Ira: «fizz!
. LA.
.. .
.1}
(’II.‘
'rltl
\il .l:r.k.n\

.! 2:13,:
.0? kittlni

.. ”Mann...“ ‘r‘
4 r0 .. . Ll
.‘l‘ .
... Sultana-Hum
LI, - 1'! In

. . Hisill. .I... .f
». IllgLvJ’ . riluo
1“ I.’

 

I...

91.2%! In it
2.1!).

pm, 9

I
- .ui‘y’v.yl
Mafia“ 4.; 3 i.
[3
. , . '.
afiriilra
FEW“.
i. .. k
and .

.«nl.

: K - u. 3 l":
57 Ovll. . ‘1'. lslfillluflnuu). u
zit ix»... \... . J.. V «i .1
. .-YIV \“ In. 9... It
‘1'”. I.‘ .I...’ AIL! I4
R )‘l‘l‘
-. .1.
a .
1‘13?!“ 6.1- IllEHCI.‘ 1..
5.}...1. 31:5! .Jlivtts. s ,v.
. . 51.. iihhscnlrn
.4 it... iuflhfl‘ u
I... .
.A .n

"UV 3.11
. {.lmLFE. 13.! ulr
. IQt.“JHqu\..Vul?l ‘C. ‘
: s...n£r..uufl.rnfiar Xi
. firm“ A.“ I \a
[Ac-abstnraqfh
lyixll) . .

1‘.
L. «I I... .A .n
. 115;... . .e. wt.
. «.52...

. 19..
A .2 1‘!

kklumwmww .

‘..I

 

’0
I.’
II

THESIS

0:000

 

 

 

 

 

 

 

—=

Illll llillllfilllfllllllfliflll

3 1293 02 7

 

 

 

 

This is to certify that the
thesis entitled

Expressive Vocabulary
of

German-English Toddlers
presented by

Dorte Antje Junker

has been accepted towards fulfillment
of the requirements for

__M.A___. degree in Wguage
Pathology

Jlola .5100ka

Ida J. Stockmam Ph.D.

Major professor

Date_&24-gs————

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

PLACE IN RETURN Box to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

' DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

11m macros-Damn

 

EXPRESSIVE VOCABULARY OF GERMAN-ENGLISH

BILINGUAL TODDLERS

By

Dorte Antje Junker

A THESIS

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

MASTER OF ARTS

Department of Audiology and Speech Sciences

1 999

ABSTRACT

EXPRESSIVE VOCABULARY OF GERMAN-ENGLISH
BILINGUAL TODDLERS

By

Ddrte Antje Junker

This study investigates whether young children learning two languages
simultaneously are inherently weak language learners compared to their
monolingual counterparts. The vocabulary skills of 10 German-English
bilingual toddlers were compared with monolingual German- and English-
speaking peers around 24 months of age using Rescorla's (1989) Language
Development Survey, a vocabulary checklist based on parental report, in its
original English and a German translated version.

Findings revealed that bilingual toddlers were not inferior in conceptual
vocabulary size and verb diversity when words in both languages were
pooled. In addition, almost half of the bilingual conceptual vocabulary
(43%) was associated with lexical forms in both languages, from which it is
inferred that language separation is possible at age 2.

These findings suggest that no special norms for vocabulary skills need
be developed for German-English bilinguals, and that existing instruments may

be exploited in the future for speakers of other languages.

COPYRIGHT BY
Dorte Antje Junker
1 999

ACKNOWLEDGMENTS

I would like to thank Dr. Susan Gass and Dr. John Eulenberg of Michigan
State University for their comments and helpful discussions. My special thanks
go to my advisor, Dr. Ida Stockman of Michigan State University, for her
unflagging help with suggested readings, critical discussions, editing, and, most
important, inspiration. I would also like to thank the College of Communication
Arts and Sciences, Office of the Dean, and the Department of Audiology and
Speech Sciences for their generous financial support of my research.

Finally, to all my friends and parents, who encouraged and assisted me in
the difficult task of finding subjects on two continents; to Susan Miller for her
major support concerning all formal aspects of this work; and to my husband,
Michael Rodemer, who stood by me with his belief in me and last but not least

his technical help-my sincerest thanks.

TABLE OF CONTENTS

Page

LIST OF TABLES ............................................... viii

LIST OF FIGURES ............................................... ix

INTRODUCTION ................................................ 1

Aspects of Bilingualism ...................................... 2

Prevalence of Bilingualism in the United States ................... 3

What Can Be Learned From the Study of Bilingualism? ............. 4

Bilingualism and Clinical Issues ............................... 7

STATEMENT OF THE PROBLEM .................................. 10

Need for Research on Early Bilingual Acquisition ................. 10

The Early Bilingual Lexicon as the Focus of This Investigation ....... 11

What Parameters Have Been Studied? ......................... 13

Monolingual Vocabulary Development .................... 13

Bilingual Vocabulary Development ....................... 17

Conceptual Issues ......................................... 18

Bilingualism Deficit Hypothesis .......................... 18

Unitary Lexicon Hypothesis ............................. 20

Clinical Issues ....................................... 23

RESEARCH QUESTIONS ........................................ 27

METHODS AND PROCEDURES ................................... 28

Subjects ................................................. 28

Subject-Selection Criteria .............................. 29

Additional Selection Criteria for Bilingual Subjects ........... 33

Recruitment of Subjects ............................... 37

Description of Vocabulary Measures ........................... 37
Original Monolingual English Version of the Language

Development Survey ................................ 37

Page

German Translated Version of the Language Development

Survey ........................................... 40
Data-Collection Procedures .................................. 44
Parent Interview ..................................... 44
Informal Observation of Child’s Spontaneous Play ........... 44
Collection of Vocabulary Data ........................... 45
Data Analysis ............................................. 47
Methodological Issues ................................ 47
Calculation of Vocabulary Size .......................... 50
Calculation of Vocabulary Richness ...................... 52
Calculation of Proportion of Translation Equivalents in
The Child’s Vocabulary .............................. 52
Reliability of Data .................................... 54
Statistical Treatment of Data ............................ 56
RESULTS ..................................................... 57
Vocabulary Size ........................................... 57
Total Conceptual Vocabulary ........................... 57
Total Lexical Vocabulary ............................... 60
Vocabulary Richness ....................................... 60
Distribution of Bilingual Vocabulary Over Two Languages .......... 62
Proportion of Translation Equivalents in Conceptual
Vocabulary ....................................... 62
Distribution of Bilingual Lexical Forms Relative to
Monolingual Groups ................................ 63
DISCUSSION .................................................. 67
Theoretical Implications ..................................... 68
Clinical Implications and Contributions ......................... 72
Future Research .......................................... 73
SUMMARY AND CONCLUSIONS .................................. 77
APPENDICES
A. Tables—Raw Data .................................... 80

B. Items on Rescorla’s Language Development Survey:
Frequency of Occurrence in Children’s Vocabularies
and Phonetic Transcriptions ............................ 86

vi

Page
C. Instruments ........................................ 1 1 1

D. Letter to Parents, Parent Consent Form, and Parent
Questionnaire ...................................... 1 17

REFERENCES ................................................ 124

vii

I... x

I

 

i.-..-

 

LIST OF TABLES

Page
Subject Pool: Overview of Age, Gender, and Birth Order ........... 81
Productive Vocabulary: Bilingual Raw Data ..................... 82
Monolingual German Raw Data ............................... 83
Monolingual English Raw Data ............................... 84
Number of Different Verbs Across the Bilingual and
Monolingual Subject Groups ................................. 85
Frequency of Occurrence, Rank Order, and Number of Syllables
for the Sample of 50 Randomly Selected LDS Items
(English Version) .......................................... 87
Frequency of Occurrence, Rank Order, and Number of Syllables
for the Sample of 50 Randomly Selected LDS Items
(German Version) ......................................... 89

viii

LIST OF FIGURES

Page
Mean Number of Lexical Forms and Concepts Across the
Bilingual, the Monolingual German, and the Monolingual
English Subject Groups ..................................... 58
Individual Vocabulary Scores Within and Across the Bilingual,
the Monolingual German, and the Monolingual English
Groups .................................................. 59
Average Number of Different Verbs Across the Bilingual, the
Monolingual German, and the Monolingual English Groups ......... 61
Number of Different Verbs Produced by Each Subject
Within and Across the Bilingual, the Monolingual German,
and the Monolingual English Groups ........................... 62
Proportion of Translation Equivalents Within the Conceptual
Bilingual Vocabulary ....................................... 63
Number of Lexical Forms for German and English Within the
Bilingual Group ........................................... 64
Average Monolingual English Vocabulary and Individual
English-Only Vocabularies of Bilingual Subjects .................. 65
Average Monolingual German Vocabulary and Individual
German-Only Vocabularies of Bilingual Subjects ................. 66

INTRODUCTION

De Bot (1992) estimated that more than half of the world’s population can
be considered bi- or multilingual. Therefore, he regarded monolingualism as an
exception rather than the norm. Despite the widespread nature of the
phenomenon, there is no consensus in the literature on what bilingualism is,
given the tremendous diversity among bilingual speakers. For example, a child
growing up in a Spanish-English bilingual community in Florida might have a
different language experience than a child growing up in a French-English
community in Quebec, and both may differ from children Ieaming the same
languages, but with little or no access to bilingual communities. Cultural values,
social and emotional attitudes, socioeconomic status, age of acquisition, and
opportunities to speak the languages all influence proficiency as well as the
motivation of the individual to use different languages (Ardila, 1998). Different
environmental contexts also pose demands on the proficiency of the speaker.
For example, a school setting might require competency in written language,
which is generally not the case in a home environment. Thus, different
proficiencies in each language-from low to full fluency in the spoken and/or
written language—are conceivable. Some researchers have even claimed that

there is no clear-cut dichotomy between monolingual and bilingual speakers and

that language skills have to be judged on a continuum (Ardila, 1998). This
variability among speakers and environments necessitates defining bilingualism

as a prerequisite to studying the phenomenon.

Aspects of Bilingualism

Schreuder and Weltens (1993) pointed out that the extent to which people
are considered bilingual largely depends on the definition of bilingualism. In the
literature, different uses of the term have coexisted. Whereas Grosjean (1982)
defined bilingualism simply as the regular usage of two languages, other authors
have had a more restrictive understanding. Bloomfield (1933), for example,
expected native-like proficiency in both languages, and Weinreich (1974)
thought that the ability to speak both languages interchangeably and to switch
from one language to the other according to situational demands (code
switching) characterizes bilingual competence. Other authors, such as
McLaughlin (1978) and Singleton and Lengyel (1995), introduced the factor of
age and differentiated between simultaneous bilingualism, when a child is
exposed to two languages from birth or at least before 3 years of age, and
sequential or consecutive bilingualism, when the child acquires a second
language (L2) in early childhood, but after the basic acquisition of a first
language (L1). Weiss (1959) differentiated natural bilingualism, when
languages are informally learned in a natural context, from cultural bilingualism,
where the second language is learned in a formal setting, such as through

instruction at school. As is evident from these examples, no agreement exists in

 

 

the literature about the usage of the term bilingualism. What makes the situation
even more confusing is that some definitions, such as the notion of sequential
bilingualism, or cultural bilingualism, overlap with definitions of the term second
language acquisition.

In this study, the term bilingualism refers exclusively to children being
simultaneously exposed to two languages in a natural environment from infancy
during their primary language Ieaming years (Genesee, 1993). The exposure to
both languages can be provided either in the home only or through additional
language contacts outside of the home in cases where the language spoken at

home is not the language of the speech community.

Prevalence of Bilin ualism in the nited tates

Although no “hard statistics” (Genesee, 1993) are available, the majority
of the world’s population might be considered bilingual if one regards various
degrees of competence of linguistic skills in more than one language as a
minimal common denominator found in definitions of bilingualism. Bilingualism
is not limited to specific regions of the world, nor to a specific socioeconomic
group. It is found in recent immigrants as well as in indigenous people.
Therefore, bilingualism is far from being an exceptional phenomenon.

According to the United States Census of 1990 (US. Bureau of the
Census, 1998, p. 56), 31.84 million people, which represents about 14% of the
total United States population, were identified as language minorities or families

in which non-English languages are spoken, predominantly Spanish (17 million),

followed by French (1.7 million), German (1.5 million), Italian (1.3 million),
Chinese (1.2 million), and several smaller language groups. Of this population,
6.3 million are school children who speak languages other than English at home
(US. Bureau of the Census, 1998, p. 55).

Presently, the proportion of children between 5 and 17 years of age who
speak languages other than English represents 14% of the total American
school-age population (45 million). Due to childbirth and immigration, their
numbers are projected to rise to 9 million in 2010, when they will amount to 22%
of the school-age population (Fix & Passel, 1994). Admittedly, these numbers
might not exactly represent the number of polyglot speakers in the United
States; however, they suggest that a significant part of the American population
may be confronted with aspects of bilingualism.

Despite its significant prevalence, interest in bilingualism has only
recently been emerging in the research literature. This seems odd because
valuable insights into the mechanisms of language Ieaming might be gained

from the study of bilingualism.

What Can Be Learned From the Study of Bilingualism?

From a theoretical point of view, the study of bilingualism offers unique
research opportunities because bilinguals may be regarded as their own
“matched pair,” according to De Houwer (1990), a fact that is especially useful in
child language research. Studies on language acquisition commonly have

reported huge variables in children’s performances, especially when they are

younger than 3 years of age. As Pearson, Fernandez, Lewedeg, and Oller
(1997) pointed out, this variability makes it difficult to separate influences of
outside variables from differences among children. Bilingual speakers, on the
other hand, make it possible to hold between-child effects constant and observe
influences of outside variables on the child. One would like to know, for
example, whether certain concepts, e.g., nouns or verbs, are easier or more
difficult to learn than are others. Studies of the simultaneous exposure to
different input patterns in different languages or cultures, such as Korean
emphasizing verbs and English emphasizing nouns in parental speech
addressed to young children, might shed light on this issue (see, for example,
Choi, 1997; GOpnik & Choi, 1995). This debate touches on the fundamental
controversy in the literature regarding whether specific input factors or universal
Ieaming strategies are prevalent in language Ieaming. Research on bilingualism
seems to offer an ideal opportunity to test whether universal principles apply
across languages, and thus provide insights into the nature of language Ieaming
in general.

Universalists look for shared strategies in learners, regardless of whether
they are mono-, bi-, or multilingual speakers. However, one aspect that
bilinguals do not share with monolinguals is the very fact that the former are
confronted with more than one language to learn. At the very least, the
demands made by the sheer quantity of words and morpho—syntactic rules to be
learned appear to be greater for bilinguals than for monolinguals. This prompts

the question of whether more effort might be required to process the additional

5

bl

De

 

 

J'fu.il'lm9b u. v ’9. , t .
l I \ I

information necessary to learn more than one language. MacNamara (1967)
raised the possibility that major milestones of language development might be
delayed in bilinguals because it may be more difficult, and thus take more time,
to learn two languages simultaneously than just one. This is an important
question to ask because language development seems to be related to other
areas of development, especially cognition. Although discussions about whether
cognition influences language development (e.g., MacNamara, 1972) or vice
versa (e.g., Gopnik & Choi, 1995) are far from being resolved, there is
agreement that at least part of cognitive development is mediated through
language (Johnston, 1992; Kahmi, 1993). Numerous studies have documented
that children with delayed expressive language abilities are more likely to
develop Ieaming disabilities or behavioral problems than children with normal
language development (see review in Carson, Klee, Perry, Muskina, & Donaghy,
1998,p.61)

If bilingual children lag behind in their language development, as is widely
believed, they might be at a disadvantage developmentally when compared to
same-age monolinguals because of the central role of language acquisition in
general development. One goal of this study is to shed light on the so-called
bilingualism deficit hypothesis (Oller, Eilers, Urbano, & Cobo-Lewis, 1997)
because professionals as well as parents have struggled with this concern in the

past

Bilingualism and Clinical Issues

As indicated above, the situation of children being exposed to more than
one language is not new, but it has only recently been addressed in the research
literature. The reason for this neglect might be that, in the past, tremendous
pressure to adapt to mainstream culture was exerted on immigrant and minority
populations. More recently, the institutional and political emphasis on diversity
in the United States has reflected the willingness of the society to legitimize
cultural differences among groups within the population, including those who
speak different languages and dialects (Battle, 1997).

The awareness of the possibility of speaking more than one language has
created the professional issue of how to assess speech-language abilities in
bilingual speakers. This is a problem because, according to a survey conducted
by Campbell and Taylor (1992), more than 80% of speech-language
professionals perceive themselves as either incompetent or only partly
competent to deliver speech and language services to bilingual speakers. Of
713 respondents, 594 self-reported a lack of competence to perform evaluations
and 570 self-reported a lack of competence to perform treatment with bilingual
speakers. Among the 18 clinical skills inquired about on the survey, evaluation
and treatment of bilingual speakers received the lowest ratings. The sense of
unpreparedness in dealing with bilingual populations might in part be due to the
scarcity of tools available to evaluate and treat these speakers.

Although a bilingual speaker, by definition, has knowledge in two

languages, assessing only one (in this case, English), as has been done in the

past, is not acceptable because the client’s knowledge might not be spread
equally between the languages. In fact, being completely balanced is rather
exceptional, even in adults (Grosjean, 1982). Grosjean claimed that, especially
in the lexicon, there will always be a proportion of words known only in one
language. Therefore, to simply test only in one language and apply norms that
have been derived from monolingual to bilingual populations would
underestimate the competence of bilingual speakers, even if it is their strong
language. To adequately judge language skills in bilingual children, their
specific language context should be taken into consideration.

Whereas a few Spanish and Italian versions of vocabulary tests exist,
assessment tools in other languages are practically nonexistent. This situation
puts pressure on speech-language pathologists because they have a legal
mandate to serve this growing diverse population of bilingual speakers. The
Bilingual Education Act of 1968, growing out of the civil rights movement, grants
equal educational access to all children in the United States regardless of race,
color, or national origin (Taylor, 1986). The Education of All Handicapped
Children Act of 1975 and its revisions, Public Law 94-142 and Public Law 98-
199, specifically addressees children with special needs and explicitly requires
clinicians to provide speech-language assessment and implement treatment
programs “in the child’s native language . . . unless it is clearly not feasible to do
so” (Taylor, 1986, p. 9).

This study was undertaken to ease the discomfort professionals feel when

faced with clients who acquire languages other than English by documenting

how the language skills of German-English bilingual toddlers can be measured.
As mentioned above, German constitutes the third largest foreign language
community in the United States to date, according to the United States Census

of 1990 (US. Bureau of the Census, 1998, p. 56).

STATEMENT OF THE PROBLEM

Need for Research on Early Bilingual Agguisition

Research on early simultaneous acquisition of two languages is scarce
and mostly derived from case studies. De Houwer (1990) compiled an overview
of existing case studies involving 33 bi- or multilingual subjects with language
combinations such as English plus French, Spanish, or Hebrew, as well as
German plus English, French, Spanish, or Italian. However, case studies allow
only limited generalizability of findings. It is even more problematic that data
have been collected and reported in different ways, which causes difficulties in
comparing results.

Only recently has work involved systematic group studies, and so far
these studies have focused mainly on Spanish-English bilinguals. Examples
include Pearson, Fernandez, and Oller’s (1995b) study of 25 Spanish-English-
speaking toddlers and Patterson’s (1998) study of 102 Spanish-English-
speaking toddlers from various social backgrounds. However, the above-
mentioned researchers did not compare bilinguals with both groups of
monolingual counterparts. This is surprising because different language
structures might place different demands on the learner. For example, bilinguals
may compare differently to age-matched peers in each language. Therefore,

more studies, including large numbers of subjects from diverse bilingual

10

populations of speakers, are necessary, as is the application of a more rigorous
methodology.
The Early Bilingual Lexicon as the
Focu_s of This Investigation
The huge information gap in the literature on early bilingualism is
surprising, given the fact that speech-language pathologists have found it
increasingly important to address the young language learner. When one
focuses on early language, the central role of vocabulary development becomes
evident. The initial stage of language acquisition, from the first words to the
emergence of word combinations, is basically lexical, with morpho—syntactic
rules just emerging. According to Clark (1993), the early lexicon offers a “unique
window on the process of acquisition for language as a whole” (p. 1). She was
especially interested in investigating cognitive-perceptual processes and
strategies involved in developing early semantic concepts, as well as processes
explaining how children attach meanings to words. If cognitive-perceptive
principles can be regarded as one of the forces driving early vocabulary
development, then semantic development should be more similar across
languages than, for example, grammatical development. Whereas the
acquisition of language forms clearly needs to focus on the syntactical and
morphological rules of a given language (Slobin, 1992), early semantic
development seems to be less language or culture specific, which would make it

especially well suited for cross-linguistic analysis.

