‘ -‘“-

 

THE ACQUISHTON AND COMPREHENSION 0F
DIRECT f END-[REM OBJECT
CONSTRUCTIONS

Thesis for the Degree of M. A.

MffiHlGAN STATE UNWERSETY

AGAVNI ZAMBAK YERAMYAN
1972

THE ACQUISITION AND COMPREHENSION OF
DIRECT/INDIRECT OBJECT

CONSTRUCTIONS

BY

Agavni Zambak Yeramyan

A THESIS

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

MASTER-OF ARTS

Department of Linguistics and Oriental
and African Languages

1972

ACKNOWLEDGMENT S

I would like to express special appreciation to
Dr. Julia S. Falk. Without her constant encouragement,
assistance and inspiration, this thesis could not have
been achieved.

Appreciations are extended to Profs. Seok Song and
John Ritter for their helpful comments.

The enthusiastic cooperation of Miss Brooke
Bartholomae and the children in her class, at the Family-

Child Research Institute, is also gratefully acknowledged.

ii

TABLE OF CONTENTS

Page
LIST OF TABLES. . . . . . . . . . . . . . vi
Chapter
I. INTRODUCTION . . . . . . . . . . . . . 1
II 0 BACKGROUND 0 O O O 0 O O O O O O O O O 7
2.1. Linguistics and Child Language
AcquiSition O O O O O O O O O O O 7
2.1.1. Linguistic Universals. . . . . 8
2.1.2. Competence . . . . . . . 9
2.1.3. Overgeneralization. . . . . . 10
2.2. Psychology and Child Language
Acquisition. . . . . . . . . . 11
2.3. Theories of Child Language
Acquisition. . . . . . . . . 12
2.4. Summary of Studies Relevant to
the Present Thesis . . . . . . . . 16
2.4.1. Active-Passive Sentences . . . . 16
2. 4. 2. Datives. . . . . . . . . . 19
III. EXPERIMENTAL DESIGN. . . . . . . . . . . 20
3.1. Introduction. . . . . . . . . . . 20
3.2. subjeCtS O O O O O O I O O C O O 22
3.3. Interview Procedure . . . . . . . . 22
3 O 3 O l O InterViews O O O O O O O O 22
3 O 3 O 2 0 Toys 0 O O O O O O O O I 27
3.3.3. Instructions. . . . . . . . 27
3.4. Test Sentences . . . . . . . . . . 28
3.4.1. Variables Involved. . . . . . 28
3.4.2. List of Sentences .. . . . . . 28
3.4.3. Order of Presentation. . . . . 30
3.4.4. Recording of the Data. . . . . 31

iii

Chapter Page

IV. RESULTS AND DISCUSSION. . . . . . . . . . 32
4.1. Introduction . . . . . . . . . . 32
4.1.1. Correct and Incorrect
Responses. . . . . . . . 32
4.1.2. Possible vs. Impossible
Sentences. . . . . . . . 33
4.1.3. Reversibility . . . . . . . 35
4.1.4. General Trend in the Results . . 35
4.2. Linguistic Variables . . . . . . . . 37
4.2.1. Word Order . . . . . . . . 37
4.2.2. Reversibility . . . . . . . 41
4.2.3. Role of Features . . . . . . 42
4.3. Nonlinguistic Variables . . . . . '. . 43
4.3.1. Age . O O . O O C O . C 43
4.3.2. Other Nonlinguistic
Variables. . . . . . . . 47
v. VARIABLES . . . . . . . . . . . . . . 48
5.1. Introduction . . . . . . . . . . 48
5.2. Experimental Design . . . . . . . . 48
5.2.1. Subject . . . . . . . . . 48
5.2.2. Interview Procedure . . . . . 49
502.3. TOYS o o o o o o o o o o 50
5.2.4. Other Differences. . . . . . 50
5.3. Comparison of Results and Discussion. . . 50
5.3.1. Introduction . . . . . . . 50
5.3.1.1. Correct and Incorrect
Responses . . . . 52
5.3.1.2. General Trends . . . 53
5.3.2. Variables . . . . . . . . 53
5.3.2.1. Word Order . . . . 53
5.3.2.2. Reversibility and
Semantic Features . 55
5.3.2.3. Age . . . . . . 57
VI. SUMMARY. 0 O O C O O O O C O O C C O 61

Chapter Page

APPENDICES
A. Correct, Reversed and Paired Responses for
Child in the Sample ' ° ° ' ‘ ’ ° ° ' 63
B. A Sample Interview . . . . . . . . . . 64
BIBLIOGRAPHY . . . . . . . . . . . . . . 66

LIST OF TABLES

Table Page

1. Family and Socio-Economic Background of the
Children in the Sample Listed in Order of

Aqe. O O O O O O O I O O O O O 23
2. Summary of Test Sentences . . . . . . . 31
3. Total Number and Percent of Responses . . . 34

4. Total Number and Percent of Responses for
each Subject Ordered According to Age in
Terms of the Features of the Indirect
Object. . . . . . . . . . . . . 44

5. Summary of the Results of the Pilot Study
with Yanna . . . . . . . . . . . 51

6. Comparison of the Correct Responses of Three
Different Age Groups with Respect to Word
order I O O O O O O O O O O O O 58

7. Comparison of the Correct Responses of Three
Different Age Groups with Respect to
Indirect Object Features Including Both
0-1 and 1-D Object Order. . . . . . . 58

CHAPTER I

INTRODUCTION

This study is concerned with several aspects of
the acquisition and comprehension of Direct/Indirect object
relations. Most studies in the past on child language
acquisition, have focused their attention primarily on
the production and not the comprehension of utterances.
These are usually observations of spontaneous speech or
elicited information, often characterized by data col-
lection techniques which record all utterances made by the
children including false starts, repetitions and other
distortions. Among others, Weir (1962), Braine (1963),
Bellugi and Brown (1964» Miller and Ervin (1964), McNeill
(1966), Klima and Bellugi (1966), Gruber (1967) and Bloom
(1970) have contributed in varying amounts to this liter-
ature.

Production is only a partial indicator of a speaker's
knowledge of his language. Studies such as those cited
above have not been fully successful in capturing the
linguistic competence of children, since they posit for

the child's grammar only those aspects of language that

the child actually produces. Yet, particularly for children,
production is, at early stages of development, severely
limited by such factors as memory span and neurophysiolog1ca1
development. As McNeill points out, the results of production
are easy to observe, but it is not always obvious what the
observations mean. The fact that the child systematically
leaves out some grammatical formatives from his speech, for
example, does not necessarily signify that he does not

have these forms in his grammar; the child, McNeill states,
may in fact exclude them because of limitations of memory
Span or other factors in actual behavior which interfere
with speech (McNeillzl970, pp. 11-12). Both production

and comprehension are instances of performance, but they

are not equally subject to the same nonlinguistic factors
(either in quantity or in kind) and therefore the investi-
gation of comprehension may provide evidence of competence
which is not reflected by production.

Noam Chomsky notes that one cannot determine compe-
tence without studying performance. It seems that, as
mentioned above, studying comprehension provides a clearer
image of a child's internalized rules than a recorded sample
of his utterances. Chomsky also holds that the description

of competence has the greatest psychological relevance

because it is concerned with structures that a person has
mastered and internalized whether or not he or she uses
them (Chomsky:1970, pp. 43-44).

Studies of comprehension are more fruitful particu-
larly with younger children, such as two or three-year-olds,
since their speech is in the early stages of development
and their comprehension is far ahead of their production,
as shown by N. Chomsky and many other researchers. Berko
(1958), Slobin (1966, a), Shipley, Smith and Gleitman
(1969), Bever, Mehler and Valian (1967), Carol Chomsky
(1969) and Fischer (1971) provide studies of comprehension
with the aim of understanding children's linguistic compe-
tence.

Most researchers in the past, whether they have
investigated the competence or the performance of the
speaker, have generally worked either with the first and
second years of development or with the ages where most
of language is already acquired (five years and above).

The experiments that involve the early ages have dealt
primarily with phonological develoPment, the holophrastic
stage, and two-word utterances (pivot-open class etc.)

(see Weir:1962, Klima and Bellugizl966, McNeill:1966 etc.).

There have been a few longitudinal works such as

Bellugi and Brown (1964) covering approximately the years

two to four; but the rest have been mainly studies of school
children and have examined word formation (Berko:1958), pas-
sives (Slobin:1966 a), and other complex grammatical struc-
tures such as the verbs 'promise' and 'ask' (C. Chomsky:1969).
With the exception of a few studies (such as Fischerzl971),
the ages of three and four have been neglected by most
psycholinguists. One reason why this age has been generally
ignored may be explained in terms of the difficulty of
eliciting information from the child. In the present work
we found that children of about four years of age are very
restless and possessed with an extremely limited attention
span. The attention span of children of one to three years
is by no means any greater; but the work done with them has
involved observing Spontaneous production as opposed to
testing comprehension. Children in the holophrastic and
very early telegraphic stages produce a limited range of
utterances, so spontaneous production provides a limited,
but analyzable set of data. Four-year-olds have acquired

a more complex grammar and even though not all of their
grammar is reflected in production, what is produced is
both highly varied in kind and great in quantity. They
must, therefore, be examined in a controlled situation
whereas younger children may merely proceed with normal

activities while an investigator records whatever they

happen to say. These intermediary years, however, should
also be considered important in language development since
it is in these ages that most syntactic structures are
acquired (see Slobinzl966 b and Bellugi and Brown:l964).

A large number of studies on child language acqui-
sition in the current literature have adopted the theo-
retical assumptions of transformational grammar as the basis
for the interpretation of the results. Although other
linguistic theories such as tagmemic or stratificational
grammar could be utilized, transformational grammar has
had particular appeal in that a) the results of earlier
works based on this framework have been insightful and
encouraging and above all, b) the various concepts such
as the relation of syntax to semantics, deep and surface
structure and linguistic analysis of competence are found
to be applicable and relevant to the study of child language.

As we shall see in the following chapters, the
role of semantics in the linguistic develOpment of the
child is indispensible. Furthermore, a theoretical study
of the deep structure of the constructions used in the
experiment is essential for an understanding of the child's
persistent preference of certain structures over others.

The benefits of studying comprehension as opposed to pro-
duction have been indicated above as well as in other sources

in the literature.

