categorization errors, as suggested previously, but of a transfer of the Choctaw pattern of inflecting adjectives and locatives when they are the final constituent of an NP. The fact that articles are prefixed to locatives is, pos sibly, further reinforced by the peculiar attributes of English prepositional locatives, to be discussed in the following section. In conclusion, patterns of the Ll, Choctaw, do appear to be realized in the emerging L2 in several minor ways, but the major L2 difficulty experienced by the children, the obligatory inclusion of articles, cannot be related to Ll patterns. 3.2.1.3.2.2. The role of cognitive analysis The discussion of errors indicates that the major error, after simple omission of articles, is the extension of articles to inappropriate categories of nouns and to inappropriate lexical classes. It was argued in the

PAGE 123

109 preceding section that the extension of articles to prepo sitions and to adjectives is a result of the transfer of the Choctaw classificatory system and the inflectional privileges of modifiers in that language rather than the overgeneralization of articles. However, this conclusion is suspect if it is the case that other English L2 learners also attach articles to adjectives and prepositions. An examination of the literature reveals little de tailed treatment of article errors. Brown and Bellugi (1970:75-99), in a general discussion of developing re strictions of privileges of occurrence within NP's among English Ll learners, note that overgeneralization of articles to mass and plural nouns is a common error. There are, however, no examples of the extension of articles to adjectives and prepositions in their sample. The children in that study did not place these in the class of words that allowed modification. Tentatively, then, we ma y conclude that overgeneral ization accounts for article errors involving the nondif ferentiation of noun subclasses, i.e., mass and proper nouns; and that the conclusion in the preceding section with regard to article errors and adjectives and preposi tions is valid. Certainly more extensive data, especially from English L2 acquisition studies, is necessary to evaluate these conclusions. It should be noted that, although the Ll transfer analysis adequately accounts for the locative errors, there

PAGE 124

110 is precedent in the L2, as well as in the Ll for the pre positional confusion. The following examples illustrate the problem. 1. In back of X. In the back. 2. In front of x. In the front. 3. On top of X. On the top. The logical extension of this pattern, of course, is: "The beside the sheep" (22341981) "A back of me" (321193) So, while arguing that the classification of locatives as modifiers which are inflectionally privileged is an Ll transfer, we might also argue that the pattern is reinforced by the eccentricities of the L2 locative system. 3.2.1.3.3. The article prefixes: summary and conclusion The children in this study have achieved a proficiency level of 47% correct usage of English articles by Grades 2/3. Imitatively, at that level, they are somewhat more successful, achieving a success rate of 58%. Their chief error is that of omission, accounting for 88% of their errors in Grades 2/3. Other errors, although much less frequent, indicate that the children employ, to limited extents, imitation as well as transfer and cognitive

PAGE 125

111 analysis as learning strategies. The question of whether or not the children reported here acquire articles in the same sequential context, and as successfully, as other English Ll and L2 learners will be discussed in the con cluding section of this chapter. 3.2.2. Verb inflections The three verbal inflections discussed in this section are. the third person singular, present tense {-s}, the progressive tense marker {-ing}, and the simple past tense {-D}. All of these inflections are bound and suffix to verb roots. 3.2.2.1. The present tense {-s}. The English present tense includes not only ''present" time events (for which, in fact, it is rarely used), but also all affirmations that are true at the time of the speaker's statement (Jespersen 1964:228). The tense is marked by the absence of inflection with the exception of the third person singular form which is marked by the suffix { -s } The children's acquisition of this inflection on "main" verbs, i.e., noncopular, nonauxiliary, is examined below. Onl y four opportunities arise during the SWCEL test for the children to respond with a third person singular verb. Each of these is designed to elicit the verb "has" which, unfortunatel y is also an auxiliar y verb in other

PAGE 126

112 contexts. However, other third person singular verbs which occurred in casual conversation are included in Table 21 below. Nonetheless, these data may not provide an accurate reflection of the children's acquisition of the inflection, since not only may ''has" acquisition intersect with the acquisition of auxiliaries, but also "has" does not conform to the regular rule for the formation of the third person in that it undergoes a root-final change rather than simple suffixation of {-s}. Therefore the figures in Table 21 should be regarded with caution. Table 21. Third Person Singular Present Tense {-s} Grade K 1 2/3 # Opportunities 14 50 66 # Successful 7 22 24 % Successful 50% 44% 36% Although the sample size for Grade K is very small, it nonetheless does not contradict a trend of decreasing success with third person singular verbs. Rather than being acquired, the inflection is losing ground in the children's inventory of English morphemes. Imitatively, the situation is not a great deal different, despite the fact that, in contrast to the limited number of spontaneous contexts for the third person

PAGE 127

singular inflection, the number of imitative contexts is quite large. Table 22 presents the children's imitative success and Figure 4 compares it with their spontaneous success. Table 22. Third Person Singular, Imitative Success 113 Grade # Opportunities # Successful % Successful K 1 2/3 72 115 146 39 71 84 54% 62% 58% Table 22 indicates that the children experience a relatively high level of initial success which increases in the second year, then declines slightly in the two fol lowing years. Despite this decline, there is an overall increase, though small, in success as the children grow older. In comparing imitative and spontaneous ability (Figure 4) we find that, at every level, imitative ability exceeds spontaneous ability. Moreover, the children's imitative ability between Grades Kand 1 increases, whereas their spontaneous productive ability decreases. Between Grades 1 and 2/3, imitative as well as spontaneous abilities decrease, although the spontaneous decrease is more marked than the imitative decline. There is no evidence for a

PAGE 128

100% 90 80 70 60 ---------------I----50 s 40 30 20 10 I K 1 2/3 Figure 4. Third Person Singular, Spontaneous and Imitative Success consistent relationship between the two abilities with regard to this inflection. 114

PAGE 129

115 3.2.2.1.1. Discussion of errors On verbs which form the third person in the regular fashion, the most common error, found at all grade levels, is simple omission of the inflection. Examples, not ex haustive but in proportion to their grade level distribu tion, are: Grade K: "He like it coke" (2K71149) "He like ice cream and eat it" (5K70035) "He want cat" (5K56035) Grade 1: "He eat ice cream" (4168118) Grade 2/3: "She go school" (3262186) As the examples indicate, the proportion of errors on regular third person verbs decreases as the children grow older. A related error is the use of uninflected "have" for "has". This error is distributed evenly among the grade levels, an indication that the root-changing form creates learning difficulties at all ages. Examples are: Grade K: "He have books" (1K51001) Grade 1: "He have a book" (6151108) Grade 2/3: "Kitten, he have kitten in the box" (6255161) Another source of error is the substitution of other tenses for the present, for example:

PAGE 130

116 Grade K : "He having ice cream" (3K8132) Grade 1: "A ball she's having" (2128093) Grade 2/3: "He had a kitten in the box" (5255156) As the above examples suggest, the most frequent substi tution is a form of the progressive for the present tense. In addition, there are errors of the type "he's have" which appear to belong to this category. However, they are not included since it is impossible to determine whether these should be analyzed as intended (if inappropriate) progres sives or as unanalyzed routines, i.e., "he's" as a single, fossilized structure followed by the verb. A third type of error is the apparent contraction of a nonauxiliary verb, in this case, "has". There are no examples from Grade K. The examples from Grades 1 and 2/3 below, may be either "his" or "he has", contracted. The latter response is that which the test item attempts to elicit. G rade 1: "[hiz] kittens in that box" (1155045) "[hiz] kittens" (1155064) Grade 2/3: "soup, [hiz] hot soup" (3269186) A fourth type of error is the inappropriate e x tension of the third person singular inflection or the third person singular verb form. E x amples are:

PAGE 131

Grade K: Grade 1: Grade 2/3: None. "He haves a book in his hand" (3151079) "She haves a ball" ( 512812 6) "He haves kits in the box" (4255165) "I has fingers on my hand" (5274156) 117 Within a single utterance, there may be variation between overgeneralization and grammatical usage: Grade 1: "Tony have~, he has kit in the box" (5155126) Finally, the use of "got" must be mentioned. The non-traditional use of this verb is becoming very wide spread and, while its syntactic status and semantic scope are debatable (and probably dialectally variable, to acer tain extent), its appearance among these English learners, whose exposure to the language is quite circumscribed, demonstrates its ubiquitousness and its absorptive power. 15 Examples are: Grade K: Grade 1: Grade 2/3: "He got a book" (4K51028) "He gots kitten" (4K550210) "Tony got a water" (3160083) "He gots ice cream" (5168126) "One, he got one book" (3362196)

PAGE 132

118 In brief summary, the major errors found in the use of the third person singular verb inflection are: the omission of the inflection on regular verbs; the use of uninflected "have"; the substitution of another tense for the simple present; the apparent contraction of "he has"; and overgeneralization of the inflection. 3.2.2.1.2. Learning strategies Although the sample size for this inflection is rela tively small, the variety of errors allows some insight into the children's approach to learning the inflection. 3.2.2.1.2.1. The role of the Ll The general form of the Choctaw verb is discussed in more detail in later sections. Here it should be noted that, in terms of person, Choctaw verbs are unmarked in the third person, and that all persons are similarly in flected for the present tense. The present tense in Choctaw is formed by suffixing {-h } to the verb root: 16 Tamaha pit balilih. 'S/he is running to town.' ~amaha pit balili + h town toward run PT With regard to the present tense, Choctaw and English differ in a very important respect: the Choctaw present

PAGE 133

119 tense includes both action that exists in present time and action that is in the immediate past, i.e., action that has just been completed, so that the example above could equally well be translated 'S/he has just run to town'. A final difference between the two languages is lexical, and specific to the subset of the third person verbs ap pearing in the data, the various forms of "have". In Choctaw there is no single verb which corresponds to this English predicate. Depending on the situation, different lexemes are available and obligatory to describe the rela tionships which are subsumed under the rubric ''have" in English. For example, the contexts which occur in the SWCEL test could be expressed in a variety of ways according to the speaker's desire to indicate possession; the position of the object referred to in relation to the speaker; or the manner in which the object is being held by an individual. There is, however, no indication that these lexical contrasts affect the children's acquisition. The verbs "have" and "got" are produced in appropriate contexts and there is no tendency to substitute positional or circum locutory expressions for these verbs. On the other hand, the fact that the third person singular present tense is unmarked in Choctaw may account for the high frequency of uninflected third person forms in the corpus. However, because other English Ll and L2 learners also omit this inflection, we must compare these

PAGE 134

120 learners with those in order to determine if these speakers learn the inflection later or omit it with a higher fre quency than others. Either of these patterns would support an Ll transfer and interference analysis. This comparison will be made in the conclusion of this chapter. 3.2.2.1.2.2. The role of cognitive analysis If the children are to any extent employing induction in the acquisition of this inflection, we should find in the corpus instances of their ''trying out" this inflection in order to determine its boundaries of application. Be sides such inappropriate usages of the suffix, other evi dence for the acquisition of the inflection by analysis is the appearance of the suffix on third person forms which do not follow the general rule, e.g., "has" appearing as "haves". Both forms of overgeneralization are found, parti cularly in the last two levels tested. Examples of each, in the order discussed above, are: Grade K: Grade 1: Grade 2/3: None None "He haves a book in his hand" (3151079) "I has fingers on my hand" (5274156) "He haves kits in the box" (4255165)

PAGE 135

121 The frequent appearance of "haves" in Grades 1-3 suggests that the children are in the process of acquiring the third person inflection, despite their success rates, and have yet to subcategorize verbs which are irregular in this person. It should be mentioned that the omission of the in flection on third person singular verbs in the present tense may be reinforced by a dialect pattern prevalent in that geographical area. Speakers of Black English, and some Southern Whites as well, do not ordinarily distinguish the present tense form by inflecting it (Fasold 1972; Wolfram and Clarke 1971). In other words, some portion of the L2 input to which the children are exposed provides an uninflected model. 3.2.2.1.3. The present tense {-s}: summary and conclusion The total number of obligatory contexts in the SWCEL for the use of the third person singular present tense inflection is relatively small and represents, predomi nantly, the irregular form ''has". The children's success with this form and occasional other, regular third person verb forms is at the 36% level by Grades 2/3. Rather than an increase in success with age, the learners experience a decrease, although the Grade K sample is so small that no significance can be attached to that age's performance. The decline iD spontaneous abilit y between Grades 1 and 2/3 occurs in imitative ability as well, although the

PAGE 136

latter's decline is not so extreme as the farmer's. At all grade levels, imitative ability exceeds spontaneous ability. 122 The analysis of errors suggests, tentatively, that interference from the children's Ll may account to some extent for acquisition problems; and that, to the extent the children are acquiring the inflection, they have hypo thesized a rule whose limitations they have only partially discovered. Their acquisition problems may be compounded by the fact that some Southern dialect patterns to which they are exposed do not consistently inflect third person singular present tense verbs. 3.2.2.2. The progressive inflection { -ing} In English, the progressive tenses are used to indi cate ongoing action in present, past, and future time. These tenses are formed by inserting the auxiliary "be" (in the appropriate tense, person, and number) before the verb whose root the progressive inflection {-ing } is suf fixed to. The present progressive is, in fact, the true present tense of the language, the present tense itself being much more frequently used to indicate conditions of fact rather than contemporaneity of time and speaker's statement. The SWCEL test includes six items (Questions 29, 54, 73, and 81-83) designed to elicit the present progressive. The figures in Table 23, below, indicate the children's

PAGE 137

spontaneous success with the inflection {-ing} only; errors pertaining to the auxiliary will be discussed in a later section. Table 23. The Progressive Inflection {-ing} 123 Grade # Opportunities # Successful % Successful K 1 2/3 118 351 241 73 219 163 63% 62% 68% Table 23 indicates that, although this inflection is fairly successfully produced by the children, even in Grade K, there is relatively little increase in successful usage as the children age; that is, their acquisition of the inflection is slow despite their initial, successful production. Comparing this performance with their imitative pro ductions (Table 24 and Figure 5, below) we : find that imi tatively they are both more and less successful. Initially, the children are less successful than th~y are in the spon taneous use of this inflection, but their imitative success experiences a steady increase until it exceeds spontaneous success. We therefore cannot conclude that there is a consistent relationship between the two abilities.

PAGE 138

100% 90 80 70 s ---60 I --50 40 30 20 10 K 1 2/3 Figure 5. The Progressive Inflection, Spontaneous and Imitative Success Table 24. The Progressive Inflection, Imitative Success 124 Grade # Opportunities # Successful % Successful K 1 2/3 70 137 108 42 88 79 60 % 64% 73%

PAGE 139

~ ---------------------------~ 125 3.2.2.2.1. Discussion of errors 17 The most frequent error is the omission of the inflectional suffix, as shown in the examples below: Grade K: Grade 1: Grade 2/3: "He draw" (1K50008) "They~ that a school" (3K65132) "He's~ there in my hand" (2136087) "You stand my back" (1181075) "I am stand" (2283157) "You are sit" (6283157) At every grade level, there is variability of inclusion: Grade K: Grade 1: Grade 2/3: II [wast]' [was] ("watch") the house burnt, watch'Ing house and burning" (5K9035) "Clapping hand, I clap my hand" (1173046) "She is bounce his, is playing with his ball" (6229162) The above examples and discussion exhaust the pro gressive suffix errors found in the corpus. 3.2.2.2.2. Learning strategies Because of the paucity of errors of any other type than omission, it is difficult to extrapolate the children's approach to learning the progressive inflectional suffix. It may well be the case, given the data presented in Table 23, that there is very little active acquisition

PAGE 140

126 occurring for this feature over the age levels tested. The following two sections will briefly discuss the evidence available. 3.2.2.2.2.1. The role of the Ll The progressive tense in Choctaw is formed by the inclusion of one of a set of positional verbs (Jacob et al. 1977), marked for the appropriate person and tense, with a verbal denoting the action itself. Compare the following examples, the first in the simple present tense, and the second in the present progressive: Kobaffilih. 'I break (it) up.' Kobaffi + li + h Break I PT Kobaffihos ~ttalih. 'I am breaking (it) up.' Kobaffi + h + os ~tta + li h Break ? focus exist I In the second example, the positional verb ~tta, in combination with the preceding derived verbal, indicates progressive aspect. The example demonstrates that there is a very major difference in syntax between Choctaw and English progressive constructions. Assuming that the positional verb functions similarly to an auxiliary in the

PAGE 141

Choctaw progressive, then the order of auxiliary and main verb is exactly opposite in the two languages. Secondly, a fact not illustrated by the example is that in Choctaw the main verbal may be optionally and redundantly marked for person: kobaffilihos (kobaffi + li + h + os) (break I ? focus) 127 Despite these differences between the languages, there is no specific evidence of interference, unless the chil dren's slow rate of acquisition is partially accounted for by the contrasts. There is no evidence of errors that can be labeled as transfers from the children's Ll. 3.2.2.2.2.2. The role of cognitive analysis There is, furthermore, no direct evidence that cog nitive analysis is utilized by the children in learning the progressive inflection. There are no occurrences in the corpus of overgeneralization; the inflection does not occur in inappropriate contexts. 3.2.2.3. The progressive inflection: summary and conclusion Considering only the children's use of the progressive suffix and ignoring, at this point, the auxiliary consti tuent of the construction, we find that the children use the inflection in Grade K with a high rate of success,

PAGE 142

but that their success increases very little between Grades Kand 2/3. 128 The only significant error is omission of the inflec tion. With regard to imitation, the children's success rate is only slightly better in Grades 1 and 2/3 than their spontaneous success; and it is lower than their spontaneous success in Grade K. Thus there is no consistent relationship between the two abilities. Furthermore, there is no specific evidence that the Ll affects acquisition, nor do the children overgeneralize the inflection. These findings are best explained as the result of a presumably temporary stasis in the children's acquisition of the progressive suffix. If the inflection were being more actively acquired, we would expect a greater growth curve and a variety in error types encoun tered. 3.2.2.3. The past tense inflection {-D} The English past tense, formed by suffixing the mor pheme (usually symbolized as {-D}) to the verb root, is used to express time previous to the present and occurrences which are completed. The inflection has three allomorphs whose shapes are determined by the voice and point of arti culation of the verb's root-final phone However, there is an extensive class of verbs that forms the past tense irre g ularly through alterations of the root vowel nucleus, e.g., "give/gave", through suppletion, e.g. "go/went",

PAGE 143

and other devices. Since these are not overtly rulegoverned there is no question but that the learner must master them by some mechanism other than analogy. 129 Three questions on the SWCEL test elicit past tense verbs (Questions 39, 42, and 46); in every case, the verb forms its past tense irregularly: "gave", "were", and "took"; moreover, "were" is marked for plural number. Although the children occasionally used regular past tense forms during the test (and these are included in the sum mary), Table 25, below, reflects primarily their success with irregular past tense formations. Table 25. The Past Tense Grade K 1 2/3 # Occurrences 26 73 72 # Successful 10 17 20 % Successful 39% 23% 28% The number of occurrences for Grade K is too small to regard the percentage figure for that grade with any degree of confidence. The figures for Grades 1 and 2/3 indicate that the irregular past tense is used at a relatively low level of success which increases only slightly with age. Comparing spontaneous and imitative success (Table 26, below), we find that the children are more successful

PAGE 144

130 100% 90 80 70 60 so 40 s 30 20 10 K 1 2/3 Figure 6. The Past Tense, Spontaneous and Imitative Success Table 26. The Past Tense, Imitative Success Grade K 1 2/3 # Occurrences 25 44 52 # Successful 18 22 29 % Successful 72% 50% 56%

PAGE 145

131 in the latter mode and that, moreover (Figure 6), the two success curves are quite similar, even with regard to the Grade 1 decline. We can conclude, then, that in the case of the irregu lar past tense, imitative and spontaneous production undergo parallel development. The obvious explanation for this fact is that these forms, which cannot be pre dicted from hypothesis, must be learned by memorization and imitation. 3.2.2.3.1. Discussion of errors The following discussion is divided into two sections: the first is the presentation of irregular past tense error data; and the second, of errors involving the regular past tense, although these data are extremely limited. The most common error found among irregular past tense verbs is the use of the simple present tense in place of the past. This error occurs at all grade levels: Grade K: Grade 1: Grade 2/3: "I give it to you up you hand" (4K39027) "I give you a marble" (5K39036) "You take marbles" (1146044) "I give you the blue" (3146082) "I give away the pig" (6239162) "I give it to you the pig" (6339160)

PAGE 146

Other errors involve the use of the past tense in place of the appropriate tense: Grade K: Grade 1: "He made his draw" ("drawing"; 1K50001; past for future tense) "He had ice cream" (1168062; past for present tense) 132 Grade 2/3: "He had a ball" (5228152; past for present tenser-Finally, some copular verbs, correctly marked for past tense, are incorrectly marked for number: Grade K: "They was in here" (4K42028) Grade 1: No examples Grade 2/3: "The marbles was in the box" (4342171) The status of these items, however, is not unequivocable since these forms are perfectly acceptable in the dominant dialect of the region. For that reason, they will be ex cluded from further discussion. There are very few regular past tense constructions in the corpus. In Grade K, there are none at all. In Grade 1, there are six occurrences, two of which are well formed. Two of those remaining are clearly overgeneral izations: "Something deed it" ("did"; 5150129) "Putted it in your hand" ("put"; 1146054)

PAGE 147

and two are inappropriate for lexical-semantic reasons: "He sharped it" ("sharpened"; 1149056) "Tony sported water" ("squirted"?; 1159048) In Grades 2/3, there are only three occurrences, all produced by the same speaker and all well-formed. The above examples exhaust the types of irregular and regular past tense error types. 3.2.2.3.2. Learning strategies 133 It has already been suggested that imitation and memorization are probably active strategies in learning the irregular past tense, and that, due to the nature of the verb formation, cognitive analysis would be an ineffi cient strategy. The following discussion, therefore, is primarily concerned with the assistance or interference the children may receive from their Ll. 3.2.2.3.2.1. The role of the Ll In casual conversation, two past tenses are regularly d h 19 employe in C octaw. The Recent Past Tense is formed by suffixing the morpheme {-tok} to the verb stem: Ohoyohm~ pisalitok. 'I saw that woman/those women.' Ohoyoh woman/women pisa + li + tok that see I RPT

PAGE 148

134 The Remote Past Tense is formed by suffixing first the morpheme {-t-} to the verb stem, then the Recent Past Tense morpheme: Ohoyohm~ pisali!tok. 'I saw that woman (some time ago).' The time depth of the two tenses varies depending upon a variety of circumstances, including whether or not the speaker wishes to emphasize or minimize the length of time since the event reported occurred. Generally, the Recent Past Tense reports events between Immediate Past time and three months previous to the statement. The Remote Past Tense may overlap with this period, but is usually reserved for events occurring more than a year previously. The parameters of neither tense are specific and are dependent upon discourse as well as more strictly linguistic vari ables. It should be remembered that immediate past events employ the same morpheme as present tense events. Despite the obvious differences between Choctaw and English with regard to past tense expression, there are definite similarities. Both languages express past events by the use of verb suffixes which, with the exception of the first person singular and third person dual in Choctaw, suffix directly to the verb root. The semantic notion of past completed action is consistent in the two languages, although in Choctaw the time elapse may be more finely delineated. This latter fact, and the fact that Choctaw

PAGE 149

has no class of verbs which forms the past tense irregu larly, are two major contrasts with English. 135 Finally, there is the fact that the Choctaw "present tense" includes events in the recent past, and this may explain the frequent use of the present tense for the past tense in English, especially in view of the fact that the events referred to have just occurred in the interview con text. Furthermore, the slow rate of past tense acquisition may be exacerbated by the large number of high frequency verbs which are irregular in English in this tense; so that the L2 itself interferes with the learning process. 3.2 2.3.2.2. The role of cognitive analysis Because the SWCEL test supplies little information on regular verbs in the past tense, it is difficult to deter mine to what extent, if any, the children actively engage in hypothesis formation and testing. The few occurrences of overgeneralization in the corpus suggest that some children employ this strategy, but that it certainly is not widespread 3.2.2.2.3. The past tense: summary and conclusion The past tense of irregular verbs is not produced by the children with a great deal of success. By Grades 2/3, the children produce these forms spontaneously and correctly at the 28% level of success. They imitate these forms much more successfully--the 56% level in the same

PAGE 150

136 grade--and their imitative ability parallels, at a higher success rate, the trend of their spontaneous ability. Three related factors in the Ll and the L2 appear to play a role in the learning of the irregular past tense forms. First, in the Ll, the past tense is never irregular, so there is no precedent in the children's first language for these forms. Secondly, in the L2, a great many high frequency verbs have irregular past tense forms, so that the children are not exposed to regularized input data. Thirdly, the Choctaw "present tense'' includes recent past events and the children may be transferring this relative time frame to English, which is much more rigid in its treatment of time. If these factors do indeed affect acquisition, we would expect the regular past tense inflec tion to appear slowly in the children's speech; unfor tunately, the corpus provides insufficient contexts in which to examine this hypothesis. Finally, in view of the correlation between imitative and spontaneous success rates, it would appear that imita tion plays an active role in learning the irregular past tense verbs. 3. 3. Pronoun Morphology The English pronominal system systematically recognizes person, number, and to a limited extent, gender and case. There are three groups of pronouns: pre-verb, or subject pronouns; post-verb, or object pronouns; and possessives,

PAGE 151

137 which have both dependent and independent paradigms. Each group has three persons; the first and third persons have singular and plural number, but the second does not (ex cept in some dialect patterns). Only the third person singular pronouns are marked for gender: feminine, masculine, and neutral. The following discussion will include data for all three groups. However, due to inherent limitations in the test itself, data on the post-verb pronouns are not extensive and discussion is consequently brief. Because the test does provide a great deal of information on gender distinction, much of the information below is con cerned with this aspect of pronoun morphology. 3.3.1. Pre-verb pronouns This section presents data on the third person singular pronouns and on "other" pre-verb pronouns. Information on the former is extensive, so the "others" will be dis cussed initially and briefly. The remainder of the dis cussion, including the analyses of errors and of learning strategies, will be devoted to the gendered pronouns. Although the SWCEL offers ample opportunity for the use of pre-verb pronouns, the tendency of the children (and probably of most beginning second language learners) is to answer with the shortest possible phrase which usually, in English, is an object complement. While this is certainly acceptable and, in fact, typical of casual

PAGE 152

138 conversation, it diminishes the amount of information available on subject and verb structures in the children's spontaneous speech. The test offers 28 opportunities for the production of non-third person singular subject pronouns. Of these, only two are plural forms; the rest are either first or second person singular pronouns. The various forms are not distinguished in Table 27, below, which presents the children's performance as a whole. Table 27. Pronoun Subjects, Non-third Person Singular Grade K 1 2/3 # Opportunities 120 324 497 # Successful 73 231 332 % Successful 61% 71% 67% Obviously the road to acquisition of subject pronouns is not a smooth one. The children experience, initially, a high level of success which increases rapidly in Grade 1, then decreases in Grades 2/3 so that they are, after two and three years, only slightly ahead of their Grade K success. There is no satisfactory explanation for this "backsliding", although examination of the data on gendered pronouns suggests that the children's attention may be focused on those pronouns rather than these.