11

Various findings support this hypothesis. It is generally believed that the
child’s perceptions of and experiences with people, objects, routines, and so on,
in his or her immediate environment form the basis of the cognitive concepts that
underlie the meanings of words (Nelson, 1974). To some extent, small children
may be assumed to have similar experience with their immediate physical and
social environments across cultures. For example, children of all cultures
interact with people in their immediate environment, mostly their families. They
explore their bodies and objects; they are fed, dressed, and so on. Therefore,
similar kinds of concepts or meanings should be accessible to all children
regardless of the language they are Ieaming. This does not imply that children
acquire the same lexical forms or represent exactly the same meanings with the
words they produce, but rather that they have words for referents in similar
categories, such as types of food, clothing, people, body parts, and different
types of activities. For example, a child growing up in the tropics would eat
different foods from a child growing up in the arctic, but they both would be
expected to have words referring to foods.

Other similarities were observed by Clark (1979), who found over- and
underextensions of meanings in early child vocabularies in many different
languages. Gentner (1982) and Caselli and colleagues (1995) found similarities
in the composition of the early lexicon across languages, namely a
predominance of nouns, whereas other word types, such as lexical verbs,
emerged later. Gentner suggested that children prefer nouns, particularly object

names, to words for encoding events or relations because they refer to stable

12

entities and are perceptually most salient in the input. But these findings
recently have been challenged in other studies. Femald and Hiromi (1993) and
Gopnik and Choi (1995), for example, claimed that noun bias found in Indo-
European languages, namely English, is language or culture specific. In
comparing studies on this issue, methodological differences become evident,
which might in part have shaped their results. Whereas Caselli et al. used
vocabulary checklists based on parental report, Femald and Hiromi’s and
Gopnik and Choi’s studies were based on free—speech samples. When the
checklist methodology was used to assess Korean toddlers, a verb dominance
as found by Gopnik and Choi disappeared. (See Pae, cited in Caselli et al.,
1995)

The literature has revealed some general similarities in the kinds and
types of meanings that are encoded by words in young children. Therefore, the
early vocabulary is suggested as a domain for assessment across languages.
This has practical value: If assessment relies on shared properties of languages,
instruments can be developed for use with more than one population of

speakers.

What Parameters Have Been Studied?

Monolingual Vocabulary Development
Vocabulary size. Expressive vocabulary size of at least 50 different

words by age 2 is one of the main parameters that has been used as a predictor

for later language development in monolinguals. Numerous studies and

13

developmental test norms of English-speaking toddlers have indicated that, by
age 2, most children have acquired a productive vocabulary consisting of at
least 50 words (Benedict, 1979; Clark, 1995; Coplan, Gleason, Ryan, Burke, &
Williams, 1982; Rescorla, 1989). Similar results were reported in Grimm, Doil,
Mueller, and Wilde’s (1996) pilot study on monolingual German 2-year-olds.

Follow-up studies on children who had not reached the 50-word mark by 2
years of age, so-called “late talkers,” indicated that language delay tended to
persist into later years (Ellis Weismer, Murray-Branch, & Miller, 1994; Paul,
1996; Rescorla, 1989; Resnick, Allen, 8 Rapin, 1984). However, prevalence of
later language delay in late talkers varied between 25% and 60%. According to
Kelly (1998), this variation in findings might be due in part to different measures
used. In general, spontaneous speech samples allow more detailed analysis
and generate higher positive findings than standardized measures. Another
reason might be that vocabulary size alone may not provide enough evidence to
identify late talkers. Klee et al. (1998) suggested that nonlinguistic factors, such
as use of symbolic gestures or cognitive measures, also be considered when
identifying at-risk language development.

The 50+ word mark is one of the milestones of the minimum vocabulary
skills expected by age 2, as measured by vocabulary checklists, and 95% of
normally developing children statistically perform at or above this threshold.
Numerous authors have reported, for example, that many children reach a
vocabulary size of 50 words as early as 18 months of age (Bloom, 1973; Fenson

et al., 1993; Nelson, 1974). Whereas the rate of vocabulary Ieaming seems to

14

be slow in the beginning (Goldfield & Resnick, 1990, 1996; Nelson, 1981 ), it
typically accelerates toward the end of the second year of life, a phenomenon
that is referred to as the vocabulary spurt. In the literature, it traditionally has
been assumed that the onset of the vocabulary spurt occurred when the child’s
lexicon contained about 50 words (Benedict, 1979; Bloom, 1973), but the “late
spurters" observed by Mervis and Bertrand (1995) had a vocabulary of more
than 100 words before their lexical growth accelerated. Mervis and Bertrand
provided evidence that all children spurt, but at different times with different
patterns of lexical growth, such as a sharp rise after a period of slow growth
(Goldfield & Resnick, 1990; Mervis & Bertrand, 1995), slow linear growth
(Goldfield & Resnick, 1990), or periods of acceleration with plateaus in between
spurts (Goldfield & Resnick, 1996). By the age of 24 months, children typically
can produce between 50 and 500 words (Fenson et al., 1994), with 50 words
being the minimal skill level separating typical from at-risk vocabulary
development.

Word combinations. The 50-word mark seems to coincide with the onset
of grammatical development, in its most basic stage: the emergence of word
combinations. Nelson (1974) found that a certain critical mass of 50 words
seems to be necessary before the child can begin to put words together and
thus combine semantic concepts into semantic relations that lead to sentences.

Numerous studies have indicated that, by 2 years of age, 85% to 95% of
English-speaking children begin to combine two to three words (Coplan, 1987;

Fenson et al., 1993). Similar findings were reported for German monolingual 2-

15

year-olds in Grimm et al.’s (1996) pilot study. Therefore, the emergence of word
combinations is considered another milestone that is reached by normally
developing monolingual toddlers toward the end of their second year of life, at
the latest by 24 months of age.

Vocabulary richness. Finally, Kauschke (1999) suggested that the
richness of vocabulary, as indicated by the ability to produce a variety of
different words and word types, be considered in efforts to identify at-risk
behavior. In studies on the composition of the early vocabulary, children with
language impairment tended to use more nouns and fewer verbs than MLU or
age-matched normally developing peers (Watkins, Rice, & Moltz, 1993). Rice
and Bode (1993) observed that a limited number of so-called general all-purpose
verbs (GAP verbs), such as have, make, and do, accounted for more than 50%
of verb usage by children with specific language impairments. Because the
authors did not include normally developing control groups, they could not
determine whether the reliance on GAP verbs was a compensatory strategy
employed by language-impaired children, or whether this was age typical for
preschool children in general. This question was investigated by Kelly (1997),
who found that although GAP verbs were frequently used by normally
developing peers as well, they appeared to use a greater variety of verbs than
did children with language impairment. Therefore, verb diversity may be
regarded as an indicator of vocabulary richness.

In summary, vocabulary size, emergence of word combinations, and

richness of vocabulary have been used to separate adequate from delayed or

16

deviant language development. Minimal skill levels have been established for

vocabulary size and the emergence of word combinations by age 2.

Bilingual Vocabulary Development

Vocabulag size. Patterson (1998), who investigated whether milestones
established for monolingual children also apply to bilingual children, found
positive evidence for a vocabulary size of 50+ words in her sample. Eighty-one
percent of her 102 Spanish-English bilingual subjects, ranging in age between
23 and 25 months, produced vocabularies consisting of 50+ words when
different words in both languages were combined. Between 26 and 27 months
of age, all subjects reached the 50+ word mark. Although no follow-up studies
have been performed to date to follow later language development of low-
performing bilingual children, it seems probable that a significant number of
these children might be at risk as well.

Porsche’s (1983) and Leopold’s (1939) longitudinal case studies provide
additional data for German-English bilingual vocabulary development. Both
Porsche’s son and Leopold’s daughter, Hildegard, reached the 50+ word mark
within their second year of life, according to their fathers’ diary reports.
Porsche’s son was reported to have a combined German-English vocabulary of
66 words by 22 months of age. Leopold’s daughter, by 16 months of age,
already produced 74 words spread over both languages. By 2 years of age, her
productive German-English vocabulary consisted of 241 words. More data,

preferably by systematic group studies on a variety of language combinations,

17

are needed to further illuminate this issue. This was one goal of the present
empirical study.

Word combinations. For her bilingual Spanish-English-speaking sample,
Patterson (1998) demonstrated that 84% of her subjects produced multiword
utterances by the age of 24 months, as did 100% of the subjects by the age of
26 months. Leopold (1949) reported the emergence of word combinations for
Hildegard around 18 months of age. No data on this topic were supplied by
Porsche. Again, more information about diverse bilingual populations is needed,
indicating the need for further investigation of preliminary findings.

Vocabulary richness. Vocabulary richness has been investigated across
languages (e.g., Caselli et al., 1995), but not within bilingual speakers. This
study was intended to provide first pilot data to shed light on qualitative aspects

of the bilingual’s vocabulary.

Conceptual Issues

Bilingualism Deficit Hypothesis

Considering the scarcity of information available on young bilingual
learners, practical as well as conceptual reasons motivated the undertaking of
this comparative study exploring early monolingual and bilingual vocabulary
skills.

Popular views of early bilingualism have been shaped by a widespread
belief that bilinguals are weak or inferior language learners. It is assumed that

learning more than one language at the same time places higher information-

18

processing demands on language acquisition, due to the child’s having to learn
two languages simultaneously. If this assumption, which Oller et al. (1997)
dubbed the bilingualism deficit hypothesis, is true, bilingual children would be at
a serious disadvantage. Implications of depressed language skills for cognitive
development have been discussed (Johnson, 1992; Kahmi, 1993). In regard to
vocabulary development, the bilingualism deficit hypothesis predicts that
bilinguals reach characteristic milestones, such as the 50+ word mark and
emerging word combinations, later, on average, than their monolingual peers
and that vocabulary size and richness should be reduced.

Preliminary findings from previous studies have suggested that bilingual
children produce fewer words than monolingual speakers in a given language,
but that they do not have a deficit when the vocabulary in both languages is
combined (Patterson, 1998; Pearson, Feméndez, & Oller, 1995b). Given that
both of these studies focused on Spanish-English speakers, it is not known
whether their results would hold for speakers of other language combinations.

In this study, a different language combination, German and English, was
investigated.

Further research on comparing monolingual and bilingual vocabulary
skills will not only shed light on the validity of the deficit hypothesis; it might also
contribute to an understanding of the bilingual process, if and how bilingual
children learn to separate languages. The deficit hypothesis feeds into another
debate in the bilingual-acquisition literature, namely, whether one starts out with

one or two languages. The information-overload assumption predicts a need for

19

maturity in order to separate languages, implying that the cognitive constraints of
the developing brain might not allow an immature speaker to switch between

languages voluntarily.

Unitagy Lexicon Hypothesis

The above-mentioned studies on the emerging bilingual vocabulary of
young children suggest that it is distributed over the two languages to some
extent. Early researchers on bilingualism argued that young children exposed to
two languages from birth go through an initial stage of language mixing in their
language acquisition, in which they do not distinguish between the languages
(Grosjean, 1982; Leopold, 1949; Saunders, 1988; Vihman, 1985; Volterra &
Taeschner, 1978). This involuntary language mixing or lack of language choice,
if genuine, would reduce communicative effectiveness, at least in monolingual
contexts. Thus, this theory, which Genesee (1989) referred to as the unitary
language system hypothesis, also supports the notion that the bilingual learner
is less competent than a monolingual speaker.

One type of evidence used in support of the theory of a single language
system comes from the study of the child’s lexicon: A child might have one
single, undifferentiated lexicon if different words in each language are produced,
with no evidence of cross-language synonyms or “translation equivalents”
(Pearson et al., 1995a).

The unitary lexicon hypothesis is supported by Clark’s (1987) earlier claim

that the principle of contrast works not only within, but also across, languages in

20

the early stage of vocabulary acquisition, until the child reaches a vocabulary
size of 150 words. In Clark’s theory, the principle of contrast can be thought of
as one of several cognitive strategies that enable the child to process language
within the cognitive constraints of the developing brain. In effect, the principle of
contrast predicts that all young children-even monolingual children—expect each
lexical form to have a different meaning. Therefore, synonyms are not possible.
Even though children get exposed to different labels, they produce only one and
reject others because, in the view of the child, they are synonymous. Applying
the same rationale across languages, a child’s Ieaming a word form in one
language prevents his learning the synonymous label in the other language.
This would result in a single lexicon. Under this premise, no or very few
translation equivalents should occur in the early lexicon.

However, recent researchers have questioned the hypothesis of a single
lexicon in early bilingual vocabulary development. Pearson et al. (1995a)
argued, for example, that the evidence of 30.8% translation equivalents, or
cross-language synonyms, produced on average in their study by Spanish-
English subjects suggested that the children were developing two different
language systems from the start. However, one could dispute whether the
evidence of translation equivalents can be taken as evidence for the capacity to
separate languages. Volterra and Taeschner (1978) argued that the lexical
equivalents found in their subjects were actually used by the children to refer to
different concepts. For example, one of their ltalian-German-speaking subjects,

Liza, used the word la (meaning there) for things she could not see while

21

speaking and the German equivalent da for things that were present during
speaking.

This argument was taken to heart by Baum (1997), who followed two
Spanish-English-speaking males, one from 18 to 30 months and the other from
15 to 19 months. She systematically used the same stimulus materials for play
and conversational sequences (toys, picture books) in both languages and thus
provided proof for “true” translation equivalents, in that Spanish-English doublets
observed in the children’s productive vocabularies really referred to the same
referent.

The fact that bilingual children appear to be able to functionally separate
both languages from early on has been used as further evidence refuting the
single-language hypothesis (Baum, 1997; Keppe, 1996; Quay, 1995). In her
case study, Quay found that her Spanish-English bilingual subject was able to
functionally separate the two languages as early as 12 months. The girl was
able to speak English predominantly with her monolingual English-speaking
grandmother and Spanish with her monolingual Spanish-speaking grandfather.
Quay’s results were consistent with Baum’s (1997) findings.

In the recent literature, there is evidence that increasingly speaks against
the hypothesis of a single lexicon in the early bilingual vocabulary. However,
researchers have focused on Spanish-English bilinguals. Evidence from other
language populations is necessary to generalize results. Furthermore, the study
of other languages might shed light on the question of whether the rate of

Ieaming or the proportion of translation equivalents in the vocabulary might be

22

affected by how similar the two languages are. For example, Spanish and
English are morphologically more different than alike, in comparison to German

and English.

Clinicflssggs

Collecting data about bilingual speakers becomes relevant to speech-
language pathologists because they are concerned with the diagnosis of what is
typical and what is exceptional. If the bilingual child is viewed as naturally or
normally slow at Ieaming languages, it would become a challenge to differentiate
normal and delayed bilingual development. This task is made all the more
difficult because few procedures exist that are targeted at bilingual populations.

But in the case of vocabulary assessment, instruments are available that
yield useful results, namely vocabulary checklists. The MacArthur
Communicative Development Inventories (CDI; Fenson et al., 1993) and the
Rescorla Language Development Survey (LDS; Rescorla, 1989) are commonly
used to assess vocabulary size in English-speaking toddlers, and both are
valued as highly reliable and valid assessment tools (Dale, 1991; Klee et al.,
1998). A checklist modeled after the above-mentioned instruments is being
developed for German monolingual speakers (Grimm et al., 1996).

Among the checklists available, the LDS taps two of the parameters
previously discussed to separate normal from at-risk language and vocabulary
development. These parameters are (a) evidence of a productive vocabulary

containing at least 50 different words and (b) emergence of word combinations

23

by 2 years of age (Rescorla, 1991 b). Neither of the checklists explicitly
measures composition of vocabulary or vocabulary richness, but these can be
easily inferred from the list.

Using vocabulary checklists for bilingual assessment has an added
advantage. Because the judgment of vocabulary skills commonly relies on
parental reports, the clinician would not be required to be a speaker of
languages other than English. Relying on parental judgment is useful because
experience has shown that small children do not demonstrate their potentials
readily in front of an unfamiliar adult, such as a speech-language pathologist,
even in informal play interactions (Dale, 1991). Furthermore, parental judgments
are based on experience with their children in many different situations and
settings, whereas a clinician typically sees the child once for a short period of
time. Therefore, parents are increasingly regarded as a valuable source of
information that should be used when trying to evaluate a young child.

Several researchers (Costarides 8 Shulman, 1998; Dale, 1991, 1996;
Rescorla, 1993) have found that vocabulary checklists are valid and effective
measures for assessing language skills in young children, even through the lists
are not without flaws (Pearson, 1998). Vocabulary lists do not represent the
child’s vocabulary in its entirety. Rescorla (1989), for example, demonstrated
that scores of vocabulary size in the upper and middle ranges have to be
regarded as a function of the size of the list. The larger the lists provided, the
larger the vocabulary measured. However, at the low end of the spectrum, size

of the vocabulary list does not seem to matter, which means that different-sized

24

lists appear to be equally sensitive in detecting at-risk populations (Rescorla,
1989)

This does not necessarily have to be regarded as a drawback. If one
expects bilingual children to lag behind in language development, as is widely
believed, then, logically, more bilingual children should fall under the at-risk
category and their performance should in general be tending toward the lower
range of the performance spectrum. Thus, using instruments that are sensitive
to low performance levels, such as vocabulary checklists, appears to be
especially appropriate for this study.

Pearson (1995b) and Patterson (1998) successfully used two-language
measures that were based on existing checklists developed for monolingual
English populations. Therefore, it seemed justified to develop a German-English
vocabulary checklist that also made use of existing tools. Among the checklists
available, Rescorla’s LDS was preferred for this study because it taps previously
discussed milestones of vocabulary development at one point in time. The
second reason is that the LDS is an open list, unlike the CDI, which allows the
parents to add words not appearing on the test form. This reduces bias and
makes it especially appropriate for cross-cultural assessment.

Although it appears that one can rightly assume that a bilingual checklist
developed for German-English bilingual toddlers will also prove to be useful,
research is needed to test its usefulness. Because German speakers comprise

the third-largest foreign language minority in the United States according to US.

25

Census data (US. Bureau of the Census, 1998, no. 59), it appears especially

relevant to develop diagnostic standards and tools to serve this population.

26

RESEARCH QUESTIONS

The general goal of this study was to determine whether 2-year-old
German-English bilingual toddlers exhibit comparable vocabulary skills relative
to their respective monolingual peers, as measured by Rescorla’s (1989) LDS.
The descriptive outcome might allow inferences about hypotheses on the
underpinnings of bilingual vocabulary acquisition. The following questions were
addressed:

1. Is the size of the early German-English bilingual vocabulary
comparable to that of respective monolingual peers, when different words in both
languages are added up?

2. Is the number of verbs in the lexicon similar or different among
German-English bilinguals and their monolingual peers?

3. How is the bilingual vocabulary distributed over the two languages?
That is, how much semantic overlap in the form of translation equivalents is
evident at age 2, and what proportion of words is produced in one language

only?

27

METHODS AND PROCEDURES

Subjects

Data were collected from 10 normally developing bilingual subjects
acquiring German and English simultaneously from birth or infancy. Nine
children were being raised in German-English-speaking homes in the United
States and Germany; one child was being reared in a German-speaking home
and had been exposed to English during daycare since 3 months of age. Eight
children resided in the area of Ann Arbor, Michigan, United States; one child in
the area of Ttibingen, Germany; and another one in the area of Aachen,
Germany.

For each language, a monolingual control group was recruited. The
monolingual English participants resided in Ann Arbor, Michigan, and the
monolingual German participants in Tiibingen, Aachen, and Kiel, Germany.

Each of the groups consisted of 5 males and 5 females, with 7 flrst-bom
and 3 later-bom children. At the time of assessment, the ages for all participants
ranged between 24 and 27 months (for bilinguals: M = 24.7, S_[_)_ = 1.06; for
German monolinguals: M = 24.9, S_D = 1.28; for English monolinguals: _M_ = 24.8,
S_D = 1.22). Statistical analysis using one-way analysis of variance (ANOVA)
revealed no significant age differences across groups (E [2,27] = 0.07, p > .05).

(See Table 1, Appendix A.)

28

Subject-Selection Criteria

Because this study was based on a relatively small sample, it was
important to ensure homogeneity to control for confounding variables that could
influence the results. The following selection criteria were applied.

Ag; Rescorla’s (1989) Language Development Survey (LDS), the
measure used for this project, is standardized to assess milestones of early
vocabulary development at 24 months of age. Consistent with Patterson (1998),
a narrow age range of :I: 3 months was permitted. In each group, 3 males and 3
females were exactly 24 months old. Of the remaining 4 children in each group,
2 ranged in age between 25 and 26 months and 2 had reached 27 months of
age. No participant was younger than 24 months of age.

five—Mg Each group required and included an equal number of males
and females in order to counterbalance gender effects.

Developmental status. Because the goal of this study was to investigate
the typically developing bilingual toddler, participants were included only if they
were developing normally. The developmental status was determined using
age-relevant questions from the Rosetti Infant-Toddler Language Scale (Rosetti,
1990), which is routinely used in the Ann Arbor school district (see Appendix D
for parent interview form). The Rosetti Infant-Toddler Language Scale screens
for receptive language abilities, as well as motor, cognitive, pragmatic, and play
skills, based on parents' reports or through observation.

Uneventful birth history and pregnancy were checked by inquiring about

Apgar values, birth weight, and whether the child was born at term date.

29

Children who failed the Rosetti Infant-Toddler Language Scale were excluded
from the sample, as were children with at-risk birth history, as determined by
Apgar values below 7, low birth weight (below 5 lb), or birth outside a 2-week
range of expected full-term date. Other exclusionary criteria were hearing loss
or repeated history of otitis media with effusion (more than two bouts within the
preceding year, according to Leonard, 1998). All children participating in this
study were typically developing toddlers with unremarkable births, normal
medical and developmental histories, and no known hearing problems.

Comgtence of language models. A child acquiring a language should be
exposed to competent speakers. To ensure competence of language models,
the participating parents were required to be native speakers (ideally) or to
demonstrate native-like fluency, with their skills having been enhanced by their
having spent a minimum of 2 years in the country being exposed to the language
used in child-parent communication.

In the bilingual group, 17 of 20 parents and grandparents reportedly
addressed their children in their native language. Only 3 parents chose their
weaker or second language to communicate with their child. Of these, 2 parents
had been raised bilingually themselves by immigrant parents until they reached
school age. Both parents took advanced German-language classes in high
school and college. They both thought that they were stronger speakers of one
language; however, they were able to easily follow and keep up conversations in

their other language. The third parent used her second language regularly at

30

work. At the time of the study, she and her child attended playgroups once a
week that, aside from her, consisted only of native speakers.

In the monolingual control groups, all parents were native speakers of the
main dialect of the majority language. Some had acquired knowledge of other
languages, but none of them reported using languages other than their native
languages to communicate with their children.

Educational background of parents. According to Hart and Risley’s
(1995) long-term study on 42 American families with children between 13 and 30
months of age, the amount of adult time spent interacting with a young child
appears to be the single most positive factor influencing vocabulary growth. It
was more important than birth order, ethnicity, or even gender. Hart and Risley
also reported that the higher the parents’ level of education, the higher the
amount of child-parent interaction. More time spent with the child offers
opportunities to provide mature language models. In their study, Hart and Risley
found that the amount of input, tied to educational level of the parents, was
significantly correlated with the productive vocabulary size of their children.

Families who elect to offer their child the opportunity of a bilingual
education, especially in a country where the language community is relatively
small, are likely to be motivated to spend time with their children. Extra effort is
necessary to provide the children with the books, games, videotapes, and
conversational partners of the minority language and culture. Therefore, the
sample was expected to be biased toward higher levels of education and middle

class.

31

Patterson (1998) criticized most studies on language acquisition for
drawing on samples that do not consider diverse socioeconomic backgrounds.
However, this was not seen as a drawback for answering the questions raised in
this particular study. If one wants to know whether learning two languages is, in
principle, as easy as or more difficult than learning just one, then it seems to be
advantageous to select participants for whom conditions for Ieaming are as
favorable as possible so that, in the end, negative environmental factors could
not be seen as responsible for possibly confounding the results.

To counterbalance the expected bias of the bilingual group toward middle
class, an above-average educational level, consisting of at least some
secondary training beyond high school, and professional employment of at least
one parent, were required of, and met by, all families.

In 9 out of 10 families in the bilingual group, one parent or grandparent
was able to stay home part or full time (8 mothers and 1 father) and had the
opportunity to spend time with the child. All mothers in the monolingual German
group stayed home full or part time, as did 9 mothers in the monolingual English-
speaking control group. Consistent with Hart and Risley’s (1995) findings, all
parents reported that they allocated time and attention for their child every day.
Of course, not only play but also everyday life activities, such as doing chores
together, may provide opportunities for child-parent interactions and dialogues.
But because families could not be observed in real life, as did Hart and Risley,
the amount of playtime was taken as an indicator of parents’ involvement with

their child. Estimations of time were based on parents’ reports of typical family

32

routines during weekdays and weekends. All households in the study provided
good conditions for active language Ieaming. It was striking to learn that book
reading and play were encouraged, whereas watching television, which also
provides opportunities for language input but is not interactive, was either not an
option or limited. None of the children were said to watch more than 30 minutes
of television per day, but they spent at least 1 hour per day engaging in play or
reading with their parents. To make a comparison: A 1999 United States survey
revealed that children between 2 and 17 years of age spent 2-1/2 hours
watching television every day on average (see national survey conducted by the
Annenberg Public Policy Center at the University of Pennsylvania, cited by Dart,

1999)

Additional election riteria for Bilin ual ub'ects

As discussed in the introduction, bilinguals are a heterogeneous
population. To increase homogeneity in the bilingual cohort of this sample, the
following additional criteria were applied.

Minimum exposure requirement. As Pearson et al. (1997) pointed out,
most parents who raise their children bilingually intend to expose them equally to
both languages, but a fully balanced exposure is rarely observed in reality.
However, they found that a certain minimum amount of input appears to be
necessary for successful bilingual development. In their longitudinal study of 25
Spanish-English bilingual toddlers, the above-mentioned authors observed that

children who received less than 20% exposure to a language (which equals

33

about 15 hours per week) were reluctant to use this language when interacting
with their parents. On the other hand, children who were reported to receive
higher amounts of language input were observed to use both languages
according to language context. Only 1 out of 19 children refused to use one
language despite having received almost balanced exposure (60% English and
40% Spanish). Therefore, Patterson’s (1998) requirement of 8 hours of
language contact is regarded as being too weak to select true bilingual
speakers, and the more rigorous criterion determined by Pearson et al. (1997) is
preferred.

Pearson et al. estimated that a child was awake about 12 hours a day on
average, which adds up to 84 hours per week. Hours spent with Spanish or
English speakers per week represent a fraction or percentage of 84 hours. As
the authors admitted, these guidelines are relatively crude because (a) they are
based on what parents estimated from memory and (b) they do not take the
quality of interaction into account. Time spent reading books, for example, has
the same value as time spent doing chores without necessarily engaging the
child in a dialogue. To increase accuracy, the parents were asked to describe
typical weekday and weekend activities and routines.

All 10 bilingual children included in this study met a minimum exposure
requirement of at least 15 hours per week for either language. Three additional
children who received less than the 15-hour-per-week minimum input in one

language were excluded.

34

Consistency of language model. Although a certain amount of language
mixing or code switching might be inevitable, even normal for bilingual speakers
(Grosjean, 1989), excessive mixing by language models has been found to be
disadvantageous for the child. The main concern is that it encourages language
mixing by the child and thereby reduces incentive to voluntarily separate
languages where necessary, as in the case of communicating with monolingual
speakers. A child whose parents constantly code switch might be likely to
imitate the model and also mix languages (Goodz, 1989; Kielhdfer & Jonekeit,
1998), which could be a confounding factor when analyzing whether the child
develops one or two separate vocabularies. Another disadvantage of language
mixing in the input language is that the amount of input provided in a particular
language is reduced by the amount of mixing, which could be significant,
depending on the extent of mixing.

Besides consistency of parental input during parent-child dialogues,
another type of consistency applies to rules of language use in different
contexts. There are different strategies reported in the literature, the most
popular one being the one-person-one-language approach (Kielhdfer &
Jonekeit, 1998). This strategy requires each parent to communicate with the
child in a different language. Other rules might allocate particular languages to
different settings, situations, times, and so on.

To meet the consistency criterion, clear-cut rules about contexts of
language use were required, although it was not critical to adhere to a particular

strategy. Consistency of language use during conversations with the child was

35

observed by the writer during free-play sequences. Families who reported
inconsistent language use or were unable to communicate consistently with their
child in one language for 95% of the time during videotaped play interactions
were excluded from the sample.

The language-modeling criterion was met by all parents participating.
One additional family had to be excluded due to frequent language mixing during
videotaped parent-child interaction.

All families in this study reported applying rules specifying contexts in
which a certain language was to be used. The one-person-one-language
strategy was reported by 80% of the parents; one family alternated speaking
German or English on different days of the week, and one family spoke German
at home and the child spoke English during daycare. When company was
present or during activities outside of the home, 8 out of 10 families reported
using the majority language even when communicating with the child. According
to the parents, none of the children had reacted negatively to rules of language
separation and none had so far rejected speaking one of the languages. In one
family, older siblings (3 years and 5 years of age) were reported as not speaking
their weaker language even though they were consistently addressed in it by
their mother. For both siblings, the amount of exposure to their mother's
language was reduced after attending all-day kindergarten programs that were

taught in the other language.

36

Recruitment of Subjects

Participants for this study were recruited in part by contacting
organizations that promote German language and institutions that target young
children in general. Advertisements asking for participation in the project were
posted at the German Cultural Center (Goethe Institute), the Steiner School,
playgroups, pediatricians’ offices, and daycare centers. Families who contacted
the writer received a one-page informational letter about the project and
conditions of participation (see letter to parents in Appendix D). The families
were asked to give or withhold written consent within a 2-week period (see
consent form in Appendix D). All families who contacted the writer were willing

to participate in the study.

Description of Vocabulagy Measures

Original Monolingual English Version of
the Language Development Survey

 

A standardized parent report instrument, the Language Development
Survey (LDS) developed by Rescorla (1989), was used to assess expressive
vocabulary at a single point in time (see Appendix C). The LDS was preferred
over the MacArthur Communicative Development Inventories (Fenson et al.,
1993) because it does not require the child to know specific words. Unlike the
MacArthur Communicative Development Inventories, the LDS allows the parents
to add words that are not included on the checklist. This reduces possible bias
and makes it more appropriate to use for cross-linguistic and cross-cultural

assessment.

37

Content of the LDS. The LDS was specifically designed to screen
expressive vocabulary of monolingual English-speaking toddlers at 24 months of
age. It provides a checklist of 309 words organized into 14 semantic fields
containing 31 words for types of food, 11 words for toys, 10 words for outdoor
objects, 9 words for places, 21 words for types of animals, 21 words for body
parts, 32 words for household items, 14 words for personal items, 15 words for
people, 17 words for types of clothing, 10 words for vehicles, 52 action words, 33
modifiers, and 33 miscellaneous words, such as hi, bye, please, meow, night,
and so on. The words were drawn from Rescorla’s previous research on lexical
development with normal English-speaking toddlers (see review in LDS manual,
Rescorla, 1991b). For each semantic category, Rescorla chose a core set of
highly frequent words as well as some words less frequently used by the
children in the sample. As mentioned before, the LDS additionally provides a
section for the parents to note words within the child’s vocabulary that are not on
the checklist. Parents are also asked whether word combinations are emerging
within their child’s vocabulary and, if so, to provide three examples. Children are
identified as being at risk when their vocabulary is smaller than 50 words and/or
no word combinations are emerging.

Reliability and validity of the LDS. Since initial publication of the LDS, a
number of studies have investigated its validity as a diagnostic screening tool.
Strong reliability was reported by Rescorla (1991b), with internal consistency

and test-retest reliability coefficients as high as g = .99. Point-to-point reliability

38

for individual words was somewhat lower, with r > .90 for 32% of the items and
r > .70 for another 55% of the words (Rescorla, 1991b).

Concurrent validity of the LDS has been tested with several established
measures. In a recent study administered by Klee et al. (1998), the LDS was
correlated with a 20-minute spontaneous language sample, plus the Infant
Mullen Scales of Early Learning. A high concurrent validity with levels of
sensitivity and specificity of 91% and 87%, respectively, was reported, which
means that the LDS identifies nearly all language-delayed children correctly
without falsely identifying too large a proportion of normal children as delayed.
This outcome is consistent with two other studies in which concurrent validity of
the LDS with the Reynell Development Language Scales (Rescorla, 1991b) and
with the Bayley Scales of Infant Development (Rescorla, Hadick-Wiley, &
Escarce, 1993) were tested. Levels of sensitivity and specificity reported were
at 89% and 86% in the first study and at 90% and 95% in the second study.

In contrast to concurrent validity, the predictive validity of the LDS was
somewhat lower. To calculate predictive validity, children screened at age 2
using the LDS were reevaluated at age 3. Depending on the prevalence of the
disorder, the positive predictive value was reported to be between 18% and 37%
(Klee et al., 1998). Based on a prevalence of 3%, only one in five children who
initially screened positive remained so at later assessment. If the prevalence
was assumed to be 7%, then one in three children remained positive. The
negative predictive value was 99%, which means that almost all children

identified with normal language development remained so at later assessment.

39

According to Klee et al., these somewhat lower levels of positive predictive
validity might indicate that a vocabulary checklist alone is not enough to identify
late talkers. Other criteria, such as use of symbolic gestures, language
comprehension, cognitive skills, and so on, should be considered in addition to
the LDS score to home in on truly delayed children at age 2. However, this does
not minimize the value of the LDS for assessing expressive language skills, as
has been demonstrated by its high concurrent validity with other tests that
identify delay, such as the Infant Mullen Scales of Early Learning, the Reynell
Development Language Scales, and the Bayley Scales of Infant Development.
Because of its strong validity, the LDS has been used as a criterion
measure for subject selection in two major studies of English-speaking late
talkers (Paul et al., 1993, 1995; Rescorla, Roberts, & Dahlsgaard, 1997). As
mentioned before, it also was tested in Patterson’s (1998) cross-linguistic

validation study for Spanish-English bilingual toddlers.

German Translated Version of the
Language Development Survey

 

Because no vocabulary measure existed at the time of this study to
assess German speakers at age 2, the writer translated the LDS into German,
based on the assumption that lexical development may to some extent be similar
across cultures for young children, as discussed earlier.

Somparison of English and German LDS items. To verify this assumption
and to check whether the English and German versions of the LDS are of

comparable difficulty, 50 items were randomly selected from the list and

40

analyzed separately for each language in regard to frequency of occurrence in
children’s vocabularies. Unfortunately, no published corpora describing
children’s vocabulary as young as 2 years of age exist to date. Corpora that are
available sample older children and thus have to be used with caution when
implications are made for younger ages. Based on the American Heritage Word
Frequency Book (Carroll, Davies, & Richman, 1971 ), which contains 86,741
different words (types) drawn from a sample of 5,088,721 words (tokens) in more
than 1,000 published texts used in schools for grades 3 through 9, it could be
determined that all 50 items sampled on the English version of the LDS were
among the 10,000 words most frequently occurring. (See Word Frequency for
English LDS items, Appendix 8.) Detailed analysis revealed that roughly two
thirds of the items (68%) were among the 2,000 most frequent words and almost
half of the items (42%) were among the first 1,000 words most frequently
occurring in children’s texts used in schools.

Pregel and Rickert (1987), on the other hand, collected samples of
productive speech to analyze word frequency in vocabularies of young German
speakers. (See Table 2, Appendix A.) Their data pool, consisting of 260,374
words (tokens), was based on elicited oral and written discourse of German
Elementary School students from grades 1 through 4. Analysis revealed that
70% of the German LDS equivalents sampled were highly frequent in first
graders’ vocabularies, and all items were present in third graders” vocabularies.

(See Word Frequency for German LDS items, Appendix B.)

41

Furthermore, English and German LDS items of the random sample were
ranked, based on frequency of occurrence in children’s vocabularies. Their
ranks were compared using the Spearman rank difference correlation.

Statistical analysis revealed a moderate significant correlation between ranked
items across languages with r (30) = .64 (p < .05).

Translation issues. The translation of LDS items into German was
straightforward for the majority of words. Only a few words did not have an exact
equivalent in German. Some simple English verbs, e.g., [to] nap or [to] lunch
require a phrase in German, such as Mittagsschlaf machen or Mittag essen. In
order not to expect a higher performance from a German than from an English
speaker, the German child was expected to produce only the noun part of the
phrase. This decision is consistent with the LDS, which lists the adult phrase [to]
eat breakfast as breakfast under the rubric of event words. Another difference
between the languages is the German partiality for compound nouns where
simple nouns are used in English. For example, floor translates to FuB-boden,
which contains three times as many syllables as the English equivalent. Again,
in order not to expect higher performance from a German speaker, only the main
part of the German compound had to be produced in order to be counted, unless
the whole compound was necessary to discriminate it from other items, such as
Hausschuhe versus Turnschuhe versus Handschuhe (house slippers, tennis
shoes, gloves). If only part of the compound was expected, it was indicated on
the list by putting the less important part of the compound in parentheses, such

as (FuB)boden.

42

Finally, for three words on the English vocabulary list, dialectal
alternatives were listed on the German version because they are typical for
different regions of Germany, even for speakers of High German. For example,
for bye the alternatives tschr'iB and ade were listed.

All items considered, only 17 out of 309 words (about 5% of the total
number of words provided) posed some translation difficulties as described
above. The translation of more than 90% of the words was straightfonrvard.

The German version of the LDS was tested in pilot studies with one 24-
month-old monolingual German-speaking boy residing in Kiel, Germany, and
one bilingual German-English-speaking 17-month-old boy residing in the United
States at the time. Minor changes were made following suggestions of the
parents who filled out the German vocabulary checklists. (See Appendix C.)

In summary, the LDS appears to be a highly reliable and valid instrument
for assessing lexical skills of English speakers at age 2. Word-frequency
analysis revealed that items used on the LDS are found in children’s
vocabularies of both languages. About two thirds of the items were among the
most commonly occurring words, whereas one third were less frequent. For the
translation, minor adaptations were made to ensure similar levels of difficulty.
Because the level of difficulty appears to be comparable across languages, it

seems legitimate to use the LDS for both English and German.

43

Data-Collection Procediures
Data were collected during one or two home visits to the families no more

than 1 week apart. Both visits together did not exceed 2 hours.

P_arent Interview

A 30-minute interview with one parent was conducted to verify the child’s
uneventful medical history (no premature birth, no major illnesses or disabilities,
no history of recurring ear infections), and normal sensory-cognitive
development.

Furthermore, information about possible environmental factors, such as
educational background of the parents, socioeconomic status (here: employment
status), patterns of language usage, and duration and type of parental
interaction with the child were requested (see parent interview form in Appendix

D).

Informal Observation of Child’s Spontaneous Play

A 15- to 30-minute sequence of spontaneous play between the child and
the parent was videotaped as an additional source of information about the child.
It aimed to verify whether the child was normally developed, and to check
parental consistency of language usage. However, only the data collected on
the vocabulary checklist were used for this study.

All video recordings were made using a Panasonic AG-EZZO digital

camcorder with a built-in microphone. Because the video recordings served as

44

additional evidence for subject selection and no transcriptions of speech

samples were made, no external microphones were used.

Collection of Vocabulam Data

About 30 minutes were needed for one parent to complete the LDS. In
the bilingual group, a second checklist in the other language had to be
completed between 3 and 7 days later. The order in which the LDS was
completed for each language was randomized.

As per the original LDS instructions, the parents were asked to mark each
word within the child’s production regardless of the child’s pronunciation as long
as the referent was stable and clear. If the parent was not sure whether the
child’s word production was recognizable for somebody outside of the family, he
or she was asked to note it on the form, so that it could be decided later whether
to count the word or not. For example, in one family the child said [nana] when It
wanted to be nursed, which was not counted. But [gak] for duck or [fant] for
elephant was counted because the pattern of articulatory simplification was
consistent with common observations of young children’s speech.