The constructions used in this experiment consist
of simple commands containing the verb to give, a direct
and an indirect object. It has been found that sentences
such as Give the book to the girl, Give the girl the book,
*Give the girl to the book and *Give the book the girl elicit
different responses at different stages in development and
that the order of words and reversibility of the construc-
tions as well as semantic features of nouns play important
roles in children's comprehension. These different vari—

ables are discussed in detail in the following chapters.

' D
u;
_. wry

CHAPTER II
BACKGROUND
2.1. Linguistics and Child Language
Acquisition

In recognizing linguistics as a "branch of cognitive
psychology," N. Chomsky (1968) has clearly indicated the
need for the integration of the two disciplines which had
been considered independent ever since the rise of struc-
tural linguistics. Although this recognition was theo-
retically accepted by many linguists, they still viewed
psychological studies of language as devoid of linguistic
information and irrelevant to competence. Fischer (1971,
p. 1) notes that recently linguists have been realizing
more and more how much the structure of language is influ-
enced by limitations of human thought and she cites works
of linguists such as Ross (1967), Klima (to appear}, and
Jackendoff and Culicover (1970) that bear references to
concepts in child language acquisition.

Language development is the area where linguistics
and psychology are most closely interrelated. Bever (1970
a, b) has tried to show that a child's treatment of language
at a particular stage in development is highly related to

the ways he or she interacts with the environment. He has

7

 

 

la
1
3

E .-r .id- at u.
_ i.

found, for example, that at a particular stage, the child
makes the same kinds of hypotheses about mathematical con-
cepts (judging numerosity by length) as about grammatical
relations (such as word order in a sentence). Furthermore,
some important notions (such as linguistic universals and
a speaker's competence) which form the core of transfor-
mational theory can be refined through research in child

language acquisition.

2.1.1. Linguistic Universals

A major interest in child language acquisition
stems from the hypothesis that the universal aspects of
language may in fact be innate to the child. If this
hypothesis is correct, then the study of child language
acquisition will lead to further insight about universals.

Even before the advent of transformational grammar,
linguists such as Jespersen (1922) had studied child
language acquisition. Jespersen was particularly inter-
ested in historical aspects, that is, the role children
played in causing sound change. Investigations more
relevant to current issues in language acquisition were
performed by linguists of the Prague school who were also
interested in language universals. Among them, Jakobson in
particular found that the order in which children acquire
language is the reverse of the order of loss of language

in aphasia (Jakobson, 1942).

In grammar as well as in phonology, it does not
seem unreasonable to hypothesize, then, that the degree of
universality of a linguistic phenomenon (e.g. notions,
construction or rule) is directly proportional to the time
of occurrence of that phenomenon in the child's development.
That is, the more universal a phenomenon, the earlier is
its appearance in the linguistic development of the child.
Evidence for such a hypothesis must be determined by a
comparison of the order of acquisition in a number of

children learning different languages.

2.1.2. Competence

Another source of the linguist's interest in the
child is the fact that studying the development of language
in different stages makes it possible to analyze competence
at these different stages and can serve as a cue to formu-
lating base structures.

Fischer (1971, p. 2) rightfully argues that gener-
ative linguists for a long time have been content to
idealize the child as a little black box (LAD) and have
been interested only in the outcome of this black box.

They have left the processes that go on inside it to the
psychologists. As mentioned above, recent researchers
such as Bever (1970 a, b) are convinced that one cannot

divorce the outcome from the processes that go on in the

 

lO

mind. A person should be viewed as a whole, rather than

being examined in isolated parts. An examination of the

linguistic competence of the child at different stages in
development can provide substantial information about the
processing of linguistic information in the mind.

Attempts to study grammatical development of the
child at different stages of acquisition, for example, have
shown that combinations of words and parts of words in child
speech seem to be systematic rather than random, productive
rather than merely imitative or memorized. The child's
language is structured from the start; it soon assumes a
hierarchical structure and tends to be regular. One thing
that should also be kept in mind is that the structures
change in the course of development and they do not always

correspond to adult structures (Slobin: 1971 a, p. 3).

2.1.3. Overgeneralization

A phenomenon most commonly observed by psycholin-
guists (e.g. Ervinzl964, Slobin:l966 b) is that of overgener-
alization (sometimes referred to as overregularization).
This is a particular stage where the child consistently
regularizes grammatical structures such as inflectional
endings, irregular plurals and past tenses (e.g. ill use

such forms as foots for feet,goed for went). Curiously,

 

however, just prior to this stage the child is heard using

a!

11

the correct forms and seems to have mastered them. This
process, seemingly regressive in nature, is easily explained.
Since the words that receive irregular inflectional endings
(such as the strong verbs) are the ones most commonly used
in the language, the child hears them first and memorizes
them at the beginning. Later, when he hears more words with
regular endings, he formulates a rule and imposes this rule
on all the words in the same category. Overgeneralization
is a relevant process for linguists as well as psychologists
in that it clearly indicates the time and the way a speaker
formulates linguistic rules.
2.2. Psychology and Child Language
Acquisition

For psychologists, language is simply an aspect of
human behavior and they are interested in child language
acquisition as much as they are interested in any other
kind of behavior. Recently, psychologists such as McNeill
and Slobin have realized the uniqueness of language be-
havior and have concentrated on language alone. They have
by and large been influenced by the principles of trans-
formational grammar; for it was transformational grammar
that revealed the complexity of language and its issue of

linguistic universals became a challenge to psychologists.

12

As mentioned in section 2.1.2., linguists had accepted
the task of examining the output of the black box, but had
left the study of the processes inside the box to psycholo-
gists. Until the recent attempts of Bever and others to
integrate the two disciplines, Fischer (1971, p. 2) shows
that psychologists were then left with questions as to why
grammar is not learned all at once, and what kinds of things
are easier and harder for the child to learn. Some of the
recent psychological theories about child language acquisition
which are relevant to the present work are discussed in the
next section.

2.3. Theories of Child Language
Acquisition

Since the development of transformational grammar,

a number of different theories about child language acqui-
sition have been proposed. Among these, within the framework
of transformational grammar, the most prevalent ones have
been those which advocate a correlation of some sort between
the complexity of linguistic structures (i.e. those postu-
lated by the grammarian) and ease and time of learning.
Beyond this general assumption, the views of psychologists
differ as to the degree and other aspects of the correlation.
Fischer calls the strongest of these theories the "Deri-

vational Theory of Complexity" (DTC). According to DTC.

13

"the complexity (hence difficulty) of a sentence is directly
proportional to the length of its derivational history"
(Fischer, 197l:p. 70). According to Fischer, DTC fails as
a general account of complexity when different sorts of
transformations such as relative clause reduction, adjective
shift, particle movement etc. are taken into consideration.
Fischer also mentions a modified version of DTC by

Brown and Hanlon (1968) which proposes a notion of 'cumulative'
complexity. In this theory

if a structure in its derivation requires N transfor-

mations to be derived, and if one adds one more rule

to this derivation, then the resulting structure 5'

will be acquired later than S, (Fischer, 197l:p. 70).
According to the predictions of this modified version, an
affirmative sentence, for example, will be acquired before
the corresponding negative, and this same affirmative
declarative will be acquired before the corresponding yes-
no question; but the prediction will not be extended as to
the relative order of negatives with respect to yes-no
questions since neither is "included in the derivation of
the other." This allows "the cumulative theory of complexity
to avoid one pitfall of DTC since it is not assuming that
one transformation is equal to another in contributing to
complexity" (Ibidzp. 71).

Fischer criticizes Brown and Hanlon for restricting

their discussion primarily to the same kinds of transfor-

mations as were used in the original DTC studies, namely

.‘L—ofi. a _.__-

".4 ‘fl .

 

,. magi
.’

14

question, negation, tag question formation and tag transfor-
mation. Thus, Brown and Hanlon are subject to the same kinds
of criticisms as the criginal DTC. "For the constructions
considered, the predictions of the theory are borne out,

but this may be due to reasons different from those which
Brown and Hanlon suggest" (p. 71). Furthermore, their
theory like all others does not account for factors such

as reversibility (to be discussed in the next section),
directional adverbials (as opposed to datives), and bene-
fectives (again as opposed to datives). "These factors
present serious difficulties for any such theory and they
have yet to my knowledge to be evaluated" (p. 72).

The basic assumption of the cumulative theory of
complexity, namely that children learn first aspects of
grammar that are closest to their base structure, is a
reasonable one as a starting point until it can be invali-
dated by further research. Fischer notes, however, that
DTC, by implicitly assuming that the child is possessed
with adult base structures from the beginning, does not
allow any changes in the base nor in the transformations once
they are formed. Only the addition of new rules seems to
be allowed. This makes the theory too rigid. It is unlikely
that children would be capable of only adding to their syn—
tactic component, since it has been hypothesized that they

are capable of restructuring phonological rules. In studies

15

on phonology, both Kiparsky (1968) and Halle (1964) mention
that children differ from adults precisely because children
can reformulate or simplify but adults can only add.

Thus we can relate the concepts of rule addition and
simplification to adult and child language respectively.
The typical form of rule addition is the borrowing of
rules among adults; simplification typically occurs in
the learning of language by children (Kiparskyzl968, p.
195).

The ability to master a language like a native, which
children possess to an extraordinary degree is almost
completely lacking in the adult. I propose to explain
this as being due to deterioration or loss in the adult
of the ability to construct optimal (simplest) grammars
on the basis of restricted corpus of examples. The
language of the adult--and hence also the grammar that
he has internalized--need not, however, remain static:
it can and does in fact change. I conjecture that
changes in later life are restricted to the addition of
a few rules in the grammar and that the elimination of
rules and hence a wholesale restructuring of his grammar
is beyond the capabilities of the average adult (Halle:
1964, p. 344).

Although their claim primarily refers to phonology, there
is no reason to believe that this ability would not be
available in syntax as well.