PAGE 153

139 There are 17 opportunities in the SWCEL test for producing the third person singular subject pronouns, 14 of which require the masculine form and three of which require the feminine. The children's success is reported in Table 28. Table 28. Third Person Singular Subject Pronouns Grade K 1 2/3 # Opportunities 106 207 270 # Successful 54 121 184 % Successful 51% 58% 68% These figures, again, demonstrate a relatively high rate of initial success, but unlike the non-gendered sub ject pronouns, the children's success increases consis tently with their age. Moreover, when we look at the figures on a per child basis, we find that the use of these pronouns also increases with age: children in Grade Kuse one of these pronouns only 3.4 times during the test, compared with Grade 1 children who use one 5.3 times and Grades 2/3 children who use one almost 11 times each during the test. The children's ability to imitate subject pronouns not marked for gender is summarized in Table 29, below, and graphically compared with their spontaneous ability in Figure 7.

PAGE 154

100% 90 80 70 60 50 40 30 20 10 140 1-----s K. 1 2/3 Figure 7. Subject Pronouns, Spontaneous and Imitative Success Table 29. Pronoun Subjects, Non-Third Person, Imitative Success Grade K 1 2/3 # Opportunities 138 300 315 # Successful 130 290 305 % Successful 94% 97% 97%

PAGE 155

141 If successful acquisition were to be judged on imi tative ability alone, the children's grammars could cer tainly be characterized as having a fully developed, non third person singular, subject pronominal system. From Grade Kon, they imitate well above the 90% criterion level for acquisition. However, their spontaneous ability (Table 27 and Figure 7) is much less proficient. Until Grades 2/3, the two abilities have a relatively consistent relationship. At that grade level, imitative ability remains constant with Grade 1 ability while spon taneous success slightly declines. Again, the only explan ation for this is that the children have reached a plateau and their attentions are focused on other second language structures. Finally, whether or not imitative ability affects or reflects the acquisition of these pronouns is, in this case, difficult to determine. These pronouns are, both spontaneously and imitatively, the most successfully pro duced structures to be examined thus far~ Is the fact that the children are capable of imitating these structures well assisting their spontaneous production? Or, is the fact that they imitate well a simple consequence of the fact that their spontaneous production is relatively fluent? At this point, there is no basis for selecting one or the other of the two possible interpretations; despite their divergent implications for language learning, they are both reasonable conclusions.

PAGE 156

100% 90 80 70 60 50 40 30 20 10 I--..-...... s K 142 -......~ ------------........ 1 2/3 Figure 8. Third Person Subject Pronouns, Spontaneous and Imitative Success Table 30. Third Person Singular Su~ject Pronouns, Imitative Success Grade # Opportunities # Successful % Successful K 1 2/3 117 276 275 114 237 223 97% 86% 81%

PAGE 157

143 Turning now to the third person singular subject pro nouns, we find a slightly different situation, reported in Table 30 and Figure 8, above. Here we find that the children's ability to imitate pronoun gender is extremely high at the youngest ages tested, but this ability rapidly declines as the children's age and spontaneous ability increase. Nonetheless, by Grades 2/3, imitative ability still exceeds spontaneous success. Imitative ability, in this case, is not only independent of but non-reflective of spontaneous ability. 3.3.1.1. Discussion of errors Errors in non-gendered subject pronouns consist almost entirely of omissions and will be discussed in the fol lowing chapter. Errors in the third person singular pro nouns consist mainly of simple substitution of the opposite gender. Grade K. "She's pulling his wagon" (5K271292) Grade 1. "She gonna get the marble" (2154103) Grade 2/3. "He is pushing the ball" (1129052) There is no indication that the children tend to use one gender exclusively or predominantly; both feminine and masculine pronouns are freely supplied for the appro priate opposite. Nor do any other factors appear to cor relate with the children's selection of gender. We conclude

PAGE 158

144 that the children are learning both genders simultaneously. Further evidence for this is the presence of variability in the upper grade levels. The following examples catch the children in the process of sorting out gender (and in one instance, case) distinctions. The appropriate gender is underlined. Grade K. Grade 1. Grade 2/3. No examples "She didn't, she, he don't have that ball" (3127082) "She, he is a yellow dog" (3163082, gender optional) "Her, he is, [hIYr] is going schoo1 1120 (33601%) "She have a, he have the book" (6251159) There are a few occurrences of gender confusion with the neutral pronoun "it", which is an optional selection for some responses, i.e., the animals may be referred to neutrally. Most of these errors, however, do not occur with those questions but replace "he" in the expected re sponse to "Who is he?" (Question 48). There are only two errors involving the substitution of a gender pronoun for the neutral pronoun. Besides gender errors, there are a few errors in case, one of which has already been illustrated above. Since the majority of these errors involve possessive case substitu tions, they will be discussed in Section 3.3.3.

PAGE 159

145 3.3.1.2. Learning strategies The following two sections deal exclusively with third person singular pronoun subjects. There is unfortunately little data, as the error discussion indicates, for deducing strategies. 3.3.1.2.1. The role of the Ll The Choctaw personal affix system, in general, recog nizes case, number, and person. The subject, or agent case has three persons and singular, dual, and plural number. However, the paradigms are not always realized in full, for both linguistic and sociolinguistic reasons. The plural affixes, for example, are optional in some contexts, and the dual is either not used or used interchangeably with the plural by some speakers in some communities. The complete paradigm is illustrated below, using the verb root pisa, "to see", in the Recent Past tense, {-tok}. Person Singular 1 . p1sa + li + tok 'I saw' see I RPT 2 is + pisa + tok 'you saw' you see RPT 3 pisa + tok 's/he saw' see RPT

PAGE 160

Person 1 2 3 1 2 3 Dual + pisa + tok 1 we 2 see RPT has + pisa + tok you 2 see RPT pisa + tokla + tok see two RPT Plural okli + pisa + tok we see RPT oklahas + pisa + tok you pl. okla they + see pisa see + tok RPT RPT 'we 2 saw' 'you 2 saw' 'they 2 saw' 'we saw' 'you pl. saw' 'they saw' We find both points of similarity and of difference between the Choctaw and English subject systems. The basic similarities are the recognition of three persons 146 in both languages, and the general, but not absolute, rule of word and morpheme order; that is, subject followed by verb root. The most obvious difference is that the Choctaw number system discriminates more specifically than does the English system in that a dual number exists for all persons in Choctaw. Secondly, and very importantly in terms of

PAGE 161

this discussion, the third person singular verb is not marked with a personal affix. Thirdly, the Choctaw lan guage makes no gender distinctions in this system. 147 Finally, and this is not observable from the illus trations, the Choctaw affixes are absolutely bound and cannot be used independently as can the English pronouns in elliptical expressions, e.g., "Who saw that? You?" A separate, independent series of pronouns exists for this purpose in Choctaw. Despite the differences between the languages, in view of the children's success rates with gendered and non gendered pronouns (68% and 67%, respectively, in Grades 2/3), we cannot conclude that the lack of a precedent for gender in the Ll makes an impact on the children's learning this L2 principle. It is obvious from the errors cited that the children are working on gender distinctions, but they are nonetheless making them, and with greater suc cess than they are, for example, making distinctions between singular and plural nouns, and with slightly more success, by Grades 2/3, than they are producing other pre-verb pronouns. We therefore conclude that, in the case of gender and the third person singular pronouns in English, there is no evidence of Ll influence. With regard to omission errors, on the other hand, we will find in the next chapter's discussion that their source may well be founded in the children's Ll.

PAGE 162

148 3.3.1.2.2. The role of cognitive analysis The children's increasing success with third person subject pronouns suggests, in the absence of evidence for other types of learning strategies, that the children de velop increasingly accurate hypotheses with regard to the effect of the sex of the third person referent on the selection of the appropriate pronouns. Their success can not be attributed to the transfer of an Ll pattern, since none exists, nor can it be ascribed to successful imitation, since their ability in that mode unaccountably declines with age. This conclusion is tentative and will be reviewed and compared with other L2 learners' experiences with gen dered pronouns in the chapter's conclusion. 3.3.1.3. Pre-verb pronouns: summary and conclusion Non-third person pre-verb pronouns are acquired unevenly by the children; that is, after the first two years, the children's spontaneous production suffers a decline in success. However, the children's ability to correctly imitate these pronouns is extremely proficient and does not decrease. This fact suggests that the chil dren are not actually losing ground with respect to the acquisition of these pronouns, but are temporarily dis tracted by other aspects of the L2 learning task. Specifically, and quite likely in view of their acqui sition history, it is the third person singular, gendered pronouns that absorb the children's attention since they

PAGE 163

149 develop rapidly and consistently. By the middle of Grade K, children produce gender correctly in subject pronouns slightly more than 50% of the time, and by the middle of Grades 2/3, almost 70% of the time. The children seem to learn these pronouns through observation and analysis since neither imitation nor Ll patterns appear to have any demonstrable effect on their acquisition. The children's overall performance on subject pro nouns, regardless of gender, is summarized in Table 31 and Figure 9, below. Table 31. Subject Pronouns, Summary Grade K 1 2/3 # Opportunities s 226 531 767 I 255 576 590 # Successful s 127 352 516 I 244 527 528 % Successful s 56% 66% 67% I 96% 92% 90% In sum, the children's success with subject pronouns increases with age, dramatically in the first two years of L2 learning, and slightly in the last two years tested. Their ability to imitate the pronouns, on the other hand, declines from an initial level of 96% success to a final level of 89%. This rate also changes only slightly

PAGE 164

100% 90 80 70 60 50 40 30 20 10 Figure 9. s K 1 2/3 Subject Pronouns, Spontaneous and Imitative Success 150 between the second and the last two levels. While the leveling off of spontaneous success may be accounted for by the fact that the learners' attentions are preoccupied with acquiring gender pronouns, as suggested above, the decline in imitative ability has no satisfactory explana tion and, in fact, violates both theory and common sense expectations concerning the relationship between these

PAGE 165

two modes of production. Clearly, more data are needed in this area of research. 3.3.2. Post-verb pronouns Post-verb pronouns in English may present special problems for the learner in that "you" and "it" are also pre-verb pronouns, and "her" is also a possessive form. Alternatively, of course, the lexical redundancy may, to 151 some extent, simplfy the learning task. The data below suggest that this is, in fact, the case for these children. Because Chapter Four treats object pronouns in greater depth, the following discussion simply summarizes the quantitative information. Although the SWCEL test requires only one object pronoun response, the children frequently substitute pro nouns for noun objects so that a relatively large number occur in the corpus. The most frequently used objects are "you" and "it". No distinction between direct and indirect objects is made in the tables below, the first of which reflects the children's spontaneous success and the second their imitative success. These are graphically compared in Figure 10. As Table 32 demonstrates, object pronouns are produced increasingl y successfully by the children and are, in fact, ap p roaching a le v el of 80 % success b y Grades 2 / 3, that is only 10 points below the acquisition criterion. Irnitatively (Table 33), the children are even more successful,

PAGE 166

although again we observe a repetition of the declining success rate found for non-third person subject pronouns. Table 32. Object Pronouns, Spontaneous Success 152 Grade # Opportunities # Successful % Successful K 1 2/3 151 318 317 91 211 244 Table 33. Object Pronouns, Imitative Success Grade K 1 2/3 # Opportunities 30 45 60 # Successful 22 39 49 60% 66% 77% % Successful 73% 87% 82% Errors consist primarily of omission and redundance ("He carrying it a box", 5254205) and therefore are within the domain of syntax, to be discussed in the following chapter.

PAGE 167

100% 90 80 70 60 so 30 20 10 K 1 2/3 Figure 10. Object Pronouns, Spontaneous and Imitative Success 3.3.3. Possessive case pronouns 153 English has both dependent and independent possessive pronouns. The two series recognize gender in the third person singular forms. The independent pronouns, with the exception of the first person and the masculine and neutral third person singular forms, are formed by suffixing (-s)

PAGE 168

154 to the dependent pronouns. The first person singular suffixes (-n) and the third person independent forms are, with the exception of the feminine, the same as the depen dent. In the following discussion, both independent and dependent forms are reported together, although the majo rity of occurrences represents dependent pronouns. There are 11 opportunities for the use of possessive pronouns throughout the SWCEL test. Six responses require "my" or "mine"; two require "your" or "yours"; one requires "his"; one requires "her" or "hers"; and one requires "our". Data for the pronouns having gender are presented first, below, followed by data for the non gendered forms. Table 34. Grade K 1 2/3 Third Person Singular Possessive Pronouns # Opportunities 42 112 79 # Successful 22 43 30 % Successful 52% 38% 38% This table suggests that the gender of possessive pronouns may cause difficulties for the learners. While they produce gender correctly approximately half the time in Grade K, their ability decreases to 38% in Grade 1 and remains at that level in Grades 2/3. The only mitigating

PAGE 169

155 statistic is that the children in the latter grades produce approximately three times the number of pronouns as the children in Grade K: .6 and 1.6 pronouns per child per test in Grades Kand 2/3, respectively; so we can conclude that the children, as they increase in age, attempt to use the pronouns more frequently and that their success, while not increasing, is at least remaining stable. The data for the other possessive pronouns are not differentiated in terms of person, so that Table 35, below, summarizes the children's success with all non-third person singular possessive forms. Table 35. Non-Third Person Singular Possessive Pronouns Grade K 1 2/3 # Opportunities 130 278 244 # Successful 78 167 157 % Successful 60% 60% 64% This table indicates that the learners experience a high level of initial ability and very little increase in spontaneous correct productions after that. Nonetheless, in contrast with the gendered pronouns, the children's use of these possessives does increase in succe~s, and they are a great deal more successful than they are in using the gendered possessives.

PAGE 170

156 Unfortunately, data on the children's ability to imi tate possessive pronoun gender are very limited. The learners rarely attempted to imitate these forms when requested to do so. Because of the small numbers, the percentages reported in Table 36, below, cannot be regarded with a great deal of confidence. The data for other pos sessive pronouns, Table 37, are more extensive and there fore more reliable. Table 36. Third Person Singular Possessive Pronouns, Imitative Success Grade K 1 2/3 # Opportunities 14 32 35 # Successful 12 17 23 % Successful 86% 53% 66% To the extent that these figures reflect the chil dren's ability to imitate these pronouns, we find that, despite a decline in Grade 1, their imitative ability exceeds their spontaneous ability at every grade level. Table 37 demonstrates that the learners experience a continuing and rapid increase in their ability to imi tate the non-gendered possessive pronouns, and imitate these more successfully than they do the gendered forms.

PAGE 171

157 Table 37. Non-Third Person Singular Possessive Pronouns, Imitative Success Grade K 1 2/3 # Opportunities 81 86 101 # Successful 40 57 79 % Successful 49% 66% 78% In order to obtain a composite picture of possessive pronoun acquisition, Table 38 and Figure 11 combine and summarize the data for all possessive pronouns, gendered and non-gendered. This summary indicates that the children achieve no gain in their overall use of possessive pronouns sponta neously, although their imitative abilities steadily increase. Previous discussion points out that it is the gendered pronouns which skew the summary figures, since non-gendered forms are spontaneously produced with some increase in success. 3.3.3.1. Discussion of errors Errors involving possessive pronouns are much more various that those associated with subject pronouns. One frequent error is the substitution of articles for pos sessors with body parts, or the conjunction of both article and possessive pronoun with body parts. For example:

PAGE 172

158 100% 90 80 70 -----60 s 50 I 40 30 20 10 K 1 2/3 Figure 11, Possessive Pronouns, Summary of Spontaneous and Imitative Success Table 38. Possessive Pronouns, Summary Grade K 1 2/3 # Opportunities s 172 390 32 3 I 95 118 136 # Successful s 100 210 187 I 52 74 102 % Successful s 58% 54% 58% I 55% 63% 75%

PAGE 173

Grade K: "Go up the top of the~ hand" (3K36132) Grade 1: "You have it in the hand" (1142055) Error in gender is another source of difficulty. With these errors, as was the case with the subject pro nouns, there is no tendency to select one or the other gender. Masculine and feminine forms are equally, and wrongly, chosen. 159 Problems with case also account for a number of errors. Some of these are clearly extensions of pre-verb pronouns, e.g. : Grade K: "You nose" (1K80020) Grade 1: "She book" (1161066) Grade 2/3: "You have in you hand" (6257158) Other extensions also involve the subject case but are not so clearly simple overgeneralizations. Examples are: Grade K: "[iiz] book" (4K61028) Grade 1: "[hiz] kitten" (1155064) Grade 2/3: "[iiz] mother" (6257158) These will be carefully examined in the following section.

PAGE 174

160 Finally, a few errors indicate that independent and dependent pronouns are not completely distinguished from each other. Examples of dependent pronouns used indepen dently are: Grade K: "your" II [m~y] 11 (2K74194; for "your nose") ( 5K9 0 33; for "my hand") Examples of independent pronouns used dependently are : Grade K: "Mine hand" (4K97021) Grade 3: "This hand is mine hand" (4379171) In sum, the majority of possessive case errors involve the use of inappropriate gender. After deducting these from the total, there are some 122 "other" errors which fall into the various types illustrated above. 3.3.3.2. Learning strategies The variety in error types allows ample opportunity to speculate on the children's approaches to learning the possessive pronouns. In the following sections, we will point out that factors in both the Ll and the L2 appear to affect the children's acquisition of the pronouns. 3.3.3.2.1. The role of the Ll The brief discussion of possession in Choctaw (Sec tion 3.2.1.2.2.1.) pointed out that the major features of

PAGE 175

161 the system are bound morphemes prefixed to possessed items. These prefixes are marked for person (first, second, and third) and number (singular, dual, and plural) as illus trated below. The possessed item is cokka 'house'. Person Singular 1 q. + cokka 'my house' my house 2 ci + cokka 'your house' your house 3 -i+ cokka 'her/his house' 3p.s. house Dual 1 Pi + cokka 'our 2 house' our 2 house 2 haci+ cokka 'your 2 house' your 2 house 3 Plural 1 hapi + cokka 'our house' our house 2 oklahaci + cokka 'your pl. house' your pl. house

PAGE 176

162 3 okli + cokka 'their house' their house If the possessor is mentioned in the sentence, it precedes the possessed item which is inflected for the appropriate person and number: Mary icokka 'Mary's house' Besides the possessive affixes above, Choctaw has a series of independent pronouns which may be used posses sively and are, functionally, very similar to English independent possessive pronouns. Human body parts are inalienably possessed by another series of affixes which are, for the most part, similar to those above except for the fact that the vowels are non-nasal. Reference terms for relatives are usually inalienably possessed, the pre fixes dependent upon the relationship of ego to the rela tive referred to. In comparing the two systems, we find that the major difference between Choctaw and English possessives, other than number, is, again, in the third person singular which is not marked for gender in Choctaw, but is in English. This contrast would appear to explain the children's re duced success with the third person possessive pronouns in English as compared with the other possessive pronouns. There are precedents for the other possessive pronouns in the children's own language and, in fact, the languages'

PAGE 177

systems are structurally so similar that the children have only to identify and "plug in" the corresponding English 163 morphemes. In the third person, however, the children must learn not only the semantic scope of gender, but three separate morphemes as well. Unfortunately, as straightforward as this analysis appears, it absolutely founders on the fact that it does not hold for the subject pronoun data. The Ll and L2 subject pronoun systems are, again, structurally sufficient ly equivalent, with the exception of number and the third person singular, that we would expect gendered subject pronouns, if our analysis in the case of the possessive forms is correct, to lag behind in the acquisition process; but, in fact, they do not and are achieved at the same level as non-gendered subject pronouns by Grades 2/3. Therefore, we cannot accept an interference analysis as explanatory of the possessive case data. In fact, the only certain conclusion from the two sets of data is that contrastive analysis cannot be conducted with isolated linguistic facts; such analysis must systematically con sider the entire interface of two languages in order to arrive at valid, or predictive conclusions concerning language learning. In this case, we cannot conclude that lack of a gender precedent in the Ll interferes with L2 learning unless it does so in all structurally similar L2 situations; and it does not.

PAGE 178

164 Because phonological systems are more easily deli neable than syntactic systems, we are on somewhat more secure ground in this area of contrastive analysis, and we can conclude with reasonable certainty that one of the error types discussed in Section 3.3.3.1. is explicable from the point of view of contrasting phonologies. The use of [ moy] for "mine" has its probable origin in the chil dren's Ll which has a phonemic series of nasal vowels and no word-final nasal consonants. Thus, these productions may result from the learners' hearing the final nasal consonant in English as a nasal vowel. If this is the case, then the children are actually producing the appro priate possessive pronoun but pronouncing it in accordance with Choctaw, rather than English, phonological patterns. There are other errors that exhibit Ll influence but are only partially explicable from that perspective. These will be discussed in the next section. To conclude this discussion, we have found little evidence that Ll transfer either facilitates or hinders the acquisition of English possessive pronouns. A preliminary analysis of interfer ence in the learning of the gendered possessive pronouns was rejected since it failed to account for the acquisition facts of similar subject pronouns. On the other hand, contrasts in phonological systems were shown to play a part in the non-differentiation of the first person singu lar independent and dependent possessive pronouns "my" and "mine".

PAGE 179

------------------165 3.3.3.2.2. The role of cognitive analysis Three types of errors may be partially accounted for by the cognitive approach to acquisition. All three also require reference to the children's Ll for satisfactory explication. The errors are: 1) the use of subject for possessive pronouns; 2) the pattern of Article+ Possessive + Body Part; and 3) the use of subject pronouns suffixed by (-s) as possessives. Errors involving the use of subject pronouns as pos sessives ("you nose"; 1K80020) are undoubtedly overgeneralizations of subject pronouns. Support for this analysis is found in the fact that subject pronouns are acquired more successfully than possessive pronouns (compare Tables 31 and 38). A factor which may reinforce the overgeneral ization of subject pronouns is the nonexistence of pho nemic or allophonic post-vocalic [r] in Choctaw. Because this sound is the sole distinction between the subject and possessive cases of the second person English pronouns, "you" and "your", the children's use of "you" may reflect phonemic rather than morphological acquisition factors. However, there is no information available on the acqui sition of English phonemes by the children, so this obser. l 21 vation remains specu ative. The concatenative use of articles and possessives with body parts may also be accounted for by two learning strategies: the transfer from the Ll of inalienable possessors, and the overgeneralization of articles to body parts,

PAGE 180

166 resulting in such productions as: "Go up the top of the hand" (3K36132). Examination of the literature proves this error to be unique, reinforcing our conclusion that the error is a combination of Ll transfer and overgeneral ization. Turning to the use of [siz] and [hiz] for possessive pronouns, we are on less firm ground. There are several interpretations or analyses of these errors and no sure means of selecting the most valid among them. One possible explanation is that these forms are analogic errors; that is, the children have observed the final (-s) pattern of the English independent paradigm ("yours", "hers", etc.) and hypothesized that dependent possessive pronouns consist of subject pronouns plus word-final (-s). There is further reinforcement for this analysis from the children's Ll in which [IJ and [i] are in allophonic distribution so that "he" + (-s) (hiyz) and "his" (hiz) are homophonous to the children's ears and consequently in their speech. Alternatively, there is the very strong possibility that the acquisition of possessive and subject pronouns in the third person intersects with the acquisition of contracted third ~erson auxiliaries, "is", "does", and "has", creating segmentation confusion in both pronoun and auxiliary systems. Two factors further compound the confusion: the phonology of Choctaw, mentioned above, in which [I] and [iJ are in allophonic distribution; and a

PAGE 181

167 feature of southern dialects, "hit" for "it 11 22 The fol lowing examples, with possible glosses, illustrate the problem. [hiz hit] cold? (6K67043) "Is it cold?" [hits] hot. (1K670110) "It s hot" ( ? ) "His hot" (?) [hiz] kittens in that box. (1155045) "His" (?) "He has" improperly contracted? What [diyz] have in his other hand? (2154102) "Does he is" (?) "Does he has" improperly con tracted? [Iz] girlfriend. (4261175) "Is" (?) "his" (?) Soup, [hiz] hot soup. (3269186) "It's" (?) "his" (?) "he's" (?) [Itsa] having box. (2261200) "It's+ a" (?) "he s + a" ( ? ) [hiz] do play with ball. (3229182) "He's"(?) "his"(?) [hiz] not green. (1143066) "His" (?) "it's" (?)