Parents were instructed to identify only words that the child
spontaneously produced on its own without any adult modeling. Words or
phrases that were merely imitated, or words that were understood but not
actively produced, were not to be checked on the form. The parents also were
asked if their child combined words and, if so, to provide examples. This

instruction deviated slightly from Rescorla’s original instructions, which asked for

45

examples of the child’s best and longest sentences. Following Patterson’s
(1998) suggestion, any examples of word combinations were accepted because,
otherwise, the parents tended to dwell on the task.

In contrast to Patterson (1998), who condensed the Spanish and English
vocabulary into one checklist, two separate lists, one in English and one in
German, were handed to the parents on two different days, no more than 1 week
apart. This procedure was used to avoid bias from one language to the other
and to help the parent keep the languages separate. When the parent was
asked to fill out the checklist in German, only German was spoken with the
family on that occasion, and vice versa. English and German checklists were
presented during the first or the second home visit. Due to the time constraints
of the examiner, two bilingual families mailed back their second checklist.

As in Patterson’s (1998) study, rater bias was controlled by having the
same parent complete the LDS for both languages. But in contrast to Patterson,
who reported having to include raters with only partial proficiency in one of the
languages, this study required native competency in one language and native-
like fluency in the other language in order to provide valid ratings.

The mothers most often supplied the data for this study: 7 mothers, 2
fathers, and 1 grandmother participated in the interview in the bilingual group; 8
mothers, 1 father, and 1 grandmother in the monolingual English control group;
and 9 mothers plus 1 father in the monolingual German control group. Reliability

data were collected from 18 parents (see Reliability of Data section).

46

Data Analysis
Methodological Issues

Definition of what counts as a word in child language. Before vocabulary

size can be calculated, it has to be clear which types of utterances in the child's
expressive repertoire can be counted as words. There are two factors that have
to be taken into account when analyzing child language at age 2: (a) the child’s
phonetic inventory is not fully developed, and (b) word meanings in the child
language might not be congruent with word meanings in the adult language.
Anybody listening to young children will notice simplification of the
phonetic structure in words, such as [laid] for slide, or [nana] for banana, as well
as consonant substitution, for example, [9, k] for [d, t], as in [gak] for duck (see
review of phonological development in English- and German-speaking children
by Fuhring, Lettmayer, Elstner, & Lang, 1985; Vihman, 1996). Because phonetic
deviations are commonly observed in young children’s language production, the
authors of both available vocabulary checklists (Fenson et al., 1993; Rescorla,
1991b) accepted “baby talk” and counted these utterances as words.
Unfortunately, they are sketchy when defining what they mean by baby talk. In
this study, children’s utterances were counted as words, as long as the parents
perceived the word forms as stable, regardless of whether the child’s utterances
contained consonant substitutions, simplification, or distortions or omissions.
Semantically, young children are often observed to over- or underextend
words, e.g., dog might refer to dogs and wolves, or dog might only be used to

refer to the child’s own pet. Again, the parents were asked to note whether the

47

child’s word deviated from the target, so that it could be decided later whether to
count it in the analysis. Similar to phonetic approximations, over— and
underextensions of words might be interpreted as semantic approximations of
target words. They are counted as words, but only once regardless of the
number of referents. In the above example, in which dog referred to dogs and
wolves, dog would only be counted once as one word.

Following Rescorla (1998), onomatopoeias, such as [tju-tju] for train,
were also counted as words. The parents were asked to write these types of
utterances next to the target word, as well as utterances that they were not sure
could be regarded as a word. However, proto-words that had no obvious
semantic relationship to the target word were not counted. One example is the
lexical form [nana] to express a request to be nursed.

As in the original LDS, proper names and names of letters and numbers
were counted as only one item, regardless of how many the child produced.

Discrimination of the two languages. Considering that small children tend
to simplify words phonetically, they might use the same phonetic form in both
language contexts, especially given the fact that German and English are highly
related languages. Of the 309 items used in both versions of the LDS, 162
consist of cognates, which are words that can be traced back to similar
etymological roots, such as Hand and hand. The concern that arises is whether
parents can make valid judgments about their child’s language choice when
cognates are used. On the checklist the parents might check off target words

only in one language without a verifiable basis for their language choice, or they

48

may possibly credit the child’s utterance on both checklists, which might inflate
the results.

Unfortunately, to date there is no information available about phonological
development in bilingual toddlers to address the concern discussed above.
Therefore, inferences have to be made from data that are available. One source
of data is studies investigating young children’s perceptive abilities because
perception is a prerequisite for speech production. Various studies on auditory
skills in young children have documented a high sensitivity to the ambient
language from early on (see review in Locke, 1993, chapter 4). Mehler,
Lambertz, Jusczyk, and Amiel-Tison (1986), for example, found that 4-day-old
infants from French-speaking families reacted differently when addressed in
French than in Russian. Eilers, Gavin, and Oller (1981) reported that
monolingual English and Spanish infants were able to discriminate minimal
phonological contrasts in novel stimuli (VCV), following rules of their ambient
languages, at 6 to 8 months of age. Even more relevant are data on productive
skills of young children across languages. Boysson-Bardies, Halle, Sagart, and
Durand (1989), who examined spontaneous babbling of 10-month-old infants
reared in monolingual English, French, Arabic, and Cantonese environments,
found that vocalic elements in the children’s babbling already resembled vowel
contours of their ambient languages. An explanation might be that vowels are
motorically easier than consonants, which is supported by the fact that
language-impaired children rarely produce vowel errors (FIJhring et al., 1985).

Despite the lack of data on bilingual phonological development, it might be

49

inferred that the 2-year-old children targeted in this study should at least be able
to produce proper vowel-quality words.

This is important because the vowel spaces for German and English are
distinctly different (Kohler, 1977; Ladefoged, 1982). When the phonological
structures of items used on the LDS are compared, it becomes evident that
among cognates it is particularly vowel quality that accounts for differences,
whereas some consonants, mostly initial consonants, are retained, such as in
Brat/bread. See phonetic transcription of English and German LDS items in
Appendix 8. Because 2-year-old children are expected to produce proper
vowels as well as some consonants, they are likely to make distinctions between
cognates, which should be detectable by their parents. Levels of above 80%
agreement of parental judgment in test-retest as well as inter-rater reliability

ratings performed by this researcher support this assumption.

Calculation of Vocabulary Size

Monolingual children. For monolingual children, vocabulary size was
determined by simply counting the number of different words checked off on the
LDS. Following Pearson et al. (1995), the number of different words was viewed
as a rough measure of the number of semantic concepts for which the
monolingual child has a lexical representation.

Bilingual children. With bilingual children, the relationship between words
and semantic concepts is more complex. Ideally, two different word forms

should be associated with each semantic concept. Therefore, the number of

50

word forms and concepts is not expected to be congruent, as is the case with
monolingual children. A complicating factor is whether so-called “translation
equivalents” really do evoke the same mental image in the child (Pearson,
1998). For example, the semantic pair dinner/Abendbrot illustrates that the
meanings of some words may be culture specific and not fully equivalent.
Following this argument, it would be necessary for a bilingual child to acquire
twice as many concepts as a monolingual child in order to be fully
communicative in both cultures. On the other hand, a monolingual English-
speaking child growing up in the United States, for example, might also be
confronted with different cultural influences from each parent due to the fact that
American society is diverse. Different religious, ethnic, or geographical roots
within the family might shape meanings of words in different ways. For example,
the phrase holiday season might evoke very different images of rituals
depending on the background of each parent. Without minimizing the existence
of cultural differences, especially between different language communities, one
should also not overemphasize them. The very fact that it seems possible to
translate from one language into another implies that there has to be at least
some semantic overlap between concepts across languages. Therefore, the
term translation equivalents was used with the implied understanding of two
phonetically different word forms sharing a core meaning, at the same time that
they might represent different connotations across languages (and cultures).

Vocabulary size in bilingual children is calculated in two different ways:

51

1. The numbers of different word forms checked off on the English
and the German vocabulary list were added up to calculate the total vocabulary.
In this calculation, translation equivalents were counted as two words, with the
exception of phonetically indistinguishable pairs, such as fish/Fisch. Words of
foreign origin, or English words used on both lists, were also counted only once,
such as pizza/Pizza or MacDonald’s. Also counted only once were the name of
the child, and a pet name. Unpaired items made up 4.5% (14 items) of the
sample and were counted only once when calculating the total vocabulary.

2. To assess semantic knowledge, or total conceptual vocabulary, the
number of different concepts was added up. Under this premise, translation

equivalents were counted as one concept.

Qalculation of Vocabulary Richness

As discussed before, the number of verbs is seen as an indicator of
vocabulary richness. For all groups, the number of different verbs recorded was
summed up.

Calculation of Proportion of Translation
Equivalents in the Chilgg Vocabulagy

All word pairs associated with a single referent were counted as
translation equivalents. Not counted were word pairs with indistinguishable
phonological structures, such as fish/Fisch. Calculated was the proportion of

concepts for which the child was reported to produce labels in each language.

52

An example below (for vocabulary of a fictitious child) illustrates the calculation

method used.

English Serman

fish F isch (Phonetically indistinguishable)
dog Hund (translation equivalent)
car Auto (translation equivalent)
duck
turkey
Blume
Vogel
F lugzeug

The pair fish/Fisch is phonetically indistinguishable; therefore, it is
ambiguous and not counted. The remaining vocabulary consists of two
translation equivalents: dog/Hund and car/Auto and five single lexical forms with
no equivalent in the other language. So the child’s vocabulary amounts to two
paired and five unpaired words. The proportion of paired words or translation
equivalents in the vocabulary is therefore 28%.

The preceding calculation deviates from Pearson et al. (1995a), who did
not calculate for how many concepts the child had double lexical
representations, but they used total vocabulary as a reference point. In the
above example they would calculate a total vocabulary of 11 words. Of these 11
words, four (dog/Hund and car/Auto) would be counted as translation

equivalents. They suggested counting fish/Fisch, the phonetically

53

indistinguishable pair, as one word, reducing the total vocabulary to 10 words.
In Pearson et al.’s view, 5110 words would be described as translation
equivalents, amounting to 50% of the total vocabulary. In this calculation, the
proportion of overlap between languages appears higher because translation
equivalents are counted twice (as two words versus one pair). Because this
seems to inflate the amount of overlap, especially in the low ranges, the present
researcher preferred the other calculation method. But to make comparisons to

Pearson et al.’s study, percentages were calculated both ways.

Relianility of Data

Test-retest and inter-rater reliability data were collected from three sets of
parents in each group.

Inter-@ter (test-retest) relia_bi|itv. There was a high agreement on
reported vocabulary size for all groups between the initial and the second
measurement, which was repeated 5 to 10 days after the initial administration by
the same parent. Levels of I = .92 for the bilingual subject group, _r; = .93 for the
monolingual English group, and r for the monolingual German control group
were somewhat lower than levels of _r_ = .99 reported by Rescorla (1991b) and
Patterson (1998), but they still demonstrate a high stability of rating over a short
period of time.

Despite the fact that item-by-item analyses are rarely reported in the
literature, they are even more revealing in regard to reliability of an instrument.

In this study the agreement between items reported during initial testing and

reassessment continued to be very high: A mean point-to-point agreement of r =
.87 was demonstrated by the bilingual group, 1 = .90 by the monolingual English
group, and r = .92 by the monolingual German control group. In comparison,
Rescorla (1991b) only reported levels of point-to-point agreement between 5 =
.70 and g = .89 for 83% of the items used on the LDS, which might be due to a
greater diversity in regard to socioeconomic background and parental level of
education in her sample as compared to the present sample.

Of the 9 children who were reassessed, 6 had a larger vocabulary at the
second reporting, whereas the vocabulary size for 3 children was reported to
have decreased. A dominant tendency of a vocabulary increase was expected
because children continue to develop during assessment intervals. Repeated
measurements might also have increased the parents’ awareness of the child’s
language development, which might have increased their ratings.

In_tegflter relia_bi|itv. To measure the stability of the overall LDS
vocabulary score across different raters, both parents were asked to assess their
child’s vocabulary during the first visit; their scores were then compared.
Reported vocabulary size revealed moderate levels of agreement for the
bilingual group (i; = .74) and high inter-rater agreement for both monolingual
control groups (monolingual English control group 5 = .87, monolingual German
control group E = .92). The lower levels of reliability in the bilingual group were
in part caused by one parent, whose assessment of vocabulary size significantly
deviated from the spouse’s in regard to vocabulary size of one language (25

words versus 95 words reported by the spouse). Nevertheless, item-by-item

55

analysis revealed that 22 of the 25 words reported by the parent were also
checked off by the spouse, which corresponds to 88% agreement. The parent
who had reported the low vocabulary score in one language was also reported
not to communicate much in that language with the child, which may in part
explain his rating. Omitting this particular parent’s rating, the level of inter-rater
agreement would rise to g = .84, which is only slightly below the levels reported
for the control groups.

ltem-by—item analysis again revealed slightly higher inter-rater agreement
for the control groups than for the subject group (5 = .76), with levels of r = .84
for the monolingual English group and r = .82 for the monolingual German
group.

In summary, reliability testing revealed high stability of vocabulary scores
over time and moderate to high agreement between different raters. The
general tendency of somewhat lower agreement in ratings observed within the

bilingual group might be due to the fact that more variables had to be judged.

Statistical Treatment of Data

Mean values and standard deviations were calculated for all dependent
variables. Variations of scores among the three subject groups were subjected
to statistical analysis using one-way ANOVA, and further analyzed with Tukey t-
tests. Chi-square correlations were used to analyze individual scores within

groups.

56

RESULTS

Vocabulapy Size

Total once tual Vocabula

The conceptual vocabulary is defined as the number of different semantic
concepts for which the child has a lexical representation. For a monolingual
child, the number of different concepts is assumed to be congruent with the
number of lexical forms. Because a bilingual child may associate two labels,
one in each language, with a single conceptual referent, the conceptual
vocabulary size may differ from the number of lexical forms. For the bilingual
group, different concepts in both languages were pooled.

Figure 1, depicting average conceptual vocabulary size across groups,
shows that the conceptual vocabulary reported for the bilingual group was close
to that of both monolingual control groups. Compared to the bilingual group
(BL), the average conceptual vocabulary was slightly higher for the monolingual
German group (MLG) and slightly lower for the monolingual English group (MLE)
(BL: M = 210.6, S_D = 48.15; M = 224, Q = 50.6; MLE: M = 205.7, M = 10 for
each group). One-way ANOVA revealed no significant differences between
average conceptual vocabulary across groups (_E [2,27] = 0.22, p > .05). I-tests

performed to compare individual groups with each other confirmed the above

57

results (BL and MLG: I [qr = 17] = 0.16, p > .05; MLE and MLG; ILd_f = 12) = 0.58,

p > .05).

400
375
350
325
300
275
250
225
2C!)
1 75
1 50
1 25
1 00
75
50
25
0

 

NUMBER OF WORDS

 

 

 

 

 

‘-
db
-

BL lexical BL
forms concepts
(N=10) (N=10)

Subject group

MLE
(N=10)

 

Figure 1. Mean number of lexical. forms and concepts across the bilingual (BL),
the monolingual German (MLG), and the monolingual English (MLE) subject
groups.

Note. For the monolingual groups, the number of lexical forms was congruent
with the number of concepts.

Individual conceptual vocabulary scores depicted in Figure 2 show wide
ranges of individual performances within and across groups with no group bias.
Each group contained low and high performers. Whereas 1 subject performed
above and 1 below one standard deviation from the mean (M = 213.43, fl
pooled = 64.03) in the bilingual (BL 6 = 286; BL 1 = 143) and in the monolingual
German group (MLG 8 = 308; MLG 3 = 117), 1 subject performed above and 5 at

or below one standard deviation from the mean in the monolingual English group

(MLE1= 363 vs. MLE 8 = 149; MLE 9 = 148; MLE 2 = 133; MLE 3 = 124).

58

0 BL I MLG A MLE

 

 

4001-

350» A

300-- o '

A 9% a . +150
75%. 250" I I - A ‘ o 0
ES 2004. ---------- .— —'i_T—;_T_— M08"
S5 0 o o o
o
6

100» A

50....

o r r 9 fi- : .L I r : :

 

 

0 1 2 3 4 5 6 7 8 '9 10
Subject number

Figure 2. Individual vocabulary scores within and across the bilingual (BL), the
monolingual German (MLG), and the monolingual English (MLE) groups.

Note. For the bilingual group, conceptual vocabulary scores were used.

 

Comparison of individual scores across groups revealed the maximum
and that the minimum score was produced by subjects from the monolingual
English group, which performed lowest on average ( maximum number of
different concepts for MLE 1 = 363; minimum number of different concepts for
MLE 10 = 102). Three monolingual English subjects showed performance levels
below the lowest bilingual scores (MLE 10 = 102; MLE 3 = 124; for MLE 2 = 133
vs. BL 1 = 143), and 1 subject from each monolingual control group showed
higher scores than the highest performer in the bilingual group (MLE 1 = 363,

MLG 8 = 308 vs. BL 8 = 268).

59

Total Lexical Vocabulagy

Total vocabulary is defined as the total number of lexical forms produced
by the child. In contrast to the close proximity of average conceptual vocabulary
across groups, Figure 1 shows much larger differences between average total
vocabulary of the bilingual group versus average vocabulary size of both
monolingual control groups (BL: M = 303, SQ = 92.1 ). Statistical ANOVA
revealed significant differences between the bilingual across groups (1: [2,26] =
3.48, p < .05). However, further analyses with multiple comparison Tukey (a)
tests revealed that only one pair of means between the bilingual and
monolingual English groups was significant (Tukey [a] HSD = 87.5, BL [total
vocabulary] vs. MLE [total vocabulary] = 98.2, p < .05). The differences between
the bilingual and monolingual German groups, although sizable, did not reach
significance at conventional levels (Tukey [a], HSD = 87.5, BL [total vocabulary]

vs. MLE [total vocabulary] = 79.9).

Vocabulary Richness

As discussed earlier, the number of different verbs or event words in the
vocabulary is regarded as an indicator of vocabulary richness. Figure 3 depicts
the average number of different verbs calculated for each group (BL: M = 32,
S_D= 11.65; MLG: M = 37.2, SQ = 9.61; MLE: M = 30.78; SQ = 15.45). As with
conceptual vocabulary, the monolingual German group demonstrated the

highest average verb score, followed by the bilingual group and the monolingual

60

English group, but the differences observed did not reach significance when

statistically analyzed using ANOVA (E [2.26] = 1.07, p > .05).

“T
‘5" up
40.
35.
30..

 

NUMBER OF DIFFERENT VERBS
N
"I

 

     

 

BL (Na-10) MLG (N-IO) . MLE (Nu-10)

Subject group

Em Average number of different verbs across the bilingual (BL), the
monolingual German (MLG), and the monolingual English (MLE) groups.
Distribution of individual scores, displayed in Figure 4, again shows a
wide range of performance with no group bias. Four monolingual subjects
performed at or below the lowest performing bilingual subject (number of
different verbs for MLG 3 = 13, for MLE10 = 13, for MLE 2 = 12, for MLE 3 =12,
vs. for BL 1 = 13) and only 1 monolingual subject (number of different verbs for

MLE 1 = 54) had a higher verb score.

61

 

 

 

60 «-
A
50 -~ 0
2:: l ‘ a I +180
; 40 -- - O
a: ' I
:5 O _________._____l ________ O
a -- 0 Mean
'5 ‘5 30 “r o o o
s. >
o A
a -1SD
20
5 ° A
z
o A
10% A I
o r : : I : r r : I .-

 

 

Subject number

Figure 4. Number of different verbs produced by each subject within and across
the bilingual (BL), the monolingual German (MLG), and the monolingual English
(MLE) groups.

Distribution of Bilin ualVocabula ver Two Lan ua es

Promrtion of Translation Equivalents
in Sonceptual Vocabulaqr

Translation equivalents are defined as cross-language synonyms
associated with one referent. The proportion of semantic overlap in the bilingual
vocabulary was calculated in relation to total conceptual vocabulary, which is
depicted in Figure 5. It shows that all bilinguals produced translation
equivalents (M = 43.7%, S_D = 19.81 ), but that none of the children was 100%
bilingual. On average, the number of double representations of concepts was

smaller than the number of single representations, either in English or in

62

German. Individual proportions of translation equivalents in relation to

conceptual vocabulary ranged between 10.7% and 84.2%.