The present work as well as all the recent research
based on generative principles claims that the child's deep
structure is different from the adult's and accepts the
further modified assumptions of Bates (1969). Bates contends
that the structures which the child acquires first or which
are easiest for him or her, are closest to the child's

deep structure. According to Bates, the child can restructure

«J,
‘H_'

16

the base component of the grammar and therefore the possible
transformations, in order to account for new generalizations
more efficiently.
2.4. Summary of Studies Relevant to
the Present Thesis
2.4.1. Active-Passive Sentences
With the exception of Fischer's dissertation which
will be mentioned in Section 2.4.2., experiments most closely
related to the present one have dealt with passive and
negative constructions. The most important work in this
area was by Slobin (1966 a, b) who studied children between
five and eleven years. In his experiment the children had
to judge the truth of sentences against pictured sqenes.
Similar experiments were also performed by Turner and
Rommetveit (1967) with children four to nine years, and
by Bever, Mehler and Valian (1967) with children two to
four years. Turner and Rommetveit required the children
to describe the pictures as well as to judge their truth
or falsity. The results of these later studies corresponded
with those of Slobin in that:
a) passive sentences required more time to evaluate
and were less accurately evaluated than active ones;
b) negatives required more time and were evaluated less
accurately than affirmatives (either active or passive);

and

17

c) passive negatives were the hardest to evaluate and
were less accurately evaluated.

These studies advocate a correlation between the
length of time it takes to evaluate a picture and ease of
comprehension. It is important to note that passive af-
firmatives were easier to comprehend than active negatives,
although passives seem to be syntactically more complex
than actives. In terms of comprehension, the most important
variable seemed to be whether the sentence was affirmative
or negative. It is also evident in studies of comprehension,
however, that more than syntax is involved in processing
everyday sentences. In fact, Slobin's experiment indicates
that the role of syntax can be dramatically altered by
manipulating the semantics of the situation in such a way
as to eliminate the distinction between active and passive
in terms of difficulty of comprehension. That is, there were
some cases in which passives were no harder to understand
than actives. These were sentences (called 'non-reversible'
by Slobin) where it was clear from the semantics which noun
is the subject and which the object. (An example of a non-
reversible sentence would be: the girl is picking straw-
berries, where *strawberries are picking the girl is un-
grammatical.) The passive forms were harder than their

corresponding active forms only when either noun could be

 

18

the subject or object (called 'reversible' by Slobin). (An

example of a reversible sentence would be: the boy is

 

chasing the girl, where the girl is chasing the boy is also

 

a grammatical sentence.) Apparently, then, part of the
difficulty in understanding passives lies in determing the
subject noun. Non-reversible passives are in a sense de-
termined by the semantics, hence the hearer's task is simpler
and his choices for subject are 'defined'.

Bever et. a1. (1967), who also tested for reversi-
bility using younger children (two to four years), compared
their work with Slobin's results. The difference in age
had a considerable impact on the outcome. At early ages
reversible and non-reversible situations are treated alike
when passive sentences are used. Non-reversibility did
not help or hinder performance at three years, although
even children of two to four years comprehended some passives
more than half the time. By four years, comprehension of
reversible passive sentences began to regress. At this time
the children performed similarly to Slobin's and Turner
and Rommetviet's subjects.

Like Slobin, Bever et a1. maintain that children
utilize strategies in understanding the structure of sen-
tences. They argue, for example, that sentences are assumed
to be semantically coherent and to describe an actor, action,

and recipient of an action. Such information is distinct

19

from the strictly grammatical information about the underlying
relations in a sentence--subject, verb and object (McNeill:

197o,p. 124).

2.4.2 Datives
Probably the most relevant study for this thesis is
the work of Fischer which has been referred to several

times in the previous sections: The Acquisition of Verb-

 

Particle and Dative Constructions (1971, Doctoral Disser-
tation, MIT). By the use of pictures and oral discussions
with children of three and four years, Fischer investigated
some of the processes involved in the acquisition of these
structures. Her work includes a comparison of the results
in the two ages as well as parameters such as stressed
versus unstressed direct object pronouns, to- versus for-
datives, directional adverbials versus datives. Her com-
parison of the two ages is designed to replace a longitudinal
study.

Her results show that, over time, the child pays
more attention to semantic plausibility, pays somewhat less
attention to order and learns to use surface structure cues
to deep structure grammatical relations. Other details of
her results will be mentioned in relation to the results of

the present study.

CHAPTER I I I
EXPERIMENTAL DES IGN

3.1. Introduction

Although Fischer's dissertation (1971) is most
relevant to the present work, it was not used as a guideline
due to its unavailability until after research had been
completed. The works of Slobin (1966 a) and Bever et al.
(1967) have been most influential in providing motivation
for this project. Since reversibility of subject and object
play a fundamental role in comprehension and acquisition
of some aspects of syntax (as indicated in their results),
it is assumed that the reversibility of direct and indirect
objects plays an equally important role. As in the acqui-
sition of passives and negatives there may be a progression,
a series of stages, perhaps correlated with age.

The main goal of this thesis, then, is the determi-
nation of those aspects of direct/indirect objects which
are most prevalent in the child's comprehension. Theoretical
questions that arise from the study include:

a) Whether a change in the syntactic order of the objects
{the inversion of direct and indirect objects) affects

comprehension; if it does, which construction (the

20

21

direct/indirect or indirect/direct order) is easier
for young ch1ldren to understand and which is ac-
quired earlier.

b) Whether the notion of reversibility plays a signifi-
cant role in learning and comprehension of direct
and indirect objects, or whether it is confined to
the use of passives as shown by Slobin and others.
It is hypothesized that sentences with reversible
object relations (where the nouns for direct and
indirect objects can be interchanged without making
the sentence ungrammatical) will be harder to under-
stand and will be acquired at a later stage than
non-reversible ones.

c) How the child will respond, if at all, to semantically
incongruent sentences such as *Give the girl to the
candy. .Although according to McNeill (1970), semantic
anomaly should not play any role for young children
(especially below five years), such a hypothesis was
not made in this experiment.

The study also attempts to determine which of the
three variables above plays the greater role in the child's
comprehension at this particular stage.

Before the actual experiments for this thesis were
conducted, a pilot study was carried out with one of the

children in the group, four months prior to the major

 

 

22

experiment. The results of the pilot work provide a longi-
tudinal study and a basis for comparison of constructions in
different ages. The details of the pilot work are discussed

below in Chapter V.

F
3.2. Subjects 1
The sample consisted of ten children ranging from

3-5 (that is, three years and five months old) to 4-2, at

 

the beginning of the project. They all attended the preschool
at the Family-Child Research Institute at Michigan State
University. As the chart in Table 1 indicates, the socio-
economic background of the children represents a cross section
of the Lansing community. There were seven girls and three

boys in the sample.

3.3. Interview Procedure
3.3.1. Interviews

The children were interviewed at the institute during
their 'free play' time in twelve sessions over a two month
period, from January to March 1972. Since the preschool
was used as a laboratory for research by many students, the
children were accustomed to interviews (as well as obser-
vation) during their play and considered this as a part of
their daily activity. Each session lasted approximately

sixty minutes and an average of five children were interviewed

23

.mfios um smeMDH on muswomxm wEoma

 

 

 

 

mo\om\a m
mm\ma\s s
mm\mm\m s
em\om\e A
~m\ma\m m mmceanam
mmaSmmsom Haoo H» N\H a mmmmmcsma Hm\ma\n Q nonuoz
masseuse on mm msmnmaa Hm\ma\k a genome
cmmaaoflz mm\sm\m a a «mama
mm\ma\oa E
ms\m~\m m mmaAHnAm
emazmmoom mm mfioswaaH vm\oa\a 3 negro:
.o.m.z nonwomoum arm mommno mm\ma\m 3 Heaven
guesses: mm\ma\m 3 A azzas
oa\AH\m s
km\m\m s mecflanam
cohonEmsD mpmnm read smmfinoaz mm\om\a Q “ergo:
nmxno3 coeuoounmsoo mpmum auaa Atmono>flpv mv\mm\m Q nmnumm
cmmeroflz mm\m\m n A razoe as
Hk\eH\HH L
adage nmumom .. A mmqnaaflm
mmfismmoom mm smmflnowz oe\m\m 3 umnuoz
.Hasm Laomixooo mNNAA mm SAEBH sv\mm\m 3 Heroes
cameras: mm\m\m 3 s zmoeg
coaummoooo coaumospm momam .m when .m worm xmm
.mmm mo Hmpuo ca pmumfla
mamamm onu ca smuoawno or» mo pcsoumxomn anOQOUEIOHUOm paw MHHEmm .H wands

24

 

 

 

 

 

 

moxeaxm m mmqeanem
mme3mmaom mm ameflaoez mskaxk 3 sensor
.D.m.z HOmmmmoum and smuH mm\em\m 3 umnumm
sameness noxmaxma 3 m 42H:
msxmxm s mecasnem
mm cameras: se\oa\oa 3 sensor
cmEnHmmmm owns mm cmmflnowz vv\m\m 3 Hmnumm
sameness mm\ss\m 3 m Hozmz
mm\om\n

me\OH\~H A
mmxma\e L mmaaanam
mma3mmoom «m savanna: ov\w\aa 3 Hespoz
.ommm mEHmHo mosmuomsH .Haoo mum m demarcaz ov\ma\m 3 Hmnumm
cmmaroaz mm\m\m 3 a zaHemHmmo

mmxmxm s

Hm\oa\h E

mm\H\a e
mm\m~\a e meananem
Lameaumumme3mmsom am cameras: ~m\m\m 3 sensor

z>o .
Hmeoeumo ruammm oeanam .mmz .mm .oa Emmerofls om\mm\m 3 umnumm
essences mm\ma\m 3 A meemzza
woos mmnfianwm
Hoorom amoecnomu ca usm©5pm hapcmmmum .Hmmsmm mv\mm\m Q Hmauoz
Apoqowusme pocv Hmsumm
cemenoaz mm\em\e n s wmmom
coaummzooo soaumosom woman .m mama .m worm xmm

A.ucouv .H manme

25

 

 

 

 

mo\mm\m a mmcflanem
mea3mmaom m2 usagesm kmxmxmfl 3 Amauoz
.D.m.2 Hommmwoum .oommd and mxmmunmz vm\h\aa 3 Hmnomm
guesses: nm\ma\ma 3 L mqqmoz

coaummsooo coflumootm momHm .m when .m momm xmm
x.ucoov .H magma

26

at each session. Due to the young age of the subjects, the
interviews were conducted in the form of a game and the
length of each was determined by the interest span of the
particular child that particular day. The range was from
three to fifteen minutes. A special effort was made to end I
the interview before the subject was bored so that he or 7
she would be willing to 'play' again the next time.

Prior to the actual interviews, the investigator "

 

spent three hours on each of five afternoons with the chil- ' i

1 I

dren in order to become acquainted with them and to acquaint
them with the interviewer and the tape-recorder. Since the
acquaintance period coincided with the beginning of a new
term, the interviewer was introduced to the children as one
of the helpers to the main teacher.

The experimental sessions started only after it was
ascertained that each child felt comfortable in the presence
of the interviewer and was willing to participate. The
subjects for a particular session were selected sometimes
with the question: "Who would like to play with me first
(or next) today?" which created a feeling of competition
among the children and rendered them more than willing to
c00perate. At other times, when only one child was free to
participate at a given period (i.e. when he or she had just

finished an activity) the child was directed to the place

27

of the experiment with the words: "Now it is your turn to
play, (child's name)." Each subject was interviewed

alone in a room with prearranged toys.

3.3.2. Toys

Toys were preferred to the use of pictures because
the children showed more enthusiasm in playing with differ-
ent toys than in seeing pictures. The toys were carefully
selected to be prOportional in size to real objects and were
changed after every two sessions in order to retain the
children's interest. The toys always included dolls, animals
and inanimate objects, such as a truck, fruits etc. It was.
also part of the daily procedure to ask each subject the
names of all the toys in the play room and to include in the
game only those toys that the particular child knew the

names of.

3.3.3. Instructions

The following instructions were given at each inter-
view: "Now we are going to play a game. I will close my
eyes and will tell you to give one toy to another toy. Then
I will Open my eyes and see if you gave the right toy."

Then a test sentence such as "Give the pencil to the boy doll"

 

would be introduced to see whether the child comprehended the
instructions correctly. The experimenter closed her eyes

only for the first few commands, and then only very briefly.

28

Although in the introductory sessions the [+ human] toys
were referred to as "boy doll" or "girl doll," later, for
reasons to be discussed in section 4.1. below, they were

used interchangeably with "boy," or "girl," "mommy" etc.

3.4. Test Sentences
3.4.1. Variables Involved
Before the interviews started, typical sentences
were selected to include each variable. The variables
discussed in section 3.1. are summarized below with examples
given for each type.
a) Syntactic variable--Object Inversion
(1) Give the book to the boy.
(2) Give the boy the book.
b) Semantic variablel--Reversibility
(3) Give the boy doll to the girl doll.
(4) Give the girl doll to the boy doll.
c) Semantic variable2--Anoma1y
(5) *Give the truck to the cow

(6) *Give the cow to the truck.

3.4.2. List of Sentences
It was observed during the pilot study that the child
responded differently to objects (direct or indirect) which

had the semantic feature [+ human] as opposed to [+ animal].

29

Because of this distinction, the original binary division
of L: animate] had to be further distinguished, thereby
necessitating the use of the three features: [- animate],
[+ animal] and I+ human]. Accordingly, all the sentences
in the experiment were based on the following eighteen
construction types: (Also see the summary chart in Table
2.)
(l) D[-animate] I[-animate]
Give the book to the couch.
(2) I[-animate] D[-an1mate]
Give the couch the book.
(3) DI-animate] I[+human]
Give the pencil to the girl.
(4) I[+human] D[-animate]
Give the girl the pencil.
(5) D[-animate] I[+anima1]
Give the bone to the dog.
(6) I[+anima1] D[-animate]
Give the dog the bone.
(7) D[+human] I[-animate]
Give the girl to the pencil.
(8) I[-animate] D[+human]
Give the pencil the girl.
(9) D[+human] I[+human]

Give the girl to the boy.

 

30

(10) I[+human] D[+human]

Give the boy the girl.
(11) D[+human] I[+anima1]

Give the boy to the tiger.
(12) I[+anima1] D[+human] 1

Give the tiger the boy.
(13) D[+animal] I[-animate]

Give the dog to the bone.

 

—" m ‘ m..§i"‘-v_ .‘nl-
—.

(14) I[-animal] D[+animate]
Give the bone the dog.
(15) D[+animal] I[+human]
Give the dog to the boy.
(16) I[+human] D[+animal]
Give the boy the dog.
(17) D[+animal] I[+anima1]
Give the tiger to the elephant.
(l8) I[+anima1] D[+animal]

Give the elephant the tiger.

3.4.3. Order of Presentation

No particular order in the presentation of different
constructions was used. The first few times at the beginning
of the game, only semantically logical and non-reversible
sentences were presented, although their order of presentation
was not systematic. Later in the interview, all sentence

types were mixed and presented in a random fashion.

31

 

 

 

 

 

 

DIRECT OBJECT

 

 

 

 

 

 

 

 

 

 

Table 2.* Summary of test sentences.
Indirect Object
give [-animate] [+human] [+animal]
[+to] [-to] [+to] [-to] [+to] [-to]
{-animate] (l) (2) (3) (4) (5) (6)
[+human] (7) (8) (9) (10) (ll) (12) i
[+anima1] (13) (14) (15) (16) (17) (18) '

 

 

*Numbers refer to the examples of each sentence type given.

[+to]: Direct/Indirect Object constructions.
[—to]: Indirect/Direct Object constructions.
3.4.4. Recording of the Data

The entire session for each interview was tape-
recorded. The instructions were given as described in section
3.3.3. The performance of the children was recorded on the

tape by the use of various code words in accordance to the
child's nonverbal response. For example:
a) "Very good"--for correct responses;
b) "O.K."--for reversed responses; and
c) "Match"--for paired responses.
All the comments made by the child, and the dialogues between
the investigator and the child were also tape-recorded.
After the interviews, the experimenter decoded the tape and

transcribed the results for each subject on separate data

sheets. (See Appendix B for a sample transcription.)

CHAPTER IV
RESULTS AND DISCUSSION

4.1. Introduction
4.1.1 Correct and Incorrect Responses

The data were analyzed in terms of correct and in-
correct responses. Responses were considered correct only
when the children carried out, exactly and without hesitation,
the instruction reflected in the command. Repetition of the
command by the investigator had no effect on the responses;
that is, once the children interpreted the sentence in a
certain way, they did not change their interpretation.

Most incorrect responses were in the form of reversals:
confusion of direct and indirect objects. For example, to
the command Give the cow the car, the child would take the
cow and give it to the car. In some instances the two objects
were paired without any indication as to which was perceived
as direct object and which as indirect. For example, to the
command Give the duck the chicken, the child would take the
duck in one hand and the chicken in the other and would hold
them close to one another. In the computation of the data
only reversals were counted as the incorrect responses.

Paired items were not included as incorrect responses since

32

33

they did not indicate what the child comprehended. Paired
responses are, nevertheless, included in the discussion of
the results. Since the reversals are the only incorrect
responses counted in the data, the two terms (i.e. reversals
and incorrect responses) will be used interchangeably in the
rest of the paper.

The summary chart of the results in percentages for
the entire group is shown on Table 3 (for more detailed tables

see the appendix).

4.1.2. Possible vs. Impossible Sentences

In this paper, the notion of implausibility (or
anomaly) is used to define those sentences which are se-
mantically illogical for real life situations as well as for
toys. A typical implausible (impossible, anomalous) sentence,
for example, is Give the truck to the pig or Give the pencil

to the fence where the direct and indirect objects in the

 

sentences violate the cooccurrence restrictions of English.
Since the experiment for this study was restricted
to the use of toys which are, in a sense, inanimate objects,
one may question the validity of using the semantic features
[-animate], [+animal] and [+human]. As previously mentioned
in section 3.4.2., however, the pilot study showed that the
child did in fact respond differently to animal toys, dolls

and other toys such as telephone, truck etc. This early

 

34

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 3. Total number and percent of responses.
Indirect Object
[-Animate] [+Animal] [+Human]
[+to] [-to] [+to] [-to] [+to] [-to]
c R c R c R c R c R c R
"a?
4,; Total 29 1 11 28 22 1 13 19 31 4 21 6
.5
’7? Total 97 3 28 72 96 4 41 59 89 11 57 43
4:8)“
‘5'“ Total 29 1 9 19 36 l 10 42 29 1 9 20“
oi
4.)
81% Total 97 3 3o 70 97 3 19 81 97 3 31 69L
"gt—a
g Total 30 5 9 20 24 13 7 24 28 2 19 26
:13
+ Total 86 14 31 69 65 35 27 77 93 7 42 58

 

 

35

work as well as discussions with the children's teacher led
to the conclusion that children treat toys as real objects,
nevertheless recognizing the fact that dolls and animal toys
cannot go to places on their own and therefore must be 91223
(or taken) to one another. It was on the basis of this
somewhat paradoxical assumption that the present sentence
types were selected. Hence, sentences that would seem 'im-
possible' to adults, such as Give the girl (doll) to the boy
(doll) were interpreted as normal by the children because
they knew that toys usually do not walk (eat, or act in any

other way) on their own. Sentences such as Give the truck

 

to the chair, on the other hand, were received with surprise

 

and sometimes produced giggles or such comments as "that's
funny."

The exclusive use of the verb to give, for the sake
of consistency, limited the number of plausible sentences in
the data; that is, there were more implausible sentences in

the data than plausible ones.

4.1.3. Reversibility

It was impossible to establish the exact criteria
that the children used for judging the reversibility of
sentences. For the purposes of analyzing the data in this
study, reversible sentences were defined as those sentences

in which both direct and indirect objects have the same

36

semantic features (i.e. both {-animate], both [+animal] or
both [+human]). Examples of reverisble sentences are:

l. a) D[-animate] - I[-animate] e.g. Give the book to

 

the pencil.

 

b) I[-animate] - Dl-animate] e.g. Give the pencil

 

the book.

2. a) D[+animal] - I[+anima1] e.g. Give the dog to the

 

cat.

b) I[+anima1] - D[+animal] e.g. Give the cat the

 

dog.
3. a) D[+human] - I[+human] e.g. Give the girl to the

 

boy.
b) I[+human] - D[+human] e.g. Give the boy the girl.

 

The only exception to this criterion were sentences such as

Give the baby to daddy where the size difference affected

 

reversibility even though both objects have a [+human] noun.
All other construction types not cited in the list above were

classified as non-reversible.

4.1.4. General Trend in the Results
As Table 3 shows, there were four major groups in
the distribution of correct and incorrect (reversed) responses:
a) Very high percentage of correct (C) responses and
very few reversals (R), such as 97 per cent C to 3

per cent R.

 

 

37

b) High percentage of R responses with low C responses,
such as 81 per cent R to 19 per cent C.

c) About one third C responses with two thirds R, such
as 30 per cent C to 70 per cent R.

d) More or less equal distributions of C and R responses,

such as 57 per cent C and 43 per cent R.

4.2. Linguistic Variables
4.2.1. Word Order

Word order in this experiment was found to have an
enormous effect on comprehension. This finding provides
further support for one of Slobin's principles about word

order:

Operating Principle C: Pay agtgntion to the ordgr g:
wordguand morphemes.

Universal C1: The standard order of functor morphemes
in the input language is preserved in child speech
(Slobin:l97l b, p. 348).

 

Sentences with direct/indirect (D-I) order (i.e.
[NP V NP to NPJ) had the highest percentage of accuracy
in comprehension, yielding a mean of 90 per cent correct
responses. Sentences with indirect/direct (I—D) order (i.e.
[NP V NP NP]) had a mean of only 33 per cent correct re-
sponses (see Table 3). These scores support the view that
for children of 3-5 to 4-2:

a) order of words is very important

 

38

b) the [NP V NP to NP] construction is more basic (i.e.

learned earlier) than the [NP V NP NP] construction.

As the low mean of 33 per cent C responses implies,
all construction types, except one, with the I-D object
order had more incorrect responses than correct ones. The
exception was the construction I[+human] - D[-animate] type
with 57 per cent C responses. This type corresponds to
sentences such as Give the boy (doll) the truck or Give the
boy (doll) the apple, where the semantic cues facilitate
comprehension.

The predominance of the D-1 object constructions
at this age was also reflected in the fact that two of the
children (John and Chris) usually interpreted the first NP
after the verb as the direct object before the experimenter
had completed the sentence. For example to the command

Give the cow the dog, Chris picked up the cow and was ready

 

to give it to the other object before the experimenter said
"the dog." This shows that some children do not really
consider any alternative to the D-I construction.

In addition to the order of NP constituents, the
presence of the preposition tg ([+to]) also seems to provide

a clue for the child. To sentences such as Give the banana

 

the corn, some of the children asked "to the?" and after
the exPerimenter repeated the sentence with the same I-D

construction, the child still gave the banana to the corn.

‘1'-

 

. ‘ I“ Luv
‘

39

Instances such as this (which typically occurred With Chris
and Robby) indicate both the subject's confusion when faced
with a sentence Without a preposition and his desire to
adapt such sentences to his own grammar--a grammar which

at this stage of linguistic development includes only D-I
object constructions.

Repetition of the command by the subject (prior to
his or her nonverbal response) was also a common occurrence.
All the children who repeated, with the exception of John,
always inserted the preposition between the two objects
whether or not it was present in the original command. If
one accepts the view of Ervin (1964) and Slobin and Welsh
(1969) that a child can repeat a sentence only in terms of
his or her own grammar, then the above examples further show
the importance of order and the use of the preposition as
a cue. It should be added here that Fischer (1971) has made
extensive use of repetition in her experiments (she asked
the children to repeat) and found it an important tool for
eliciting information.

It was also interesting to find that the pairing or
matching of two objects (mentioned in section 4.1.1.) only
occurred as responses to I-D object constructions. Although
some children never had paired responses, this phenomenon

\

once more indicates that sentences with I-D order or with

40

the absence of the preposition £2 at times caused confusion
for the subjects.

Fischer (1971) used approximately the same age group
(3-9 to 4-2) for some of her experiments. Although her
methods of elicitation differed from the present study (she
used repetition, picture choice, paraphrase and picture
description), her results are in agreement with the present
study in that her subjects responded more accurately to
those sentences with the structure [NP V NP to NP], in both
repetition and choice tasks.

These results therefore suggest a possible rule that
a child unconsciously utilizes: "always give the first object
in the sentence to the second object" or more abstractly
"the first NP after the V" is the direct object and the second
is the indirect." D-I object constructions are in conformity
with this rule (in fact, they served as the data based on which
the child formulated the rule). Thus, with such constructions,
the rule works, but I-D constructions violate the rule and
therefore yield a high percentage of incorrect responses.
The situation is further illustrated by another of Slobin's

principles:

    

Universal D1: Structures requiring permutation of elements
will first appear in nonepermuted form( Slobin:l97l b, p.
352).

 

41

4.2.2. Reversibility

Reversible sentences did not play a major role in
the results for this particular age group. This conclusion
can be quite misleading, however, unless another factor,
plausibility, is taken into account.

Reversibility had no effect in sentences with D-I
object structure. Sentences like Give the tiger to the
elephant resulted in the same frequency of error as ELIE

the shovel to the man. Reversibility had an effect on sen-

 

tences with I-D object structure.

Plausibility of sentences was a factor as important
as reversibility in I-D object constructions. Non-reversible
objects aided the comprehension of plausible I-D object con-
structions. The children always responded correctly to non-

reversible sentences such as Give mommy the baby and only

 

30 per cent of the time to reversible sentences as Give the

girl the boy.

 

In implausible I-D object constructions, the effect
of reversibility was minimally in favor of non-reversible
sentences. Implausible reversible sentences like Give the

corn the cookie (which having {-animate] feature for both

 

objects conforms to the definition of a reversible sentence
given in section 4.3.1.) yielded 28 per cent C responses,

whereas implausible non-reversible sentences such as Give

 

42

the telephone the alligator reached 33 per cent accuracy
(i.e. correct responses).

In one of Fischer's (1971) experiments, where she
tested choice accuracy of pictures with children of approxi-
mately-the same age as the present study, she also found
that reversibility was significant only after the sentences
were divided into semantically possible and impossible ones.
Accuracy for reversible sentences was only slightly more
than chance (56 per cent) but the semantically plausible
non-reversible sentences gave an accuracy of 76.3 per cent.
Those non-reversible sentences which were semantically
implausible, on the other hand, yielded 33.7 per cent accuracy.
Fischer concluded "these results show strongly that children
of this age pay a great deal of attention to semantic cues,
many times superseding contradictory syntactic cues" (p. 106)-

To summarize, we see that the results generally
correspond to Fischer's in that reversibility plays a role
in plausible sentences. In her analysis of reversibility,
Fischer does not distinguish between D-I and I-D object con-
structions. In the present study, with this particular age

group, only [-to] sentences were influenced by reversibility.

4.2.3. Role of Features
Besides influencing the plausibility of sentences,

the features were found to be important only in I-D

 

43

constructions. Further details of features will be discussed

in section 4.3.1. in connection with age.

4.3. Nonlinguistic Variables

4.3.1. Age

 

 

 

The difference in age between the youngest and the FE
oldest child in the study was nine months. Despite the 1 “3
individual differences in rate of development, the data
(see Table 4) showed a change in responses between the first &j i

three youngest children and the rest. (The turning point
is marked with a double line in Table 4.) In this experiment
3-9 years marked the turning point and the following differ-
ences between the two age categories were observed:
1. The most striking difference is in the order, or
D-I object constructions. In the older children,
almost all had a perfect score of accuracy, yielding
a mean score of 93 per cent. The younger children,
however, only reached a mean of 78 per cent accuracy.
This shows that attention to order of words and the
presence of the preposition increased with age.
2. In I-D object constructions, where the older children
achieved a mean of 29 per cent accuracy (i.e. 71
per cent R), the younger group performed with almost
chance (47 per cent) accuracy. This may indicate

an overgeneralization process among the older (i.e.

44

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

«O OH I OOH SO OO I OOH HO O I OOH O Hence mHHooz
O N I O O O I O OH H I O a Hence OIO
OO as I OOH OO OH O OO «O OO I OOH O Hmooe OOH:
O O. I HHIIIIHH- m H HH. O m I O # Hmuoe NIO
OO OO OH NO OO OO O OO OO OO I OOH O Hmooe Oooo3
O O O OH O O H OH O O I O O Hence HHIO
OO O I OOH OO OH I OOH OO HH I OOH O Hence mango
OH H I O .O HH m I O O H I OH O Hence HHIO
I- am me I OqH .IOO OH I OOH NO Om I OOH O Hmuoe muumcca
O O I OH O- O I O O O I OH O Hence HHIO
OO ON I OOH OOH I Om OS OO OO I OOH O Hence Onoom
O H I O O I O O O O I O . O Hmooe OHIO
IIIInN:O~I I OOH OO ON I OOH I OOH I OOH O Hmuoe moomH
II m H [HI O O H I w I m I m # HMHOB IIWHM
, Om NO IHH.OO OO OH OO OO mm mm HH OO O Hmuoe mccow
O O H O O H O O O O H O O Hmooe OIO
II- OOIOO OHIoO. OO OO OH OO OO SO an OO O Hmooe Ooooema
O O H O O O H O O O O O O Hmooe OIO
Om OO OH OO OO OO HO OO OS NO ON OH O Hmooe goon
O OH O OH O O O HH O O O O O Houoe OIO
m U m U m U m U m U m U TEMZIomd
HooI_ Hoo+O HOHIO HoHIO HOHIO Hou+O
Hcmfinm+a HHOEHQ¢+H Hmumsflcdla
HooOoo HoouHooH
.uoohoo
uomuwpsH on» mo mousummm may no mason cH .mmm ou mchHooom
Umumpmo Downnsm comm MOM noncommmn mo uswo son was Hogans Hmuoa .O THQmB

qo noelta

E

1139

45

they were applying the rule mentioned in section
4.2.1.-—always give the first object to the second
object--to all the sentences they heard) and, still

a period of confusion With the younger because they
had not yet completely formulated a rule and there-
fore reached only chance accuracy.

Furthermore, when the responses of the two age groups
were compared with respect to the three features
([-animate], [+animal] [+human]) in I-D object con-

structions, more differences became apparent:

 

(The data here are analyzed only in terms of the
indirect object, since no patterns were noticed in
the direct object features other than the general
trend of interpreting the {—animate] feature as the
direct object.)

a) [+human]. When the indirect object was a human,
the older children had a mean score of only 28
per cent accuracy; but for the younger group
this score was 67 per cent.

b) [+an1mal]. When the indirect object was an animal
the older subjects had a score of 17 per cent and
the younger ones 47 per cent accuracy.

c) [-animate]. For this feature, the score of accuracy

for both groups was quite similar: older age 32

 

46

per cent, younger 28 per cent accuracy. [-animate]
for the indirect object was the only feature

which yielded higher {although very little) scores
for the older group.

From the above observations it can be tentatively

concluded that:

a) younger children depend much more on the inherent

b)

semantic features rather than on the syntactic form
of the construction.

There is a hierarchy within the three features for
the Younger group. The noun with a [+human] feature
is most frequently interpreted as the indirect object.
This also occurs, but to a lesser extent, with
[+animal] nouns, and least of all with [-animate]
objects. As mentioned above, the reverse was the
case for the direct object. This heirarchy may have
been due to the use of the verb 'to give' but cannot
be determined until more research is done. No par-
ticular hierarchy of features was observed in the
responses of the older children which is presumably
due to their attention to word order at this later
age.

These results were further supported by the preliminary

study which will be discussed in the following chapter (V).

 

47

They do,however, contradict Fischer's (l97l) conclusion.

As mentioned in section 2.4.2., her results show that, 9323
time, children pay more attention to semantic cues and some-
what less attention to order. This conclusion is in fact
the opposite of what was found in this work. This contra-
diction may be due to the method of experimentation and
indicates the need for further research in this particular

area.

4.3.2. Other Nonlinguistic Variables

Other than age, no variable seemed to have influenced
the responses. No marked difference was noticed between
girls and boys, nor between blacks and whites. The family
and socio-economic background were also unimportant with
respect to the experiment: the only observed difference was
that lower class children had more lexical limitations (i.e.
cbuld name fewer fruits or animals etc-) than middle class

children.

 

CHAPTER V

COMPARISON OF THE PILOT STUDY

WITH THE MAIN STUDY

5.1. Introduction

Four months prior to the research for the experiment

reported in Chapter IV, a pilot study was carried out with

 

one of the children in the preschool (the same school where
the group study was done). This chapter will primarily
deal with a comparison of the two studies. In addition,
comparative references will be made to the informant's re-
sults in the group experiment wherever they are considered
to be necessary. Besides comparing the results, the dif-
ferences in the experimental design will be mentioned in

order to show the scope of the preliminary work.

5.2. Experimental Design
5.2.1. Subject

One child, Yanna, was used as the informant for the
pilot study. Yanna was 3-4 at the time of the first inter-
view. That is, she was one month younger than the youngest
<:hild in the group and four months younger than she herself

tmas at the time of the group study.

48

49

5.2.2. Interview Procedure
Contrary to the group study where all the elicitations
took place at school (an environment where the children were
already exposed to different experiments), Yanna was inter-
viewed at home, on an average of two interviews per week. FE
Each interview lasted approximately two hours, although the i

experimentally relevant period varied between three and seven J

 

minutes each time. .=1
In both studies, prior to the actual experiment, the Li

investigator Spent several afternoons with the children in

order to become acquainted with them. At school the experi-

menter was introduced to the children as a helper to the main

teacher, whereas Yanna knew the experimenter as a friend of

her mother.
The experimenter either picked up Yanna from school

or waited for her at home, and started to play with her soon

after her snack. At first this was a novelty, but after

the second session it became extremely difficult to attract.

her to the game. She behaved very informally and usually

stopped COOperating after about the first three minutes.

At the preschool, however, the children were more than willing

to cooperate due to the Spirit of competition among the

children established by the school situation. Curiously

enough, in the group experiment Yanna was one of the most

50

enthusiastic children in 'playing' the ‘game' and at times

her attention span laSted as long as fifteen minutes.

5.2.3. Toys
There were two main differences in the toys between
the two studies. In the preliminary work:
a} the toys were not proportional in size to real life
objects; and
b) they were Yanna's own toys, and therefore provided no
novelty as the toys did in the second study.
For one session, the use of pictures was also tried,
but the pictures were quickly abandoned due to Yanna's total

disinterest in them.

5.2.4. Other Differences
Except for those mentioned above, all other method-

ological details were the same in the two experiments.

5.3. Comparison of Results and Discussion
5.3.1. Introduction

For reasons given in the preceding sections of this

chapter (e.g. interviewing at home, use of one informant etc.

compiling of the data was very difficult with Yanna. The
results of the study are summarized in Table 5. The total

number of sentences tested for Yanna was smaller than that

of the average number for each child in the group (51 to 61).

 

Direct Object

Table 5.

51

Summary of the results of the pilot study with
Yanna

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Indirect Object
[-animate] [+animal] [+human]
[+to} {-to} [+to} [-t01 [+to] [-to]
{-animate] 88% C - 50% C 60% C 100% C
12% R 34% R - 20% R -
- 66% X 50% X ? 20% X -
8 expls 3 expls 2 expls 5 expls 5 expla
[+animal] - - 60% C - H’
100% R 100% R - -
- - 40% X 100% X ? ?
_~ 2 expls 1 expl 5 expls 6 expls
H+human1 20% c 100% c - - -
40% R - 50% R - -
40% x ? - 50% X 100% X 100% X
5 expls 2 expls 2 expls 3 expls 2 explfl
C - the correct response
R - reversal of what is actually said
X - no response due to confusion
? - sentence type not investigated

In the group study a mean of 3.5 sentences per child for each

construction was used, whereas for Yanna 2.9 sentences were

tested.

distribution of the sentences.

As the Table 5 shows,

Perhaps more important, however, was the uneven

some

constructions had as many as eight examples and some others

had as few as one, or in some instances had no examples.

This

was due to the increasing unwillingness of the subject to

 

52

cooperate; eventually it became impossible to elicit any
further information. The constructions which were not tested
in the pilot study are:

l) D[+animal] - I[+human] e.g. Give the cat to the boy.

 

2) I[+anima1] - D[-animate] e.g. Give the dog the bone. rt?

3) I[+human] - D[+animal] e.g. Give the boy the cat.

 

4) I[-animate] - D[+human] e.g. Give the truck the boy.

 

Due to these gaps and other constructions with one or J

two examples, the results remain only suggestive in value.

 

5.3.1.1. Correct and Incorrect Responses

In the group study, the sentences were analyzed in
terms of correct and incorrect responses. In the first study,
however, there was an additional category (marked by X in
the table) to indicate the absence of a response due to
confusion. In such cases Yanna usually stOpped playing or
focused her attention to another place or even left the room
instead of responding to the command. No responses (X)
were included in the results because their occurrence was
very frequent.

As in the group study, in the cases of reversed or
no responses, the command was repeated several times and the
object relations were made very clear {with stress and pointing
to the objects) by the experimenter. The repetitions did

not seem to have any effect on the child‘s responses.

53

5.3.1.2. General Trends

Unlike the group work, the pilot study did not show
any particular patterns in the results. It is assumed that
the use of one subject and the gaps in the data are responsi-

ble for this.

5.3.2. Variables
5.3.2.1. Word Order

Contrary to the group study, in the pilot work, the
word order was not found to be particularly functional for
comprehension.

In D-I object constructions, the mean per cent of
accuracy was 48 per cent and the remaining 52 per cent was
divided between 31 per cent X and 21 per cent R. This ratio
of accuracy differed considerably from the mean of 90 per cent
C of the group study and even from that of Yanna's own results
four months later (82 per cent).

In I-D object constructions the mean per cent of
accuracy was only 13 per cent. The reSponses were mainly
confusions (57 per cent) and to a lesser extent reversals
(30 per cent). A more careful look at the results shows
that this 13 per cent accuracy was obtained from one con-
struction type which had only 100 per cent C responses, all
others had no correct responses. Interestingly this exception

has I[+human] - D[-animate] structure (e.g. Give the boy the

 

54

apple or Give the boy the truck) and therefore corresponds

 

to the exception in the group work (see section 4.2.1.).
Isolating Yanna's results in the second experiment from the
group showed that she had not changed her rule for this con-
struction within the four month time period between experi-
ments.

As pointed out in section 4.2.1., 100 per cent

accuracy for the I[+human] - D[-animate] structure indicates
the importance of semantic cues for comprehension at this
age. Particularly at this early age, then, wherever the
semantic cues are sufficient for correct interpretation,
the syntactic cues (or the absence of the preposition E3)
play no apparent role in the child's interpretation.
The above exception provides additional support for the
inference that children first interpret as the direct object
the noun with a [-animate] feature, and as the indirect object
the noun with the [+human] feature (also see section 4.3.1.).

This inference is further justified by the results

of the construction D[+human] - I[-animate] (e.g. Give the man

 

to the apple) where Yanna was confused 40 per cent of the

 

time, reversed the command 40 per cent of the time (giving
the apple to the man) and was correct 20 per cent of the time.
The confusion may have been caused by implausible sentences

such as Give the man to the truck, where the reverse is also

 

 

55

anomalous. It can be concluded from the above discussion
that:

a) the child was at a very early stage, possibly just
beginning the process of formulating a rule for word
order such as the one suggested in section 4.2.1. 'rfi

b) because of the absence of a fixed rule for syntax, the H
child at this age relies heavily on semantic cues.

c) when neither the semantic nor the syntactic cues (if

 

 

formulated at all) help, the child is totally confused ti

and prefers not to respond.
5.3.2.2. Reversibility and Semantic

Features

The role of reversibility was not very clear from
the results because of the gaps in certain constructions and
other limitations discussed at the beginning of this chapter.

In reversible sentences (where according to the
definition in section 4.3.1. both objects have the same
semantic features) with D-I object constructions, it was
noticed that [-animateJ feature had 88 per cent C responses;
[+animate] had the next highest (60 per cent) and [+human]
had no correct responses. It can be inferred from these
figures that there may be a hierarchy in the acquisition
of features for reversible relations: [-animate] feature is

learned first (despite the anomaly caused by the use of the

56

particular verb 'to give‘), the child then tries to master
the [+animal] feature, and finally will develop the [+human]
feature at a later stage. Looking at the results for all of
the children in the group study, the above hypothesis is in
fact justified, since in D-I reversible constructions the
[+human] feature has reached 47 per cent accuracy and the
other two features both have close to perfect responses

(97 per cent C).