PAGE 182

168 These examples, occurring at all grade levels, empha size the uncertainty of ascribing the errors to any single acquisition factor. Clearly, auxiliaries cause problems, probably southern "hit'' is being learned and slows down the differentiation of "his" and "it's"; 23 and probably, also, the allophonic variation of [I] and [i] in Choctaw contri butes to the problem. More extensive data on the acquisition of independent pronouns in order to determine if the children are making analogic errors, and more information on the acquisition of auxiliaries, would greatly assist the analysis of the errors. The first are not available, but the second will be found in following sections, so we will return to this discussion there. At this point, we can only conclude that factors in both languages appear to affect the learning of the third person singular possessive pronouns. 3.3.3.3. Possessive pronouns: summary and conclusion Possessive pronouns are successfully produced by the children at a level of 58% in Grade Kand at the same level in Grades 2/3. There is no increase in the rate of suc cess. The data suggest that the learning of the gendered pronouns affects the overall rate of success and an analysis of errors demonstrates that these pronouns present special problems for the children. However, no one factor can be singled out as preeminent in the learning process. Factors

PAGE 183

in both languages combine to inhibit the children's acquisition of the possessive pronouns. 3.4. Auxiliary Verbs This section examines the children's success with three major English auxiliaries, "do", "be", and "will". 169 The SWCEL provides three contexts for the production of these auxiliaries: progressive verb tenses, interrogatives, and negative statements. Although these contexts do not fully test the range of the English auxiliary system, they do require the learners to manipulate a great many factors, including person, number, and tense; contraction; and the placement of the auxiliary in negative and inter rogative sentences. 3.4.1. The children's auxiliary system, an overview A summary of the children's overall performance by grade level is presented below. This summary collapses a variety of factors which will, afterwards, be discussed separately. Included in the summary are: copular and non-copular auxiliaries ("be" and "do"); third and non third person present tense forms ("is" and "are", "does" and "do''); contracted and non-contracted forms: the future "will"; and affirmative, interrogative, and nega tive modes. Table 39, below, presents the children's success in the use of all forms of the auxiliaries.

PAGE 184

Table 39. All Auxiliary Verbs Grade K 1 2/3 # Opportunities # Successful 131 67 412 220 422 283 % Successful 51% 53% 67% 170 The most notable feature of auxiliary acquisition, in general, is the increase in use between grade levels. In Grade K, each child employs only two auxiliary verbs during the test; children in Grade 1 use between four and five auxiliaries per test; and children in Grades 2/3 use be tween eight and nine auxiliaries. Not only is their use of these forms increasing, but also their successful pro duction, particularly in the last level tested. We can therefore conclude that the children are actively acquiring auxiliary verbs. Turning to the children's ability to imitate auxil iaries, a slightly different picture emerges, reflected in Table 40, below. The gradual increase in success from grade level to grade level found in the children's spon taneous success is not exhibited in the children's imitative productions. Again, it is Grade 1 in which the lag occurs. Nonetheless, the children's imitative ability exceeds, at every level, their spontaneous ability, as shown in Figure 12, below.

PAGE 185

100% 90 80 70 60 50 40 30 20 10 s K 1 2/3 Figure 12. Auxiliary Verbs, Spontaneous and Imitative Success Table 40. All Auxiliary Verbs, Imitative Success Grade K 1 2/3 # Opportunities 246 596 500 # Successful 170 395 385 % Successful 69% 66% 77% 171

PAGE 186

172 The following sections consider each of the auxiliaries, their various forms, and their development in the children's grammars. 3.4.2. "Do" auxiliaries This auxiliary has several different functions in English. The following are found in these data. First, the auxiliary may emphasize the truth-content of an affirmative statement: "I do want to go". Secondly, it may take the place of an object complement which has been ellipsed: "I do (want to go) ". Thirdly, it is obliga torily inserted to negate a statement: "I do not want to go". Finally, and also obligatorily, it is placed before the subject in interrogatives: "Do you want to go?". In each of these cases, the auxiliary, rather than the main verb, is marked for person, number, and tense inflec tions. This fact might be expected to pose problems for ESL learners. The SWCEL provides 11 opportunities altogether for the production of a form of "do''. Four of these are in statements, both affirmative and negative (Questions #27, 47, 49, and 68), and the rest are in interrogatives (Ques tions #40, 41, 45, 51, 55, 71, and 75). Table 41, below, summarizes the children's overall success in these contexts.

PAGE 187

Table 41. The Auxiliary "Do'', Overall Success Grade K 1 2/3 # Opportunities 79 243 2 30 # Successful 42 123 146 173 % Successful 53% 51% 64% As Table 41 demonstrates, the children's spontaneous success in producing forms of the auxiliary "do" is some what less successful than their production of auxiliaries in general. A slight decline in Grade 1 is followed by a sharp increase in successful production in Grades 2/3. In comparing the children's imitative and spontaneous abilities, we find that their imitative ability exceeds, in every case, their spontaneous ability. Table 42, below, summarizes their imitative success with all forms of the auxiliary. Table 42. The Auxiliary "Do", Overall Imitative Success Grade K 1 2/3 # Opportunities 246 596 500 # Successful 170 39 5 385 % Successful 69% 66% 77%

PAGE 188

174 100% 90 80 70 --I -----.,,,,,,,..,,,,,,,..,,,,,,,. ______ .,,,,,,,..,,,,,,,,. 60 50 s 40 30 20 10 K 1 2/3 Figure 13. Auxiliary "Do", Spontaneous and Imitative Success Figure 13 compares their imitative and spontaneous success and demonstrates that the two modes closely parallel each other in development, suggesting that they are, in this case, related. In order to determine which aspects of the auxiliary are acquired most readily, the following tables separate various forms and functions of "do". 24 Table 43 compares third and non-third person forms; Table 44 presents the

PAGE 189

175 h d f d h h 25 t ir person orm contracte wit t e negative and Table 45 presents "do" contracted in Wh-Interrogatives, i.e., "What do you produced as [hw : \'cuw]. Table 43. "Do" and "Does 1126 Grade K 1 2/3 # Opportunities do 18 73 93 does 3 23 20 Table 44. "Doesn't" Grade K 1 2/3 Grade K 1 2/3 # Opportunities 12 26 35 # Opportunities 46 121 82 # Successful do 16 47 80 does 2 10 16 # Successful i 7 12 # Successful 24 59 38 % Successful do 89% 64% 86% does 13% 43% 80% % Successful 0 27% 34% % Successful 1 52% 49% 46%

PAGE 190

176 Separation of the various forms allows us to observe that by Grades 2/3 the children are most successful in producing "do" and "does" in both statements and Yes/No Interrogatives (Table 43); less successful when contracting in Wh-Interrogatives (Table 45); and least successful in contracting the third person singular with a negative (Table 4 4) Imitative success is presented in tables 46 and 47, below, which separate the auxiliary into three forms of occurrence, "do" and "does", and "doesn't". (There are too few occurrences of [hw A 'cuw] on which to base percen tages.) Table 46. Grade K 1 2/3 "Do" and "Does", Imitative Success 27 # Opportunities do 30 76 87 does 34 96 48 # Successful do 28 63 74 does 21 57 32 % Successful do 9 3% 83% 85% does 62% 59% 67%

PAGE 191

177 Table 47. "Doesn't", Imitative Success Grade # Opportunities # Successful % Successful K 1 2/3 16 44 42 15 38 42 94% 86% 100% Figure 14, below, compares imitative with spontaneous success for the different forms of the auxiliary. Here we see that the summary presented in Figure 13 is deceptive in that it obscures a great deal of variability in the rela tionship between the two modes of production with respect to specific forms of the auxiliary. For example, there appears to be little relationship between the two produc tive modes in the case of "doesn't"; and spontaneous ability actually exceeds imitative ability in the cases of "do" and "does", although in the former there is only a percentage point difference. We must conclude, there fore, that the consistent relationship observed in the summary of spontaneous and imitative productions is misleading and that the relationship between the chil dren's success rates in the two modes is extremely variable when individual forms of the auxiliary are compared.

PAGE 192

100% 90 80 70 60 50 40 30 20 10 I / ,, ,, ,, ,, ,, / --_.:-I / / s / s K / / / ,, ,, / / / / ,, / / "Do" "Does" "Doesn't" / / / ,, 1 ,, ,, / / / / 2/3 Figure 14. Forms of "Do", Spontaneous and Imitative Success 178

PAGE 193

179 3.4.2.1. Discussion of Errors The most common error for every form and function of the auxiliary "do", at all grade levels, is omission. In Grade K, in fact, omission is the only error type that occurs. Examples are: Grade K: "How much finger I got on my hand?" (4K75021) "You have marbles is there?" (3K41132; "is" for "in") Grade 1: Grades 2/3: "How much finger I have?" (6175109) "How many finger you have?" (4175119) "What you want?" (3245191) "What Tony has in the hand?" (6251162) As use of the auxiliary incr~ases in Grades 1-2/3, so does the variety in errors committed. The most common error, following omission, is the use of "do" for the copular auxiliary in intended progressive structures: Grade K: None Grade 1: "What do you do?" (3176079) "What do you doing?" (2176094) Grades 2/3: "What do you do?" (5276205) "What do you doing?" (5276156) The next most common error is the use of "do" without inflections or with inappropriate inflections. Examples are:

PAGE 194

Grade K: Grade 1: Grades 2/3: 180 None "Do he gots kitten in the box?" (3155079) "Do he has a kitten in the box?" (4355172) "How man y fingers did I have?" (j375183) Least common is an error which suggests that the data in Table 36 may not accurately reflect the children's degree of mastery of the auxiliary in interrogatives. Table 36 counts all occurrences of [hwA'cuw] as correct, but, in view of the examples below, these utterances may actually be unanalyzed routines: Grade K: Grade l: Grades 2/3: None "[hw A 'cuw] do you want?" (2145104) "[hw A 'cuw] do you like?" (2171104) None These examples indicate that, at least for some children, interrogatives are learned as routines; then, as the auxiliary system develops during Grade 1, there is a period when the auxiliary and the routine appear in con junction. While the majority of these conjunctions occur in Grade 1, auxiliary delineation is still active in Grades 2/3, as evidenced by the example below involving "do" and "be": Grades 2/3 : "What [r Jiw] want?" (4245165; "What are do you want?"

PAGE 195

181 Finally, there are, in Grades 2/3, a few errors which involve the formation of interrogatives by the expedient of prefacing a statement with an interrogative word, leaving the auxiliary in post-subject position: Grades 2/3: "Which one you do want?" (3345196) "What he's has in that other hand?" (3351184) 28 One other error appears to be explicable in terms of the collapsing of competing phonological systems. The children frequently produce [dttJ for "doesn't"; this form is probably an approximation of the common local pronunciation of "doesn I t", [ d A dn J. Because this form seems to be simply a variant pronunciation, it will not be discussed below. In summary, the most common type of error at all grade levels is the omission of the auxiliary. Following this is the use of "do" in place of the copular auxiliary. The next most frequent error is the use of uninflected or inappropriately inflected "do" Next in frequency is the use of two auxiliaries, one of which is embedded in a routine or fossilization The final error type is found in interrogatives in which the auxiliary does not undergo movement to pre-subject position.

PAGE 196

182 3.4.2.2. Learning strategies The vast majority of errors discussed above involve interrogatives, indicating that the primary learning task facing the children is the use of the auxiliary in such structures. The following discussion will focus on the multiple bases for this difficulty. 3.4.2.2.1. The role of the Ll A brief examination of interrogative formation in Choctaw reveals that the language accomplishes interroga tion very differently than does English. Yes/No questions are formed by suffixing an interrogative mo~pheme to the verb. Compare the two examples below: statement: Allayat oka bannah. 'The child wants some water.' Alla + (y)at oka banna + h Child SCM water want PT interrogative: Allayat oka bannah~? 'Does the child want some water?' Alla + (y)at oka banna + h + Q Child SCM water want PT INT Information interrogatives are formed by the inclusion of an interrogative word and the suffixation of the inter rogative morpheme; the latter becomes redundant and is frequently dropped in casual speech.

PAGE 197

Natah allayat bannah~? 'What does the child want?' Natah alla + (y)at banna + h + Q What child SCM want PT INT 183 These examples illustrate that the relationship between statements and interrogatives is much more direct in Choctaw than it is in English. The major differences between the two languages are in the use of a question particle in Choctaw and the insertion of an auxiliary in English; the most obvious parallel is the use of inter rogative words in both languages. The latter may facili tate positive transfer for Wh-Interrogatives, but the fact that there is no structural precedent in the chil dren's Ll for auxiliaries or for word order changes be tween statements and interrogatives suggests that, while there is thus no possibility of conflict between language systems, neither is there any opportunity for positive transfer. Other functions of the auxiliary "do" in English are accomplished as differently in Choctaw as are interrog atives. Non-emphatic negatives, for example, are formed by prefixing a negative morpheme to the verb stem which, in this case, takes a different set of personal subject affixes than does an affirmative verb; the final vowel of the verb root changes to /o/ in the negative. Compare the following examples:

PAGE 198

184 affirmative: Allayat oka bannah. 'The child wants some water. I Alla + (y) at oka banna + h Child SCM water want PT negative: Allayat oka ikbannoh. 'The child does not want any water. I Alla + (y)at oka ik + banno + h Child SCM water NEG want NEG PT Obviously the patterns of the children's Ll are dis similar to those of the L2. We can conclude from this discussion that the children must learn completely new and different structures in order to form interrogatives and negatives in English. There are neither facilitating nor contrasting patterns in their Ll to influence the acquisi tion of the auxiliary "do". 3.4.2.2.2. The role of cognitive analysis Cognitive analysis is a factor in the acquisition of "do" auxiliaries in interrogatives to the extent that some children have apparently analyzed the structure of inter rogatives with respect to auxiliary placement but have not yet distinguished among auxiliary types, e.g., "What do you do?" (3176079). The frequency of such errors, however, is insufficient to claim that this strategy is universal among the children. There is no other evidence for active analysis.

PAGE 199

185 3.4.2.2.3. The role of imitation Imitation plays an important role in the early stages of interrogative development resulting in the embedding and subsequent fossilization of auxiliaries and interrog ative words. Whether or not this inhibits the children's acquisition of auxiliaries is impossible to determine. It should be noted that this stage also occurs among English Ll learners (Brown 1973:181). In any event, there is a definite stage in the children's developing L2 in which we find fossilization and auxiliaries coexisting. These become sorted out as auxiliaries extend through the chil dren's grammars. 3.4.2.3. The auxiliary "do": summary and conclusion The children produce this auxiliary at a success rate of 64% by Grades 2/3; however, the success rates for specific forms and functions of the auxiliary vary between 34% and 86% at that same grade level. Most difficult, or least successful, is the inflected, contracted negative, "doesn't", and most successful is uninflected "do" in both statements and interrogatives. The chief error at all grade levels is omission of the auxiliary. During Grade 1, the auxiliary begins to appear frequentl y in the children's speech and the children con comitantl y experience a temporar y decline in success. An examination of errors other than omission does not point conclusively to any single learning strategy. With regard

PAGE 200

186 to the Ll, there appears to be insufficient correlation or contrast with the L2 to either facilitate or inhibit acquisition. Imitation plays an early and obvious role in interrogative structures, after which some analysis of the general structure of questions occurs, assisting chil dren to determine where to place auxiliaries in these structures. 3.4.3. Copular auxiliaries Copular auxiliaries are those which employ a form of "be". They appear in progressive tenses, e.g., "She is working", and are inflected for person, number, and tense. Because the present progressive is the true present tense in English, the frequency of copular auxiliaries in casual speech is very high. They are, in other words, a very handy structure to learn. The SWCEL presents seven opportunities for the pro duction of the copular auxiliary in present progressive verbs (Questions #29, 54, 73, 76, and 81-83). There are five statements, one of which is negative, and one affir mative interrogative. Of the seven, two require the third person singular inflection, four require the second per son, and two require the first person, singular in one instance and plural in the other. Table 48, below, summarizes the children's success with all copular auxiliary forms.

PAGE 201

187 Table 48. Copular Auxiliaries Grade # Opportunities # Successful % Successful K 1 2/3 24 106 128 13 73 102 54% 69% 80% It is obvious from Table 48 that the copular auxiliary is steadily and successfully entering the children's lin guistic systems. Not only does successful production increase across grade levels, but so does frequency of production: children in Grades Kand 1 produce, on the average, less than one auxiliary each during the test (.2 and .8 respectively), whereas children in Grades 2/3 produce 2.6 auxiliaries per child, per test. In comparing spontaneous production with imitative production (Table 49 and Figure 15, below), we find a great deal of difference. First, although imitative abilities exceed spontaneous abilities in Grade K, from there on spontaneous produc tions are more successful than imitative ones. Secondly, while imitative ability suffers a slight decline in Grade 1, no such decline occurs in spontaneous ability. It is strange that the children should produce spontaneously more successfully than they imitate, but that is, in fact,

PAGE 202

100% 90 80 70 60 50 40 30 20 10 I s K 1 2/3 Figure 15. Copular Auxiliaries, Spontaneous and Imitative Success Table 49. Copular Auxiliaries, Imitative Success 188 Grade # Opportunities # Successful % Successful K 1 2/3 77 158 120 45 88 80 58% 56% 67%

PAGE 203

the case with the copular auxiliaries. The data do not yield a satisfactory explanation for this circumstance; 189 it is quite possible that less attention is focused on a structure one already produces proficiently; it is also possible that many items counted as correct are, in fact, unanalyzed routines as discussed in Section 3.5.3.2, below. The following discussion examines differential ac quisition of the various copula forms. Unfortunately, coding procedures preclude complete separation of all forms of the copular auxiliary. Four groups may be dis tinguished: third and non-third person forms (Tables 50 and 51, below), and contracted and non-contracted forms for all persons (Tables 52 and 53, below). Table 50. Third Person Singular Copular Auxiliary Grade K 1 2/3 # Opportunities 13 52 51 # Successful 5 32 29 % Successful 38% 62% 57%

PAGE 204

Table 51. Non-Third Person Copular Auxiliary Grade K 1 2/3 # Opportunities 11 54 77 # Successful 8 41 74 190 % Successful 73% 76% 96% Comparison of these two tables indicates that the children have much less difficulty with non-third person singular forms ("am", "are") than they do with the third person singular ("is"). In fact, the former are produced well above the criterion level of acquisition. In comparing contracted and non-contracted forms, we find both approaching the criterion level, the non contracted auxiliaries slightly ahead of the contracted forms. Table 52. Contracted Copular Auxiliary, All Persons Grade K 1 2/3 # Opportunities 6 31 50 *N too small for significance # Successful 5 28 39 % Successful 83%* 90% 78%

PAGE 205

191 Table 53. Non-Contracted Copular Auxiliary, All Persons Grade K 1 2/3 # Opportunities 18 75 78 # Successful 8 45 63 % Successful 44% 60% 81% In summarizing this information, we can postulate the following order of acquisition based on success rate. Beginning with the earliest, i.e., most successful, form, the children acquire: 1) non-third person; 2) non contracted, all persons; 3) contracted, all persons; and 4) third person singular. Imitatively, a slightly different order emerges, as demonstrated in Tables 54 57, below. Here we find that the contracted forms are acquired earliest, and that third and non-third person forms are produced at essen tially the same level of success. The only correspondence between the two modes is in the third person forms which are produced with the least success in both cases~

PAGE 206

192 Table 54. Third Person Singular, Imitative Success Grade # Opportunities # Successful % Successful K 1 2/3 34 87 59 18 45 39 53% 52% 66% Table 55. Non-Third Person, Imitative Success Grade K 1 2/3 # Opportunities 37 71 61 # Successful 28 43 41 % Successful 76% 61% 67% Table 56. Contracted, All Persons, Imitative Success Grade K 1 2/3 # Opportunities 23 39 32 # Successful 19 35 26 % Successful 83% 90% 81%

PAGE 207

193 Table 57. Non-Contracted, All Persons, Imitative Success Grade K 1 2/3 # Opportunities 47 99 75 3.4.3.1. Discussion of errors # Successful 26 53 54 % Successful 55% 54 % 72% Because grammatical, but non-sentential, responses are frequently given (e.g., in ~ response to "What's she doing?" a child may respond "Playing"), it is difficult to thoroughly evaluate the children's use of the copular auxiliary. For those responses which provide an obliga tory context for the auxiliary, the following types of errors are found: omission of the auxiliary; the use of copular auxiliaries with non-progressive verbs; lack of agreement in person and/or number; and the substitution of an inappropriate auxiliary for the copula. Omission errors are the most common and are found at all grade levels. Examples are: Grade K: Grade 1: "He going to school" (1K60017) "You sitting down" (2K82150) "He hiding" (1135060) "Where you standing?" (4181115)

PAGE 208

194 Grades 2/3: "She walking" (2265199) "He carrying it a box" (5254205) These forms, while acceptable in local Black English, are not found among White speakers. Because the children have more contact with Anglos than with Blacks, it is unlikely that these sentences are imitative. The use of copulas with non-progressive verbs is also found at all grade levels and may, in fact, represent unanalyzed forms or routines consisting of subject pro nouns or interrogative words in which a contracted copula is embedded. Examples are: Grade K: Grade 1: "He's have kitten [h,i] ('in'?) his box" (4K550281) "He's have a ice cream" (1K68017) "He's have a book" (1151045) "He's not have it" (1127064) Grades 2/3: "He's working. He's write a pencil" (3250186) "He's have big box" (3255182) That these are probably best interpreted as routines is supported by the presence, in Grades 2/3, of a number of structures that clearly demonstrate fossilization has taken place: "What [r jfw] want?" (4245165) "We're are clapping our hand" (5273153)

PAGE 209

"Where're you're standing?" (4281168) "I'm are running" (6283157) Lack of agreement in person or number is the next 195 most frequent error, though much less common. Examples are: Grade K: Grade 1: Grades 2/3: "Because they are, both is lying" (2K34147) "Is stand in back of me" (3181079~ second person referent) "Writing, he are writing paper" (3350196) Finally, the use of an inappropriate auxiliary occurs occasionally in Grades 1 and 2/3. Examples are: Grade 1: "What do you doing?" (2176094) Grades 2/3: "What do you doing?" (5276154) Of these error types, Grade K produces the highest frequency of omissions while the errors produced by Grades 1 and 2/3 are much more variable, indicating that it is after Grade K that the copular auxiliary begins to enter the children's grammars. 3.4.3.2. Learning strategies Because omission is the most frequent error, the following discussion is brief.