I Translation Equivalents El German El English

 

 

 

100% -
r 90% " >
I
3 80% -
.9
g 70% -
O
> 60% .
'6 50% -
g 40%J
2 o
a :30/0 ~
6
2 20% .
O
“ 10% -

0% -

 

BL7 BL3 8L9 BL2 BL4 8L8 BL5 BL10 8L6 BL1
Blllngual subjects

Figure 5. Proportion of translation equivalents within the conceptual bilingual
vocabulary.

Distribution of Bilingual Lexical Forms
Relative to Monolingual Groups

The bilingual group contained children who were stronger in English as
well as children who were reported to have a larger vocabulary in German, as
displayed in Figure 6.

Chi-square calculations used to determine language dominance revealed
that five bilinguals were reported to produce significantly more lexical forms in
English than in German (for BL2, x2 [d_f 1, fl = 269] = 6.24, g < .05, $2; for BL4,

x2 [911,14 = 255], = 13.65, 9 < .05, _S_Q; for BL6, x2 @ 1, fl = 380] = 25.27, Q <

63

.05, §_[_); for BL9, 12 [(111, H = 389] = 23.2, p < .05, fig; and for BL10, x2 [g1 1, fl
= 337] = 42.02, p < .05, 5Q). Therefore, they were regarded as English

 

 

 

 

 

 

    

 

  

 

 

 

 

 

 

 

 

 

 

 

 

dominant
IGERMAN aENGLISH
300
250
a E =-
2 200 E E
3 E E
53150 - E E '- '-
‘6
E100 "
3
z
50 a
04
BL9 BL6 BL10 BL4 BL2 BL7 BL3 BL5 BLB BL1
E E E E E B B B G G
English dominant Balanced German

dominant
Bilingual subjects

Eigure Q, Number of lexical forms for German and English within the bilingual
(BL) group.

Two bilinguals were reported to produce significantly more lexical forms in
German than English word forms and were regarded as German dominant (for
BL1, 12 [df 1, fl =161]= 40.73, p < .05, st; for 8L8, 12 [d_f 1, _N, = 387] = 4.18, p
< .05, S_Q). For 3 bilinguals, chi-square calculations revealed no significant
differences between the number of word forms produced in the two languages.

These children were regarded as balanced bilinguals (for Bl3, x2 [gt 1, fl = 295]

64

= 1.66, p < .05, ns; for BL5, x2 [1f 1, N = 254] = 1.57, p < .05, ns; for BL7, x2 [1f
1, N = 494] = 1.36, p < .05, ns).

Figure 7, depicting English-only vocabulary of individual bilingual subjects
relative to average monolingual vocabulary, shows that English-only vocabulary
scores of 4 bilingual subjects (BL 6, 7, 9, and 10) were above the average score
of the monolingual English group (number for Englishme words for BL7 = 260;
for 8L6 = 239; for BL9 = 242, and for BL10 = 228 vs. MLE: M = 205.7, §Q =
86.1 ). Of these bilinguals, only 3 were English dominant (BL 6, 9, and 10),

whereas 1 was a balanced bilingual (BL7).

3007

250-
200‘
150«
100-
so.
0« T

Avg. BL7 BL9 8L6 BL10 BL4 BL2 8L3 8L8 BL5 BL1
MLE

Number of Words

 

 

Bilingual Subjects

Figure 7. Average monolingual English (Avg. MLE) vocabulary and individual
English-only vocabularies of bilingual (BL) subjects.

Note. For bilingual subjects, numbers of lexical forms are depicted.

65

Two more English-dominant subjects (BL 2 and 4) and 1 additional
balanced bilingual (BL3) had scores within one standard deviation relative to the
mean of the monolingual English control group (number of English-only words
for BL4 = 157, for BL2 = 155, and for BL3 = 151 ). in summary, English-only
vocabulary scores of 7 bilingual subjects fell within one standard deviation of the
mean.

Figure 8 shows that 1 balanced bilingual subject (BL7) had a higher
German-only score than was reported on average for German monolinguals
(number of German-only words for BL7 = 234 vs. MLG: M = 224, §_Q = 50.6). All
other German-only scores were below one standard deviation relative to the

average vocabulary size for German monolinguals.

300 '

limuuu

Avg. BL7 8L8 BL9 BL3 8L6 BL5 BL1 BL2 BL10 BL4

.. .. re to
o u: c on
o o o o
l l L I

Number o1 Words

at
O

 

 

 

O
L

Bilingual Subjects

Figure 8. Average monolingual German vocabulary (Avg. MLG) and individual
German-only vocabularies of bilingual (BL) subjects.

Note. For bilingual subjects, numbers of lexical forms are depicted.

66

DISCUSSION

This empirical study was in part motivated out of concerns about the lack
of clinical tools available to serve a growing population of language learners in
the United States who acquire more than one language. The task to identify
bilingual speakers at risk is made even more difficult by the widely held
assumption that bilinguals develop language skills more slowly than their
monolingual counterparts.

To ieam more about the early bilingual language acquisition process and
to contribute information necessary to develop clinical standards to better serve
bilingual populations in the future, vocabulary skills of 2-year-old German-
English bilingual toddlers were compared to their respective monolingual peers.

Specific questions addressed were: (a) how German-English bilingual
toddlers compared to their respective monolingual peers in regard to size and
richness of vocabulary at 24 months of age, and (b) how the bilingual vocabulary
was spread over the two languages.

To answer these questions, the productive vocabulary of 10 German-
English bilingual toddlers was compared to that of 10 age-matched monolingual
peers in each language, using a German and an English version of Rescorla’s

(1989) Language Development Survey.

67

The findings of this study have theoretical and practical implications that

are discussed in the following pages.

Theoretical implications

The widely held belief that regards young bilinguals as inherently weak
ieamers has been supported by the deficit hypothesis (Oller et al., 1997). The
rationale of this hypothesis is that the task of having to learn two languages
simultaneously at a young age, as opposed to one, slows down the acquisition
process due to the limited processing capacities of the developing brain.

To compare the educational potential in a bilingual child relative to
monolingual peers, Pearson et al. (1995b) suggested determining for how many
different concepts the bilingual child can produce lexical representations. When
different concepts were pooled across languages in the present study,
vocabulary size was comparable between the bilingual and monolingual groups.
This is consistent with findings reported by Pearson et al.

When the number of lexical forms (total vocabulary) was calculated, the
bilingual group had a somewhat larger vocabulary, but this difference was only
significant between the bilingual and monolingual English groups and not
between the bilingual and monolingual German groups. Pearson et al. (1995b)
found no significant differences, but because they used one control group
containing both Spanish and English monolingual toddlers, their results may not

accurately reflect the differences.

68

No studies exist in the literature to date investigating the composition of
the early bilingual vocabulary. First findings, gained through this study, show
that vocabulary richness, expressed by the number of different verbs, was
comparable for bilinguals and their monolingual counterparts, as was conceptual
vocabulary size. The bilingual vocabulary differed neither in regard to quantity
nor quality.

Post—hoc analysis revealed further similarities between monolingual and
bilingual toddlers. Vocabulary sizes of all bilingual subjects was well above the
50+ word mark, and all parents reported emerging word combinations. This
shows that characteristic milestones identified in the literature to determine
adequate monolingual development were met by all bilinguals, too. This result is
consistent with Patterson’s (1998) finding on 102 Spanish-English toddlers.

These findings have profound implications for views of the general
educational potential of the young bilingual learner. If language skills in a
bilingual child are comparable to those of monolingual peers, and providing
there is a correlation between language and cognitive development, then the
bilingual child should not be at a disadvantage developmentally.

However, for the above evaluation of bilingual vocabulary skills,
performance in both languages was combined. if one wants to know how well a
bilingual child functions in the majority environment, one would want to know
how strong the child is in one or the other language relative to the majority
language (see Figures 7 and 8). This is important with respect to education

because most bilinguals in the United States receive their education in the

69

majority language. Therefore, low skills in the majority language would still put
them at a disadvantage. As discussed earlier, as many as 40% of the bilingual
subjects in this sample were reported to have higher than average scores for
their English-only vocabulary relative to monolingual English controls, and about
70% had English-only scores comparable to monolingual English controls. in
German only, 10% of the scores were above the monolingual German average.
All other German-only scores fell below one standard deviation from the
monolingual German average, but they were still well above the 50+ word mark.
The relatively lower German-only scores might in part be due to the fact that for
only 20% of the bilinguals was German the dominant language.

in summary, the above data speak against the foregone conclusion that
all bilinguals are weak performers, even when only one language is concerned.
A bilingual child may perform as well as, or even above, the monolingual
average in one language only. Therefore, it cannot be assumed that the
bilingual child necessarily communicates less effectively in a monolingual
environment. These findings are again consistent with Pearson et al. (1995b).

Besides relative strength of each language in the bilingual vocabulary,
this researcher examined how the bilingual vocabulary was distributed across
languages in order to test another hypothesis, the so-called unitary language
system hypothesis, which, similar to the deficit hypothesis, assumes inferior
communication capacities in the young bilingual, namely the lack of voluntary
language separation. On the basis of this hypothesis, no or only very little

semantic overlap in the form of translation equivalents should be expected.

70

However, the data gathered in this study show that bilinguals do not avoid
translation equivalents. For 43.8% or almost half of the bilingual conceptual
vocabulary, their double lexical representations were evident. Pearson et al.’s
(1995a) calculation method yields an even higher proportion of 63.9%. This
finding demonstrates a definite potential for language choice, even though it
cannot predict how the child uses words in communicative contexts. For their
Spanish-English sample, Pearson et al. (1995) reported somewhat lower
proportions of translation equivalents of around 30.8%, which might in part be
due to the fact that Spanish and English are less similar languages than German
and English.

On the other hand, bilinguals in this study showed a wide range of
translation equivalents in their vocabularies, ranging from little (10.7%) to almost
complete semantic overlap (84.2%) of the languages. This variety is consistent
with findings reported by Pearson et al. (1995a). Therefore, other influences,
e.g., environmental factors, such as language models or time of exposure, might
affect the child’s ability to associate two labels with one referent. in this study,
some parents were observed to practically train translation equivalents by asking
for names of objects in both languages.

Pearson et al. (19953) suggested size of vocabulary as another possible
influence. Children with vocabulary sizes of more than 500 words in their
sample were reported to produce almost twice as many translation equivalents
(59% vs. 30.8%) than children with smaller vocabularies. in this sample, the

child with the largest total vocabulary of 488 lexical forms was reported to

71

produce the largest proportion of translation equivalents (84.2%), which is
consistent with the earlier finding. However, some children with lower
vocabularies also produced high amounts of translation equivalents; e.g., one
child whose vocabulary contained only 288 lexical forms also contained 60.5%
translation equivalents. Logically, the amount of potential semantic overlap is in
part a function of vocabulary size of the weak language; in other words, the
number of translation equivalents cannot exceed the number of words in the
weak language. Further research is necessary to investigate how children learn

to associate two labels with one referent.

Clinical Implications and Contributions

These results have several implications for clinical praxis. if bilingual and
monolingual children have comparable vocabulary skills, then bilinguals do not
need to be regarded as an exceptional population for which separate evaluation
standards and tools have to be developed. Tools that are already developed to
assess vocabulary skills in monolinguals may be used for bilingual populations
as well, as long as vocabulary in both languages is assessed.

To assess vocabulary skills in young learners, vocabulary checklists have
been favored by clinicians. Some foreign language adaptations of existing
vocabulary lists have already been developed, such as a Spanish version of the
MacArthur Communicative Development inventories (CDl) (Jackson-Maldonado
& Bates, 1988) or a Spanish-English version of the LDS (Patterson, 1998). This

study contributes a German version of Rescorla’s (1991b) LDS that, together

72

with her original English version, can be used as a double language measure
and thus increases the number of bilingual populations that can be addressed by
speech-language pathologists. All of the above tools are based on parental
report, which is especially useful for clinical practice because it allows the
clinician to assess vocabulary skills in languages other than English without
being required to know these languages.

The use of double-language measures to assess bilingual populations is
necessary because only a portion of the bilingual vocabulary is represented in
each of the two languages. A large part of the vocabulary is represented in
either one or the other language (see Figure 5). Therefore, any measure that

assesses only one language will tap only part of the potential of a bilingual child.

Future Research

One reason that studies on simultaneous language acquisition are rare
might be the complex methodology that is required to do research in this area.
Besides the lack of instruments available to describe skills in languages other
than English, there is a lack of agreement on many of the methodological issues.
One problem is that any bilingual population of speakers is naturally
heterogeneous. For example, this sample includes balanced bilinguals and
bilinguals who demonstrate dominance in one language. A question that arises
is: Can these subjects be compared with each other, or should subgroups be
formed? It has been discussed earlier, for example, that the proportion of

translation equivalents in the bilingual vocabulary is in part a function of

73

vocabulary size of the weak language. This means that, theoretically, balanced
bilinguals may be at an advantage compared to bilingual speakers with distinct
dominance in one language, which, in turn, creates the danger of influencing
group means in different directions. This is further complicated by the fact that
the bilingual language status may be dynamic and language dominance might
change over time.

Another problem is the issue of control groups. Should bilingual
performance be related to monolingual counterparts in each language, or is one
language sufficient? is it even legitimate to mix monolinguals from each
respective language in one control group, as has been done by Pearson et al.
(1995a)? Because it is not known yet whether different languages are easier or
more difficult to learn, it may be possible that bilinguals compare differently to
different monolingual control groups. Standards are necessary in order to
compare results gained in different studies.

Even more fundamental is the possibility of an implicit disadvantage for
the monolingual participants observed in this comparative study of vocabulary
skills. Whereas monolingual parents were asked to complete only one
vocabulary checklist, parents of bilingual children were asked to complete two,
which theoretically doubles the number of opportunities to identify words, even
though they are semantically equivalent. Because one group performs a task
twice as opposed to once, differential test effects have to be considered.
Because no standards exist about comparability of monolingual and bilingual

populations, it is possible that an implicit disadvantage of the monolingual group

74

might have skewed the results of the present study. However, rigorous reliability
testing was performed in this study to control for these effects.

Besides clarifying methodological issues, further research needs to
explore different populations of learners. This study focused on vocabulary
skills of normally developing German-English-speaking toddlers. No data are
available yet on young bilinguals whose skills fall below or around characteristic
milestones. Do they meet the challenge of handling two language systems
simultaneously as do normally developing bilinguals, or are they especially
disadvantaged due to the demands of processing a higher information load in
the face of limited information-processing capacities? Further research is
necessary to explore this issue.

Another open question is whether similarities or differences in the target
languages facilitate or hamper the language-leaming process. Evidence from
this study and from literature suggests the possibility that similarities between
languages might facilitate the learning process and/or retention and recall of
translation equivalents. The proportion of translation equivalents in the German-
English vocabulary was higher than measured in the Spanish-English
vocabulary of same-age bilinguals. More research, especially from populations
including speakers of unrelated languages, is necessary to increase our
knowledge about the nature of bilingual language learning.

Finally, there is the issue that vocabulary lists, although clinically useful,
do not consider language context. To understand better the process of how two

languages are learned simultaneously at an early age, we need to know more

75

about possible influences affecting this learning process. Environmental
influences need to be investigated, such as overall exposure patterns in the
ambient languages during everyday-life contexts, attitudes toward different
languages, different types of language models, and so on. Just to name an
example, some parents in this study were observed to overtly train language
awareness by asking the child for “the word in the other language," which directs
the child’s attention to looking for synonymy between languages at a very early
age. This again leads to the fundamental issue in the literature of whether
cognitive strategies or environmental factors exert more influence on the
learning process. Research into this question might shed more light on the

nature of language learning in general.

76

SUMMARY AND CONCLUSIONS

There are many open questions in the field of early bilingual acquisition
that need further research. One question this empirical study tried to answer
was how vocabulary skills of German-English bilingual toddlers compare to their
respective monolingual counterparts at one point in time. It was found that
bilingual subjects had comparable skills in regard to vocabulary size and verb
diversity, which was regarded as an indicator of vocabulary richness. Because
no signs of inferiority could be found either in quantity or quality of vocabulary, it
is concluded that the bilingual vocabulary, taken as a whole, does not differ from
the monolingual.

Considering the similarities found between the bilingual and monolingual
vocabularies and the fact that all bilinguals met characteristic milestones of
language development at the same rate as their monolingual counterparts, as
indicated by a productive vocabulary of at least 50 different words and emerging
word combinations by age 2, the bilingual deficit hypothesis, which labels young
bilinguals as deficient learners, cannot be upheld. Whatever information load
has to be processed, the bilingual child appears to be able to handle it.

The other question that was raised was whether the bilingual child would

be capable of representing concepts in more than one language at an early

77

stage of the acquisition process. To answer this question, the proportion of
translation equivalents in the child’s vocabulary was assessed, which was found
to amount to almost half of the child’s conceptual vocabulary. Given the
simultaneous representation of concepts in two languages, it is inferred that
young bilinguals are capable of building two separate lexicons, even as early as
2 years of age. This finding does not justify the single lexicon theory.

In summary, findings from this study contribute to the growing body of
literature that does not regard early simultaneous acquisition of more than one
language as an inherent disadvantage for the child. This has consequences for
clinical practice. if vocabulary skills are essentially similar regardless of whether
one or two languages are learned, at least when they are as similar as German
and English or Spanish and English, than existing assessment tools can be
exploited. Clinicians have favored vocabulary checklists for which double
language measures can be easily developed. Rescorla’s (1989) LDS appears to
be especially promising to adapt for bilingual populations because it is an
instrument that contains very little inherent bias. A German-English version of
the LDS was used successfully in this study; the usefulness of this instrument for

studying other languages should be investigated in the future.

78

APPENDICES

79

APPENDIX A

TABLES—RAW DATA

80

Table 1: Subject pool: Overview of age, gender, and birth order.

 

 

Bilingual Monolingual Monolingual
Subjects German Subjects English Subjects
BL1 MLG1 MLE1
27 months, male, 3rd born 24 months, male, tst born 24 months, male, 151 born
BL2 MLG2 MLE2
25 months, male, 1st born 24 months, male, 1st born 24 months, male, 1st born
BL3 MLG3 MLE3
24 months, male, 1st born 24 months, male, 1st born 24 months, male, 2nd bom
BL4 MLG4 MLE4
24 months, male, 1st born 26 months, male, 1st born 25 months, male, tst born
BL5 MLGS MLE5
24 months, male, 1st born 27 months, male, 3nd born 27 months, male, 1st born
BL6 MLGG MLE6
24 months, female, 3rd born 24 months, female, 2nd born 24 months, female, 151 bom
BL7 MLG7 MLE7
25 months, female, 1st born 24 months, female, 1st born 25 months, female, 3rd bom
BL8 MLGB MLEB
28 months, female, 2nd born 24 months, female, 1st born 27 months, female, 7th born
BL9 MLGO MLE9
24 months, female, 1st born 25 months, female, 151 born 24 months, female, 1st born
BL10 MLG1O MLE10

24 months, female, 1st born

27 months, female, 2nd born

81

24 months, female, 1st born

Table 2: Productive vocabulary: Bilingual raw data.

 

subject No. of German No. of English phonet. distinct vocab.: vocab.:

 

words words Translation total no. total no.
Equivalents of of different
concepts words

BL1 103G+18TE 22E+18TE 15(10.7%) 143 158
(27 months) (121) (40)

BL2 23 6+ 91 TE 64 E + 91 TE 83 (46.6%) 178 261
25 months) (114) (155)

BL3 31 G+113TE 38E+113TE 106 (60.5%) 182 288
(24 months) (144) (151)

BL4 17G +81 TE 76E + 81 TE 74 (44.3%) 174 249
(24 months) (98) (157)

BL5 66 + 71 TE 46 + 71 TE 61 (35.2%) 183 244
(24 months) (137) (1 17)

BL6 47 G + 94 TE 145 E + 94 TE 86 (30.9%) 286 372
(24 months) (141 ) (239)

BL7 7 G + 227 TE 32 E + 227 TE 220 (84.2%) 268 488
(25 months) (234) (260)

BL8 77 G + 78 TE 43 E + 78 TE 71 (38.2%) 193 269
(26 months) (155) (121)

BL9 8 G +139 TE 103 E + 139 TE 132 (54.3%) 250 381
(24 months) (147) (242)

BL10 21G + 88 TE ( 140E + 88 TE 80 (32.2%) 249 329
(24 months) 109) (228)

 

M: 43.71% M = 210.6 M = 303.9
SD: 19.81 SD = 148.2 SD = $92.1

 

Note. G = German only
E = English only

TE = Translation Equivalent

82

Table 3: Monolingual German raw data.

 

 

 

Subject Vocabulary Vocabulary
German English
MLG1 238
(24 months)
MLGZ 252
(24 months)
MLG3 117
(24 months)
MLG4 246
(26 months)
MLGS 198
(27 months)
MLG6 203
(24 months)
MLG7 207
(24 months)
MLG8 308
(24 months)
MLGQ 209
(25 months)
MLG10 262
(27 months)
M = 224
SD = 150.6

 

83

Table 4: Monolingual English raw data.

 

 

 

Subject Vocabulary Vocabulary
German English
MLE1 363
(24 months) A
MLE2 133
(24 months)
MLE3 124
(24 months)
MLE4 269
(25 months) (*64tidnanes)
MLE5 281
(27 months)
MLE6 253
(24 months)
MLE7 235
(25 months)
MLE8 149
(27 months)
MLE9 148
(24 months)
MLE10 102
(24 months)
M = 205.7
SD = :86.1

 

Table 5: Number of different verbs across the bilingual and monolingual

 

 

 

subject groups.
No. of Bilingual German English
Subject Monolingual Monolingual
1 13 38 54
2 29 41 12
3 19 13 12
4 29 45 44
5 29 33 45
6 40 36 36
7 52 34 34
8 31 45 18
9 46 44 23
10 34 43 13
mean 32.21 37.2 29.1
SD 11.65 9.61 15.53

 

85

APPENDIX 8

ITEMS ON RESCORLA’S LANGUAGE DEVELOPMENT SURVEY:
FREQUENCY OF OCCURRENCE IN CHILDREN’S VOCABULARIES
AND PHONETIC TRANSCRIPTIONS

Table 6: Frequency of occurrence, rank order, and number of syllables for the
sample of 50 randomly selected LDS items (English version).

 

 

No. LDS item Frequency Rank Number of
Syllables
1 coffee 280 1518 2
2 cookie 33 7899 2
3 juice 170 3049 1
4 orange 221 2058 2
5 soda 1 12 3689 2
6 tea 223 2291 1
7 blocks 260 1694 1
8 house 2705 195 1
9 cow 263 1705 1
10 dog 1380 387 1
1 1 horse 1263 431 1
12 bottom 858 584 2
13 ear 331 1522 1
14 eye 707 720 1
15 foot 849 571 1
16 knee 93 3987 1
17 teeth 538 943 1
18 store 681 737 1
19 catch 679 743 1
20 clap 89 9520 1
21 dance 608 1182 1
22 feed 372 1259
23 have 22337 25 1
24 knock 85 4241 1
25 make 8333 71 1
26 run 1473 333 1
27 bottle 346 1442 2
28 chair 421 1139 1
29 clock 330 1474 1
30 cup 364 1684 1
31 towel 48 5961 1
32 money 1694 301 2
33 pencil 331 1636 1
34 aunt 109 3762 1

87

Table 6: Continued.

 

 

No. Q3 Item Frequency Rank Number of
Syllables

35 Grandpa 71 5908 2

36 belt 174 2726 1

37 boots 176 2451 1

38 bus 345 1375 1

39 train 556 905 1

40 boy 2529 213 1

41 dirty 143 2725 2

42 mine 378 1318 1

43 open (adj) 1416 328 2

44 wet 418 1253 1

45 white 2085 252 1

46 away 3814 145 2

47 hi 85 4704 1

48 there 15194 37 1

49 where 561 1 96 1

50 yes 1317 399 1

 

Note. Word frequency and rank order based on a corpus of 86,741 types and
5,088,721 tokens derived from published written materials used in US. school
from grades 3 through 9.

This corpus was published by J. B. Carroll, P. Davies, & B. Richman. (1971).
The American Heritage Word Freguency Book. New York: Houghton Mifflin.

88

Table 7: Frequency of occurrence, rank order, and number of syllables for the

sample of 50 randomly selected LDS items (German version).

 

 

No. Item Frequency Rank by Grade Level Syllables
1 Kaffee (coffee) 38 224 (first grade) 2
2 Keks (cookie) 5 272 (first grade) 1
3 Saft (juice) 3 623 (first grade) 1

here: Apfelsaft

4 Apfelsine (orange) 4 726 (second grade) 4

5 Brause (soda) 6 506 (second grade) 2

6 Tee (tea) 7 669 (second grade) 1

7 (Bau)klotz (blocks) 3 488 (first grade) 1

8 Haus (house) 477 4 (first grade) 1

9 Kuh (cow) 31 196 (first grade) 1
10 Hund (dog) 362 3 (first grade) 1
1 1 Pferd (horse) 78 90 (first grade) 1
12 Po(po) (bottom) 6 998 (first grade) 1
13 Ohr (ear) 125 25 (first grade) 1
14 Auge (eye) 399 1 1 (first grade) 2
15 Full (foot) 116 44 (first grade) 1
16 Knie (knee) 100 112 (first grade) 1
17 Zéhne (teeth) 68 102 (first grade) 2
18 Laden (store) 10 455 (first grade) 2
19 fang(en) (catch) 114 55 (first grade) 2
20 Klatsch(en) (clap) 5 594 (first grade) 2
21 tanz(en) (dance) 18 403 (first grade) 2
22 fatter(n) (feed) 83 42 ( first grade) 2
23 hab(en) (have) 3633 2 (first grade) 2
24 kiopf(en) (knock) 12 746 (second grade) 2
25 mach(en) (make) 631 10 (first grade) 2
26 renn(en) (run) 55 17 (first grade) 2
27 Flasche (bottle) 27 334 (first grade) 2
28 Stuhl (chair) 52 124 (first grade) 1
29 Uhr (clock) 77 46 (first grade) 1
30 Tasse (cup) 23 206 (first grade) 2
31 (Hand)tuch (towel) 26 246 (first grade) 1
32 Geld (money) 28 223 (first grade) 1
33 (Blei)stift (pencil) 3 1062 (second grade) 1
34 Tante (aunt) 139 75 (first grade) 2

89

Table 7: Continued.

 

 

No. Item Frequency Rank by Grade Level Syllables
35 Opa (grandpa) 94 114 (first grade) 2
36 Stiefel (boots) 1 1 303 (first grade) 2
37 Junge (boy) 225 16 (first grade) 2
38 Githel (belt) 3 711 (third grade) 2
39 Bus (bus) 41 134 (first grade) 1
40 Zug (train) 79 186 (first grade) 1
41 schmutzigldreckig (dirty) 12 93 (first grade) 2
43 auf (open) 31 57 (first grade) 2

hier: offen
44 nals (wet) 56 60 (first grade) 1
45 weiB (white) 448 9 (first grade) 1

46 weg (away) —
Frequency was not assessed for:

42 mein(s) (mine)

46 weg (away)

47 Hallo (hi, hello)

48 da, dort (there)

49 wo (where)

50 ja (yes)

 

 

Note. Word frequency, rank order and number of syllables of the sample of 50
randomly selected l_.l_)§ items is based on a corpus of 260,374 words (tokens) of
which 12.5% are nouns (3,772 types, 32,600 tokens), 16% are verbs (2,345
types, 41,889 tokens), and 5.9% are adjectives (804 types, 15,437 tokens).
Whereas word frequency was calculated for the total corpus, the words were rank
ordered for each grade level separately.

 

The corpus was published by D. Pregel 8. G. Rickheit. (1987). Der Wortschatz im

Qrundschulalter. Haufigkeitsworterbuch zum verbalen, substantivischen und
adjektiviichen Wortgebrauch. Hildesheim: Olms.

9O

Phonetic Transcription of English and German Items on
Rescorla's (1989) Language Development Survey

(based on guidelines of the lntemational Phonetic Association)

Food

apple ['88 pl]
banana [be'n as n e ]
bread ['b red ]
butter ['b At 3“]
cake ['k eI k]

candy ['kaandI]
cereal ['SIrIal]
cheese ['t I i: 2]
coffee ['kofI ]
cookie ['kUkI]
crackers ['krtek a“ z ]
drink ['drIIJk]

egg ['eg]

food ['fuzd]

91

Lebensnfiflel

Apfel ['a pf eI]
Banane [ba'nazne]
Brot ['b Rozt ]
Butter ['buteR ]
Kuchen ['kuzx n]
Bonbon ['bonbon]
Miisli ['m yzslI]
Kase ['kezze ]
Kaffee ['kafe ] [ 'kafez]
Keks ['ke:ks]
Zwieback ['tsvizbak]
Trinken ['tRIIJkn]
Ei ['ae]

Essen ['esn]

grapes ['greIps]

gum I 'g A m]
Hamburger ['haemb 39 at]
Hot Dog ['hcxdog]

ice cream ['aIs 'krizm]
juice ['d3u:s]

meat ['mizt]

milk ['mIlk] '

orange ['orInd3]
pizza ['piztsa]

pretzel ['prstsl]

soda ['sode]

soup ['suzp]

spaghetti [spe'get I]
tea ['ti:]

toast ['tost]

water ['wozt 3“]

92

(Wein)trauben [(vaIn) 'tRaobn]
Kaugummi [kao'gUmI]
Hamburger ['hamburge R]

Hot Dog ['hodok]

Eis ['aes]

Saft ['zaft]
Fleisch [‘f la e I ]

Milch ['mIlg]
Apfelsine [apfel 'zi:ne]
Pizza ['pItsa]

Bretzel ['bReztsl]
Brause ['bRaoze]
Suppe ['zupe]
Spaghetti [ I pa'get I]

Tee ['te:]
Toast ['tost]

Wasser ['vaseR]

Toys

ball ['bol]
balloon [ bel'uzn]
blocks ['blaks]
book ['bUk]
bubble ['b A b1]
crayons ['kreens]
doll ['dal] ['dol]
present ['preznt]
slide ['slaId]
swing ['SWIIJ]

teddy bear ['tedI bar]

Outdoors

flower ['f laU 9“]

house ['haus]

moon ['muzn]

rain ['reIn] ['ren]

93

S pielzeu g

Ball [' bal]

(Luft)ballon [(‘lU f t)ba'1013]
(Bau)klotz [(bao)'klots]
Buch ['buzx]

(Seifen)blase [(zaefn)'bla:ze]
(Bunt)stifte [(bunt) ' J' tIfte]
Puppe ['pUpe]

Geschenk [ge' jenk]
Rutsche ['Ru tje]
Schaukel [' ] aokl]

Teddy(bar) ['tedI (be:R)]

drauBen

Blume ['bluzme]

Haus [' haos]

Mond ['moznt]

Regen ['Rezgn]

sidewalk ['saIdwok]

snow [ 'sno]

star ['star]
street ['strizt]
sun [' s A n]

tree ['tri:]

AmmMS

bear ['bsr‘]

bee ['bi:]

bird [' b 3 d ]
bug I' b A g]
bunny [' b A nI]
cat ['kaet]

chicken [' if IkIn]
cow ['kaU]

dog ['dog] ['dag]

duck ['d A k]

94

Gehweg ['gezvezk]
Schnee [ ' _l' nez]

Stern [' I IS Rn]
StraBe [' ItRazse]
Sonne ['zone]

Baum [' baom]

Tiere

Bar ['beza]

Biene ['b i:ne]

Vogel [' f ozgel]

Miicke/ Fliege [‘myke] ['flizge]
Hase ['hazze]

Katze ['katse]

Huhn ['huzn]

Kuh ['ku:]

Hund ['hunt]

Ente [' ante]

elephant ['Slefent]
fish ['fI ] ]

frog ['frag] ['frog]
horse ['hors]
monkey [' m A nkI]
pig ['pIgl

PUPPY ['p A le
snake ['sneIk]

tiger ['taIg 3‘]
turkey ['t 3‘kI]

turtle ['t at 1]

Body Parts
arm ['arm]

belly ['belI ]
bottom ['batem]
chin ['tjxn] .

ear ['Ir]

95

Elefant [ele ’ f ant ]
Fish ['fI j ]

Frosch ['fRo ] ]

Pferd ['(p)fe:Rt]

Affe [‘ afe]

Schwein [' j vaen]
Hiindchen ['hyntg In]
Schlange [' I lane]
Tiger ['tizgSR]
(Trut)hahn [(tRUt) hazn]

Schildkrote [' j Ilthete]

Korperteile
Arm ['aRm]

Bauch ['baox]
Po(po) ['po(po)]
Kinn ['kIn]

Ohr ['ozR]

elbow ['elbo]
eye ['aI]

face ['feIs]
finger ['fIng 8“]
foot ['fUt]

hair ['her]
hand ['hmnd]
knee ['niz]

leg l'leg}
mouth ['maUG]
neck ['nek]
nose ['noz]
teeth ['ti:6]
thumb ['6 A m]

toe ['to]

tummy ['t A mI]

Ellenbogen ['slenbogn]
Auge ['aoge]
Gesicht [ge 'ZIgt]
Finger ['fIgaR]
FuB ['fuzs]

Haar ['hazR]

Hand ['hant]

Knie ['kniz]

Bein ['baen]

Mund ['munt]
Hals ['hals]

Nase ['nazze]
Zéihne ['tsezne]
Daumen ['daomen]

Zeh ['tse:]

Magen ['magn]

Places

church ['tI 3‘tI]

home ['hom]

hospital ['haspItl]

library ['laIbrerI]
McDonalds [mek'danelds]
park l'parkl

school ['skuzl]

store ['stor]

zoo ['zu:]

Actions

bath ['bme]
breakfast ['brekfest]
bring ['brIIJ]

catch ['kae t j]

clap ['klEp]

close ['kloz]

97

Orte

Kirche ['kIRge]

zu Hause/daheim [tsuz 'haoze]
Krankenhaus ['kRathIJ'haoz]
Biicherei [byge'Rae]
McDonalds [msk‘do-ne 11'. s ]
Park ['pazak]

Schule [' I uzle]

Laden ['lazdn]
Zoo [tso:]

Tatigkeiten
bad(en) ['ba:d(n)]
friihstiick(en) ['fRy j tyk(n)]
bring(en) ['bRIIJ(n)]
fang(en) ['fa13(n)]

klatsch(en) ['kla tI(n)]

zumach(en) ['tsumax(n)]

come ['k A m]
cough ['kof]

cut ['k A t]

dance ['daens]
dinner ['dIn 3‘]

doodoo ['duzduz]

eat ['izt]

feed ['fizd]
finish ['fInI I]
fix ['fIks]

get ['gct]

give ['gIv]

go ['go]

have ['hav]
help ['help]

hit ['hIt]

hug ['h A g]
jump Ms A mp]
kick ['ka]

98

komm(en) ['kom(n)]
hust(en) ['hUst(n)]

schneid(en) [' I naed(n)]]
tanz(en) [‘tants(n)]

Abendbrot ['a:bentbRo:t]

(k)a-(k)a ['(k)a (k)a]

esse(n) ['es(n)l
fiitter(n) ['fyt(n)8R]
fertig ['fthIc]
heil(e) ['haele]
kfieg(cn) ['kRi=g(n)]
geb(en) ['ge:b(n)]

geh(en) ['ge=(n)l
hab(en) ['ha:b(n)]

helf(en) ['h81f(n)]

' hau(en)/schlag(en) ['hao(n)]

[' l la=g(n)]
schmus(en) [' I mu:z(n)]
hfipfien) ['hypf(n)]

schieB(en)/stoB(en) [' I i:s(n)]

kiss ['kIs]
knock [‘nak]

look ['lUk]

love ['1 A v]

lunch ['1 A n tI]
make ['meIk]

nap ['nwp]
outside ['aUt'saId]

pattycake ['pedI k eI k]

peekaboo ['pIke bu:]]
peepee ['pizpiz]
push ['pU I]

read ['rizd]

'ride ['raId]
run ['run]

see ['si:]

show [' I o]

99

ktiss(en) ['kys(n)]
klopf(en) ['kIOpf(n)]
guck(en)/schau( en) ['gUk(n)]

[' I ao(n)]
lieb(en) ['li:b(n)l]

Mittag ['mItazk]
mach(en) ['max(n)]
Mittagsschlaf [mItaks' I lazfl
drauBen ['draosen]
Backe-Backe-Kuchen
['bake bake 'ku:x(e)n]
Guck-Guck ['gukuk]
pischerm) ['pit l 8R(n)]
schieb(en) [' l i:b(n)]
les(en) ['1e:z(n)]
fahr(en) ['fa:R(n)]
renn(en) ['Ren(n)]

seh(en) ['ze:(n)]

zeig(en) ['tsaeg(n)]

sing ['SIIJ]
sit ['SIt]

sleep ['slizp]

stop ['stap]
take ['teIk]
throw ['Gro]
tickle ['tIkl]
walk ['wok]
want ['want]

wash ['wa I ]

Household

bathtub ['baet A b]
bed ['bed]

blanket [blankIt]
bottle ['botl]

‘bowl ['bol]

100

sing(en) ['SIrJ(n)]

sitz(en) ['ZIts(n)]
schlaf(en) [' I la:f(n)]

halt(en) ['halt(n)]

nehm(en) ['ne:m(n)]
werf(en) [IveRf(n)]

kitzel(n) ['kItsel(n)]
lauf(en)/spazier(en) ['laof(n)]
wollen/will ['vol(n)] ['vrl]

wasch(en) ['va I (n)]

Haushalt

(Bade)wanne [(ba:de)'vane]
Bett ['ber]

Decke ['dcka]

Flasche ['fla j a]

Schiissel [' I ysel]

chair ['tIer]
clock ['klak]
crib ['krIb]
CUP ['k A P]
door ['dor]
floor ['flor]
fork ['fork]
glass ['glaes]
knife ['naIf]
light ['laIt]
mirror ['mIr 9‘]
pillow ['leo]
plate ['pleIt]
potty l'padIl
radio ['redIo]
room ['ruzm]
sink ['sxnk]
soap ['sop]

sofa ['sofe]

101

Stuhl [' I tu:l]

Uhr ['uR]

Wiege ['vizge]

Tasse ['tase]

Tiir ['tyR]
(FuB)boden [(fu:s)'bod(e)n]
Gabel ['ga:b(e)l]
Glas ['glazs]

Messer ['meseR]
Lampe ['lampe]
Spiegel [' I pi:gl]
Spiegel [' I pi: g1]
Teller ['tsleR]
pofchen ['tepf can]
Radio ['RadIo]
Zimmer ['tSImcR]
Spfile l' I pylel

Seife ['zaefe]

Sofa ['zofa]

spoon ['spuzn]

stairs ['sterz]

table ['terl]
telephone ['te lefon]
towel ['taUl]

trash ['tras I ]

TV [ti:'vi:]

window ['WIndo] ['WInde]

Personal

brush ['br A I]

comb ['kom]

glasses [’glzeSIz]
key ['ki:]
money ['m A M]
paper ['pep 9“]
pen l'penl

pencil ['pensl]

102

Loffel ['lefel]

Treppe ['trepe]

Tisch ['tI I]

Telefon ['telefon]
(Hand)tuch [(han)'tUx]
Miill ['myl]

Fernseher ['fCRn zezeR]

Fenster ['fensteR]

, Persdnliches

Biirste ['byRste]

Kamm ['kam]

Brille ['bRIle]
Schliissel [' I lysel]
Geld ['gelt]

Papier [pa'pizR]
Kuli ['kUlI] ‘

(Blei)stift [(blae)’ I tIft]

penny ['penI]
pocketbook ['pakIthk]
tissue ['tI I U]
toothbrush ['m 6 'br A I ]
umbrella [ A m'brele]

watch ['wa tI]

People

aunt ['zent]

baby ['berI]

boy ['bOI]

daddy ['dadI]
doctor ['dakt er]
girl ['g 31]
grandma ['grama]
grandpa ['grzempa]
lady ['leIdI]

man ['man]

103

Pfennig ['(p)fenIg]
Notizbuch [no'tiztstx]
Tempo ['tempo]
Zahnbiirste ['tsa:n byRste]
Schirrn [' I IRm]

Armband(uhr) ['(aRmbant)u:R]

Personen

Tante ['tante]

Baby ['besz]
Junge ['june]

Papa [‘papa]
Doktor ['dOktOR]
Madchen ['metgen]
Oma ['ozma]

Opa ['ozpa]

Frau ['frao]

Mann ['man]

mommy ['mamI]

own name

pet name

uncle [' A IJkl]

Ernie, etc.

Clothes

belt ['bslt]

boots ['buzts]
coat ['kot]

diaper ['daIpe]

dress ['dres]

gloves ['gl A vz]

hat ['hat]

jacket ['d3aekIt]
mittens ['mItns]
pajamas [pe'ogzemez]

pants ['pmnts]

104

Mama ['mama]

eigener Name

Name e. Haustiers

Onkel ['ogkel]

Peter, etc.

Kleidung

Giirtel [‘gthl]
Stiefel [' I tizfl]
Mantel ['mantl]

Windel ['VIndl]

Kleid ['klaIt]

Handschuhe ['han tIu: e]
Hut ['hUt]

Jacke ['jake]

Faustlinge ['foastlIne]
Schlafanzug [' I lafantsuk]

Hose ['hoze]

shirt [' I 3t]
shoes [' I uzz]
slippers ['sle 92]
sneakers ['sni:k a‘z]
socks ['saks]

sweater ['svet 9“]

Vehicles

bike ['baIk]
boat ['bot]
bus ['b A 5]

car ['kar]

motorbike ['mot a‘baIk]
plane ['pleIn]

stroller ['strol 9‘]

train ['treIn]

trolley ['tralI]

truck ['tr A k]

105

Hemd ['hamt]

Schuhe [' I uze]
Hausschuhe ['haos I uze]
Turnschuhe ['tURn I we]
Socken ['zokn]

Pulli ['pUlI]

’ Fahrzeuge

(Fahr)rad [(fa)'Rat]
Schiff [' I If]
Bus ['bUs]

Auto ['aoto]

Motorrad ['motoaat]
Flugzeug ['flUktsoIk]
Kinderwagen ['kIndeRvazgn]
Zug ['tsuzk] ['tsux]
StraBenbahn [' I tRasenbazn]

Laster ['lasteR]

Modifyers

Bestimmungen

(d. Ortes, Art u. Weise, etc.)

all gone ['0] gen]
all right ['01 ran]
bad [bad]

big ['bIgl

black ['blmk]

blue ['blu:]

broken ['broken]
clean ['klizn]

cold ['kold]
dark ['dark]

dirty ['d 3‘tI]
down ['daUn]
good ['guzdl
happy l'haele
heavy ['hSVI]
hot ['hCIt]

hungry ['h A gan]

106

alle (balle) ['ale ('bale)]
(alles) klar [(‘ales) klaR]
schlecht [' I lext]

groB [gRozs]

schwarz [' I vaRts]

blau ['blao]

kaputt ['kapUt]
sauber ['zaobeR]

kalt ['kalt]
dunkel ['dUIJkel]

schmutzig [' I mUtSIk]

unten ['Unten]

gut ['guzt]

gliicklich ['glyklIc]
schwer [' I vezR]
heiB ['haes]

hungrig ['hungRIg]

little ['lIt l]

mine ['maIn]

more ['mor] ['mor]
Open ['Open] ['OPm]
pretty ['prItI]

red ['red]

shut [' I At]

stinky ['stInkI]
that [' 6 at]

this [' 0 Is]

tired ['taIrd]

up ['Ap]

wet ['wet]
white ['hwaIt]
yellow ['jelo] ['jele]

yucky ['j A RI]

107

klein ['klaen]
mein ['maen]
mehr ['mezR]
auf ['aof]
hiibsch ['hyb I ]
rot ['Rozt]

zu ['tsuz]

eklig ['ezkllq]

das ['das]

. dies [‘dizs]

miide ['myde]
hoch ['hox]
naB ['nas]
weiB ['vaes]
gelb l'galp]

igitt [i:‘gIt]

Others

A, B. C, etc.

away [e'we]

booboo ['bu: buz]

byebye [baI'baI]

cursewords

here ['hIr]

hi, hello ['haI] [he'lo]
in ['In]

me ['mi:]

meow ['mIao]

my ['maI]

myself [maI'self]
night-night ['naIt'naIt]

no ['no]

off ['of]
on ['on] ['an]
out ['aUt]

please ['pli:z]

108

Verschiedenes

A, B, C, etc.

weg ['vek]

buh ['bu:]

tschiiB/ade [' t I ys] [a'dez]
Schimpfworte
MaFMR]

hallo [ha'lo]

in ['In]

mir/mich ['mIR] ['mIc]
miau ['mIao]

meins ['maens]

ich/ selber ['Ig] ['zelbe(R)]
Nachti ['naxtiz]

nein ['naen]

runter lab ['RUnte(R)] ['ap]
an/ auf ['an] ['aof]
(r)aus ['(R)aos]

bitte ['bIte]

Sesame St. ['scsemI 'strizt]
scuse me ['sijs 'mi:]
shutup ['I At Ap]
thank you [' 9 2813K jU]
there [' 6 8r]

under [' A nd 9‘]
welcome ['welkem]
what ['hwat]

where ['hwer]

why ['hwaI]
woofwoof ['wufwuf]
yestfiSSI

you ['juz}
yumyum ['j A m'j A m]

1,2,3,etc. ['w A n] ['tuz] [' 8 Riz]

109

Sesam StraBe ['ze:zam ' I tRazse]
Tschuldigung [' tI UldI gunk]
Ruhe! ['Ruze]

danke ['danke]

da/ dort ['da] ['dORt]

unter ['Unte(R)]

iyas[wcs]
wu>Fvo]

warum [va' Rum]
wauwau ['vao'vao]
jalfial

(thdufl
rnnaluun]

1,2,3, etc. ['aens] ['tsvae] ['dRae]

References

De Boor, H., Moser, H., & Winkler, C. (1969). Siebs: Deutsche
Aussprache. Reine und gemaBigte Hochlautung mit
Aussprachewdrterbuch, Berlin: Walter de Gruyter & Co

Kenyon, J. S., & Knott, T. A. (1953). A pronouncing dictionary of
American English, Springfield, Mass: G. & C. Merriam Company

Wéingler, H.-H. (1963). Rangwdrterbuch deutscher Umgangssprache,

Marburg, Germany: N. G. Elwert Verlag

110

APPENDIX C

INSTRUMENTS

111

The Language Development Survey

Please circle each word your child says spontaneously. Do not include words that your child
imitates. Furthermore do not circle words that your child understands, but does not say.

Remember, only circle words that your child uses on his/her own.

FOOD
apple
banana
bread
butter
cake
candy
cereal
cheese
coffee
cookie
crackers
drink
food
grapes
gum
hamburger
hot dog
ice cream
juice
meat
milk
orange

pizza
pretzel
soda
soup
spaghetti
tea

toast
water

TOYS
ball
balloon
blocks
book
bubble
crayons
doll
present
slide
swine
teddy bear

OUTDOORS
flower
house
moon
rain
sidewalk
snow
star
street
sun

tree

ANIMALS
bee
bird
bus
bunny
cat
chicken
cow
doe
duck
elephant
fish
frog
horse
monkey
pig
PUPPY
snake
tiger
turkey
turtle

BODY PARTS

arm
belly
bottom
chin
ear
elbow
eye
face
finger

BODY PARTS
(Cont'd)
foot
hair
hand
knee
lea
mouth
neck
nose
teeth
thumb
toe
mmmy

PLACES
church
home
hospital
library
McDonalds
park
school
store

zoo

ACTIONS
bath
breakfast
brine
catch
clap
close
come
cough
cut
dance
dinner
doodoo
eat

feed
finish

fix

get

112

ACTIONS
(Cont'd)
give

go
have
help

hit

hue
iump
kick
kiss
knock
look
love
lunch
make
nap
outside
pattycake
peekaboo
peepee
push
read
ride

run

see
show
sing

sit
sleep
stop
take
throw
tickle
walk
want
wash

HOUSEHOLD

bathtub
bed
blanket
bottle
bowl
chair
clock
crib
CUP
door
floor
fork
glass
knife
light
mirror
pillow
plate
potty
radio
room
sink
soap
sofa
spoon
stairs
table
telephone
towel
trash
TV
window

Please list any other words your child uses here:

PERSONAL
brush
comb
glasses
key

money
paper

pen

pencil
penny
pocketbook
tissue
toothbmsh
umbrella
watch

PEOPLE
aunt

baby

boy
daddy
doctor

girl
grandma
grandpa
lady

man
mommy
own name
pet name
uncle
Ernie, etc.

CLOTHES
belt
boots
coat
diaper
dress
gloves
hat
jacket
mittens
pajamas
pants
shirt
shoes
slippers
sneakers
socks
sweater

VEHICLES
bike

boat

bus

car
motorbike
plane
stroller
train
trolley
truck

W

allgone
all right
bad
big
black
blue
broken
clean
cold
dark
dirty
down
good
happy
heavy
hot
hungry
little
mine
more
open
pretty
red
shut
stinky
that
this
tired
"9

wet
white
yellow
yucky

OTHERS
A, B, C, etc.
away
booboo
byebye
curse words
here

hi, hello

in

me

meow

my

myself
nightnight
no

off

on

out

please
Sesame Street
scuse me
shut up
thank you
there
under
welcome
what
where

why
woofwoof
yes

you
yumyum

1, 2, 3, etc.

 

 

Does your child combine two or more words in phrases? (e.g., more cookie, car byebye, etc.)

Yes

Please list below THREE of your child's longest and best sentences or phrases.

No

 

 

 

Child’s name:

 

Date of birth:

 

Filled out by:

 

This survey instrument was developed by Leslie Rescorla, Ph.D. (1989).

113

Checkliste ftir Wortschatzentwicklung

Bitte krelsen Sie jedes Wort ein, das lhr Kind spontan spricht. Wtirter, die lhr Kind imitiert bitte
nicht markieren. Ebenso markieren Sie Wdrter nicht, die lhr Kind versteht, aber nicht spricht.
Denken Sie daran, kreisen Sie nur Wdrter ein, die lhr Kind von selbst spricht.

Lebensmittel
Apfel
Banane
Brot

Butter
Kuchen
Bonbon
Mllsli

KAse

Kaffee

Keks
Zwieback
Trinken

El

Essen
(Wein)traube
(Kau)gummi
Hamburger
Hot Dot

Eis

Saft

Fieisch
Milch
Apfelsine
Pizza
Brezei
Brause/Limo(nade)
Suppe
Spaghetti
Tee

Toast
Wasser

SQelzegg
Ball

(Lufl)bailon
(Bau)klotz
Buch
(Seifen)blase
(Bunt)stifte
Puppe
Geschenk
Rutsche
Schaukel
Teddy

11310;

BAr

Biene
Vogel
MiickelFliege
Hase
Katze
Huhn

Kuh

Hund
Ente
Elefant
Fisch
Frosch
Pferd

Affe
Schwein
Hiindchen
Schlange
Tiger

(T rut)hahn
Schildkrdte

Kdmrteile
Arm

Bauch
Popoll-iintem
Kinn

Ohr
Ellenbogen
Auge
Gesicht
Finger
FuB

Haar

Hand

Knie

Bein

Mund

Hals

Nase
ZAhne
Daumen
Zeh
Magen

Tatigkeiten
bad(en)
frtihstiick(en)
Daumen)
fang(en)
klatsch(en)
zumach(en)
komm(en)
hust(en)
schneid(en)
tanz(en)
Abendbrot
(Isle—(km
ess(en)
filtter(n)
fertig/end(en)
heil(e)
Krieg(en)
geb(en)
geh(en)
hab(en)
helf(en)
hau(en)lschlag(en)
schmus(en)
hiipf(en)
schieB(en)
Kiiss(en)
kiopf(en)
guck(en)lschau(en)
lieb(en)
Mittag
mach(en)
Mittagsschalf
drauBen
Backe, backe Kuchen
Guck-Guck
pischer(n)
schieb(en)
les(en)
fahr(en)
renn(en)
seh(en)
zeig(en)
5509(80)
sitz(en)
schlaf(en)

114

Tatigkeiten

(Cont'g)
halt(en)

nehm(en)

werf(en)

kitzel(n)
lauf(en)lspazier(en)
woll(en)lwili
wasch(en)

Haushalt
(Bade)wanne
Bett
Decke
Flasche
Schiissel
Stuhl

Uhr
Wiege
Tasse

Tar
(FuB)Boden
Gabel
Glas
Messer
Lampe
Spiegel
Kissen
Teller
pof(chen)
Radio
Zimmer
SplilelWaschbecken
Seife

Sofa

LOffel
Treppe
Tisch
Telefon
Handtuch
Miill
Femseher
Fenster

draulsen
Blume
l-laus
Mond
Regen
Gehweg/Burgersteig
Schnee
Stem
StraBe
Sonne
Baum

Persdnliches
Biirste
Kamm

Brille
Schliissel
Geld

Papier
KulilFiiller
(BleDstift
Pfennig
Notizbuch
Tempo
Zahnbiirste
Schinn
(Annband)Uhr

Personen
Tante

Baby

Junge

Papa

Doktor
Madchen
Oma

Opa

Frau

Mann

Mama
eigener Name
Name e. Haustiers
Onkel

Peter, etc.

Qrfi

Kirche

zu Hauseldaheim
Krankenhaus
Blicherel
McDonalds

Park

Schule

Laden

Zoo

file—Rims
Giirtel

Stiefel
Mantel
Windel
Kleid
Handschuh(e)
Hut/Miitze
Jacke
Fauslling
Schlafanzug
Hose

Hemd
Schuh
Hausschuhe
Turnschuhe
Socken

Pulli

Fahrzeuge
(Fahr)rad

Schiff

Bus

Auto
Motorrad
Flugzeug
Kindenrvagen
Zug
StraBenbahn
Laster

Bestimmungen
(d. Ortes, d.

Art u. Weise, etc.)
alle (b)alle

(alles) klar
schlecht

grols

schwarz

blau

kaputt

115

Bestimmungen
(Cont'd)
sauber
kalt
dunkel
schmutzig
unten

gut
froh/gliicklich
schwer
heiB
hunofio
klein

mein
mehr

auf
htibsch

rot

zu
stinktleklig
das

dies

miide
hoch

naB

weiB

gelb

igitt

Verschiedenes
A, B, C, etc.
W39

buh
tschiiB/ade
Schimpfworte
hier

Hallo

in

mirlmich
miau

mein(s)
ichlselber
Nachti

nein
mnterlab
anlaufldran
(r)aus

bitte

Sesam StraBe
Tschuldigung
Ruhei

danke

Verschiedenes
M
da/dort
unter
was

we
warum
wauwau
ja

du

mm
1,2,3, etc.

Bitte notieren Sie hier alle anderen W0rter, die lhr Kind benutzt:

 

 

Enthalten die AuBerungen lhres Kindes Phrasen bestehend aus zwei oder mehr Wdrten? (z.B.
mehr Kekse, Auto tschtils etc.) ja nein

Bitte notieren Sie D R E l satze oder AuBemngen lhres Kindes

 

 

 

Name des Kindes:

 

Geburtsdatum:

 

ausgefiillt von:

 

Datum:

 

Dieser Fragebogen wurde entwickelt von Dr. Leslie Rescorla und tibersetzt von Ddrte Junker.

116

APPENDIX D

LETTER TO PARENTS, PARENT CONSENT FORM, AND
PARENT QUESTIONNAIRE

117

Dear Parent,

I would like to ask for your permission to have your child participate in a
research project on early language acquisition. 1 am specifically interested in
Ieaming more about how normal children build their vocabulary in the beginning
of their language development. Typically we only study children when they are
older and have a sizable vocabulary. Presently there is little information available
about how young children normally develop before they have lots of words at
their disposal. In the future, information about normal vocabulary development
might contribute to knowledge necessary to identify children at risk early in life
so they can receive appropriate support to enhance their language development.

My study will be based on parents’ reports. If you decide to participate you will
be asked to fill out a questionnaire about your child’s developmental history. I
will videotape a 15- to 20-minute free play interaction with your child. While I
play with your child, you will be asked to fill out the Language Development
Survey (LDS), developed by Leslie Rescorla in 1989. To accomplish this, I will

make one or two visits to your home, which will take not more than two hours of
your time (total time).

Participation by you or your child will be strictly voluntary and participation can
be declined at any time. All information about your child will be kept
confidential and will not be revealed in the results from this project.

If you are a parent of a toddler between 23 and 24 months of age who is
beginning to learn English. or English and German simultaneously, and would be
interested in supporting my research project. please contact me as soon as

possible latest by June 15 via phone or email as listed below. I will also be happy
to answer any further questions.

Respectfully,

Ddrte Junker

Graduate Student at Michigan State University
Audiology and Speech Sciences

email: junkerdo@pilot.msu.edu

phone: 734-761-8255

118

Parent Consent Form
Dear Parent,

I would like to ask for your permission to have your child participate in a
research project on early language acquisition. I am specifically interested in

Ieaming more about how normal children build their vocabulary in the beginning
of their language development.

My study will be based on parents’ reports. If you decide to participate you will
be asked to fill out a questionnaire about your child’s developmental history. I
will videotape a 15- to ZO-minute free-play interaction with your child. While I
play with your child. you will be asked to fill out the Language Development
Survey (LDS). developed by Leslie Rescorla in 1989. To accomplish this, I will

make one or two visits to your home, which will take not more than two hours of
your time (total time).

If you would be interested in supporting my research project, please fill out the

consent statement below and return it to me as soon as possible, latest by June 15.
1999.

Parent Consent Statement

Title of the Project: Expressive Vocabulary Development and Word
Combinations of English-Speaking, German-Speaking, and
German-English Bilingual Toddlers

Investigator: Dorie Junker
Michigan State University
Department of Audiology and Speech Sciences
East Lansing, MI 49924

My child
and I WILL/ WILL NOT agree to
participate in the research project on analysis of my child’s vocabulary.

 

1 have been informed by the investigator of the procedures and goals of the study.
I understand that my child or I may refuse to participate or may discontinue
participation at any time without penalty. I understand further that my child’s
identity will be kept confidential. Neither I nor my insurance company wrll incur
any expense as a result of my child’s participation in this project.

 

 

(Signature of parent! guardian) Date

119

Parents’ Interview - Questionnaire

Child’s name:
Child’s birth date:
address:

 

 

 

 

mother’s name:
profession
education: high school diploma. college degree: BA, MA, PhD, other
foreign languages spoken

father’s name:
profession
education: high school diploma college degree: BA, MA, PhD, other
foreign languages spoken

 

 

 

 

siblings:
Medical History

1. Did any complications or health problems occur during pregnancy? Yes - No
if yes. please describe

 

 

2. Did any complications occur during birth? Yes - No
if yes, please describe
3. Was your child born at term date? Yes - No

What was your child’s birth weight?
5. What were his Apgar values?

 

 

 

6. Has your child had any serious health problems or concerns? Yes - No
if yes, please describe
7. Has your child been hospitalized? Yes - No

if yes, please describe
8. Has your child participated in regular check-up visits at your pediatrician?

 

 

Yes - No

9. Has your child’s hearing been checked? Yes - No
What was the result? '

10. Has your child had repeated ear infections? Yes - No

If yes, how many within the last year?

 

Developmental History continued (based on The Rosetti Infant-Toddler
Language Scale and Bang & Dodson, Birth to Three, Developmental Scale)

Pragmatic Skills
11. Mainly uses words to interact with others? Yes - No
12. Takes turns with others during conversations? Yes - No

120

Functional/Conventional and Symbolic Gestures
a) 21-24 months

13a) Pretend to pour from one container into another.
(imitates adult object use)
4a) Flies a toy airplane in the air.

b) 24-27 months

13b) Pretends to talk on the telephone.

14b) Wipes his hands and and face with a napkin. bib or washcloth.

Problem Solving
18-24 months

places three shapes in form box
(places six pegs in pegboard)

24-30 months
places three shapes in rotated form box
(builds six block tower)

Motor
18-24 months

pulls off own shoes, socks
walks up stairs alone

24-30 months

jumps in place two or more times
descends stairs alone

Play Skills
a) 21-24 months

15a) Attempts to repair broken toys. (also cognition)
16a) Stacks or assembles play objects. (also fine motor)

b) 24-27 months

15b) Performs many related activities with the same toy during
a play sequence.
16b) chooses toys selectively.