It should be noted that Yanna's 88 per cent C re-
sponses for the [-animate] reversible construction was ob-
tained from sentences in which the direct object was a
smaller item than the indirect object. For the remaining
12 per cent R responses the direct object was a bigger item
than the indirect object. The size of items, therefore,
was an important factor influencing the judgment of the
child and is linked to semantic plausibility. Thus, the
sentence Give the book to the couch was considered less

anomalous than Give the couch to the book.

 

For animals size was not a critical factor, pri-
marily because all animals used as toys were approximately
the same size.

The learning of the [+animal] feature before the
[+human] feature in reversible D-I constructions, despite

the fact that children are usually more in contact with

 

EU. 4-..] -.n ‘ ‘
1.

57

humans than with animals, may be due to the view of Jespersen

{1922) who claims that young children show a special inter-

est in animals. "Among the child's first passions are animals
and pictures of animals . . ." (p. 114).
Reversible sentences in I-D constructions mainly “3

produced confusions, particularly with the [+human] and
[+animal] features both of which yielded 100 per cent X.

For the [-animate] feature Yanna had 34 per cent reversals

 

which shows that she was beginning to formualte the rule *5
(according to which the first noun is always the direct ob-
ject) applying it first to sentences with [-animate] feature
nouns.
In reversible sentences the use of the preposition
pg played a role in nouns with {-animate] and [+animal]
features. Yanna gave some correct responses to constructions
involving these features. Here too, as in other situations
mentioned previously, Yanna tried to resort to syntax only

when the semantic cues were insufficient.

5.5.2.3. Age

In this section, a comparison is made of results
from Yanna (3-4) in the pilot study with that of older (3-9
to 4-2) and younger (3-5 to 3-8) children in the group study.
The results are summarized in Tables 6 and 7. In most in-

stances the tables show a progression, either increasing or

58

Table 6. Comparison of the correct responses of three
different age groups with respect to word order.

 

 

 

Older Group Younger Group Yanna
3 D-I
“g [+to] 93% C 78% C 48% C
O
.U I-D
S [-to] 23% C 47% C 13% C
3

 

Table 7. Comparison of the correct responses of three
different age groups with respect to indirect
object features, including both D-I and I-D
object order.

 

 

 

 

Older Group Younger Group Yanna
# a) [+human] 28% C 67% C 75% C
Egg [+animal] 17% c 47% c 0% c
Egg I-animate] 32% C 28% C 20% C

 

 

59

decreasing from the youngest to the oldest group. For example,
Table 6 clearly shows that the older group have almost com-
pletely mastered the D-I object constructions (93 per cent C).
Moreover, they have overgeneralized the rule (always give

the first object to the second object) to I-D object con-
structions where a mean of only 29 per cent C responses are
obtained. The younger group, on the other hand, have not
quite mastered the rule yet (78 per cent C), and did not
overgeneralize it to I-D structures (47 per cent C). Yanna
in the pilot study had only started to formulate a rule for
the D-I structures (48 per cent C), and her low accuracy

for I-D object constructions (13 per cent C) is due to the
large percentage of confusions.

In Table 7, when the indirect object has a [+human]
feature, there is a decreasing progression of correct re-
Sponses from the younger to the older groups--75 per cent C,
67 per cent C and 28 per cent C respectively. It should be
noted that most of the command sentences with the [+human]
feature for the indirect object are plausible, such as give

the book to the man or Give the man the book. The progression

 

therefore shows that dependence on semantic cues is correlated
with young age. Thus, Yanna and the younger group had a high
percentage of correct responses for plausible sentences with

[+human] indirect objects. The older group, however, who do

 

ll-"VL I" «Is * u
t

60

not attend to semantic cues such as plausibility but instead
rely on word order, had a lower percentage of correct re-
Sponses.

The present results do not lend themselves to a
meaningful interpretation for the [+animal] feature. Based
on her comprehension of other linguistic variables, Yanna's
0 per cent C responses for the [+animal] feature is suspect.
Zero per cent C for any feature is highly unlikely, since
there is no explanation to support it and even if the child
had a rule, some errors would still occur.

For the [-animate] feature, there is a slight in-
creasing progression from the younger to the older ages--20

per cent C, 28 per cent C and 32 per cent C respectively.

Yanna has the lowest per cent accuracy since at this stage she

expects the indirect object to be an animate noun.' The
accuracy for the [~animate] feature increases with older age
because of the lesser reliance on the features.
To conclude then, three points seem to be significant
for the three age groups:
a) attention to syntax (word order) increases with age;
b) attention to syntactic and semantic cues are inversely
related; and
c) children first associate the indirect object with the

[+human] feature.

CHAPTER IV

SUMMARY

In this paper the acquisition of direct-indirect
object constructions was studied with children ranging from
3-5 to 4-2 years. The comprehension of the various structures
were tested by direct interviewing. The results were analyzed
in terms of the word order, reversibility and plausibility
of sentences. A comparison was made between the responses
of the lower age group (3-5 to 3-8) and the older group
(3-9 to 4-2). These results were further compared to the
results of a pilot study carried out with one of the children
in the group (3-4), four months prior to the main study.

It was found that constructions with D-I object
structure were mastered earlier than those with I—D structure.
The most striking difference between the three age groups
was with respect to word order and semantic plausibility.
Older children paid more attention to the word order and
syntactic relations, whereas the younger children were more
influenced by the semantic cues of a sentence. The younger
children most frequently interpreted the [-animate] feature
as the direct object and the [+human] feature as the indirect
object.

61

 

62

With regard to word order, the younger children in
the study were only beginning to formulate the rule which the
older children already possessed: always give the first object
to the second object. The older group overgeneralized the
rule for the D-I object constructions and applied it to I-D
structures as well.

For both age groups in the main study, reversi-
bility played a role only in plausible sentences with I-D
object construction. In the pilot study which covered a
still younger age, a hierarchy of features in reversible
sentences was noticed. Reversible relations which have
[-animate] objects are learned first; the child then tries
to master the [+animal] feature and eventually is expected
to develop the [+human] feature at a later stage.

It can be concluded from this study that the second
half of the third year is a crucial stage for the development
of syntactic relations. More research is needed with three-
year-old children in other aspects of syntax (such as loca-
tives and other prepositional phrases), as well as in the
use of different verbs, with special attention to the semantic
features of the nouns in the sentences. An experiment,
similar to the present one, should also be performed with
children above and below the age groups studied in this paper
Moreover, dative constructions should be studied with children
speaking different languages to provide evidence for linguistic

universals.

 

APPENDICES

 

APPENDIX A

CORRECT, REVERSED AND PAIRED RESPONSES

OF EACH CHILD IN>THE SAMPLE

 

 

 

 

Indirect Object

 

 

 

 

 

.m. _._....._.11_..1.1...11_.1...3
1m 1... ._.1124273 .121132333 4231123433
m f. 432._4.242323..11___4.3112134.
H

.I.

[.w ._l.._.3__1__.__._..11..........

+ 3421143733 4231233353 3332133433

[

.m. __.._1._1_.2..___12_.1......12

I... _.2122245152523755441121134353
]

Ha. .I. 3411.1.1226...._22_.12__.1._12
m

.1

n .1 , .

Mum 1_._._..___.__...1._.415.2._.1.
.l. + 2211231433 5332243833 41.1223353

[

]
1n]. 0 2332143343 2.23.323_4 3211113143

t I-
t _ 2.2.131_2_ _2_.11_.5. .1_.11_411
a .l. .

m

i

n 1

Jam .l.___._...1........311...._._

[ + 3432233333 3323234333 3133443333
[

 

 

 

 

 

 

 

a e a e a e

d +t e d rt e d l. e

naaYtsY l naaYtsYl naaYtsYl
nonrbeidal nonrbeidalnonrbeidal
thubnrnne.nTnubnMnne.thubnrnne
oaaaonheio oaaaon eio oaaatheio
JLYLRACWMN JLYLRACWMN JLYLR CWMN.

Hmamefic¢na HHmEflc<+H Hawssm+_

 

 

 

pomnno uomnwo

63

APPENDIX B*

A SAMPLE INTERVIEW

Now Mina, let me see if you know the names of all these toys:

 

What's this? -- (a truck) -- right:
What's this? -- (a book) -- right)
What's this? -- (a dress) -- right:
What's this? -- (a banana)-- right)
What's this? -- (a pear) -- righti
What's this? -- (a sheep) -- right!
What's this? -- (a cow) -- righti
What's this? -- (a pig) -- righti
What's this? -- (a duck) -- right!
What's this? -- (a horsie)-- right;
What's this? -- (daddy) -- righti
What's this? -- (mommy) -- right!
What's this? -- (a baby) -- right:

What's this? -

(a boy) -- right!
That was very good. Now let me see if you will remember how
to play the game. I'll close my eyes and will say Give the

book to the boy, that's right Mina, that's very good. Now

 

we'll play the game.

 

*Parentheses are used for the child's speech. Square
brackets enclose a description of the child's nonlinguistic
behavior. Codes: Very good--the correct response; O.K.--
the reversed response; Match--the paired response (i.e. the
child paires the two toys instead of giving oneto the other).

64

65

Give the banana to the boy. -- Very good!

Give the boy the pear. -- Very good Mina!

Give the cow to the truck. -- Very good?

Give daddy to the truck. -- Very good!

Give the cow to the horsie. -- Very good!

Give the truck the sheep. -- O.K.

Give mommy the truck. (That's funny) What, what I said?
(yes) why? (because) [the child giggles] -- O.K.

Give the pear (to the?) to the banana. -- Very good!

Give the banana the baby. -- O.K.

Give the pig the duck. -- Match

Give daddy to mommy. -- Very good;

Give the banana the pig. -- O.K.

Give mommy daddy. -- O.K.

Give daddy the baby. -- Very good!

That was all very good Mina, we'll continue the game next

time.

BIBLIOGRAPHY

 

BIBLIOGRAPHY

Bach, E. and R. Harms, eds. 1968. Universals in Linguistic
Theory, New York: Holt, Rinehart and Winston Inc.

 

 

IT:
Bates, R. Reed. 1969. A Study in the Acquisition of Syntax, g ‘
Doctoral Dissertation, University of Texas. (As i
quoted by Fischer, 1971). ‘
Bellugi, Ursula. 1965. "The Development of Interrogative
Structures in Children's Speech," in Reigel, 1965, :
pp. 103-138. 5.
L4

 

Bellugi, Ursula and Roger Brown. eds. 1964. The Acquisition
of Language, Monograph Social Research Child Develop-
ment, 29, No. 92.

 

Berko, Jean. 1958. "The Child's Learning of English
Morphology," Word, 14, pp. 150-177.

Bever, T. G. 1970 (a). "The Cognitive Basis for Linguistic
Structures," in Hayes, 1970, pp. 91-113.

Bever, T. G. 1970 (b). "The Nature of Cerebral Dominance
in Speech Behavior of the Child and the Adult," in
Huxley and Ingham, 1970, pp. 100-120.

Bever, T. G., J. Mehler and V. V. Valian. 1967. "Linguistic
Capacity of Very Young Children," Lecture at Graduate
School of Education, Harvard University. (As quoted
by McNeill, 1970).

Bloom, Lois. 1970. Language Development: Form and Function
in Emerging Grammars, Cambridge Mass: M.I.T. Press.

 

Blumenthal, A. L. 1970. Language and Psychology, New York:
Wiley.

Braine, Martin. 1963. "The Ontogeny of English Phrase
Structure: The First Phase," Language, 39, pp. 1-13.

Brown, R. and C. Fraser. 1964. "The Acquisition of Syntax,"
in Bellugi and Brown, 1964, pp. 43-78.

66

67

Brown, R. and C. Hanlon. 1968. "Derivational Complexity
and Order of Acquisition in Child Speech," Paper
delivered to Carnegie Mellon Symposium on Cognitive
Psychology (as quoted by Fischer, 1971).

Carmichael, L. Ed. 1954. Manual of Child Psychology, New
York: Wiley.

Chomsky, Carol. 1969. The Acquisition of Syntax in Children
from 5 to 10, Cambridge, Mass: M.I.T. Press. F1.

 

 

Chomsky, Noam. 1965. Aspects of the Theory of Syntax,
Cambridge, Mass: M.I.T. Press. '

 

Chomsky, Noam. 1968. Language and Mind, New York: Harcourt,
Brace and World Inc.

 

 

Chomsky, Noam. 1970. "Formal Discussion of 'The Development 1}
of Grammar in Child Language,'" by Wick Miller and
Susan Ervin, in Lester, pp. 41-50.

Dingwell, William 0. ed. 1971. A Survey of Linguistic
Science, Maryland: University of Maryland.

 

Ervin, Susan and Wick Miller. 1964. "Language Development,"
Child Psychology, The Sixty-Second Yearbook of the
NatiOnal Society for the Study of Education, Part I,
pp. 108-143.

 

Ervin, Susan. 1964. "Imitation and Structural Change in
Children's Language," in Lenneberg, 1964, pp. 163-
191.

Fillmore, Charles J. 1965. Indirect Object ConStructions
in English and the Ordering of Transformations,
The Hague: Mouton.

 

 

Fillmore, Charles L. 1968. "Case for Case," in Bach and
Harms, 1968, pp. 1-88.

Fischer, S. D. 1971. The Acquisition of Verb—Particle and
Dative ConstructiOns, Doctoral DiSsertation, M.I.T.

 

 

Fodor, J. A. and J. J. Katz. eds. 1964. The Structure 9f
Language, Englewood Cliffs, New Jersey: Prentice-Hall.

 

Gough, P. B. 1965. "Grammatical Transformations and Speed
of Understanding," Journal of Verbal Learning and
Verbal Behavior, 4, pp. 107-111.

 

 

68
Gruber, J. 1967. "Topicalization in Child Language,"
Foundations of Language, 3, pp. 37-65.

Halle, Morris. 1964. "Phonology in Generative Grammar," in
Fodor and Katz, 1964, pp. 334-353.

Hayes, J. R., ed. 1971. Cognitive and Human Learning, New
York: Wiley.

 

Huttenlocker, J. and S. Strauss. 1968. "Comprehension and
a StatementSRelation to the Situation it Describes,"
Journal of Verbal Learning and Verbal Behavior, 7,
pp. 300-304.

Huxley, J. and D. Ingham, eds. 1970. Mechanisms of Language
Development, New York: Academic Press. !

 

 

Jakendoff, R. and Peter Culicover. 1970. "A Consideration
of Dative Movement," Rand Corporation (As quoted
by Fischer, 1971).

Jacobs, R. and P. S. Rosenbaum. 1968. English Transforma-
tional Grammar, Boston: Blaisdell Publishing
Company.

 

Jacobs, R. and P. S. Rosenbaum, eds. 1970. Readings in
English Transformational Grammar, Boston: Ginn and
Company.

 

 

Jakobovits, Leon, A. and Murray S. Miron, eds. 1967.
Readings in the Psychology of Language, Englewood
Cliffs, New Jersey: Prentice-Hall.

Jakobson, R. 1942. "Kindersprache, Aphasie and Allgemeine
Lautgesetze," in Jakobson, 1962, pp. 328-401.

Jespersen, Otto. 1922. Language: It's Nature, Development
and Origin, London: George, Allen and Unwin Ltd.

 

Kiparsky, Paul. 1968. "Linguistic Universals and Linguistic
Change," in Bach and Harms, 1968, pp. 171-205.

Klima, Edward. (to appear. "Regulatory Devices Against
Functional Ambiguity." (as quoted by Fischer, 1971).

Klima, Edward and Ursula Bellugi. 1966. "Syntactic Regu-
larities in the Speech of Children," in Lyons and
Wales, 1966, pp. 183-207.

69

Lenneberg, Eric. 1964. ed. New Directions in the Study of
Language, Cambridge, Mass: M.I.T. Press.

 

Lester, Mark, ed. 1970. Readings in Applied Transforma-
tional Grammar, New York: Holt, Rinehart and Winston
Inc.

 

 

Lyons, J. and R. J. Wales, eds. 1966. Psycholinguistic
Papers, Edinburgh: Edinburgh University Press.

 

McCarthy, Dorothea. 1954. "Language Development in Children,"

in Charmichael, 1954.

McNeill, David. 1966. "Some Universals of Language Acqui-
sition," in Lane, H., ed. 1966. Studies of Language
and Language Behavior: Centr. Reasr. Lang. Lang.
Bhvr., Rept. No. 2.

 

McNeill, David. 1966. "The Creation of Language in Children,"

in Lyons and Wales, 1966, pp. 94-114.

McNeill, David. 1970. The Acquisition of Language, New
York: Harper and Row.

 

Mehler, J. 1963. "Some Effects of Grammatical Transfor-
mations on the Recall of English Sentences," Journal
ngerbal Learning and Verbal Behavior, 2, pp. 346-
351.

Menyuk, Paula. 1969. Sentences Children Use, Cambridge,
Mass: M.I.T. Press.

Miller, G. A. 1962. "Some Psychological Studies of Grammar,"
American Psychologist, 17, pp. 748-762.

 

Reibel, David A. and Sanford A. Schane, eds. 1969. Modern
Studies in English, Englewood Cliffs, New Jersey:
Prentice-Hall.

 

Reigel, K., ed. 1965. The Development of Language Functions,
Univ. Mich. Lang. Develop. Program, Rept. No. 8.

Reed, Carroll, ed. 1970. The Learning of Language: Essays
in Honor of D. H. Russel, New York: Appleton-Century
Crofts.

 

Rosenbaum, Peter. 1967. The Grammar of English Predicate
Complement Constructions, Research Monograph Series
No. 47, Cambridge, Mass: M.I.T. Press.

 

 

4‘: =E-m‘ .lel

 

70

Ross, J. R. 1967. "On the Cyclic Nature of English
Pronominalization," in To Honor Roman Jakobson,
The Hague: Mouton.

 

Saporta, Sol, ed. 1961. Peycholinguistics, New York:
Holt, Rinehart and Winston Inc.

 

Savin, H. and E. Perchonock. 1965. "Grammatical Structure
of Immediate Recall of English Sentences," Journal
of Verbal Learning and Verbal Behavior, 4, pp. 348-
353.

 

Shipley, E., E. Smith and L. Gleitman. 1969. "A Study in
the Acquisition of Language," Language, 45, No. 2.
pp. 322-342.

Slobin, Dan. 1966 (a). "Grammatical Transformations and
Sentence Comprehension in Childhood and Adulthood,"
Journal of Verbal Learning and Verbal Behavior, 5,

 

pp. 219-227.

Slobin, Dan. 1966 (b). "The Acquisition of Russian as a
Native Language," in Smith and Miller, 1966, pp.
129-149.

Slobin, Dan. 1970. Psycholinguistics, Glenview, 111: Scott,
Foresman and Company.

 

Slobin, Dan. 1971 (a). The Ontogensis of Grammar, New York:
Academic Press.

 

Slobin, Dan. 1971 (b). "Developmental Psycholinguistics,"
in Dingwall, 1971, pp. 298-410.

Slobin, Dan and C. K. Welsh. 1969. "Elicited Imitation as
a Research Tool in Developmental Psycholinguistics,"
Language Behavior Research Laboratory Working Paper
No. 10, University of California, Berkely, Calif.
(as quoted by Fischer, 1971).

Smith, Frank and George A. Miller, eds. 1966. The Genesis
of Language, Cambridge, Mass: M.I.T. Press.

 

 

Turner, E. and R. Rommetveit. 1967. "The Acquisition of
Sentence Voice and Reversibility," Child Development
38, pp. 469-660.

 

 

71

Turner, E. and R. Rommetveit. 1968. "Focus of Attention in
Recall of Active and Passive Sentences," Journal of
Verbal Learning and Verbal Behavior, 7, pp. 543-548.

 

Velten, H. V.. 1943. "The Growth of Phonemic and Lexical
Patterns in Infant Language," Language, 19, pp. 281-
292.

Weir, Ruth. 1962. Language in the Crib, The Hague: Mouton.
Weksel, W. 1965. "Review of 'The Acquisition of Language'

by Ursula Bellugi and Roger Brown eds.," Language,
41, pp. 692-709.

 

 

 

III
III
III