PAGE 210

196 3.4.3.2.1. The role of the Ll The Choctaw expression of ongoing action is presented in Section 3.2.2.2.2.1. To summarize, there is little similarity, other than semantic, between Choctaw and English progressives. Specifically, the auxiliary (or, more correctly, positional in Choctaw) follows the verb in Choctaw and precedes it in English; the main verb receives a suffix in Choctaw and does not in English; the main verb may optionally retain agent pronoun mor phemes in Choctaw, but obligatorily loses the third person singular morpheme in English. Finally, when sentences are changed to their corresponding Yes/No Interrogatives, the Choctaw auxiliary does not change position in the question, but the English auxiliary does. There is no evidence that the children's Ll pattern appears in their developing progressive system in English. The differences between the two languages are so great in the case of auxiliaries and progressive tense formation that there appears to be little possibility for either transfer or structural interference. 3.4.3.2.2. The role of cognitive analysi~ The evidence for an analytical approach to learning copular auxiliaries is the same as that cited for the auxiliary ''do". Overgeneralization in the form of substi tution of one auxiliary for another suggests that, to a limited extent, the children approach English auxiliaries

PAGE 211

in interrogatives systematically and holistically, but have not yet differentiated among lexical and functional subsets. 3.4.3.3. Copular auxiliaries: summary and conclusion 197 Copular auxiliaries are produced at the 80% level of success by children in Grades 2/3. This high rate of suc cess as compared with, e.g., "do", may be a result of a high rate of input frequency in classroom conversation. In comparing different forms of the auxiliary, we find that children have mastered the non-third person forms most successfully, and the third person singular forms least successfully. Because omission is the principal error, little can be deduced concerning the children's learning strategies; it does appear, however, that some children approach the learning task analytically. 3.4.4. The auxiliary "will" Although the auxiliary "will" is known as the future tense marker in English, in fact, i.e., in the local dialect of English being learned by these children, future time is usually expressed by the use of "gonna", e.g., "I'm gonna do it", or by contracted "will", e.g., "I'll do it". Non-contracted "will" is reserved for emphatic intention. It is not surprising, then, that those SWCEL items (#37, 50, and 78) designed to elicit "will" usually are

PAGE 212

198 responded to with "gonna". As a result the data base for the auxiliary "will" is extremely limited. The children's spontaneous use of this form is presented in Table 58, below, but because the numbers are so small, no signifi cance should be attached to the high success rates. Table 58. The Auxiliary "Will" Grade K 1 2/3 # Opportunities 11 14 9 # Successful 10 7 7 % Successful 91% 50% 78% Imitatively, however, the children are much more active. Table 59, below, reports their success. Table 59. The Auxiliary "Will", Imitative Success Grade K 1 2/3 # Opportunities 36 69 88 # Successful 20 49 69 % Successful 56% 71% 78%

PAGE 213

199 Although the imitative data are much more extensive, the success rate in Grades 2/3 is the same for both modes. The only conclusion to be drawn from these figures is that the children imitate a great deal more readily than they spontaneously generate this auxiliary. 3.4.4.1. Discussion of errors The only error is that of omission. Be~ause of the widespread use of "gonna" in casual speech, this form is not considered erroneous. 3.4.4.2. Learning strategies The following discussion will present, primarily, the Choctaw future tense. There is no evidence for extra polating learning strategies since the only error is that of omission, and since the children's use of this tense is limited. 3.4.4.2.1. The role of the Ll The semantic parameters of the future tense in Choctaw correspond to those of English. Any event not in process at the moment of speaking must be referred to as future. For example, if someone is walking out the door, it is correct to ask her where she is going in the present tense. However, if someone is verbally contemplating leaving, one must ask the equivalent of "Where will you go?" The future tense in Choctaw is formed by dropping the

PAGE 214

200 final vowel of the verb stern and suffixing the future tense morpheme {-acih}. As illustration, using the root balili 'to run', compare each of the members of the fol lowing pairs. The first is in the present tense and the second in the future. Balililih. I am running. Balili + li + h run I Balililac.th. 'I will run. Balili + l(i) run I Isbalilih. PT + acih FT 'You are running.' Is + balili + h you run Isbalilac.th. 'You will run .' PT Is + balil(i) + acih you run FT The differences between the two languages are quite obvious. Choctaw employs a morpheme suffix to express the future tense, whereas English preposes "gonna" or

PAGE 215

contracted "will". Choctaw marks person and number in this tense, whereas these attributes are not marked on the English verb root. It is possible that these differences result in the slow rate of acquisition demonstrated by the learners. It is more likely, though, that "will" is simply not a frequent morpheme in the local dialect; nor, it should be noted, is this form presented in the ESL teaching materials used in the school. These materials present future time as "going to". 3.4.4.2.2. The role of cognitive analysis Because error data are limited to omission, there is no evidence that the children utilize, at this point in the acquisition process, cognitive analysis as a learning strategy. 3.4.4.3. The auxiliary "will": summary and conclusion Spontaneous use of the auxiliary "will" is rare in the corpus. Apparently this auxiliary has not entered the children's productive competence. Imitative use, on the other hand, is much more frequent and increasingly successful as the children age. Support for the conclu sion that this auxiliary is just beginning to enter the children's grammars is found in the facts that the sole error is omission and yet, at the same time, imitative production is relatively successful, suggesting that at 201

PAGE 216

202 the passive level the children are acquiring the auxiliary. It is probable that as the auxiliary develops errors of usage will diversify. At this point, it is impossible to determine learning strategies. 3.5. Morphology: Summary and Conclusions This section summarizes and compares the success rates of the inflections examined in the preceding discussion. The rates for Grades 2/3 are then compared with acquisi tion orders of English Ll learners and with English L2 learners from different language backgrounds in order to determine if the children who are the topic of this study are similar to other learners of English. The chapter concludes by examining the limits and difficulties of determining the role of the children's Ll in the develop ment of their L2. 3.5.1. The rate and sequence of acquisition The first part of this discussion examines individual inflections in terms of their developmental rate, or "acti vity" level over the age groups tested. This examination attempts to determine if there is any relationship between this index and the acquisition history of the individual morphemes. The second part presents and discusses the success rates for all grades and all inflections in both spontaneous and imitative productions.

PAGE 217

203 3.5.1.1. The rate of acquisition As has been pointed out, not all inflections made steady, upward progress across grade levels. Many exper ienced a temporary period of decline. The rate of acqui sition is examined below from the perspective of over-all change, as measured by the difference in percentage points, between Grades Kand 2/3. Table 60 is an index of inflec tional activity between those grades only. Grade 1 fluc tuations are omitted from the summary in order to compare over-all gain and loss. When we compare the inflections' activity levels with the success rates exhibited by Grades 2/3, there is only sporadic correlation between the degree of activity and the degree of success. Table 61 arranges the inflections in order of their success in Grades 2/3, beginning with the most successful. Those demonstrating an increase of more than 8 points are marked with a ''+", and those with a de crease of more than 8 points with a"-". Although there is a loose correlation between amount of change and success, it is not sufficiently strong that we could conclude that the most successful inflections, without exception, are those that undergo the greatest growth, nor the least successful those that decline the most. Nonetheless, the index, by reflecting the morphemes' relative progress, reflects those inflections which the children find relatively more and less difficult to learn.

PAGE 218

204 Table 6. Percentage Point Change in Successful Spontaneous Production between Grades Kand 2/3 1. 2. 3. 4. 5. 6. 7. 8. Inflection Copula*,+ [Contract] Present Tense "s" AUX "will" Past Tense, Irregular AUX "be",+ [Contract] AUX "do" Pronoun, Possessive Progressive "-ing" 9/10. Articles 9/10. Pronoun, Post-Verb 11. Noun Possessive "s" 12. Pronoun, Pre-Verb 13. Copula*, [Contract] 14. Noun Plural "s" 15. AUX "be", [Contract] Change -34 -14 -13** -11 5 3 0 + 6 + 7 + 7 + 8 +11 +12 +17 +37 *Copula data, from Chapter IV, included here for comparison with other studies. **N too small for significance; hereafter excluded from consideration.

PAGE 219

Table 61. Grades 2/3 Spontaneous Success Order and Activity Indices Order 1. AUX "do" 2. AUX "be", [Contract] 3. AUX "be", + [Contract] 4. Pronoun, Post-Verb 5/6. Copula, [Contract] 5/6. Progressive 11 -ing" 7. Pronoun, Pre-Verb 8. Pronoun, Possessive 9/10. Copula,+ [Contract] 9/10. Noun Plural "s" 11. Articles Change over 8 points + + + + 12. Present Tense "s 11 Irregular 13. Noun Possessive "s" 14. Past Tense, Irregular 205

PAGE 220

The following section presents the resulting sequences of acquisition. 3.5.1.2. The sequence of acquisition 206 One of the more interesting findings of this investi gation is that there is no consistent sequence of acquisi tion across grade levels; that is, there is no orderly emergence and regular increase in success of the structures examined. Although variability has been found in other studies, it is limited compared with the variability found in this population. Figure 16, below, which is arranged in the order of success achieved by Grades 2/3, demonstrates that, had the chart been arranged according to the ascending success rates of either Grades Kor 1, different sequences would have resulted. To a certain extent, the variability indicated in Figure 16 can be categorized by grouping inflections ac cording to their acquisition histories. The first category contains those inflections that experience progressive, if uneven, increases between each of Grades K through 2/3. These are: Articles, Noun Plural "s'', Pre-Verb Pronouns, Post-Verb Pronouns, Non-contracted Copulas, and Non contracted "be'' Auxiliaries. For some of these inflections, however, the change between levels is so minimal that these items are, for all productive purposes, on an acquisition plateau. The items whose change is minimal

PAGE 221

1-tj I-' lQ C H ro I-' O'I tJ::l C/) Ill 'd H 0 {/) ::i -rt .. Ill ::i ...., ro 0 -c {/) 0 H CJ) I-' C N 0 0 0 ::i ro rt {/) Ill Cl) I-' C/) tJ::l C Ul Ill -H .. .. Ill ::i G) p. H Ill N p. '-ro w {/) -:,<;: C/) o1-' C/) I-' I-' p. Ill ::i tJ::l rt Ill H {/) Noun Possessive "s" Present Tense "s" Articles Noun Plural "s" Copula +[Contract] Pronoun Possessive Pronoun Pre-verb Progressive 11 -ing" Copula -[Contract] Pronoun Post-verb Aux "be" +[Contract] Aux "be" -[Contract] Aux "do" I-' 0 N 0 w 0 ,i,. 0 lJ1 0 O"I 0 --.J 00 0 \.0 ~ltal.ll1b1b1b1h1hll1ti1111111 "''''''''''''~ I-' 0 N 0 --.J

PAGE 222

208 between Grades Kand 1 are Articles and Noun Plural "s"; and between Grades 1 and 2/3, Pre-Verb Pronouns. A second category is filled by only one inflection, the Present Tense "s". This inflection is less successfully produced with each higher age level, reaching its lowest success rate in the highest level tested. A third category contains those inflections which suffer a decline in Grade 1, followed by a Grade 2/3 re covery which in some cases does and in other cases does not exceed the initial level of success. These inflec tions are: the Irregular Past Tense; Noun Possessive "s"; Possessive Pronouns; and Auxiliary "do". A smaller category includes those inflections which suffer a decline in Grade 2/3; these are the Contracted Copula and the Contracted Auxiliary "be". A final category includes those inflections which while not declining, either progress very little (Articles and Noun Plural "s") or not at all (Progressive "-ing"). An interesting observation which emerges from these categories is that Grade 1, or the second year of learning, is a difficult one for the children. Approximately one third of the inflections suffer a relapse in spontaneous success at that stage, and several others are stranded on a developmental plateau. Because a similar decline does not occur for any significant group of inflections in Grades 2/3, it must be assumed that there is some factor (or factors) specific to that stage of the acquisition

PAGE 223

209 process that has a negative effect on the children's pro gress. However, the factor is not obvious: the declining inflections are not, for example, structurally similar; nor do they all lack precedent in the children's Ll. Declines and plateaus are certainly not unknown in language learning and do not necessarily signify a complete halt in the acquisition process; passive competence may continue to grow or at least not decline. Although it is difficult to measure passive competence, comprehension tasks in which the children are asked to manipulate objects rather than speak is one method, but unfortunately impos sible with these data. A second approach, and one that is possible with the data available, is based on the assumption that imitative ability reflects, to some extent, passive competence. We can then compare the two productive modes (Figure 17) and we find that three of the four inflections which experience a decline in spontaneous success also decline imitatively: Noun Possessive "s", Irregular Past Tense, and Auxiliary "do". Thus, if imitative ability does indeed reflect pas sive competence, we can conclude that the children's passive competence in most cases parallels their productive bilities, i.e., they both suffer in Grade 1. Interestingly, this is not the case for those inflec tions which plateau or minimally increase in Grade 1 (Figure 18). The children's imitative ability for these inflections (Articles, Noun Plural "s''. and Progressive 11 -ing") continues to increase.

PAGE 224

210 100% -90 -80 _._ I 70 .... iiiii 60 .... 50 -40 -30 --20 -10 -I --....__ K 1 K 1 K 1 K 1 Noun Irreg Poss AUX Poss Past Pro"do 11 II S II Tense nouns Figure 17. Spontaneous (Solid) and Imitative (Striped) Success of Inflections Declining in Grade 1.

PAGE 225

211 100 % -90 80 70 60 50 40 I 30 i-20 ..... 10 K 1 K 1 K 1 Articles Noun Prog. Plural II -ing" Figure 18. Spontaneous (Solid) and Imitative (Striped) Success of Inflections that Gain Minimally or Plateau in Grade 1.

PAGE 226

212 In sum, we do not find any explanation for Grade l's decline in success, but we do find a strong correlation be tween productive and passive competence. Moreover, we can observe that, of all the inflections that decline or remain relative static, only one, the Present Tense "s", suffers further decline the following year; the rest demonstrate an increase in productive success so that we can at least conclude that the Grade 1 phenomenon does not have a perma nent adverse effect on acquisition of the inflections in volved. Returning to Grade 2/3's acquisition sequence (Figure 16), it is important to note that none of the inflections can be characterized as completely within the children's productive or oral competence. Two structures are in the 80-90% range of success, Auxiliary "be" non-contracted, and Auxiliary "do"; two are in the 70-80% range, Contracted Auxiliary "be" and Post-Verb Pronouns; three are in the 60-70% range, Progressive 11 -ing", Pre-Verb Pronouns, and Non-Contracted Copula; and the remaining four, Articles, Present Tense "s", Noun Possessive "s", and the Irregular Past Tense, fall below the 50% level of success. From this, we can conclude that three to four years of exposure to and instruction in the English language is insufficient if the goal for the children is oral fluency. In comparing spontaneous and imitative success, the children are generally more successful when imitating the inflections. The mean success rate for imitation is 70%,

PAGE 227

213 and the mean spontaneous success rate is 60%. Figure 19 summarizes the imitative development of the inflections considered. Again the order of success is arranged according to Grades 2/3's performance and again, ranking by other grade levels would result in different sequences of success rates. In comparing Figures 19 and 16, we find that there is some general agreement between spontaneous and imitative success and difficulty. With the exception of spontaneous Possessive Pronouns and Contracted Copulas, the same in flections fall above and beneath the means for spontaneous and imitative success. In other words, the children find the Irregular Past Tense, Noun Possessive "s'', Present Tense "s", Articles, and Noun Plural "s" the most difficult both spontaneously and imitatively, although the ranking of difficulty varies between the two modes. Spontaneously, but not imitatively, the children find Possessive Pronouns and Contracted Copulas relatively difficult. The remainder of the inflections falls above the mean for both modes and for these, as well, the ranking of suc cess varies. The greatest variance is shown in the spon taneous and imitative rankings of Pronoun Subjects, Auxiliary "be" Non-contracted, Contracted Copulas, and Auxiliary "do". We therefore conclude that, although there is no exact correlation between spontaneous and imitative productive ability for each inflection inasmuch as the success sequences do not demonstrate the same rank orders, there is, in fact,

PAGE 228

t-rj I-' LO C ti (D ...... tJ' .,_. H PJ I-' s ti ::::1 I-' [/) (D rt --[/) PJ --rt .. I-'
PAGE 229

215 an overwhelming agreement as to relative difficulty since the majority of the same inflections, spontaneously and imitatively, fall above and below the mean levels of suc cess for the two modes. A second question of interest to be addressed with the imitative data is whether or not imitative success at one level of development predicts productive increase in sueceeding levels; i.e., if imitative success at one level exceeds spontaneous success, is there a subsequent (and presumably consequent) increase in spontaneous ability? Figure 20 charts the spontaneous developmental history of all inflections whose Grade K imitative ability exceeds spontaneous ability, and Figure 21 does the same for Grade 1 inflections. The graphs are arranged by grade level for each inflection, Grades Kand 1 to the left in Figures 20 and 21, respectively, and Grades 2/3 to the right; imitative success is indicated by the striped areas and spon taneous success by the darkened areas. Here we find that the ability to imitate an inflection in Grade K is not necessarily indicative of subsequent suc cess. However, this is definitely not the case for Grade 1. With the exception of the Present Tense "s'', an excess of imitative over spontaneous success in Grade 1 absolutely predicts an increase in spontaneous success the following year. Th . . f. d. 29 is is an interesting in ing. Unfortunately, at this point, it cannot be determined whether it is the

PAGE 230

216 100 % 90 80 70 60 50 40 30 20 10 Q) ,.Q ,.Q .......... .......... .......... ti) ti) ti) H H +l ro +l ro +l Q) Q) Q) u MU MU Q) :> :> .o ro ::l ro ::l ro ti) Q) 8 01 I I H 0.l--1 0.. H Ul Ul Q) Q) +l +l 0 +l 0 +l 0 +l H H Ul >: u u P. Q) Ul H P. 0 ::l 0 0 0 8 ro H P. u u u P. + ::l Ul 0 0 Q) H 0 z H P. H P. P. Figure 20. Spontaneous Developm e nt of Same Inflections, Gr a de K Imitative Abilit y (Striped) and Grades K -3 Spontaneous Abilit y (Solid).

PAGE 231

1-Tj f-' I.Q s:::: Ii CD N ..... NG) Ii WP, p. cn CD '"d 0 ..... :::, rt H PJ ;:l :::, f-' CD rt 0 PJ s:::: rt en f-' <: cn CD s:::: () Cl) (1 s:::: CD 0 en o en CD t/l ........ en Cl) 0 ........ I-' Cl) f-' rt p. Ii .__, f-' '"d CD p. PJ :::, p. G) Ii PJ p. CD en Noun Plural "s" Article Present Tense "s" Progressive "-ing" Past Tense Irregular Pronoun Pre-Verb Pronoun Post-Verb Pronoun Possessive Aux "do" Copula -[Contract] I-' N w O'I 0 --.J 0 co 0 \D 0 1111111111111111111111 11111111111111 I-' 0 0 N I-' --.J

PAGE 232

maturational level of the children or the stage in the acquisition process which is critical. 218 Summarizing the findings of this research thus far, it has been shown that: 1. There is no regular sequence of acquisition be tween Grades Kand 2/3. The inflections do not uniformly demonstrate increased proficiency in successive grade/age levels. 2. A significant number of inflections declines in successful spontaneous production between Grades Kand 1. Although the decline is not permanent, there is no obvious explanation for this phenomenon. 3. There is a general but not exact, correlation between spontaneous and imitative success. 4. Imitative ability is, in general, more successful than spontaneous ability. 5. In those cases in which imitative ability exceeds spontaneous ability in Grade 1, an increase in spontaneous ability occurs in the following year. 6. Finally, none of the inflections is spontaneously produced at the criterion level of acquisition (90%) in Grades 2/3. Four years is insufficient time for these children to acquire ESL.

PAGE 233

219 3.5.2. Comparative acquisition sequences Comparison with other acquisition studies is made in order to determine if the L2 development of these children is similar to that experienced by others. In comparing these data with the data from children learning English as a first language, we hope to discover if first and second language learners acquire English inflections in relatively the same sequence. If so, that is strong evidence that both the learners' Ll and their more advanced age have negligible consequences on the learning process. In comparing these data with other ESL learners we can investigate more specifically the possible effects of the learners' first languages and thereby address the ques tion of whether or not universal cognitive strategies, in combination with the configuration of the L2 being acquired, are preeminent in L2 development. In order to compare this study with others, Grade 2/3 has been arbitrarily designated as the standard. Although the success ranking for that level is not completely simi lar to those of Grades 1 and 2, the data suggest that the ranking is more stable in terms of the maintenance of rela tive order than that of Grade 1. Grade K is eliminated from consideration since it is the children's first exposure to the language.

PAGE 234

220 3.5.2.1. Comparison with English Ll acquisition English Ll studies differ methodologically from this research in two important respects. First, they are, by and large, longitudinal rather than cross-sectional. Secondly, while the data base of a longitudinal study may be as extensive as that of a cross-sectional study, the former is characteristically composed of utterances from a single speaker or a small number of speakers rather than many; and these utterances, furthermore, are collected as casual, communicative contexts rather than in formal eli citation circumstances. Longitudinal studies ordinarily correlate acquisition stages with the children's mean length of utterance (MLU) (Cazden 1972:39), a measure reflecting the average length of an utterance in morphemes as an inflection develops to ward the criterion. Although MLU is a more language intrinsic yardstick than is age or grade level, it was decided that the measure would be unreliable in this study due to the formal nature of the data collection context. Before examining the data presented below, a final note of explanation is necessary. The acquisition se quences appear in rank order, but these are not statistical ranks. As a result, the intervals between ranks are not observable and it is impossible to determine more than relative sequence. It should be noted that the intervals between ranks may be quite different for different groups of learners and this is significant information; however,

PAGE 235

221 this information is not always available in the literature : It is to be hoped that this brief preamble will answer questions of method and comparability which might cloud conclusions. One of the most extensive English Ll acquisition studies is that of Roger Brown and his colleagues (Brown 1970, 1973). The speech of three children, Adam, Eve, and Sarah, was collected regularly in a naturalistic setting over a period of five years. The children's speech has been subjected to intensive analysis and forms the basis of much of our knowledge of English Ll acquisition. Brown contends that other studies of English Ll devel opment (deVilliers and deVilliers 1973; Ervin and Miller in Miller and Ervin 1964; Leopold 1937; and Menyuk 1969) confirm, substantially, the order of acquisition he reports, and concludes that this order therefore is "fairly independent of criterion of acquisition, children studied, and investi gator" (1973:282). The acquisition order found by Brown is reported in Table 62, below, and compared where possible with the suc cess levels of the Grade 2/3 children of this study. The most successful inflections (or in Brown's study, those which achieved the criterion level first) appear at the top of the table. It is immediatel y apparent that the two sequences are dissimilar. On closer examination, howe v er, some simi larities emerge i n relative sequencing. In both studies,

PAGE 236

222 Table 62. English Ll and L2 Acquisition Sequences English= Ll 1. Progressive 11 -ing" 2/3. In, On 4. Noun Plural "s" 5. Past Tense, Irregular 6. Noun Possessive "s" 7. Copula, -[Contract] 8. Articles 9. Past Tense, Regular 10. Present Tense "s", Regular 11. Present Tense "s", Irregular 12. AUX "be" -[Contract] 13. Copula +[Contract] 14. AUX "be" +[Contract] ' English= L2 1. AUX "do" 2. AUX "be" -[Contract] 3. AUX "be" +[Contract] 4. Pronoun, Post-Verb 5/6. Copula, -[Contract] 5/6. Progressive "-ing" 7. Pronoun, Pre-Verb 8. Pronoun, Possessive 9/10. Copula, +[Contract] 9/10. Noun Plural "s" 11. Articles 12. Present Tense "s", Irregular 13. Noun Possessive "s" 14. Past Tense, Irregular Progressive "-ing" is ordered before the following inflec tions: Contracted Copulas, Noun Plural "s'', Articles, Present Tense "s'', and the Irregular Past Tense. Secondly, Noun Plural "s'' is ordered before: Articles, Present Tense "s", and Noun Possessive "s". Articles appear before Present Tense "s" and, finally, the non-contracted form of the AUX "be" is ordered before the contracted form, as the

PAGE 237

non-contracted form of the Copula is ordered before the contracted. 223 These similarities imply a hierarchy of difficulty for all English language learners, whether Ll or L2. The dif ferences, on the other hand, suggest that the Choctaw children's greater sophistication with the concept of language itself, and/or their own first language, affects their L2 development. The most important difference is in the early emergence of both the contracted and non-contracted forms of the AUX "be" among the Choctaw children. Since auxiliaries have little or no precedent (depending on the analysis of Choctaw) in the children's Ll, it is difficult to explain their early appearance in L2, as compared with Ll, acquisition. The somewhat vague assumption that L2 learners possess greater metalinguistic sophistication may, in this case, have the concrete effect of learners focusing on obligatory, non-redundant L2 features which are new to h 1 . l . 30 t em, resu ting in ear y acquisition. We will return to this point in the discussion of other L2 sequences. The second major difference between the Ll and L2 learners is in the acquisition of the Irregular Past Tense. For Ll children it is a structure that is mastered early in their development; for the L2 children, it is the most difficult of all items e x amined. Although again we must speculate, it seems quite possible that the explanation for this lies in the fact that these forms are not rule-governed, and the L2 children with their much more limited exposure

PAGE 238

224 to the language are consequently much more dependent on hypothesis formation and the generalization of rules than are children who enjoy natural, casual input and feedback. Thus, forms which must be learned through rote memoriza tion are more difficult for these children since their op portunity to hear them is limited. Moreover, there is no Ll precedent for these forms, although some Choctaw b t 1 1 h b 31 ver s are qui e irregu ar a ong t e num er axis. The final difference is in the ordering of Noun Pos sessive "s" which appears relatively early in the Ll se quence and quite late in the L2 order, although in both sequences it follows the acquisition of the homophonous Noun Plural. In an attempt to generalize the similarities and dif ferences, the inflections can be grouped according to whether they fall above of below the developmental median 32 which, for the Ll children falls between the Non-contracted Copula and Articles, and for the L2 children, between Pre-Verb and Possessive Pronouns. On this basis, it can be stated that for both Ll and L2 learners, Non-contracted Copulas and Progressive "-ing" are relatively "easier" and that Articles and the Present Tense Irregular "s'' relatively more "difficult" than all other inflections. Beyond these facts and certain sequencing similarities mentioned above, the two groups are not comparable. We therefore conclude that the children in this study learn English quite differ ently than do children learning English as their first language.

