-
FICG' 0-.

ASPECTS OF THE ENGLISH SYNTAX
or cmcmo SCHOOL CHILDREN

Thesis for the Degree of M. A.
MICHIGAN STATE UNIVERSITY
BARBARA MAY HORVATH

1971 ’

......

......

  
 

 

LIB R A 7’ 1"
Michigan :‘ ' 1.3
Univ“. Si .2 y

"f

 

 

M!““‘l“fl‘£“fl““ W!“ “l!“ W

 

 

 

 

ASPECTS OF THE ENGLISH SYNTAX

OF CHICANO SCHOOL CHILDREN

BY

Barbara May Horvath

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

I'IASTER OF ARTS

Department of Linguistics and Oriental
and African Languages

1971

Chapter

Contents

I INTRODUCTION

1.0
2.0

The Chicano Situation
Overview of the Present Study

II APPROACHES TO THE STUDY OF CHILD LANGUAGE

1.0
2.0

Introduction

Theoretical Vs. Applied Approaches

to the Study of Child Language

Goals of the Research

Competence Vs. Performance

The Concept of Maturity
Statistical Studies

Data Collection
Approach Taken in the Present Study
ummary and Evaluation of the O'Donnell

MN
on

gNNN
coo
OUDUQNH

n
O.
‘4

Objectives

Theoretical Background

The Use of the T-Unit

Data Collection Techniques
Inadequacies of Structural Analyses
Results of the O'Donnell Study

b§b§bbmm
c
C‘U‘wal-‘C

III A STUDY OF CHICANO ENGLISH

1.0
2.0

3.0

Introduction

Background to the Present Study

2.1 The Children

2.2 The Types of Data Collected

A Comparison of the Syntactic Structures

Used by CE and MCE Speakers

3.1 Overall Results of the Study

3.2 Results of the Study of Specific
Construction Types
3.2.1 Nominal Constructions
3.2.2 Adverbial Constructions
3.2.3 Coordinate Constructions

3.3 Structural Patterns of Main Clauses

3.4 Predictive Ability of the Number of
words per T-Unit measure

3.5 Conclusions

ii

Page

oocn

12
15
17
19
20

20
21
23
24
25

28

34

34
36
36
36
39

39
50

52
58
61
62
65

68

Chapter
III (Con'd)

IV

Bibliography

Appendices

4.0 Further Investigation of CE
4.1 Introduction
4.2 Results of the Sentence Repetition
Exercise
4.3 Conclusions
4.4 Critique of the Sentence Repetition
Exercise

SUMMARY AND CONCLUSIONS

A.

C.
C.

1.0 Summary
2.0 Conclusions
3.0 The Role of the Applied Linguist

Chicano English Data

Chicano English Worksheet

A List of Nonstandard Grammatical Forms
Results of the Sentence Repetition Exercise
Given to Three Anglo Children

iii

Page

72
72
74

80
81

84

84
87
89

91

94
96
97
103

Table

10

ll

12

l3

14

15

16

List of Tables

Title

A4 _. MA

Background Information on the Children in the Study
Group

Length of Responses

Mean Number of Words per T-Unit by Grade

Number of Words per T-Unit: A Comparison
Percentage of T-Units Having Less Than Nine Words

Mean Number of Sentence-Combining Transformations per
T-Unit

Percentage of T-Units Having No Sentence-Combining
Transformations (CE Only)

Types of Constructions Formed by Sentence-Combining
Transformations

Headed Nominal Constructions Formed by Sentence-
Combining Transformations

Non-Headed Nominal Constructions Formed by Sentence-
Combining Transformations

Grammatical Functions of Nominal Constructions

Adverbial Structures Formed by Sentence-Combining
Transformations

Coordinate Structures Formed by Sentence-Combining
Transformations

Structural Patterns of Main Clauses
Individual Rank (Speech)

Individual Rank (Writing)

iv

40

41

44

46

48

49

51

53

57

59

60

63

64

67

69

Acknowledgments

I would like to express my appreciation to the following people.

The members of my committee, Dr. Meyer Wolf, Dr. Ruth Brend,
and Dr. Julia Falk, for assisting me in the preparation of

this thesis.

iB. Toni Sims for her enthusiastic c00peration with this

project.

Eddie, Robert, Laura, Ana, Ceferino, David, Rene, Kathy, and
Becky, members of Ms. Sims' class, for making the data

collection a joy.

Chapter I.

INTRODUCTION

1.0 The Chicano Situation

Within the last decade educators have become interested in the
educational and social development of the so-called 'disadvantaged'
child,and they often include Chicano children in this category. Like
Black Americans, Mexican Americans are separated from the mainstream
of American society; about three-quarters of all Chicanos live in
barrios, i.e., in neighborhoods where the residents are predominantly
Chicanos (Heller: p. 21). The 1960 United States Census revealed that
in the five states of the Southwest where the majority of Chicanos live
(Arizona, New Mexico, Texas, Colorado, and California), 36% of adult
Mexican Americans were either illiterate or functionally illiterate,
that is, had only one to four years of schooling, as compared to only
5% of the Anglo Americans in these states (Heller: p. 15).

Chicanos have a high drop-out rate (or perhaps more accurately
stated, a high push-out rate) from schools. In 1960 the median number
of school years completed by Chicano males in urban areas was 8.4 years
as compared with 11.0 years for the total U.S. male urban population

(Heller: pp. 14-15). One of the reasons given for the high drop-out

rate is that the difficulties these students have with English prevent
them from doing well in other subjects (Manuel: p. 115; Heller: p. 14 ff).
One educator, W. R. Sinninger, has observed that the Chicano student
begins to drop below norms in the fourth grade and continues to drop

in each succeeding grade (Manuel: p. 115). In addition little or no
instruction is given Chicanos in their native language, Spanish, until
they reach high school, and then it is often taught as though it were

a foreign language for them. As a result, some educators assert that

most Chicano children know neither Spanish nor English well (Manuel:

p. 117). These assertions by educators reveal a lack of understanding

of the nature of language; they are doubtless referring to the fact

that the Chicanos are not taught to read and write in Spanish and also,
perhaps, that they speak nonstandard dialects of both Spanish and English.
Nevertheless, they do point out that the needs of the Chicano community
are not being met by the educational system. The general lack of interest
in the bilingual child is reflected in the paucity of published material
available on the languages spoken by Chicanos. What is published often
deals with the 'problem' of language, but actual grammatical analyses of
English or Spanish as spoken by Chicanos have not yet been published.

When the Chicano child comes to kindergarten, he speaks either
Spanish only or Spanish and, to a limited extent, English (Manuel: p. 123).
He is expected generally to pick up English on his own; often there are
no specific classes organized for teaching him English. The language
spoken in the home, on the playground, and elsewhere in the community is
normally Spanish and the majority of the children's practice in English

comes in the classroom (Manuel: p. 123; Heller: pp. 29-30). With

classes of thirty-five children, where the teacher does most of the
talking, it is not strange that these children will not develop a native
facility for English. The difficulties resulting from not being able
to use English effectively are cumulative, so that Chicano students begin
to fall back in all school subjects and eventually are pushed out of the
educational system.

Within the past few years there has been increased interest in these
children, as evidenced by an increase in federally aided programs and
in publications on topics related to Mexican Americans .* A great deal
of discussion is given to language difficulties and to possible solutions.
Although these programs vary from one school district to another, there
seem to be two basic approaches now being considered: the Teaching English
as a Second Language (TESL) approach and the bilingual approach. In the
bilingual program, instruction is generally given in Spanish for the first
three or four years and then English is used, whereas in the TESL approach,
English is the medium of instruction from the start (Del Buono: personal
communication). Some communities are also emphasizing the need for
bicultural programs to be included in their bilingual programs, recognizing

that different languages reflect more than just different grammars.

w

Several recently published bibliographies reflect this increased
interest in the Chicano community. Among them the following seem to be
the most general and complete: University of California at Los Angeles.

Mexicgg-Americgg Study Project Revised Bibliography, Advance Report No. 3,

Los Angeles: 1967; Stanford University, Center for Latin American Studies.

The Mexican-American: A Selected and égnotated Bibliography. Stanford,
California: 1969.

No matter which approach is taken, however, it seems fairly obvious
that there must be some analysis of the English and Spanish grammar of
Chicano children. Yet in recent years no systematic studies have been
made of the languages used by these children (Heller: p. 30). Courses
designed to help them learn either language must take into account what
they know about that language and what aspects of that language they

are likely to have difficulty with.

2.0 Overview of the Present Study

This thesis is a comparative study of the grammar of English used
by middle-class Anglo children and Chicano children, most of whom are
approximately ten years old and in the fifth grade of school. The
middle-class Anglo children were chosen as a comparative base because it
was assumed that curriculum design and textbooks were aimed at this
particular group. Given the fact of the high school drop-out rate and the
observation by many educators that one of the problems which Chicanos
have in school stems from their lack of facility with English, it may prove
interesting to describe the grammatical structures used by Chicanos and
compare them with the structures used by Anglos.

How one goes about describing the language use of children becomes
a central problem in this thesis. Theoretical grammarians, particularly
transformational grammarians, focus their attention on competence, that
is, what a speaker knows about his language. The study of language use
is often left to researchers whose primary goals are not the theoretical
explication of grammar per se. The people who study language use often

have more practical aims in describing the grammar of children and usually

expect to see the results of their research reflected in direct changes
within the educational structure. Chapter II of this thesis discusses
in detail some theoretical and applied approaches to the study of child
language.

A study of the syntax of middle-class Anglo children by O'Donnell,
Griffin, and Norris entitled Syntax of Kindergarten and Elementayy
School Chgldren: A Trggsformational Analysis was chosen as a model for

collecting data and describing the syntax used by the Chicano children,
as well as providing the comparative data. A summary and critical evaluation
of the O'Donnell study is given in Chapter II.

The data collection techniques, some general background of the
Chicano children used in the study, and results of the comparative analysis
form the major part of Chapter III. Appendix A is a tabulation of the
statistical information obtained in the analysis; Appendix B is a sample

data sheet.

Because it was not the purpose of the O'Donnell study to analyze
nonstandard grammatical structures, using this study as a model did not
allow such structures to be isolated in the Chicano data. In the O'Donnell
study, nonstandard utterances were either classified within the categories
set up or were not included as part of the data. Obviously, this situation
is not satisfactory, particularly in the study of English as spoken by
Chicanos where nonstandard forms may prove interesting and provide some
answers as to how to design courses for these children. It also may be
that there exists a Chicano dialect of English, analogous to Black English.
The Chicanos' isolation in barrios would provide similar conditions for the

possible development of such a dialect. The implications for teaching

then would be profound: not only would the children be required by
the schools to be bilingual, but bidialectal in English as well.

In an attempt to investigate the possibility of such a dialect, a
sentence repetition exercise was given to the Chicano children based on
observed use of nonstandard grammatical structures. Since the examination
of Chicano English as a separate dialect was not one of the major goals
of this thesis, only tentative and speculative results are given in
Chapter 111. Appendix C gives in tabular form all of the observed non-
standard forms used by the Chicano children along with a tentative explana-
tion for such nonstandardisms, i.e., whether it may be the result of
interference from Spanish, or whether it is probably a widely-used
nonstandard form, or whether it is a nonstandard form typically used by
all children in the acquisition of English.

The sentence repetition exercise was given to three Anglo children in
an attempt to evaluate the validity of the test. The results are given
in Appendix D.

very little analysis was done in this thesis on problems related to
interference from Spanish; however, this is an area which must also be
taken into account when looking for solutions to the problems of Chicano
school children. Meaningful English courses must take into account those
areas of English which would predictably be difficult for Spanish-speakers
learning English. Some work has been done in phonology (Olguin), but
very little has been done with aspects of syntax (however, see Stockwell,
Bowen, and Martin), especially the particular problems that children

have in acquiring a second language.

Conclusions to the comparative study and the sentence repetition

exercise are given in Chapter III. Chapter IV contains a summary and

conclusions concerning the type of study done, that is, an applied linguistic
study of the use of language. The role of the applied linguist and his

relationship to the field of linguistics as a whole is also discussed.

Chapter II.

APPROACHES TO THE STUDY OF THE LANGUAGE
OF CHILDREN

1.0 Introduction

The language of children has been studied by both educators and
linguists for many years. The approaches taken to these studies have
differed widely based primarily on the goals of the researchers. It will
be useful for the purposes of this thesis to identify two major approaches
to research on child language, hereafter referred to as the applied and
the theoretical approaches. By contrasting the goals and methodologies
of these two approaches, a better understanding can be obtained of the
place of this thesis within the larger framework of research on child

language.

2.0 Theoretical Vs. Applied Approaches to the Study of Child Language
2.1 Goals of the Research - The differences between the theoretical
and applied approaches become apparent upon examining the respective
goals. The applied approach is most often taken by researchers whose
educational backgrounds are either in the fields of education or English.
Their goals, which are often not explicitly stated, appear to be to

describe the language of older children ranging from kindergarten through

high school. The object of research can be to find means of measuring

the linguistic capabilities of the children, or to design curriculum to

suit various levels of linguistic maturity, or to develop teaching techniques
for improving language-related skills such as reading and writing.

The goals of the theoretical researcher are quite different. He
is most often a linguist interested in advancing the development of a
theory of grammar, often transformational grammar. The goals of his
research are to provide evidence for claims made by the theory suCh as
the innate capacity of human beings to acquire language, the universal
nature of language, or the creative aspect of language. He is also
interested in what children learn about their native language as well as
how they learn it. The children studied are usually young, that is,
below five years of age, because the focus is on the acquisition of
language rather than on the use of language.

The place of theory in relation to the goals of research is also an
important difference. As the name implies, the development of theory
plays a central role in research using the theoretical approach. However,
in the applied studies, grammatical theories may or may not be a part
of the research. Where theory is brought in, it often plays a peripheral
role used when convenient to provide an explanation or to provide a very
general framework for the study.

2.2 Competence Vs. Performance - Chomsky has established a dichotomy
which is useful in contrasting these two approaches to child language
research. This dichotomy, competence and performance, fairly well sums

up the differences in focus between theoretical and applied research.

10

According to Chomsky's definitions, competence is the speaker-hearer's
primarily subconscious knowledge of his language and performance is the
actual use of language in concrete situations (Chomsky: 1965, p. 4).

Applied research focuses on performance and is often characterized
by data collection techniques which record all utterances made by the
children, including garbles such as false starts, repetitions, etc. Given
the emphasis on school language, much of the data used in the studies
come from the written work of children. The analysis of the data is often
statistical in nature, either in terms of counting words, e.g., the number
of words in a sentence, in a written composition, or in a story told by
the child, or in tabulating the frequency of occurrence of various
grammatical structures.

One basic criticism of the emphasis of applied research on performance
is that there is little explicit discussion of the underlying assumptions
upon which the studies are based, particularly regarding the difficulties
in studying performance and its relation to competence. In fact, some
of the conclusions which have been arrived at in these studies make it
apparent that there is very little understanding on the part of the
applied researchers of the differences between competence and performance
and the relationship of their studies to studies of competence.

Theoretical researchers, on the other hand, are quite explicit about
the differences between competence and performance and emphasize the study
of competence as the primary goal of their studies. HcNeill points out
quite clearly that, particularly in the study of child language, the
distinction between competence and performance must be maintained. The

object of the study of child language, as he sees it, is to account for

11

the emergence of competence itself and eventually to be able to account
for performance, and this cannot be accomplished, he feels, by confusing
competence and performance (MtNeill: p, 17),

Chomsky has noted that the description of competence has the greatest
psychological relevance because it is concerned with the kinds of struc-
tures which are internalized, whether or not they are used. To illustrate
the notion that failure to use particular grammatical features does not
necessarily indicate that the structure is not a part of the speaker's
internalized grammar, Chomsky cites a phonological example. In acquiring
the sound patterns of his language, a child will immediately put in a
newly acquired sound all across the board, i.e., with the acquisition of
the initial sound /s/, /kuwl/ becomes /skuW1/, /puwn/ becomes /squn/,
etc., indicating that there was internalized knowledge of the occurrence
of such sounds before the child was able to pronounce them. Competence,
he claims, exceeds performance (Chomsky: 1970, p. 43).

The emphasis on performance and the failure to examine the underlying
assumptions of their research have led the applied researchers to describe
child language in terms which would be inappropriate in a theoretical
study of language based on transformational grammar. Such terms as
'manipulation' and 'control' of syntax reflect an underlying difference
in the view of what language is. The use of these terms implies, at least
to a linguist, the psychological reality of rules and perhaps even conscious
control in the generation of grammatical structures. No such claim is made

by theoretical grammarians.

12

2.3 The Concept of Maturity - One of the aims of applied research
is to devise means of measuring the linguistic maturity of children.
Statistical measures are sought in terms of the frequency of occurrence
of certain grammatical structures in the spoken and written work of
children of various age/grade levels. The increase in the frequency
of occurrence over time (or in some cases to be mentioned below - the
decrease in frequency) is used as a measure of growth, maturity, or
control.

The emphasis on maturity, in particular the measurement of maturity
in children, is one which seems peculiar to the applied researchers. But
in examining the studies by Hunt, Laban, (VDonnell, Griffin, and Norris,
and Mellon, who attempt to measure maturity, we find they do not seem to
have a firm notion about what maturity is. Frequently in the literature
of the applied researchers, maturity is referred to as an increased
'control' in the use of language. As previously mentioned, this term is
a difficult one for linguists to accept. O'Donnell, Griffin and Norris
attempt to deal with the problem of what 'control' means. They say that
they do not claim that the increased use of any particular syntactic
structure is indicative of maturity and they recognize that circumstances
affect the appropriateness of the structures used. However, they say,
if statistics show that older children,under similar circumstances, make
use of certain syntactic features more frequently than younger children,
then they assume that the use of such features demonstrates firmer
comand of the resources of the language (O'Donnell et al: pp. 93-94).

In reviewing some of the ideas which have been put forth regarding

maturity, one finds what seems to be a search for one aspect of language

13

which will serve as an indicator of maturity. Riling focused on the
decrease in frequency of garbles, that is false starts, abnormal redun-
dancies, or word tangles (O'Donnell et al: p. 39), but found this to

be an unreliable index of growth, a conclusion which was further confirmed
in a study by O'Donnell, Griffin, and Norris (O'Donnell et al: p. 40).

Cross word count has also been examined as a measure of maturity.
Gross word count generally refers to the number of words in a written
composition. It can also, however, refer to the number of words used
in a spoken monologue. McCarthy has reviewed studies which indicated
that gross word count might be a fair measure of maturity, but then
Observation and common sense indicate that this is no measure of the adult
model and supposedly the adult is the ideal mature language user. Once
again, the O'Donnell study does not show a significant correlation between
gross word count and increased use of various complex syntactic structures
(O'Donnell et al: p. 97).

The subordination ratio index, which calculates the relative frequency
of the use of subordinate clauses, has also been used as a measure of
maturity (Hunt: 1970, p. 189), but again the reliability of this index was
not confirmed by the O'Donnell study. O'Donnell's children show that
kindergarten children use nominal, adjectival, and adverbial clauses quite
often and no significant increase from grade level to grade level in the
use of subordinate clauses was shown (O'Donnell et al: p. 98). Finally,
Loban has come to believe that it is not pattern, that is main clause
structure, but a demonstration of flexibility with patterns which indicates
control in the use of language (Loban: 1963, p. 84). Hunt and O'Donnel,

Griffin, and Norris further refine Loban's notion by pointing out that it

14

is not the appearance of certain syntactic structures that marks maturity,
since all the syntactic structures they counted in their data were used

by the youngest children in the groups studied. However, the most impor-
tant mark of maturity is frequency of occurrence of these structures,
particularly those structures produced by sentence-combining transformations
(Hunt: 1970, p. 189; O'Donnell et al: p. 89).

Another feature of maturity pointed by O'Donnell et a1 is the decreased
use of certain structures such as coordination of main clauses with 'and'
and the use of the sentence pattern Subject-Verb-Predicate Nominative.
First graders reduced by about forty percent, i.e., used fewer such
patterns than did kindergarteners, their use of main clause patterns in
which these predicate nominals follow linking verbs. The authors argue
that this fact reflects growth toward maturity, for the simple pattern
serves the very elementary purposes of indicating identification or equa-
tion. They give such examples as 'It was an ant.'; 'He was a hunter.';
and 'It was a rainy day.‘ (O'Donnell et al: p. 94). Presumably, adult
speakers indicate identification or equation by using other, more complex,
grammatical structures.

That introduces the prdblem, obviously, of describing the adult model
of language use so that we can know when a child is approaching maturity.
If we are to use frequency of occurrence as a measure of maturity, we should
have some idea of how frequently various structures are used by adults.
O'Donnell et al admit that they are making assumptions based on a loosely
defined notion of what adult performance would be and this influences their
notion of what constitutes development (O'Donnell et al: pp. 26-27). In

fact, they note that "restraint in the use of some syntactic possibilities

15

is a demonstration of control" (O'Donnell et al: p. 94). There must be
a limit to the nunber of times one can use, say, embedding, because too
much embedding will result in a lack of communication and will not be
regarded as mastery or control of language.

The frequency of occurrence of particular grammatical structures, then,
appears to define the notion of maturity for applied researchers such as
O'Donnell, Griffin, and Norris. But just what does the notion of frequency
of occurrence mean when applied to the concept of maturity? And how does
frequency of occurrence fit 'into Chomsky's division of grammar into
competence and performance? If the youngest child of all those studied
used a particular syntactic structure with a certain degree of frequency
and if the child uses this syntactic structure creatively, that is, in
ways which he could not have heard before, thereby ruling out the
possibility of direct imitation, do we not have to say that the child
knows the rules needed to generate such a structure? In other words, there
is a difference between maturity in the context of a theory of competence
and maturity in relation to a theory of performance. This distinction
is a necessary one and one which the applied researchers have failed to

take into account.

2.4 Statistical Studies - That frequency of occurrence should come
to define the concept of maturity is the obvious outcome of the methods
used by the applied researchers. Earlier it was noted that statistical
studies were used frequently by the applied researchers whose interests
are in performance, whereas this is not the case with theoretical researchers,

who focus their studies on competence.

16

The glorification of statistics is part of the myth of 'objective
scientific' studies. lees has noted that what is thought to be scientific
is marked by precise measurement, resulting in statistical analyses of
masses of data (Lees: p. 375); however, as he goes on to point out, precise
measurement is not nearly so characteristic of the scientific method as
is theory construction and validation. There is a basic misunderstanding,
he feels, ". . . that statistical methods, that is, elaborate counting
techniques, will not only reveal the correct analysis but even explicate
linguistic behavior" (Lees: p. 379). Statistics are mechanical and are
good for computers but they do not explain anything (Lees: pp. 379-380).
More recently, Chomsky has observed that some linguists and psychologists
see the relationship between competence and performance to be a probabilistic
one and, he asserts, this is not the case.(Chomsky: 1970, p. 44).

Among current applied researchers, Hunt talks about having certain
intuitive feelings about maturity which he hopes to be able to measure
quantitatively (Hunt: 1970, p. 189). O'Donnell et al and Leban are examples
of other applied researchers doing statistical studies. A more mature -
though doubtless just as ill-conceived - attitude toward statistical
studies is represented by another applied researcher, Ohmann, who
attempts to study literary style. He notes that while statistical counts
of grammatical features seem pertinent, significant results are highly
elusive. But he places the blame for this on the crudeness of the
categories made available by traditional grammarians. These categories,
he says, are numerous and lack a system which relates them in meaningful,

formally motivated ways. A count is nothing more than that without a theory

l7

and he contends that transformational theory can be used as the basis
for another statistical approach to a corpus of data, in this case, the
works of several novelists (Ohmann: p. 425).

Despite the fact that statistics have not been able to lead us to an
understanding of the competence of the native speaker, Hunt's concern
with intuitive feelings about the nature of maturity cannot be easily
brushed aside. The fact of the matter is that applied research has shown
that older children do use more complex grammatical constructions more
often than younger children (O'Donnell et a1: Hunt). It is this finding
which necessitates defining the concept of maturity within the context of

a theory of performance.

2.5 Data Collection - There is a significant difference in data
collection techniques among applied and theoretical researchers. These
differences are, once again, related to the focus on the study of performance
by the applied researchers as opposed to competence.

The data upon which applied research is based is usually the written
compositions or the spoken monologues of great numbers of children. The
attempt is made to obtain natural use of language without interference
on the part of the ivestigator. The kinds of information which such
techniques yield have been questioned by transformational grammarians who
emphasize the primacy of developing a theory of competence upon which a
theory of performance must be based (Chomsky: 1965, p. 10—15).

Chomsky has made it quite clear that this amassing of data will not
in his opinion yield the kind of results necessary to develop a theory of

grammar. The direct record of Speech, he contends, leads only to superficial

18

linguistic analysis (Chomsky: 1970, p. 44) and more exact description of
recorded speech will not lead to further explanation of the underlying
ability of a speaker to use and understand his language (Chomsky: 1965,

pp. 20-21). Although some people assume that the underlying knowledge

of the speaker can be derived from a corpus of data by well-developed

data processing techniques (Chomsky: 1970, p. 47), in actual fact, Chomsky
claims, no adequate, formalizable techniques are known for obtaining reliable
information concerning the facts of linguistic structure (Chomsky: 1965,

p. 19).

New experimental methods must be developed, Chomsky feels, which go
beyond mere observation of Speech in normal situations. In order to get
at competence, we need to study the ability of the Speaker to understand
and interpret sentences as well as actual output. But it must always
be remembered that data serve only as a ground for inference (Chomsky:
1970, p. 45). Since the study of actual output - or performance - is
the only way to study a speaker's competence, that is, to get at his
subconscious knowledge of his language, and direct record is unsatisfactory
as a data collection technique, Chomsky suggests that we carry out the
study of performance in "devious and clever ways" if we want to obtain
serious results (Chomsky: 1970, pp. 43-44). Particularly with regard to
child language acquisition, he suggests four specific areas in which
tests can be devised to determine the underlying ability of children to
1) use and comprehend sentences; 2) detect deviance and compensate for it;
3) apply rules in new situations; and 4) form highly specific concepts

from scattered bits of information (Chomsky: 1970, p. 47).

19

3.0 The Approach Taken in the Present Study

The applied researcher who is aware of the theoretical advancements
made by the proponents of transformational grammar is faced with a dilemma.
If the theory of performance must, in fact, be based on the theory of
competence, should he wait until the theory of competence is better
understood before proceeding with his research? It seems obvious to
this writer that the answer to this question must be a resounding no.
The studies done by these applied researchers are in response to needs
expressed by the educational community to which they can at least provide
partial solutions, although these solutions will need to be reevaluated
from time to time. The criticism of the applied researcher must not be
that he shouldforsake his research and do more theoretical studies, but
rather that he heed the advances that have been made in grammatical theory
when doing his studies.

The present study of the use of English by Chicano children is an

applied study which uses as a model the work of O'Donnell, Griffin, and

Norris entitled Szgtgx of Kindergarten ggd Elegantagy Sghool Children:
A‘Trgggfgggatiogal égalxsig. These investigators take the approach of the

applied researchers who attempt to place their work within the framework
of a grammatical theory, in this case, transformational grammar.

This study was used as a model in order to be able to compare the
Chicano children with native speakers of English. It seemed fruitless
to describe Chicano English in a vacuum, i.e., it would have helped little
more than previous studies of Chicano children which have noted that these

children have special problems with English. The important thing to

20

understand is how these children's problems differ from the problems of

other children of the same age. In order to have some comparative

base, then, the O'Donnell, Griffin, and Norris study was chosen. In this

study the authors measured the linguistic development of middle class

Anglo children from kindergarten through grade 7.

4.0 Summary and Evaluation of the O'Donnell Study

4.1 Objectives - The objectives of the study, essentially statistical

in nature, are given as follows (O'Donnell et al: p. l).

1.

3.

To find ways of quantfying differences in grammatical

structures used by children in various age/grade levels.

. To discover, if possible, the sequential acquisition of

syntactic structures.

To determine the pattern of growth, 1,3r, are there gradual
and consistent trends or periodic spurts.

To look for differences in the use of syntactic structures

in written and oral expression.

To note sex differences in the use of syntactic structures.
To discover a simple measure to use as an indicator of

syntactic control.

All of these objectives are easily understood except for the last one. As

wasnoted earlier in this chapter, the concept of 'syntactic control' is not

one readily acceptable to transformational linguists because it implies

some kind of psychological reality of rules or conscious control in the

generation of syntactic structures. In speaking of the development of

21

'syntactic control', it should be understood that the authors are referring
to the increased use from one age level to the next of complex syntactic
structures. A more specific definition of these complex syntactic structures
is given below.

4.2 Theoretical Background - The tenets of transformational grammar
are claimed as the theoretical underpinnings of this study. The authors
distinguish three types of transformational rules. The first type are
those which operate on elements within the sentence, e.g., those which
transform PAST TENSE + VERB into VERB + PAST TENSE. The second type are
optional rules which transform a simple, active, declarative, affirmative
sentence into a different sentence type. The third type of transformation
works on two underlying strings so as to join or embed one in the other.
These they refer to as sentence-combining transformations and it is this
third type of transformational rule which is investigated to meet the
stated objectives of their study (O'Donnell et al: p. 15).

According to the authors, sentence-combining transformational rules
make use of three types of operations: deletions, substitutions, and
expansions. The authors do not discuss these operations in detail and
give only example sentences showing the use of these subtypes in combining
two so-called kernel sentences such as 'The man was poor.‘ and 'The man
bought an automdbile.‘ into single sentences. According to their analysis)
an example of the use of the deletion operation would be: 'The poor man
bought an automobile.' Substitution would yield the sentence 'The man
who was poor bought an automobile.' Finally, an example of the expansion

operation would be 'The man bought an automobile though he was poor.‘

(O'Donnell et al: p. 16). These subtypes become important divisions in the

22

study when the authors attempt to show that the use of the deletion Operation
marks mature use of language.

It must be pointed out here that this summary of transformational
grammar presented by O'Donnell et a1 is an outdated version of the theory.
Since Chomsky's publication of Syntactic Structures in 1957, various
scholars have applied transformational theory to devising curricula for
teaching high school English grammar (Roberts), or improving the ability
of school children to use complex syntactic structures in their writing
(Mellon), or measuring the ability of children to use syntax (Hunt;
O'Donnell et a1). One of the obvious drawbacks of attempts to apply the
findings of transformational theory has been that the theory itself has
undergone basic revision since its inception, leaving many of the applied
researchers without firm theoretical support.

There are two major deficiencies in the O'Donnell study, however,
which cannot be easily excused on the basis that the theory changed while
these authors were in the middle of their study. First of all, despite
the fact that the O'Donnell study was published in 1967, they used
Sygtactic Structures for the theoretical background to their work
although they recognized that the theory had been revised in Aspects
of the Theogy of Syntax published in 1965 (O'Donnell et al: pp. 15-16).
Even this version of the theory has, of course, been subjected to revision.

The second deficiency in the overall approach used by O'Donnell et al
is that they tended to overlook many of the points that Chomsky and other
transformationalists have made about the nature of language and, in
particular, the nature of linguistics, including criticism of taxonomic

linguistics. Instead, they selected some parts of the theory which lent

23

some degree of explanation to their findings, ignoring such basic
notions as the distinction between competence and performance. The result
is that transformational grammar is used as a kind of veneer overlaid on

a basically structuralist or taxonomic study.

4.3 The Use of the T-Unit - One of the first problems which many
applied researchers encounter is determining what the boundary of a
sentence is. Most studies of child language, particularly those of
school age children, have noted that these children use an inordinate number
of coordinate sentences, usually with 'and' as the conjunction. Since many
of the studies were attempting to prove that sentence length was an
indicator of linguistic growth, this naturally led to problems. One obvious
problem is that sentences used by younger children were longer in terms of
word count than those used by older children but were not as complex in
grammatical structure. In other words, the younger children would often
string together short, simple sentences connected by 'and', whereas
older children used transformations to combine these sentences (O'Donnell
et al: pp. 21-22).

In order to solve this problem, Kellogg Hunt in his study of the
written work of school children introduced what he called the T-unit,
the minimal terminal syntactic unit. Hunt defined the T-unit as "one
main clause with all the subordinate clauses attached to it." (Hunt: 1965,
p. 201). In other words, simple and complex sentences can be T—units,
but compound sentences would be divided into two or more T-units. The
T-unit was considered an objective measure because there was a steady,

significant increase in the mean length of T-units from grade to grade

24

and the individual ranges showed less overlapping when compared either to
other indicators of maturity such as the subordination ratio used by
Loban (1963), that is the number of subordinate clauses used, or the
length of sentences previously used by Hunt (1964) (O'Donnell et al:

p. 23).

Although part of the increase in length of T-units could be explained
by an increase in the use of dependent clauses, Hunt also found that the
length of these clauses themselves increased as did the length of T-units
containing no dependent clauses. The cause of this increase, he hypothesized,
was the more frequent use of sentence-combining transformations (O'Donnell
et al: pp. 23-24). It is with this information as a starting point that
O'Donnell, Griffin, and Norris begin their study of sentence-combining
transformations, expecting to be able to further refine the notion of

linguistic maturity.

4.4 Data Collection Techniques - The data used in the O'Donnell
investigation were based on interviews of 180 white middle-class children
living in Murfreesboro, Tennessee; thirty children each in kindergarten
and in Grades 1, 2, 3, 5, and 7. The children were shown two short films
in which the sound was cut off to prevent bias. They were then asked
to relate the plot of the films to an interviewer and their speech was
transcribed. In addition, children in Grades 3, 5, and 7 were asked to
write compositions about the films. The transcriptions and compositions
were then divided into T-units and each T-unit was analyzed by assigning
a sentence pattern type to the main clause and then counting the number

of nominal, adverbial, and coordinate structures used in forming the

25

T-unit. Each of these three structural types was subdivided into a number
of subtypes. (See Appendix B for a sample worksheet which lists all the
categories counted.) A computer was then used to calculate a series of
statistics based on the analyzed data.(0'Donnell et al: pp. 29-36).

The data collection techniques, the granmatical analysis, and finally
the statistical analysis used in the O'Donnell study once again clearly
show how few of the tenets of transformational grammar actually were used
as basis for this study. The pitfalls of the data collection method and
of statistical studies have been discussed earlier in this chapter, but
the grammatical analysis, which is essentially structural in nature, has

not been discussed.

4.5 The Inadequacies of Structural Analyses - Applied researchers
are often structuralists in disguise, often disregarding the methods and
goals of transformational grammar. In the O'Donnell report, the analysis
of the data uses the methods of structural grammar and transformational
grammar is brought in only in an attempt to interpret the results of the
analysis.

A point at which transformational interpretations of structural
analysis led to misunderstandings was in the noun adjunct classification.
The authors contend that compound nouns are treated as structures produced
by sentence-combining transformations (O'Donnell et al: p. 35). They give
as an example 'North Wind', one of the characters in the film story they
used in the interviews. To treat all compound nouns as though they were
produced transformationally is to misunderstand the theory. Some, perhaps

many, that were included in their data are to be considered as single lexical

26

items. This particular misunderstanding caused a number of problems
in an attempt to compare the results of their study and the Chicano
English study.

Another goal of transformational grammar which seems to have been
overlooked is the attempt to show the relationship among syntactic structures
which may look very different in their surface structures but which are
in fact related at a deeper level. The authors, as we can see in examining
their structural classification scheme shown on the worksheet in Appendix B,
have looked only at the surface structures of the T-units. They come up
with such conclusions as: nouns modified by prepositional phrases,
participles, and participial phrases increase in usage with age/grade
level. In the light of two other conclusions, 1) that the use of relative
clauses decreases with age/grade level and 2) that the use of deletion
rules seems to indicate maturity, there seems to be some generalization
which could be made but which is not being made. If the authors were
concerned about showing the relationship among these various structures,
they might have had more evidence for the conclusion that the use of
deletion rules was a mark of maturity.

In addition to looking only at the surface structure of the data,
it appears that the authors also looked only at the "surface structure"
of the theory. They seem to be primarily interested in the three general
types of rules outlined: substitution, expansion, and deletion. They
recognize that some sentence-combining transformations are more complex
than others (O'Donnell et al: p. 50), but this does not seem to play an

important part either in the analysis of the data or in the interpretation

27

of the results. To state simply that deletion rules are used more often
by more mature speakers of the language is so general that it does not
really tell us very much. Chomsky has also pointed out this general
tendency to oversimplify the facts of linguistic structure (Chomsky:
1970, p. 47). These three types of rules are used in generating almost
all sentences and are present in very early stages in the development of
grammar (Menyuk: p. 1071.80 that the kinds of statements made about them
are not very useful. Besides, if the authors were attempting to arrive
at such conclusions, one would have expected them to design their study
in such a.way as to count the types of rules used in generating various
structures and to report their findings in this manner. Instead they
designed a structural study and appended the transformational analysis.

The futility of applying the methods of structural grammar in the
description, let alone its explanation, of grammar has been pointed out
in the past by such men as Chomsky and Lace. The infinitely creative
aspect of language provides evidence that the taxonomic views of linguistic
structure are inadequate to explain the knowledge a child has about his
language. Such knowledge cannot arise by application of step-by-step
inductive operations such as segmentation, classification, etc. (Chomsky:
1965, p. 57).

Classifying grammatical constructions does not necessarily lead to a
deeper understanding of that category. Transformationalists believe that
a Study of the underlying transformational history of a grammatical
constzruction is more likely to yield a correct grammatical statement than
15 CJLassifying such a construction according to some arbitrarily chosen

diffeerentia (Lees: p. 392)-

28

The failure of O'Donnell et al to examine the transformational
history and thereby to show the interrelatedness of the grammatical categories
chosen has been pointed out as well as the pitfalls of examining only
surface structure. Lees notes that using the techniques of structuralism,
two or more analyses can be made from a single unambiguous construction
such as:
The dog is barking.
This sentence may be analyzed either as NP AUX VP or NP COPULA ADJ
since the sentence patterns like
The dog is a friendly animal.
The latter is counter-intuitive, he points out, but still is possible using
these techniques (Lees: p. 386).
Chomsky's example sentences
I persuaded John to leave.
I expected John to leave.
rather clearly show how unrevealing structural analyses of surface structure

can be as to underlying deep structure (Chomsky: 1965, pp. 23-24).

4.6 Results of the O'Donnell Study - The authors make an initial
disclaimer about the results of their data, saying that the language used
by this particular set of children may not reflect the language used by
children elsewhere and that the special conditions under which data were
collected may in some way bias the type of data collected. Nevertheless,
the results do seem to agree faVorably, they say, with published results
ofcother investigations of children's use of language (O'Donnell et a1:
P- 37) . The results of their study are given here in term of the six

Objectives noted above (p. 20).

29

Three basic observations were made by the authors regarding the ques-
tion of discovering measurable differences in the use of grammatical
structures by children of various age/grade levels.

1. Little diversity was found in basic structure patterns of
main clauses. The use of the patterns Subject-Verb-Object
and Subject-Verb were most common; Subject-Verb-Object-
Object Complement and Adverbial-Verb-Subject were rare
(O'Donnell et al: p. 88). (A list of all the types of
main clause patterns counted is given in Appendix B.)

2. There was significant increase in the use of whole classes
of transformation-produced nominals and adverbials in both
speech and writing and in the use of coordination within
T-units in speech (O'Donnell et al: p. 89).

3. The greatest overall increases occurred in the use of
adverbial infinitives, sentence adverbials, coordinations
within T-units, and modifications of nouns by adjectives,
participles, and prepositional phrases. All of these
structures involve the application of deletion rules
(O'Donnell et al: p. 90).

Because no instances were found in which a particular structure was
absent in the speech of kindergarten children and present in the speech of
older children taken as groups, any findings about the sequence of the
acquisition of syntactic structures are inconclusive. However, certain
structures (nouns modified by a participle or participial phrase, the

Serund phrase, adverbial infinitives, sentence adverbials, coordinate

30

predicates and transformally-produced nominals functioning as objects of
prepositions) appeared only rarely among kindergarteners' speech but
were used from three to ten times more frequently by the oldest children
in the sample (O'Donnell et al: pp. 91-92).

Linguistic control, by which the authors mean the increased use of
complex syntactic structures, seems to be characterized by rapid growth
at particular stages rather than by gradual development. In speech, the
time between kindergarten and the end of the first grade and that between.
the end of fifth grade and the end of the seventh grade were observed to
be stages of rapid development in syntactic control (O'Donnell et al:
p. 93).

The children in grades 5 and 7 advanced in their control of syntax
in writing far beyond that reflected in their Speech (O'Donnell et al:
p. 95).

Generally there were no differences in syntactic control which could
be attributed to sex (O'Donnell et al: p. 96).

In looking for a simple, objective measure of the develOpment of
syntactic control in children, the authors note that the best measure
is to calculate, as they did in this study, the relative frequency of all
sentence-combining transformations, especially those involving deletion
transformations. An objection, however, can be made that this task is
far-from a simple one. A more readily performed calculation is to determine
the: mean length of T-units. This figure gives a close approximation to
the results of the more complicated accounting of sentence-combining
transformations. Increases in the length of T-units is reflective of the

varying degrees of expansions in the use of syntactic resources (O'Donnell

at al : pp. 97‘99).

31

An interesting further observation made by the authors is that the variety
of sentence patterns and the variety of constructions filling the pattern
slots were fairly restricted at all grade levels (O'Donnell et al: p. 100).

Assuming for the present that the goals of the investigators are
consistent with the goals of transformational grammar as the title of the
study would lead us to believe, the authors want to study the underlying
knowledge a speaker has of his language. Theoretical grammarians who are
studying child language acquisition, like Menyuk and McNeill, argue that
we must write grammars which are descriptive of what children know about
their language. If we were to do this for the children studied byO'Donnell
et a1, we would very likely find that the grammar written for the children
in kindergarten was not too different from the one written for seventh
graders - that is, given the analysis provided by the study. Using statis-
tical methods, these investigators were not able to show that any
grammatical structures were absent in the youngest speaker and present
in the oldest (O'Donnell et al: p. 21). Their research results, which
are questionable given the criticisms made here, indicate that the
competence of the children does not change after they reach kindergarten
age.

The results of the O'Donnell study can also be evaluated from a
different set of goals. As was discussed earlier in this chapter, the
goals of the applied researchers are often not clearly formulated; one
obvious difference, however, is that the applied researchers are far
more interested than are theoretical grammarians in actual use of

language - both spoken and written - and the ways in which the use of

32

language changes as the child matures. By putting aside for the moment
Chomsky's argument that we cannot arrive at meaningful conclusions about
a theory of performance until there is a fairly well-formed theory of
competence, the more limited goals of determining in which ways language
use changes with age can be evaluated. Whether these goals were attained
will be examined by looking more closely at the results of the comparison
of the written and spoken data.

Throughout the study the authors point out that the children exhibit
more control of syntax, that is, use more complex syntactic structures,
in written work than they do in speech (O'Donnell et al: p. 53). Other
investigators have also noticed that children have greater control of
syntax in writing than in speech (Harrell: p. 70). In fact, O'Donnell
et al state that "mastery of syntax in writing developed so rapidly in
upper grades that by the end of Grade 7 it outran the acquisition of such
control in speech" (O'Donnell et al: p. 57). They note the less frequent
use in written work of garbles (p. 40), sentence adverbials (p. 69), the
Subject-verb-Predicate Nominative clause type (p. 75), and the
Subject-Verb clause type (p. 74) as indicating more mature control of
language. Once more we have to consider what maturity means. If
:naturity is a feature of competence, that is, what the child knows about
his language, the fact that he uses a particular structure at all, whether
in speech or written work, would indicate that the rules for generating
that structure are in fact part of his gramar. If a child is counted
linguistically mature only when he uses them.in certain contexts, then it is
time contextual use of a language that is being measured and not knowledge

of a language.

33

It seems apparent that O'Donnell et al confused what may be stylistic
devices used in writing and those used in speech and because of the
assumptions they had made about what constituted maturity, i.e., the more
frequent use of complex grammatical structures, were forced to say that the
children controlled syntax more in writing than in speech. The conclusion
which should have been reached is simply that certain syntactic structures
are used more often in writing than in speech, that is, that they represent
differences in styles but do not necessarily represent differences in
ability to control syntax, except in some limited sense of the term 'control'.
The child can have more 'control' in the sense that he can spend time
contemplating about exactly how he wants to say something when writing a
composition, which is not usually the case with speech.

If we take this argument one step further, we may find some reasons to
conclude that the differences between the use of spoken language by
kindergarteners and seventh graders may also be viewed as a difference in
style. The argument here would be that as the child becomes more familiar
with adult style - through reading as well as having adults talk to him
in a more adult fashion - he changes his own style so that it becomes
more like the adults. The conclusion would be that competence does not
change much after a child reaches school age but his use - or style -
does change. Essentially the child has within his grammar a variety of
optional transformations and he learns to choose certain options as

preferable.

Chapter III.

A STUDY OF CHICANO ENGLISH

1.0 Introduction

The major portion of this thesis is concerned with investigating
'the English syntactic structures used by a particular group of Chicano
school children. The O'Donnell study was used as a model for both
data collection and analysis and the figures from this study have been
adopted as a base for comparing the Chicano children with Anglo children.
Whether the two sets of data are really comparable is difficult to ascertain.
In the first place, the size of the samples differs: O'Donnell et al used
30 fifth grade children, 16 boys and 14 girls ranging in age from 10:2 to
11:8. The sample of Chicano children used nine children, five boys and
four girls with ages ranging from 9 - 12, most of whom were. fifth graders,
but one was in the fourth grade and one was in the sixth grade. Another
problem of comparability might be that the film used to stimulate discussion
was different for this study. Yet another difficulty comes from the fact
that in attempting to analyze the data following O'Donnell et al, there
may have been instances where the interpretation of the Chicano data differed
from what theirs would have been because the grammatical analysis they used

- basically a structural analysis - often allows for more than one

34

35

interpretation of the surface structure of a sentence. (cf. discussion
of the criticism of this kind of analysis by Lees and Chomsky, Chapter II,
pp. 25-28.)

The theoretical weaknesses of the O'Donnell study have been pointed
out in Chapter II; the present study is obviously inadequate in the same
way. The import of these weaknesses became increasingly apparent as the
analysis of the data progressed. Although the value of the present study
is limited by the theoretical shortcomings, it can at least give some
idea of where Chicano children stand relative to middle class Anglo
children and it can also point out those areas of widely divergent uses
of various English syntactic structures.

In order to get closer to an understanding of the competence of the
Chicano children than the methods of the O'Donnell model allowed, a
sentence repetition exercise was devised. Throughout the data collection
phase of the study, certain nonstandard grammatical structures were
observed. (A list of these nonstandard structures is given in Appendix C.)
The sentence repetition exercise was designed to test whether these 9;
fairly common nonstandard structures were systematic and possibly might
form a dialect of English which could be described as Chicano English.

This chapter describes the background of the children used in this
study, the results of the comparative study of the Chicano and Anglo
children, and the results of the sentence repetition exercise. Throughout
this chapter, CE will be used to refer to the English used by the Chicano
children; MCE refers to the English used by the middle class Anglo

children from the O'Donnell study.

36
2.0 Background to the Present Study

2.1 The Children - The children used in the sample were all students
in the fourwweek summer session of the Nottingham School in the La Mirada
School District in Norwalk, California. The school is in a Chicano
neighborhood and the student body is largely Chicano. Nottingham School
is one of two Norwalk schools chosen to participate in a federally Sponsored
program relating specifically to Mexican Americans.

Table 1 gives some information about the children. Spanish is the
home language of all the children in the sample except for one, Ceferino.
The table also shows that all except one student, Laura, claimed to use
English in speaking to their friends. This is perhaps not a realistic
picture; during the course of this study these children were often observed
speaking to each other in Spanish. It seems likely that they speak SpaniSh
more often than they are aware of. VHowever, while the boys played team
sports, they were observed using English which may indicate that in groups a?

larger than one or two intimate friends, English is the language most

gene rally useda

r
2.2 The Types of Data Collected - Several types of data were
collected during the four-week summer term. Much of the time was spent
in getting the children to feel comfortable in the interview situation,
tfliam is in talking with the interviewer and in using the tape recorder.
As Labov points out in his article "The Logic of Nonstandard English",
many mistaken notions have been made about the linguistic ability of

Bla<ikLchildren who were placed in threatening interview situations because

37

 

 

 

 

 

 

 

 

 

 

 

 

Table 1.
Background Information on the
Children in the Study Group

Name Age/ Birthplace Time of Home Playground

Grade Arrival in Language Language

U.S.

Ana 9/4th Mexico Kindergarten Spanish English
Kathy 9/4th California Spanish English
Becky lOISth California Spanish English
Robert 10/5th U.S. Spanish English
Eddie lO/Sth Cuba 3rd Grade Spanish English
David lO/Sth Texas Spanish English
Rene ll/6th Mexico Kindergarten Spanish English
Laura 12/5th Mexico 3rd Grade Spanish Spanish
Ceferino 10/5th U.S. English English

 

 

 

 

 

 

 

38

the speech samples collected did not reflect reasonable normal speech

‘.

‘av
«‘..

habits (Labov: 1970: pp. 157-163). To avoid this, the interviewer made (y
it a habit, during the four-week session, to arrive on the school ground
during recess time to talk informally with the children. The interviews
were conducted outside sitting under a tree with groups of four or five
children. Some time was spent playing with the tape recorder and listening
to the playback. It was not until the fourth week that the major
sources of data for this study were collected and at that time the inter-
viewer worked only with individuals. 373'

Conversations and folk tales as told by the children were recorded
and the nonstandard utterances used formed the base for the sentence
repetition exercise.

Another source of data, and the one used in comparing these students
with O'Donnell's students, was the retelling of the plot of a film, v;
The Little Mariner, which was shown to the children. The film, which lasted.»
about twenty minutes, had only musical background. There was no narrative
to influence the children's retelling. In addition, the children were
asked to write a paragraph retelling the plot of the film.

The data from the film accounts were transcribed, divided into T-units,
and analyzed according to O'Donnell"s fairly detailed techniques.
Appendix A contains the individual and group figures resulting from this
analysis. O'Donnell et al report their statistics according to sex as
well as giving the combined Statistics. They found little evidence of
Significant differences in linguistic development based on sex and
neither did this investigator in the Chicano data; therefore, the figures

rePorted will include both boys and girls.

39

3.0 A Comparison of the Syntactic Structures Used by GE and MCE Speakers

3.1 Overall Results of the Study - Table 2 shows the means and
individual ranges in word length of the total response in both speech and
writing. The length of response here, which shows a much lower rate for
Chicanos, is probably not indicative of a lack of verbal response on the
part of the Chicanos but more likely is situational. The O'Donnell
children had viewed two films and probably had more to talk about. It is
given here only to compare the amount of data analyzed.

It should be pointed out that since all statistics were calculated
on the base of rate per hundred T-units, the figures for both groups should
be roughly comparable.

As was noted earlier in the summary of the conclusions drawn from
the O'Donnell study, and as had previously been reported by Hunt in his
studies of the written work of school children, a simple, objective
measure of linguistic maturity is thought to be the number of words per
T-unit. Hunt found positive correlations between the increase in the
number of words in a T-unit and the advancing age of the child (Hunt: 1965,
p. 141). Loban (p. 64) has also observed that the average number of
xaords in a communication unit, a base, comparable to the T-unit, which he
uses, is a significant measure of maturity. He noted that a high average
‘words-per-unit figure correlated highly with increased complexity of
sentence structure.

All of the O'Donnell data are reproduced here in Table 3 in order to
ShOW’hOW the Chicano children compare. If we can use this figure as a

“Ea-sure of syntactic control, as O'Donnell et al suggest, then we can say

40

Table 2.

Length 0 f Responses

 

 

 

 

CE MCE (p. 43)
11.--,“ 13;}: Range, M ‘ - _ Mean Range
Speech 357.5 271-449 656.0 254-1160

 

 

Writing 123.6 76-152 % 387.5 237-861

 

 

 

 

 

 

weapons

41

 

 

 

 

 

 

 

 

 

 

m.ml~.\. ed
m.o|m.o w.“ nuoomm
omsmm some.“
mo
mg." 0.: 9:
um.“ mm.m lie 31¢ and $4. memory:
5.3 Q: w.OH w.m HA: m5
11w w.m no.5 m.w leg 26 Inc mm.w I~.m no.5 10$ no.5 museum
owned coo: ow com on a: om mom new: om new so on ow com and: am one sees
m N H m
m2,“

 

upmuo .3 35:9 use mouoz mo moansz smmz

.m magma

 

42

that the Chicano children are not nearly as mature in their control of 3
English syntax as are the middle class children whose native language is
English. The CE mean coincides with the MCE mean number reported for
children in Grade 1 and the individual range of the CE mean numbers

indicates that even the most mature of the Chicano children does not equal \

‘9‘,

the most mature first grader.

The figure for the written data shows the Chicano children to be
closer to their middle class counterparts in that the CE mean falls .3
somewhere between the MCE third and fifth grades. Again the most mature
Chicano does not equal the most mature informant in O'Donnell's group. It
is not unexpected that the syntactic control exercised in written work is
greater than that in Speech. O'Donnell et al also noted that,other than
third grade where the children are just learning to write, children in the?
upper grades "control their writing more strictly than their speech"
(O'Donnell et al: p. 46). An interesting point made by O'Donnell is
that the writing ability of third graders shows less syntactic control
than first grade speech. The Chicano evidence Shows that although
their spoken ability falls within the first grade range, their written
ability is much closer to their middle class peers.

The Chicano evidence lends weight to the argument presented in
Chapter II concerning written vs. Spoken 'control' of language.‘ If by
'control' we mean, as O'Donnell et a1 did, the increased use of complex
syntactic structures, then it seems reasonable to assume that the kind of
conscious generation of sentences, which may include editing and revision,

will allow for the use of more complex structures. However, it must

43

also be remembered that the differences that one intuitively knows exist
between written and spoken language may also be a difference in the kind
of syntactic structures apprOpriate to either written or spoken language.
This study - and the O'Donnell one - indicate that a greater number of
complex structures are used in writing, but do not provide clear evidence
for an adequate description of these differences.

Additional data are given in Table 4 on the comparison of T-unit
length reported by other researchers. Loban's data were collected in
Oakland, California, from the Speech of Black children and White children.
He divided his data among four groups: a high and low group of Caucasians
(based on performance), a low Negro group, and a random group selected
from among the three other groups. Hunt's data were collected from the
written work of middle class children in Tallahassee, Florida. Loban
studied the same children over a number of years whereas Hunt, like
O'Donnell et a1, studied different children from various grades.

In comparing Laban's high Caucasian group with the Chicanos, the results
agree fairly well with O'Donnell"s, i.e., the Chicanos on this scale
fall somewhere between the second and third grade. Interestingly, by
comparing the Chicano children with the children used in the Loban study,
the Chicano children would be considered linguistically more mature in
their use of English syntax than either the low Caucasian or the Negro
group. The fourth group, the one chosen randomly, shows that the Chicanos
are not on the same level as their peers, though not quite as far behind
as they are when compared to O'Donnell"s middle class group. It must be
remembered that Loban's figure included many so—called 'disadvantaged'

children.

44

 

 

 

 

 

m.HH m
mm.m Ho.m om.m mm.oa a
ma.w mm.a «0.x oa.m o
s.m m.a OH.m mm.o mN.e sw.m m
o.w mm.k se.e me.e me.w s
mm.e om.m oh.o mo.m m
sm.e em.m No.e «0.x N
No.e ww.s mm.m Ho.a a
Hm.m m~.s mo.a ew.m a

9.6.5 ouwoz usso sumo

sesame sou sob swam memos
has .a .meaev
mo scan as Ame .ev canon
uzeeeez seesaw

 

 

83.3980 <

uuficala won menoz no women:

.e man—ma.

 

 

45

Again the comparison with Hunt's data indicates that the Chicano 3;
children are closer to their middle class peers in writing than they are
in speech.

Another measure of linguistic maturity used by O'Donnell and Hunt
is the percentage of T-units containing less than nine words. They found
that this percentage decreases with advance in age/grade. Table 5 compares
the O'Donnell and the Chicano data. The comparison further confirms
the notion that in speech the Chicanos fall within the first grade range,
but that they are much closer to their middle-class peers in their ability
to manipulate English syntax in writing, where they fall somewhere between
the third and fifth grade levels.

Counting the number of words per T-unit is obviously only looking
at the very surface of the data. As Loban and O'Donnell point out, this
figure is significant as an indicator of linguistic maturity only because
it correlates highly with more interesting features of maturity, i.e., the
increased use of syntactic structures of greater complexity (Loban, p. 64;
O'Donnel et al: p- 44, 46).

O'Donnell et al correlated the word-count per T-unit figure with a
series of other statistics in an effort to delve more deeply into the
syntactic structure of the T-units. Their purpose was to understand
more objectively just what linguistic maturity involved. First of all
they reported the mean number and individual range in the mean number of
transformations per T-unit, which was determined by counting the number
of structures in the T-unit which were formed by what O'Donnell et a1

call sentence-combining transformations (cf. Chapter II, pp. 21-22, for

46

 

 

 

 

 

 

m.wm om.as me.Hm ee.mo masses:
mo.wo -.sm wa.oo Hm.oo NN.SS om.ao aa.ma summam
a m m N a a
mu Ame .ev mo:

 

 

mono? mcwz none

II.“ II I.].|.JIIII...I¢I|!II'.(.§I.I:..! .5 . t u

mom; msw>mz muwsbnh mo smousouuom .m manna

 

47

further discussion). For example, the T-unit:

'He got in the big boat and sailed to the beach and told the man

something.'
contains three sentence-combining transformations: a nominal structure
(N Adj), 'the big boat', and two coordinate verb structures, 'sailed to
the beach' and 'told the man something'.

Table 6 illustrates the first occurrence of a discrepancy in the CE
data. The CE Speakers are clearly at the same level of maturity as their g;
peers in speech and in writing using the figure for mean number of
sentence-combining transformations per T-unit. It is interesting to note
that the highest individual mean number of sentence-combining transforma-
tions in written work among the Chicanos would indicate that this individual
is at a higher level of maturity than any of his HCE peers and even higher
than any individual in Grade 7 as reported by O'Donnell (2.7, O'Donnell
et al: p. 51). This fact would certainly not be expected if we assume
the number of words-per-T-unit figure to be a valid and objective measure
of maturity. We would expect to find the CE Speakers falling somewhere
between the first and second grade MCE level.

A perhaps more revealing figure might be to determine the percentage
of the total number of T-units which contained no sentence-combining
transformations. Unfortunately, O'Donnell et al did not give this
figure for their data, but the evidence taken from comparing individuals
within the Chicano group indicates that this figure may be useful. These

percentages are shown in Table 7. Once again the greater degree of

48

Table 6. Mean Number of Sentence-Combining
Transformations per T-Unit

 

 

 

 

CE HCE (p. 51) *
Than Individual Than Individual
Range Range
Speech 1.08 .7 - 1.7 1.05 .6 - 1.7
Writing 1.6 .8 - 3.0 1.41 .7 - 2.5

 

 

 

* Unless otherwise indicated, all figures are based on rate per
100 T-units; the MCE figures are those for the fifth grade.

 

49

 

 

 

 

Table 7. Percentage of T-Units Having No
Sentence-Combining Transformations
(CE Only)

Name Speech Writing
Ceferino 28% 33%
Eddie 31 13
Ana 32 30
Laura 34.3 -
RObert 41.6 -
Becky 45.7 28
Kathy 54.5 60
Rene 54.8 35
David 56 -

Mean: 42 33
Range in Mean: 28-56 13-60

 

 

 

50

written maturity over spoken maturity can be seen. Also the individual
range supports both subjective judgements made by this investigator

and other objective data, e.g., Ceferino and Eddie are more mature than
David, Rene, and Kathy by other measures. (See pp. 66-69 for more
discussion of individual differences within the group. Appendix A gives

a detailed accounting of these variations.)

3.2 Results of the Study of Specific Construction Types - In further
breaking the sentence-combining transformation category down into sub-
categories, that is, into nominal, adverbial, and coordinate structures
which result from these kinds of transformations, the CE speakers are
shown to be more advanced in their use of written and Spoken use of nominal
constructions than the HCE group and again more mature than even the
oldest Grade 6 informant. (See Table 8.) This discrepancy will be
discussed further when the topic of nominal structures is presented. In
the use of adverbial constructions in speech, the CE speakers fall at
about an HCE Grade 2 level (11.77; O'Donnell et al: p. 56), while in
written work, they fall below the MCE Grade 3 level (13.01; O'Donnell
et al: p. 56), the lowest grade at which written work was analyzed. The
use of coordinate constructions shows an even wider gap between CE and
MCE Speakers. Here the speech of the Chicano group falls below the MCE
kindergarten group (14.67; O'Donnell et al: p. 56) and the writing
once again falls below the MOE Grade 3 level (20.35; O'Donnell et al:

p. 56). In all cases except one, the use of adverbial constructions,

the written skills appear to be more deve10ped than Speech.

51

Table 8. Types of Constructions Formed by
Sentence-Combining Transformations

 

Type of Construction CE MCE
(p. 56)

 

 

~——_--_._—.

 

 

NOMINAL
Speech 80.2 62.27
Writing 85.4 76.43
ADVERBIAL
Speech 12.8 14.89
Writing 9.0 19.03

 

COORDINATE
Speech 12.8 28.03

Writing 17 44.61

 

 

 

 

52

3.2.1 Nominal Constructions - Table 9 lists the headed subtypes
of nominal constructions formed by sentence-combining transformations.
Whether the nominal constructions counted by O'Donnell would be considered
to be produced by transformations by current transformationalists is
open to question. In order to obtain comparative results, however,
they were counted in the CE data as though they were. But, it is at
this point in the analysis that gross inconsistencies appear. The
figures show the CE Speakers to be far more mature in almost all categories /?
than the more general statistics which have been presented. When the
Simple, objective measure proposed by O'Donnell et al was used, that
is, the mean number of words per T-unit, CE Speakers were judged to be
at the first grade level. How then can we explain the difference between
the high degree of correlation between this Simple measure and the more
complex analysis of syntactic structures reported by O'Donnell (and
previously reported by Loban and Hunt) and the fact that this correlation,
at least in nominal structures, does not exist in the Chicano English
data?

One reason is that there were differences in the way the data were
analyzed and this becomes increasingly apparent as the study becomes
more detailed. The reasons for the possibility of arriving at different
analyses of identical structures has been discussed in Chapter 11 (cf.
Lees' and Chomsky's discussion of the deficiencies of structural analyses,
pp. 27-28).

Other factors may also have contributed to the discrepancies: it

may be that the data-collecting situations were sufficiently different to

53

 

 

 

Hm.m mw.H H.H m.H
no.0 no.0 H.H e.o
oo.m m~.m a m.e
am.m em.m m.s e.q
He.H~ ma.mo o.ma e.ma
ms.oa m~.m H.e~ e.oH
mm.ma ee.m m.ws m.ea

wannabe eommdm wannabe sommdm

Rom .mv me: so

 

 

.EH: passes
msfieou umon wan menu comm on soaH

 

oEHfi afloumu Ou CNS mfiu U53 uecfifiv ME

 

.xmop ecu mo no» so ow use om

 

.mumoo
deudmw mm: oozlmoe mficu ou uses m:

 

.umoo ma: ou wagons» uso3 o:

 

AN mm pouGSOUV .zmww1wao oBu mm: o:

 

.mcoam memo umooaflmm wan <

Auu Eouhv
mud $.2va

 

spouse
Hmadnuwuume
\oamaowuusm

ommusm
o>HufisHucH

ensues
HchHuwmoeoum

omsoao
o>aumamz

u>wuacwo
m>auumfie<

z

+

z

mZOHHUDMHmZCU
A< nude 2

 

we reason msowuosuumsoo HmswEoz poems:

mcoaumEuowmcme wsficfiosouioocoucom

.m pansy

 

54

call forth more of certain subtypes of constructions in the Chicano English
case than in the MCE case. Again, it may be that the high correlation
found between the mean number of words per T-unit and Specific syntactic
structures as reported by O'Donnell et al was a random feature of

their study which is not applicable to other studies.

The first explanation would help in understanding the discrepancies
apparent between the figures obtained for the N + N and the N + Adjective
subtypes. The figure would lead one to assume that the CE Speakers
were far more mature than the MCE Speakers. But the example given by
O'Donnell for N + N constructions was 'North Wind', one of the characters
in the film they used entitled "The North Wind and the Sun". Furthermore,
words were counted according to the number of bases involved, e.g.,
'snowball' was counted as two words in the O'Donnell data. That 'North
Wind' or 'snowball' is derived from a sentence-combining transformation
is not at all clear. The N + Adjective example given by O'Donnell,

'cold rain', is more clearly acceptable as having been derived by such

transformations. There is every reason, however, to regard words such

as 'popcorn', 'hot dog', and 'North Wind' as being single lexical items
(cf. Chapter II, pp. 25-26).

If the analysis implicit in these examples was to be followed, it
seemed obligatory to include such expressions from the Chicano English
data as 'hot dog', 'popcorn', and 'Long Beach' in either one of these
two categories in order to arrive at comparable figures. It may be that
the authors would not have analyzed them in this way, but the single
example 'North Wind' given by them does not Seem to allow any other

interpretation for these items.

55

The second explanation, that the Situation called for the use of
more N + N or N + Adjective units and that it is not at all a measure
of maturity but a situational feature, might then explain the higher
figures obtained for the Chicano data. It may be that expressions such
as 'hot dog' or 'popcorn' were used more because they were part of the
plot of the film and the films used by O'Donnell et al simply did not
have as objects as many expressions which might be classified as nominal
constructions formed by sentence-combining transformations but simply as
single word lexical items.

The fact that the figures for the two subtypes discussed are
inordinately high has a cumulative effect on the rest of the figures.
That is, it causes the figure for the number of sentence-combining
transformations in nominal constructions to be much higher than would
be expected. Additionally, it affects the mean number of sentence-
combining transformations per T-unit which indicated that the Chicano
children were at about the same level of maturity as Anglo children.

As has been noted, this would not be expected to be the case given the
mean words per T-unit figure.

Interesting too are the figures in Table 9 for the subtypes
N + Participle or Participial Phrase and N + Prepositional Phrase.
O'Donnell et a1 concluded that these subtypes of nominal constructions,
along with N + Adjective, showed the greatest overall increase with
age/grade. In truth there is a great deal of fluctuation in the spoken
data given for the N + Prepositional Phrase subtype. (O'Donnell et al,

p. 59, report the following figures for this subtype: K: 3.90,

56

Grade 1: 2.97; 2: 4.33; 3: 3.57; 5: 5.27; 7: 7.31.) A comparison

of the kindergarten and grade 7 figures does Show increase in frequency
of use. However, the variations for the other grades reported does not
allow the Chicano English data to be assigned a level. The subtype

N + Participle or Participial Phrase shows the CE Speakers to be at the
5th grade level in Speech but falling below even the third grade level
in writing.

The fact that the CE figure for the use of relative clauses is
higher than the Anglo figure is consistent with the findings of the
O'Donnell study. They found a higher incidence of the use of relative
clauses among kindergarten children than among children in all the other
grades studied.

The non-headed nominal constructions in the O'Donnell data Show
little overall increase in frequency of usage with age/grade in speech.
As can be seen in Table 10, the figures for the CE speakers once again
do not show what would have been predicted given the mean words per T-unit
measure. Again the reasons may well be inconsistencies in the analysis.
The use of gerund phrases, for instance,is so much higher than any of
the figures reported by O'Donnell for any grade level that one can only
suspect a difference in the criteria used to determine gerund phrases.

The one consistency which does come out in comparing both sets of€%'
data is that all of the children exhibit more mature syntactic control
in their writing than in their speech. Disregarding the actual figures
for the moment, this trend becomes apparent when the figures for speech

and those for writing are compared and the increase in the mean number of

57

 

Hm.H um. ‘ m.q q.N .ufi mofiumo cmuumum mooowcwm one

 

xoum‘ou hon
Ho.o mo.q I ~.o ozu uama u.oowv mumamo Ono . . .

 

 

mm.~ mm.~ e.mH a.a .xums as sow cu emauu a:
on; SA a; m; .m.m.3_ .ula .05.: e835: um
Ame.av Ase.av
masons: seesaw massage soomam Ame sense
mo: mo mm4a2<xm

 

 

 

ommuzm cashew

uoofinsm
an“: m>auscsmcH

unsung
m>auacseco

omsmao coon

monHUDmHmzoo
A<szcz

 

assaquuomoeuue measfinaoo:ouoousom
an vuauom neowuusuumcou HooHEoz novmomlooz .OH magma

58

times a particular nominal construction is used per one hundred T-units
is noted. For further discussion of the differences between the results
of the written and Spoken data,see Chapter II, pp. 32-33.

Table 11 displays the grammatical functions of those nominal
constructions which have been reported as being produced by sentence-
combining transformations. The figures for the Chicano children are .fi
higher, which is to be expected because they are based on information
given in Tables 9 and 10. O'Donnell et a1 point out that the most
important feature observed here is the prominence of nominal constructions
used as direct objects (O'Donnell et al: p. 66). This is certainly
clear in the Chicano data as well. The only other feature which showed
steady growth was the use of nominal constructions as objects of preposi-
tions. Again, the trend in the Chicano data, if not actual numbers,

coincides with the O'Donnell data.

3.2.2 Adverbial Constructions - Three subcategories of adverbial
constructions are distinguished by O'Donnell et al. Table 12 shows
the figures for these categories.

Sentence adverbials are of two types: 1) interjected clauses
such as 'I think' or 'you know' and 2) absolute constructions as in the
sentence 'The rain over, we went outdoors'. As might intuitively be
expected, the interjected clause is used more often in speech than in
writing; in the case of the Chicano group, it was used only in Speech.
There were no cases of absolute constructions in the Chicano data in

either speech or writing.

59

m.s~ swap eases magmas was” .maeeo oneness.

 

 

 

NN.N mm.H m.HH H.N
mN.mH em.ms «m.oq H.~N
«N. NN. N.N H.m
mo. so. H.H a.o
me. mm.H . «.0
He.Hs Hw.sm o.m~ s.om
Hw.HH Hm.m m.H~ N.oa
Ane.av Ase.av

means“: sumoam mcaufiuz euooam

mo: so

 

 

Aommunn m>wufisfimcflv
.Mflda ma: nuw3.Nwwm;mw ucm3 o:

 

A0>Huflcou + 2V
.cflmwm umon was cw omawmm oz

 

Am>auumhe< + 2v sass um>o
can .umon was we» .umop use» was

33an v5.33 .923 now o.u_
Ba: .39» cu mcamu Sun: 3mm 55 9S.

Ammmucm
Hmcowufimoeoud pmnmosoozv
.mmo m usonm mos Eafim och

 

Am>auuuhu< + 2 can m>auacoo + zv
.umop odwmwa mam new on

 

Auumfinsm mm omamHo

u>aomaum + 2 was m>auuono< + 2V
.wuoa mumon was mswmmuonmidsoummofisou

mm3 own» on mauufia m was muonu mono

 

Amu Ecumv
mmgmz<xm

 

m>fiumaw502
Haasum>e<

coaufimommum
so uumflsc
o>HuHmooa<
ucmEmHmEoo

o>suumfino

ucoEmHeEou
m>auoofiasm

uuoflnc bemoan

uuofisam

ZOHHUZDh
120 Hhsezmm

 

maoauosuumeou Hmcasoz mo macauucsm Hmowumaawuu

.HH canoe

 

60

 

nw.H mm.H

Hq.H mo.~

mo.mH mo.OH

80.3 83:
mofiufiuz seesaw

w.o

masons:

:ommam

m.w

 

 

mo:

 

mu

 

 

.opoummm,nwm uuwusnamz
meow uow,ou woodcum m: cmnu wo<

 

.wcHmeB mmz .Bocx DON .uoHHmm mflLH

.wuonu
unwfiu won msu mo mmB uw .ousuuw
osu scamsfi voxoodlwonu c053 oozH

 

Amp Ecumv
mmqu<xm

 

mm>aoacaecH
Hawsuo>u<

mamsnuo>w<
mucousmm

momsoau
Hmwpum>v<

mmeHUDmHm
A<Hmzm>m<

 

mcofiumEMOMmcmue wchHnEooloocmuomm

An weapon mousuosuum Hownuo>w< .NH manna

 

61

Adverbial infinitives, as in the sentence 'To drive well, you
must be alert', were used infrequently by the Chicano children in speech
and not at all in writing. Nevertheless, the frequency of use is comparable
to the MCE fifth grade figure. This particular grammatical item, along
with sentence adverbials, is not used frequently in either speech or
writing and shows little increase in frequency with age/grade level
(O'Donnell et al: p. 68). Neither category would seem to be useful, then,
in determining maturity in the use of language.

The study of the use of adverbial clauses, on the other hand, shows
that MCE seventh graders use these grammatical structures almost twice
as often as kindergarten children, as reported by O'Donnell (p. 68).
Adverbial clauses, then would seem to be a good syntactic construction
to use to compare the development of control evidenced by Chicano and
Anglo children. The Chicanos fall between the second and third grade
level in comparison to the Anglo children in Speech (2: 7.85; 7: 10.12;
p. 68) and below the MCE third grade level in writing (3: 8.93; p. 68). 7%
It is also noteworthy that the trend which has been observed with most
of the data reported, that is, the more frequent use of the constructions
formed by sentence-combining transformations in writing than in speech,

does not hold true for the Chicano children in this case.

3.2.3 Coordinate Constructions — The coordination accounted for
here involves only that between nominals, modifiers, and predicates, i.e.,
only coordination within T-units. These structures are particularly
interesting to use as a basis of comparing the performance of the

Chicano and Anglo children. First of all there is very little possibility

62

of inconsistencies between investigators' analyses because the coordinate
surface structures are easily identifiable. Secondly, there was a
significant increase in the use of these structures in Speech from
kindergarten to seventh grade as reported in O'Donnell et al. Thirdly,

and perhaps most significantly, Hunt (1964, p. 81) has noted that the

use of coordination within T-units requires considerable ability to control
syntactic structure. If this is so, then a comparison of the use of these
structures might better show the level of maturity of the speaker.

Table 13 shows the rates of occurrence for both groups. The Chicano
3%

a

children fall below the MCE kindergarten level for both coordinate
nominals and modifiers in speech and at the kindergarten level in their
use of coordinate predicates. In writing, the Chicano children's
performance falls below the third grade in their use of coordinate
nominals and at about the third grade level when comparing the figure
obtained for coordinate predicates (O'Donnell et al: p. 70). There

were no instances of coordinate modifiers in the written Chicano data.

3.3 Structural Patterns of Main Clauses - Several interesting
observations result from a statistical count of the frequency pattern
of the structure of main clauses. O'Donnell reported that Subject-Verb
and Subject-Verb-Object patterns appeared in 80% of the T-units used by
kindergarten, first, and second graders and in 85% of the T-units of
the third, fifth, and seventh graders in both speech and writing (p. 74).
Table 14 gives a comparison of these figures, including example sentences,
indicating that these two patterns account for about 83% of the main

clause types used by the Chicano children.

63

 

m¢.mm Hw.mH

mm.m no.m

mm.“ mm.o

Aca.av Aca.av
mafia; nausea.

m.qH

massage

5.0

m.m

seesaw

 

 

NU:

 

 

.580 93 cu

 

 

. «no can no muocuou on» : souzu
.u_so 1mm museums“. o5 gov HAM. ass: on

 

.mumon

Hafiam one amanm was moons 3mm 0:

Ann sonny
mmqmz<xm

moumoaomum

€333“

mamoweoz

mmMDHUDMHm
MH<ZHQKOOU

 

meoaumEuommemuH wofioaneoolousuuoem

up weapon mousuusuum oumcavuoou .ma macaw

 

64

 

mq.H mo.~
Hm.n mm.m
«w. Hm.H

co.aq o~.N¢

nm.mm ow.mq

Ama.av Awa.av
masons: nuooam

o.m H.N
¢.m m.q
m m.~

m.m~ m.~m

m.nm o.me

massage zoomam

.waaamww owe meme muo3 mumnfi

.ooumom uow on .aamudmu
m was on son on cons noumo saga

.mcom ma mama m2
.umop masons a can son <
.mumnsoo

mo uso memo umon msoauwumms <

Amo Ecumv

 

 

mu:

 

mu

mamz<xm

 

 

goofinsm
Inuo>lmuonh

m>suomfie<
oumowvmum

unuo>uuomhs=m
Hands 2
oumuapmum

unuo>uuoonnam
uuufioo

unuo>uooofin=m

num>nuumhnsm

ZKMHH<A
A<MDHUDMHw

 

woodman can:

mo monouumm amusuusuum .QH nanny

 

65

One of the more interesting clause types for purposes of assessing
maturity in the use of syntactic structures may be the Subject-Verb-
Predicate Nominal type. O'Donnell et a1 noted that it was used more
frequently by kindergarten children and its use decreased with age in
both speech and writing. They also observed that this type is used less
frequently in writing than in speech, thereby confirming their hypothesis
that children mature more rapidly in writing than in speech (O'Donnell
et al: p. 75). The Chicano children use this type of main clause
structure somewhat less often than kindergarteners but more than the second
graders in the O'Donnell sample of speech patterns (K: 5.07; 2: 2.19;

p. 72).

The main clause pattern 'There-Verb-Subject' shows that the Chicano
children use this structure as often as their Anglo peers in speech, but
use it much more frequently in writing. O'Donnell et a1, as well as
other investigators,such as Loban and Strickland, noted that this pattern
was used infrequently by most children at all grade levels. Riling,
however, observed it fairly often and suggested that regional language
habits might account for the discrepancies (O'Donnell et a1: p. 95).
While this might account for the Chicano data as well, little can be

concluded on the basis of such scanty evidence.

3.4 Predictive Ability of the Number of Words per T-Unit Measure -
O'Donnell et a1 claim that they have discovered a simple, objective
measurement of linguistic maturity: the mean number of words per T-unit.
This figure "reflects the varying degrees of expansions in the exploitation

of syntactic resources" (O'Donnell et al: p. 98). Presumably this

66

measure could be used to determine the standing of a particular individual
within a group. Let us examine the predictive capability of this

measure by looking at the performance in speech and writing of the nine
children studied in the Chicano group.

Table 15 ranks the children according to the mean words per T—unit
score in Spoken English. The children are then divided into three
groups: the first is a grouping of children scoring 9 words per T-unit
or above, the second 7 to 9, and the third 6 to 7. Five other variables
were then correlated with this figure by ranking the student's performance
relative to each other. The variables were: (1), the mean number of
sentence-combining transformations per T-unit; (2), the number of
sentences containing no sentence—combining transformations; (3) , (4) ,
and (5), the mean number of nominal, adverbial, and coordinate structures
produced by sentence-combining transformations per T-unit. Within the
rankings, the first two students were said to be in the first class,
the next five in the second class, and the last two in the third class.

A figure was then obtained as to whether the children fell above (A),
below (B), or within (8) the same class as that determined by the mean
words per T-unit score. As the table shows, only in the case of Becky
was the measure possibly predictive; it might even be considered somewhat
predictive for Eddie and Ana, but in three cases it definitely did not
predict rank within the other variables (Laura, Ceferino, and Rene), and
in the remaining three cases there seems to be a random correlation
between the measure and the other variables. Therefore, we can conclude

that the measure does not predict for speech how the individuals within

 

67

 

 

 

 

 

 

 

 

 

 

<H . <N <N <N U <m m _
as an , mm mm r mm 5
”ml. ”up. ”my. .nwl ”Hr. i
n < < < < M m.e ocaumemo
n h u < < < M m.o snows
_ W HHH mmeao
- . u m m m m H.k ua>mc
n
I M < ..I... I I y m 0N “2
_ .
a W a m m m «.5 made
< I I I < m oh ““050“
I I I I I nth shawuom
HH mm<ao
m m m m e m.a manna
I m I I I new QHVUN
H 3.30
meowumEuom uwGDIH\mcowu uHoDIH
Insane 010. oz Inseam meme“. cum \mouoz mo
moumaavuoou mamwnuo>n< mamofieoz sue? muueala mo nonEsz emu: umnEsz emu: uEmz

 

 

Annamamv acme Haasa>aece .mH oases

 

68

the group will be ranked compared to one another on the other measures
of syntactic control. It is interesting to note the high correlation
between the maturity measure and the frequency of the use of coordinate
structures. Hunt and others have pointed out that the more frequent use
of coordinate structures is apparently a feature of maturity. Further
investigation in this area will be necessary before any conclusions can
be made.

The written data were analyzed in the same way (see Table 16). In
this case two of the six were possibly predictive (Becky and Kathy), but
the other four showed a random correlation. Once again we must conclude
that the maturity measure did not predict for this particular study group
the comparative rank within specific syntactic categories. Also, the
high correlation found in the spoken data between the maturity measure
and the use of coordinate structures is not found in the written data;
in fact, it shows the least amount of correlation. Instead, a high
correlation here exists between the maturity measure and the percentage

of T-units having no sentence-combining transformations.

3.5 Conclusions - The following conclusions can be drawn from
comparing the linguistic data collected from Chicano children and
middle-class Anglo children.

1. Using the objective measure suggested by O'Donnell )2
et a1, i.e., the mean number of words per T-unit,
the Chicano children are less mature in their use of

syntactic structures than are their middle-class peers.

69

 

 

 

 

 

 

 

en es <H ea <H M
mm ma ma ma M
.a u« n« an n« l
< u .n < n «.5 meme a
< I I I I Ooh ehfiumvm L.“
n I < n < m.n osfiummoo w
_
HHH mm<qo m
< n u I n ¢.w mxumm
m < m n m n .w m5
HH mm<au
LWF
m m a I I m m.m swoon
H mm<qo
meoaumsuom oweDIH\nsoHu sweaty
Insane via 02 Imauommemua 01m \mouoa mo
moumcwouoou mamanum5o< mamofieoz suwz muqeblh mo nonesz sou: Honssz oomz memz

 

 

 

 

 

 

 

 

“masseuse

scum Hmaea>auae

.ea «Haas

2.

70

According to this statistic, the ability of the
Chicano group to Speak English is comparable to the“
middle class first grader. The written ability of the
Chicano children studied lies somewhere between the
third and fifth grade ability of the middle class
child.

In almost all cases the Chicano children studied
evidenced a greater degree of syntactic control

in writing than in speech. This fact concurs with

the findings for the Anglo children. It has been
argued here and in Chapter 11, however, that this

is probably reflective of the fact that style or

the use of language in a particular context

was measured.

Inconclusive results must be reported on the finding
regarding nominal constructions formed by sentence-
combining transformations. The finding, which showed
the Chicano group as having more mature control over
syntax than their middle class peers, is not supported
by any other finding in the study. The unexpectedly
high number of nominal constructions may have resulted
from a lack of consistency between the O'Donnell
grammatical analysis and this investigator's analysis.
On the other hand, it is possible that situational
features biased the data so that more nominal construc-
tions were produced in the Chicano study than in the

O'Donnell study.

4.

71

Because the figure for the mean number of sentence-
combining transformations includes the figure for
nominal constructions, this finding, which showed

the Chicano group at the same level of maturity as

the O'Donnell group, must also be considered
inconclusive.

In their use of adverbial constructions, the Chicano
children studied were between the second and third

grade level in Speech and below the third grade level
in writing as compared to the O'Donnell group. In this
instance adverbial constructions were used less often

in writing than speech, with no occurrences of two of
the three subtypes, i.e., sentence adverbials or
adverbial infinitives. These subtypes, it must be
noted, occurred infrequently in the O'Donnell data.

That Chicano children studied exhibit less control in 'E
syntax becomes most apparent in their use of coordinate
structures. In Speech they fall below the kindergarten
level in their use of coordinate nominals and modifiers
and at the kindergarten level in their use of coordinate
predicates. In writing they are below the third grade
level in their use of coordinate nominals and at the
third grade level in frequency of coordinate predicates.
There were no instances of coordinate modifiers in the

written data for the Chicano group.

72

7. The overall patterns or trends displayed by the
Chicano English data concur with the trends seen in
the O'Donnell data. That is, the proportional rates
of occurrence per hundred T-units among subtypes of
the nominal, adverbial, and coordinate constructions
were similar. For example, within coordinate structures,
the O'Donnell findings show that coordinate modifiers
are used less frequently than coordinate nominals,
which occur nearly twice as often as modifiers. There
is considerable fluctuation in the frequency of
occurrence of coordinate predicates. This same general
trend is apparent in the Chicano data.

8. The maturity measure proposed by O'Donnell et a1, i.e.,
the mean number of words per-Thunit, was not predictive
within the Chicano group of their ranking relative to
one another on specific syntactic operations. In other
words, the maturity measure was not reflective of an

individual's ability to control syntax.

4.0 Further Investigation of CE

4.1 Introduction - In Chapter II of this thesis criticism was made
of the use of spontaneous speech as primary data in linguistic research.
It was shown that the competence a child has may often exceed his
spontaneous performance (cf. Chomsky's phonological example, p. 11).

A study of child language acquisition by Shipley, Smith, and Cleitman

73

". . . a description of the child's spontaneous

has also noted that
utterances does not do justice to his linguistic organization" (p. 336).

In order to investigate further the competence of the CE speakers,

a sentence repetition exercise was devised to test whether certain

a;

nonstandard forms which had been spontaneously produced would be
repeated by the children when given the standard version of that form.
Many of the people who investigate the acquisition of language by
children obtain the data used in their analysis through the sentence
repetition technique.

Paula Menyuk asserts that the child's reproduction of structures is
limited by the rules that are part of his internalized grammar, i.e.,
that he will not repeat certain grammatical structures if the rules
required to generate these structufes have not been internalized. However,
given the memory aid of a sentence to repeat, children may exceed the
level of competence determined from spontaneous Speech which may be
indicative of rules which they are acquiring at that particular stage.
She concludes that structural descriptions of spontaneously-produced
sentences may lag behind the level of competence of the child in under-
standing syntactic structures, i.e., encoding lags behind decoding, but
the lag is not too great (Henyuk: p. 154).

Joan Baratz has also made use of sentence repetition in a study of
the speech of both Black children and White children and has found that
the children tended to give back their own dialect (pp. 18-19). Others
have also found the sentence repetition technique useful in studying

dialectal differences (Labov and Cohen: 1967; Jensen: 1970).

74

The sentence repetition exercise was the last task the children
were asked to perform. Because the entire time of the study was limited
to four weeks (the duration of the summer school session), only a
preliminary and cursory examination of the data could be made prior to the
construction of this exercise and,hence, only the more obvious nonstandard'
forms used by the children were examined in this test procedure. The
children were asked to repeat twenty-six sentences given by the investiga-
tor; eleven of the sentences were standard English and fifteen were
nonstandard forms, which the children had been observed using. Nine
nonstandard grammatical structures were selected to be tested. Various
numbers of sentences were devised to test each of the nine structures
and these were then randomly arranged within the whole exercise of
twenty-six sentences. In reporting the results here, the number of the
sentence will be given (indicating its place within the exercise), followed
by N or S, that is, whether the sentence was given in nonstandard or
standard form. In parentheses following, the response of the children $3
is given, indicating how many repeated the sentences using standard or
nonstandard forms. Occasionally, there were no repetitions made by the
students for various reasons; this will be indicated by showing how many

failed to repeat.

4.2 Results of the Sentence Repetition Exercise
4.2.1 Warmrup and Copula Deletion
V #1 S (118) Robert is going to day camp tomorrow.

#2 S (118) The girl's making a dress to wear to the party.

75

4.2.2 Warm-up and Subject-Predicate Agreement
#3 S (118) When his father gives him money, he goes to the
movies and buys some candy.

The first three sentences in the exercise were devised as warm-up
sentences to get the children used to the form of the test. Secondarily,
the children were tested to see whether any features commonly described
as being part of the dialect of Black English (see Jensen) were also
used by Chicano children. Sentences l and 2, then, tested for copula
deletion and sentence 3 examined subject-predicate agreement. All of
the children repeated these three sentences as they were given, i.e., s.
using standard English. Sentence 2 proved rather effectively that copula
deletion is not a feature of Chicano English. Several children inter-
preted the subject as a plural noun rather than a contraction and inserted
'are' appropriately.

4.2.3 Topicalization

r“ ~' #6 N (6S,5N) The boy he wanted to be the captain of a big ship.
#11 N (58,6N) The boy who found the boat he had many
exciting adventures.
#15 N (95,2N) After feeding the seals, the boy he went back
to the boat and sailed away.
#19 N (83,2N,10) Maria and Carmen they are going to be in
the fifth grade next year.
#25 S (103,10) Mary went to the store for her mother and

bought some food for supper.

76

The results of this set of sentences show that there seems to be no
decided preference for repeating subjects; no one child used the nonstandard \3
form exclusively, and only Eddie, who has been shown previously to be
close to middle class Anglo children in his use of English, consistently
used the standard form.

An interesting intonation curve patterned frequently though not
invariably with occurrences of topicalization. The pattern was such
that the final syllable of the subject noun phrase ended on a rising
intonation followed by a pause and then the anaphoric pronoun. Some
examples from the data include:

The warlord's brother: he pushed him in the water

DSh Coyote: he saw this other rabbit.
It is as though the subject noun phrase is separated from the sentence
with the intonational features calling attention to it as the topic of
the sentence and then followed by a comment on that topic, the comment
itself having its own subject noun phrase.

Jensen in her work on Black English has pointed out that this type
of nonstandard utterance is common also in that dialect. She also cites
other works which discuss topicalization in such languages as Arabic,
Spanish, and French. It has also been seen as a common feature of the
language of children during the acquisition period. These facts lead
Jensen to conclude that this type of sentence seems to be fundamental

and perhaps even innate (Jensen: pp. 25-32).

77

4.2.4 Embedded WH Statements
#5 N (28,9N) The boy saw some people and asked them whose
boat was it.
#10 N (25,9N) I saw a strange animal and I didn't know
what was it.
A - #13 N (llN) The boy who took the boat wanted to know whose
was it.
#18 N (lS,9N,10) We learned something about Jalisco but
I don't remember where is it.
#21 S (7S,3N,10) I can go to the movies alone because I
know where it is.
#24 S (88,2N,10) I found a jacket in the room and asked the
teacher whose it was.
Here an attempt was made to test whether, in generating embedded WH
statements, particularly following the lexical items 'ask and 'know',
the WH form was consistently followed by the copula and noun phrase in
the English used by children in this group. There is more consistency

here than could be shown for the topicalization test when given a

:9

nonstandard form to repeat. On the other hand, when given a standard
form, most of the children repeated the standard form.

Jensen, in examining similar constructions in Black English, has
suggested that this type of construction be considered as two separate
sentences simply joined together into a single sentence with no adaptation
taking place in the conjoined sentences (p. 19). Further investigation

involving more complex constructions, i.e., expansions of the verb

78

phrase and/or the noun phrase, and related constructions appear to
be needed in order to determine whether Jensen's explanation would

hold for this case.

4.2.5 Deletion of the Past Tense Morpheme
v #7 S (108,1N) The boy sat down at the table and dropped
his popcorn.on the ground.
#26 S (8S,lN,20) The boat floated toward the boy and
stopped right in front of him.
The test here was to see whether the children deleted the past tense
morpheme particularly following verbs having a final voiceless stop.
Both sentences were given in standard English and repeated by the majority
of the group as given. The two instances of nonstandard reproduction
were by two different children. Clearly, then, although examples of
this type of nonstandard formation were found in the spontaneous speech
of these children, it would not seem to be a consistent feature of their 3?

internalized grammar.

4.2.6 Negative Formation
i - #9 N (4S,7N) When there isn't no wind, sailboats can't
move very fast.
Lt ' #16 S (lOS,lN) If there aren't any pencils left, we can't
finish writing our stories.
Once again the pattern seems to indicate a lack of consistency and
a tendency to repeat what is given as it is given. When a nonstandard 9

form was given, most of the children gave it back; when the standard form

79

was given, then it was given back. Some importance might be given to the
fact that so many children repeated the standard form, perhaps indicating
that that form is dominant.

The sentences chosen were poorly designed to test the use of double
negatives and are not typical of the ones produced in spontaneous speech,
e.g., 'A boat was coming without nobody.' or 'we don't have nothing to
eat.’ Spanish allows double negatives in these situations and their use ;y
in English by Chicano children is probably a reflection of interference
from Spanish, whereas in the sentences chosen for the exercise, double
negatives would not be allowed in Spanish.

4.2.7 Omission of Relative Pronoun

v #4 N (83,3N) In the film there was this man caught some
fish and had his picture taken.
x, #12 S (88,1N,20) In the film there was this boy who
sailed in a boat and had many adventures.

The only time the omission of the relative pronoun was found in the
data was when it occurred in the sentence pattern type shown, i.e.,
there-be-subject. The results fairly clearly indicate that this is not

a consistent feature in the Speech of these children.

4.2.8 Idiomatic Expressions
#8 N (58,6N) The cameraman set his camera so that he
could take a picture to the man.
#17 N (4S,6N,10) The man sat down so that the cameraman
could take him a picture.

#22 S (9S,lN,10) Ana took a picture of the man while he was

working.

In their retelling of the plot of the film, several children
exhibited difficulty in forming the expression 'to take a picture of
someone'. In comparing the results of these sentences, once again the
conclusion must be drawn that when given the standard form, it is likely

to be repeated as such by the children.

4.2.9 Pronunciation of 'his' as lhiyz/
#14 N (108,1N) In the end the boy was happy because he
got lhiyz/ boat back.
#20 N (8S,2N,10) The boy ate his hot dog but the pidgeons
ate all of /hiyz/ popcorn.
#23 N (9S,lN,lO) manual said he's going to take [hiyz/ bike
to the shop to get it fixed.

This last item indicates that there is little difficulty with this
particular feature. The investigator was motivated to test the pronunciar
tion feature when it appeared in the written work of one of the students
Spelled 'he's'; in this exercise, this particular student consistently
used the [hisz pronunciation. This type of nonstandard pronunciation
would be expected of native Spanish speakers because in Spanish there is

no /i/:/I/ contrast.

4.3 Conclusions - The results of the sentence repetition exercise
are difficult to assess, but some tentative speculations can be made. Of
the nine grammatical items tested, two, copula deletion and subject-
predicate agreement, are definitely not problem areas for the Chicanos

in the study group. Two other areas might, upon further investigation,

81

prove to be candidates for dialectal differences, that is, topicalization
and nonstandard embedded NH statements. However, even with these two
items, the reSponse seemed to be dictated more by the form in which a
sentence was given, i.e., either standard or nonstandard, than by any
internalized set of rules.

The results of the other five items tested show only that there is no
consistent pattern in the repetition of these forms, either standard or

nonstandard.

4.4 Critique of the Sentence Repetition Technique - One of the
specific techniques Chomsky mentions as a good method of getting at the
underlying system of a child's language is to test his ability to repeat
sentences and nonsentences (Chomsky: 1970, p. 47). He makes no mention
of the limitations of such a technique with regard to age, degree of well-
formedness of the nonsentences, etc. It does, however, seem important
to ascertain the limitations of the sentence repetition technique since
the use of sentence repetition, at least the one designed for this Study,
was not considered to be fruitful.

In the study of Chicano children, the interpretation of the results
was fairly unclear. The only result which can be stated is that when the
children were given a standard sentence to repeat, there was a decided
tendency to repeat the standard form. On the other hand, a nonstandard
form would most likely elicit a nonstandard repetition on the part of the

informant. One reason for this might be that neither the standard nor <5
7

Kt;
/‘
nonstandard formmwas dominant in their individual grammars and, hence, they

were bidialectal and were able to repeat either naturally.

82

In order to investigate this problem further, the sentence repetition
exercise used in this study was given to three more children, none of
when used in spontaneous speech the nonstandard forms used by the Chicano
children in the main study. The subjects included a ten-year-old boy,
and two girls, aged ten and six. They were middle class Anglo children.
The boy repeated every sentence, except those involving nonstandard
pronunciations typical of native Spanish-speakers, exactly as they were
given, that is standard forms were repeated when standard forms were
given and nonstandard forms were repeated when they were given. At the
completion of the exercise, he volunteered information about the 'bad'
grammar in many of the sentences. The ten-year—old girl responded quite
differently. By and large, she changed the nonstandard forms into her
own, Standard version. She too, however, pointed out that some of the
sentences given her were incorrect. The six-year—old girl responded by
giving her own, most often standard, version of the sentences. She did on
occasion repeat a nonstandard form that was probably not part of her
grammar. (The data for this exercise are given in Appendix D.)

The kinds of conclusions that one can draw from such a small sample
are doubtless tentative. It may be that there is an age limit in using
this kind of test and ten years may be too old. That is, the children
may have had so much practice in repeating what the teacher says in a
classroom situation that the sentence repetition test is not as spontaneous
an exercise for them as it is with younger children. It also seems to be
the case that certain subjects are more aware of the intricacies of

language, as in the case of the ten-year-old boy, and hence, are not good

83

subjects for this kind of exercise. The question of when one can begin
to ask children direct questions about language has been investigated by
Russian linguists. In his review of Russian studies of child language,
Slobin cited a study by L. E. Zhurova, which indicated that school age
children (he mentions only one specific case of an eight-year-old) can
make words the object of their attention and begin to make initial
generalizations about linguistic material (Slobin: p. 384).

It must be pointed out that the technique used here and the ones
used in the studies mentioned are not good examples of what Chomsky
suggested. In our study we used standard and nonstandard (although
generally presumed acceptable within the limitations of the grammars
of the subjects) versions of sentences. Once again, if we want to get
at what the child knows about his language, his performance in tests
involving the repetition of nonsentences,as opposed to simply nonstandard

sentences, may be more revealing.

Chapter IV.

S IJTCMRY AND CONCLUS IONS

1.0 Summary

Claims are often made by nonlinguists - usually educators - that
the Chicano child drOps out of school because he fails to master English.
Yet very little linguistic investigation has been done of the English
used by Chicano children either to support or refute this claim.

The goal of this thesis was to compare the English used by ten
Chicano children who were approximately ten years old and in the fifth
grade with that used by middle class Anglo children of the same age/grade
level. A study by O'Donnell, Griffin, and Norris provided a model for
both data collection and analysis as well as the comparative figures
for the fifth grade Anglo children.

The study is basically statistical in nature, using the T-unit
identified by Hunt as a basic unit. Figures were calculated and compared
for all nominal, adverbial, and coordinate constructions produced by
sentence-combining transformations for both spoken and written data. By
and large the results were that the Chicano children did about as well
as first grade Anglo children in spoken language and fell at the thirdfié

grade level or below in written work.

84

85

It was interesting to note that using O'Donnell's definition of
linguistic maturity, i.e., the increased use of complex syntactic
structures, the children, both Anglo and Chicano, were shown to be more
mature in written rather than in spoken language. A number of eXplana-
tions were put forth in this thesis for this phenomenon. One possibility
is that there are certain complex syntactic structures which are more
appropriate to writing than to speaking. If this is so, then it does not
necessarily reflect more mature use of language, but simply differences
in style. The point was also made here that, in terms of the child's
competence or internalized knowledge of his grammar, whether he uses the
grammatical structures in writing or speech does not affect this competence.
That is, if the child can generate the structure creatively - whether in
writing or speech - then it is a part of his competence and it is the
definition of linguistic maturity which is weak.

Another consideration in comparing written and spoken results, is
that the idea of 'control' of language is given some meaning, at least
within a theory of performance. That is, given time, explicit instruction
in writing techniques, and a conscious awareness of sentence generation,
it is not unreasonable to expect that the syntactic structures used in
writing would be more complex than those generated in spontaneous Speech.
In this instance then, 'control' or 'manipulation' become meaningful.

O'Donnell, Griffin, and Norris proposed a simple, objective measure
of linguistic maturity which they claimed reflected increased use of
complex syntactic structures. That measure was to count the mean ?

number of words in a T—unit: the higher the number, the more linguistically

86

mature the speaker. The Chicano children were rated using this measure
as well as other measures such as mean number of nominal, adverbial,
and coordinate structures produced by sentence-combining transforma-
tions per hundred T-units. It was found that the simple, objective
measure was not predictive for the Chicano children.

A sentence repetition exercise was given to the Chicano children
to test whether nonstandard forms which they had spontaneously produced
might be part of a Chicano dialect. The sentences in the exercise
contained both standard and nonstandard versions of nine grammatical
items that had been selected to be tested. The children's repetition
tended to follow the version that was given to them. The results of
this exercise are difficult to assess; however, it may be that the
children are bidialectal and that neither the standard nor the non-
standard form is dominant in their grammar. Although no clearcut
evidence of the development of a separate dialect of English by Chicanos
was given, a more intensive study along these lines might prove worth-
while.

A further investigation into the particular sentence repetition
exercise given to the Chicano children was made to determine the
validity of using such a technique. Three middle class Anglo children
were chosen for the study - a ten-year-old boy and two girls, aged
ten and six. The boy repeated the sentences as they were given, but the
girls generally changed the nonstandard to standard. Both ten-year—olds,
however, volunteered information about the 'bad' grammar in some of the

sentences. It was concluded that it may be that the sentence repetition

 

87

technique is not fruitful to use with children over a particular age
and it seems that ten is too old for such a test. Further investiga-
tion of the use of sentence repetition as a tool for investigation of r'
the acquisition of language by children seems called for, particularly
to determine the age at which such tests are no longer fruitful.

The study done of the Chicano children, as well as like studies
of the child's use of language by O'Donnell et a1, Hunt, Loban and
others, calls for serious reconsideration of the goals and methodologies
of such studies. Too often not enough thought is given to the concept
of linguistic maturity and what it means within the framework of
grammatical theory. It was suggested here that there is a difference
between linguistic maturity in the context of a theory of performance
and maturity in the context of a theory of competence. Too often these
investigators do not seem to be aware of the differences between competence
and performance and their studies suffer from such an oversight. They
must ask themselves what the concept of the frequency of occurrence of
complex syntactic structures has to do with the development of the
child's competence. Further, they must consider what underlying
assumptions are involved in such concepts as the 'use of language',

'control of language', and 'manipulation of language'.

2.0 Conclusions

If our interest in the study of the language of children is to
reveal the development of the knowledge the child has of the grammar
of his language, then studies of this type will not lead to any

theoretical developments in this area. Statistical counts, such as were

 

88

done in this study, which show only that the child increases the
number of times he generates a syntactic structure but not that he #3
adds any new ones to his grammar from the age of five to twelve, tell
us little about the development of competence during this time. And
yet we know from Carol Chomsky's work on children from ages five to
ten that competence is not fully developed at age five.

This study confirms the need expressed by Noam Chomsky for
linguists to develOp new techniques for Studying the competence of
children. The use of spontaneous speech has been shown by Shipley,
Smith, and Cleitman not to be a good reflection of competence. However,
the use of direct recording of spontaneous speech should not be ruled
out entirely in linguistic research. In preliminary studies, such
as the one done here on the use of English by Chicano children where
no published linguistic material was available, such data can be useful.
However, it must be remembered that such information is limited in its
usefulness. It is not interesting of itself, but should be used in
designing procedures to investigate the linguistic competence of these
children and to provide the investigator with insights into their
competence.

Much more research needs to be done with Chicano children. We
need to study the development of their competence in Spanish as well
as in English and to determine if possible the interrelation, particu-

larly during the time of acquisition, of these two languages.

89

3.0 The Role of the Applied Linguist

Only negative criticism has been given of the type of study
exemplified by the O'Donnell one. We have seen that their attempt to
develop an index of maturity was not satisfactory. Although their
reasons for searching for such an index are not stated explicitly, it
seems to be the case that such an index is thought to be useful to
educators in their attempt to measure the capabilities of the children
under their care in order to aid in the design of curriculum. And so
the need arises for a measure. O'Donnell et a1 Show how, by using an
IBM computer and many professional staff pe0ple, such a measure may be
arrived at. But this is obviously much too complex and costly an
operation for most school districts to perform, so they devise what
the call a Simple, objective measure: the number of words per T-unit.
Once again we are back to word counting - perhaps more refined than
simple gross word counts - but word counting nonetheless. They claim
to have shown that the word count per T-unit correlates highly with the
increased use of other syntactic structures. However, the correlation
did not show, in the study of the Chicano children, that there was
anything but a random association between the measure and these
structures.

Just what the implications of such an index of maturity might be
has received too little attention on the part of most applied linguists.
Presumably curriculum would be designed so that a child might be explicitly
taught to use more complex grammatical structures in his Speech and in

his written work. Hunt and Mellon have both suggested such classroom

90

techniques (Hunt: 1970, p. 191; Philon). But to what avail? If maturity
is seen as a natural growth process, at least for native speakers of a
language, what is the need for this kind of eXplicit instruction? As

far back as 1836, Humboldt concluded that one cannot teach language but
can only present conditions under which it will develop Spontaneously

in the mind in its own way. Experience sets the language-forming
process into operation (cited in Chomsky: 1965, p. 51).

The work of applied linguists suffers from a lack of theoretical
perSpective and its usefulness at the present time is difficult to assess.
Certainly no clear-cut use can be shown for the studies which have been
discussed in this thesis. Perhaps the most effective thing linguists
can do which would have immediate, practical value is to try to educate
the educators on what we know about the development of language as a
natural growth process, about dialectal differences being simply
differences and not defective attempts to speak correctly - more in the

style of Labov than has been done here.

I) IBLI ()CRAPUY

BIBLIOGRAPHY

Baratz, Joan. 1970. "Teaching Reading in an Urban Negro School System' ,
in Williams, 1970, pp. 11-24.

Chomsky, Carol. 1969. The Acquisition of Sygtax in Children From 5
to 10, Research Monograph No. 57, M.I.T. Press, Cambridge, Mass.

Chomsky, Noam. 1957. S at ctic Structures, Mouton & Co., The Hague.

Chomsky, Noam. 1965. Aspects of the Theogy of Sygtax, M.I.T. Press,
Cambridge, Mass.

Chomsky, Noam. 1970. 'Formal Discussion of "The Development of Grammar
in Child Language" by Wick Miller and Susan Ervin', in Lester,
1970, pp. 41-50.

Harrell, Tester E., Jr. 1957. 'A Comparison of oral and Written
Language in School-Age Children', Monographs of the Society for
Rese rch in Child Develo ment, 22:3.

Heller, Celia S. 1966. Mexican American Youth, Forgotten Youth at the
Crossroads, Random House, New York.

Hunt, Kellogg W. 1964. Differences in Grammatical Structures Written
at Three Grade Levels, the Structures to be Analyzed by
Tr sfo tional Methods, Report to 0.8. Office of Education,
Cooperative Research Project No. 1998, Tallahassee, Florida.

Hunt, Kellogg W. 1965. Grammatical Structures Written at Three Grade
Levelg, Research Report No. 3, National Council of Teachers of
English, Champaign, Ill.

Hunt, Kellogg W. 1970. 'Recent Measures in Syntactic Development',
in Lester, 1970, pp. 187-200.

Jensen, Louise. 1970. An Investigation of Four Aspects of Black English
Within a Tr sformational Framework, unpublished M.A. thesis,
Department of Linguistics and Oriental and African Languages,
Michigan State University, East Lansing.

Labov, Wm. 1970. 'The Logic of Nonstandard English', in Williams, 1970,
pp. 153-1890

Labov, Wm., and Paul Cohen. 1967. 'Systematic Relations of Standard
and Nonstandard Rules in the Grammars of Negro Speakers', from

Prgject Literacy Reports No. 8, Cornell University, Ithaca, N.Y.

91

92

Lees, Robert B. 1957. 'Review of Chomsky's S tactic Str ctu es',

W 33:375-408.

Lester, Mark, (ed.). 1970. gppdipgp 1p Applied Trppsfprmatipppl
m, Holt, Rinehart, and Winston, _Inc., New Yo .

Loban, Walter D. 1963. mg magge g; Elempppn fiflool Children,

Research Report No. 1, National Council of Teachers of English,
mmdm. I11.

Manuel, Herschel T. 1965. S anish-S eakin Children of the Southwest, .
University of Texas, Austin. '

McCarthy, Dorothea A. 1954. 'Language Development in Children', in
Manual of Child Psychology, ed. by Leonard Carmichael, John -

Wihy 5 8“. “W730“. PPS 492’630o ‘

|

McNeill, David. 1966. 'Developmental Psycholinguistics', in Smith
and Miller, 1966, pp.15-84.

Mellon, John C. 1969. Transformational Sentence Combinin , Research
Raport No. 10, National Council of Teachers of English,
Chamaign, Ill.

Menyuk, Paula. 1969. Sentences Qildrep Use, Research Monograph No. 52,
MoIsTs Press, C‘d’tidg‘. “38’s

O'Donnell, Roy C., William J. Griffin, and Raymond C. Norris. 1967.
Syntax of Kindergarten and Elementapy School Children: A
Transformational Analysis, Research Report No. 8, National
Council of Teachers of English, Champaign, Ill.

Ohmann, Richard. 1970. 'Generative Gramars and the Concept of Literary
Style', in Lester, 1970, pp. 117-136.

Olguin, Ieonard. 1968. Shuck Loves Chirle , Golden West Publishing
House, Huntington Beach, Calif.

Roberts, Paul. 1964. En lish S t x, Harcourt, Brace, and World,
New York.

Shipley, Elizabeth, Carlota Smith, and Lila Gleitman. 1969. 'A Study
in the Acquisition of Language' , m, 45:322-342.

Slobin, Dan I. 1966. 'Abstracts of Soviet Studies of Child Language',
in Smith and Miller, 1966, pp. 363-386.

Smith, Franklyn L., and George A. Miller. 1966. The Genesis of Len as e,
M.I.T. Press, Cartridge, Mass.

93

Stanford University, Center for Latin American Studies. 1969. The

mxican-Ameriggg A selected and Annotated Bibliography,
BmJMCO. California.

Stockwell, Robert B., J. Donald Bowen, and John W. Martin. 1965.

The Grammatical Structures of English and Spanish, University
of Chicago Press, Chicago.

Strickland, Ruth C. 1962. 'The Language of Elementary School Children:
Its Relationship to the Language of Reading Textbooks and the
Quality of Reading of Selected Children' , Bulletin of the

School pf Egugation, Indiana University, 38:41

University of California at Los Angeles. 1967. Mexican-American Study

Project Revised Bibliography, Advance Report No. 3, Los Angeles,
Calif.

Williams, Frederick, (ed.). 1970. Lan ua e and Povert , Markham Publishing
' CO0. Ch1C880s

 

APPENDI CES

 

€911

 

 

 

 

 

 

H.H N.o H H H H assoc
«.0 H H song
H.H N.o H N H H H H soHuHesoo
H.H N.H H N H N H H N H sauce
:6 N H H Hoses:
N.o n n sosHm
N.S N.S S H N N H N N N N N N n csHa
N.N N.N N N N S N S S N S S N N N cos-Ho H.Hnsosoa
o.N N.NH m NN H N S S S N S N N N n N Numaaoameu HSHNNNSNS
N.NN N.NH N NN N N N N N N N HH N N N N N oH Hoeuo
N.HH H.N oH N H H H N H H S N H so: H.Hnuosoa
N.oS H.NN SN HN N HH H N H N N N N4 S NH N N NH NN noun co SooHSo
N.N H.N N NH N H N H N opHSH-osaa
H.H N.o H n H H H H Sac-SHNSoo .Hno
S.o N H H H Secs-Hasoo .HSSN
N.o H H Soonno SooHHocH
o.NN S.oN NN NNH N NH NH N N HH S NH N NH NH NH N NH SooHao
N.HN N.oH NH NS N N H N N N N H S N N N oH N N Scones»
guuogh
N.o n N Hoeuo
N.S S.N S OH H H N H H N N H N conned season
N.o H sooHnSn SSH: :HNSH
N.NH N.N NH NN H N H S S N N N N N N oouuem SHeaH
N.o n H H H cussed nous
QK a; N om N m n m n N H n H H ens-Ho use:
uses-mug:
S.o N H H Hoeuo
M.H N H N H N H.Hnnopua o -
H.H .H H N H N H N H H H causes .HH-N o -
H.H N.o H N H causes nHNSH . n
a n3 0 3 a H H n N H m H N n eschew noun a. a
N.~ o.~ S NH N H H H S N H N m H ooanHo Has o a
o. H o. H NH NN N N NH H S N N NH N N N N N noon S -
H.NN N.NH NN HN N N H HH N N N H S N SH 0 HH N NH Hoa . a
N.NS N.NH NS No oH SH N N H S S N N N S S N oN oN - o s
coo-on
S.NN N.oN NNH oNN NN on N on N NH NN NH HN NH HS SN Nn NH MN Nuuauosman SHIN-as
N.H No.H NNH NSS HN NS N NS HH on NH NN NN NH NN HN NN NS N .ususs scHeHnsoo cocoSnoN
N.N NH H H H H S N N H H.HSHSN
M.N SH H H H N N eoHSSSSa
.o S N H H N> SH
N.N H.N N N H N N H H H S H H N> Shoes
NS SSS
H H uuo>n
N.o S N H coo>N
H.H N.H H N N N H H N oHSN
S.N N.S n NH N N H H N N H N N SSS»
N.N NH N H H H H N N H H no>n
N.NN N.NN 0N NNH N NH H NH S SH NH N HN N NH 0H NH S SH o>n
N.NN N.NS HN SoN N NH N NN N nN HN NH NN N N NH NN NH NN >N
aauaaam mosses:
i# I
sansuu ooH use a N a N a N 3 N S N 2 N S a m a N 3 N
3:23.08 no 38 3.92. :38 F 23. 25m 325 2.288 s: n— 93! H.Honom .2ch Eamon-o SSS-IL

 

 

 

 

 

 

 

 

 

HcooHHus on. unseen. cuss e-HHNcu ocuoHco .< «Hosumma

 

 

 

.. .N. i
A v. A .. a.
as- - H . S HN
_. ,H sale. a. c‘o-sso
. . i.——~
—.-.. Hm-v—.~-..~H "A. .-e —...—..-N -v -

Mum “\

-~ .—...1

m... J.

Hw .— H»
- -. . -H.
, Q .
- . . SS-
... ..~4
N-.-s ..-.--- ... .
N ‘«- H ’--o—-—‘ m.-

*.fl,

 

”A... 4

“NH .

.»~-—‘— 1*...“ ,

—._a—-‘H~— H...

of.

vi-

4.-
one.
.vc.
«u. .

95

 

 

 

 

 

 

 

 

 

 

 

 

 

 

echo: a slew mood
N.oN o.oo HN new NH wN N Nn m an Nn 0H NN N N N HS o NN uposu suHeous no sense.
n.H H.H N..H n.H o.o m6 0.0 N5 5.0 SH.H N.H n.H. H. SH. men con N.H “Sgt“. Mom
seoHussuoeseshH uoHoHnsoo
season-n no Hos-s: ass:
NN HHS NH NN SH SS SH NS SN NH SN NH 3 NN 3. NH NS 3:54. co Non-S.
Noo H H Hosea
MoH v N H n Ob
N.SH .N NH NN N N N N S N N N H N N N S
N.SH o.m NH an N N N N S N N N n a n o sousoHoskm
N.N N. N oH H H H H N H N o N sHsoHsoa
N.o n H H H HsapHSSed
N6 n H H H 23:23:
o.NH N.NH NH NN S N N H N N N N N 3 NH N HH 81:32.3 339.80
N.H N N H N H H .3253 H392»:
H.H 3.N H OH H H H H n H 0.518 cones—nus»:
H.H S.N H OH H H H H a H Hannah—He sense—Hon
2:88 23.82.... Hansen:
3 n
sud—SIB OOH ask 3 n 3 n 3 m 3 a 3 o 3 m 3 a I n a a 3 a
3:93.58 No and nag mason Nessa soon 3.55 2.3800 see Essa uuosom oHeoa huououso HooHusIIe

 

 

 

Appendix B. Chicano English Worksheet

Name ‘__ Ana#h‘ Oral X Standard X
Written Nonstandard

T-UNIT: (221__the man looked up and stopped the 'uh' ride#_

Words in T-Unit 8

Sentence Patterns:

5 v s v 0 Ca No Vp
svo SVOCa :1in
S V Cn Adv V S Question
S V Ca There V S Request,Command
S V I 0 It V S Partial
SENTENCE COMBINING TRANSFORMATIONS 1
Nominal Structures
Headed Non-Headed Function
N + N Noun Clause Subject
N + Adjective Prep Phrase Object
N + Possessive Infin Phrase Indirect Object
N + Relative Clause Infin with S Subject Compl.
N + G Rel Clause Gerund Phrase Object Compl.
N + Prep Phrase Other Appositive
N + Infin Phrase Object of Prep
N + Part. Phrase Adverbial Noun
N + Adverbial Other
Other
Adverbial Structures
Adverbial Clause‘___ Sentence Adverbial__. Adverbial Infin___
Time Reduced Comp Absolute Construction
Place The More the Interjected Clause
Manner merrier Other
Cause Adj Compl
Condition Other
Comparison
Coordinate Structures 1
Modifiers Nominals Predicates
Adjectival
Adverbial V V I O
V O V O Cn
V Cn V 0 Ca
V Ca Other

96

Appendix C.

A List of Nonstandard Grammatical Forms

1.0 Introduction

The fact that there might exist a Chicano dialect of English has
been mentioned in this thesis, although proof of such a dialect was not
given. It was pointed out that the social conditions, at least, were
amenable to the development of such a dialect.

In order to provide evidence, one will have to provide a description
of grammatical, that is, phonological and syntactic, features which
are not part of standard English and which are used systematically by
a number of Chicanos, adults and children. As a first step in this
direction, it is necessary to have some knowledge of the nonstandard
forms used. Thereafter, investigations will have to be made to see
how widespread these nonstandard forms are, that is, how many people
in the Speech community use them and how consistently they are used.
In the case of Chicano English, it will be interesting to know which
nonstandard forms result from interference from Spanish.

As an initial step in the investigation of Chicano English, then,
a list of the nonstandard syntactic structures included in the data
collected for this thesis is presented here. The data were taken from
twelve children, three more than were included in the major portion
of this thesis. The data come from many sources, but only spoken

language is used.

97

 

98

The data used here came from taped conversations with groups of
four and five children, from their retelling of folk tales, from their
summaries of paragraphs read to them by the investigator, and from their
retelling of the plot of the film, The_Little Mariner.

It is the intention here to point out those areas in which the
students produced sentences unacceptable to Speakers of standard English.
Little can be said about the frequency of occurrence of these nonstandard
forms, except to give some notion of relative frequency within the
group. Those nonstandard forms which were used most frequently will be
listed first, followed successively by the less frequently used non—
standard forms.

An attempt is made to provide some degree of explanation for these
forms by suggesting whether the form might be caused by interference
from Spanish (indicated on the table by an 'S'), or whether it might
be a feature used by all children in the acquisition of English (C),
or whether it is a form which can be singled out as a possible feature
of a nonstandard dialect of English (N). This last category may contain
nonstandard forms which are not necessarily peculiar to Chicanos, but

may be a part of the speech of Black or working class Americans as well.

 

99

 

.53235: was some: 33 on one 35. .N
.uH no: moons zone ou causes 0: .c
Nmom use n.0u033 30:3 30% on .m

 

 

.uH no:
moons 306x u.eva menu vam has» new H.3ocx. one .xmw.
NH mm: mecca mHeooa «Eon end on News o: .m wcHsoHH0m NHumH=UHuuuev
o .Nu.m .oemo use use use: no: buxom m: .H mcowumoso :3 usesonfim .N
esosoue
.EHe No use: we umonu0uoe an w cH one < .N oHuonamno men one case
.uuHem ago he one umeu uow o: ova m.vuoHun3 uomnnsm on» awesome
2 .o .m men so I one» eH no: umeu one on» eonu vn< .H unsung wcHeo>uoueH .v
.umonHHom
Noe mHeu one on son mHuuHH mHeu Now see .N :soc uumnesm oeu :uHs
z .o .m .emHm meow newsmu a: one mHeu was whose .H .mwzu. o>HumuumcoEov och .o
esocoue
oHuocemem new
:50: uuonnsm emoBNon
.muHm nuHs xUHum use» on: on I E: I one 059 .N A.Es. Ne ooNHHonuo>
z .0 .m emHm can unwaou on I a: I one saw .H couwov cOHumuHmo: .2
.osou on nausea on Nos 059 .N NHoumHomEEH onBOHHom
.uHenmu asocoue UHNOeawem
z 10 .m ecu vomwnu we ouohou oeu done one . . . .H euHs :50: uumnesm .m
eOHumNHHmoHeoa .H
cOHuocmHmwm -
oHnHmmom wmwmamxm Ewom cheweMuonoz we make

mason vumvswumooz mo momma o.~

100

 

.um: cuHs omHuumE umm ou meHow mums Nona .q
.o ocsw :H xomuum ou GOHMHumv m meme 0: .m
.uuHem oeu Eouw a: EH: voJUHe om .N
m .oouu onu mo csov now «3 .H soHuHmoemum woo»: mo on: .S
noncou
.oouwum Hmmu on 0» wnHow mas o: .N onQEou once no on:
2 .0 .von NE 3H wcHeomHm some Nvopueom .H oumoemumeoc Hmsonmooo .o
.oumn eyes
one no: How mcHeuanm has cu pauses None .N
o .m .cmHm Nona ouoez oomHa w :« noun 0: .H uI Am
.EHn ou
uH seem men new use «Emu oHHouoocHo nose .N
o .m .onmn Noumea oueHue ooze .H NI AN
.wcHensu uumum o: .N
o .N .Nmo use: as aw .H 3.. HH

mnuo> umHomou
mo Show uuouuouoH .u

.EHHm use some H .m
.nau one was whose one son mam .m mnuo> umHawouuH

z .o .xumn NH wsHun men coeu was uH oxmu mam .H no snow uuouuoucH .n

Esau oumocmumeoc
z .o .xume EH; means 0: .H .omcmu uoouuoo .m

mmcoH euo> uuouuoooH .m

maOHuwanmmm -III
oHnHmmom mommwmxm Euom bumpemumeoz.wo omme

101

o .m

z .o .N

III). 111'

COHumcmHmmm
oHNHmmoN

 

.sme. pom .mcme.

ovum usom

uchm cu vow: woweumEom wchooE .muwfimumu.
.coNoum. mcHemoe .oouoHomH.

wchcou ego vHU mosh

.uHmm n no? uH muon3 oumHe m :H acum on case

chHnu pr umsu HHou 30% on sex
.oum menu
so: soox u.cov H mwchu an mmoeu mo mac

.mwmumw ozu usem you on» use on
.o00w on» HHw use .mumom

.ouooeoe m was won as: o Newson o:
.Nos Nasow me Na<

.umon

mHn omen co Eoou anon om cu panama m:

.938 0: EH: £83 3536 o:

.umo ou meHnuoc ops: u.aov oz
.mvonoc unocuHB mcHEou mm3 umon m . . .

monmme

.H

COHu
IEHOM HmMSHm uowuuoudH

COUonH «0
upon uoc no vuoz moon?

.ouoSN.
new wousuHumpsm .uH.

.uwfi3.
pom vmusuHumnsm .30:.

.use. ouo3 mo mmsmHz

.0

.U

.N

.M.

mGOHomouexm uwquOHwH
HEN was”; HoonQH

oHUHuu<
osu we on: vuovewumeoz

GOHuowoz oHnsoc

it‘ll-o"! ll..- ‘ I- ‘

.m

 

Euom bumpooumooz mo omNH

102

.mwmuww on» usem you one use on
.umememsoc a Non macaw ou News
0 on guns 0» News on usHu use case vc<

u .m Nmomo an o>os so» he: .ponuoEvcmuo

.bHo HHsn m mm3 ouoeH
.moonm
m we uHmm won m use ou mecca ewaoco uow hoeH

.HHmu uooH co>om me omen umHHmu m.w:
.oum mesa so:
u .m 30oz u.oov H I mwcHeu chH mHn amoeu mo moo

.mmooe moo» wcHH ou has one I mono:

usom ca 30% eHm: HH.H I has oeu so» eHme HH.H
.ouoSN umon NH: omHHmm on whose mnu umoz

.c30w once an oHuuHH w nEHHo muses we» seek

m .umu m was on» so vesoH Noeu cosh

.HEHE no: no wcdemouv Pane.
.muquHe m EH: use» on wcHow mums away
m .anvcH no HHmn ou.o:

 

coHuwcmHexm - I; III
oHnHmmom Nonmem

0
HNMQ

mEoHnoum GOHumcHouooo .o
soHumEuom aOHumoac .o

mEoHnoue
ucoEouon 0>Huuono< .o

mEoHeoum GOHuoHoQ .2

muHcs
oHuomuezm Ho moHuowuo .m

mGOHumEuom UHuomuGNm

vumoeoumooz‘waoosmHHoomH: .m

m:0Hmmoumxo
oHuoEOHVH uuouuoucH .m

IIII'IIIIIII‘..

Each bumvdmumeoz we MNMH

Appendix D.

Results of the Sentence Repetition Exercise
Given to Three Anglo Children

Item Tested Responses
(No. of Sentences Given/
Standard or Nonstandard) Dirk Jenny Jane
1. Harm—up and Cepula ZS ZS ZS

Deletion (ZS)

2. Warm-up and Subject 18 18 IS
Predicate Agreement
(13)

3. Topicalization (4N,lS) 4N,lS 1N,AS 1N,4S

4. Embedded NH Questions 4N,ZS 6S ZN,AS
(4N,ZS)

5. Past Tense Formation ZS ZS ZS
(23)

6. Negative Formation lN,lS lN,lS 1N,lS
(1N,lS)

7. Omission of Relative ZS ZS ZS

Pronoun (1N,lS)

8. Idiomatic Expressions ZN,lS 13,28 28,10
(ZN,lS)
9. Pronunciation (3H) 38 BS 38

103

 

"7'11 11!? lilifilfllflll 1“