17. Uses toys appropriately

If no, describe type of toy and quantity of toys used in an
unusual manner

121

Yes - No

Yes- No

Yes-No
Yes-No

Yes- No

Yes-No

Yes-No
Yes-No

Yes-No
Yes-No

Yes-No
Yes-No

Yes-No
Yes-No

Yes-No

Language Comprehension
a) 21-24 months

18a) Chooses one object from a group of five upon request.
19a) Follows novel commands given without gestures.

20a) Follows a two-step related command given without gestures.

21a) Understand new words rapidly.

b) 24-27 months

18b) Points to four action words in picture.

19b) Recognizes family members by their given name.
20b) Understands the concept of one.

21b) Understands size concepts

Language Expression

a) 21-24 months

22a) Uses two-word phrases frequently.

22a) Uses 50 different words.

23a) Uses new words regularily.

24a) Relates personal experience.

25a) Uses three word phrases occasionally.

26a) Refers to self by name.

27a) Uses early pronouns occasionally.

28a) Mean length of 1.25-1.50 morphemes per utterance.

b) 24—27 months

Yes-No
Yes-No
Yes-No
Yes-No

Yes-No
Yes-No
Yes-No
Yes-No

Yes-No
Yes-No
Yes-No
Yes-No
Yes-No
Yes-No
Yes-No
Yes-No

22b) Can ask questions (intonation pattern, two word question form).Yes-No

23b) Uses three-word phrases frequently.
24b) Asks for assistance with personal needs.
25b) Uses action words.

26b) Mean length of 1.50-2.00 morphemes per utterance.

Family History
29) At what age did your child say his/her first words?

Yes-No
Yes-No
Yes-No

Yes-No

 

30) Family History of Speech, Language and/or Learning Problems Yes-No

For Bilingual Children

 

31) Who communicates with your child in English?
32) Who communicates with your child in German?

 

33) how many hours per week does your child communicate in English?
34) how many hours per week does your child communicate in German?

122

35) Are there certain settings, situations, play themes specific for one language?
How do you separate the languages?

 

 

36) What is the native language of the mother father ?

 

37) Which language is spoken within the family?

38)Why do you want your child to learn two languages?

 

39) Have there been any problems with your child Ieaming two languages
simultaneously? If yes, describe.

 

For All Children
40) Do you or your spouse play with your child? Yes—No

if yes: a) How many minutes/hours per day?

 

b) What type of games?

 

41) Do you do special activities/classes with your child? Yes-No

if yes: what type of activities?

 

42) Do you read books with your child? Yes-No

if yes: How many minutes/hours?

 

43) Does your child watch TV? Yes-No

if yes: How many minutes/hours?

 

44) Describe a typical day/ routines + weekend activities

123

REFERENCES

124

REFERENCES

Ardila, A. (1998). Bilingualism: A neglected and chaotic area. Aghasiolggy, 12,
131 -1 34.

Bates, E., Marchman, V., Thai, D., Fenson, L., Dale, R, Reznick, J. S., Reilly, J.,
& Hurting, J. (1994). Development and stylistic variation in the

composition of early vocabulary. Journal of Child Language, 12, 85-124.

Battle, D. E. (1997). Communication disorders in a multicultural society. In D.

E. Battle (Ed.), Communication disorders in multicultural mpulations (2"
ed., pp. 3-30). Boston: Butterworth-Heinemann.

Baum, S. (1997). Early bilingual language agguisition: A sociolinguistic
pgrsgective. Paper presented at the 5" Annual Symposium About

Language and Society in America, Austin, TX.

Benedict, H. (1979). Early lexical development: Comprehension and

Production. Journal of Child Language, 6, 183-200.
Blackwell Rosetti, R. (1990). The Rosetti infant-Toddler Language Scale: A

measure of communication and interaction. East Moline, lL: Lingui
Systems.

Bloom, L. (1973). Qne word at a time: The use of single word utterances
before syntax. The Hague: Mouton.

Bloomfield, L. (1933). Language. New York: Holt.

Boysson-Bardies, B. de, Halle, P., Sagart, L., & Durand, C. (1989). A
crosslinguistic investigation of vowel forrnants in babbling. Journal of

Child Language, 16, 1-18.

Campbell, L. R., 8 Taylor, O. L. (1992). ASHA-certified speech-language
pathologists: Perceived competency levels with selected skills. Ina

Howard Journal of Communications, 3(3-4), 163-176.

125

Carroll, J. 8., Davies, P., 8. Richman, B. (1971). The American Heritage word
freguenpy book. New York: Houghton Mifflin.

Carson, D. K, Klee, T., Perry, C. K, Muskina, G., 8. Donaghy, T. (1998).
Comparison of children with delayed and normal language development
at 24 months of age on measures of behavioral difficulties, social and
cognitive development. Infant Mental Health Journal 19(1), 58-75.

 

Caselli, M. C., Bates, E., Casadio, P., Fenson, J., Frenson, L., Sanderl, L., &
Weir, J. (1995). A cross-linguistic study of early lexical development.

anitive Development, 10(2), 159-199.

Choi, S. (1997). Language-specific input and early semantic development:
Evidence from children ieaming Korean. In D. l. Slovin (Ed.), _‘Lh_e_

Qrossiinguistic Study of Language Aguisition (pp. 41 -1 33). Mahwah, NJ:
Erlbaum.

Clark, E. (1979). The ontogenesis of meaning. Wiesbaden, Germany:
Akademische Verlagsgesellschaft Athenaion.

Clark, E. (1987). The principle of contrast: A constraint on language
acquisition. In B. MacWhinney (Ed.), Mechanisms of languagp
amuisition. Hillsdale, NJ: Erlbaum.

Clark, E. (1993). The lexicon in aguisition. Cambridge, UK' Cambridge
University Press.

Clark, E. (1995). Language acquisition: Lexicon and syntax. In J. L. Miller &
P. D. Eimas (Eds), Speech, language, and communication (pp. 303-337).

San Diego: Academic Press.

Coplan, J., Gleason, J. R., Ryan, R., Burke, M. G., 8. Williams, M. L. (1982).
Validation of an early language milestone scale in a high-risk population.
Pediatrics 70 677-683.

 

Costarides, A H., & Shulman, B. B. (1998). Nomi-referenced language
measures: implications for assessment of infants and toddlers. Topics in
Language Disorders, 18(3), 26-33.

Dale, P. S. (1991). The validity of a parent report measure of vocabulary and

syntax at 24 months. Journal of Speech and Hearing Research, 34, 565-
571.

126

Dale, P. S. (1996). Parent report assessment of language and communication.
In K N. Cole, P. S. Dale, & Thai, D. J. (Eds), Assessment of
communication and language (pp. 161 -1 82). Baltimore: Paul H. Brooks.

Dart, B. (1999). Youth spending more time on TV, videos than school. The Ann
Arbor [Michigan] News, pp. A1, A6.

De Bot, K (1992). A bilingual production model: Levelt’s "speaking' model
adapted. Appligg Linguistics, 13, 1-24.

De Houwer, A (1990). The aguisition of two languages from birth: A case
study. Cambridge, UK' Cambridge University Press.

Eilers, R. E., Gavin, W. F., & Oller, D. K (1981). Cross-linguistic perception in
infancy: Early effects of linguistic experience. Journal of thld

Language, 9, 289-302.

Ellis Weismer, S., Murray-Branch, J., 8 Miller, J. (1994). A prospective
longitudinal study of language development in late talkers. Joumai of

Sflm and Hearing Research, 34, 852-867.

Fenson, L., Dale, R, Reznick, 8., Thai, 0., Bates, E., Hartung, J., Pethick, S., &

Reilly, J. (1993). The McArthur communicative development inventories.
San Diego, CA: Singular Publishing.

Fenson, L., Dale, R, Reznick, 8., Thai, D., & Pethick, S. F. (1994). Variability in
early communicative development. Monographs of the Socieg for
Research in thld Development, 59 (5, Serial No. 242).

Femald, A, 8. Hiromi, M. (1993). Common themes and cultural variations in
Japanese and American mothers” speech to infants. Shild Development,
S4, 637-656.

Fix, M., &Pasel, J. (1994). Immigration and immigrants: Setting the record
straight. Washington, DC: Urban institute.

Fi‘ihring, M., Lettmayer, O., Elstner, W., & Lang, H. (1985). Die Sprachfehler
des Kindes und ihre Beseitigung. Wien: Osterreichischer Bundesverlag.

Gass, S. (1994). Second language aguisition: An introduction. Hillsdale, NJ:

Erlbaum.

Genesee, F. (1989). Early bilingual development: One language or two?
Joumai of Shild Language, 16, 161-179.

127

Genesee, F. (1993). Bilingual development in preschool children. in D. Bishop

8. K Mogford (Eds.), Language development in exceptional circumstances
(pp. 62-78). Hillsdale, NJ: Erlbaum.

Genesee, F., Nicoladis, E., 8. Paradis, J. (1995). Language differentiation in
early bilingual development. Joumai of Shild Language, 22(3), 611-631.

Gentner, D. (1982). Why nouns are learned before verbs: Linguistic relativity
versus natural partitioning. In S. A. Kuczaj (Ed.), Language development:

Vol. 2. Language, cognition, and culture. Hillsdale, NJ: Erlbaum.

Goldfield, B. A, & Reznick, J. S. (1990). Early lexical acquisition: Rate,

content, and the vocabulary spurt. Joumai pf Shild Language, 17, 171 -
183.

Goldfield, B. A, & Reznick, J. S. (1996). Measuring the vocabulary spurt: A
reply to Mevis and Bertrand. Joumai of Shild Language, 23, 241-246.

Goodz, N. S. (1989). Parental language mixing in bilingual families. Infant
Mental Health Joumai, 10, 25-43.

Gopnik, A, & Choi, S. (1995). Names, relational words, and cognitive
development in English and Korean speakers: Nouns are not always
learned before verbs. In M. Tomasello & W. Merriman (Eds), Seyond

names for things: Young children’s aguisition of verbs (pp. 63-80).
Hillsdale, NJ: Erlbaum.

Grimm, H., Doil, H., Mueller, C., 8. Wilde, S. (1996). Eltemfragebogen for die
differentielle Erfassung fruher sprachlicher Fahigkeiten. Sprache &
Kognition, 15(1-2), 32-45.

Grosjean, F. (1982). Life with two languages. Cambridge, MA Harvard
University Press.

Grosjean, F. (1989). Neurolinguists, beware! The bilingual is not two
monolinguals in one person. Brain and Language, 36, 1-15.

Hart, B., & Risley, T. R. (1995). Meaningfpl differences in the everyday
exmrience of young American children. Baltimore: Paul H. Brookes

Publishing.

Jackson-Maldonado, D., & Bates, E. (1988). Inventario dei Dessarrollo de las
Habilidades omunicativas ommunicative Develo ment kills
inventopy]. San Diego: University of California, Center of Research on
Language.

128

Johnston, J. R. (1992). Cognitive abilities of language-impaired children. In P.

Fletcher 8. D. Hall (Eds), Specific speech and language disorders in

children: Qorrelates, characteristics and outcomes (pp. 105-116). San
Diego, CA: Singular Publishing Group.

Kamhi, A G. (1993). Children with specific language impairments: Perceptual
and cognitive aspects. In G. Blanken, J. Dittmann, H. Grimm, J. D.
Marshall, & C. W. Wailesch (Eds). Linguistic disorders and pathologies:
An international handbook (pp. 625-640). Berlin: Walter de Gruyter.

Kauschke, C. (1999). Fruher Wortschatzerwerb im Deutschen: Eine
empirische Studie zum Entwicklungsverlauf und zur Komposition des
kindlichen Lexikons. in J. Meibauer 8. M. Rothweiler (Eds), Das Lexikon

im Spracherwerb, Tilbingen (pp. 128-156). Tabingen, Germany:
Francke.

Kelly, D. J. (1997). Patterns of verb usage by preschoolers with normal
language and specific language impairment. Applied Psycholinguistics,

1S, 199-218.

Kelly, D. J. (1998). A clinical synthesis of the late talker literature: implication
for service delivery. an ua e oh and Hearin ervices in
Schools, 29, 76-84.

Kielhdfer, B., & Jonekeit, S. (1998). Zweisprachige Kindererziehung.
Ti‘ibingen, Germany: Stauffenburg Verlag.

Klee, T., Carson, D. K, Garvin, W. J., Hall, L., Kent, A, & Reece, S. (1998).
Concurrent and predictive validity of an early language screening

program. Joumai of Speech, Language, and Hearing Research, 41, 627-
641.

Kohler, K J. (1977). Einftihrung in die Phonetik des Deutschen. Berlin: Erich
Schmidt Verlag.

Koppe, R. (1996). Language differentiation in bilingual children: The
development of grammatical and pragmatic competence. Linguistics, 34,
927-954.

Ladefoged, P. (1982). A course in phonetics. New York: Harcourt Brace
Jovanovich.

Leonard, L. B. (1998). thldren with sgcific language impairment. Cambridge,
MA: MIT Press.

129

Leopold, W. F. (1939). Speech development of a bilingual child: A linguist’s

r rd: Vol. 1. Vocabula rowth in the firsttwo ears. Evanston, IL:
Northwestern University Press.

Leopold, W. F. (1949). Sfiech development of a bilingual child: A linguist’s
rflrg: Vol. 3. grammar and general problems in the first two years.

Evanston, IL: Northwestern University Press.

Locke, J. L. (1993). The child’s path to sppken language. Cambridge, MA:
Harvard University Press.

MacNamara, J. (1967). The bilingual’s linguistic perfonnance—A psychological
overview. Joumai of Social issues, 23(2), 58-77.

MacNamara, J. (1972). The cognitive basis for language ieaming in infants.
Psychological Review, 79, 1-13.

McLaughlin, B. (1984). econd lan ua ea uisition in childhood: Vol. 1.
W Hillsdale, NJ: Erlbaum.

Mehler, J., Lambertz, G., Jusczyk, P., & AmieI-Tison, C. (1986). Discrimination
de la langue matemelle par le nouveau-né. Académie des Sciences, S,
637-640.

Meibauer, J., & Rothweiler, M. (1999). Das Lexikon im Spracherwerb.
Tubingen, Germany: Francke.

Mervis, C. B., & Bertrand, J. (1995). Early lexical acquisition and the
vocabulary spurt: A responsive to Goldfield and Reznick. qumal of

thld Language, 22(2), 461 -468.

Nelson, K E. (1974). Concept, word, sentence: interrelations in acquisition
and development. Psycholpgical Review, S1, 267-285.

Nelson, K E. (1981). Individual differences in language development:
Implications for development and language. Qevelopmental Psychology,
1_7_, 1 70-187.

Oller, D. K, Eilers, R. E., Urbano, R., & Cobo-Lewis, A B. (1997).
Developmental precursors to speech in infants exposed to two languages.

Journal of thld Language, 24(2), 407-425.

Patterson, J. L. (1998). Expressive vocabulary development and word
combinations of Spanish-English bilingual toddlers. American Joumai of

speech-Language Pathology, 4, 46-56.

130

Paul, R. (1991). Profiles of toddlers with slow expressive language
development. Topics of Language Disorders, 11(4), 1-13.

Paul, R. (1996). Clinical implications of the natural history of slow expressive
language development. American oumal of eech-Lan ua e

Pathology, 5(2), 5-21.

Paul, R., 8 Smith, R. (1993). Narrative skills in 4-year-olds with normal,

impaired, and late-developing language. Joumai of Sp_ee_ch and Hearing
Epsearch, 5, 592-598.

Pearson, B. Z. (1998). Assessing lexical development in bilingual babies and

toddlers. lntemational Joumai of Silingualism, 2(3), 247-372.

Pearson, B. 2., Fernandez, S. C., Lewedeg, V., 8 Oller, D. K (1997). The
relation of input factors to lexical Ieaming in bilingual infants. Applied

Psycholinguistics, 18, 41 -58.

Pearson, B. 2., Fernandez, S. C., 8 Oller, D. K (1995a). Cross-language
synonyms in the Iexicons of bilingual infants: One language or two?

Joumai of thld Language, 22(2), 345-368.

Pearson, B. Z., Fernandez, S. C., 8 Oller, D. K (1995b). Lexical development
in bilingual infants and toddlers: Comparison to monolingual norms. In B.

Harley (Ed.), Lexical issues in language ieaming (pp. 31 -57). Ann Arbor,
MI: Research Club in Language Learning.

Porsche, D. C. (1983). Die Zweisprachigkeit wahrend des primaren
Spracherwerbs. Tubingen, Germany: Gunter Narr Verlag.

Pregel, D., 8 Rickheit. G. (1987). Qer Wortschatz im Srundschulalter,

Haufigkeitsworterbuch zum verbalen, substantivischen und adiektivischpn
Wortgebrauch. Hildesheim: Olms.

Quay, S. (1995). The bilingual lexicon: Implications for studies of language

choice. Joumai of thld Language, 22(2), 369-387.

Rescorla, L. (1989). The Language Development Survey: A screening tool for
delayed language in toddlers. Joumai of Speech and Hearing Disorders,
S4, 587-599.

Rescorla, L. (1991a). identifying expressive language delay at age two. Topics
of Language Disorders, 11(4), 14-20.

131

Rescorla, L. (1991b). The Language Development Survey user’s manual. Bryn
Mawr, PA' Author.

Rescorla, L. (1993). Use of parental report in the identification of

communicatively delayed toddlers. Seminars in Speech and Language,
14, 264-277.

Rescorla, L., Hadicke-Wiley, M., 8 Escarce, E. (1993). Epidemiological
investigation of expressive language delay at two. First Language, 13, 5-
22.

Rescorla, L., Roberts, J., 8 Dahlsgaard, K (1997). Late talkers at 2: Outcome

at age 3. Joumai of Sm, Lgnggage, am flearing Reseprch, 49, 556-
566.

Resnick, T., Allen, D., 8 Rapin, I. (1984). Disorders of language development:
Diagnosis and intervention. Pediatrics in Review 6 85-92.

 

Rice, M., 8 Bode, J. (1993). Gaps in the verb lexicons of children with specific
language impairment. First Language, 13, 1 13-131.

Rothweiler, M., 8 Meibauer, J. (1999). Dax Lexikon im Spracherwerb-Ein
éUberblick. in J. Meibauer 8 M. Rothweiler (Eds), Das Lexikon im
Spracherwerb (pp. 9-31). Tubingen, Germany: Francke.

Saunders, G. (1988). Silingual children: From birth to teens. Philadelphia:
Multilingual Matters.

Schreuder, R., 8 Weltens, B. (1993). The bilingual lexicon. Philadelphia, PA:
John Benjamins North America.

Schreuder, R., 8Weltens, B. (1993). The bilingual lexicon. Philadelphia: John
Benjamins North America.

Singleton, D., 8Lengyel, Z. (1995). Thea factori econ lan
aguisition. Clevedon, UK: Cromwell Press.

Slobin, D. (Ed.). (1992). Ihg crosslinguistic study of language aguisition. (Vol.
3). Hillsdale, NJ: Erlbaum.

US. Bureau of Census. (1998). Statistical abstract of the united States: The
national datg book (118“ ed.). Washington, DC: US. Government

Printing Office, pp. 55-56.

132

Vihman, M. M. (1985). Language differentiation by the bilingual infant. Journal
of thid Language, 12, 297-324.

Vihman, M. M. (1996). Phonolpgical development The origins of language in
the child. Cambridge, MA: Blackwell.

Volterra, V., 8 Taeschner, T. (1978). The acquisition and development of
language by bilingual children. Joumai of Shild Language, 5, 311-326.

Watkins, R., Rice, M., 8 Moltz, C. (1993). Verb use by language-impaired and
normally developing children. First Language, 13, 133-143.

Weinreich, U. (1974). Languages in Qntact. The Hague: Mouton.

Weiss, A von. (1959). Hauptprobleme der Zweisprachigkeit. Heidelberg:
Stauffenburg Verlag.

Wendlandt, W. (1992). Sprachstorungen im Kindesalter. Materialien zur
Fruherkennung und Beratung. Stuttgart: Georg Thieme Verlag.

133

 

 

I'll...
I‘I“

rttli‘lol'

"IIllllllllllllllllllllllls