PAGE 239

225 3.5.2.2. Comparison with English L2 acquisition The most comparable data available for child ESL learners are from studies of children speaking Japanese (Hakuta 1978:132-147), and Spanish and Chinese (Dulay and Burt 1974b) as their first languages. Because each study considered, to a certain extent, different inflections, discussion below is partially limited by this fact. This study's data for Copulas, AUX "be", and Pronouns have been 11 d f 1 33 co apse to aci itate comparison. Comparisons are made in terms of the orders in which the inflections achieved the acquisition criterion or in terms of their success rate order. Again, it is relative sequences that are being com pared; the data do not reflect the time depth of the developmental process nor the intervals between ranks. Hakuta reported the language development of a five year old Japanese girl transplanted to the u. S. and ex posed to English in both school and social contexts. One of his purposes was to compare Ll and L2 acquisition se quences so he therefore considered several of the morphemes examined in Brown's study and followed a similar methodology in scoring utterances. Hakuta concluded that the order in which criterion level was achieved substantially replicated the order found by Brown for English Ll learners. The two orders are presented in Table 63 below. numbers are preserved.) (Original rank There are, indeed, similarities in early and late emerging inflections, but there are also differences and

PAGE 240

Table 63. English Ll and L2 Acquisition Inflection Progressive -ing Copula AUX "be" in, to Noun Possessive "s" Past, Irregular Noun Plural "s" Article Present Tense "s"* Past, Regular L2 (Hakuta) 2 2 2 4/5 6 7 8 9 10 11 Ll (Brown) 1 7/12 11 2/3 6 5 4 8 10 9 *Regular for Brown, Regular and Irregular for Hakuta 226 these are of particular interest to this discussion. The major difference is in the L2 child's early acquisition of the Copula and the AUX "be"; we find this difference replicated in the experience of the Choctaw children. Table 64 compares the Japanese child and the Choctaw chil dren. Only morphemes considered in common are listed. Original rank numbering is preserved. Relative to English Ll learners, both the Japanese and the Choctaw children acquire the Copula and the AUX "be" very early in the acquisition process. The two groups

PAGE 241

Table 64. Comparative English L2 Sequences Inflection Progressive -ing Copula AUX "be" Noun Possessive "s" Past, Irregular Noun Plural "s" Article Present Tense "s", primarily Irregular Japanese 2 2 2 8 9 10 11 12 Choctaw 4 5 2 9 10 6 7 8 227 are also similar to each other and different from Ll learners in that they acquire the Noun Possessive "s" before they acquire the Irregular Past Tense. Nonetheless, these same inflections also serve to discriminate between the two L2 groups in that, relative to the sequential group of Noun Plural Article Present Tense ''s", the Japanese child acquired the Noun Possessive and Irregular Past Tense before she acquired the sequential group, and the Choctaw children acquired these two structures after this group. The explanation or the relatively early success with au x iliaries and copulas displayed by both groups of L2 learners is not clear. While there is a precedent in Japanese for these structures there is, as mentioned earlier,

PAGE 242

228 no real precedent in Choctaw; yet the latter group ac quires the structures almost as early in the order as the Japanese child. In contrast, English Ll learners are rela tively late in acquiring these items. Turning to a study of Chinese and Spanish children, we again find that the Copula and AU X "be" emerge earlier, compared with Ll children, but not, in the case of the AUX, as early as in the sequences for the Japanese child and the Choctaw children. (Table 65) Table 65. Three English L2 Acquisition Sequences Inflection AUX "be" Pronoun Placement* Progressive -ing Copula Noun Plural "s" Article Present Tense "s", Irregular** Noun Possessive "s" Past, Irregular Choctaw Japanese Chinese and Spanish 2 2 6 3 1 4 2 4 5 2 3 6 7 8 9 10 5 6 7 3 4 5 2 11 10 8 *Pre-Verb and Post-Verb; not included in Japanese study. **Japanese data include both regular and irregular forms.

PAGE 243

229 The Chinese and Spanish children were found by Dulay and Burt (1974b) to have an accuracy order (success rate) sufficiently similar that a single acquisition sequence de scribed both groups of learners. Therefore they appear as a single group in the table above and in discussion below. In comparing the three groups of data, there is rela tively little specific correlation. It can be stated that Progressive "-ing", the Copula, and Noun Plural "s" all emerge early, but the sequencing of the three inflections, relative to each other, is not similar. The Choctaw and Spanish/Chinese children resemble each other in the early emergence of Pronoun placement, a feature unfortunately not included in Hakuta's study. The Choctaw and Spanish/Chinese children also resemble each other and differ from the Japa nese child in the relatively late emergence of the Irregular Past Tense and Noun Possessive "s". However, the correla tion among the three groups is so general and marred by major exceptions, that we cannot conclude, as did Dulay and Burt, that "universal cognitive mechanisms are the basis for the child's organization of a target language, and that it is the L2 system, rather than the Ll system that guides the acquisition process" (1974b:2). On the basis of these data, that conclusion must be rejected as too encompassing. Although Dula y and Burt's hypothesis ma y account for the similarities among d ifferent groups of learners, it d o e s not allow for the differences which are undeniabl y present.

PAGE 244

3.5.2.3. English Ll and L2 acquisition: summary and conclusion 230 The learners of this study are in many ways unique when compared with English Ll and other English L2 learners. We cannot support the position that English is learned as a second language in much the same way as it is learned as a first, basing our comparison on relative sequences of morpheme appearance and success. Nor do we find strong similarities among groups of children from unrelated Ll backgrounds learning ESL. We must therefore conclude that the strategies employed by older children, as these chil dren are, may be, if not different, at least applied in different circumstances; that differences in learning con texts may be relevant to acquisition; and that the role of the learners' Ll must be considered. The last of these factors is the topic of the next and final section of this chapter. 3.5.3. On determining the role of Choctaw in the acquisition of English This discussion deals primarily with the children's Ll. The role of cognitive analysis as a learning strategy has been demonstrated where data exist in the discussion of individual inflections, and the previous section has shown that cognitive analysis can be only one of the factors affecting the various groups' acquisition of English since their acquisition sequences are different from each other.

PAGE 245

231 While the role of the children's Ll must definitely be taken into account, it is not at all clear how to do so. Contrastive analysis has been most successfully applied to phonological systems. In the areas of morphology and syntax, which necessarily entail semantics, the limits of "precedent" and "contrast" become less easy to demarcate. Do we consider semantically similar concepts with different morpho-syntactic realization to be capable of providing potential learning benefits through transfer? As a con crete and relevant case, consider the question of the semantic concept "number". Choctaw not only recognizes number as a delineation device but obligatorily expresses singular, dual, and plural number on verb stems, and sin gular and plural in one particular class of adjectivals. However, nominals are not inflected for number in the language, but are in English. Does this constitute con trast? Or, does the fact that number is recognized else where in the learners' Ll constitute precedent? If we decide that only structurally and semantically similar features have the potential to facilitate transfer, then the question is, how similar do they have to be? For example, there are many similarities between Choctaw and English pre-verb pronouns. However, Choctaw personal affixes are (a) bound, and (b) sometimes suffixed and sometimes prefixed to the verb; English pronouns are (a) free, and (b) precede the verb. Are the two systems

PAGE 246

sufficiently similar to place them on a Stockwell-Bowen type hierarchy in a relatively low position? 232 Finally, there is the issue, already raised, con cerning the extent to which lack of precedent can be con sidered detrimental to learning. For example, if English speakers were learning Choctaw, there is no precedent in their language for a dual number in the verb system. Will the learners find this difficult because there is no Ll precedent? Or, alternatively, will they find it easy to acquire because it is an addition to rather than a dif ference from their own linguistic system? Stockwell and Bowen (1965:9-18) label such cases instances of "zero transfer" and rank them, in a hierarchy of difficulty, as representing the first magnitude of difficulty for the learner. Data from this study, however, suggest that our knowledge of learning processes is insufficient to assume "new" learning to be the most difficult. In an attempt to determine whether or not predictions based on the most general definitions of precedent predict transfer and interference, Hakuta (based on Brown 1973) com pared Japanese and English on the basis of salient semantic notions of English inflections (1978:145). Table 66 presents this information for Choctaw and English and furthermore notes whether there is a similar structural precedent or an y structural precedent in Choctaw for the English inflection.

PAGE 247

233 Table 66. Ll Precedent for L2 Inflections Inflection Any Semantic Similar Any Structural Precedent Structural Precedent Precedent AUX "do" X X AUX "be" X X Pronoun Placement and Case X X X Progressive -ing X X Copula X Noun Plural X X Article X X Present Tense II s" X X Noun Possessive X X Past, Irregular X X On the basis of Table 66, we would predict that Pronoun Placement and Case would be achieved more success fully than all other inflections and the copula least suc ~essfully. Neither of these predictions is, in fact, met by the data, although pronominal placement and case is achieved relatively successfully; but then so is the copula. Nor is the table helpful in sorting out sequences of dif ficulty since, structurally, the two languages are extremely dissimilar and since, semantically, concepts may be ex pressed in different domains and through different surface mechanisms.

PAGE 248

234 A second approach, only partially explored in the pre ceding section, is to compare the acquisition of the same inflection by different Ll children, with specific refer ence to that inflection's precedent in the various first languages. We will focus on those inflections which were suspected of demonstrating Ll influence and for which fur ther evidence was sought. These are: Noun Plural "s", Noun Possessive "s", Articles, Present Tense "s", and the Irregular Past Tense. The Noun Plural "s" appears more or less in the middle of all three sequences considered earlier. It would appear therefore that the children's Ll has little influence with regard to this inflection, since Spanish, Japanese, and Choctaw (Chinese is omitted from discussion) offer quite different opportunities for transfer. Spanish does inflect nouns for number, but neither Japanese (Hakuta 1978:143) nor Choctaw does. Noun Possessive "s" appears very low in the Choctaw and Spanish order, and in the middle of the Japanese se quence. None of the three languages inflects nouns for possession in the same manner as English. Spanish employs a prepositional phrase following the possessed noun; Japa nese employs a particle linking the possessor to the pos sessed in that order (Brown 1973:11); and Choctaw inflects the possessed item with a personal affix referent to the possessor. Without a more precise definition of what con stitutes similarity, it seems premature to conclude, in

PAGE 249

235 order to explain the earlier emergence in the Japanese sequence, that the Japanese system resembles the English system more closely than do the others. Articles appear early among the Spanish children and emerge in the lower range of inflections for the Japanese and Choctaw children. Spanish has an article system simi lar to that of English, Choctaw has a system minimally similar, and Japanese has none (Hakuta 1978:144). Although the Spanish precedent may give these speakers the advantage of transfer, it is not clear why Choctaw and Japanese chil dren should be similar, given the differences in their first languages compared to English. Present Tense "s" appears low in all three sequences, and lowest among the Spanish children. Spanish marks per son on verbs (in all tenses); Japanese marks neither person nor number; and, in Choctaw, the third person singular is unmarked. Ll facts, therefore, are not reflected in the comparative orders of acquisition. Finally, the Irregular Past Tense appears in the middle range of the acquisition sequence for the Japanese child, somewhat below the mid range for the Spanish children, and at the lowest point of the Choctaw children's sequence. Spanish is replete with irregular verbs in the past tense, but there is no precedent in Choctaw. This difference may explain the difference in success between these two groups In sum, two items, Noun Plural "s" and Present Tense "s", do not appear to reflect Ll influence; two items, Noun

PAGE 250

236 Possessive "s" and the Irregular Past Tense, possibly re flect Ll influence; and one item, Articles, seems to par tially reflect Ll influence. This exercise attests to the difficulty of practicing contrastive analysis in the areas of morphology and syntax among unrelated languages. Without greater precision in the cognitive and linguistic parameters of "precedent", "contrast", and "similar", claims and predictions are dif ficult to either substantiate or trust. This research can do no more in this area than it has so far done, and that is, essentially, make data available for further study of the question which so obviously deserves investigation. The data have provided no immediate conclusions concerning the role of the children's Ll, but careful examination of the implications has suggested that current notions of Ll influence may not do justice to the complexity and capacity of the brain, i.e. comparing simple structural disparity and/or isomorphism between two languages may be too sim plistic and mechanistic to describe, let alone explain, the complex interactions which occur between languages and brains during the learning process. We are left with the fact that the children of this study are, when compared with others, in many ways unique in their acquisition of English inflections. Their first language has been found to account for some facts of their developing L2; the L2 itself appears to have caused and compounded occasional confusion; and the children have

PAGE 251

237 attempted, through hypothesis formation and generalization, to reduce the learning task to a proposition of manageable proportions. There is no evidence whatsoever that the L2 has imposed a learning logic resulting in an acquisition sequence common to all ESL learners. 3.6. Notes 1. The test provided no opportunity for eliciting the mor phophonemic alternate of the indefinite article ("an") and an insufficient number for analysis was produced in casual conversation during the testing. The plural and possessive suffixes have voiced and voiceless allomorphs, but these were not distinguished by the coding system since, in ge neral, the quality of the recording was not sufficient to the task. The third allomorph of the plural, syllabic [-fz] (as in "horses") was not elicited by the test and did not occur in casual conversation. 2. No implication of inherent latency is here intended. 3. See Chapter Two, Note 15. 4. These and following examples are adapted from Gibson and Kwachka (1978) and from Nicklas (1972). Any errors are the responsibility of this writer. 5. Abbreviations used for Choctaw data are: SCM, Subject Case Marker; PT, Past Tense; RPT, Recent Past Tense; FT, Future Tense; 3 p.s., Third Person Singular; 3 p.d., Third Person Dual; 3 p.pl., Third Person Plural; INT, Interroga tive; and NEG, Negative. 6. A limited class of adjectives furthermore requires a plural or distributive prefix when the referent is plural. 7. It is conceivable and likely that different strategies predominate at different stages of learning. Cazden (1972) notes that, as an inflectio~ approaches the acquisition criterion, overgeneralizations are more widely produced than previously.

PAGE 252

238 8. The possessive suffix is, in this item, attached to a proper noun. Although I know of no research comparing the acquisition of suffixes with reference to this noun class variable, it is a possible, if unlikely, factor in the acquisition process. 9. The number of such errors is greater in imitative productions. 10. The omission percentage may be slightly inflated. A change was made in the coding system midway through the analysis, so these figures are a result of a hand-sort of article errors and a subtraction of the latter from the total number of errors. If there is some inflation, it is probably nullified by the fact that the majority of re sponses to Question 70 are counted as "correct". Question 70 requires a mass or generic noun response (i.e., no article) and the correct omission of the article in these responses is probably not a result of children having sub classified mass and generic nouns and restricted articles from co-occurring with them, but rather from their not having fully acquired article prefixes. It is to be hoped that the count is thereby balanced. 11. Included in this count are instances where the refer ent is visibly plural (the stimulus picture is three cats) and the plural suffix is omitted, e.g., "a cat" (3K56137). 12. With respect to the body part and kin term nouns, the testing context clearly calls for possessive pronouns rather than articles. 13. Again, the context demands a noun. The classification of these as article subcategorization errors, rather than errors of noun omission, is discussed in Section 3.2.1.3.2.1. 14. Forms such as "write-a" (p. 97) are not counter examples to this claim. English loan verbs are regularly suffixed in this fashion, e.g., kala 'call (by telephone).' 15. "Got" also appears with other than third person forms, but it is notable that a common southern dialect form, "gots" with non-third person subjects, is not found in the corpus. 16. This suffix also functions as a derivative morpheme, partially serving to distinguish verbs from nouns, e.g., nowa 'a walker'; nowah 'she / he / it walks'. 17. A number of utterances are excluded from this dis cussion. There are of the type: "She have (N) the box" (6252159). Because of the common reduction of [-ing] to a

PAGE 253

239 syllabic nasal (Fischer 1964) and the resulting homophony with the preposition "in'', a structure which has not been fully mastered by the children and therefore may appear in unexpected contexts, it is impossible to classify these and similar utterances, i.e., there are two possible inter pretations: "She have in the box" and "She having the box". 18. This e x ample is from Nicklas (1972:189), but my analysis is unsatisfactory. As mentioned in Note 14, above, the inflection in question functions as a verbalizer as well as a tense marker, but its function is this example is not clear. 19. There is also a Very Remote Past Tense, formed quite differently, which is reserved for reportative and legen dary information. 20. Here the child seems to have solved gender and case problems by the ingenious solution of creating a combina tory pronoun. 21. It is a fact, however, that many Choctaw speakers ex perience difficulty with this English phoneme. For example, one community, Tucker, is ordinarily pronounced [tAk'e]. Obviously further reinforcement is provided by southern dialect pa tterns which replace word final post-vocalic (r) with an upgliding mid-central vowel. 22. This form itself may persist in the language as an overgeneralization of the (h) of other English third person pronouns; since it is an archaic form of "it", it may, on the other hand, persist simply as a relic. 23. A careful analysis of the voicing of the final sibilant might solve the classification problem of these segments, assuming, of course, that the children observe voicing dis tinctions, i.e., "hits" (hits) and "his" (hiz); however, there is no certainty that the children make-this distinc tion since sibilants are -[Voice] in Choctaw. 24. It is, unfortunately, impossible to separate inter rogative and statement modes completely due to data coding procedures. 25. A large number of occurrences of "don't" accompanying a third person singular subject are omitted from the error count since this form is standard in the local variety of English. 26. The combined success rates will be referred to later and are: K, 88 % ; 1, 59 % ; 2/3, 85%.

PAGE 254

240 27. The combined imitative success rates will be referred to later and are: K, 77 % ; 1, 70 % ; 2/3, 79%. 28. This e x ample is included because the appropriate auxiliary is "does", although it would appear "is" was intended. 29. Whether or not this finding has pedagogical applica tions is debatable. While the obvious teaching implication is to increase patterned drill during this stage, recent research (Seliger 1978:11-18) suggests that holistic learn ing is a right brain function and not directly usable by the language generating mechanisms of the left brain. 30. This very tenable speculation flies in the face of an equally tenable assumption held by many language teachers and behaviorists: that is, L2 learners do not notice com pletely new structures because, it is thought, they have no basis, i.e., Ll precedent, for recognition. I rather suspect both positions to be valid and the interesting question is whether or not there is a consistency to the categorization of noticed/unnoticed structures by learners of the same L2 from different Ll backgrounds. 31. It is very doubtful this irregularity in the Choctaw number and verb system constitutes precedent inasmuch as the relationship between singular, dual, and plural forms of the same verb is opaque to most speakers. The three forms are regarded as separate verbs. 32. There is a question of comparability here, since the sequences are not ranked. (Eliminating non-mutual morphemes does not affect conclusions.) On the bases of similar longitudinal exposure to the language, the medians may be designated as comparable. In fact, however, they are not because of the differentes in language exposure context. This is one of the many vexing problems which arise in attempting to compare data measured by different "yard sticks". 33. This results in the following acquisition order: 1 AU X "do" ( 8 5 % ) 2 AU X "be" ( 8 0 % ) 3. Pronoun Case (70 % ) 4. Progressive -ing (68%) 5. Copula (64 % ) 6. Noun Plural (47 % ) 7. Article (47 % ) 8. Present Tense, Irregular (36%) 9. Noun Possessive (30 % ) 10. Past, Irregular (28 % )

PAGE 255

CHAPTER FOUR SOME OBSERVATIONS ON SYNTAX 4.1. Introduction No matter what aspect of language is selected for study, it is the ability to generate sentences which is the final criterion of adequacy both of speakers and of gram mars. This chapter examines the children's sentences, but for both sociolinguistic and linguistic reasons the data are not extensive. Socially, phrasal responses are accep table in casual contexts and, in general, the test inter viewers succeeded in achieving this atmosphere, desirable in terms of promoting speech but not necessarily sentence level speech. Linguistically, sentence-level integration is extremely complex and many of the children simply have had insufficient experience with the English language to permit this degree of grammatical fluency. There are two major areas of discussion. The first is the children's performance on the various sentence types included in the test. These are: 1) Transitive; 2) In transitive; 3) Interrogative; and 4) Copular. Within each of these types are several variants, i.e., transitive sentences may have one or two objects, may be affirmative 241

PAGE 256

or negative, etc., and where data permit subtypes are evaluated. 242 The second section of this chapter will consider some general difficulties experienced by the children in pro ducing sentences. Included in this discussion will be an examination of obligatory inclusion of specific grammatical role realizations, and the use of adjectives and locative prepositions in copular sentences. Unfortunately, comparison with other acquisition re search is not possible with these data. It is at the sen tence level that the method of categorizing linguistic development adopted by this study is most inadequate. Grade levels simply fail to describe the individual steps in the developing stages of syntax. For this reason, it is neither possible to observe sentence evolution closely nor to compare these children's development with that of others. 4.2. Sentence Types and Success In order to facilitate data presentation, the following labels will be used in discussion. Transitive sentences are labeled Type I; Interrogatives, Type II; Copular, Type III; and Intransitive Non-copular, Type IV. The SWCEL includes 22 opportunities for Type I sentences, 13 for Type II, 19 for Type III, and 3 for Type IV. The sentences are scored according to the following hierarchy of difficulty judged by mode of reply. The most

PAGE 257

243 difficult sentence types are those resulting in no response or inappropriate responses. The ne x t order of difficulty is represented by those types which elicited one-word re sponses. Following this are sentence types resulting in phrasal responses. The least difficult are those types to which children responded with complete and grammatical sen tences. It is to be understood that this hierarchy does not claim to represent degree of grammatical complexity; it reflects degree of difficulty from the learners' collective point of view. These two perspectives are not necessarily contingent or even related. 4.2.1. Overview of sentence success The data below provide a summary of the children's performance on the four sentence types across grade levels. To some extent, Table 67 provides a measure of passive as well as oral competence, as will be discussed below. Table 67. Sentence Types Ranked by Three Response Variables Difficulty No Response Inappropriate R. Score I + Difficult II ( 6 6) l II (73.7) II (96.4) Difficult I (41.2) IV ( 16) I ( 2 81) Easy IV (32.7) III (11. 6) III (297 .1) + Easy III (32) I (10.1) IV (335.3)

PAGE 258

244 The columns headed "No Response" and "Inappropriate Response" provide an indirect measure of the children's comprehension, or passive competence. The measure is not straightforward because what the children respond to in the testing situation are questions, or requests for ques tions, so that Type II questions, which they find most dif ficult overall, are of the form: "Ask me if/what X". Type III questions, which they find relatively easy, are of the form: "Is it X?" The order of comprehension difficulty, as reflected by the children's tendency not to respond at all, is: requests for interrogatives; transitive-based interrogatives ("Does she have a ball?"); followed by copular and intran sitive-based interrogatives, which have almost equal ranking. Turning to the second column, we find that tran sitive-based questions which are the second most likely not to be responded to are least likely to be responded to inappropriately. In other words, the children, although they have difficulty understanding transitive sentences in interrogative form, are most likely to produce appropriate responses to this type if they do respond. The third column, Score I, measures the children's actual productive competence. Interrogatives are the most difficult type for the children by far, and intransitive, non-copular sentences the least difficult, followed closely by copular sentences. Thus, their productive and passive abilities are closely correlated at this level of language.

PAGE 259

245 The correlation between their spontaneous and imita tive abilities is also quite close, as shown in Table 68, below. Table 68. Sentence Types, Spontaneous and Imitative Success Difficulty + Difficult Difficult Easy + Easy Spontaneous Ranking II (96.4) I (281) III (297.1) IV (335.3) Imitative Ranking II (1. 1) III (25.6) I (34.3) IV ( 53. 3) With only minor reversal (note the indices) of Types III and I, the two modes parallel each other; that is, intran sitive, non-copular sentences are most successfully pro duced both spontaneously and imitatively, and interrogatives are least successfully produced. We can therefore conclude that, at the sentence level, there is a close relationship between spontaneous and imitative abilities. 4.2.2. Sentence success, grade level analysis An examination of sentence success by grade level provides a very general picture of the development of each type. Table 69 compares each level's success with the four sentence types on the basis of Score I.

PAGE 260

Table 69. Sentence Success, Grades K 2/3 Difficulty + Difficult Difficult Easy + Easy Grade K II (12.2) IV ( 55. 7) I (58.7) III (70.5) Grade 1 II (30.2) I (126.2) III (137. 6) IV (150) Grade 2/3 II (54) I ( 96) III (97.1) IV (129. 7) At all levels, interrogatives trail significantly behind in development. The rank orders for Grades 1 and 246 2/3 are exactly the same, but there is an important differ ence in that transitive sentences, in Grade 2/3, are catching up with copular sentences. In fact, the difference in scores is not significant. Four different kinds of transitive sentences are found in the SWCEL test. These are examined for relative success by grade level in Table 70. The forms distinguished and labeled are: A) affirmative sentences with one object; B) affirmative sentences with two objects, direct and in direct; C) affirmative sentences with infinitive complement objects; and D) negative sentences with one object. At all grade levels, the children produce one-object sentences and negative sentences most successfully, the negative subtype taking precedence after Grade K. The most difficult sentences are those with two objects and those

PAGE 261

247 with infinitive complement objects, the exact ranking of these two switching from grade level to grade level. Table 70. Transitive Sentences Difficulty + Difficult Difficult Easy + Easy Grade K C ( 12) B ( 21) D (55.5) A (57.1) Grade 1 B ( 49) C (53) A (113.3) D (209.5) Grade 2/3 C ( 44) B ( 57) A (85.9) D (165.5) As for interrogatives, the type with which the chil dren have the most difficulty, there are two forms: A) Yes/No interrogatives; and B) Wh-Interrogatives. Table 71 presents the children's success with these. We find that, until Grade 2/3, Yes/No interrogatives are produced more successfully than Wh-Interrogatives, but in Grade 2/3, the order reverses. Table 71. Interrogatives Difficulty + Difficult + Easy Grade K B (10.6) A (15.8) Grade 1 B (2 8) A (35.3) Grade 2/3 A (51) B (55)

PAGE 262

248 Four types of copular sentences are distinguished in Table 72, below. They are those with: A) locative prepo sitions; B) adjectives; C) nominals; and D) negative plus nominals. Table 72. Copular Sentences Difficulty + Difficult Difficult Easy + Easy Grade K A (39.6) D ( 65) C ( 72) B (140. 7) Grade 1 A (91.1) D (112) C (142. 6) B (241.3) Grade 2/3 A ( 63) C (101.8) D ( 110) B (160) For all grades, copular sentences with adjectives are produced most successfully and those with locative prepo sitions least successfully. In Grade 2/3, the children's success with negative copular sentences increases, displacing their previous success with sentences containing predicate nouns. Non-copular, intransitive sentences fall into two groups: A) Affirmative; and B) Negative. Table 73 demon strates that, with the exception of Grade K, the children produce negative sentences of this type more successfully than they do affirmatives.

PAGE 263

249 Table 73. Intransitive Sentences, Affirmative and Negative Difficulty + Difficult + Easy Grade K B ( 3 9) A ( 6 4) Grade 1 A (14 2) B (166) Grade 2/3 A (113) B (163) This brief summary of the children's production of L2 sentences certainly does not completely reflect the chil dren's competence at this level, but it does suggest that superficially less complex sentence types, e.g., transitive, affirmative sentences, may not, from the learners' perspec tive, be the most easily acquired. The next section ex plores some of the difficulties the children experience in the acquisition of transitive and copular sentences. 4.3. Sentence Errors Errors at the sentence level resist classification even more so than do inflectional errors. In general, the errors in this corpus have to do with the misplacing or omission of lexical items resulting in a confusion of, or failure to express grammatical role relationships. In order to discuss the errors systematically, they are presented in association with the types of sentences in which they usually occur. One error, the omission of subject nouns, occurs in all sentences and therefore will be examined first.

PAGE 264

250 4.3.1. The omission of subjects While the children's responses quite often consist of single words only, the production of sentences and almost sentences is not uncommon. The omission of subject nouns in the latter is not only a relatively frequent error, but a relatively unique one as well. The data below reflect only those instances in which subject inclusion is obli gatory within the context of the child's response itself. Table 74, below, presents a summary of subject omission, and Tables 75 and 76 compare pronoun with noun omission rates. Table 74. Subject Omission, Summary Grade K 1 2/3 # Opportunities 316 691 993 # Omitted 84 140 243 % Omitted 27% 20% 24% As Table 74 indicates, the children provide increas ingly frequent opportunities for the inclusion of subjects as they increase in age. Children in Grade 1 produce only 7.6 contexts per test per child, whereas children in Grades 2/3 produce almost 20 contexts. Thus, the possibility for error increases almost threefold, but the omission rate increases by only 4 percentage points, and actually decreases by 3 points overall between Grades Kand 2/3.

PAGE 265

251 Contrary to the history of many inflections examined in the preceding chapter whose success declines in Grade 1, subject inclusion is most successful at that level. Table 75. Pronoun Subject Omission Grade K 1 2/3 # Opportunities 264 581 854 Table 76. Noun Subject Omission Grade K 1 2/3 # Opportunities 52 110 139 # Omitted 73 125 214 # Omitted 84 15 29 % Omitted 28% 34% 25% % Omitted 21% 14% 20% Comparing Tables 75 and 76, we find that the increased success shown in Grade 1 is accounted for by an increase in the inclusion of noun, rather than pronoun subjects. In general, pronoun subjects are omitted at a greater frequen cy than are noun subjects, but by Grades 2/3 the difference in rates is diminishing.

PAGE 266

On an imitative basis, the children are slightly more successful than they are spontaneously (Table 77). Here we find that their inclusion of noun and pronoun subjects is most successful in Grades 2/3, the rate of omission remaining unchanged in Grades Kand 1. Table 77. Subject Omission, Imitative Rate 252 Grade # Opportunities # Omitted % Omitted K 1 2/3 420 848 798 4.3.1.1. Discussion of errors 351 713 699 16% 16% 12% The following examples briefly illustrate the omission of subjects. Grade K: "Look like chile" (4K67029) Grade 1: 2 "Is draw" (2150089) Grade 2/3: "[i] kitten, has a kitten" (3255197) All subject omission errors are similar to these. 4.3.1.2. Learning strategies Because subject omission is a relatively unusual error among ESL learners, one immediately suspects the children's

PAGE 267

253 Ll as the source. Discussion below suggests, however, that it is a combination of sources and strategies that produces the error. 4.3.1.2.1. The role of the Ll The children's Ll does, in fact, offer some precedent for the omission of pronoun subjects. The analysis in Sec tion 3.3. pointed out that there is no third person singu lar personal subject affix in Choctaw. There are, however, bound personal affi x es in the other persons. Because the majority of sentences in the SWCEL test call for the third person singular subject pronoun, it may well be that the omission of pronoun subjects is a case of pattern transfer from the children's Ll. While transfer from the Ll may well explain the omis sion of pronoun subjects, it does not explain the omission of noun subjects which are as obligatory in Choctaw as they are in English. Moreover, although the children do omit a higher percentage of pronoun subjects than noun subjects, the difference is not so great that we would expect two separate explanations of the data. Therefore, we cannot conclude that the children's Ll is responsible for this L2 error, although it may, to some extent, contribute to it. 4.3.1.3.2. The role of cognitive analysis If we consider casual, English communication, we can observe that man y conversational turns are phrasal, omit

PAGE 268

254 subjects, and are nonetheless grammatical in context. The responses to the following questions illustrate this. "Where are they going?" "To town." "What do they have?" "A bicycle." "What are they doing?" "Working at the hospital." "What do they like to do?" "Play cards." None of the responses is a citation sentence, yet each is not only an acceptable but a common form of ellipsis. The last two are of particular interest to this discussion in that they involve verbal, or superficially verbal phrases as do all of the children's phrasal errors. English is very specific about which constituents of the VP may be dropped in such response sentences: either the entire predicate can be dropped, leaving the NP or PP constituent, as in the first two examples above; or the auxiliary can be dropped, leaving the main predicate and NP or PP consti tuents, as in the second two examples. It is impossible to maintain the auxiliary in elliptical responses. The children's responses frequently violate both of these constraints. It seems likely that the children attempt to imitate subjectless phrasal responses, for which they have some precedent in their own language in the third person singular, but that their hypothesis concerning the maintenance of verb constituents in phrases is as yet too inclusive and

PAGE 269

nonspecific for grammaticality, i.e., "keep something verbal in the phrase." 4.3.1.4. Subject omission: summary and conclusion The omission of subjects, although it occurs at relatively low frequencies in the overall corpus, is a sufficiently unique error among ESL learners to merit consideration. 255 In examining the factors which may contribute to this error in the children's developing English, it was concluded that structural models in both the Ll and the L2 serve as the basis for an hypothesis which produces ungrammatical (as well as grammatical) phrasal responses. 4.3.2. Objects in transitive sentences The children's tendency with objects is rather the opposite of their tendency with subjects; instead of omitting them, they supply too many. Although, again, the error is not characteristic of every child's speech, it is sufficiently widespread to deserve comment. The SWCEL test contains 22 items which elicit transi tive sentence responses. In general, the children supply the objects successfully. Table 78 describes their over all success with both noun and pronoun objects. Although there is a slight decline across grade levels, it is accompanied by a very large increase in the number of attempts to produce objects. Children in Grade K attempt,

PAGE 270

individually, approximately 13 objects per test, whereas children in Grades 2/3 attempt 31 objects per test. Table 78. Objects in Transitive Sentences 256 Grade # Opportunities # Successful % Successful K 1 2/3 868 2090 1550 668 80% 1619 77% 1174 76% Imitatively (Table 80) the children are even more suc cessful, and their ability increases rather than decreases across grade levels. Table 79. Objects in transitive Sentences, Imitative Success Grade K 1 2/3 # Opportunities # Successful % Successful 362 313 86% 807 710 88% 758 677 89% 4.3.2.1. Discussion of errors The major problem with objects is the inclusion of a neutral pronoun in the NP constituent. Examples are:

PAGE 271

Grade K: Grade 1: "I' 11 take it cow" (2K38149) "She gonna do it that picture" (4K50026) "She has it one book" (2162097) "I got it my pig" (1137058) 257 "You have it marbles in your hand?" (1141063) Grade 2/3: "She have it ball" (6228162) "I give it to you the pig" (6339160) "He have it book" (4351166) "He carrying it a box" (5254205) As the children increase in age, the frequency with which this error occurs also increases, as indicated by the pro portions of examples above. 4.3.2.2. Learning strategies The origin of redundant object pronouns in the chil dren's speech is difficult to explain. The following dis cussion is therefore brief and somewhat speculative. 4.3.2.2.1. The role of the Ll Both noun and pronoun objects in Choctaw transitive sentences precede the verb, and the latter are bound pre fixes. Examples of first noun, then pronoun objects are:

PAGE 272

Ohoyohat tapossik cQpatok. 'A woman bought a basket.' Ohoyoh + at tapossik CQpa + tok Woman SCM basket buy RPT Cipisalitok. 'I saw you.' Ci + pisa + li + tok You see I RPT Thus, there is nothing in the structure of Ll transitive sentences that might lead the children to insert neutral pronouns before the noun object. Moreover, there is no neutral pronoun in the language. 4.3.2.2.2. The role of cognitive analysis 258 The most satisfactory explanation of the redundant pronoun is rather speculative; and th~t is that the chil dren have overgeneralized a structure which is new to them, neutral third person singular pronouns, and include this pronoun redundantly and inappropriately. It is possible that the complete lack of precedent in the Ll may contri bute to "over-learning'' and consequently to overgeneraliza tion. Comparative data from other ESL learners and other L2 learners would be helpful in establishing whether or not such strategies and results occur elsewhere.

PAGE 273

4.3.2.3. Objects in transitive sentences: summary and conclusions 259 The inclusion of both pronoun and noun objects in transitive sentences increases slightly as the children age, and indicates, it is suggested, the overgeneralization of a newly learned structure. Neither transfer nor inter ference appear to be relevant. Despite this error, the children are quite successful in supplying transitive sen tence objects, producing them 76% of the time correctly in Grades 2/3. 4.3.3. Copular sentences There are three developmental anomalies which occur in copular sentences. All involve aspects of the VP. The first involves the verb itself; the second adjectival constituents; and the third, prepositional phrase constituents. Neither the adjective nor propositional phrase anomaly is necessarily a structural concomitant of copular sentences. However, it is in this context that they appear in the corpus so they will be discussed here. The major problem involving copular verbs is inclu sion. Number and tense errors occur, but are infrequent. The major problem involving adjectives is their order both in relationship to themselves, when more than one is pre sent, and in relationship to the head noun. The chief prob lem with prepositions appears to be lexical. Each of the problems will be discussed separately below.

PAGE 274

260 4.3.3.1. Copular verbs There are 19 test items on the SWCEL designed to eli cit copular sentence responses. Onl y two of these require non-third person singular present tense forms. Table 80 summarizes the children's omission of all copular verbs in these sentences. Table 80. Copular Verbs, Omission Rate Grade K 1 2/3 # Opportunities 85 172 341 # Omitted 25 51 82 % Omitted 29% 30% 24 % The tendency to omit copular verbs peaks in Grade 1 and decreases rather rapidly in Grades 2/3, indicating that the children are beginning to include these verbs more successfully. Imitatively (Table 81), the children show a definite increase in success across grade levels and are, by Grades 2/3, including copular verbs 90 % of the time. Because there are no errors other than omission, the following discussion deals with possible influence from the children's Ll. First, there are no copular verbs in Choctaw. In order to express the equi v alent of the English predicate nominative copular sentence, Choctaw simply

PAGE 275

Table 81. Copular Verbs, Imitative Omission Rate Grade K 1 2/3 # Opportunities 163 322 326 # Omitted 45 41 34 % Omitted 28% 13% 10% suffixes a subject marker to the topic noun and follows 261 it with a co-referential nominal complement. For example: Hattakmat alikci. 'That man is a doctor.' Hattak + ma + (a)t Man that SCM alikci doctor The Choctaw equivalent of English predicate adjective constructions is formed by suffixing a verb tense morpheme to an adjectival, thereby deriving a verb and a sentence: Allamat acokmah. 'That child is pretty.' Alla + ma + (a)t acokrna + h Child that SCM pretty PT Because there is no precedent in the children's Ll for copular verbs, it is quite possible that the children's omission of these structures is the result of the conflicting structural patterns in their Ll. Evidence from other

PAGE 276

English Ll and L2 studies can be compared but it is some what contradictory. 262 Among English Ll learners (Hatch 1978:38), copulas begin to appear midway among the morphemes reported; whereas, in two studies of English L2 learners, the copula appears quite early in the acquisition of morphemes and, in a third study, the copula's emergence is similar in rank order to that of English Ll learners. In the absence of clear evidence from other research, and in view of the children's general success with copular verbs, we can conclude that, while there may be some conflict from the children's Ll, it does not interfere to a great degree with the children's acquisition of copulas. 4.3.3.2. Adjectives in co p ular sentences The majority of adjective opportunities in the SWCEL test require predicate adjectives or adjectives associated with predicate nouns. Only one item elicits a quantifying adjective for the object noun of a transitive sentence, and it is included in this summary. Table 82 reports the children's success. Initially the children are much more successful than they are at the last level of testing. The explanation for this decline probably lies in the children's increasing volubility; that is, in Grade K, they frequently respond with one word which, in the case of adjectives, is usually correct and grammatical and therefore counted as successful.

PAGE 277

263 As their vocabularies and fluency develop they begin to connect adjectives and nouns, thus producing more oppor tunities for error. Imitatively, their success is above the 90% level at all levels tested (Table 83). Table 82. Adjectives Grade K 1 2/3 # Opportunities 260 526 489 # Successful 245 487 374 % Successful 94% 93% 76% Table 83. Adjectives, Imitative Success Grade K 1 2/3 # Opportunities 176 312 244 # Successful % Successful 162 92% 296 95% 225 92% The majority of errors involve the ordering of adjec tives with respect to each other and to the modified noun. (A previous section has presented another type of error, the apparent classification of adjectives as nouns, evidenced by the prefixing of articles to adjectives.) In English,

PAGE 278

adjectives precede nouns and, in cases where a noun is qualified by more than one item, the order is: 264 Article/Demonstrative+ Quantifier+ Size+ Color+ Noun The two major errors are reversal of color and size ad jectives in concatenated phrases, and reversal of the order of adjectives and nouns. Examples of each are: Grade K: Grade 1: Grade 2/3: "Brown big doggie brown" (4K64026) "A dog big, big dog" (6K65043) "Brown big dog" (4164121) "Big dog red" 4 (316408) "I have fingers five on my hand" (5275153) It is difficult, on the basis of these errors, to determine if the children have formed any hypotheses about English adjectives. In fact, the errors appear very di rectly explicable in terms of the children's Ll. In Choc taw, adjectives follow the noun they modify in the order: Noun+ Color+ Size+ Quantifier+ Demonstrative/Article For example: Ofi losa cito 'a big, black dog' ofi dog losa black cito big

PAGE 279

265 Moreover, as mentioned in the previous section, adjectivals (and also quantifiers) may be verbal: Ofi citomat losah. 'That big dog is black.' Ofi cito + ma + (a)t losa + h Dog big that SCM black PT In the absence of comparable errors from other ESL learners, the most satisfactory explanation for this L2 error is the transfer of the Ll pattern to the L2, a con flict of structures which becomes increasingly obvious as the children, in the older levels, become increasingly verbal and attempt longer L2 utterances. 4.3.3.3. Prepositional phrases in copular sentences Locative expressions comprise a semantic domain which is notoriously eccentric in languages and English is no exception. Some of the peculiarities of English locative prepositions have already been pointed out. Here, the children's success and errors will be summarized and brief ly discussed. The SWCEL test contains 12 opportunities for the chil dren to produce locati v e prepositions, the majority of which occur in copular sentences. Their spontaneous success is presented in Table 84, and their imitative success in Table 85.

PAGE 280

266 Table 84. Locative Prepositions Grade K 1 2/3 # Opportunities # Successful % Successful 247 214 87% 535 347 65% 432 280 65% After an initial success, the children's production of locative prepositions suffers a decline which remains unchanged through Grades 2/3. Imitatively, the children are even less successful. Table 95. Locative Prepositions, Imitative Success Grade K 1 2/3 # Opportunities # Successful % Successful 187 106 57% 391 215 55% 379 241 64% Errors are numerous and varied. One category involves the omission of the preposition. Examples are: Grade K: Grade 1: "A fingers hand" (2K4149; "fingers on my hand 0 ) "They go a that school" (3K65132) "Got kittens his box" (6155110)

PAGE 281

Grade 2/3: "He's going his house" (2160101) "Take it his home" (3354195) "He's looking the cat" (3254186) A second category of errors, and by far the largest group, involves the substitution of one preposition for another, suggesting that the semantic correlates of the lexemes are unclear. Examples are: Grade K: Grade 1: "on the box" "Up my hand" (1K41007; for "in") (1K36005; for "on") "In the hand" (3135079; for "under") "In top of my hand" (3136076; for "on") 267 Grade 2/3: "Side on the cow" (3234193; "beside the cow") "In, on my, this" (223520) The last example illustrates an effort to select the cor rect locative, evidence that the semantic fields are sepa rating and becoming associated with specific lexical items. There is no variability in or vacillation between lexemes before the Grade 2/3 level. Finally, there is a group of errors which is difficult to typify. All have to do with the ordering of phrasal con stituents. Some reverse the normal order of PP's and NP's, for example: Grade K: "Tony in hand a book" (2K53150)

PAGE 282

268 Grade 1: "I see in the box is cat" (1156058) Grade 2/3: "I have in my hand a pig" (3331183) The last of these is perfectly grammatical but usually found only in formal speech or written English. A second type of error involving constituent order results in a confusion of locative relationships. For example: Grade K: "Box in the kitten" (3K55132) Grade 1: None Grade 2/3: "I have a hand in my pig" (5231205) These errors are relatively common in Grades Kand 2/3. It would appear that the children realize there is a structural necessity for a preposition but have not yet organized the relationship between the nouns involved. It is difficult to determine how the children approach learning locative prepositions. It is doubtful they have adopted any general hypotheses, in view of the variety of errors that occur. Moreover, the influence of their Ll, if present, is impossible to establish. Choctaw has several comple x systems for indicating position, location, direction, and orientation, and each of these systems may interact with the others. Although these systems have been at least partially explored (Jacob et al. 1977:137-140; Nicklas 1973: 206-210), their explication is by no means complete.

PAGE 283

269 In sum, the locative prepositions pose problems for the children, but it is difficult to determine to what extent these problems are due to the peculiarities of the L2 locatives themselves and/or to the transfer of complex ities from the Ll locative systems. Inasmuch as errors similar to those discussed above have not been reported for other children learning ESL, transfer and structural conflict are the most probable explanations for the developmental difficulties. 4.4. Notes 1. In the first two columns, the numbers in parentheses are weight indices. Because of the different number of occur rences of each sentence type, it was necessary to weight re sponses to equalize disparity and rank types. The weighting was derived by totaling the number of occurrences of the response under investigation for each type and dividing the sum by the number of occurrences of the sentence type per test. 2. It cannot be argued here that the children produced "is" for "it's". Only those cases where the sibilant was clearly voiced are included in the analysis. 3. For an alternative analysis, see Jacob et al. 1977:18. 4. A clear case for imitation and fossilization, in the language domain, and prejudice, in the social domain, is the response of one child, during the previous year's testing: "That's a big, redneck dog.

PAGE 284

5.1. Introduction CHAPTER FIVE BEYOND DATA It is all too easy, in a study of this nature, to lose sight of the "forest'' the research set out to describe, that is, the children's L2 development. In order to pro vide perspective, this chapter will first evaluate speci fic aspects of the research methodology used to elicit and describe that language; and, secondly, briefly review significant factors in the children's language development and speculate on their implications for the L2 classroom. 5.2. Evaluation of Research Methodology Almost every aspect of language research has its frus trating moments, but two components of this study contri buted more than normal quotas of aggravation; they were the test used to elicit the data and the computer used to analyze them. Nonetheless, with the suggestions and limi tations presented below, the writer would maintain both in future, similar research. Given the fact that structured tests are probably the least effective means of realistically assessing real 270

PAGE 285

271 language, they nonetheless have specific advantages, parti cularly for research with large populations. The major advantage is in the comparability of the data base created. Thus, the language behavior of a group of speakers is ac cessible; and language, after all, is defined by its speakers' cumulative idiolects. In this study, for example, the use of a structured test has allowed us to state, with reasonable assurances of accuracy, that the children's language includes Feature V, but not Feature Y, and, more over, we can calculate, on the basis of success rates, the degree of inclusion and exclusion at various develop mental points. A derivative advantage of the comparability of a large data base is the credibility of the conclusions. Because data are uniformly gathered, the results may be considered predictive for the population. Finally, in cases such as this study, where the para meters of the field of observation are completely unknown, a structured test allows description and delineation of some, relatively static, boundaries. These not only pro vide baselines but suggest directions for future study. Nonetheless, it would be erroneous to conclude, on the basis of data from a structured evaluation instrument, that we have a comprehensive understanding of the children's L2 proficiency. What is obviously needed, in addition, is observation and analysis of the children's L2 in non structured communicative settings. Such data would vastly

PAGE 286

272 inform the interpretation of language data presented in this study. For example, the writer suspects that inter rogatives are difficult for the children for more than linguistic reasons. Questioning may simply not be as common in Choctaw verbal communication as it is in English. Obser vation of children talking in both languages in unsupervised groups in the classroom would quickly confirm or reject this suspicion. The value of the analysis of natural lan guage use in L2 research cannot be exaggerated, even though the data are less easily recorded and organized than are those from structured interviews. In sum, initial research is greatly facilitated by the use of a structured approach which permits comparison of a great many specific, isolable language features. Ideally, and on the basis of the limitations of and information from this approach, a study of casual L2 speech would follow. The resulting profile of the learners' L2 would be far more accurate and complete than either approach, by itself, offers. Turning to the specific instrument used in this study, certain features, both linguistic and nonlinguistic, deserve comment. There is no perfect test, but tests can certainly be modified to more nearly approximate that goal, and the SWCEL is no exception. The worst defect, and one for which there is probably no remediation other than observation of casual speech, is that larger syntactic units are not generated. Because of the question-answer format of the test, few spontaneous

PAGE 287

273 responses are more complex than phrases, and many responses are single words. While both are usually appropriate in the testing context, the children's proficiency with larger syntactic units is probably underestimated as a result. Secondly, one of the presumed advantages of a struc tured test is that it can elicit items not likely to be observed in casual conversation; however, unless the eli= citations are shrewdly constructed, this theoretical edge vanishes in practice. The SWCEL contains several clumsy attempts to elicit uncommon or formal speech, e.g., "Yes, I do" (Question 47) instead of the more frequent "Yes" or "Uh huh". Moreover, some forms expected as responses are simply ungrammatical in spoken, non-written language, e.g., the non-contracted auxiliary "will" as the future tense instead of "going to" or "gonna". These awkward items seem to result from a confusion on the part of the test construc tors between prescriptive and descriptive models of language use. A further problem is created by these unnatural or ungrammatical items. When the children supply the normal response, they are asked to imitate the unnatural response, causing them to suspect, erroneously, that there was some thing wrong with their original, spontaneous answer. For some, there results a cumulative sense of confusion and con sequent loss of confidence, so that their responses become

PAGE 288

increasingly minimal as the test proceeds and requests for imitation are met with silence. 274 The solution is to remove, modify, or replace such items. Any factor which focuses the children's attention on their own speech, especially in a negative way, detracts from the credibility of test conclusions. A similar but nonlinguistic problem with the SWCEL is that some of the "games" are not very play-like. This is particularly a problem for beginning learners who have to decide whether the tester is a little crazy or their new L2 has failed to make sense of the situation. The older children, whose language skills are more advanced and who are accustomed to the absurdities of tests, comply with the occasional bizarre request with remarkable patience. A case in point is Question 74, in which children must respond to the inane sequence: "I have fingers on my hand. What do you have on yours?" However, the majority of games approach real play, so that the few awkwardly artificial sequences do not disrupt the flow of the test or distract the children. The requests for unnatural speech, mentioned earlier, are a much more serious defect in the test. A final criticism, and one as easily corrected as the awkward game situations, is of the pictures used as stimuli. There are six pictures featuring children; four of these are of a boy and only two of a girl; she, moreover, is always in his company. Whether or not this produces

PAGE 289

275 differential responses from the two sexes because of dif ferential interest and/or identification is not clear; but in the interest of removing possible prejudice from the instrument and especially in the interest of removing sexist models, girls should receive equal pictorial exposure. Ideally, a test used in program evaluation should assess both overall L2 proficiency and the effectiveness of the language curriculum and pedagogy, but, because the SWCEL is not linked in any systematic fashion with the curriculum sequence, it cannot meet the latter desideratum. However, in view of the difficulties of developing such an instrument, and within the limitations discussed above, this study concludes that the SWCEL satisfactorily fulfills its primary purpose, that of assessing general L2 profi ciency and is, moreover, as this study demonstrates, suitable for research purposes. Turning now to the second procedure to be discussed, computer analysis, it should be noted at the outset that this research could not have been accomplished within any reasonable length of time without the use of the computer. All language analysis is labor-intensive, particularly de scriptive research, and a study of this size, conducted by a single researcher, simply exceeds the capacities of the traditional file catalogue procedure, or even of punch card file systems. Moreover, given that there is a large quantity of data to be organized, the inevitability of human error and possible,

PAGE 290

276 if unconscious, prejudice of the researcher's expectations, the computer's manipulation of figures can be accepted with a great deal more confidence than can an investi gator's. Despite its advantages, computer analysis presents certain problems. In this study, for example, the process of developing a coding system for the language data, in conjunction with programs to extract the necessary varieties of information from the codes, became so lengthy and dif ficult that there were periods when the writer doubted the efficacy, in terms of time, of computer analysis. For the linguist, the major difficulty in effectively utilizing a computer is programming. Without sophisticated and creative programming, a computer is no more useful than a card file. Unfortunately, few programmers are prepared to work with language data, and even fewer linguists are prepared to develop complex programs. It is to be hoped that as the interest in and need for language research with large popu lations grow, so will course options in computer analysis for graduate students. Finally, there is no question but that the final print-out for this study, containing every conceivable cross-correlation, summative counts, individual breakdowns, comparisons, and significance levels of all features, justified the effort that produced it. Because the coding system is, to a certain extent, language general and the

PAGE 291

277 programs flexible, the writer intends to use both in other descriptive research. 5.3. Some Implications of the Children's L2 Development One of the more unexpected findings of this study is the variability of success rates found for individual lan guage features over a four-year time period; no delineable, incremental sequence of acquisition is evidenced. Thus, the ESL development of these children is very different from that reported for other ESL learners, as well as for English Ll learners. The writer believes this variability is associated with two correlated factors in the children's environment: the first is the English spoken in the community; and the second is the community's goals and needs for speaking that language. By any measure, English in the Choctaw community is a subordinate language; it is not only spoken a great deal less than Choctaw, but it is, with rare and individual ex ceptions, spoken a great deal less proficiently. While this study has followed English development only through the first four years of school, observation and casual testing of adult high school graduates suggests that development after the fourth year of schooling is relatively limited. Specifically, and for example, a full range of noun feature specifications (e.g., plural, mass, definite) is not present in the speech of most young Choctaw adults; nor are relative

PAGE 292

278 clauses, complements, or other extended syntactic struc tures common in L2 conversation. Because of a generally poor educational system (among other factors), there is no development of a more extensive passive competence through L2 literacy. Thus, speakers' active L2 vocabularies are almost coterminous with their recognition vocabularies, and their comprehension of written sentences parallels sentences they generate orally. Perhaps the best way to characterize the L2 situation is in terms of language styles, and from this perspective, Choctaw speakers control only one style of the L2 and that style, while functionally equivalent to its counterpart in local Ll English, is marked by various minor departures (phonological, morphological, and syntactic) from the local style. From the point of view of the second factor mentioned earlier, the speakers' L2 needs and goals, the lack of stylistic diversity is easily understood. The relationship between L2 development and communicative needs cannot be minimized. Need, however defined, is probably the primary variable in L2 development, and there is very little need, from the Choctaw perspective, for English. It is therefore interesting that learning English is articulated as a goal b y all age groups. However, there is little consensus among age groups concerning why the lan guage should be learned. Both older and middle-aged people regard English as a means of ameliorating their economic

PAGE 293

279 circumstances, but, while older people still believe that the relationship between English and employment opportuni ties is an inevitable one, middle-aged people are less naive since they are the second generation of English speakers to have not been benefited in any appreciable eco nomic terms by learning the language. In fact, employment patterns have changed between the working years of the two generations. Older people had to leave the reservation in order to work; today, with the advent of federal programs and tribal efforts to locate light industries on reservation lands, this is no longer the case. The majority of middle aged persons are employed on the reservation itself and are, unlike their parents, economically independent of the sur rounding English-speaking world. Young adults have grown up with the expectation of on-reservation employment and it is this group, paradoxi cally the most proficient L2 speakers, who have the least positive attitude toward learning and speaking English. By and large, this group regards the language as the prover bial necessary evil. Thus, there is no consensus regarding L2 acquisition, and very little is accomplished in the Choctaw community without consensus. Despite the fact that learning English is considered, in vague terms, as necessary and desirable, social and economic isolation and independence preclude the necessity for proficiency. Perhaps consequently, there has been little concern over its teaching, a task considered

PAGE 294

280 to be solely the responsibility of the school system. Children somehow learn enough to survive this system, and later, the Anglo world with which they are in contact. In effect, the Choctaw pay the same attention to learning and teaching English as they do to English speakers; that is, the minimum necessary. What has happened in the classroom, then, has largely been ad hoc adaptations by monolingual Anglo teachers of traditional English Language Arts curricula for monolingual Choctaw-speaking children. Whether or not there would have eventually been more profound changes in the class room and in L2 acquisition as a result of a bilingual and ESL teaching program is now a moot point, since both the bilingual program in the reservation schools and the bi lingual teacher degree program at nearby Mississippi State University have been abandoned. Although it is therefore impossible to evaluate either the "ad hoc" or the ESL approach to English teaching, it is important to note that, if the efficacy of a pedagogical program is to be judged by whether or not it meets the learners' needs, then the various programs of English in struction on the reservation have been adequate; 1 Choctaw people do speak English when it is necessary. In the event that contexts for L2 use broaden in the future, as they almost certainly will, this research offers some information that may be helpful to curriculum develop ment. First, and most importantly, this study presents very

PAGE 295

281 specific information concerning the children's L2 strengths and weaknesses at each grade level. These data can be translated by teachers to supplement, or even supplant, existing curriculum content in order to take advantage of students' abilities and assist them with their weaknesses. Reference to grade level success rates would be parti cularly helpful in mitigating two pervasive problems in L2 classrooms. The first is the tendency of teachers to focus on those L2 errors which, for one reason or another, strike their ears as very un-English, and to attribute the aber rant features to an entire classroom when, in fact, only a small percentage of voluble children may exhibit the traits. Success rates would assist teachers in making realistic assessments of the prevalence of specific deviant features. A second, rather inevitable tendency in L2 classrooms is the unconscious adjustment, on the part of teachers, to certain non-English features in the children's speech. The result is that some aberrant patterns remain fixed, grade after grade, in the children's L2. These patterns even tually become the norm for what is sometimes called "reser vation" or "Indian" English. Again, success rate compari sons across grade levels would pinpoint those features not affected by ESL instruction so that compensations in curricula could be made. Another pedagogically applicable finding of this study is that four years of English instruction do not result in L2 proficiency; none of the inflections examined is

PAGE 296

282 produced at the acquisition criterion level by children in Grades 2/3. There are two alternative solutions for this problem: either intensify and expand instruction during the first four years, or extend the duration of ESL in struction.2 Based on the fact that there is little or no non-classroom opportunity to use the L2, and on the assump tion that there are upper neurological limits for pro cessing "useless" information, it is doubtful that inter nally expanding the four-year curriculum would significantly accelerate L2 development. Therefore, extending ESL instruc tion to upper grade levels appears to be the more effective approach. One very appealing possibility of such an ex tension, for example, is the presentation of a diversity of L2 styles, both written and oral. In the past, high school English has been largely taught through workbook exercises; little or no attention has been given to dis course, debate, rhetoric, or literature. For those students who will go to college and for those who will eventually negotiate with the Anglo world in the interests of the tribe, the benefits of an extended program are obvious. Whether the L2 curriculum is extended or intensified is probably less crucial a question than the curriculum itself, but the findings of this research do not appear to offer straightforward implications for curriculum design. Most ESL curricula are sequenced in a presumed hierarchy of difficulty (e.g., Paulston and Bruder 1976:47-54), the ordering independent of the learners' Ll. Had this research

PAGE 297

283 discovered a regular developmental sequence among the learners, then a curriculum to facilitate emergence of features might have been proposed. However, because of the diversity of language behavior found among these learners and the absence of a ''natural" acquisition sequence, such a proposal is impossible. This study can only suggest, on the basis of success rates, that specific structures not being acquired be emphasized in future curricula. A complicating factor in proposing curriculum recom mendations is that it is difficult, from these data, to determine if structures are successful because they are being taught and/or because they are high frequency features in the children's L2 environment. Here, again, observational data, if available, could contribute to the resolution of this question. A final point to be made concerning curricula is that any language lesson contains a great deal of cultural as well as linguistic information, i.e., the appropriateness of topic and participants, the roles of participants, the ordering and timing of dialogue and conversation, the use of humor, politeness, redundance, implication, and a myriad other more or less covert clues to the organization of the L2 culture. From the learner's perspective, these aspects of the curriculum can be at least as comple x and confusing as the language itself. As pointed out in Chapters One and Two, many Choctaw speakers of English e x perience discomfort when using the

PAGE 298

284 language, and it is largely due to the fact that, although they are more or less proficient in the L2, they are not pro ficient in the L2 culture communicative style. Therefore, this study strongly recommends that curricula be based on contrastive analyses of communicative styles so that where cultural differences, and particularly cultural dissonances, occur, they may be explicated. Learners can no more be ex pected to "absorb" disparate cultural patterns than they can covert grammatical classes; in other words, Anglo eye con tact, for example, should be taught as consciously and 3 carefully as English mass nouns. The recognition and understanding of difficult, per haps repugnant communicative style features of the L2 will relieve the students of a learning task they are now expected to blunder through on their own. As a result, the children may come to appreciate cultural differences rather than experience a discomfort that hinders not only their own L2 development but, ultimately, their relationships with the speakers of that language. 5.4. Notes 1. However, they have not been sufficient for those stu dents attempting college. The sheer quantity of language they must process, in conjunction with unfamiliar styles, is so overwhelming that relatively few survive to graduate. Language is, of course, not the only factor, but it is often cited as a major reason for difficulty with course work.

PAGE 299

285 2. Solutions for a school program that does not include ESL instruction are obscure; in fact, such a program is so inappropriate that it is excluded from consideration. 3. It is by no means being suggested here that the L2 teaching style be changed to match the students' learning style. One of the more undesirable consequences of this is presented in Phillips' discussion of Warm Springs Indian children (1972:383).

PAGE 300

APPENDIX ONE THE SWCEL TEST --The SWCEL test consists of two sections, pronunciation and grammar. Only those items included in this research, that is, the grammar items, are presented below. The item numbers are as in the original test and as referred to in the text of this study. Interviewer instructions are in cluded where necessary for the reader's understanding. The props referred to are: a toy pig, cow, and sheep; a small lidded box containing one blue and one white marble; eight pictures of: 1) a parental couple with a son pulling a wagon and a daughter bouncing a ball; 2) a boy ("Tony") writing at a table while looking at a book he holds; 3) Tony carrying a large, covered box; 4) a box with three cats peering over the rim; 5) Tony dressed as a policeman and directing traffic; 6) Tony dressed as a fireman holding a hose; 7) Tony and a girl ("Mary") followed by a large brown dog and a small yellow dog walking toward a building labeled "school"; 8) Tony and Mary eating bowls of dif ferent substances. 27. DOES HE HAVE A BALL? (Hold up family picture. Point to boy.) 286

PAGE 301

28. WHAT DOES SHE HAVE? 29. HE IS PULLING HIS WAGON? WHAT IS SHE DOING? 30. WHICH ONE DO YOU WANT? (Place animals on table.) 31. WHAT DO YOU HAVE IN YOUR HAND? 32. WHERE IS THE SHEEP? (Line up animals, one in back of the other in the order: pig, cow, sheep.) 33. WHERE IS THE PIG? 34. NOW WHERE IS THE PIG? (Move sheep to one side. Place pig and cow side by side.) 35. WHERE IS THE PIG? (Place child's hand over the pig.) 36. WHERE IS THE PIG? (Place pig on top of child's hand.) 37. I'LL TAKE THIS ONE: WHICH ONE WILL YOU TAKE? (Select one of the animals.) 38. DO YOU WANT THIS ONE OR THAT ONE? 39. GIVE ME THE SHEEP. WHAT DID YOU DO? 287

PAGE 302

288 40. I HAVE SOMETHING IN THIS BOX. DO YOU KNOW WHAT IT IS? DO YOU WANT TO GUESS? ASK ME WHAT I HAVE IN THIS BOX. (Put animals away. Take out box containing marbles and shake it.) 41. ASK ME IF I HAVE A MARBLE IN MY HAND? (Sleight of hand shuffle.) 42. HERE ARE THE MARBLES NOW. WHERE WERE THE MARBLES? (Show child marbles; point to empty box.) 43. IS THIS A WHITE MARBLE? (Give child blue marble; hold up white one.) 44. WHAT COLOR IS YOUR MARBLE? 45. ASK ME IF I WANT THIS ONE OR THAT ONE. (Give child both marbles, placing one in each hand.) 46. I'LL TAKE THIS ONE. WHAT DID I DO? 47. DO YOU WANT TO LOOK IN THIS BOOK? (Hold up book containing stimulus pictures.) 48. HERE'S A PICTURE OF A BOY. ASK ME WHO HE IS? 49. DOES HE HAVE A PENCIL? 50. WHAT WILL TONY DO WITH HIS PENCIL? 51. LOOK AT TONY'S OTHER HAND. HE HAS SOMETHING IN HIS OTHER HAND. PLAY LIKE YOU CAN'T SEE WHAT IT IS. NOW

PAGE 303

ASK ME WHAT TONY HAS IN HIS OTHER HAND. 52. HERE'S TONY AGAIN. HE'S GOT SOMETHING ELSE IN HIS HANDS. WHAT DOES HE HAVE? 53. IS THE BOX BIG OR LITTLE? 54. WHAT IS TONY DOING? 55. ASK ME IF HE HAS KITTENS IN HIS BOX. 56. HERE IS TONY'S BOX. WHAT DO YOU SEE IN IT? 289 57. LOOK AT ALL THOSE KITTENS. THAT MUST BE A HEAVY BOX. WHO CAN CARRY IT? 58. HERE IS TONY AGAIN AND HE IS PLAYING THAT HE IS A POLICEMAN. NOW TONY IS GOING TO CHANGE HIS CLOTHES AND BE SOMEBODY ELSE. CLOSE YOUR EYES. CLOSE YOUR EYES. WHAT DO YOU THINK HE IS NOW? ASK ME WHAT HE IS. 59. OPEN YOUR EYES NOW. WHAT IS HE? (Picture of fireman.) 60. TONY IS FINISHED PRETENDING HE'S A FIREMAN. NOW HE IS GOING SOMEWHERE ELSE. I WONDER WHERE HE IS. I'M GOING TO PEEK. NOW I KNOW. ASK ME WHERE HE IS. 61. OH, LOOK. HE'S AT '.1 SCHOOL WITH A FRIEND. THEY HAVE BOOKS. THIS IS HIS BOOK AND THAT'S 62. HOW MANY BOOKS DOES SHE HAVE?

PAGE 304

63. THE DOGS HAVE FOLLOWED THEM TO SCHOOL. IS HE A BIG DOG? (Point to little dog.) 64. HE'S A LITTLE YELLOW DOG. WHAT IS HE? (Point to large dog.) 65. WHERE ARE THEY? 66. HERE ARE TONY AND MARY IN THE CAFETERIA. THIS IS MARY'S LUNCH. WHOSE LUNCH IS THIS? 290 67. TONY HAS A BOWL OF ICE CREAM. IT'S COLD. MARY HAS SOMETHING TOO. IT DOESN'T LOOK LIKE ICE CREAM. LET'S FIND OUT WHAT IT IS. ASK ME IF IT'S HOT OR COLD. 68. I THINK IT'S SOUP. DOES TONY HAVE ANY SOUP. 69. THIS IS A BOWL OF ICE CREAM. WHAT'S THIS? (Point to bowl of soup.) 70. TONY LIKES ICE CREAM. WHAT DO YOU LIKE TO EAT? 71. ASK ME WHAT I LIKE. 72. WHO IS ? (Supply name of child being tested.) 73. LET'S PLAY A NEW GAME. I'M CLAPPING MY HANDS. YOU CLAP YOUR HANDS TOO. WHAT ARE WE DOING? 74. I HAVE FINGERS ON MY HAND. WHAT DO YOU HAVE ON YOUR HAND?

PAGE 305

291 75. LET'S PLAY ANOTHER GAME. PUT YOUR HANDS BEHIND YOUR BACK AND I'LL GUESS HOW MANY FINGERS YOU HAVE. BUT YOU HAVE TO ASK ME, OK? ASK .ME HOW MANY FINGERS YOU HAVE ON YOUR HAND. 76. ASK ME WHAT I'M DOING. (Tap knee with one finger.) 77. CAN YOU TAP YOUR FINGER? 78. WILL YOU TAP TWO FINGERS? 79. THIS IS MY HAND. WHOSE HAND IS THIS? (Point to child's hand.) 80. WHOSE NOSE IS THIS? (Point to own nose.) 81. I'M GOING TO STAND SOMEWHERE. CLOSE YOUR EYES. OK, ASK ME WHERE I'M STANDING. (Stand behind child.) 82. AM I RUNNING? (Sit down.) 83. NOW YOU STAND UP. I'M SITTING. WHAT ARE YOU DOING?

PAGE 306

APPENDIX TWO COMPUTER RESPONSE FORM The following sample illustrates how the data were re corded for later analyses. These forms were filled in as the writer listened to the children's tape recorded test responses. Once the forms for each question had been de veloped, it was actually possible to record responses al most as quickly as the children talked. The data were stored in the computer exactly as they appear here, so there were no intermediate steps, other than programming, before analyses. Although modifications to eliminate cer tain redundancies should be made for future use, these forms proved to be extremely time-saving; moreover, they are simple enough to be understood by the very limited capacities of computer intelligence. In order to "read" the sample, the following informa tion is necessary. The first seven columns identify the school and community, grade level, test item, and student. All the responses on the sample are from School 2, Grade 1, Question 50. Columns 11 16 provide general character istics of the child's response. If the child did not re spond, either spontaneously or imitatively, Column 11 was marked "l". If a spontaneous response was grammatical 292

PAGE 307

293 (whether or not it was appropriate in context), Column 12 was marked "l"; if ungrammatical, "2". If the child's response was appropriate (whether or not it was grammatical), Column 13 was marked "l"; if inappropriate, "2". If the child responded with only a single, appropriate, grammatical word, Column 14 was marked "l"; if with a single, inappropriate and/or ungrammatical word, "2". If the child's imitative response was grammatical and was, in fact, imitative, Column 15 was marked "l"; if non-imitative and/or ungrammatical, "2". If the child's spontaneous response was a complete, grammatical sentence, column 16 was marked "l"; if the imitative response was a complete, grammatical sentence, "2". Columns 17 32 contain the expected response to Question 50 ("What will Tony do with his pencil?"). For each lexical item in the response (and, for some questions, for individual morphemes), there are two columns; the first is for spontaneous responses ("x" means the item was produced as shown in the column heading; "O" means the item was incorrectly used, see "Additional Information" for form given by child); the second is for imitative responses ("3" means the item was used; "4" means the item was incorrectly used). Columns 33 and 34 indicate a frequent error for this particular question, the omission of the auxiliary. Here an "x" or a "3" simply mean the error occurred. From Column 35 on, other information and explanation

PAGE 308

294 is recorded. If the information is spontaneous and orthographic, no symbol precedes it; if spontaneous and coded (see Appendix C), a"+" precedes it; if the information is imitative, a 11 11 precedes it. Thus, for example, in the second response on the sample, the child's imitative re sponse could not be completely recorded in Columns 17 32, so it appears in the "Additional Information" section. (Redundance with Columns 17 32 was eliminated through programming.) Errors in the child's imitative response were recorded in code; "7A" indicates the absence of an obligatory preposition (in this case, "on"), and "13" indicates an article error. The grammatical aspects of the child's responses were automatically picked up by the computer from Columns 17 32. Although the response form is intended to be "read" by the computer, the above information should allow the reader to interpret the ten responses illustrated.

PAGE 309

5 1 , ~ ll 1t s .. o, v n ~ -1 2 -0 \u Ii C.,J,, l . {) D i z a: ( l ~d 1. .J L>r --__ c) e,/ 0,.11 <: oJl t..5 H --.,;"\ ,__.._...._ I l J 5 7 I I 215O0Cff i.,ts-ooqq ~1S'D10O 1 l1S01D1 7.150107. ; t 1~D10 8 l.15"0 /()4 Z15oto5' !I 11500 :/-b 111 A15"Do11 11 ~Iii~ l 2,.;_.~i~ ? C: ... ..:i 2 Ot~f;;lll ::! '!? i :2.~{i~ :J: -.,J,:) c.) .c r, r:: w :r. t I4' ~c.:-1--~....1 :S 1-i I Al>l>\ll OU l'II-U,f-01-:111\Tl O ~ t 1 u J11 ll I J u I) 10 111 11 1 1' l0 ) 1 1 n l ll J -. I J) i. 1 11 J bJ H J v! J 1 H i ll Jt l n 11 l ; Jli n , 1 H ~ 1 <4J H > h, ,. ,a )u ~ .,, ~' ,. . ) ... ... 1 z 1 2 l'j.g t1 "l1111:<. 3 Z1X Z 1 ID l 1 1 /4 1 z 1. ZJ.. 31 3 3 x' 3 X X X 0 '~ 3X3 0 )( ')( x x X "x X -Hf's qOIJ w/(IT )I 'ff',,PP. -1A~ 1J wRITINf;i t--1AD Wll11E-~ t~AD ll.IIW )I fJJ e.ruRE t1A 14ft,B Iq. llE ~ oTS A ~DOH t4ADfJ4[+14fb8 LITTl tU TfEAJS f7../pfbC SAMPLE COMPUTER RESPONSE FORM N I.D lJl

PAGE 310

Category VERB Present Tense Past Tense Future Tense APPENDIX THREE DATA CODING SYSTEM Code 1 Incorrect 2 Correct A B C Present Progressive D Third Person Singular E Auxiliary F Copula G Omission H Contraction I PRONOUN 3 Incorrect 4 Correct Subject A Object B Possessive C Gender D Omission E Demonstrative F Number G NOUN 5 Incorrect 6 Correct Subject A Object B Plural C Possessive D Omission E Mass F PREPOSITION 7 Incorrect 8 Correct Omission A 296

PAGE 311

297 NEGATIVE 9 Incorrect 10 Correct QUANTIFIER 11 Incorrect 12 Correct ARTICLE 13 Incorrect 14 Correct INTERROGATIVE 15 Incorrect 16 Correct OTHER 17 Incorrect 18 Correct LOCATIVE 19 Incorrect 20 Correct CONJUNCTION 22 Incorrect 23 Correct WORD ORDER 24 ADJECTIVE 25 Incorrect 26 Correct

PAGE 312

APPENDIX FOUR CHOCTAW ORTHOGRAPHY There is no satisfactory published account of Choctaw phonology, so that the phonemes posited below are specula tive and based on the writer's analysis. This analysis is sufficient for practical purposes, but a more abstract analysis would be more appealing from a theoretical point of view. As in many minority language communities, orthography is a sensitive issue and the so-called "school'' orthography presented below has not been unanimously well received. Its practical merit (one symbol for each phoneme, thus facilitating literacy instruction) is somewhat offset by the fact that it is different from a ''missionary" ortho graphy which, while lacking phonemic accuracy, has a vener able tradition and is used proficiently by most older people. These object to the non-standard characters in the school orthography and to the implied inadequacy of the system in which sacred texts are written. By emotional extension, the new orthography is regarded by some as bordering on blas phemy. The phonemes are represented by IPA symbols (Lehmann 1972) in the left-hand column below and the school 298

PAGE 313

orthography equivalents are listed in the right-hand column. I. Vowels. There are nine vowel phonemes. Phoneme Letter / a ,/ a /o/ 0 /i/ i I a_ :/ a /o:/ 0 /i:/ 1 / a, / /6/ Q /i/ :j, II. Consonants. There are thirty consonant phonemes, fifteen short consonants each having a long counterpart. /p/ /p:/ p pp /b/ /b:/ b bb /t/ /t:/ t tt /k/ /k:/ k kk / / / : / f ff /1/ /1:/* l ll /s/ /s :/ s ss /s/ /s:/ s ss /h/ /h:/ h hh *An alveolar, voiceless, lateral fricative. 299

PAGE 314

300 le/ /c:/ ... cc C /1/ /1 :/ 1 11 /m/ /m:/ m mm /n/ /n:/ n nn /w/ /w:/ w WW /y/ /y:/ y yy

PAGE 315

BIBLIOGRAPHY Bauer, Evelyn 1971. A History of Bilingual Education in B.I.A. Schools. In Bilingual Education for American Indians. Indian Education, Curriculum Bulletin No. 3, pp. 29-32. Wash ington, D. C.: u. S. Bureau of Indian Affairs. Beilin, H., ed. 1975. Studies in the Cognitive Basis of Language Development. New York: Academic Press. Bereiter, c. and S. Engelmann 1966. Teaching Disadvantaged Children in the Pre school. Englewood Cliffs, N.J.: Prentice-Hall. Bernstein, Basil 1964. Elaborated and restricted codes: Their social origins and some consequences. In J. Gumperz and D. Hymes, eds., The Ethnography of Communication. Ameri can Anthropologist, Special Publication, 66:6, Part 2, pp. 55-69. Bloom, Lois, L. Hood and P. Lightbrown 1974. Imitation in language development: If, when and why. Cognitive Psychology, 6:380-420. Braine, Martin 1971. On two types of models of the internalization of grammars. In D. Slobin, ed., The Ontogenesis of Grammar, A Theoretical Symposium, pp. 155-186. New York: Academic Press. Bricker, W. A. and Diane D. Bricker 1974. An early language training strategy. In Richard L. Schiefelbusch and Lyle L. Lloyd, eds., Language Perspectives, Acquisition, Retardation, and Intervention, pp. 431-468. Baltimore: University Park Press. Brinton, D. G. 1870. Contributions to a grammar of the Muskokee lan guage. American Philosophical Society, 11, l0E: 301-309. 301

PAGE 316

302 Brown, Roger 1970. Psycholinguistics, Selected Papers. New York: The Free Press. Brown, Roger 1973. A First Language, The Early Stages. Cambridge: Harvard University Press. Brown, Roger and Ursula Bellugi 1970. Three processes in the child's acquisition of syntax. In R. Brown, ed., Psycholinguistics, Selected Papers, pp. 75-99. New York: The Free Press. Brown, Roger, C. Cazden, and U. Bellugi 1970. The child's grammar I to III. In R. Brown, ed., Psycholinguistics, Selected Papers, pp. 100-154. New York: The Free Press. Brown, Roger and C. Hanlon 1970. Derivational complexity and order of acquisi tion in child speech. In R. Brown, ed., Psycholin guistics, Selected Papers, pp. 155-207. New York: The Free Press. Burt, Marina, H. Dulay, and E. Hernandez Ch. 1973. Bilingual Syntax Measure (Restricted Edition). New York: Harcourt Brace Jovanovich. Burt, Marina and Carol Kiparsky 1972. The Gooficon, A Repair Manual for English. Rowley, Mass.: Newbury House Publishers. Byington, Cyrus 1915. A Dictionary of the Choctaw Language. John R. Swanton and Henry S. Halbert, eds., Smithsonian Insti tution, Bureau of American Ethnology, Bulletin 46. Washington, D. C.: U. S. Government Printing Office. Carterette, Edward C. and Morton P. Friedman, eds. 1976. Handbook of Perception, Vol. VII. Language and Speech. New York: Academic Press. Cazden, Courtney B. 1972. Child Language and Education. New York: Holt, Rinehart and Winston. Cazden, Courtney and Vera P. John 1968. Learning in American Indian children. In St y les of Learning among Indians, An Outline for Research. Report and Recommendations of a conference held at Stanford University, August, 1968. Washington, D.C.: Center for Applied Linguistics.

PAGE 317

303 Chomsky, Carol 1969. The Acquisition of Syntax in Children from 5 to 10. Cambridge, Mass.: The M.I.T. Press. Chomsky, Noam 1964. Current issues in linguistic theory. In J. Fodor and J. Katz, eds., The Structure of Language, Readings in the Philosophy of Language, pp. 50-118. Englewood Cliffs, N.J.: Prentice-Hall. Chomsky, Noam 1965. Aspects of the Theory of Syntax. Cambridge, Mass.: The M.I.T. Press. Corder, S. P. 1978. Language-learner language. In Jack Richards, ed., Understanding Second and Foreign Language Learning, Issues and Approaches, pp. 71-93. Rowley, Mass.: Newbury House Publishers. Cornejo, Ricardo J. 1974. A Synthesis of Theories and Research on the Effects of Teaching in First and Second Languages: Im plications for Bilingual Education. Austin, Texas: National Education Laboratory Publishers. Crothers, Edward and Patrick Suppes 1967. Experiments in Second-Language Learning. New York: Academic Press. Debo, Angie 1934. The Rise and Fall of the Choctaw Republic. Norman: University of Oklahoma Press. DeRosier, Arthur H. 1970. The Removal of the Choctaw Indians. Knoxville: University of Tennessee Press. deVilliers, J. and P. deVilliers 1973. A cross-sectional study of the acquisition of grammatical morphemes in child speech. Journal of Psycholinguistic Research, 2, 3:267-278. Dillard, Karl C. 1978. The Language Teaching Controversy. Rowley, Mass.: Newbury House Publishers. DiPietro, Robert J. 1971. Language Structures in Contrast. Rowley, M ass.: Newbury House Publishers.

PAGE 318

Drach, K. 1969. The language of the parent: A pilot study. Working Paper No. 14, The Structure of Linguistic to Children. University of California, Berkeley: Language Behavior Research Laboratory. Dubin, F. 1970. The Sociolinguistic Dimension of Dormitory English. ERIC 038637. Dulay, Heidi C. and Marina K. Burt 1973. Should we teach children syntax? Language Learning, 23,2:245-258. Dulay, Heidi C. and Marina K. Burt 304 In Input 1974a. An analysis of children's second language errors. In J. Richards, ed., Error Analysis, Per spectives on Second Language Acquisition, pp. 95-123. London: Longman Group. Dulay, Heidi C. and Marina K. Burt 1974b. Natural sequences in child second language acquisition. Language Learning, 24,1:37-53. Dumont, Robert V., Jr. 1972. Learning English and how to be silent: Studies in Sioux and Cherokee classrooms. In C. Cazden, V. John, and D. Hymes, eds., Functions of Language in the Classroom, pp. 344-369. New York: Teachers College Press. Ervin-Tripp, Susan 1971. An overview of theories of grammatical develop ment. In D. Slobin, ed., The Ontogenesis of Grammar, A Theoretical Symposium, pp. 189-212. New York: Aca demic Press. Ervin-Tripp, Susan 1973. Some strategies for the first two years. Timothy E. Moore, ed., Cognitive Development and Acquisition of Language, pp. 261-268. New York: Academic Press. Fasold, Ralph In the 1972. Tense Marking in Black English. Arlington, Va.: Center for Applied Linguistics. Ferguson, Charles A. 1965. General introduction to the Contrasti v e Struc ture Series. In Robert P. Stoc k well and J. D. Bowen, The Sounds of English and Spanish, pp. vv i. Chicago: The Universit y of Chicago Press.

PAGE 319

305 Ferguson, Charles A. 1971. Absence of copula and the notion of simplicity: A study of normal speech, baby talk, foreigner talk, and pidgins. In D. Hymes, ed., Pidginization and Creolization of Languages. pp. 141-150. Cambridge: Cambridge University Press. Fischer, John L. 1964. Social influence in the choice of a linguistic variant. In Dell Hymes, ed., Language in Culture and Society, pp. 483-488. New York: Harper and Row. Fishman, Joshua A. 1972. The Sociology of Language: An Interdisciplinary Social Science Approach to Language in Society. Rowley, Mass.: Newbury House Publishers. Fodor, J. and J. Katz 1964. The Structure of Philosophy of Language. Prentice-Hall. Gibson, A. and P. Kwachka Language, Readings in the Englewood Cliffs, N.J.: 1977. Choctaw and Anglo Teaching and Learning Styles. A paper presented to the National Indian Education Association Conference at St. Paul, Minn., November, 1977. Gibson, A. and P. Kwachka 1978 . Choctaw Language and Culture, II. Unpublished materials developed for the Bilingual-Bicultural Tea cher Education Project, Mississippi State: Mississippi State University. Haas, Mary 1941. The classification of the Muskogean languages. In Leslie Spier, ed., Language, Culture, and Person ality, pp. 41-56. Menasha, Wis.: University of Wis consin Press. Haas, Mary 1971. Southeastern Indian linguistics. In Charles M. Hudson, ed., Red, White, and Black: Symposium on Indians in the Old South. Southern Anthropological Society Proceedings, 5:44-54. Athens, Ga.: Southern Anthropological Society. Hakuta, K. 1978. A report on the development of grammatical mor phemes in a Japanese girl learning English as a second language. In E. Hatch, ed., Second Language Acquisi tion, A Book of Readings, pp. 148-154. Rowley, Mass.: Newbury House Publishers.

PAGE 320

306 Hatch, Evelyn 1978a. Second Language Acquisition, A Book of Readings. Rowley, Mass.: Newbury House Publishers. Hatch, Evelyn 1978b. Acquisition of syntax in a second language. In Jack C. Richards, ed., Understanding Second and Foreign Language Learning, Issues and Approaches, pp. 34-70. Rowley, Mass.: Newbury House Publishers. Harvey, Gina Cantoni n.d. Dormitory English: Implications for the Teacher. Mimeo ms. Jacob, Betty, Dale Nicklas, and Betty 1977. Introduction to Choctaw. semination and Assessment Center tion. Jakobovits, Leon Lou Spencer Austin, Texas: Dis for Bilingual Educa1970. Application of recent psycholinguistic devel opments for the teaching of a second language. In Mark Lester, ed., Readings in Applied Transformational Grammar, pp. 245-268. New York: Holt, Rinehart, and Winston. Jespersen, Otto 1964. Essentials of English Grammar. University, Ala.: University of Alabama Press. Original publi cation, 1933. John, Vera P. 1972. Styles of learning, styles of teaching: Reflec tions on the education of Navaho children. Inc. Caz den, V. John, and D. Hymes, eds., Functions of Language in the Classroom, pp. 331-343. New York: Teachers College Press. Katz, J. 1966. The Philosophy of Language. New York: Harper. Labov, William 1972. Sociolinguistic Patterns. Philadelphia: University of Pennsylvania Press. Lambert, Wallace E., R. C. Gardner, H. C. Barick, and K. Tunstall 1963. Attitudinal and cognitive aspects of intense study of a second language. Journal of Abnormal and Social Psychology, 66:358-368.

PAGE 321

Langendoen, D. Terence 1970. Essentials of English Grammar. New York: Holt, Rinehart and Winston. Leachman, D. and Robert A. Hall 1955. American Indian pidgin English: Attestations and grammatical peculiarities. American Speech, 30: 163-171. Lehmann, Winfred P. 1972. Descriptive Linguistics, An Introduction. 2nd ed. New York: Random House. Leibowitz, Arnold H. 307 1971. A history of language instruction in American Indian schools: The imposition of English by government policy. In Bilingual Education for American Indians, Indian Education, Curriculum Bulletin No. 3:1-6. Washington, D.C.: U.S. Bureau of Indian Affairs. Lenneberg, Eric H. 1964. The capacity for language acquisition. In J. Fodor and J. Katz, eds., The Structure of Language, Readings in the Philosophy of Language, pp. 579-603. Englewood Cliffs, N.J.: Prentice-Hall. Lenneberg, Eric H. 1973. On explaining language. In Mark Lester, ed., Readings in Applied Transformational Grammar, pp. 83-105. New York: Holt, Rinehart and Winston. Leopold, W. F. 1937. The Speech Development of a Bilingual Child. 4 vols. Evanston, Ill,: Northwestern University Press. Lewis, Edmond, J. R. Scott, L. J. Mardis, and Paul Liberty 1977. Annual Evaluation Report, 1976-1977. Bilingual Education for Choctaws of Mississippi. Philadelphia, Miss.: Mississippi Band of Choctaw Indians. Livero, James L. n.d. Developing the Oral Language Program. Albu querque, N.M.: Southwestern Cooperative Education Laboratory. Locks, Nancy A., B. Pletcher, and D. Reynolds 1978. Language Assessment Instruments for Limited English-Speaking Students, A Needs Analysis. Wash ington, D.C.: U. S. Department of Health, Education, and Welfare and the National Institute of Education. Luria, A. R. 1969. Speech development and the formation of mental

PAGE 322

308 processes. In M. Cole and I. Maltzman, eds., A Hand book of Contemporary Soviet Psychology. New York: Basic Books. Martin, Jeanette P. 1975. A Survey of the Current Study and Teaching of North American Indian Languages in the United States and Canada. Arlington, Va.: Center for Applied Lin guistics. McNeill, David 1966. Developmental psycholinguistics. In F. Smith and G. A. Miller, eds., The Genesis of Language: A Psycholinguistic Approach, pp. 15-84. Cambridge, Mass.: The M.I.T. Press. McNeill, David 1970. The Acquisition of Language: The Study of De velopmental Psycholinguistics. New York: Harper and Row, Publishers. Menyuk, Paula 1969. Sentences Children Use. Research Monograph No. 52. Cambridge, Mass.: The M.I.T. Press. Miller, Wick and Susan Ervin 1964. The development of grammar in child language. In U. Bellugi and R. Brown, eds., The Acquisition of Language. Monographs of the Society for Research in Child Development, 29.92:9-34. Milon, J. 1972. A Japanese Child Learns English. A paper presented at the TESOL Convention, San Francisco. Mithun, Marianne and Wallace L. Chafe 1979. Recapturing the Mohawk language. Shopen, ed., Languages and their Status, Cambridge, Mass.: Winthrop Publishers. Moore, Timothy E. In Timothy pp. 3-33. 1973. Cognitive Development and the Acquisition of Language. New York: Academic Press. Morrison, James D. 1975. Schools for the Choctaws. Durant, Okla.: Choc taw Bilingual Education Program, Southeastern Oklahoma State University. Moskowitz, Breyne A. 1978. The acquisition of language. Scientific Ameri can, 239,5:92-108.

PAGE 323

Natalicio, D. C. and L. F. S. Natalicio 1971. A comparative study of English pluralization by native and non-native English speakers. Child Development, 42:1302-1306. Nicklas, Dale 1972. The Elements of Choctaw. Ph.D. dissertation, University of Michigan, Ann Arbor. Nicklas, Dale 309 1975. Choctaw morphophonemics. In James Crawford, ed., Studies in Southeastern Indian Languages, pp. 237-249. Athens: University of Georgia Press. Ohannessian, Sirarpi, ed. 1967. The Study of the Problems of Teaching English to American Indians. Washington, D.C.: Center for Applied Linguistics. Ohannessian, S. and W. Gage 1969. Teaching English to Speakers of Choctaw, Navajo and Papago, A Contrastive Approach. Indian Education, Curriculum Bulletin No. 6. Washington, D.C.: U. S. Bureau of Indian Affairs. Oller, John, Lori Baca, and Fred Vigil 1977. Attitudes and attained proficiency in ESL: A sociolinguistic study of Mexican Americans in the Southwest. TESOL Quarterly, 11, 2:173-183. Osborn, Lynn R. 1970. Language, poverty, and the North American Indian. In Frederick Williams, ed., Language and Poverty, Perspectives on a Theme, pp. 229-247. Chi cago: Markham Publishing Co. Palermo, David S. 1971. On learning to talk: Are principles derived from the learning laboratory applicable? In Dan I. Slobin, ed., The Ontogenesis of Grammar, A Theoretical Sym posium, pp. 41-62. New York: Academic Press. Paulston, Christina B. and Mary N. Bruder 1976. Teaching English as a Second Language: ~ Teehniques and Procedures. Cambridge, Mass.: Winthrop Publishers. Peters, Ann 1977. Language learning strategies: Does the whole equal the sum of the parts? Language, 53,3:560-573.

PAGE 324

Peterson, John H. 1970. The Mississippi Band of Choctaw Indians: Their Recent History and Current Social Relations. Ph.D. dissertation, University of Georgia, Athens. Pfuderer, C. 1969. Some suggestions for a syntactic characteriza tion of baby talk styles. In Working Paper No. 14, The Structure of Linguistic Input to Children. Lan guage Behavior Research Laboratory. Berkeley: Uni versity of California Press. Philips, Susan u. 310 1972. Participant structures and communicative com petence: Warm Springs children in community and class room. In C. Cazden, V. John and D. Hymes, eds., Func tions of Language in the Classroom, pp. 370-394. New York: Teachers College Press. Plumer, Davenport 1970. A summary of environmentalist views and some educational implications. In Frederick Williams, ed., Language and Poverty, Perspectives on a Theme, pp. 265-308. Chicago: Markham Publishing Company. Powell, J. W. 1966. Indian linguistic families of America north of Mexico. In Franz Boas and J. W. Powell, eds., Ameri can Indian Languages, pp. 83-221. Lincoln: Univer sity of Nebraska Press. Original publication, 1891. Ravem, Roar 1974a. Acquisition in a second language environment. In J. Richards, ed., Error Analysis, Perspectives on Second Language Acquisition, pp. 124-133. London: Longman Group. Ravem, Roar 1974b. The development of Wh-questions in first and second language learners. In J. Richards, ed., Error Analysis, Perspectives on Second Language Acquisition, pp. 134-155. London: Longman Group. Richards, Jack C., ed. 1974a. Error Analysis, Perspectives on Second Lan guage Acquisition. London: Longman Group. Richards, Jack C. 1974b. A non-contrastive approach to error analysis. In Jack C. Richards, ed., Understanding Second and Foreign Language Learning, Issues and Approaches, pp. 172-188. Rowley, Mass.: Newbury House Publishers.

PAGE 325

311 Richards, Jack C., ed. 1978. Understanding Second and Foreign Language Learning, Issues and Approaches. Rowley, Mass.: Newbury House Publishers Sachs, Jacqueline 1976. The development of speech. In Edward c. Car terette and M. P. Friedman, eds., Handbook of Per ception, Vol. 7, Language and Speech, pp. 145-172. New York: Academic Press. Schiefelbusch, R. L. and L. L. Lloyd 1974. Language Perspectives--Acquisition, Retardation, and Intervention. Baltimore: University Park Press. Scollon, Ronald 1976. Conversations with a One Year Old: A Case Study of the Foundations of Syntax. Honolulu: The University of Hawaii Press. Seliger, Herbert W. 1978. Implications of a multiple critical periods hypothesis for second language learning. In William C. Ritchie, ed., Second Language Acquisition Research, Issues and Implications, pp. 11-19. New York: Academic Press. Selinker, Larry 1974. Interlanguage. In J. Richards, ed., Error Analysis, Perspectives on Second Language Acquisition, pp. 31-54. London: Longman Group. Shipley, E. R., C. S. Smith, and L. R. A. Gleitman 1969. A study in the acquisition of language: Free responses to commands. Language, 45:322-342. Shuy, Roger W., W. Wolfram, and William K. Riley 1968. Field Techniques in an Urban Language Study. Washington, D.C.: Center for Applied Linguistics. Silverman, Robert J., Joslyn K. Noa, and Randall H. Russell 1976. Oral Language Tests for Bilingual Students: An Evaluation of Language Dominance and Proficiency Instruments. Portland, Or : Northwest Regional Edu cational Laboratory, 1976. Reprinted as "Issues in Language Testing," Bilingual Resources 1,2:36-39. Skinner, B. F. 1957. Verbal Beha v ior. New York: Appleton-Century Crofts. Slobin, Dan I., ed. 1971. The Ontogenesis of Grammar, A Theoretical Symposium. New York: Academic Press.

PAGE 326

312 Snow, C. 1972. Mothers' speech to children learning language. Child Development, 43:549-565. Southwestern Cooperative Education Laboratory, Inc. 1971. The SWCEL Test of Oral English Proficiency. Albuquerque, N.M.: Southwestern Cooperative Education Laboratory. Spencer, Barbara, J. H. Peterson, and Choon S. Kim 1975. Choctaw Manpower and Demographic Survey, 1974. Philadelphia, Miss.: Mississippi Band of Choctaw Indians. Spolsky, Bernard, ed. 1972. The language education of minority children. In Bernard Spolsky, ed., The Language Education of Minority Children, Selected Readings, pp. 1-10. Rowley, Mass.: Newbury House Publishers. Staats, Arthur W. 1971. Linguistic-mentalistic theory versus an explana tory S-R learning theory of language development. In Dan I. Slobin, ed. The Ontogenesis of Grammar, A Theoretical Symposium, pp. 103-150. New York: Aca demic Press. Staats, Arthur W. 1974. Behaviorism and cognitive theory in the study of language: a neopsycholinguistics. In R. L. Schiefelbusch and L. L. Lloyd, eds., Language Perspec tives: Acquisition, Retardation, and Intervention, pp. 615-646. Baltimore: University Park Press. Stockwell, Robert P. and J. D. Bowen 1965. The Sounds of English and Spanish. Chicago: University of Chicago Press. Swanton, John R. 1931. Source Material for the Social and Ceremonial Life of the Choctaw Indians. U. S. Bureau of American Ethnology, Bulletin No. 103. Washington, D. C.: U.S. Government Printing Office. United States Commission on Civil Rights 1975. A Better Chance to Learn: Bilingual Bicultural Education. Clearinghouse Publication No. 51. Washing ton, D.C.: u. S. Government Printing Office. Valdman, A. and J. S. Phillips 1975. Pidginization, creolization and the elaboration of learner systems. In A. Valdman and H. Gradman, eds., Studies in Second Language Acquisition, Proceedings

PAGE 327

313 of the Fourth Neuchatel Colloquium in Applied Lin guistics, pp. 21-40. Bloomington: Indiana University Linguistics Club. Vildomec, Veroboj 1963. Multilingualism. Leyden, The Netherlands: A. W. Sythof f. Wax, Murray, Rosalie Wax, and Robert Dumont 1964. Formal education in an American community. Social Problems 2,4. Williams, Frederick. 1970. Language, attitude, and social change. In Frederick Williams, ed., Language and Poverty, Per spectives on a Theme, pp. 380-399. Chicago: Markham Publishing Co. Witherspoon, Gary 1977. Language and Art in the Navaho Universe. Ann Arbor: University of Michigan Press. Wisconsin Native American Language Project 1974. Lessons in Basic Menominee. Milwaukee: Great Lakes Inter-Tribal Council and the University of Wisconsin-Milwaukee. Wolfram, Walt 1971. Black-White speech differences revisited. In W. Wolfram and Nona Clarke, eds., Black-White Speech Relationships, pp. 139-161. Washington, D.C.: Center for Applied Linguistics. Wolfram, Walt and Nona H. Clarke, eds. 1971. Black-White Speech Relationships. Washington, D.C.: Center for Applied Linguistics. York, Kenneth and Robert J. Scott 1976. Annual Evaluation Report, 1975-1976, Bilingual Education for Choctaws of Mississippi. Philadelphia, Miss.: Mississippi Band of Choctaw Indians. ---------------

PAGE 328

BIOGRAPHICAL SKETCH Patricia Butler Kwachka is a native of Florida where she attended college, receiving her B.A., M.A. and Ph.D. degrees in English, anthropology, and linguistics, respec tively, from the University of Florida, Gainesville. A son, Alexis, was born in mid-M.A., and is now in his teens. Kwachka has worked as an anthropological linguist in southern prisons and rural communities, and for the Mis sissippi Band of Choctaw Indians as well as other Native American groups. She teaches in the areas of ESL, anthropology, and linguistics, and currently is Acting Head of the Department of Cross-Cultural Communication at the University of Alaska, Fairbanks. 314

PAGE 329

I certify that I have read this study and that in my op1n1on it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. I certify that I have read this study and that in my op1n1on it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. ;de C. Scott Assistant Professor of English and Linguistics I certify that I have read this study and that in my op1n1on it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy Ruthellen Crews Professor of Instructional Leadership and Support J

PAGE 330

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Harry B. Shaw Associate Professor of English I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. ..... Milanich Professor of Anthropology This dissertation was presented to the Graduate Faculty of the lnterdiscipl inary Graduate Program in Linguistics in the College of Liberal Arts and Sciences, and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. March, 1981 Dean, Graduate School I /

PAGE 331

r ~ --UNIVERSIT Y O F FL O RIDA II I II IIIIII Ill Ill lllll lllll II IIIII I III I II lllll l 11111111111111111 3 1262 08553 5903

PAGE 332

r ~ --UNIVERSIT Y O F FL O RIDA II I II IIIIII Ill Ill lllll lllll II IIIII I III I II lllll l 11111111111111111 3 1262 08553 5903


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EVGEW3O24_BZSWOV INGEST_TIME 2011-07-01T15:31:48Z PACKAGE UF00102859_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES