02whole PDF

Download as pdf or txt
Download as pdf or txt
You are on page 1of 302

‘Mami, lietadlo! Aeroplane, daddy!

’:
A Case Study Exploring Bilingual First
Language Acquisition in a Mixed-
Lingual Family

Margareta Rebelos

A thesis submitted in fulfilment of the degree of Doctor of Philosophy

Discipline of Linguistics, School of Humanities

The University of Adelaide

August 2012
Table of Contents
Table of Contents ................................................................................................................... i
List of tables .......................................................................................................................... v
List of figures ........................................................................................................................ v
List of abbreviations ............................................................................................................. vi
Abstract................................................................................................................................ vii
Declaration .........................................................................................................................viii
Acknowledgements .............................................................................................................. ix
1 Introduction ................................................................................................................... 1
1.1 Theoretical background .......................................................................................... 2
1.2 Research questions .................................................................................................. 5
1.3 Thesis structure ....................................................................................................... 6
2 Literature review ........................................................................................................... 7
2.1 Bilingual First Language Acquisition (BFLA) ....................................................... 8
2.1.1 Early vocal development and sound production ........................................... 11
2.1.2 Bilingual Comprehension .............................................................................. 12
2.1.3 First words and early phonological development .......................................... 13
2.1.4 Assessing bilingual vocabulary ..................................................................... 16
2.1.5 Combinatorial speech and morphosyntactic development ............................ 17
2.2 Language differentiation ....................................................................................... 18
2.3 Language mixing .................................................................................................. 22
2.4 Language choice ................................................................................................... 25
2.5 Parental discourse strategies ................................................................................. 26
2.6 Theory guiding this thesis ..................................................................................... 29
3 Methodology and study design .................................................................................... 35
3.1 Data collection ...................................................................................................... 36
3.1.1 Diary records ................................................................................................. 36
3.1.2 Audio-video recordings ................................................................................. 38
3.2 Data analysis ......................................................................................................... 41
3.2.1 Transcription.................................................................................................. 43
3.2.1.1 Selection for transcription ................................................................................... 43
3.2.1.2 Phonetic and orthographic transcription ............................................................ 44
3.2.1.3 Conventions of CHAT transcriptions .................................................................... 46
3.3 Data presentation .................................................................................................. 48
4 Language learning environment in the family and linguistic input ............................. 52
4.1 Child’s parents ...................................................................................................... 52
i
4.1.1 The mother..................................................................................................... 53
4.1.2 The father....................................................................................................... 54
4.1.3 From monolingual to bilingual home ............................................................ 55
4.2 The child ............................................................................................................... 56
4.2.1 Ria’s SES and health ..................................................................................... 57
4.3 Parental language presentation ............................................................................. 58
4.3.1 Language presentation within domains of use .............................................. 68
4.3.2 Linguistic soundscape (audio-media and other media) ................................. 75
4.4 Language attitudes and impact belief in the family .............................................. 76
4.4.1 Attitudes outside the family........................................................................... 78
4.5 Language models .................................................................................................. 80
4.5.1 Ria’s languages-in-acquisition: Slovak and English ..................................... 81
4.5.2 Child Directed Speech (CDS) ....................................................................... 82
4.5.3 Parental language choice and code-switching ............................................... 91
4.5.4 Parental discourse strategies .......................................................................... 93
5 Becoming Familiar with the Languages (birth-1;3) .................................................. 101
5.1 Prenatal stage ...................................................................................................... 101
5.2 Prestages ............................................................................................................. 102
5.2.1 Bilingual sensitivity ..................................................................................... 103
5.2.2 Recognizing linguistic forms in the input ................................................... 103
5.3 Transition to language ........................................................................................ 106
5.3.1 Pure babbling ............................................................................................... 107
5.3.2 Protolanguage .............................................................................................. 109
5.3.3 Using gestures to communicate intentions .................................................. 112
5.3.4 Emerging Word Comprehension ................................................................. 112
5.3.4.1 Comprehension in bilingual environment ......................................................... 116
5.3.5 First words ................................................................................................... 122
5.3.5.1 The role of onomatopoeia ................................................................................. 123
5.3.5.2 Producing words ................................................................................................ 124
6 Using words (1;4-1;6)................................................................................................ 130
6.1 Early word development ..................................................................................... 131
6.1.1 Rate of vocabulary growth .......................................................................... 132
6.1.2 Reproduced words ....................................................................................... 134
6.1.3 Spontaneous productions ............................................................................. 134
6.1.4 Attention to language .................................................................................. 136
6.1.4.1 Strategies of production .................................................................................... 137
6.1.4.2 Vocabulary composition .................................................................................... 143
6.1.4.3 Fixed expressions ............................................................................................... 146
ii
6.2 Bilingual Repertoire ............................................................................................ 147
6.2.1 Using TEs .................................................................................................... 152
7 Using structures (1;7-2;0) .......................................................................................... 161
7.1 Changing words .................................................................................................. 163
7.2 New words .......................................................................................................... 164
7.2.1 Overextension .............................................................................................. 165
7.3 Combining words ................................................................................................ 166
7.3.1 Word order................................................................................................... 168
7.3.2 Fixed expressions ........................................................................................ 171
7.3.3 Expressing negation..................................................................................... 175
7.4 Inflectional morphology ..................................................................................... 177
7.4.1 Nominal morphology................................................................................... 178
7.4.1.1 Slovak nominal morphology .............................................................................. 180
7.4.2 Gender and number agreement .................................................................... 183
7.4.2.1 Comparatives ..................................................................................................... 185
7.4.3 Slovak verb morphology ............................................................................. 185
7.4.3.1 Other verb structures ........................................................................................ 191
7.4.4 The verbs ‘byť’ and ‘to be’ ......................................................................... 194
7.4.5 English verb morphology ............................................................................ 195
8 Emerging Bilingualism .............................................................................................. 199
8.1 Ria’s language mixing ........................................................................................ 201
8.1.1 Intra-sentential mixing................................................................................. 208
8.1.1.1 Mixing with the mother ..................................................................................... 210
8.1.1.2 Mixing with the father ....................................................................................... 213
8.1.1.3 Mixing in triadic interactions ............................................................................. 214
8.1.2 Bilingual blends ........................................................................................... 216
8.2 The developing sense of appropriate language choice ....................................... 218
8.2.1 Repairing language choice .......................................................................... 228
8.3 The developing sense of bilingualism ................................................................ 240
9 Conclusion ................................................................................................................. 245
9.1 Study Limitations and future research ................................................................ 251
APPENDIX A Transcription conventions......................................................................... 253
APPENDIX B Universal proto-word ................................................................................ 260
APPENDIX C Onomatopoeia ........................................................................................... 262
APPENDIX D First 50 words ........................................................................................... 263
APPENDIX E Fixed pivot schemas and item-based frames ............................................. 265
APPENDIX F Early word combinations ........................................................................... 269

iii
APPENDIX G Ria’s Mixed Utterances ............................................................................ 271
Bibliography ...................................................................................................................... 286

iv
List of tables
TABLE 2.1 BFLA CASE STUDIES .............................................................................................................................. 30
TABLE 4.1 FAMILY COMPOSITION............................................................................................................................ 52
TABLE 4.2 PATTERNS OF HOME LANGUAGE USE (ADAPTED FROM DE HOUWER 2009)....................................................... 56
TABLE 4.3 RIA’S RESIDENCE AND LONG-TERM TRAVEL (>1 WEEK) SINCE BIRTH ................................................................. 57
TABLE 4.4 RIA’S CARE HISTORY IN THE HOME ............................................................................................................. 69
TABLE 4.5 WEEKLY PATTERNS OF LANGUAGE CONTEXT EXPOSURE .................................................................................. 70
TABLE 4.6 EXPOSURE TO BILINGUAL LANGUAGE MIXING............................................................................................... 72
TABLE 5.1 FIRST 10 WORDS ................................................................................................................................. 128
TABLE 6.1 MONOSYLLABIC HOMOPHONES .............................................................................................................. 141
TABLE 6.2 FIXED EXPRESSIONS .............................................................................................................................. 146
TABLE 6.3 TRANSLATION EQUIVALENTS IN EXPRESSIVE VOCABULARY............................................................................. 149
TABLE 6.4 TRANSLATION EQUIVALENTS IN EXPRESSIVE VOCABULARY - CONTINUED .......................................................... 150
TABLE 6.5 LANGUAGE NEUTRAL WORDS ................................................................................................................. 150
TABLE 6.6 INDETERMINATE WORDS ....................................................................................................................... 151
TABLE 7.1 ENGLISH FIXED PIVOT SCHEMAS ............................................................................................................. 172
TABLE 7.2 ENGLISH FIXED ITEM-BASED FRAMES........................................................................................................ 172
TABLE 7.3 SLOVAK FIXED PIVOT SCHEMAS ............................................................................................................... 173
TABLE 7.4 SLOVAK FIXED ITEM-BASED FRAMES......................................................................................................... 173
TABLE 7.5 SLOVAK NOMINAL CASES ....................................................................................................................... 180
TABLE 8.1 USE OF MIXED ONE-WORD UTTERANCES IN DIFFERENT LANGUAGE CONTEXTS (INAPPROPRIATE LANGUAGE CHOICE) . 202
TABLE 8.2 USE OF DIFFERENT TYPES OF COMBINED UTTERANCES (IN PROPORTION TO TOTAL COMBINED UTTERANCES) ........... 203
TABLE 8.3 USE OF MIXED COMBINED UTTERANCES IN DIFFERENT LANGUAGE CONTEXTS.................................................... 204
TABLE 8.4 USE OF UNILINGUAL COMBINED UTTERANCES IN INAPPROPRIATE CONTEXTS..................................................... 204
TABLE 8.5 MIXED WORD TYPES ACROSS LANGUAGES (ONE-WORD STAGE) ..................................................................... 209

List of figures
FIGURE 1 DEVELOPMENTAL STAGES ......................................................................................................................... 50
FIGURE 2 USE OF ONE-WORD UTTERANCES WITHIN CONTEXTS..................................................................................... 202
FIGURE 3 USE OF TWO/MULTIPLE-WORD UTTERANCES WITHIN CONTEXTS ..................................................................... 204

v
List of abbreviations
BFLA Bilingual First Language Acquisition

CDI MacArthur Communicative Development Inventory

CDS Child Directed Speech

CS Code-switching

ENI English Noun Inflection

ESLA Early Second Language Acquisition

IDS Infant Directed Speech

IPA International Phonetic Alphabet

MFLA Monolingual First Language Acquisition

OPOL One parent-one language

SG Slovak Gender morpheme

SNI Slovak Noun Inflection

SPA Slovak Phonetic Alphabet

SVI Slovak Verb Inflection

TE(s) Translation Equivalent(s)

1P/1L 1 parent/1 language

1P/2L 1 parent/2 languages

1P/1L & 1P/2L 1 parent/1 language & 1 parent/2 languages

1sg 1st person singular

2sg 2nd person singular

3sg 3rd person singular

1pl 1st person plural

2pl 2nd person plural

3pl 3rd person plural

vi
Abstract
Family environment plays a crucial role in bilingual language socialization in early
childhood. The bilingual family introduces the child not only to the languages-in-
acquisition, but also to the preferred language use patterns. In recent years the discussion
on how and when a bilingual child comes to use her two languages in contextually
appropriate ways has become central to Bilingual First Language Acquisition (BFLA).
While evidence for language differentiation and sensitivity to interlocutor’s preferred
language is available in the two-word stage, few studies consider the language learning
environment and its impact on bilingual development in the one-word stage and early
combinatorial speech.

This longitudinal case study reports on linguistic developments from birth to 2;0 in a child
who was acquiring Slovak and English simultaneously in the home. The effects of the
child’s language learning environment on linguistic development from the onset of speech
were considered, focusing on lexical development, word combinations, emerging morpho-
syntax, and pragmatic aspects such as language choice and mixing.

The child’s two languages developed separately in a side-by-side fashion, as shown by use
of translation equivalents and language specific morphological markers from the
beginning. She used the two languages in contextually sensitive ways from the one-word
stage, relying on several pragmatic language choice strategies. Mixing was productive and
accounted only for a small proportion of productions. It was explained by sociolinguistic
as well as psycholinguistic factors. Language differentiation thus emerged as grammatical
as well as pragmatic differentiation at the end of the one word stage.

vii
Declaration
I certify that this work contains no material which has been accepted for the award of any
other degree or diploma in any university or other tertiary institution and, to the best of my
knowledge and belief, contains no material previously published or written by another
person, except where due reference has been made in the text. In addition, I certify that no
part of this work will, in the future, be used in a submission for any other degree or
diploma in any university or other tertiary institution without the prior approval of the
University of Adelaide.

I give consent to this copy of my thesis, when deposited in the University Library, being
made available for loan and photocopying, subject to the provisions of the Copyright Act
1968.

I also give permission for the digital version of my thesis to be made available on the web,
via the University’s digital research repository, the Library catalogue and also through web
search engines, unless permission has been granted by the University to restrict access for
a period of time.

viii
Acknowledgements
I would like to express gratitude to my daughter Ria for being the very centre of my
research and for making this case study possible. I am thankful to her for being a delightful
participant who made my research enjoyable. I also wish to thank other participants,
members of the extended family, who allowed me to observe their interactions with Ria,
and interactions between their children and Ria. Without their participation my research
would not have been complete.

I owe gratitude to my husband John for allowing me to observe and analyse his
interactions with our daughter, for his support and encouragement throughout my study
and for being ready to help when I encountered IT problems. I am also thankful to John for
proofreading my draft and providing helpful editing suggestions.

I am grateful to my principal supervisor Dr Peter Mickan for his guidance throughout my


research, for discussing my ideas and perspectives, for providing critical and constructive
feedback on numerous chapter drafts, for tireless support and encouragement, and for his
enthusiasm for my research.

I also wish to extend gratitude to my co-supervisor Professor Peter Mühlhäusler for


providing critical and constructive feedback on my draft and for his support and
encouragement throughout my research.

I am sincerely thankful to Professor Daniela Velichová and Nikoleta Tomečko for


invaluable editing suggestions and comments on my draft. I am also indebted to Roberta
Velichová and Leonard Velich for meticulous proofreading of my draft.

Last but not least I would like to thank my parents for believing in me, for being an
inspiration, for setting a high standard, and for giving their children love of learning.

ix
1 Introduction
The aim of this study was to document and describe the early years (birth-2;0) of linguistic
development of a child growing up as a simultaneous bilingual, with the view to contribute
to the growing number of studies of varied language combinations in the Bilingual First
Language Acquisition (BFLA) field. Special focus was to fill the gap in childhood
bilingualism research on lexical development and the development of the pragmatic skill
of language choice in the one-word and early two-word stages, as well as to examine
language learning environment in the early years of a child growing up with two
languages.

Exploratory ethnographic approach was employed as the method to examine the child’s
initial bilingual acquisition. A longitudinal descriptive case study allowed long-term
regular observation and documentation of data on both the language learning environment
and gradual linguistic development. It allowed exploring particular phenomena in depth
while assessing the validity of existing models of bilingual linguistic development.

Since the child in the case-study, Ria, is my own child, observations of language
development and linguistic environment could be carried out on daily basis. Being a parent
researcher brought the advantage of being present across various social situations, thus
capturing novel developments and changes as they unfolded. Regular observations
provided insights into Ria’s prelinguistic and linguistic developments, as well as
sociolinguistic factors such as language acquisition history, linguistic environment, and
language usage by the people in Ria’s environment, and thus a more complete picture of
language development.

The methodologies of several case studies, all of which contributed to the knowledge base
in BFLA with findings supporting separate development of the two languages, guided my
research. Three longitudinal case studies, which were also carried out by linguist-parents
and examined BFLA in child/children less than two years old, were used as basis for data
collection and analysis: Deuchar & Quay (2000), Cruz-Ferreira (2006) and Štefánik
(2000). These studies assessed the development of lexicon, emergence of morpho-syntax,
mixing and pragmatic language choice.

Other case studies with children aged 2 and above carried out by researchers that were not
the subjects’ parents were also drawn on. Sociolinguistic aspects such as parental input and
analysis of interactions in the family were based on Lanza’s (1992, 2004) framework of

1
parental discourse strategies. Descriptors of language learning environment were based on
De Houwer (2009a, p. 83). Aspects such as separate development of the two languages,
language differentiation, general theoretical background and methodological guidelines
relevant for BFLA were sourced from De Houwer (1990, 2009a).

The above case studies collected data in naturalistic interactions sourced in the subjects’
homes. Linguistic development was assessed from two data sources – diary and
audio/video-recordings. In this thesis I employed the same data collection procedures. Two
kinds of data sets were collected and analysed. A diary captured general progressions in
development, and more specifically the construction of cumulative vocabulary, emergence
of morpho-syntax, combinatorial speech, language choice and mixing. Weekly video-
recordings throughout the first two years allowed detailed analyses of interactions in the
family in varied situational contexts, and documented changes in language use over a
longer period of time. Regular recordings were also useful to confirm and demonstrate the
developments noted in the diary.

1.1 Theoretical background


Simultaneous development of two languages in young children has been of interest to
researchers since the first published study by Ronjat in 1913. Many studies followed
Ronjat’s, however, it was not until the 1990s that the BFLA field was established (De
Houwer, 2009a, p. 13). In the earlier studies the question of development of one or two
linguistic systems in developing bilinguals dominated research. It produced two opposing
hypotheses – the unitary language system hypothesis versus separate development
hypothesis. More recent studies provided empirical evidence in favour of the separate
development hypothesis, since children raised as simultaneous bilinguals from birth were
found to develop their linguistic capacities in both languages of exposure, and were able to
differentiate their languages grammatically and pragmatically from early stages of
linguistic development (Cruz-Ferreira, 2006; De Houwer, 1990; Deuchar & Quay, 2000;
Genesee, 2007; Meisel, 2001; Nicoladis, 1998).

Most recently the focus of the discussion changed into the question of how early in
development and through what processes does a bilingual child come to differentiate her
two languages. Evidence for differentiation is available in the stage of morpho-syntactic
development when children start combining words into multiple-word utterances and use
morphological markers, which occurs around 24 months. Focus on the development of
morpho-syntax in BFLA suggests that bilingual infants do not show signs of language

2
differentiation prior to combinatorial speech. Indeed, studies on bilingual differentiation in
the early stages of linguistic development, including early lexical development, babbling,
even speech perception, are considerably fewer (Genesee & Cenoz, 2001, p. 5). Only a few
studies address this question at an age earlier than 2 (Deuchar & Quay, 2000; Nicoladis &
Genesee, 1996). Nicoladis & Genesee (1996) found there is an initial period when
bilingual children do not differentiate their languages pragmatically, and that emergence of
differentiation varied considerably among children. Another study found differentiation
according to parental preferred language between the one-word and two-word stage, and it
was observed even in situations with both parents present (Genesee, Nicoladis, & Paradis,
1995).

Bilingual differentiation also remains relatively unexplored from a sociolinguistic


perspective, since few studies focus on the family environment as a crucial factor shaping
the development of bilingualism in the early stages of life (Goodz, 1989; Lanza, 2004).
The current body of research in BFLA also lacks studies with mixed-lingual families in
which one parent is a bilingual speaker proficient in both family languages, while the other
parent is a monolingual speaker able to interact only in one language, and how these
families solve the issue of language choice in triadic interactions.

Although BFLA field grew rapidly and was enriched with new language combinations in
the last decade, western European languages still dominate the research (Lanza, 2001b, p.
241). Combinations with a Slavic language as one of the languages-in-acquisition remain
underrepresented. To date only one case study is available concerning Slovak-English
language combination. However, it is a case of intentional bilingualism in a child growing
up in Slovak majority environment, acquiring English from the father, who is a non-native
speaker, while both parents are bilingual (Štefánik, 2000, p. 25). No case studies of
children growing up with Slovak and English in an English majority environment are
available.

The theoretical foundations of this study stem from the functional or usage-based language
acquisition theories (Halliday, 1975; Painter, 2006; Tomasello, 2003), which consider
learning of a child’s initial languages as context based. Family environment plays a crucial
role in language socialization in early childhood. For a bilingual child the bilingual family
provides the socializing environment and introduces the child not only to the languages-in-
acquisition, but to preferred language use patterns in the family and in the wider
community. In particular, language presentation affects how a bilingual child will learn to
use her languages appropriately in various social situations.
3
In a bilingual family each parent and child share a linguistic bond through the medium of a
particular language, called the person-language bond (Grosjean, 2010, p. 183). This bond
provides a base language for interpersonal interactions between the parent and the child,
and the child becomes sensitive to the language which is normally used by the parent.
Thus the language bond establishes a specific language for interaction, and this in turn sets
a language mode that the child will typically be in when interacting with that parent. In the
pre-linguistic stages of development this is the parents’ preferred language. Thus the base
language preferred by the parent can become the main language of communication in the
parent-child dyad. The base language can be strictly separated, where each language is
shared by the bilingual child and a particular parent, or it can be a mixed variant used by
the family. What is important is that the type of linguistic relationship which is established
at the onset will determine the later language of communication.

A popular strategy among bilingual families is the one parent-one language approach,
which suggests that a child comes to associate her languages with different people, and
therefore learns to separate them. However, research found that although consistency of
input is necessary in this strategy, in reality parents experience difficulties in keeping the
languages apart and often the language of the community becomes the dominant language
in the family (Barron-Hauwaert, 2004, p. 123; Noguchi, 1996). While parental guides to
raising children bilingually stress the importance of consistency of input, parents are left
with little advice on how to ensure consistency. The guides fail to highlight that
consistency creates an emotional and linguistic bond through a particular language
between a child and her parents.

In this thesis I consider the interpersonal principle crucial in establishing a linguistic bond,
and claim that it is this bond in the parent-child dyad that makes bilingual upbringing
possible. The linguistic bond can set the language mode in the dyad to a ‘default mode’ -
the usual language of interaction between the parent and the child. If parents consistently
address a child in their preferred language from birth or even the pre-natal period, she will
establish emotional attachment through that language. The parents will feel at ease using
that language with the child in various social situations, and it will become a natural choice
of language of communication. Moreover, an infant in her first year is yet unable to engage
in the process of negotiation of language choice, a phenomenon which is observed in all
bilingual interactions (Grosjean, 2008, p. 86). Thus the parents have a rather unlimited
opportunity to create the linguistic bond with the child based on their preferred language,
making later language use consistent.

4
1.2 Research questions
This thesis aimed to answer the following research questions, and subquestions which
framed the analysis:

1. How did the language learning environment influence Ria’s bilingual linguistic
development?
1.1. What was Ria’s language acquisition history?
1.2. How did the parents use the two languages and how did they negotiate their
preferred language context?
2. What was the course of bilingual linguistic development?
2.1. Were there any differences in the course of development between the two
languages?
3. What was Ria’s lexical and morpho-syntactic development in the two languages?
3.1. How did Ria’s cumulative lexicon develop in the two languages?
3.2. Did she use translation equivalents and how?
3.3. How did Ria progress into combinatorial speech?
3.4. When and how did morphological development emerge?
3.5. Was Ria using appropriate morphological markers in each language?
4. When and how did Ria start using her languages in contextually appropriate ways?
4.1. How early was pragmatic language choice ability observed?
4.2. What types of utterances did Ria produce?
4.3. Did Ria mix the languages in her utterances and how?
4.4. Was Ria differentiating the two languages, and if so, when and how did
differentiation emerge?
4.5. Were there any signs of language dominance?
In order to answer these questions I examined the earliest stages of Ria’s BFLA. General
developmental patterns were determined, and each stage was assessed in detail. Aspects
such as the construction of cumulative lexicon during the one-word stage, and emergence
of combinatorial speech and morpho-syntactic development during the two-word stage
were analysed. I also analysed how Ria came to use her two languages differentially, and
how parental language choice and discourse strategies facilitated this experience.

The main contribution of this thesis is in the insights into the very early stages of infant
bilingualism. By exploring sociolinguistic factors affecting language acquisition in the
early stages of linguistic development, more detailed information on the development of
pragmatic ability to use two languages appropriately in different social contexts was
5
gained. The findings taken together allowed for a general assessment of Ria’s developing
sense of bilingualism. Moreover, this thesis examined a relatively unexplored language
combination consisting of Slovak and English. Since simultaneous development of two
morphologically different languages was compared, this thesis also contributes to cross-
linguistic developmental research.

Due to space and time constraints I was not able to include detailed analysis of Ria’s
phonetic and phonological development, such as cooing and babbling, and detailed
development of phonological differentiation between the two languages. Likewise, an
analysis of paralinguistic means of communication, such as the use of gesture and kinetics,
and use of protolanguage in the months prior to the onset of speech had to be omitted.
While the basic theoretical background of the interpersonal principal in language
development as explicated by Halliday (1975) and Painter (2006) guided the theoretical
foundations of this study, a functional analysis of the data was not carried out.

1.3 Thesis structure


The content of this thesis is organised as follows. In chapter 2, I shall discuss current
research in the field of bilingual linguistic development. Key theoretical and
methodological approaches that guided this study will be acknowledged.

Chapter 3 provides details of study design, outlines procedures of data collection and
specifies how the data were analysed.

Chapters 4-8 present the data and the results of the analyses. Chapter 4 introduces the case
study by analysing the linguistic environment in the bilingual family. In Chapter 5 I
analyse prelinguistic developments and bilingual word comprehension. In Chapter 6 I
discuss the development of early productive lexicon and cumulative vocabulary repertoire.
Chapter 7 analyses the emerging morpho-syntax and combinatorial speech.

In chapter 8 I interpret the findings of the previous chapters in light of bilingual first
language acquisition, considering the theoretical principles that guided this study. In
Chapter 9 I draw conclusions from the findings, state the limitations of my study and
highlight the contributions of the thesis to the wider research.

6
2 Literature review
Traditionally bilingualism was perceived as deviation from monolingualism (Cruz-
Ferreira, 2006, p. 16; Grosjean, 2008, p. 10; Meisel, 2006, p. 93). Perhaps this perception
was caused by the fact that in the past linguistics was dominated by Western research, and
it was by and large the monolingual speakers who deemed it necessary to describe
bilingualism with reference to monolingualism, in order to gain an understanding of a
bilingual state of mind (Romaine, 1995b, pp. 7-8). This approach used the seemingly
‘common’ monolingual experience as the norm. Modern linguistics has arrived to a new
approach: bilingualism is to be taken as the norm rather than the exception, since
multilingual environments are more common in the world (Baker, 2007, p. 186; Grosjean,
1982; Kroll, Gerfen, & Dussias, 2008, p. 108). Bilingualism is now acknowledged as
another form of lingualism, a term suggested as shorthand for the ‘language faculty’
(Cruz-Ferreira, 2006, p. 16). As such, bilingualism is equal to other forms of lingualism,
namely monolingualism and multilingualism and needs to be researched under this
assumption.

Despite bilingualism being common, research in the field of child language acquisition
focuses predominantly on monolingual development. The little focus there is on bilingual
language acquisition - it has been estimated that approximately only 2% of research on
language development concerns children learning two languages - grants little attention to
bilingual development in the early years (Fernald, 2006, p. 19; Petitto et al., 2001, p. 456).
Thus only small portion of research focuses on the pre-school years when the parents or
carers have the greatest influence on language development (Arnberg, 1987, p. 3).

Given the scarcity of literature on early bilingual language development it seems necessary
to draw from two existing fields: the relatively young field of Bilingual First Language
Acquisition and the vast body of research that is available in language acquisition literature
which uses monolingual norms. Indeed, it has been demonstrated that bilingual
development resembles the overall patterns of monolingual development (Genesee, 2006,
p. 60). Genesee concludes it is evident in the fact that bilingual children reach the same
milestones at approximately the same ages as monolingual children, and that language-
specific grammatical development is also taking place at the same rate as in monolingual
children. However, there appears to be greater variation among bilingual children, because
the pathway to acquisition of two languages is influenced by many factors along the way
of bilingual experience, including age of first exposure to the two languages, amount and
7
patterns of exposure, and typological differences between the two languages (McCardle &
Hoff, 2006, p. ix).

Furthermore, it is stressed in the literature on bilingual language acquisition that any form
of comparison of bilingual development with monolingual development is neither
advisable nor useful (Baker, 2007, p. 35; Cruz-Ferreira, 2006, p. 14), since this approach
only raises monolingualism as the norm bilingualism is to be compared with. Bilinguals
need to be compared with bilinguals.

In this chapter I will first consider research in specific areas of BFLA carried out to date,
to my knowledge, within each developmental period that was identified in the data of this
thesis up to age 2;0. I will consider both the BFLA and general initial language acquisition
fields. I will then turn to phenomena specific for bilingual development, and outline the
theory guiding my research.

2.1 Bilingual First Language Acquisition (BFLA)


Literature concerning early bilingual development can be characterised as case studies
undertaken on children, often linguists’ own, growing up in high SES families. These have
been labelled cases of ‘elite bilingualism’, and much of the criticism suggests they do not
provide models for a ‘typical’ bilingual language development (Romaine, 1995a, p. 187).
However, it could be argued that at present, due to upward social mobility, there are
increasing numbers of families with stable SES with varied motivations to become
bilingual and to maintain bilingualism. Individuals now have a free choice to move to
different countries for various reasons and thus many migrant families no longer belong to
low SES parts of populations (Tokuhama-Espinosa, 2001, p. 3). It therefore appears that
the term elite bilingualism is becoming less relevant, and more research on childhood
bilingualism is needed specifically in children who are growing up in more favourable
environments. It needs to be recognized that such form of bilingual development is no
longer elite, but is becoming common.

Another point is that most studies focus on linguistic productions in infant bilingualism,
while only few case studies on bilingual development focus on the language learning
environment and patterns of language presentation. Lanza (2004, p. 49) further argues that
‘context’ in which children acquire and use languages remains neglected in studies on
infant bilingualism.

In this thesis I adopt Lanza’s (2004, p. 10) perspective that an infant’s language acquisition
experience is influenced primarily by its language learning environment. Thus the best
8
indicators of how the infant’s two languages are likely to develop at the very early stages
of life would be by examining the input and language strategies that the parents use when
interacting with children.

Ervin-Tripp & Reyes (2005, p. 87) differentiate contexts of child bilingualism depending
on whether the parents are speakers of both languages themselves, or whether there is a
bilingual and a monolingual parent in a family. The language acquisition context examined
in case studies to date is mostly a situation with both parents speaking both languages (one
parent as a native language, the other as a second language), or a family where one of the
languages is not a native language of neither of the parents, but at least one parent is fluent
in a second language, which they use with the child, thus aiming for intentional
bilingualism (Štefánik, 2000, pp. 20-21).

Mixed-lingual families where only one parent is a speaker of both languages remain
underrepresented. Mixed-lingual families and their use of two languages in the home
render the family environment different also to a migrant family, where the minority
language tends to be the only language used at home. While in a minority family relatively
clear boundaries delimiting the use of the two languages can be set –usually the private
and the public domain, in mixed-lingual families this distinction becomes blurred. There
are still two languages, a minority and a majority language, but the domains for these
languages overlap considerably. This form of family bilingualism, with one bilingual and
one monolingual parent, is termed the asymmetric parent bilingualism (Chin &
Wigglesworth, 2007, p. 11). While the bilingual parent addresses the child primarily in the
minority language, the child is aware of the parent’s bilingualism and often overhears the
parents’ communication in the shared language.

Furthermore, the term ‘childhood bilingualism’ includes different ways of becoming


bilingual without taking the start of exposure to the two languages and chronological order
in which the languages were acquired into consideration. In general, children who are
exposed to two languages from birth are termed simultaneous bilinguals (also called
infant, incipient or childhood). Children whose exposure to the second language started
later than exposure to the first are sequential bilinguals, since they acquired first language
at home and a second language later on in a different setting, typically childcare or from a
different carer (Baker, 2001, p. 221). However, De Houwer (1990, p. 11) argued that this
distinction is somewhat imprecise since it does not satisfactorily describe the
circumstances of initial language exposure. She argues that even a short delay in exposure
to the second language could affect the way languages are processed in an individual, and
9
that an infant who was exposed to the second language at a later stage may develop
differently. Since full effects of such differences in exposure are not certain, it is
recommended to make this distinction obvious. De Houwer adopted an earlier term –
Bilingual First Language Acquisition (Meisel, 1989; Swain, 1976) – and established a new
field of study, which specified the onset of acquisition of the two languages more
precisely. The BFLA field examines simultaneous acquisition of two languages from birth,
while allowing for different exposure patterns and language presentations across families.
Since a BFLA child is acquiring two first languages from birth, labels such as mother
tongue, native language or first and second language become unsuitable. The languages
are labelled Language A and Language Alpha, in order to give both equal status. De
Houwer’s (1990, p. 12) strict criteria for BFLA comprise only those studies which clearly
specify that the subjects’ exposure to language Alpha started no later than a week after the
start of exposure to language A, and exposure to both was on regular basis until the time of
the study.

Thus BFLA is distinct from other early language acquisition contexts, namely:

 Monolingual First Language Acquisition (MFLA) - a context with one first language
(Language 1),
 Early Second Language Acquisition (ESLA) - a context with first language (L1), and a
second language (L2) that is introduced considerably later to the first language (De
Houwer, 2009a, p. 4).

According to De Houwer, BFLA as a field of study is an intersection between language


acquisition and bilingualism (De Houwer, 1998, p. 250). It focuses on the first 3 years
(infancy and toddlerhood) of linguistic development. Chronological description of research
carried out within the field of BFLA has been outlined in great detail in other studies
(Barron-Hauwaert, 2004; De Houwer, 1990, 2009a; Genesee, 2006). To date the focal
points in BFLA were the following phenomena:

 general linguistic developmental patterns


 one language system vs. two language systems
 language differentiation
 language choice
 comparisons with monolingual norms (Pearson, 1998, p. 349).

10
These phenomena were explored from phonological, lexical and morpho-syntactic
developmental perspectives. A detailed discussion of the focus of various studies is
presented in De Houwer (1998, 2009a) and McCardle & Hoff (2006).

2.1.1 Early vocal development and sound production


The earliest vocalizations produced by very young infants, cooing and other non-linguistic
sounds, are generally not seen as being language specific, and it is believed that all infants
across languages tend to produce similar sounds (De Houwer, 2009a, p. 171). No studies
to date have focused specifically on vocalizations by bilingual infants. However, one study
reports the bilingual subjects used intonation patterns which resembled the two languages
the infants were acquiring (Cruz-Ferreira, 2006, p. 63).

In the second half of the first year infants start producing canonical and reduplicated
babbling productions. No studies to date focused specifically on babbling in bilingual
infants. De Boysson-Bardies & Vihman (1991) carried out a cross-linguistic study in order
to examine whether babbling sounds are the same across languages or whether infants
exhibit language specific productions. The findings indicated that by 10 months infants
showed language specific tendencies in both vowels and consonants, which developed into
more marked differences further on. It was further concluded that the phonetic structure of
the infant’s environment is an important factor that influences the shapes of babbling
productions.

Another study compared developmental patterns of monolingual and bilingual infants, in


order to assess the possibility that acquiring two languages may require substantially
greater effort, and therefore cause delays in linguistic development and in achievement of
milestones in bilingual infants (Oller, Eilers, Urbano, & Cobo-Lewis, 1997). However, no
differences were found in monolingual and bilingual infants in the average age of
canonical babbling onset. In all normally developing children the age of the onset ranged
between 4 to 10 months of age, with no delays in bilingual infants.

Another research interest has been the link between babbling and early word productions.
In a study on babbling productions in hearing and deaf infants Pettito and Marentette
(1991, p. 1495) conclude that babbling is not dependent only on the motor development of
the articulatory mechanisms as previously thought, but that it is an expression of certain
amodal language capacity, and as such it is the ‘mechanism by which infants discover the
map between structure of language and the means of producing this structure’ (Petitto &
Marentette, 1991, p. 1495). The authors found that similarities in manual and vocal

11
babbling support the hypothesis of the existence of a unitary language capacity that is the
basis of language acquisition.

The above findings have important implications for bilingual infants. In their language
environments two phonetic structures will be found which could be presented either
separately or in the form of a blend, depending on the language presentation patterns in the
family. Thus in general, in families that adopt the 1 parent/1 language (1P/1L) approach
infants hear phonetic structures mostly in separate fashion, while in families with 1
parent/2 languages (1P/2L) approach the infant receives varied phonetic input from one
person. However, to date no BFLA studies report on these aspects of infant babbling.

2.1.2 Bilingual Comprehension


It is well understood that before the onset of speech children are able to comprehend words
in speech that is directed to them (Tomasello, 2003, p. 79). Word comprehension precedes
word production by several months. For example Cruz-Ferreira (2006, p. 146) observed
the first signs of comprehension as early as seven months of age, while production did not
start until several months later. However, the exact link between comprehension and
production is not fully understood, and there can be a large variation between the number
of words an infant comprehends and the number of words she can produce. It is believed
that comprehension and production are two different processes (De Houwer, 2009a, p.
209).

As De Houwer et al. (2006, p. 331) point out studies on comprehension focus mainly on
monolingual children, with only two diary studies on bilingual children which produced a
similar finding regarding comprehension preceding production in BFLA infants (Deuchar
& Quay, 2000, p. 55; Leopold, 1939). To date only one experimental study examined
comprehension in bilingual children in detail (De Houwer, et al., 2006). It examines
whether BFLA children are able to comprehend translation equivalents (TEs), cross-
linguistic synonyms with equivalent meaning, at the age of 13 months. Comprehension of
TEs was examined through parental and carer reports on the infant form of the MacArthur
Communicative Development Inventory (CDI). Carers reported on vocabulary items
understood by the children in their two languages, Dutch and French. The findings showed
a large individual variation. All children in the study understood at least some TEs, called
doublets, but the number of TEs they understood differed considerably. Likewise, all
children understood at least some words in one language only, called singlets, again with
large variation in the number of singlets. Differences were also found in the distribution of

12
doublet and singlet comprehension, with some children understanding only singlets
overall, or mainly in one of their languages, while other children understood relatively
same number of singlets in each language. Furthermore, the more meanings a child
understood, the more likely she was to understand two words for that meaning, one in each
language. According to the authors, the same pattern of different meanings in the two
languages was also found in studies on bilingual production. It was concluded that
bilingual children know more words than meanings.

An important bilingual perspective is how an infant comes to associate one concept with
two different words, one from each language. In monolingual and bilingual development,
Clark (1987, p. 13) proposed that in early stages of lexical acquisition children apply the
Principle of Contrast, a mechanism that guides word acquisition which allows children to
accept only one label for a category, and thus reject synonymy. However, the Principle of
Contrast did not hold up in studies on comprehension in bilingual development. De
Houwer et al. (2006, p. 343) demonstrated in their study that once a child acquired a
meaning, she was able to learn the meaning in the other language as well. Thus in
comprehension, the child did not link the meaning to one word form only and the Principle
of Contrast did not hold up. Indeed, Yow & Markman (2007, as cited in De Houwer 2009,
p. 206) compared the strategies used by bilingual and monolingual children to learn new
words. The findings suggested that monolingual infants learn by Mutual Exclusivity Bias,
in other words they assume that every class of objects can only have one name. However,
this strategy was not used so extensively by bilingual infants who appeared to rely on other
strategies, such as use of gestures that helped them interpret the meanings of unfamiliar
words.

The process of TE learning has not yet been explored in a detailed case study. However, it
can be inferred from the findings outlined above that a proportion of a BFLA child’s
receptive vocabulary would consist of TEs, since the child would be naturally driven to
discover the equivalents in both languages. This, of course, cannot be true for all of the
child’s vocabulary, since it is unlikely that she would experience exactly the same kinds of
contexts with both parents.

2.1.3 First words and early phonological development


Production of first words can occur anywhere between 8 and 13 months of age, within a
span of around 5 months. When infants start producing first words the forms they produce
are often unstable and different from adult forms. It depends largely on individual

13
differences in children’s articulatory abilities to realize individual sounds, to match their
productions with the sounds they are trying to produce, and at what age their speech
becomes clearer.

Moreover, early word productions often undergo different processes that make the words
appear less like adult forms. One such process is truncation, when children produce only
one or two syllables of a longer adult word. This process is dependent on the typical stress
patterns of the input language and children were found to use it with very early word
productions in the second year of life (De Houwer, 2009a, p. 172). However, it is not fully
understood whether bilingual children use different speech rhythms in their two languages.
In fact findings to date suggest contradictory evidence. On one hand BFLA infants were
found to show language specific use of prosodic features, while other evidence showed
influence of one language on the other (De Houwer, 2009a, p. 173).

Other processes apparent in early words are phonological. Most common of these are
deletion, substitution and reduplication of sounds, which result in simplifications of adult
word forms. Natural phonological processes are found in both bilingual and monolingual
children, however, it is not yet known whether they are language specific in bilingual
children, and whether the two languages influence each other (De Houwer, 2009a, p. 182).

Thus early words pose a problem in assigning the forms to a particular adult target word,
and can be determined only within context. It is also necessary to determine the language a
bilingual child was targeting in her production. According to De Houwer (2009a, p. 177)
bilingual infants’ first speech sound productions are not yet language specific and it is not
possible to determine whether they are dependent on one or two systems. Phonological
development in a bilingual child entails acquisition of two sets of target phonemes that
likely overlap to some degree. It is a gradual process believed to emerge in the second part
of the second year. When segmental organisation of the infant’s lexicon is beginning,
productions are governed by rules of the language and the child’s vocabulary increases
rapidly (de Boysson-Bardies & Vihman, 1991, p. 317).

Prior to phonological development ambiguities may arise in the production of first words.
For example one criterion suggested in Vihman & McCune (1994, p. 523) is the exclusion
of homonyms as first words in the early word productions. The authors argue that if a child
uses a sound form in conflicting contexts such as homonymy, the sound sequence cannot
be included as a word. However, this criterion could unnecessarily exclude legitimate first
words in BFLA infants or cognates in a TE pair, which appear as the same production in
form but differ in prosody. As explained above, no system of phones was developed as yet,
14
neither are the child’s articulatory abilities fully developed. Thus it is possible that a
bilingual child could be attempting to produce two separate words, each belonging to a
different language, with different referents, while the actual productions result in the same
form. Indeed one study found that some forms produced as monosyllables by the BFLA
child in the early stages resulted in cross-linguistic homophones which designated different
concepts (Deuchar & Quay, 2000, p. 53). While the child’s productions were constrained
by a limited phonetic inventory, their meanings could be interpreted contextually and the
different target adult forms identified.

Categorization of first words is equally difficult. First words tend to have a fluid meaning
and they are used in a holophrastic way, since a single word may subsume a larger
meaning, and on many occasions it does not have to be related to the meanings of the adult
forms. Often the meanings of first words undergo semantic processes of overextension or
underextension (De Houwer, 2009a, p. 216). Therefore first words are categorized
according to semantic rather than grammatical categories.

However, as Cruz-Ferreira (2006, p. 150) argues, it is equally not easy to determine the
most commonly used semantic categories of first words. While infants’ first words would
include semantic fields with words for their interests and concerns such as requests, labels
for people, objects of interest, foods, actions, greetings, interjections, and onomatopoeiae,
the first words are necessarily related to the physical contexts, which can vary across
children and across languages. Likewise, children’s own preferences in semantic fields can
vary, as Cruz-Ferreira (2006, p. 150) found in the data of three BFLA children in her
study.

Holowka et al. (2002, p. 246) concluded that monolingual and bilingual children’s first 50
words are organized into similar conceptual domains: objects (animate and inanimate) and
nonobjects (person related and object-related). In their study infants indeed showed
preference for their favourite things when acquiring new words. Therefore a new category,
termed person-related, was created which included words that were of personal interest to
the infants. All children seemed to show preferences for person-related words in first
productions, and this preference remained prevalent throughout the first 50 words period,
even with an increase in non-person related words.

There is also a great variance in the amount and types of first words infants use. For BFLA
infants the variance is even greater, since they may use different patterns across their two
languages (De Houwer, 2009a, p. 218). The rate at which children’s lexical repertoires
develop also varies considerably. In general, children are fast learners of new words, since
15
they need to be able to quickly fill lexical gaps in order to avoid communication
breakdowns. It is often sufficient for children to hear a word in one instance and
immediately use it themselves (De Houwer, 2009a, p. 221), a process termed ‘fast
mapping’. BFLA children have an additional need to fill lexical gaps in this way, since
gaps can occur only in one of their languages, especially if some of their interlocutors are
monolingual speakers. However, BFLA children are able to build on the knowledge they
already have in language A when learning a word in language Alpha.

Other ways of filling lexical gaps were found in BFLA children, such as creating new
words in ways that reflect the productive processes of their languages, for example
compounding or adding of bound morphemes to stems (Cruz-Ferreira, 2006, p. 174). In
BFLA children this processes may result in relatively rare bilingual blends, where words
contain syllables from two words from a different language (De Houwer, 2009a, p. 222).
However, only few bilingual case studies examined lexical development with focus on
one-word utterances and the meanings bilingual children create during this stage (Sinka &
Schelletter, 1998, p. 302).

2.1.4 Assessing bilingual vocabulary


Research on bilingual lexical development confirmed that the developmental pathway and
achievement of first word milestones in bilingual infants are the same as in monolingual
children (Holowka, et al., 2002, p. 240; Pearson, 1998, p. 360). Bilingual children produce
their first words around the same age ranges as monolingual children. Likewise, the 50
word milestones were achieved at around the same age range as in monolingual children.
This first vocabulary growth spurt of 50 words occurred approximately at around 18
months of age, comparable to monolingual children. According to Pearson (1998, p. 360),
the vocabulary spurt appears to occur in one language or in the two languages combined,
but not in both of the two languages at the same time.

However, researchers warn that norms based on monolingual infants’ lexical development
do not adequately represent normal bilingual development and care needs to be taken to
assess bilingual babies’ knowledge appropriately (Pearson, 1998, p. 349). A bilingual
child’s combined conceptual vocabulary needs to be assessed, considering both languages
together rather than each language separately. A bilingual child can have two lexical
representations for each conceptual meaning, one in each language, thus two labels are
linked to the same concept. Moreover, a bilingual child’s lexicon could consist of up to
three parts: two sets of words each belonging to one language (singlets), and a set of words

16
that the child knows in both languages (doublets). They can produce either doublets
consisting of translation equivalents, or just a singlet from a TE pair. Holowka et al. (2002,
p. 245) found that TEs accounted for approximately a third of bilingual children’s total
lexicons.

There may also be different time lags between the first productions of TEs in the two
languages, with no delay at all in some children and with considerable delay in others (De
Houwer, 2009a, p. 218). Bilingual children may also produce language specific words that
do not have a TE in the other language, as well as language neutral words that could
belong to either of their two languages (De Houwer, 2009a, p. 220). Language-specificity
of words can vary considerably depending on the language combination.

According to De Houwer (2009a, p. 228) BFLA children’s two languages do not tend to
develop at the same rate but with many individual differences. Firstly, the number of
words a particular child can produce in one language can differ from the number of words
in the other language at a given age and, secondly, the number of word types produced in
one language can vary from the other. However, a link between the word productions in
each language of bilingual toddlers was found (Conboy & Thal, 2006, p. 720). Likewise,
increases in word productions in one language were followed by an increase in the other
language.

2.1.5 Combinatorial speech and morphosyntactic development


When children reach the critical mass of 50 words in active vocabulary around 18 months
of age two-word combinations start appearing in their speech. Few studies examine two-
word combinations in bilingual children (Deuchar, 1999, p. 26; Lanza, 2004, p. 84).

The early syntax of two-word and multiple-word combinations does not reflect adult like
syntactic structures as yet (De Houwer, 2009a, p. 254). However, Deuchar & Quay (2000,
p. 81) found rudimentary syntax that relied on the juxtaposition of a predicate and an
argument to carry meaning. Word order of such combinations was variable and changes in
order had no effect on the meaning. Noun phrase combinations that consisted of two nouns
were also found, but such combinations were considered equivalent to one-word utterances
rather than showing syntax.

In early word combinations the use of bound morphology also emerges (De Houwer,
2009a, p. 254). However, morphological development in young bilingual children in the
earliest productive stages also received little attention in the literature. Most studies focus
on these aspects only after the emergence of multiple-word combinations.
17
According to Sinka & Schelletter (1998, p. 323) evidence for morphosyntactic
development in bilingual children is to a great extent dependent on the language
combination. The authors suggest BFLA children with languages with high inflectional
marking could show language specific use of morphology as early as the one word stage.
They state evidence for separate development of the two languages, and adherence to
language-specific rules can be examined only on those aspects of grammar that are
different between the languages. One of the findings was that the BFLA child was using
language-specific nominal morphology in both of her languages in the one- and two-word
stages, while person and number marking on verbs were observed from the two-word
stage. However, there was a lead-lag pattern evident since the child did not start using
morphological marking in both languages at the same time. Different timing in
development of these structures was taken as evidence that the two languages were
developing independently. Deuchar & Quay (2000, p. 84) also found language-specific
morphology in the two-word stage and considered this as evidence for two syntactic
systems. Other studies also support the finding that children are able to use verb inflections
correctly when two word combinations emerge (Meisel, 1989, p. 32).

BFLA children can combine words from the two languages in various ways. They can
produce unilingual combinations in each of their languages, as well as mix words from the
two languages. In mixed combinations a child may borrow single lexical items from the
other language, while the morphosyntax of the utterance remains language specific (De
Houwer, 1990, p. 102). Such mixing is taken as evidence for language-specific
morphosyntactic development. Deuchar & Quay (2000, p. 88) also argue that the use of
mixed combinations can be attributed to lexical gaps in one of the languages in the child’s
vocabulary rather than to a single lexical system.

2.2 Language differentiation


Language differentiation is an aspect of linguistic development distinctive of bilingual
language acquisition only. There are two main theories in the research literature. The first
is the unitary language system hypothesis which emerged in early diary studies on infant
bilingualism and suggested that as a bilingual child starts linguistic production she goes
through an initial stage of one fused linguistic system. She is not able to differentiate
between her two languages and only gradually acquires this ability first through lexical
differentiation, followed by syntactic differentiation. Evidence that was to support these
claims constituted examples of the subjects’ code-mixing, seeming lack of translation
equivalents in some cases, and perceived delays in overall linguistic development
18
(Arnberg, 1987; Leopold, 1939; Redlinger & Park, 1979; Vihman, 1985; Volterra &
Taeschner, 2007) 1.

More recent research proposed an alternative stance – the separate development


hypothesis. Several case-studies confirmed that children growing up with two languages
demonstrate early differentiation and separate development of their two languages (Cruz-
Ferreira, 2006; De Houwer, 1990; Deuchar & Quay, 2000; Genesee, 2007). Differentiation
is evident even in contexts that may encourage more mixing, such as when both parents,
who normally use different languages with the child, are present and interacting with a
child in the same situation (Genesee, et al., 1995, p. 627) . Childhood mixing was found to
be a result of sociolinguistic factors rather than confusion, and it can also be related to
language dominance, since bilingual children often code-mix to fill lexical gaps in one of
their languages (Genesee, et al., 1995, p. 614; Nicoladis & Genesee, 1996, p. 443). Yet
other research showed that bilingual children use translation equivalents (TEs) very early
in development, which needs to be taken as evidence for early sensitivity to two input
languages (Cruz-Ferreira, 2006; Holowka, et al., 2002; Pearson, 1998; Petitto, et al., 2001).
Studies that examined these issues empirically with a larger group of children also proved
that early infant bilingualism does not pose any disadvantages and developmental delays
when compared to monolingual developmental norms (Petitto, et al., 2001, p. 490).

Hence there is evidence for early sensitivity to two input languages in BFLA children in
phonetic perception, lexical, syntactic, morphological, as well as pragmatic developments.
This evidence suggests that BFLA children do in fact acquire their two languages in a
separate, side-by-side fashion.

However, some studies point out that while we have evidence for early differentiation, we
do not understand how and when children come to understand that there are two input
languages in their environment (Nicoladis, 1998, p. 113). Little is known about
differentiation in comprehension and in the use of prosody in very early, pre-linguistic
stages. Similarly, while children have been observed to show signs of pragmatic
differentiation around the age of 2 years, only few studies examined pragmatic
differentiation (also called pragmatic sensitivity, interlocutor sensitivity or context
sensitivity) at an earlier age (Nicoladis & Genesee, 1996, p. 441). Nicoladis (1998, p. 106)
argues that pragmatic differentiation may develop differently to other types of
differentiation. She stresses the need to consider emergence of sensitivity to the existence

1
Although according to De Houwer’s (2009) criteria since the child in Vihman (1985) was not exposed to a
systematic input in one of her languages until 6 months of age it is not a case of BFLA but ESLA.
19
of two languages, as well as sensitivity to interlocutors’ language preference and dominant
language use, rather than focusing on when bilingual children start using their two
languages in a monolingual fashion. It is argued that bilingual language use does not mean
adherence to monolingual behaviour in two languages, rather pragmatic differentiation is
defined as the ability to use language appropriately in different pragmatic contexts.
Nicoladis (1998, p. 106) further reasons that in the early stages of linguistic development
(up to the 50 words stage) the presence of translation equivalents alone is not sufficient
evidence for emerging pragmatic differentiation, because TEs may be used more like
synonyms. Instead, it should be examined if children use their TEs with speakers of their
two languages appropriately within context.

Overall, studies to date seem to suggest that appropriate language choice according to
interlocutor emerges around 2 years of age and that there is an earlier stage in bilingual
children when pragmatic differentiation is not seen despite the presence of TEs (Nicoladis
& Genesee, 1996, p. 460). Nicoladis (1998, p. 107) hypothesized that when bilingual
children reach the 50 words stage, their prosodic systems start showing language specific
productions, which could lead to an understanding that the TEs in their lexicon are in fact
cross-language equivalents. This may then lead to the understanding of pragmatic use of
language in their environment. Children may be using the presence of TEs in their lexicons
as a cue in learning the particular language use patterns in their environments, thus
learning to use them pragmatically.

However, after a detailed analysis of one child’s language use, the results in Nicoladis’
(1998, p. 113) study showed the reverse pattern. The child was first using his two
languages appropriately when there was only a very low number of TEs in his lexicon.
Soon after the child showed understanding of two pragmatic contexts, the number of TEs
in his lexicon grew considerably. Thus Nicoladis (1998, p. 113) suggests that it is possible
children need to first come to an understanding that there are two input languages, and
then build on pragmatic differentiation abilities to learn how to use TEs in different
pragmatic contexts.

Petitto et al. (2001, p. 459) asked an important question in order to seek further empirical
evidence for the separate development hypothesis: ‘What underlies early bilingual
language acquisition?’ The study focused on achievements of early linguistic milestones,
and patterns of language use and mixing when interacting with parents in 6 hearing BFLA
children. When the results were compared to monolingual developmental milestones,
bilingual subjects achieved their milestones at the same ages and in the same
20
developmental patterns in each of their languages, thus there were no delays in
development. The results of the study confirmed that very young bilingual children
differentiate their languages at a very early stage of linguistic development, as early as the
production of first words. Findings showed that while children did mix their languages, the
rate of mixing was a reflection of the mixing patterns used by their parents rather than a
sign of confusion2. More importantly, children were able to demonstrate early sensitivity
to their interlocutor’s linguistic abilities by adjusting their language use. Another finding
that emerged was that bilingual children showed preference to one of their languages early
on. The preference was influenced by the child’s sociolinguistic environment and
corresponded to the language of the person who had the most contact with the child, and to
whom the child had developed the strongest attachment. Language preference could
change overtime with changes in the sociolinguistic environment. The authors conclude
that since bilingual children appear to be differentiating their two languages as early as the
first word stage, therefore they must already possess a representation of the two input
languages from very early stages of language acquisition, starting at birth.

Lanza (1992, p. 654) goes even further and questions the relevance of unitary vs. dual
language system issue in BFLA children’s mixing, and argues differentiation is a reflection
of language socialization, a process via which children learn to use their languages in
socially appropriate ways. She suggested a child needs to learn to navigate in different
language contexts and learn when it is appropriate to mix her languages as opposed to
situations when it is appropriate to keep them separate. The notion of bilingual awareness
was proposed (Lanza, 1992, p. 653). There are different levels of awareness, and differing
degrees of bilingual awareness are reflected in varied uses of code-switching in bilingual
children. Lanza (1992, p. 653) claims that in very young bilinguals (before the age of 3)
the participant is the main determinant in mixing, while topic and setting are secondary.
But even very young bilinguals at the age of two are able to use mixing in contextually
sensitive ways.

Vihman (2002) approached the unitary versus separate development hypothesis debate by
proposing an entirely different course of development. It was suggested that a bilingual
child has no linguistic system as such in the first months. Bilingual input provides the child
with phonetic patterns of two languages, and from these, supported by implicit learning
and speech perception capacities in the early months, the child starts paying attention to

2
Although another study found no evidence for a systematic relationship between child mixing and parental
mixing patterns, and suggests a threshold level of mixing is necessary to have an impact on children’s mixing
(Genesee, et al., 1995) .
21
the learning of first words. Such explicit learning allows reproduction of retained first
words and phrases in familiar contexts. According to Vihman (2002) the early words are
holistic, used in situational contexts and associated with the child’s interests, since they
were learnt item by item. Their phonetic shape reflects the target words with only few
changes, mainly omissions.

Once a child accumulates 50-100 words in expressive vocabulary, the phonological system
emerges. With increasing word productions first consistent phonological patterns appear
which Vihman (2002, p. 242) calls word templates. The child’s early word forms reflect
such patterns, and therefore become more assimilated to the templates (with more changes
in segments, sequence, etc.) in comparison to the first item-learnt words and adult forms.
They are thus less accurate as far as the target words go. The templates differ from child to
child and more general structures as well as subpatterns were identified (Vihman, 2002, p.
243). The two languages interact in the patterns, while at the same time the templates
reflect the phonology of the two languages. The child now starts forming representations
of particular words and starts generalizing across contexts of use.

However, Vihman’s (2002, p. 250) analysis does not extend past the first words stage and
therefore it is not explained how the child may arrive to a more adult like phonological
system. It is only suggested that within a year of the first word production a child’s
productions should reflect adult forms, except for difficult segments and clusters. The
child’s growing dual lexicon is taken to allow for further differentiation of the
phonological system as well.

As MacWhinney (2001, p. 257) points out, to date language differentiation in infancy and
early linguistic development such as the one-word stage and combinatorial speech stage
remains underexplored.

2.3 Language mixing


Researchers use the term language mixing as an umbrella term for any interaction of the
two languages in a child’s speech. They do not discriminate between switching and mixing
in bilingual children. In adult bilingualism, however, a distinction is made. Switching is
often assumed to be possible only in adults who possess the necessary pragmatic
knowledge and grammar constraints under which switching can operate. Mixing, on the
other hand, is perceived to violate such constraints, both sociolinguistic and grammatical.

According to Lanza (1992, p. 645) in the earlier literature lexical mixing is perceived as
characteristic of mature bilingual code-switching, while grammatical mixing is a reflection
22
of developing bilingualism in very young children. Bilingual infants were not considered
to have the ability to use languages in contextually sensitive ways, and therefore unable to
code-switch, hence the use of the term language mixing instead. It refers to two possible
phenomena:

 the use of lexical or morphological elements from both languages in one utterance
 the use of whole utterances in the ‘wrong’ language with an interlocutor who does not
usually address the child in that language.

Genesee et al. (1995, p. 623) referred to these two different types of mixing as intra-
utterance mixing (use of two languages in one utterance) and inter-utterance mixing
(mixing across utterances with the same speaker). Inter-utterance mixing is especially
important in young children in the one-word stage, since they do not yet produce
multiword utterances with intra-utterance mixing. However, they can use inter-utterance
mixing in their choices of language in the same conversation with a particular speaker
(Genesee, et al., 1995, p. 626).

Language mixing is a central issue in language differentiation theories in childhood


bilingualism. Recent research produced evidence that mixing in very young BFLA
children is not a sign of confusion, but reflection of sociolinguistic factors. Several
hypotheses for mixing were proposed:

 the lexical gaps hypothesis (Genesee, et al., 1995; Nicoladis & Secco, 2000; Quay,
1992)
 the modelling hypothesis (Goodz, 1989)
 child’s dominance in one language (Lanza, 1992).

In the earliest stages of linguistic development (before age 2) lexical gaps and lack of
translation equivalents in one of the languages are proposed as reasons for child mixing.
Nicoladis & Secco (2000, p. 5) found that while children may have pragmatic sensitivity to
two languages early on, due to their limited linguistic resources they may not be able to
show signs of pragmatic differentiation until later age. Evidence is found in observations
that while a child may try to use the language of their interlocutor most of the time, their
linguistic proficiency is limited and to maintain interaction they need to use whatever
resources available, which means borrowing words from the other language. Once they
acquire the right lexical resources, they can use the language of the context consistently.
Thus mixing is explained as a form of overextension of the pragmatic context (Nicoladis &
Secco, 2000, p. 24).

23
The modelling hypothesis explains child’s mixing as a reflection of the level of parental
mixing in a child’s input (Comeau, Genesee, & Lapaquette, 2003, p. 114). While there is
consensus that in general, older bilingual children adopt code-mixing patterns typical of
their speech communities, it has been questioned whether young bilingual children have
the ability to code-mix pragmatically. The assumption is that due to developing cognitive
capacities, they may not have the necessary sensitivity to mixing by their interlocutors, and
therefore may not be able to adjust mixing accordingly. However, Comeau et al. (2003, p.
124) found that young bilingual children at the one-word stage do appear to have the
necessary pragmatic sensitivity to input, and that they do adjust their rates of mixing to
interlocutors’ mixing patterns in a turn-by turn fashion.

Modelling hypothesis was also supported in Goodz’s (1989) study of parental input in
bilingual families. It found that parents did mix languages when addressing bilingual
children in the earlier stages of one, two and multiple-word utterances, even if the rates of
mixing were rather low (Goodz, 1989, p. 42). Moreover, even those parents who assessed
their own language use as a strict separation and adherence to the 1P/1L approach were
found to mix languages when addressing their children. Children’s and parental mixing
was closely related, in that parents would tend to switch in response to a child’s switch to
the other language. Explanation was sought in the nature of the interaction in parent-child
dyads. According to Goodz (1989, p. 41) parents tended to acknowledge child’s utterances
by recasting or expanding on the utterance, using the borrowed word or utterance
themselves, rather than providing a translation or requesting the child to repair the ‘wrong’
language choice. Thus the parents maintained the switch to the other language. Goodz
(1989, p. 42) hypothesised that in the early stages of language acquisition parents focus on
maintenance of communication flow rather than the form. They do not pay attention to the
word or utterance borrowed from the other language, but focus on the content and the
meaning the child was trying to convey, maintaining the child’s form. In doing so, they
essentially provide models of language mixing.

An aspect of language mixing which has not been analysed in much detail is situational
switching according to addressee (Ervin-Tripp & Reyes, 2005, p. 91). In a BFLA setting,
this form of switching accounts for the child’s need to switch languages by addressee in a
triadic interaction. According to the authors “evidence of situational switching reflects
what is salient to young children as they develop: people, place, activity, and genre.”
(2005, p. 92). Interlocutor was identified as the most important determinant of language

24
choice, and physical setting the second. Activity was also noted as further determinant
(Kwan-Terry, 1992, p. 248).

Thus in general, bilingual infants under the age of two are thought to go through a first
stage of mixing, during which they lack the necessary grammatical knowledge and violate
rules of code-switching. During the third and fourth year, young bilingual children are
believed to start developing the necessary grammatical as well as pragmatic knowledge
that will allow them to start switching in a manner similar to adults (Lanvers, 2001, p.
438). However, according to Lanvers (2001, p. 438) little is understood about how young
bilinguals may develop these abilities, apart from evidence that determinants of switching
in very young infants are the language of the child’s interlocutor and the child’s
competence. Lanvers (2001, p. 460) attempts to fill this gap by suggesting developmental
trends in switching. She identified instances of intentional switching before the age of 2,
particularly switching for emphasis and appeal, crutching and harmonization, as well as
sensitivity to the interlocutor’s language preference and switching on-demand in response
to a parental discourse strategy. Many forms that the author identified did reflect the same
properties of adult switching. However, the particular child’s competence and “psycho-
socially different interactional mode” likewise played an important role in shaping the
functions of their switching (Lanvers, 2001, p. 461).

2.4 Language choice


Language choice is defined as accommodating use of a language in a conversation
according to an interlocutor’s preferences, which assumes certain pragmatic knowledge. In
childhood bilingualism literature the child’s ability to choose an appropriate language in a
specific context is taken as a reflection of the child’s pragmatic language differentiation.

Research literature on language choice in bilingual children suggests this ability appears
before and around the second year of life, with most studies focusing on production of
multi-word utterances (Genesee, et al., 1995; Lanza, 1992; Nicoladis & Genesee, 1996;
Vihman, 1985). However, studies which examined single word utterances also
demonstrated that children can make language choices with sensitivity towards the
speaker’s language preference as early as the one word stage (around 1;7) (Deuchar &
Quay, 1999, 2000; Quay, 1992).

Several determinants of language choice were found. According to Deuchar & Quay’s
(2000, p. 113) findings in the early stages of productive language, when the child was
building up her vocabulary, failure to make appropriate choice was most likely due to

25
lexical gaps in one of the languages, thus lack of translation equivalents rather than lack of
differentiation. The child’s interlocutor and the language used to address the child was
most important, which reflects Leopold’s suggestion that consistent parental language
choice promotes the child’s ability to make an appropriate language choice within a social
context (Leopold, 1939). Deuchar & Quay (1999, p. 474) also found setting or location as
determinants of language choice. While the child in their study was mostly making an
appropriate choice in both language contexts, she was more likely to use more Spanish in
an English context at home (in conversations with English monolingual), which was
otherwise a Spanish only environment. However, what promoted appropriate language
choice was the style of interaction between parent and child - the discourse strategies a
parent used in reactions to a child’s utterance in the ‘wrong’ language, and whether the
parent accepted utterances in ‘wrong’ language.

When a bilingual child uses a language that is not the preferred language of the
interlocutor, it is likely that a breakdown in communication will follow, especially with a
monolingual interlocutor, or a bilingual interlocutor who chooses to strictly separate the
two languages, and does not accept the child’s utterances in their less preferred language.
Thus when a language choice in a bilingual child causes a breakdown in communication,
the child needs to acquire additional communicative competence. She needs to identify the
reason for the breakdown as inappropriate language choice rather than any other reason
(e.g. inaudible or unintelligible utterances). She also needs to repair such breakdowns.

Breakdowns in communication can be indicated by an interlocutor who makes a request


for clarification, or it can be noted by the child herself. Indeed one study found bilingual
children aged from 2 to 3 years were able to use both strategies appropriately (Comeau,
Genesee, & Mendelson, 2007, p. 172). If an interlocutor requested a clarification following
an inappropriate language choice by the child, most children in the study were able to
identify these requests as requiring language choice repair, even if the request was not
specifying a need to change language. These children never used the language change
strategy if the breakdown in communication was due to other reasons. They were also able
to self-repair their inappropriate language choices before the interlocutor indicated a
breakdown in communication.

2.5 Parental discourse strategies


Lanza (1992, p. 635) suggests a different approach to language mixing and language
choice in BFLA. Rather than considering mixing itself, researchers should focus on when

26
the child uses mixing, whether it is used in contextually appropriate ways, and what
influences mixing. The question should be whether a child has the pragmatic ability to mix
when it is appropriate to do so, and to separate the languages if the social context requires
separation (Lanza, 1992, p. 653). In a detailed analysis of the contextual factors Lanza
(1992, p. 644) found that the different levels of grammatical and lexical mixing by the
subject were related to the conversational discourse strategies used by the subject’s parents
(the family used the 1P/1L approach). She identified 5 discourse strategies. Each strategy
negotiated different language context with the child, either in an implicit or a more explicit
way. Lanza (1992, p. 649) situates these on a continuum from monolingual to bilingual
contexts:

 Minimal grasp
 Expressed guess
 Adult repetition
 Move on Strategy
 Code-switching.

If parents indicate to the child that they did not understand and expect a clarification of the
child’s utterance, they negotiate a monolingual context which does not allow mixing, and
thus a monolingual identity. On the other hand, if they continue in the conversation by
repeating the utterance in the other language, simply moving on or switching to the other
language completely, they negotiate a bilingual context and allow mixing in the
conversation with the child. They negotiate a bilingual identity. However, as Lanza (1992,
p. 649) points out, such identities cannot be mutually exclusive.

Lanza (1992, p. 646) also analysed the child’s reaction to a parental discourse strategy, as
a reflection of the context each parent tries to negotiate. The findings indicated that the
child was able to adjust her language use accordingly – she continued to use mixing with
the parent who negotiated bilingual context but repaired her mixed utterances with the
parent who negotiated monolingual context more explicitly. Thus the child was using her
languages in pragmatically competent ways, demonstrating sensitivity to the requirements
of the context.

Using Lanza’s (1992) categories as the basis, Hiroko (1998, p. 333) identified 6 different
strategies a parent can use, and classified them as explicit, implicit and code-switch
strategies. The study then analysed parental feedback after a child made a ‘wrong’
language choice, and the child’s language choice immediately after a parent used a

27
discourse strategy. The results showed that using one of the explicit strategies was more
successful in influencing the child’s choice to choose the parent’s language.

Juan-Garau & Pérez-Vidal (2001, p. 83) also found that their subject tended to repair
language mixes with the parent who used monolingual strategies. On the other hand, he
tended not to repair his language mixing after a request for clarification made by the parent
who used the bilingual strategies, presumably since he did not interpret these as requests to
switch language. Some parental discourse strategies were more conducive to maintaining a
certain level of proficiency in the minority language, and to develop the child’s bilingual
awareness (Juan-Garau & Pérez-Vidal, 2001, p. 77). The repetition strategy was
particularly successful because it provided the child with the missing information in the
other language, which was in many cases the reason for mixing. Thus it was avoiding
communication breakdown, which could occur in cases where the parent used the minimal
grasp strategy. Moreover, the repetition strategy invited the child to reproduce the
utterance in the target language. The expressed guess strategy on the other hand, required
only a yes/no answer, not necessarily eliciting the child’s reproduction of the utterance.

Nicoladis & Secco (2000) found considerably different strategies used by parents in
response to a child’s mixing. They found two trends in parental mixing in the input.
Firstly, much of the parental code-mixing was accounted for by use of words that the child
knew in the other language, and was actively producing (Nicoladis & Secco, 2000, p. 25).
Thus the parents chose to borrow some words from their non-native language because they
knew the child was able to use them. A possible interpretation of this strategy is that the
parents may be trying to make the meaning clear to the child. The second trend was that
the parents often used a word borrowed by the child from the other language as an
opportunity to teach the translation equivalent in their native language.

An area that remains relatively unexplored is triadic interaction between the parents and a
child, when each parent addresses the child in their respective languages, but the parents
use only the shared language to communicate (Lanza, 2001a, p. 222). Lanza (2004, p. 301)
points out such situations place different communicative demands on the child. They also
emphasize bilingual behaviour, since the bilingual parent negotiates maintenance of
monolingual conversation with the child in one language, and a monolingual conversation
with the other parent in the other language, but overall models bilingual identity. Triadic
contexts were analysed in more detail in a case of trilingual development (Quay, 2008).

28
2.6 Theory guiding this thesis
This thesis explored and described the course of bilingual first language acquisition
through a holistic view of bilingualism:

“Because bilinguals, like monolinguals, have an innate capacity for language and are, by
essence, communicators, they will develop competence in each of their languages to the
extent needed by the environment…but they will always maintain a necessary level of
communicative competence (Grosjean, 2008, p. 16).”

Thus bilinguals were not viewed as two monolinguals in one. The development of
bilingualism was approached and studied with bilingual expectations of language
acquisition.

The design of the case study was data driven. An exploratory ethnographic approach, as
suggested by Cruz-Ferreira (2006, p. 5), was suitable for data collection and descriptive
analysis, since it allowed the data to dictate the findings of the study, rather than grappling
to fit existing analytical categories of a particular framework onto the data. In the process
of literature review no single framework was found that adequately reflected the findings
in the data. Instead a combination of certain theoretical aspects derived from existing
frameworks appeared to describe most appropriately the process of becoming a bilingual
within the particular language learning environment. Language is not a static phenomenon,
which is especially relevant in the early years of language acquisition, whether bilingual or
monolingual. Therefore it should not be restricted to and studied under one perspective
alone.

The study of childhood bilingualism in this thesis was guided by the principles of BFLA as
described in De Houwer (2009a), Deuchar & Quay (2000) and Cruz-Ferreira (2006). The
theory on early language development and typical milestones in BFLA children were
drawn on De Houwer (2009a).

This thesis also operates with developmental stages found in models of initial language
acquisition by Oksaar (1983), de Boysson-Bardies (2001), Halliday (1975) and Tomasello
(2003). Descriptions of linguistic developments in these works are based on findings
across languages and thus avoid purely monolingual norms. Moreover, the developmental
progression which emerged from my data reflected the stages proposed in these works, and
thus allowed for logical organisation of the data. Specific developmental patterns in Slovak
speaking children are drawn on Horňáková et al. (2005), and in Slovak-English bilingual
children on Štefánik (2000).

29
The methods for this thesis are based on several book-length multilingual case studies
where the children under study fell in the BFLA category, and in which the authors
collected naturalistic data in the child’s home, as shown in Table 2.1. However, in some
studies later age ranges were studied.

Table 2.1 BFLA Case Studies

Parental
Child’s age Data
Source Research focus Languages language
range source
competence
Morphosyntax,
De Houwer separate Kate Dutch- Both Audio/
1990 development 2;7-3;4 English bilingual Diary
hypothesis
Phonology,
Diary/
Deuchar & lexicon, Manuela Spanish- Both
Video/
Quay 2000 morphosyntax, 1;8-2;3 English bilingual
Audio
language choice
Karin
1;1- teen
Cruz- Lexical Portuguese- Diary/
Sofia Both
Ferreira development, Swedish- Audio/
3;1- teen trilingual
2006 morphosyntax English Video
Mikael
3;1-teen
Siri
Language
1;11-2;8 English- Both Audio/
Lanza 2004 mixing,
Tomas Norwegian bilingual Diary
language choice
1;9-2;8
Lexical
development, Diary/
Štefánik Natalie Slovak- Both
morphosyntax, Unspecified
2000 1;3-5;7 English bilingual
Language recordings
mixing

The difference in this thesis is that the family is mixed-lingual, while one of the parents is
bilingual in Slovak and English and the other parent is monolingual in English. In a mixed-
lingual family a specific aspect of developing bilingualism emerges as central, the learning
of appropriate language choice in triadic interactions as described in Lanza (2004, p. 292).
This thesis focuses on the course of BFLA under the particular circumstances resulting
from the mixed-lingual family’s language learning environment, and verifies the relevance
of current BFLA theories to the data.

A further difference stems from the fact that the above case studies are concerned with
bilingual children’s linguistic production, while only a few studies focus on the language
environment, input and the child’s developing communication abilities as she is learning to

30
make sense of her two languages in the first two years of life. In first language acquisition
a child’s language learning environment is crucial since it provides the basis for language
acquisition, that which is to be learnt. As such, the language learning environment is both
the source of language(s), the tool for communication and expression, but most
importantly, the tool for social contact, the purpose for using language (Hoffmann, 1991,
p. 35). As Hoffmann points out, a child learns not only the language itself and its formal
aspects, but also the importance of language in forming relationships with her
environment. Thus it is the child’s emotional engagement with the environment that drives
language acquisition.

For the first few months the child’s emotional engagement is with the important people,
while later it extends to the immediate contexts including favourite objects, activities, and
routines. Even later, emotional engagement can extend to the wider environment outside
the family. But most importantly, the infant’s contexts stem from the interpersonal
communication in which she participates with the familiar people in numerous reciprocal
interactions. Initially, the infant is interested primarily in communicating with people and
responds to stimuli addressed to her by people over any other outside stimuli (de Boysson-
Bardies, 2001, p. 38). Even at a later stage when the child becomes interested in other
aspects of her environment interactions with people remain most important. Thus linguistic
development is embedded in emotional engagement with people. This theoretical basis
stems from the functional or usage-based language acquisition theories (Halliday, 1975;
Tomasello, 2003) which consider acquisition of a child’s initial languages as context
based.

Research on childhood development demonstrated that the key factor in the development
of emotional and social intelligence is face-to-face social interaction with other human
beings (Grille, 2008, pp. 10-12; Sunderland, 2006). Language acquisition, interlinked with
other developments, necessarily develops under the same conditions. During the first 18
months the infant is learning to communicate with the people in her environment and
participates in meaningful communication before the onset of speech. The use of symbolic
gestures indicates the onset of intention-reading communication, and the ability to
communicate by means of symbolic representation is believed to aid in development of
verbal communication (Goodwyn, Acredolo, & Brown, 2000, p. 84; Tomasello, 2003, p.
35). Child’s early language is meaningfully supported by certain social-communicative
interactions, or routines, constructed by the carers and the child (Horňáková, et al., 2005,
p. 19; Tamis-Lemonda, Bornstein, & Baumwell, 2001, p. 763).

31
Studies show that the first important clue in the development of language is the amount of
child directed speech (Huttenlocher, Haight, Bryk, Seltzer, & Lyons, 1991, p. 245).
Secondly, the topic of the parent-child interaction is important, however, the degree to
which the infant controls the topic is crucial in the extent to which the infant’s interests are
followed-in by the adult in interactions (Goodwyn, et al., 2000, p. 99). Regular child-led
interactions allow the child to derive meaning from the context, since it is easier for her to
determine the intended referent of the interaction (Tomasello & Farrar, 1986, p. 1460).
Thus the joint attentional focus between the carer and the child is stressed as more
important than directiveness by the parent to gain the child’s attention. Tomasello & Farrar
(1986, p. 1462) suggested that if the adult is able to determine the child’s focus and then
follow the child’s lead, language acquisition is facilitated, since the child does not have to
coordinate her attention with the adult’s attention. Language acquisition can thus take
place when “the child is attentive, motivated and best able to determine the meaning of her
mother’s language” (Tomasello & Farrar, 1986, p. 1462). During interactions with joint
attention the conversations were found to be longer and the mother’s language was
simplified with shorter sentences, while it included more comments. Language
development following such interactions was related to the types of object references
directly from the context of the interactions. Finally, the child’s interest influenced the
mother’s actual input in cases where the mother tuned into the child’s attention and
followed in the conversations, which in turn led to more learning.

The socializing aspect of first language acquisition was described as paramount by


Halliday (1975), and it was developed into the interpersonal first principle by Painter
(1996, 2006). Acquisition of the native language is interpreted as ‘a process of interaction
between the child and other human beings’(Halliday, 1975, pp. 5-6). He further explains
that the infant’s social world has only a limited number of contexts in which the child can
use symbolic expressions to communicate. The child brings certain abilities to the
acquisition context in the form of ‘biological predisposition to attend and respond to
communicative behaviour addressed to him’ (Painter, 1985, p. 49) , but this predisposition
alone is not enough to acquire language. A child needs constant interaction with adults, a
model of the language on which, together with the biological predisposition, the ability to
communicate by means of language is developed.

From birth the parent, ‘mirrors her baby’s vocal and gestural behaviours’ (Painter, 1985,
p. 49), and then moves on to other forms of game playing. Mothers interpret babies’
vocalizations and thus assign some kind of meaning to them. Through these interpretations

32
the mother helps the child ‘to experience shared states of feeling’, and ‘brings the child to
meaningful action by her use of the language’ (Painter, 1985, p. 49). The parent’s
interpretations of the child’s behaviours further add to the interactional experiences of the
child.

According to Painter (1985, p. 49) at around 8-9 months (about 40 weeks) an important
development of brain function occurs, when the child has a new way of accepting people –
she starts reacting to people on a voluntary basis and ‘combining her interest in the
environment of objects with acts of communication addressed to persons’ (Painter, 1985,
p. 49). Previously these were expressed by the child as two separate interests. This
development was observed as an alternation of eye-contact between objects and people
while they vocalize, and by giving or showing objects to people. The child also realizes at
about the same time that her mother and she is not one person, but that she is her own
person.

These two developments that lead meaning making by the child are what Halliday (1975)
captured in a functional framework. According to Painter (1985, p. 50) Halliday’s
functional system accounts for the distinction between person and thing, and between self
and other. A number of interactional capacities are built into the functional framework:
achieving objects by verbal control of a person, interaction with a person as mediated by
objects, expression of interest in the environment, and non-mediated address to persons.

There appears to be a link between the interpersonal first principle and joint attention in
language acquisition as suggested by Tomasello & Farrar (1986). In joint attention it is the
mutual focus, ‘the interpersonal’ in Painter’s terminology, that allows the parent to follow
the child’s interest, and therefore feeds into what is ‘learnable’ for the child through her
experience and retention of knowledge. In this case what the child learns is also
emotionally charged. Thus language acquisition is mediated by the people and physical
environment, but it is dependent on the manner in which the environment interacts with the
child and vice versa. In language socialization the amount and quality of input are
important (Huttenlocher, et al., 1991, p. 245).

In bilingual families, where children learn their two languages from birth and each
language from a different parent, bilingual socialization is mediated through language
exposure patterns, parental language choices and discourse strategies. This is particularly
important for the development of the minority language. Studies to date accorded little
focus to the influence of linguistic environment on bilingual development. This thesis

33
offers further understanding of the role language learning environment plays in the course
of BFLA.

34
3 Methodology and study design
In this chapter the study design, procedures of data collection and data analysis will be
described.

This thesis portrays linguistic development of a child called Ria growing up as a bilingual.
The study was of an ethnographic format, and described the process of bilingual first
language acquisition (BFLA) in the early years (birth-2). Ethnographic longitudinal case
study was suitable for several reasons:

 Ria’s language learning environment could be examined allowing for identification of


factors that had an effect on linguistic development
 observations were carried out in naturalistic contexts, using naturalistic language
 developments could be better understood since they were observed over a longer
period of time
 multiple sources of data, such as audio-video recordings and diary notes and reflections
from varied times and settings were available.

There are limitations to a descriptive case study. It can be argued a case study captures the
developments of one subject only and no group generalizations can be made. Further, there
is a risk of ‘observer’s paradox’ (Lanza, 2008, p. 76), meaning that the presence of the
observer can have an influence on the participants’ behaviour, rendering the context less
naturalistic. Moreover, a parent conducting research on her own child’s language use may
convey a certain bias to the analysis.

However, a number of authors (Cruz-Ferreira, 2006, p. 46; Deuchar & Quay, 2000, p. 2)
have argued that despite its limitations a case study is in fact the most appropriate research
method in early linguistic development. Firstly, while the findings cannot be used to make
new generalizations, they can be used to refute previous generalizations or to suggest new
areas of research that need to be explored further in a bigger scale study involving more
subjects (Hua & David, 2008, p. 99). The criticism of observer’s paradox can also be
challenged, since in this thesis it was a ‘parent-researcher’, Ria’s mother, undertaking the
observations. Since a mother and a child are already familiar with each other, the need to
establish a relationship with the participants is eliminated. This allowed ready access to
Ria’s familiar contexts for observation of naturalistic data and minimized the ‘observer’s
paradox’. Moreover, as a ‘parent-researcher’ access to naturalistic data in more varied
situations was available, as opposed to a researcher obtaining limited data recorded in

35
prearranged settings. According to Deuchar & Quay (2000, p. 3), a researcher’s familiarity
with the child’s speech is in fact advantageous for correct interpretation of developing
language, since in early words the conventional forms a child targets are phonologically
modified.

3.1 Data collection


There were two sources of data: audio-video recordings and diary records kept by the
mother-researcher. As Lanvers (2001, p. 444) points out, these two data collection
methods complemented each other. However, some instances of Ria’s developing
languages were missed, since the diary could not capture distributional usage and the
recordings by no means gave an overview of the overall competence.

3.1.1 Diary records


The diary records consisted of notes and observations made throughout the period of the
study. The records were made whenever the mother was with Ria, which in the family’s
particular case and carer arrangements was the majority of Ria’s daily awake time. The
diary records started while the researcher was expecting Ria and continued until the end of
the study.

During the pregnancy diary records were kept on monthly bases. During the first 2 months
after Ria’s birth sporadic observations of the communication between Ria, her parents and
other close relatives were recorded. From approximately 0;2 months of age observations
and developments were recorded at least once a week. However, if necessary, notes were
made more often or on daily basis. Observations of Ria’s language, physical and cognitive
developments were recorded, as well as language choice and language use patterns used by
Ria’s interlocutors in various situations.

From the age of 1;1 Ria started producing first words. This became the main focus of the
diary recordings. From this age records of any new linguistic developments (such as new
words) were kept whenever they occurred. When Ria’s utterances became frequent and it
was impossible to record them daily, only new types of utterances were recorded. De
Houwer (2009a, p. 223) argued a systematic diary study can be particularly suitable for
recording of a child’s early lexical development. While transcripts of recorded sessions are
limited in this respect, since they do not capture a wide range of lexical productions, a
systematic diary can capture many various productions over an extended period of time.

From the age of 1;4 until 1;10 an additional recording of Ria’s utterances was made in
order to obtain an approximate count of word types produced and to obtain data on Ria’s
36
expressive vocabulary. An attempt was made to record all of Ria’s utterance tokens on a
monthly basis. Each month the researcher spent a number of days recording every
utterance token into a vocabulary sheet. From these lists word types were tallied up for a
total number of words. Initially 1-2 days were sufficient to record every token Ria
produced. With the increasing number of words, and to be able to capture as many new
types of words as possible, the number of recording days each month was increased to 3-4
days. With more frequent speech it became impossible to record every single token, thus
the aim was to record every utterance in the first two days, and on the subsequent days to
record only new types of words as they occurred.

Thus during this period two kinds of utterance recordings were taking place: usual
sporadic recording of all new utterance types as Ria produced them, and monthly
recording of utterance tokens produced by Ria in a 1-4 day period, in order to capture all
word types in production and to obtain an approximate word count each month.

From the age of 1;10 Ria had progressed from one word stage to two-word and multiple-
word stages. During this period Ria’s vocabulary and her ability to form many new and
different constructions grew significantly, and it became impossible to record every
utterance. Thus from this age daily notes of new utterance types were kept and the focus of
diary notes was mainly on the new word combinations produced rather than individual
new words. The word count ceased.

At the end of the study from 1;11 Ria produced multiple-word combinations as well as
first sentences with adult like syntax, first in the Slovak language and soon after in the
English language. Initially an attempt was made to record all of the early sentences, but the
frequency of such constructions was rapidly increasing and it was possible to record only
new structures as they occurred, rather than every instance of a structure or a sentence.

Throughout the study the mother did not notice that the act of taking notes during daily
activities was in any way obtrusive to Ria, nor did it seem to affect Ria’s behaviour. On the
contrary, since the practice was taking place from birth, Ria perceived the mother’s note
taking as a normal part of daily activities. When Ria was able to express herself
linguistically, she demonstrated that she understood what the mother was doing by
occasionally stating that the mother was recording what she had said, telling the mother
what to write or spontaneously repeating words for the mother, as in this example:

[mamɪ, pɪʃ faɪfen] ‘mami, píš frying fan’ “mummy, write frying-pan” 1;11.6.

However, the research purpose had no effect on Ria’s behaviour.


37
The diary records as well as Ria’s utterances were initially written by hand. For the
purpose of analysis Ria’s utterances and notes related to those utterances (such as date,
situational context, interlocutors, translation into English, comments, etc.) were transferred
to computer files of various formats depending on the intended analysis. Diary data was
used mainly to establish developmental stages, keep track of the onset and development of
individual linguistic items and for collection of cumulative vocabulary.

Ria’s utterances recorded in the diary are reported in the following format:

[phonetic transcription of utterance] ‘gloss’ “English translation” (age),

as in

[mamɪ] ‘mami’ “mummy” (1;7)

for Slovak utterances, and

[dædɪ] ‘daddy’ (1;7)

for English utterances. Some utterance sequences were transcribed using the CHAT
transcription format as used on the CHILDES system (MacWhinney, 2008).

3.1.2 Audio-video recordings


A total of 251 audio-video recordings were available in this study, of which 67 were
selected for transcription and analysis. The recordings started when Ria was aged 0;4 and
lasted for the duration of the study. Over this period several recordings were made weekly
aiming to have at least one recording per week. Their length varied between 5-40 minutes,
but on average lasted about 15-20 minutes. A JVC Everio HDD (hard disk drive) digital
camcorder supported by a tripod was used for the audio-video recordings. This allowed for
quality audio as well as video recording, with easy transfer or reproduction of data and no
loss of quality. The use of audio-video was necessary to capture verbal and non-verbal
communication as well as context.

During the recording times, it was the mother who operated the video-camera, thus there
was no concern with an external observer affecting the participants’ behaviour. As some
authors pointed out (Deuchar & Quay, 2000, p. 18), it is possible that the video-camera
might also be considered an obtrusive object for the participants, affecting their behaviour.
However, Ria was quite familiar with the video-camera as a permanent object in the
household, and although she did look at it at times and reached out to touch it, it was more
out of Ria’s natural interest to examine all objects in her environment. At a later stage,
when Ria was able to express herself linguistically she also asked about the object, and its
38
purpose of recording family videos was explained to her. Ria also enjoyed watching the
recordings, much like she was used to looking at photographs taken by the family. Thus it
was concluded that the research purposes of the recordings were not affecting Ria’s
behaviour.

With the mother being the operator, it was possible to start recording virtually at any time
and there was no need to set up regular sessions. Rather, various contexts that Ria engaged
in throughout the day were captured, on different days of the week, at different times of
the day and in different locations of the household. Activities during recording sessions
included family meal times, one-on-one play time, play with toys, pretend play, role
playing, sharing and reading books, singing nursery rhymes, Ria helping with household
duties, bath time, special occasions such as parties and celebrations, and occasional
outdoor play. It could be argued that recordings made in the home environment only are
not sufficient, but it was observed that Ria was the most productive in her own
environment, where she felt comfortable and did not need any time to adjust to a new
situation.

During the recording sessions Ria was allowed to move freely in her environment so that
any limitations on her natural behaviour and linguistic expression were minimized.
However, as she learnt to crawl and then walk, her increasing ability to move around freely
made the task of capturing every moment and utterance, without adjusting the angle of the
camera, more difficult. Thus the activities of later videos from approximately 18 months
include more play, reading and meal times, since these where situations where Ria tended
to stay mainly in one area, which allowed for a less interrupted recording.

Ria’s interactions with her regular interlocutors were recorded, primarily with the mother
and the father, with occasional recordings with other members of extended family. An
analysis of the weekly patterns of language exposure based on the waking time showed
that Ria was exposed to three possible language environments: the Slovak environment,
the English environment and a bilingual environment where she was addressed in Slovak
by the mother (or other participants) and in English by the father (or other participants),
thus required to alternate between Language A and Language Alpha during interactions.
These three contexts were thus determined by the language spoken to Ria by her
interlocutors, and these in turn defined the language context of the recordings. However,
the language preferred by the adult to address the child should not be considered the main
determinant of the language context (Tracy, 2001, p. 18). It is the target language, and it
does not have to be in line with the actual languages used in the interaction. Determining
39
language context may be problem free when analyzing interactions between an adult and a
small, pre-linguistic infant, or a young child who started producing first utterances. It is a
different matter with a child who has become more skilled in using language to
communicate. While adults use discourse strategies that allow them to maintain their
preferred language context, a child may not adjust her language choice accordingly. She is
able to impose her language preferences, which need to be considered when analyzing
interactions. A child who is able to differentiate pragmatically and thus able to make
language choices consciously is helping to establish the language context. A child’s
language choice can influence the language mode of both bilingual speakers, which can
affect their subsequent language choices further. Therefore three aspects were considered
in determining the language of the context:

 adult language choice


 child language choice
 adult’s and child’s responses to their respective choices.

It was necessary to consider if Ria adjusted her language choice or insisted on her
language preferences. Language context may be affected by both the child’s and adult’s
language preferences and discourse strategies. A child who maintains a language different
to the adult is establishing bilingual language context, whereas a child who adjusts her
language choice accordingly helps establish monolingual context. Similarly, the adult
interlocutor may or may not accept the child’s preference. If adults adjust their language
choice upon a child’s insistence to use her preferred language they trigger language shift,
e.g. from monolingual language A context to bilingual context using both languages, or to
monolingual language Alpha context. Language context thus becomes a fluid concept a
child does not accept passively, but negotiates in the same manner adult speakers do.

Considering the above factors, three possible language contexts were identified and
captured in the video recordings. Transcriptions were carried out in these proportions:

 Slovak context - 36 videos


 English context - 6 videos
 Bilingual context - 28 videos.

In 2 videos the context changed from bilingual to Slovak, and vice versa, and in 2 videos
from bilingual to English and vice versa. Overall there were more videos in the Slovak and
bilingual contexts than there were in the English context. This imbalance, however,
reflects the overall patterns of authentic language exposure, as analysed in section 4.2. Ria

40
was exposed to Slovak environment ranging from 36-51.5%, to English environment 1-4%
and to the bilingual environment ranging from 47-61% of the total waking time. In
addition there were also two periods in which Ria was exposed to bilingual environment
100% of the time, including the first 2 weeks of Ria’s life and the family’s overseas trip to
Slovakia of a 5 week duration when aged 0;11 - 1;1. Thus overall, the imbalance of the
language contexts of the video recordings reflects the actual exposure to the two languages
in the home during the study.

Setting up specific recording sessions when Ria would be interacting only with an English
speaking interlocutor would not be an entirely natural context for her, since the normal
daily arrangements of the family resulted in a natural imbalance in monolingual language
use. De Houwer (1990, p. 77) describes a similar situation where usual patterns of
language use in the family influence the balance of data collected in each language.
However, as pointed out by De Houwer (1990, p. 77), it was preferred to obtain more
naturalistic data. Care was taken to include sufficient number of interactions with English
speakers (predominantly with Ria’s father) that would allow for comparisons of Ria’s use
of both languages. Moreover, the father often reported to the mother on Ria’s language use
on occasions when he was spending time with Ria in an English context.

Audio-video recordings were used to support the developments established through diary
records. For this purpose selected recordings were transcribed.

3.2 Data analysis


The main unit of analysis was the utterance. Following Lanza’s (1992, p. 638; 2004, p.
123) definition an utterance was defined as “a single word or combination of words with a
single intonation contour”. Thus utterance could vary in length, as reflected in each
chapter.

Chapter 6 analysed one-word utterances. Due to the instability of first productions criteria
to identify possible first words vary in strictness and the type of productions included.
Some studies developed detailed mechanisms (de Boysson-Bardies & Vihman, 1991;
Vihman & McCune, 1994), while others use more general criteria (Cruz-Ferreira, 2006; de
Boysson-Bardies & Vihman, 1991; Petitto, et al., 2001; Tamis-Lemonda, Bornstein,
Kahana-Kalman, Baumwell, & Cyphers, 1998). In general it is agreed that words elicited
by an adult or reproduced after an adult are not considered part of the child’s active
vocabulary. It is not until the child can produce words spontaneously with a clear purpose
in a particular context, as pointed out by Cruz-Ferreira (2006, p. 149), that productions can

41
be considered first words. Most researchers agree on the following minimum identification
conditions for first words:

 approximation to the adult form: at least one segment matching with the adult form, the
syllable pattern and stress matching with the adult form
 use of a consistent word form (sound sequence) consistently in relation to a referent
across multiple contexts
 appropriate use in more than one situation
 identification by the interlocutor
 spontaneous use of the word in context.

In other words, to be considered a word the production must be a spontaneous, consistent


sequence of sounds that regularly corresponds to any approximation to the target adult
form. This working definition of first words was adopted in this thesis to avoid any
parental bias in interpretation.

Chapter 7 analysed two-word and multiple-word utterances. Some word combinations had
the form of telegraphic speech, with a short pause and falling intonation between, but still
formed one unit and therefore an utterance. The decision was made not to calculate MLU.
Although this measure is used in language acquisition studies to determine the level of the
child’s grammatical development, De Houwer (2009a, p. 65) argues it can be problematic
as comparison across a bilingual child’s languages. Moreover, due to the fact that
inflectional morphology in early combinations appeared at a context related surface level,
and became productive only towards the end of the study period, MLU would not
accurately reflect Ria’s competence.

Chapter 8 considered different types of utterances from two perspectives: language mixing
and language choice. In structural analysis of mixing the unit of analysis was an utterance.
Utterances were categorized according to language of utterance, language context and
structure into 5 categories:

 Unilingual Slovak
 Unilingual English
 Mixed utterances
 Indeterminate
 Bilingual.

However, when considering language choice conversational turns at talk on discourse


level became the unit of analysis.
42
Lastly, in some chapters observations were made on speech perception, segmentation,
comprehension and paralinguistic meaning making. The study of these aspects in bilingual
infants in the first few months of life is subject to methodological constraints. These
aspects were examined mainly through observations of Ria’s reactions to parental address.
Conclusions were therefore inevitably affected by the parent’s interpretation of cues given
by Ria, the meanings the parents assigned to such cues, and parental responses. Language
comprehension is also difficult to study. The CDI method was not used in comprehension
data collection because there was no translation of the inventories available in the Slovak
language. Other methods were used to measure comprehension relying on behavioural
responses to familiar words within context, such as the head turn and eye gaze. These
responses confirmed Ria’s focus on speech and thus the recognition of familiar words in a
sequence of speech.

3.2.1 Transcription
Of the total 251 recordings, 67 were used for transcriptions. Due to the high number of
recordings, the time consuming nature of the phonetic transcription and coding of other
details it was not possible to transcribe and include all recordings for a detailed analysis.
However, all the recordings available in this study were used for a general analysis of
Ria’s language learning environment, as well as to determine the approximate
developmental stages and milestones. Preference was put on obtaining a bigger volume of
recordings, recorded more frequently over a longer period of time and in various
situations. While it would be possible to observe changes and developments from less
frequent recordings, for example on a monthly basis, weekly videos allow to follow the
progression of developments through, noting subtle changes as well. The continuity of
development was regarded more important than notable changes over longer periods,
despite the fact that it was a more time consuming task.

Only selected sections of the video recordings were transcribed and analysed with the
purpose of a more detailed analysis of language use by all interlocutors in the context. The
selected recording sections were transcribed using the CHAT format of CHILDES. The
conventions used in transcription are outlined in the following sections. List of transcribed
videos is reported in APPENDIX A.

3.2.1.1 Selection for transcription


During analysis each recording was viewed and a summary of linguistic phenomena
present in Ria’s speech, developmental aspects, language choice by participants and

43
language use patterns was produced. When selecting passages for transcription, the main
purpose of the study was important - to gain an insight and understanding of the overall
process of BFLA in the early years, which naturally requires observations over a longer
period of time, rather than detailed analysis focusing on individual aspects or periods. The
following principles for selection were adhered to:

 A maximum period of 2-3 weeks between passages to ensure continuity of


development
 Passages that included linguistic productions representative of particular periods
 Passages with the largest proportion of Ria’s spontaneous talk, which would most
accurately reflect her linguistic abilities within the period
 All three language contexts must be represented.

The linguistic phenomena and Ria’s abilities for each period were cross-checked with the
diary records (with weekly recording schedule). Thus for example dyad interactions
between Ria and an adult participant were the subject of transcription, while the adult
reading to Ria whilst she was listening only and not spontaneously interacting, or extended
conversations between adults whilst not addressing Ria directly were excluded from
transcription.

Sections were selected according to the following criteria:

 Each transcribed section represented an interactional episode


 An episode was considered as ending whenever the topic or the focus of the interaction
changed, or whenever there was no response following a speaker’s turn for 10 seconds
or more.

For each episode the setting, activity, and persons present were noted. Following Lanza’s
(2004, p. 128) methods the first 50 utterances of a recording did not need to be excluded in
transcription, since the child was already familiar with the parents.

3.2.1.2 Phonetic and orthographic transcription


Using the principles of CHAT format, all child and adult utterances in a selected section
were transcribed. Since Ria’s utterances before 2 years of age were recorded, the
productions did not yet have adult forms, and approximations may had been overlooked
and not interpreted as attempts to produce an adult word. To avoid this, all of Ria’s
utterances were transcribed phonetically using broad phonetic transcription based on the
International Phonetic Alphabet (IPA) symbols. A phonetic transcription allows for all

44
utterances, including proto-words, to be interpreted within the context, not just the
utterances that closely resemble adult forms.

IPA conventions for transcription of speech are not fully adequate for transcribing infant
and child speech. A child with a developing articulation is unlikely to produce sounds
occurring in her languages consistently, but with variations in the pronunciation of a
particular sound as she is attempting to produce the target phone. However, as Cruz-
Ferreira (2006, p. 56) points out, IPA provides the most convenient alternative that
approximates the actual sound productions of infants.

A further problem with the use of IPA when transcribing Slovak utterances was the fact
that in Slovak linguistics the IPA conventions are not traditionally used for phonetic
transcription of the Slovak language. The accepted and widely used system, the Slovak
Phonetic Alphabet (SPA), is based on the symbols used in the orthography of the language
(Kráľ, 1984, p. 39). The suitability of IPA for transcription of Slavic languages has been
criticized in Slovak linguistics (Ivanecký & Nábělková, 2002, p. 86). It was pointed out
that the IPA was based primarily on non-Slavic languages, and consequently difficulties
with accurate phonetic transcription of Slavic languages arise. At the same time the need
for a solution of discrepancy in the use of two transcription conventions has been
acknowledged, especially for the purpose of wider international publications of research
focusing on the Slovak language.

Ivanecký and Nábělková (2002, p. 89) provide a conversion table from SPA to IPA and
admit that even this conversion system is not entirely problem free. The main difficulty in
representation of Slovak sounds presents itself with vowels. According to Ivanecký and
Nábělková (2002, p. 91) vowels in the Slovak language are not identical with the cardinal
vowel system, since they are not identical with open or closed vowels in the IPA. It could
be argued, however, that any form of representation of sounds with a particular symbol is
merely an approximation. Although the sound representations used in Ivanecký and
Nábělková (2002, p. 94), as pointed out by the authors, are not entirely suitable for the
Slovak language, nevertheless, they are the closest approximations available to date.

The reality that infant and child speech is not always consistent with adult forms and the
fact that only a broad phonetic transcription of Slovak utterances was necessary in this
study meant that the current available transcription system of Slovak into IPA was
sufficient. This was the transcription used throughout the thesis when transcribing Ria’s
Slovak utterances phonetically. When transcribing Ria’s English utterances, IPA for

45
English was used. A list of the IPA symbols for Slovak language and their orthographic
representation as used in the transcriptions is shown in APPENDIX A.

In transcription of adult utterances the usual orthography of the respective languages was
used. For Slovak utterances a loose translation into English is provided as well. It is also
necessary to note that the transcriptions in this study are representations of spoken
vernacular language. As such they necessarily include non-standard forms. This is the
model of language presented to Ria. When a child is acquiring language as a tool for
communication, she targets the model of languages in her environment in her utterances,
and consequently adopts the forms that are presented to her in the input. Therefore, when
transcribing adult utterances care was taken to preserve all non-standard adult forms in the
transcription.

3.2.1.3 Conventions of CHAT transcriptions


The transcriptions were produced using conventions of the CHAT transcription format
(MacWhinney, 2008), which consists of two sections:

 an introduction to the transcription contained in the file headers,


 speech and non-speech lines contained in the main and dependent tiers.

The obligatory file headers delineate the beginning and end of the file, and include
constant headers with information about the participants, date, place and situation of the
recording, age of Ria and languages used by the adult participants when addressing Ria.
There are also changeable headers showing the location of the passage (interactional
episode) in the recording.

The headers are followed by speech and non-speech lines in which participants’ utterances
and other information are reported. Only these lines were reproduced in examples in this
thesis. Each example is followed by Ria’s age at the time of recording shown in brackets,
as shown in the following example:

%sit RIA and MAM are reading a book when RIA picks up a magazine and
passes it to MAM

*RIA [mama. mamɪ ʃa.]

%add MAM

%glo mama. mami časopis.

%eng mum. mummy(’s) magazine.

46
*MAM To je mamičkin časopis, ano.

%add RIA

%eng That’s mummy’s magazine, yes.

*RIA [da ʃai mamɪ]

%glo dá časopis mami

%eng (Ria) give magazine (to) mami

(1;8.9)

Speech lines consist of main tiers, showing the utterances. These are indicated by an
asterisk and a name assigned for each participant. Following each utterance are dependent
tiers containing non speech lines. These are preceded by the symbol % and include
additional information, such as orthographic gloss of the target utterances that Ria
produced, which also indicates non-standard child forms and in some cases an idiomatic
meaning of Ria’s utterance. For all Slovak utterances an English translation is provided.
When necessary for analysis other dependent tiers are also included to indicate addressees
of utterances in a bilingual context, proxemic, gestural or paralinguistic information,
situational context and actions of the participants, addressee of the utterance, analytical
comments and explanations. This information was necessary for interpretation of Ria’s
utterances and to determine the language context and language mode of the speakers in
analysis of language mixing and language choice. Care was taken to include all relevant
aspects of the interaction that communicated some form of meaning, such as proxemics,
gestures, whispering, laughing, crying sounds, and actions.

A list of speech lines and non-speech lines used in the transcriptions, explanations for each
abbreviation and other utterance markers from CHAT manual used in the transcriptions are
summarized in APPENDIX A.

In addition, when required for a more detailed analysis of speech such as pauses, prosody
and intonation patterns, other symbols adapted from the CHAT manual were used in the
transcription. For example during the telegraphic speech stage, when Ria was producing
early two-word combinations by leaving a pause between two words with a falling
intonation, the pauses and intonation patterns were recorded and coded. Occasionally it
was necessary to mark intonation patterns used by Ria with Slovak as well as English
utterances. Symbols used for this type of analysis are also listed in APPENDIX A.

47
The transcriptions were not verified by a second transcriber, since a person that would be
proficient in both languages and trained in linguistics and phonetic transcription was not
found.

3.3 Data presentation


Linguistic development tends to be described by way of stages that are organised in a
chronological order according to milestones achieved by particular age. By choosing to
operate with stages researchers suggest that linguistic development takes place in a linear
fashion, with periods of a particular duration, uniformly following a set sequence
(Hoffmann, 1991, p. 72). Hoffmann (1991, p. 72) pointed out several problems with
describing development in terms of stages. Firstly, while general developmental patterns
are the same and normally developing healthy children will go through the same
progression of development, each child develops at a different rate, and their language
acquisition is dependent on various social, psychological and environmental factors.
Secondly, developments do not follow each other in a perfect sequence, rather it is often
difficult to determine when one finishes and another one starts. Developmental stages
seem to come in waves and often two stages overlap for different periods of time.

Nevertheless, for the purpose of this thesis the data were organised into developmental
stages. A developmental stage can be defined as a period that is most characteristic of a
development of a specific linguistic trend. A stage started shortly after the first occurrence
of a particular phenomenon, when Ria started using it regularly, and it lasted until a new
development became more dominant. Since linguistic items were learnt over time in waves
of development and several new items appeared simultaneously, the stages overlapped.
Periods where two stages overlap will be clearly specified to highlight progressions in the
development of linguistic trends.

The following overview lists the stages that were observed in the data for both languages,
and gives an approximate age range and a brief description of each:

 Prenatal stage (2nd trimester – birth):

 The auditory system becomes functional around the twenty-fifth week of gestation
 The foetus is familiarized with the prosody and rhythm of the languages spoken by
the mother and people she regularly interacts with
 Prestages (birth – 0;10) subsume two periods (Oksaar, 1983):

48
 The earliest stages (birth-0;3): crying sounds and early cooing (glottal and velar
sounds), vocal expressions are reactional
 The babbling stage (0;3-0;10): expressions become spontaneous; Ria is responding
to others interacting, playing with cooing and babbling sounds, expressing
emotional states and feelings, vocalizations are accompanied by gestures
 Ria is creating expression-content associations
 Ria starts responding to her own name at age 0;6
 Transition period (0;10- 1;4) includes rich linguistic repertoire:

 Pure babbling (e.g. [baba] [pa] [mamamam] [dadadada]) resembles adult speech in a
number of features: it combines sounds creatively into sequences similar to
syllables, words and longer stretches; it appears to have intonation patterns similar
to adult speech, most often statements and exclamations; it follows the turn taking
rules of a dialogue.
 ”Babbling words” (also termed vocables, babbling words, vocalizations) are
phonetically consistent vocalizations, often combined with paralinguistic and
kinetic means (e.g. [aː aː aː aː] + pointing and leaning towards an object, [iɟaɟ] +
laughing) which Ria invented herself. Their form and functions resemble adult
words, but since they are idiosyncratic can be understood only by the caregivers.
 Words used at this stage are approximations to adult words which Ria attempts to
produce. Their meaning can be established in numerous situational contexts (e.g.
[mama] ‘mami’ “mummy”, [dada] ‘daddy’, [dada]→[kaka]/[gaga] ‘kačka’ “duck”,
[Ɉeje] ‘kde je’ “where’s”, [ham] ‘hami’ “boobie”).
 One word stage (1;4 – 1;7):

 Words in the form of holophrastic utterances appear


 Ria uses paralinguistic and kinetic means in combination with holophrases
 Prosodic characteristics and situational context determine the meaning, since the
reference depends on the situation
 By the end of the stage Ria actively produces approximately 245 words
 Two-word stage (1;7 – 2;0):

 Ria’s repertoire consists of both one-word and two-word utterances, as well as some
multiple-word utterances, which express basic semantic relations (e.g. agent-action)

49
 Combined utterances can have the form of word-utterance blocks, which consist of
two one-word utterances uttered together but with separate intonation and a pause
between, or the form of two words uttered under single intonation
 Expressive vocabulary expands
 Productions develop further and approximate adult forms closer
 Inflectional endings appear
 Early word ordering with syntax appears in multi-word combinations.
The progression of stages is shown in Figure 1.

Figure 1 Developmental stages

Prenatal
stage
(2nd Transition Two-word
trimester- period utterances
birth) (0;10-1;4) (1;7-2;0)

Prestages One-word
(birth-0;10) stage
•The earliest (1;4-1;7)
stages (birth-
0;3)
•Babbling stage
(0;3-0;10)

Aspects of the prelinguistic stages relevant to the developing bilingualism will be


discussed briefly. The stages from the onset of speech (One-word and Two-word stages)
will be analysed in more detail showing examples from the data. First the productions
typical of each stage and then emerging communicative competence skills will be
described. Lastly, the question of emerging bilingualism will be addressed.

Several early studies in childhood bilingualism report the development of bilingualism in


stages organized before and after formal differentiation of the two languages (Redlinger &
Park, 1979; Vihman, 1985; Volterra & Taeschner, 2007). However, as shown in the
literature review in section 2.2, these models of bilingual development were opposed in
more recent literature. Moreover, since in this thesis differentiation was perceived as a
process rather than a point in time this approach was not adopted. Mixing was explained

50
by pragmatic and sociolinguistic factors rather than lack of differentiation, as discussed in
Chapter 8.

A child’s linguistic development does not take place in isolation from other developments.
Infant’s contexts provide the topics for linguistic interactions she initiates or is engaged in
by others. During the early years of life such contexts are limited in comparison to an older
child’s contexts, since they are to a great extent related to the infant’s motor skills and
readiness. The contexts of interaction stem from what the infant is able to participate in. As
she progresses through development the contexts expand and become more varied, which
also impacts on further linguistic development.

51
4 Language learning environment in the family and linguistic
input
This chapter includes a descriptive analysis of Ria’s language acquisition history. To
present this information as accurately as possible, protocol for information about BFLA
contexts developed by De Houwer (2009a, p. 334) was adapted. Apart from description of
Ria’s language acquisition experience the protocol requires the following background
information:

 Family composition
 Patterns of home language use
 Child’s care history
 Child’s residence and long-term travel
 Child’s health.

Ria’s linguistic environment consisted of two languages, Slovak and English. From the
prenatal stage the mother addressed Ria in Slovak. The father did not address the foetus
directly, he addressed Ria in English from birth. Since Ria’s first exposure to both
languages after birth occurred within a period of a week, this type of bilingualism can be
termed bilingual first language acquisition as defined by De Houwer (1990, 2009b).

4.1 Child’s parents


In a discussion paper on methodological issues in the field of bilingualism Grosjean (1998,
p. 133) suggested that it is necessary to record not only the language environment of the
child under study, but also the language background of the people from whom the child
received the most language input. Language background of a parent is a factor that
influences parental language attitudes, and therefore language acquisition. In the following
paragraphs Ria’s and her parents’ language history is described in detail.

Table 4.1 Family Composition

Family member Year of birth Country of origin Gender Employment status


Mother 1976 Slovakia Female Student
Father 1974 Australia Male Full-time
Child 2007 Australia Female -

52
4.1.1 The mother
Ria’s mother was a multilingual speaker of Slovak, English, Italian and Czech3. The
mother was born in Slovakia and her mother tongue was Slovak. She was growing up as a
Slovak native monolingual speaker and completed primary and secondary education in the
Slovak language. At the age of 8 she started studying English as a foreign language at
school and through an after school language course. She studied English formally until the
completion of secondary school at age 18, with English as one of the matriculation
subjects. She then taught English as a foreign language at a primary school for a period of
two years.

Apart from learning foreign languages through formal schooling the mother was also
learning the Czech language largely through media exposure. In the historical and social
context of Slovakia the Czech language appeared widely in all forms of mass media and a
large majority of Slovaks acquired varied degrees of communicational skills. Due to the
typological closeness of the languages (both Slavic languages) and the joint history of the
two nations even children of school age were expected to understand Czech without formal
tuition. In the literature this form of linguistic competence is referred to as passive
bilingualism (Nábělková, 2002, p. 102). While there was an awareness of the two
languages and a clear distinction was made between Czech and Slovak, it was not a usual
practice to label this form of linguistic acquisition as bilingualism and children only
developed their bilingual awareness as adults.

From the age of 16 the mother attended a course in Italian at a state school of languages
and later enrolled in a university course in Italian completing one semester in Slovakia.

At the age of 21 the mother migrated to Australia, where English became the dominant
language in both private and public spheres. Most importantly, English became the
language of her relationship with Ria’s father, a monolingual speaker of English. After
migrating to Australia the mother enrolled in a university course in languages and
linguistics and completed Honours degree in Italian.

Thus since migrating to Australia the mother experienced a major shift in her language use
- from dominant Slovak to dominant English. After her arrival in Australia her Slovak
usage initially became relatively limited. She had regular face-to-face contact with her
sister in Australia and occasional contact with some acquaintances, but her contact with all

3
In addition to the four main languages the mother also acquired Russian, German, Portuguese, Latin and
Greek as foreign languages. However, she had no contact with these languages since school and had only
passive knowledge.
53
other relatives and friends from Slovakia was limited to telephone conversations and email
exchanges, except for 2 one-month visits to Slovakia, and 2 one-month visits by Slovak
relatives in Australia. Slovak therefore shifted from being the mother’s primary language
of communication in all domains to being used mainly in the spoken form with limited
reading in the private domain. English meanwhile shifted from a foreign language to a
dominant language used for speaking, writing and reading in most domains.

This shift started reversing when the mother took up a role as a volunteer presenter in a
Slovak ethnic radio and later became a translator, thus increasing her use of the language
to other domains. She met more people of Slovak background and spent more time reading
and writing in the language on various topics.

From the mother’s pregnancy with Ria the language shift to English was reversed in the
private domain as well. She started using Slovak when addressing the foetus in-utero, and
continued to do so after Ria was born. Thus Slovak again became more dominant. Since
throughout the first two years of Ria’s life most of the time with Ria was spent at home,
Slovak became the overall dominant language for the mother again.

When the mother’s experience with Slovak as dominant language in childhood and
adolescence is compared to her experience with Slovak as the dominant language in the
first years of Ria’s life, there is a significant difference related to topics and register. While
prior to Ria’s birth the topics varied from private to academic domains with respective
registers, after Ria’s birth they tended to be limited to the private domain and concerned
the language of small children.

4.1.2 The father


Ria’s father was a monolingual speaker of Australian English. He grew up in an English
speaking family in Australia. The father’s heritage background on the paternal side was
Greek. He was exposed to limited Greek input since childhood. He was regularly
addressed in Greek by his paternal grandparents. Although he understood only some
utterances addressed to him, he was able to infer meanings from the context and answer
appropriately in English. He was not able to communicate in Greek beyond a few basic
phrases. He attempted formal study of the Greek language as a heritage language at school,
and more recently through an adult language course, but did not pursue any further study
to develop and maintain the skills.

54
The father also studied Italian as a foreign language for several years at school and at
university but was never able to communicate other than a few idiomatic expressions,
which he learnt mostly by rote.

The father also attempted to learn the Slovak language but lacked the motivation to
continue. However, he was able to understand some words and phrases which he had
acquired from exposure to the mother’s interactions with the Slovak family and friends
over the 10 years of the relationship. He was unable to express himself or to hold a
conversation in Slovak apart from occasional attempts to use familiar words/phrases in a
jovial manner.

Throughout the bilingual parenting approach and experiences in the bilingual family the
father was immersed in the Slovak language since Ria’s birth. While he was not directly
addressed in Slovak, he was overhearing Slovak interactions as a participant in the
bilingual context. One of the family’s strategies was to interpret or retell interactions to the
father, whether he requested an interpretation or not. Once Ria was able to express herself
linguistically she also used this strategy for the father’s benefit. The combination of
overhearing Slovak interactions within the situational context and subsequent
interpretation allowed the father to learn aspects of the Slovak language, such as
vocabulary, expressions and some morphology. By the end of the study, he was able to use
individual words and several simple phrases himself.

Although English was the father’s dominant language, and the only language he was able
to express in, his experiences with other languages positively shaped his language
attitudes.

4.1.3 From monolingual to bilingual home


The language of Ria’s parents’ relationship was the language they had in common -
English. It was the dominant language in the nuclear home for 10 years prior to Ria’s birth.
Slovak was also used in the home, but only in limited situations when other Slovak
speakers were present (relatives and acquaintances), or during telecommunication with
relatives and friends from Slovakia. If the father was not present during such interactions,
only Slovak was used. If the father was present, the other speakers attempted to
accommodate him linguistically when they deemed appropriate. As a result, on these
social occasions mostly Slovak was used in interactions between Slovak speaking
interlocutors, but whenever it was necessary or socially appropriate to include the father in
the conversation, all Slovak speaking interlocutors momentarily switched to English.

55
However, the mother used only Slovak to address her parents and youngest sister directly
even in front of the father4, since it felt neither natural nor right to address them in English.

Thus English was the primary language of the home before Ria’s birth. After Ria’s birth
there was a shift from English dominant to a bilingual home environment. English became
the language of usual communication between the mother and the father, and between the
father and Ria. Slovak was the usual language of communication between the mother and
Ria.

4.2 The child


Ria was born in a hospital in Adelaide, Australia, and during the one-week hospital stay
she and her Slovak-speaking mother roomed-in. The father was present at the birth and
visited daily afterwards. The hospital staff spoke Australian English in conversations with
the mother and the father. Thus Ria was exposed to both languages from birth, and
continued to receive regular input in both languages during the period of the study (birth to
2;0).

From birth each language was associated most importantly with a person. The mother, who
addressed Ria in Slovak, was the primary caregiver since birth. Thus Ria spent the greatest
proportion of time in the first 2 years with the mother. When the father, who addressed Ria
in English, was the main carer, it was the only time Ria was exposed to English only.
Periods of such monolingual English contexts varied throughout the 2 years of the study,
initially being limited to approximately 2 hours a week, later extending to 4 hours a week.

Ria’s parents spoke English to each other. Thus when the whole family was spending time
together interaction took place in the form of bilingual context. This occurred during
weekends, and on average 3 hours during the working week (in the morning and at night).
Table 4.2 outlines patterns of most common language use in the family.

Table 4.2 Patterns of home language use (adapted from De Houwer 2009)

A
NOTE:
This figure/table/image has been removed
to comply with copyright regulations.
It is included in the print copy of the thesis
held by the University of Adelaide Library.

4
The mother had three other siblings (an older sister and 2 younger brothers) who were fluent English
speakers and in Ria’s father’s presence she addressed them in English.
56
When Ria’s linguistic input is considered from her perspective, the mother used both
languages, but strongly preferred Slovak when addressing Ria, while the father used
English only.

Monolingual input in English was received also from other relatives in Australia. English
was also overheard when spoken by other speakers, including the mother, in conversations
with the father, and outside the private home domain. However, when the mother spoke
English to other interlocutors, Ria heard this language as spoken by a L2 speaker, while
influence from the mother’s L1 was apparent mainly in accent and pronunciation, with rare
examples of transference on other levels of language functioning.

Ria received monolingual Slovak input from relatives who resided in Slovakia, which was
limited mainly to telephone conversations. It also included a separate occasion of one
month when relatives from Slovakia were visiting the family in Australia in August 2007.
During this period Ria’s exposure to Slovak was from several interlocutors.

There was also a short period of time when the domain status of the two languages was
reversed. During the family’s 4-week visit in Slovakia in July 2008 the language of the
public domain became Slovak. The mother’s relatives and acquaintances used the standard
Slovak spoken in the Bratislava region to address Ria. The only exposure to English from
a native speaker that Ria received during this period was in interactions with the father,
who continued to address her in English. The dynamics of language choice within the
extended family whilst the father was present were the same as described in section 4.1.3.
Thus Ria also overheard the mother and her relatives as L2 English speakers in their
interactions with the father.

Ria’s residence during the study is summarized in Table 4.3.

Table 4.3 Ria’s residence and long-term travel (>1 week) since birth

From (m, y) Until (m, y) Residence/Place visited Majority language(s)


July 2007 June 2008 Adelaide, Australia English
June 2008 August 2008 Bratislava, Slovakia Slovak
August 2008 July 2009 Adelaide, Australia English

4.2.1 Ria’s SES and health


A child’s health and socio-economic-status of the parents are also factors influencing
linguistic development. Ria’s social background was the middle class. Both parents were
university graduates in professional jobs. The father was an IT consultant and the mother
was a linguist undertaking doctoral research.
57
Ria was a healthy, normally developing infant and did not experience any health problems
that could affect hearing, cognitive functioning or language use (De Houwer, 2009a, p.
336). Ria did not suffer from hearing or speech problems and had no developmental delays
or disorders. It is therefore assumed that her linguistic development in the two languages
should take a developmental trajectory similar to normally developing monolingual
children in the two languages.

4.3 Parental language presentation


Bilingual parenting guides recommend a child is spoken to by an adult in their native
tongue, in order to expose her to the natural form of adult language with minimal errors,
and to achieve a consistent and accurate model (Menyuk & Brisk, 2005, pp. 30-31). Thus
the model of the language a child observes is seen as crucial. Moreover, parenting guides
suggest it is more important for a parent to have a relationship with a child based on the
parent’s preferred language, rather than to accommodate other people and include them in
interactions, while neglecting the parent’s language (Cunningham-Andersson &
Andersson, 1999, p. 36). However, this aspect of linguistic relationship was not researched
from the perspective of an emotional attachment based on a particular language.

In this thesis changes in the overall language patterns in the nuclear family occurred during
the mother’s pregnancy. The mother started addressing the foetus in Slovak as early as the
prenatal period. She believed that since in the third trimester the foetus’s auditory system
is developed it was necessary to already provide input in both languages. In fact, research
confirmed the role of prenatal exposure, and it is well understood that infants get used to
the prosody of the mother’s language in utero, resulting in newborns’ preference for that
language (de Boysson-Bardies, 2001, pp. 22-26).

The mother did not have to make a decision to bring up her child in Slovak. It was self-
evident that the usual language of communication with her child would be her native
tongue. Slovak was the language that felt most comfortable and natural while interacting
with her child, since she was able to provide a linguistic bond through the use of Slovak
motherese (Barron-Hauwaert, 2004, p. 24). She had very limited awareness of CDS in
English, and thus she would not be able to provide this form of input to Ria naturally.

The mother felt that establishing a close emotional attachment based on her mother tongue
with Ria early in the relationship was crucial, since it was likely to determine the Slovak
language as natural and normal medium for communication within the relationship. This
would likely influence long-term success of Ria’s bilingualism as well. She also felt it was

58
important that Ria would be able to confidently communicate with her side of the family in
Slovak, namely Ria’s grandparents, other relatives, and mother’s friends and their children.
The mother felt she would be greatly disappointed if her relatives had to resort to English
to communicate with Ria, even though most relatives were fluent English speakers.

Thus for the mother a linguistic need arose during pregnancy to create a Slovak language
context with regular in-utero exposure to the Slovak language. However, in her daily
activities she engaged mainly in conversations with speakers of English, and lacked an
opportunity to speak Slovak. The diary records include a note on the mother’s active
attempt to engage in daily routine by addressing the foetus in Slovak. Since the yet unborn
child was expected in a relationship that was based on English, initially the mother had to
consciously focus to ensure she spoke Slovak when addressing the foetus. This was despite
the fact that Slovak was an instinctive form of interaction with children for her, due to
childhood experiences with Slovak as a first language.

In the father’s presence the mother’s approach for a short period in early pregnancy was to
interpret Slovak utterances, so as not to exclude him from address of the foetus. However,
eventually she interpreted her utterances into English only when she judged it necessary to
directly include the father in the address. Since she mostly addressed the foetus only it was
a form of intimate mother-child dyad interaction and there was no need to interpret for the
father’s benefit. By the third trimester the mother felt comfortable with her language
choices. Thus a period of approximately 6 months was necessary to establish an emotional
relationship with the yet unborn child through the medium of the Slovak language in an
otherwise English speaking environment.

The dynamics of language choices within the family changed again after Ria’s birth. The
mother used mainly Slovak and the father used English when addressing Ria, thus they
separated languages according to a person. It was self-evident that from birth each parent
would address Ria in the language that felt the most natural to them for interactions with
an infant (De Houwer, 2009b, p. 4). Thus it was not a planned decision to use this
approach, but it was intuitively suitable for the family’s linguistic needs.

When parents use an approach in which they separate the two languages completely, it is
called naturalistic separate setting. In practice the child hears one language from one
parent, and the other language from the other parent, which is a variety of Grammont’s one
parent-one language approach (also called OPOL or 1P/1L) (Barron-Hauwaert, 2004, p. 4).
When the speakers do not separate the two languages when interacting with the child, it is
called naturalistic fused setting.
59
However, it is not always the case that families are strictly either separating or fusing their
languages (Ervin-Tripp & Reyes, 2005, p. 87; Goodz, 1989, p. 32). De Houwer (2009a, p.
110) also argues this strategy is an exception rather than a rule in most families, and they
do not adhere to such systematic presentation patterns. These two options roughly
represent the opposing poles with varied preferences in between. It is more likely that at
least one, if not both parents will occasionally, and in specific social situations, address
their children in either of the two languages, thus use some kind of combination of the two
settings.

Parental language presentation can vary not only between families but within each
individual family (King & Fogle, 2006, p. 699). De Houwer (2009a, p. 111) suggests using
the terms 1 parent/1 language and 1 parent/2 languages (1P/1L and 1P/2L) to
differentiate between the two parental input principles, and proposes a third possibility,
which takes into account that while one parent may be adhering to the 1P/1L principle, the
other parent may be using both languages when interacting with the child. This combined
principle is termed 1P/1L & 1P/2L.

Similar observations in support of the combined 1P/1L & 1P/2L principle were recorded in
the diary notes in this thesis. The descriptions of language presentation pattern show that
the basic principle used by the parents was 1P/1L. However, the general observation was
that Ria did not receive uniformity but diversity in language presentation. Moreover, the
family did not settle to one final approach, but it was a fluid phenomenon that
continuously evolved with changing family and social requirements. Throughout the study
various social and linguistic situations emerged when the two languages came into contact.
On those occasions it was necessary for the family to plan ad hoc when deciding on
appropriate language choices.

While person-language separation felt natural and simple when only one parent was caring
for Ria, it was not always straightforward for the bilingual mother to adhere to Slovak in
the father’s presence. Although the father never expressed concerns about being excluded,
according to the diary notes the mother was conscious so as not to appear that she is
deliberately excluding him from interactions, and occasionally addressed Ria in English to
accommodate the father linguistically.

Thus despite the fact that a linguistic bond was established with the foetus in Slovak,
additional period of 5 months after the infant’s birth was necessary for all family members
to adjust to the approach. When Ria was aged 0;1-0;2 the mother’s relatives from Slovakia
visited the family for a period of one month. During this time the mother received support
60
for her need to bond emotionally with the infant through her native language, which
reinforced the natural need for such bond. By the time Ria turned 0;5, the diary notes state
that the mother was comfortable using exclusively Slovak when addressing Ria, even in
the presence of the father. Slovak utterances were interpreted to the father only if he
requested an interpretation, or the mother volunteered an interpretation if she wished to
include him in the interaction. English was used to address both Ria and the father
simultaneously. By the time Ria was 0;6 the diary notes state that the mother felt
uncomfortable if she had addressed Ria in English. Similar experiences were recorded by
parents in other diary studies (Leopold, 1939, p. 143; Saunders, 1984, p. 80).

The parents’ language choices and the mother’s alternation of languages according to
addressees are shown in the following excerpt:

%sit The family has just woken up, MAM is getting up, RIA and DAD are
still in bed, chatting to MAM

*MAM You’re so cute!

%add RIA & DAD

*MAM Look at that, I love that! One leg lifted, and see the way her hands are like
this underneath her. And the cheeks.

%add DAD

*DAD She’s happy.

*MAM Ty nádhera! No ahoj! Ahoj, krásna moja!

%tra Beautiful! Well hello! Hello, beautiful!

%add RIA

*RIA [ɛɛːɛ ɛː]

*MAM [ɛɛː ɛ ɛɛː]

%com Mirroring

*MAM Si veselý človiečik?

%tra Are you happy shortie?

%add RIA

*DAD Look, she almost sit her bum up.

*MAM Yeah!
61
%add DAD

*MAM Človiečik moj veselý! Prečo si včera robil také problémy mame?
Človiečik? Prečo, no? Si nechcela spať včera večer. Do ôsmej som tu s
tebou sedela. Do ôsmej, dve hodiny! Dve hodiny sme sa uspávali! Hm?
Prečo?

%tra My happy shortie! Why were you causing trouble for mummy yesterday?
Shortie? Why then? Didn’t want to sleep last night. I was sitting here with
you until eight. Until eight, for two hours! We were putting you to sleep for
two hours! Hm? Why?

%add RIA

*DAD What about two days?

*MAM Two hours, I was putting her to sleep last night.

%add DAD

*DAD Ah yeah.

*MAM And she fell asleep, I xx xx.

%add DAD

*MAM You think it’s a game? You think mami’s playing a game with you?

%add DAD & RIA

*MAM Look look look look look!

%add DAD

*DAD What?

*MAM She’s figuring it out.

%add DAD

*MAM Ty zlatko malé, ty!

%tra You little cutie, you!

%add RIA

*MAM She loves the tassels on the blanket.

%add DAD

62
*MAM To sa ti páčia tieto framforce? Framforce sa ti páčia? Páčia sa ti framforce?5

%tra Do you like these tassels? You like tassels? Do you like tassels?

%add RIA

(0;6.16)

In the example the mother switches between languages according to the addressee, using
English when addressing both the father and Ria simultaneously. When the father requests
an interpretation the mother provides it. The father uses English.

Thus three discourse strategies used by the mother emerged in the first year of Ria’s life:

 monolingual Slovak strategy when addressing Ria


 monolingual English strategy when simultaneously addressing Ria and the father
 bilingual strategy consisting of a Slovak utterance addressed to Ria followed by an
English interpretation addressed to the father.

Similar social situations with other monolingual English speakers presented some
difficulty in language choice. The mother found the need to accommodate other speakers
and include them in the interaction by addressing Ria in English in their presence. While
this often resulted in feelings of disappointment on the mother’s side, such strategies
provided Ria with a model of bilingual linguistic behaviour. Most often these occasions
occurred in the public domains (in shops or at university), but emotionally it was most
difficult with the father’s monolingual side of the family. Such situations are
psychologically difficult by nature, thus the added divide of the language the extended
family did not understand could potentially hinder communication. To resolve this issue,
the mother decided that the use of Slovak when addressing Ria was more important than
accommodating other speakers in interactions. The decision was thus to maintain the use
of Slovak when addressing Ria. The mother’s believes were reaffirmed in the bilingual
parenting literature:

“Before I would switch to English so as to not exclude anyone. Now I don’t care how they
feel. My relationship with my son is most important to me.” (Dr Edgra Monterroso, USA,
cited in Cunningham-Andersson & Andersson, 1999, p. 9.36).

Thus, when addressing other monolingual English speakers simultaneously the social
situation called for the use of both languages. In these situations the mother was modelling

5
Non-standard use of the word ‘franforce’ “tatters” with substitution of the nasal stop [n] for [m], and used
instead of the word ‘strapce’ meaning “tassels”
63
bilingual behaviour. Appropriate language choice was determined by the addressee, thus
the mother modelled whom to address in which language. Occasionally the mother judged
it more appropriate to address Ria in English so as not to offend any interlocutors by not
including them in the conversation. The use of English was dictated for various reasons:

 the need for politeness to accommodate other speakers


 to include monolingual speakers in the conversation in order to maintain harmony
 to encourage interaction with Ria
 during interactions with other children in an attempt to make Ria feel comfortable
among her peers
 mother’s intentional modelling of desirable linguistic behaviour to Ria
 when the mother was interacting with other English speakers and Ria was present and
overheard her speech, observing the mother in her role as a bilingual speaker.

The mother’s usual strategy in interactions with other speakers was to separate the
languages by switching to Slovak to address Ria individually, or switching to English to
address the English speaker. She used English when addressing Ria and other English
speakers collectively. The following example shows how the mother alternated between
collective and individual address:

%sit RIA is unwrapping present from VAN

*MAM Nah, not in your mouth. Nepapaj to, nie, nepapaj.

%eng Don’t eat that, no, don’t eat.

%add RIA

%com MAM starts addressing in English, but immediately switches to Slovak

*MAM That’s it! ## Wow! Wow! Wow! Wow wow wow!

%add RIA & VAN

*MAM Don’t! Nepapáme to. Moja. To nepapáme. Ukáž to, v pusine? Nemáš.

%eng We don’t eat that. Lovey. We don’t eat that. Show me, in your mouth?
Nothing there.

%add RIA

*MAM Look! Look!

%add RIA & VAN

64
*VAN There you go!

*MAM Á! Čo to?

%eng What is it?

%add RIA

*MAM Á! Čo to je? Jéj!

%eng What is it? Yay!

%add RIA

*VAN xx xx xx

*MAM Two-in-one crocodile keys, two ways to play.

%act MAM reads the packaging

*VAN And with that thing.

*MAM Yeah, that’s like a xylophone.

%add RIA and VAN

*VAN Or you can walk along, you can walk along with the stick and the eyes roll.

*MAM Ah, yeah!

%add VAN

*MAM Look! Look! That’s so cute, isn’t it? Isn’t that a great toy?

%add RIA & VAN

(0;5.24)

In this example the mother-child dyad and Ria’s monolingual English aunt were focusing
on the same activity. The mother addressed Ria and her aunt in English, but she switched
to Slovak when the focus was on parenting and therefore addressed to Ria only. On
occasions the mother did not switch from English to Slovak to address Ria immediately
after she had addressed the monolingual speaker. However, she realized this and repeated
her own utterance in Slovak, so as to model the TE to Ria instead. This example also
shows that the linguistic relationship was well established and the Slovak language was the
language of usual and natural communication in the dyad.

Three strategies were observed in these situations:

 addressing Ria in Slovak and then rephrasing in English;


65
 informing other speaker(s) about what was said, a strategy which emphasized that
although Ria was the addressee, she did not wish to exclude others;
 addressing Ria in English in the form of an invitation to start a conversation between
all present, used when the mother’s aim was to be inclusive, or to encourage Ria’s
interaction with others.

Another factor that further influenced the family’s language presentation approach was a
review of literature on childhood bilingualism, which allowed the mother to make
intentional choices that developed the initial free approach into a family strategy. Specific
strategies will be discussed in section 4.5.3.

In summary, the languages were presented to Ria in the 1P/1L approach in terms of
relative language exposure during the first two years, however, during the second year
there were increasingly more reasons to start separating language presentation according to
social contexts. While the mother continued to use only Slovak when addressing Ria
directly, there were social occasions when she was presenting both Slovak and English as
her two languages, therefore using the 1P/2L approach. English continued to be presented
as 1P/1L by the father throughout the study. A number of questions arise about the
suitability of OPOL as a classification for the family’s approach:

 What proportion of each language from the mother would be acceptable for this
approach to be classified as OPOL?
 If the bilingual mother showed comprehension of both languages to Ria, but responded
in one language only, is this viewed as 1P/1L or 1P/2L?
 If the monolingual father showed comprehension of some Slovak utterances used by
the mother, and responded accordingly, but did not speak Slovak to Ria or another
adult, can it be said he used the 1P/1L?

According to Barron-Hauwert (2004, p. 163), providing most of the time the parents
separate the two languages and limit addressing a child in the other language to certain
social situations, the approach can be classified as OPOL. Although OPOL was not
successful in achieving strict language separation as previously claimed in the literature
(Leopold, 1939), as a strategy it ensured Ria received maximum exposure to the minority
language that was limited to a few speakers. This was especially important in the situation
with one societal language, as was the case for Slovak and English respectively in
Australia in this case study.

66
Establishment of a consistent language presentation approach in the bilingual family
required some adjusting by the members over a considerable period of time. It appeared
crucial to establish a suitable approach for the family very early on to maintain that
approach in long term. In this way a positive shift occurred not only in actual language use
in the family, but in the bilingual mother’s perception of the monolingual father’s
experiences. It is also interesting to note it was the bilingual parent who felt unease at how
to accommodate all family members in the bilingual family, while the monolingual parent
had minimal difficulties adjusting to the patterns of language use.

Indeed, a number of diary entries mention the mother’s private conversations on issues
related to linguistic exclusion with several mixed-lingual families. They state most
families frequently experienced concerns over exclusion of the non-Slovak speaking
parent. Such concerns resulted in one of the parents abandoning consistent use of their
native language when interacting with their children, and the use of the monolingual
parent’s language or the language the parents had in common within the family (the
majority language that already received enough support from the wider community). This
approach neglected the importance of establishing a strong emotional bond with a child in
the minority language. It is also likely that the lack of impact belief in these parents caused
them to switch to the other language, and consequently the children came to prefer the
majority language. In most cases it proved very difficult to reverse such language shift in
the bilingual children. Difficulties associated with alteration of a set language pattern in a
linguistic relationship were also noted in research literature (Chin & Wigglesworth, 2007,
p. 14). According to Chin and Wigglesworth (2007) any two interlocutors have a
‘linguistic relationship’ that is based on a certain language. This relationship evolves
naturally and when the language pattern is set, it is difficult to alter. The importance of
person-language bond and the psychological implications of such bond for the child were
also emphasized by Kwan-Terry (1992, p. 258). Saunders (1984, pp. 129-130) also
stressed the status of the language used during the initial encounter with children. He
found it determined the language used in the future, since the children intuitively classified
the person as a speaker of that language. In contrast, bilingual parenting guides advise to
remain consistent and allow such issues to work out over time, stressing the parents’
impact belief as paramount (Pearson, 2008, p. 124).

However, the findings in this thesis indicate that in a bilingual family it is necessary to
establish linguistic relationship with a child as early as the antenatal stage and no later than
at birth, despite the suggestions that such relationships will naturally evolve overtime

67
(Pearson, 2008). An early establishment of linguistic relationships and discourse strategies
had a positive impact on language use and maintenance within the bilingual family, and
possibly in later life for the bilingual individuals. The language-person bond became
especially important for the minority language, since Ria did not receive input from many
speakers of Slovak. It was the establishment of person-language bond that allowed parents
to follow bilingual strategies such as 1P/1L and 1P/1L&1P/2L, since the child developed
emotional association of her languages with different persons. These findings support
Hoffmann’s (1991, p. 38) suggestion that “successful establishment of bilingualism may
well depend on psychological factors”.

4.3.1 Language presentation within domains of use


Typically there are two main domains for the use of language - private and public, with
many other subdomains such as person, topic, etc. (Chin & Wigglesworth, 2007, p. 14).
Bilinguals often use different languages in different domains. Cruz-Ferreira (2006, p. 26)
argues that a multilingual does not use several languages in exactly the same way every
day and for the same purpose, because if that was the case, one language would suffice.
While this argument is valid, for BFLA children languages cannot be as strictly separated
into domains of use. The distinction is not as clear since a very young child actually uses
both languages for the same purpose within the private domain. For a BFLA child the
purpose stems from the need to communicate with a different person in a different
language, since the family uses both languages for the same communicative purposes.
Therefore in early childhood the most important factor of domain becomes the person.
Often each parent interacts with a child about the same topic using their respective
languages. Although it is possible that during play or reading different adults may focus on
different aspects, the main argument remains that the distinction between the two
languages in terms of domains in a bilingual family becomes blurred.

This thesis reports on the first two years of life, and therefore a stage when to the
developing child individuals and their languages play a more important role than the
language of the larger society. Thus the analysis will focus on the private domain, with
some references to the use of languages in the public domains when necessary.

In the private domain Ria received input from the Slovak speaking mother during the
working week, and input in both languages in the mornings, evenings and weekends.
Exposure to English from the father was spread around short periods across different days,
with a maximum of approximately 3 hours a week.

68
Ria’s primary carer was the mother. An average amount of time spent in the mother’s care
alone was 60 hours a week. There were also periods of time when the father spent 2-3
hours per week as Ria’s carer, but overall during this period the mother remained the
primary carer for the rest of the week, with an average of 57 hours a week. Table 4.4
outlines the small changes in Ria’s care history.

Table 4.4 Ria’s care history in the home

Main Average
Language(s)
From Until caregiver/ number of
spoken to Ria by
(m, y) (m,y) Other hours per
caregiver(s)
caregiver week
Period September
July 2007 Mother 60 Slovak
1 2008
Period October November Mother/ 57/ Slovak/
2 2008 2008 Father 3 English
Period December February
Mother 60 Slovak
3 2008 2009
Period May Mother/ 57/ Slovak/
April 2009
4 2009 Father 3 English
Period June July
Mother 60 Slovak
5 2009 2009

It is important to note that although the periods when the father had the role of the carer
were short and limited in comparison to the rest of the week, it was the only time when Ria
received input in English only. At all other times Ria continued to receive input either in
Slovak or in both languages.

Ria’s overall language exposure during different periods is shown in percentages in Table
4.5, showing usual patterns, pattern changes and relative frequencies of exposure. This
calculation is based on weekly waking hours, which changed overtime. In the first 4
months waking hours totalled 35 per week, while from 11 months onwards total wake time
was 80.5 hours per week. The daytime waking hours in each period account for 100% of
language exposure time.

A weekly percentage of exposure to each language is calculated according to the number


of hours spent with parents or other speakers of each of the two languages who were
present. Thus there were 3 possibilities for language exposure: Slovak only, English only
and both languages in the same context. This would occur when both parents were present,
or when Ria and the mother were attending community based activities for toddlers, where
they were regularly interacting with English speakers. An estimate of relative frequency of

69
exposure as a proportion of total amount of input based on the waking times is presented in
Table 4.5.

Table 4.5 Weekly patterns of language context exposure

Period Slovak English Both


35h
Birth – 2 weeks
100%
12.5 h 1h 21.5h
2 weeks - 0;4
36% 3% 61%
32.5h 1h 29.5h
0;4 - 0;9
51.5% 1.5% 47%
35h 1h 34h
0;9 – 0;11
50% 1% 48.5%
80.5h
0;11 - 1;1
100%
35.5h 1h 44h
1;1-1;3
44% 1% 54.4%
33.5h 3h 44h
1;3-1;4
41.5% 4% 54.5%
35h 1h 44.5h
1;4-1;7
43% 1% 55%
33h 3h 44.5h
1;7-1;9
41% 4% 55%
80.5h
1;9 – 1;10
100%
35h 1h 44.5h
1;10-2;0
43% 1% 55%

In the initial stages, up to approximately 18 months, the English contexts consisted of


usual morning and evening routines with the father, whose time spent with Ria was limited
due to work commitments. On the other hand, since the mother was the primary carer, she
had more opportunities to model the Slovak language through varied shared routines and
games. A change in language exposure patterns took place during an overseas trip of 5
week duration, when exposure to the languages was equal.

Therefore, for example in the period from 2 weeks to the age of 0;4, Ria received Slovak
only input for a total of 12.5 hours, input from both languages simultaneously for a total of
21.5 hours and input from English only approx. for 1 hour per week. The division of carer
duties impacted on the kind of linguistic skill Ria developed in the two languages, as will
be shown in the following chapters.

Language exposure in the public domain is a complex phenomenon. Since the family
resided in Australia, Ria was exposed to English input in interactions in public domains
daily. The more meaningful contexts were when the bilingual family was spending time
70
with extended family, which occurred on average once a week throughout the study. In
these situations relatives addressed Ria in English, while the mother continued to address
her in Slovak, excluding the occasional situations when she used English, as described in
section 4.3.

Other regularly occurring social situations, such as outings to university, at the community
radio station, or during visits at other acquaintances accounted for short bursts of exposure.
Conversations between the parents and other interlocutors were not addressed directly to
Ria and consisted mainly of overheard speech. This is not to say that this form of input
was not important. As De Houwer (2009a, p. 102) claims overheard speech helps children
with some pragmatic rules, such as the use of pronouns, which cannot be seen in
conversations they are involved in. Moreover, in these situations Ria was learning
principles of bilingual behaviour.

Language use patterns were similar to interactions with extended family. The mostly
monolingual English speakers used English when addressing Ria, while bilingual Slovak
and English speakers tended to code-mix/switch before ascertaining with the mother that
Ria spoke Slovak as well. In all situations the mother continued to address Ria in Slovak,
with occasional switches to English where required by the social situation to accommodate
other interlocutors.

The main changes in the public domain occurred from approximately 0;9 when Ria and the
mother attended public social activities designed for toddlers. These kinds of situations
were more meaningful to Ria, since they were targeted at children of her age and other
children participated with their caregivers. Such activities included playgroup sessions,
swimming lessons, gym for preschoolers, library sing-along and story times, and were
taking place through the medium of English. However, even in such situations the mother
continued to address Ria in Slovak, while switching to English when addressing other
interlocutors. Ria was also receiving input from overheard speech from interactions that
took place between the mother and other interlocutors. Thus even in the monolingual
English public domain Ria was exposed to both languages. She received English input
from other speakers and input in both languages from the mother.

Nevertheless, the diary records document changes in the mother’s strategy for language
choice when interacting with Ria in the public domain. During Ria’s infancy the mother
addressed her in Slovak only, without concerns of excluding other monolingual English
interlocutors from the interaction. However, when Ria became linguistically expressive,
the mother became more conscious of excluding other people from their interactions,
71
thereby often switching to English when addressing Ria so as not to exclude others
present. The reasons for such changes are not clear; however, introspective observations
by the mother provided a possible interpretation. It appeared that a mother-infant dyad
formed a single entity whose interaction could not potentially offend other speakers. On
the other hand, a mother and a toddler capable of expressing herself linguistically were
separate individuals. The possibility that other interlocutors may be offended by exclusion
from interaction existed, since they might perceive such ‘exclusion’ as intentional. The
mother ascribed others the right to be participants in her interactions with a toddler, while
they did not have this right during her interactions with an infant. This indicates the
mother’s choice to be inclusive was triggered by the loss of privacy of interaction in the
mother-infant dyad, since the mother-toddler interaction was perceived more public,
requiring others to be included. Despite these beliefs the mother did not consistently
include others in her interactions with the toddler by means of language choice. She based
her choices on various factors ranging from emotional states to the other person’s level of
engagement in the context. Nevertheless, her choices on inclusion or exclusion were
modeling appropriate bilingual behaviour.

As Ria’s social network was extending, more frequent meetings with English speaking
monolingual acquaintances often presented situations when their inclusion was
appropriate. Such examples, however, were limited to English utterances at a time, and
were not applied to whole conversations. More frequently the mother moved between the
use of Slovak to address Ria, and the use of English to address the whole group, even if the
utterance appeared to be addressed to Ria.

In the public domain, the amount of Slovak input from other speakers was limited (apart
from one month the family spent in Slovakia). Throughout the study on average once a
week the mother and Ria spent a full day with the mother’s Slovak relatives, a family
where both parents were bilingual speakers of Slovak and English who spoke mainly
Slovak, but used free mixing as a usual communication strategy. In these situations Ria
received mainly Slovak input from the mother, and mainly Slovak input interspersed with
English code-mixing from others. The patterns of bilingual exposure are summarised in
Table 4.6.

Table 4.6 Exposure to bilingual language mixing

Family
Speaks to Speaks to Speaks to Speaks to Speaks to
member
Father Mother Ria Ria’s aunt Ria’s cousin(s)
(age)
Bilingual aunt English Slovak & Slovak - Both,
72
Family
Speaks to Speaks to Speaks to Speaks to Speaks to
member
Father Mother Ria Ria’s aunt Ria’s cousin(s)
(age)
(>35) mixed mainly Slovak
& mixed
Bilingual Both,
Slovak & Slovak & Slovak&
female cousin English mainly Slovak
mixed mixed mixed
(7-9) & mixed
Bilingual male Slovak & Slovak & Slovak & Slovak &
English
cousin (1-3) mixed mixed mixed Mixed

During interactions with the mother’s extended family the mother code-switched and
mixed her languages as in the following excerpt, where mixes are shown in bold:

%sit The two families are giving Christmas presents to each other

*MIS To má same as Romi Ria! Iné colours.

%eng Ria’s got that same as Romi. Different colours.

%add MAM

*MAM Kúk, pozri, aká rybka!

%eng Peek, look, it’s a fish!

%add RIA

*MIS Mami, Romi má iné colours!

%eng Mami, Romi’s got different colours!

%add NEL

*MAM Pozri, stacking cups!

%eng Look, stacking cups!

%add RIA

*NEL Nó!

%eng Yeah!

%add MIS

*MIS Aj Romi to má, ale Romi má different colours.

%eng Romi has that too, but Romi has different colours.

%add MAM

73
*MAM Aj Romi? Thank you!

%eng Romi too? …

%add MIS

*MIS Ano, ale different colours.

%eng Yes, but different colours.

*MAM Hej. Pozri, to je fish. Rybička. A tu je taká iná rybička. Pozri!

%eng Yeah. Look, that’s a fish. Fish. And here’s a different fish. Look!

%add RIA

*NEL Dúfam, že to nebude recalled, hey? Je to Fisher Price.

%eng I hope it’s not going to be recalled, hey? It is Fisher Price.

%add MAM

*MAM Shouldn’t be.

%add NEL

%com MAM makes a complete switch to English

*NEL Yeah, well, Fisher Price had been recalled.

*MAM Yeah I know, but,

*MIS Romi to má v piesku.

%eng Romi has that in the sand.

%add MAM

*MAM hopefully all the recalls are over now.

%add NEL

*MIS Romi to má v piesku.

%eng Romi has got it in the sand.

%add MAM

*NEL Let’s hope.

(0;5.25)

74
In this example the mother also adjusted her language choices to free mixing and even
made a complete switch into English in order to include the father in the interaction, thus
modelling switching and mixing to Ria as a communication strategy. Mixing with others
also triggered mixing when addressing Ria, however, she self-repaired a mixed utterance.
Such self-repair did not occur when addressing other bilingual speakers. While the mother
rarely code-switched when addressing Ria directly, Ria nevertheless was able to overhear
her mixing. In fact, due to the regularity of interactions with Ria’s Slovak extended
bilingual family, overheard speech was the main source of bilingual code-switching and
free mixing, and thus a plausible explanation for Ria’s bilingual language use.

There were also occasional visits among other acquaintances from the Slovak community.
Since language mixing is widely used among the bilingual Slovak community members,
the mother also tended to mix languages in her speech. Thus Ria was indirectly exposed to
language mixing as a normal bilingual communication strategy.

4.3.2 Linguistic soundscape (audio-media and other media)


Linguistic soundscape includes all forms of language exposure a person receives from
various sources. The term was suggested by De Houwer (2009a, p. 97) and it subsumes all
aspects of language exposure that play a role in linguistic development, thus it does not
limit input to persons that interact with a child regularly. Aspects such as audio-media
exposure add to the overall linguistic development and need to be considered as well.
Relatively equal audio-media exposure to both languages could enrich the development of
a language, while the lack of audio-media exposure in one of the languages may mean that
the language may lack the added support, since a child has fewer opportunities to hear the
language from varied speakers, or she may perceive the language as less useful.

According to the diary records in this thesis, indirect audio-media exposure to English
started early on due to English media of the wider community. However, Slovak media
such as children’s music CDs and community radio broadcasting were also used regularly
and intentionally by the mother to enhance Slovak language contexts. The mother started
engaging Ria in regular exposure to Slovak media as early as 0;2.13.

Ria was equally exposed to books in both languages, since sharing books was a favourite
activity from very early on. Initially the mother ‘read’ all picture books in the Slovak
language and translated simple texts into Slovak. She started reading English texts only
when Ria became aware of the fact that books were in different languages and explicitly

75
requested a story to be read in English. Thus during the study Ria received relatively equal
exposure to audio-media and written media in both languages from early on.

4.4 Language attitudes and impact belief in the family


Studies found the main reason parents choose bilingual parenting was personal experience
with language learning, followed by knowledge from popular parenting literature on the
topic, and lastly other examples from extended family (King & Fogle, 2006, p. 706).
Successful development and maintenance of bilingualism in a family is also dependent on
several factors. Three major factors were found to effect language choices within a
bilingual family, and impact subsequent development and maintenance of both languages:

 family language attitudes


 generation
 social class (Eilers, Pearson, & Cobo-Lewis, 2006, p. 83).

This study suggested that long-term consistent exposure to both languages in the family
was necessary for successful bilingualism. According to Zurer Pearson (2008, p. 125) it is
also important to make the language attractive, valuable and indispensable to children in
order to insure interest to learn, thereby feeding children’s own positive attitude towards
the language and bilingualism. De Houwer (2009a, p. 92) defined another important factor
in successful development of bilingualism, impact belief - the parents’ belief that they can
have an effect on a child’s linguistic development.

Positive attitudes allow the speakers to assign positive values to both languages, and to
freely make language choices in communicative settings, which in turn could contribute to
the development of successful bilingualism. However, few parental guides suggest
possible solutions and discuss the importance of parental attitudes and impact belief on the
outcome of bilingual parenting (Cunningham-Andersson & Andersson, 1999; De Houwer,
2009a; Pearson, 2008). This aspect was not described or addressed in the research
literature.

An individual’s language attitudes are the result of their linguistic experiences and
background. Parental language attitude is transferred to a child indirectly through their
speech behaviour, with the parents’ language choice being the most obvious display. Thus
through daily interaction a child is explicitly exposed to language ideologies and parents’
attitudes towards the two languages. Positive attitudes can be demonstrated to a great
extent by a consistent use of the languages, and can be enforced through discourse
strategies that require the use of a particular language in given social situations. It therefore
76
seems necessary to establish strategies that encourage bilingualism early on, well before a
child needs to make her own (subconscious or conscious) language choices. The process of
establishing such strategies in this thesis will be described in section 4.5.3.

The mother’s positive attitudes towards languages and bilingualism stemmed from her
experience with foreign languages as well as from her professional interests. She was
embarking bilingual parenting with good understanding of the phenomenon of
bilingualism and with knowledge on various language strategies for families. From the
prenatal stage the mother was applying her positive attitudes to establish a suitable strategy
for the family. It was a process of using a strategy, internal reflection on the strategy and
balancing between maintaining the chosen approach and accommodating all family
members both linguistically and emotionally.

The mother encountered two main concerns regarding bilingual parenting in the initial
stages. During pregnancy she was concerned that she would not find it natural to speak
Slovak to her child, since Ria was born in a relationship between the mother and the father
that was based on English. On the contrary, when Ria was born the mother found
addressing her in Slovak instinctive.

Initial concerns were also related to situations when all three family members were
present. The mother felt that there were two possible negative consequences of parenting
in a language other than the father’s language: she would have to either exclude the father
from interactions when using Slovak, or neglect the use of Slovak if opting for English in
these situations. On these occasions the mother had to consciously develop a language
choice strategy, and chose to use Slovak when addressing Ria in family situations, thus
creating a bilingual environment when both parents were present. As a result the family
came to feel comfortable with the use of both languages in such situations, and the mother
in particular became comfortable using Slovak when addressing Ria in father’s presence.

The father’s attitudes to languages and towards bilingualism remained positive despite
being unsuccessful in his attempts to acquire a second language. Moreover, the father
always regarded the ability to express oneself in two languages desirable, and believed a
child could acquire two languages in naturalistic context with minimal effort. In fact, King
& Fogle (2006, p. 704) found having experienced missed opportunities in language
learning can have a positive effect on parents’ goals for their children. The father was
supportive of the family’s bilingualism from Ria’s birth and accepted the strategies used
by the mother, later forming his own strategies as Ria was becoming linguistically
expressive.
77
The father’s only concern was related to Ria’s uneven amounts of input in the two
languages (see section 4.2). Initially he questioned whether Ria who spent most of the day
with a Slovak speaking mother would be sufficiently competent in English in order to
communicate with him. However, such concerns were minimised at a very early age, as
soon as Ria showed comprehension of both languages, and as she became linguistically
expressive the concerns were disregarded. Furthermore, the father was also able to learn
some Slovak in natural everyday settings. He was exposed to the kind of input that a child
receives when acquiring their native tongue, which allowed him to start learning language
in naturalistic interactions, not as a foreign language learner. During the second year of
Ria’s life he often attempted to respond appropriately to Slovak utterances produced either
by Ria or the mother, and occasionally reproduced Slovak utterances himself. This form of
receptive language learning allowed the father to be indirectly included in most
interactions between Ria and the mother, which in turn created a positive environment for
bilingual language use.

Despite diverse language backgrounds both parents in this thesis had a positive belief that
the use of their languages could impact Ria’s linguistic development. The mother’s
decision to address Ria in Slovak only so as to emphasize the preferred language and to
maximize the amount of input in the minority language, and the father’s perception of
language acquisition in naturalistic environments reflected their impact believes. These
findings indicate that early establishment of consistent language exposure pattern helps to
foster positive attitudes towards bilingualism.

The parents’ experiences with languages other than their native tongues in earlier life
meant they had positive attitudes towards languages in general, towards Slovak and
English in particular, as well as towards childhood bilingualism, even before Ria was born.
Both parents believed they can positively influence Ria’s language acquisition. The
mother’s professional interest in linguistics led to fostering of positive attitudes in the
relationship well prior to Ria’s birth and in the bilingual family afterwards. The linguistic
environment the parents were able to create with their positive attitudes towards
bilingualism showed to be conducive to Ria’s development of bilingualism.

4.4.1 Attitudes outside the family


Attitudes of extended family and other relatives, including English and Slovak
monolingual speakers, towards Ria’s bilingualism were positive from the start. They never
expressed concerns regarding possible negative consequences of bilingual upbringing.

78
Likewise, they did not object to the mother’s exclusive address of Ria in Slovak, or the
father’s address of Ria in English, and delighted in observing Ria interact with each parent
in a different language.

Overall it was accepted by monolingual English and Slovak speaking relatives that they
were excluded from conversations within the mother-child and father-child dyads. On
occasions English speaking relatives asked for an interpretation of Slovak utterances,
which was provided by the mother. The monolingual father was not able to provide
interpretations of English utterances to Slovak speaking relatives, thus the mother took up
the role of informing those who did not understand if necessary.

Bilingualism was also viewed positively by more distant bilingual relatives in Australia
and Slovakia. Australian relatives bilingual in Greek and English supported Ria’s bilingual
development even though Slovak was not one of their languages. Having experienced
bilingualism they understood the benefits, and also delighted in her ability to alternate
between languages appropriately.

Throughout the study Ria did not show signs of concern or discomfort to use either
language in front of monolingual speakers who might feel that they were excluded from
the conversation. This is interpreted as a result of the mother’s approach to continue using
Slovak when addressing Ria even in the presence of other monolingual speakers.
Similarly, the fact that the mother and Ria did not face disapproval of their use of Slovak
from another person meant that no concerns arose about addressing the mother in Slovak.
Thus overall, positive attitudes towards Ria’s bilingualism, mother’s use of Slovak in front
of monolingual English speakers and general positive experiences supported Ria’s
bilingualism.

When attitudes of the wider community are considered, Ria was growing up in a situation
that was somewhat deterring in terms of perceived values of her languages by the wider
society. It appears the Australian society values bilingualism as an important life skill.
Each state implemented a language-in-education policy which recommends learning of
languages other than English throughout compulsory years of education (Clyne, 2005, pp.
158-159). However, this is in selected languages of economic importance and no formal
opportunities to develop minor community languages equally are provided to families.
Thus minority language maintenance in Australia is largely dependent on the family’s
impact belief and desire to maintain the language.

79
4.5 Language models
According to Lanza (2004, p. 249) when considering the role of input in the study of
language acquisition three aspects need to be considered:

 Linguistic forms used by native speakers with the child


 Manner of presentation of the input – how are the forms made available to the child
 Metalinguistic input – feedback on the child’s productions.

The extent of the effect particular forms of language can have on a child who is exposed to
monolingual language acquisition was described by Tough (1977, p. 61):

“The particular forms and patterns of language to which the child is exposed carry values
and attitudes, and serve to focus the child’s attention on to aspects of his experience that
those who talk with him see as significant, or that are unchallenged tenets of a way of
life.”.

It can be argued that in a bilingual family the effect of the forms of language is even
stronger, in particular in cases where a minority language is associated with a limited
number of people. Few individuals have a greater opportunity to have an effect on the
language development than a wide social network of language speakers (De Houwer,
2009a, p. 101).

An important circumstance in Ria’s experience was the fact that the mother represented
one of few sources of Slovak. Therefore the mother’s speech style and the particular
variety of language she used significantly affected the variety spoken by Ria. The mother
was not the only Slovak speaker, but interactions with other Slovak speakers outside the
home were considerably less common than interactions with English speakers. Conversely,
the variety of English speakers that Ria received in her input meant that the father’s speech
style and variety of language was not as important in shaping Ria’s English.

In addition, language models in bilingual families provide information about their


preferred language use. Most of the English models were monolingual speakers. The father
modelled monolingual language behaviour, with rare diversions from this general pattern.
The mother, on the other hand, provided a model of varied bilingual language behaviour.
At home alone Ria observed the mother in Slovak monolingual contexts, as well as
bilingual contexts when alternating between languages in triadic family interactions.
Bilingual contexts in which the mother used both languages also appeared in various other
social situations with interlocutors outside the family. Within such contexts the mother
engaged in interactions that were not addressed to Ria, but still provided a bilingual model.
80
Thus overall the mother provided a bilingual model for both languages, while the father
provided a model for English only. The main point is that none of Ria’s Slovak language
models were monolingual, and not all of Ria’s main English language models were
typically monolingual.

4.5.1 Ria’s languages-in-acquisition: Slovak and English


The variety of Slovak used by the mother was standard Slovak spoken in the Bratislava
region, influenced to some extent by ‘Australian variety of Slovak’6. While there were
some differences in accent and intonation patterns (as influenced by the Australian
English), on phonetic and morpho-syntactic levels the variety was most likely not different
from the standard Slovak used in the Bratislava region, and in the media in Slovakia. The
father spoke Australian English variety spoken in the Adelaide region (Burridge &
Kortmann, 2008).

There are morphological differences between the two languages, which can be best
highlighted by a simple description and comparison. Slovak belongs to the West Slavic
family of languages and it is a highly inflectional (or synthetic) language. The different
functions of words are expressed primarily by the use of inflectional affixes attached to
lexical word classes such as nouns, verbs, adjectives, numerals and pronouns. These
affixes express grammatical relations such as case, number, gender, person, tense, mood,
degree, as well as being used for derivation of words. Word order is then free, although
most commonly SVO is found. To express emphasis on a particular word it can be placed
in the sentence initial position. In clause structures dependent clauses are numerous (Sgall,
1999, p. 55).

English, on the other hand, being an isolating (or analytic) language uses function words in
combination with lexical items to express relational meanings. Since function words
express the grammatical relationships, free word order is not possible, with English
requiring a set SVO order. Syntactically, derived clauses with the use of conjunctions are
common.

Another contrast is in the prosody and rhythmic differences. In the Slovak language, word
stress is usually on the first syllable, except when the word is linked with a preposition,
then the stress is transferred onto the preposition. However, in a stream of speech word
stress is only potential and it is typically not realized with each word, only with those
words that carry the main meaning of the utterance. The emphasis is thus placed on the

6
To date there is no formal linguistic description of this variety of Slovak.
81
syllable of the word which the speaker wishes to emphasize in the utterance (Mistrík,
1988, p. 37).

In English, on the other hand, the word accent is typically realized on all lexical items of
the utterance, usually either on the first or the second syllable, while the functional items
tend to be unaccented (Finch, 2005, p. 197). This difference accounts for different rhythm
of utterances in the two languages, with English being a stress-timed language with regular
falling and rising accent, whereas Slovak utterances tend to have a less fluctuating rhythm
with emphasis on a particular word. Sentence intonation also differs, with English
intonation rising and Slovak intonation falling at the end of the sentence, while rising
intonation in Slovak signifies a question.

4.5.2 Child Directed Speech (CDS)


Child directed speech (also termed parentese) is a form of speech used by adults when
addressing children. CDS has distinct speech patterns and was widely observed and
described in the literature. It is used by speakers of numerous languages as a special form
of speech found when adults communicate with children, and appears to be universal (de
Boysson-Bardies, 2001, p. 90). Although different cultural variations and cross-linguistic
differences were found in the forms of CDS, some common characteristics were observed.
Oksaar (1983) describes the typical features in Western societies as following:

 Typical intonation & other paralinguistic patterns, including more slowly spoken
speech, clearer articulation, exaggerated intonation contour and pitch, higher tone of
voice
 Phonological and grammatical modifications including avoiding difficult consonant
groups, using mostly two-syllable words, simple syntactic structures and duplication of
syllables and words
 Use of 3sg when referring to oneself or a child in direct speech
 Use of associative ‘we’
 Diminutives and palatalisation – particularly prominent in Slavic languages
 Use of baby words instead of standard words, which are often of onomatopoeic nature.
 Use of expansions.

According to Kuhl et al. (1997, p. 684) CDS is modified syntactically, semantically and
prosodically. They found that especially vowel sounds were exaggerated, allowing infants
to perceive differences in vowel categories. This form of speech was preferred by infants
and it was suggested it provided necessary information about the sound system of the
82
infant’s native language. In the case of a bilingual child, CDS used by both parents would
therefore provide sound information on both languages, although no studies to date
considered this aspect in bilingual infants.

In the English literature on CDS a distinction is made between motherese, which refers to
changes in voice and prosody, and baby-talk referring to simplifications of vocabulary and
syntax (de Boysson-Bardies, 2001, p. 82). According to de Boysson-Bardies (2001, p. 84)
the typical prosodic characteristics of motherese remain until the third year, however, the
style and content of the speech are reshaped reflecting the child’s developments.

In recent years the importance of CDS in language acquisition was criticized, however
such critique overlooks that the purpose of CDS is not to teach language to a child. De
Boysson-Bardies (2001, p. 83) argues:

“The first vocal messages - which are intended, on the one hand, to capture the child’s
attention – convey affective values through melodic contours.”

Thus while CDS alone cannot explain how languages are acquired, it emphasizes the focus
on context-based interpersonal communication in the parent-infant dyad. The ‘here &
nowness’ nature of CDS is embedded in the context as the parent verbalizes what the child
experiences – what she sees, hears or feels in a way that engages the child’s attention, and
thus makes language especially relevant to the child. Thus through interpersonal
interaction CDS leads language acquisition as a catalyst (Matychuk, 2005, p. 316).

The presence of CDS in the parental speech in the data in this thesis cannot be ignored.
The above features were present in both parents’ address to Ria. CDS was observed
predominantly in the private domain, especially the tone of voice and use of diminutives.
In the public domain generally simpler language was used, although it was not necessarily
the ‘tender’ language of the private domain.

The main features of the mother’s CDS in the early months were the use of soft or higher
pitched voice, repetitions, mirroring and assigning meaning to Ria’s vocalizations,
providing commentary on Ria’s movements and sounds, maintaining Ria’s attention on
objects she was looking at by providing a commentary and descriptions, and providing
commentary on the mother’s own actions. Questions were also addressed to Ria but did
not require response. The mother referred to self in 3sg, using the appellative ‘mami’
“mummy”, and referred to Ria in 3sg using her name, as well as 2sg. She also used the
associative ‘we’ when providing commentary on activities. The following example shows
some of the features of CDS as used by the mother:
83
%sit Ria is on her change table, MAM is playing with her

%act MAM walks up to RIA singing, then she starts tickling RIA while making
playful sounds to make her laugh

*MAM á! Brušiačik! Áaá Mamin mamin mamin! Mamin. Brušiak. ## Brušiak!


Brušiak mamin mamin!

%tra Tummy! Ah, mummy’s mummy’s mummy’s! Mummy’s. Tummy. ##


Tummy! Mummy’s mummy’s tummy!

%com diminutives and repetitions of short phrases

*MAM No čo ty?

%tra Well, what is it?

*RIA [aaː aː!]

*MAM [aːa]

%com mirroring RIA’s cooing

*MAM Ty moja! Ty moja. Ty moja. Moja, mamina.

%tra Mine! Mine. Mine. Mine, mummy’s.

*RIA [aːa. aːa]

...

*MAM Mamin nos papáš? Spapáš mame nos?

%tra Are you eating mummy’s nos? Are you going to eat mummy’s nose?

%com repetition

*RIA [aaaː]

*MAM Mama ti spapá pršteky.

%tra Mummy will eat up your toes.

%com diminutives

*RIA [aːaːaː! aːa! aː!]

*MAM Jaj. ## Ham. Ham. Hám. Hámamamamam.

%tra Oh. Munch. Munch. Munch. Munch munch munch.

%com baby-word
84
*RIA [aːa!]

*MAM [aːa aaa!]

*RIA [aːa a aːaː!]

*MAM [aː aː!]

%com Mirroring RIA’s cooing

*MAM Ťa spapám od ľúbosti! Hamamamamamamamam. ## hamamam.


Hamamamamam. Ham.

%tra I’ll eat you all up from all my love (formulaic expression)! Munch munch
munch munch.

*MAM Hó! Buďabuďabuďa!

%com Inventing interjections

*RIA [a! aɛːa! aːaː!]

*MAM [aːa aːa aː]

%com Mirroring

*MAM raspberry sound

*RIA [a a aa]

%act RIA is producing these sounds with a hand in her mouth

*MAM Jée! Ty vieš stáť!

%tra Yay! You can stand up!

%com describing actions

*RIA [aːa]

*MAM Ty vieš takto stáť, šikovne? Ty vieš takto stáť na tých tvojich nožulcách
tvojich šunčičkových? Ty vieš takto stáť na nich?

%tra Can you stand up like this. Clever? Can you stand like this on your chubby
wubby leggies? Can you stand like this on them?

%com repetition and expansion, diminutives

(0;6.13)

85
In this example several CDS features are present. Prominent is the use of high pitched tone
of voice with exaggerated intonation. The speech is context related. The mother is
mirroring Ria’s vocalizations, and this exchange is interspersed with her commentary on
Ria’s actions, as well as her own actions. She is referring to herself in 3sg. Words and
whole short phrases are repeated, and the mother asks questions without expecting
answers. The use of onomatopoeic baby-words and diminutives is evident as well.
However, the fact that diminutive suffixes are used to such an extent means that some
words are actually relatively long and contain many consonant clusters with palatal
sounds, typical of the Slovak diminutive suffixes. This means that difficult and long
combinations are used rather than being avoided.

Both parents used several CDS features, although there were differences. Changes of tone
were not used as often by the father as by the mother. The mother also used considerably
more diminutives since they are typical of CDS in Slovak. English diminutives are more
limited, however, the family invented some diminutive forms by affixing the –ie suffix to a
noun, such as ‘cranky wombie’ (‘cranky wombat’ referring to Ria’s grizzly moods),
‘coughie’ (referring to Ria’s typical coughing sound when she spotted an object she liked).

Main differences were noted in the style of interaction each parent employed with Ria.
Both parents reacted to Ria’s vocalizations and allowed her to follow her own interest, but
the form of talk with Ria was different with each parent. The mother talked considerably
more than the father and used a lot of repetitions of her own utterances. She allowed Ria to
follow her interest and then provided descriptions or commentary on Ria’s actions or
activities, thus providing her with the linguistic form of what was happening in the
context. Thus she was attempting to convey Ria’s attention as well as to gain reaction from
Ria, as in the above example.

The father on the other hand attempted to gain Ria’s interest. If Ria was not interested, he
started an activity himself, which triggered her interest. She gave clues in the way she was
looking at the father and vocalized, as if wanting to initiate communication, as shown in
the following excerpt:

%sit DAD is playing with RIA on the floor in her room. He is showing her a
hand puppet but RIA is looking at a book

*DAD Hello! Hello Ria! Tutututu! Tutururu! Hello! Tup tup tup tup! Tututututu!
Hello!

%com Using interjections and actions in attempt to engage Ria


86
%act DAD pretends the puppet is biting RIA, but she is not interested. DAD puts
the puppet down and cleans his glasses. RIA looks at him.

*RIA [maː]

*DAD [maː] what?

%com mirroring

%act RIA continues looking at DAD

*DAD What’s the matter? Cutie pie! You reading your book? Snowy? The little
kitten? Good book? Good book, isn’t it? Yes it is!

%com diminutive, description, expansion, repetition

%act DAD picks up a toy vacuum cleaner, RIA picks up the lion hand puppet and
they are looking at the objects independently

%act RIA looks at DAD

*RIA [a]

%act RIA becomes interested in the toy DAD is looking at and she crawls up to
him, looking at him as he is trying to figure out how to use it.

*RIA [waɟ]

%glo yyy

*DAD Ria’s Dyson.

%act DAD pretends to vacuum RIA and she smiles

*DAD Cutie!

(0;9.18)

Another difference was in parental responses to Ria’s initiation of interactions. The father
responded to her cooing by pretending to hold a conversation, he nodded his head and
responded with ‘Yeah’ as if agreeing with Ria, encouraging her to continue cooing. He did
not alter the intonation contour or tone of voice during such exchanges, making the
interaction sound phonetically more like adult speech. This form of interaction was also
observed when Ria was babbling extensively, especially when she was herself trying to
‘hold a conversation’, babbling and pausing much like in a dialogue. The father interpreted
Ria’s babbling as words, assigning a meaning to them based on the situational context.
When Ria’s babbling changed and the sound patterns resembled words, often English
87
words, the father targeted these comparable sound patterns and pretended they were words
uttered by Ria. Thus he assigned linguistic forms and meanings to sound patterns that
unintentionally resembled an adult word, assigning a specific meaning to Ria’s babbling
vocalizations. He also kept a dialogue with Ria if she continued babbling, as if she was
responding to his utterances with adult speech.

CDS changed considerably over the two years of the study with changing contexts and
Ria’s developing communicative abilities. Initially the parents used mainly statements,
descriptions and questions that did not require answers. When Ria was able to read
communicative intentions, the parents included questions that anticipated Ria’s response.
As Ria became more mobile and interested in her environment, imperatives became
prominent. With Ria’s developing comprehension and production, the tone of voice and
intonation contours of the mother’s speech approximated adult directed speech. Towards
the end of the first year CDS included more expansions, especially by the mother. Apart
from commentary and labelling of objects and actions, she provided further descriptive
terms, such as colour, size, shape and function. She also related the content to Ria’s
previous experiences. The following example demonstrates these features:

%sit RIA and MAM are playing on the lounge floor, RIA is rolling around and
playing with objects found around the room

*MAM No čo ty moja? Čo si to našla na tom stole? Čo si tam na tom stole našla?

%tra What is it love? What have you found on the table? What have you found at
that table?

%com repetition, the question seeks Ria’s reaction

*RIA [aaː a]

*MAM Takú misku drevenú?

%tra A wooden bowl?

%com labeling

*RIA [aa]

*MAM S takou si sa ešte nehrala, však? S takou miskou?

%tra You haven’t played with one like that before, have you? With a bowl like
that?

%com relating
88
*RIA [ɛəːə] [mː mː mː]

*MAM S takou si sa ešte nehrala, čo?

%tra You haven’t played with one like that before, have you?

%com repetition

*RIA [əːmː]

*MAM mm! Ešte nie? No čo? Moje dievčatko!

%tra Not yet? What is it? My little girl!

%act RIA starts fussing as she is trying to get a feed

*MAM Dievčatko mamičkine? No čo ty?

%tra Mummy’s little girl? What is it?

%act RIA fusses

*MAM No čo? Dievčatko! Ty moja! Moja!

%tra What? Little girl! Mine! Mine!

%com repetitions and diminutives, maintaining attention

%act RIA is crawling away

*MAM Ria poď sem, poď! Poď sa sem pozreť, poď. Pozri aké knižočky tu sú!
Všetko možné tu je. Pozri sa, a tu je takáto miska hore. Fíha! To ťa
nezaujíma?

%tra Ria come here, come! Come have a look here, come. Look at these books!
All different things are here. Look, and up here there’s this bowl. Wow!
You’re not interested in that?

%com imperatives, engaging interest

*MAM Mami tam má oriešky! Mami chrumká oriešky. Lieskové oriešky.

%tra Mummy has nuts in there! Mummy’s crunching nuts. Hazelnuts.

%com expansion

%act RIA crawls up to MAM

*RIA [mːm mː]

*MAM Chrum chrum!

89
%tra Crunch crunch!

%com expansion, engaging interest

(0;8.11)

In this example the mother firstly related the new object (wooden bowl) to Ria’s
experience, then explained the function of the object (used as a container for nuts), which
triggered further expansion (mother eats the nuts, new label for the type of nuts and
onomatopoeia for the sound produced when chewing). Imperatives and questions
anticipated some response from Ria - the mother was calling her to come closer and
attempted to engage her attention.

In the second year CDS changed further as Ria became familiar with daily routines and her
comprehension repertoire included words and expressions associated with such routines.
Questions requesting to identify objects, people and animals were used, such as

‘Kde je xxx/ Where is xxx?’

‘Prosíš si/Do you want xxx?’;

as well as open invitations to actions and activities:

‘Pomôžeš mi xxx/ Can you help me xxx?’.

A question aimed at comprehension ‘Čo je to?’ “What is it?” also emerged, but did not yet
expect Ria to provide an answer, since the parents provided the answer themselves. They
were, however, modelling how to request labels for objects. Comprehension related
features of CDS were modelling specific language for Ria to reproduce.

The structure of parental utterances was also changing. Utterances were longer and
contained more explanations with more complex phrases. Holophrases were still used by
the parents relatively often, but their purpose was now limited to teaching and modelling
of new words.

The parents were giving Ria more space for active communication and response, aimed at
allowing her to demonstrate comprehension of speech. They gave Ria time to respond to a
request or invitation to do something by pausing for a moment before trying to repeat the
utterance, especially if Ria was focusing on different activity, to allow time for change.

Ria’s early utterances and attempts at producing adult words were mirrored. Through
parental mirroring Ria’s productions were supported and encouraged with positive
feedback. This strategy was used especially by the mother, and it started by mirroring

90
Ria’s cooing and babbling vocalizations, thus actively encouraging her to use more
vocalizations in response. When Ria started producing linguistic forms the mother
mirrored Ria’s utterances mostly to confirm that Ria’s intended meaning was understood,
but also as a form of acknowledgment that the mother had heard the utterance, even in
instances when there were no doubts as to the intended meaning. Thus the mother was not
only showing Ria that the forms she had chosen to express were correct and meaningful,
but she also modelled a particular linguistic behaviour – repetition after the other speaker.
It is likely that the imitative behaviour itself was often imitated by Ria, and thus when the
mother had produced an utterance, this was the reason Ria would reproduce it. In fact, first
words were initially reproduced after the adults, and then used as spontaneous productions.

Thus both parents interacted with Ria in a responsive way, oriented towards maintenance
of conversation with her. Especially the mother’s more prominent use of CDS in Slovak
was largely child-centred. Döpke (1986, p. 504; 1988, p. 110) found that similar linguistic
realisations of child-centred behaviour, such as acceptance of the child’s attempts at
interaction and maintenance of verbal interaction, especially by the minority parent, were
among factors influencing acquisition of the minority language.

4.5.3 Parental language choice and code-switching


Input patterns shape the bilingual child’s use of the languages. How parents chose to
present their languages gives children important cues on acceptable and preferred language
choice. Parents also model when code-switching (CS) is considered appropriate and when
it is not.

In section 4.3 language presentation patterns were analysed, showing that the two
languages were presented to Ria both in monolingual contexts, as well as in bilingual
contexts with occasional instances of code-switching. Situations when code-switching
occurred where then analysed qualitatively following Goodz (1989, p. 31). Input aspects
such as who initiated mixing, extent of mixing by each speaker, triggers for mixing, and
parental response to child initiated mixing when addressing parents were considered.

Parental CS in the data was influenced mainly by the interlocutor and activity. The
bilingual mother was frequently required to make an appropriate language choice. When
interacting with Ria, her aim was to model Slovak in a variety as close as possible to the
variety used by monolingual speakers of Slovak. CS when addressing Ria was therefore
not frequent and was limited to specific contexts in several instances:

91
 Occasionally Ria’s mix triggered maternal code-switching. The mother followed Ria’s
mix and repeated it in her response by integrating the mixed word or phrase into the
base Slovak language. However, she realized the error and self-repaired by recasting
the utterance in the appropriate form in Slovak.
 In bilingual contexts, when the whole family was interacting together, the mother was
in a bilingual language mode and occasionally initiated code-switching when
addressing Ria in English. However, since the relationship with Ria was established
through the Slovak language, an address in English felt unnatural and she self-repaired
in an attempt to model the appropriate Slovak form.
 In bilingual contexts the mother addressed Ria and the father simultaneously in
English.

Thus the bilingual mother rarely initiated CS when addressing Ria, except for bilingual
contexts when the trigger was bilingual language mode, or when there was a need to
address both the father and Ria simultaneously.

However, for the mother CS was the normal form of interaction with other bilingual
speakers within the domains of the Slovak language in Australia, as discussed in section
4.3. CS allowed for a natural flow of communication among bilingual speakers, and
interactions with CS as appropriate language choice accounted for a proportion of
overheard speech for Ria. Conversely, when interacting with Slovak speakers from
Slovakia the mother made a conscious effort not to CS – she strived to use the variety of
language that would be considered appropriate by those interlocutors.

The monolingual father did not have the opportunity to alternate language choice, thus his
use of CS was limited to several instances of playful use of some Slovak expressions in his
speech and to borrowing of family words, such as ‘bábo’ “baby/dolly” or ‘book-ina’ “little
book”.

Thus the main language models, the parents, did not model code-switching as an
appropriate language choice pattern when addressing Ria. The bilingual mother provided
Ria with some patterns of CS when appropriate in domains of the Slovak language, and
through simultaneous address of two speakers in triadic contexts. The monolingual father
did not model CS in English contexts.

However, parental language presentation is not the only factor influencing the child’s
language use. How parents react to the child’s mixing, and how they negotiate appropriate
language use can shape the child’s choices. Such strategies may not be intentional, but

92
they play an important role in conveying which language the parents perceive as
appropriate, as discussed in the next section.

4.5.4 Parental discourse strategies


The process through which a child learns the cultural behaviours, namely the expected
bilingual behaviours in her bilingual language environment is called bilingual
socialization. It is modelled by the people in the child’s environment. While to date aspects
of language socialization were explored extensively in the field of monolingual first
language acquisition, the study of bilingual socialization in BFLA was limited to language
choice and mixing.

In a discussion on language socialization Ochs & Schieffelin (1995) discuss cross-cultural


differences in approaches to child language and different expectations of children’s
linguistic behaviour in various social situations. When interacting with children parents
use various culture specific discourse strategies. Moreover, the kinds of strategies used
change overtime with a child’s developing linguistic abilities.

The family in this thesis perceived Ria as an active participant in interactions, expecting
her to be an early interlocutor from birth (Ochs & Schieffelin, 1995, p. 78). They conveyed
these expectations of Ria’s role through several discourse strategies:

 Addressing Ria directly, describing actions and activities, labelling objects,


commenting on situations or voicing their affect
 Interpreting and talking about Ria’s emotional states or actions
 Interpreting and verbally expressing Ria’s sounds, actions or other signs of meaning
making
 Voicing Ria’s address of/response to interlocutors on her behalf, well before she was
able to speak, in interactions between the parents directly addressing each other, but in
a manner that one parent was inferring what Ria might be saying. Although they used
1sg, they indicated that they had assumed Ria’s position by speaking in a higher pitch
of voice, as if impersonating her, to voice utterances on her behalf. The parents were
actually expressing their own thoughts rather than interpreting Ria’s sounds or actions.
They were assigning these to Ria, thus including her as an active interactional partner
in conversations.

The following example demonstrates the discourse of voicing Ria’s thoughts, shown in
bold:

93
%sit The family is sitting in the lounge, chatting about Ria’s vocalizations

*MAM Ty rozprávaš! So talkative this morning!

%add RIA , then DAD

*DAD Yep, she doesn’t say much.

* RIA [ʁ ʁ]

*MAM What do you mean doesn’t say much?

*DAD She’s not making different sounds.

*MAM Yeah. Just a content…

*RIA [aqʁ]

*MAM She’s gotta explain why she was up all night, and why she woke up at 5:30.

%add DAD

*DAD She sure does.

*RIA [gʁ gʁ]

*MAM I was growing daddy, I was growing!

%add DAD & RIA

%com With a higher pitch voice MAM is assuming RIA’s role and assigning a
meaning to RIA’s cooing

*DAD Couldn’t sleep?

*MAM You even said it yourself this morning how big I was all of a sudden.

%add DAD & RIA

%com MAM is assuming RIA’s role and voicing on her behalf

*RIA [gʁ gʁ]

*MAM She’s very chatty.

%add DAD

*RIA [gʁ gʁ]

*DAD She’s on the camera, that’s why.

* RIA [ʁ]

94
*MAM [aqʁ]

%com Mirroring

(0;4.12)

In this example the mother assumes Ria’s role by using 1sg and a higher pitched voice to
mark this utterance as Ria’s. She is not interpreting her cooing sounds, but inventing
meaning that would make sense within the interaction taking place between the adults.

Other specific forms of interaction were various social routines and games played with
Ria, such as traditional rhymes and songs used with children by native speakers of the two
languages. During familiar routines the parents expected Ria to anticipate what followed.

When Ria produced speech new parental strategies emerged through which she was
regarded an equal interactional partner:

 Repeating the word Ria attempted to produce in order to model the correct adult form
 Praising Ria’s attempts at using adult words to encourage further speech production
 Expanding on Ria’s utterances, thus providing additional information.

These strategies are shown in the following example:

While helping with the washing Ria pointed at father’s shorts and uttered [∫o:]; the
mother responded ‘Ano, to sú daddyho šortky’ “Yes, that’s daddy’s shorts.” (1;7)

Thus the mother acknowledged Ria’s attempt to produce an adult word, modelled the adult
form and expanded on the holophrase.

In a bilingual family discourse strategies play an important role in successful development


and maintenance of both languages. Bilingual families make a choice about which
languages are used with which person and in what situations. These expectations need to
be conveyed to children through daily interactions. Parents achieve this through different
responses to child’s language mixing. They respond by indicating what is appropriate
language use and negotiating a specific language context with a child, giving her
opportunity to co-construct the context in her responses to their strategies (Lanza, 2004, p.
261). Several authors provided detailed analyses of such parental responses to child mixing
(Hiroko, 1998; Lanza, 1992, 2004). To analyse parental discourse strategies in this thesis
Lanza’s (2004) framework was adopted. Lanza (2004, p. 262) classifies such responses
into 5 possible discourse strategies ranging along a monolingual - bilingual context
continuum:

95
 Minimal grasp – indicating no comprehension and requesting clarification
 Expressed guess – asking a yes-no question in the other language and requesting
clarification
 Adult repetition of the content in the other language
 Move on strategy – continuing in the interaction
 Adult code-switch.

Although parental discourse strategies were not a direct focus of this thesis, they were
nevertheless considered, since they influenced Ria’s language choice. Due to space
restrictions, only a qualitative analysis of parental strategies evident in the video-
recordings and diary records was carried out. A summary of strategies used by each parent
is produced below, while examples and Ria’s response to such strategies will be discussed
in more detail in section 8.2.1.

Ria was socialized into the pragmatic skill of appropriate language choice with each
interlocutor through parental language choice when addressing her from birth. When Ria
started producing speech the parents’ discourse strategies had to include ways of
negotiating the expected language choice. During the case study the parents maintained
their preferred language choice, and never switched completely following Ria’s mix. Both
parents also accepted any language used by Ria and did not explicitly reject her language
choice with the ‘Minimal grasp’ strategy, although the monolingual father had to seek
clarification of words he did not understand. Both parents used Lanza’s (2004) bilingual
strategies in response to language mixing, but there was a difference in the type of
strategies used by each parent.

The parents developed their strategies in response to Ria’s use of the two languages in
different contexts. The father did not read guides to bilingual parenting, but intuitively
developed several strategies in his reactions to Ria’s use of Slovak, depending on his
ability to understand or infer meanings of Slovak words:

 Requesting clarification of a word, focusing on the meaning


 Requesting the English equivalent
 Explicitly stating who would normally use the word, supplying the TEs and
relating them to a particular person, e.g. “Yes, and daddy says xxx. Mami says
xxx.”; this strategy was also used when the whole family was interacting together
while focusing on a new object or activity, and was particularly evident during the
family’s stay in Slovakia

96
 Inferring the meaning from the context if Ria used a word he was not familiar with,
although such inferences were not always correct
 Continuing in the interaction if Ria used a Slovak word familiar to the father; since
by the end of the first year he was able to understand many Slovak words related to
Ria’s contexts he also understood most early Slovak words uttered by Ria
 Adopting and regularly using the Slovak word as if it was the correct word in both
languages (e.g. bábo “baby”, havo “doggie”), which occurred mainly with Ria’s
preferred and frequently used words.

Thus although the father was a monolingual speaker who at times clearly stated through
‘minimal grasp’ and ‘expressed guess’ strategies that he did not understand Slovak, he also
allowed Ria to see that he understood some Slovak words by using the bilingual ‘move on’
and ‘repetition’ strategies in responses to her mixing. In his attempts to understand Slovak
he occasionally made inference errors, which in turn emphasized his monolingual identity.
Examples of the father’s strategies are reported in section 8.2.1.

The mother was undertaking background reading on bilingual parenting throughout the
study. Although it can be argued that being knowledgeable on the topic had potentially
created an environment more conducive to successful bilingual development in
comparison to other families, previous research found it is common in families that choose
to raise their children bilingually to seek out sources of information and guidance (Barron-
Hauwaert, 2004, p. 7; King & Fogle, 2006, p. 697). It appears bilingual families establish
greater language awareness in general. Moreover, it is necessary to point out that recording
sessions for the purpose of this thesis preceded analysis, and the mother was not familiar
with the theoretical basis of discourse strategies prior to recording, and her strategies were
also developed intuitively in response to Ria’s language choice.

The mother used mainly bilingual strategies and allowed Ria to be aware that she
understands English utterances, but without explicitly pointing out which person normally
used which word. The mother relied on recasting Ria’s utterances into Slovak, which is a
form of ‘repetition’ strategy:

 Acknowledging Ria’s attempt and supplying the Slovak TE: Ria uttered “/ku/” and
pointed at the father cooking, the mother responded: ‘Ano, daddy varí.’ “Yes, daddy is
cooking”

97
 Repeating the English word and recasting with Slovak TE, especially if the English
word was a first token of use by Ria, in order to model the adult form in both
languages
 Recasting the content of Ria’s utterance in the Slovak language with question-like
intonation, thus seeking clarification. Thereby the mother modelled the repaired and
preferred form in Slovak, as well as negotiated the maintenance of the Slovak
monolingual context. It worked also as a word learning strategy since it encouraged
Ria to actively use the words or phrases in the Slovak language after the mother had
modelled the repaired utterance.

Examples of the strategies used by the mother are discussed in section 8.2.1.

As pointed out by Quay (2001, p. 182), Lanza’s framework overlooks one important
strategy. The mother in this thesis repeated the mixed lexical item by incorporating it into
her utterance in the other language. Thus a complete switch did not occur, but the child’s
utterance was accepted and then the mother moved on using the preferred language. This
strategy appears to be intermediate between a temporary ‘code-switch’, ‘repetition’ and
‘move-on’. Lanza (2001a, p. 211) indeed includes this response as ‘code-switch strategy’,
since it is used as intra-sentential switch by the parent. However, this strategy could be
incorporated in the framework in its own right, because although the item was accepted in
the ‘wrong’ language, the parent repeated the child’s utterance in both languages, thus
insisted on returning to the preferred language.

A similar strategy was described also in Saunders (1984, p. 77), where the mother often
modelled both equivalents to the child to provide new vocabulary or to check knowledge
of an equivalent. Such focus on both languages helps the child with bilingual awareness,
since it makes the use of both languages in the family explicit, and reinforces the
importance of knowing TEs to be able to communicate effectively in both languages.

Each parents’ strategies were guided by a different set of circumstances. The bilingual
mother perceived it important to establish a Slovak only context while interacting with
Ria, aiming to provide as much support for the minority language as possible. The
monolingual father established English context as the only option available. However, the
manner in which each parent maintained the desired monolingual language context
differed.

The bilingual mother’s main concern was the lack of monolingual models of the Slovak
language. Therefore in interactions with Ria she was trying to negotiate mostly

98
monolingual Slovak context. However, she was required to use both languages in daily
interactions in front of Ria (see section 4.3), and thus had to indicate which language was
appropriate when. She mostly used the ‘Adult repetition’ of the content of Ria’s mixed
utterance in the Slovak language, or repetition of the content in both languages. Although
this strategy did not strictly negotiate monolingual context in the minority language, it
modelled the mother’s preferred language. Thus overall the mother modelled bilingual
identity, since she was presented daily with situations in which the use of the majority
language, English, was appropriate. She was modelling language separation and preference
for monolingual language use, but through a bilingual identity.

As a monolingual English speaker the father modelled monolingual English identity


through daily interactions without the need to consciously implement a strategy. However,
in his attempt to avoid communication breakdowns with Ria he also indirectly allowed
bilingual context. If Ria mixed a Slovak word into an English utterance and the father was
familiar with the word, he used the ‘Move on’ strategy and continued the conversation in
English, because he understood and there was no need to request a clarification. Likewise,
adopting family Slovak words in his speech allowed somewhat bilingual context. The
father could afford accepting Slovak words, since English was the dominant language of
the community and Ria received many monolingual models of the language. However,
since he explicitly stated that he did not understand all words, and occassionaly made
errors in inference, he negotiated monolingual context in his interactions with Ria.
Bilingual strategies which specified particular speaker for each language also reinforced
his preference for monolingual English context. Thus despite the use of bilingual strategies
that accepted Slovak words he was familiar with, Ria was able to infer from the usual
input, overheard speech and the father’s discourse that he was a monolingual English
speaker who was keen to accept Slovak. Therefore she most likely did not perceive the
father’s identity as bilingual.

A similar difference in parental approaches was noted by Saunders (1984, p. 78). In this
case study from an Australian environment the English speaking mother could afford a
bilingual context, since her language was supported by the community, whereas the father,
a non-native speaker of German and the only speaker who modelled German to their
children, preferred a monolingual German context to support the minority language.

In this thesis both parents used some form of repetition. However, the mother’s repetition
in the form of a question appeared the most successful in maintaining the Slovak language
context. By phrasing the repetition as a question, the mother provided Ria with a repair
99
cue, since she elicited a response (clarification) in the Slovak language. This reinforced the
use of the language, by directly modelling the appropriate form to Ria, inviting her to
practice the form immediately and then to move on in the interaction with minimal
disruption, maintaining the use of the same language.

In fact, the repetition strategy was previously found to be the most favourable in raising a
child’s bilingual awareness, since it provided the child with the missing information in the
other language, which was in many cases the reason for mixing (Juan-Garau & Pérez-
Vidal, 2001, pp. 77-79). Thus it avoided communication breakdown, which could occur in
cases where the parent used the minimal grasp strategy. Moreover, the repetition strategy
invited the child to reproduce the utterance in the target language (Juan-Garau & Pérez-
Vidal, 2001, p. 74).

Having examined the language learning environment, in the following chapters I will turn
to descriptive analysis of Ria’s linguistic developments in the two languages. The case
study data pertinent to pre-linguistic bilingual awareness, development of cumulative
vocabulary, use of TEs, grammatical development, Ria’s developing sense of bilingualism
and aspects such as language choice and mixing will be addressed.

100
5 Becoming Familiar with the Languages (birth-1;3)
In this chapter I will present data from the prelinguistic developmental stages, focusing on
Ria’s developing awareness of two input languages.

5.1 Prenatal stage


Linguistic development starts before a child is born (Pearson, 2008, p. 45). The auditory
system is functional as early as the twenty-fifth week of gestation (de Boysson-Bardies,
2001, p. 23). According to Huotilainen (2004, p. 5), a foetus is therefore exposed to its
linguistic environment already in-utero. Sound is the only element of the outside world
that the foetus is able to perceive, and allows the foetus to become familiar with the
rhythm of the mother’s language, and to memorize this auditory information and use it
after birth (Huotilainen, 2004, p. 4).

If the mother is bilingual, the rhythms of both languages that the mother uses on a regular
basis become familiar to the foetus, thus the process of recognition of her two languages
can start prenatally, and she becomes naturally attuned to the mother’s language straight
after birth (De Houwer, 2009b, p. 29).

The most striking examples, as summarized by De Boysson-Bardies (2001, p. 25), are the
preference of the mother’s voice over another female voice, even when the mother is
speaking a foreign language, as well as recognition of voices of people who frequently
spoke to the mother during pregnancy. What is familiar to the infant are prosodic features
of the mother’s voice, the contour of the voice and alteration of speed. Another type of
auditory information which foetuses are capable of memorizing in-utero is music (de
Boysson-Bardies, 2001, p. 26; De Houwer, 2009a, p. 29).

Linguistic development therefore begins through exposure to the sounds of the language(s)
which form a child’s linguistic environment, and by learning to recognize how these
languages sound when spoken by familiar people. Huotilainen (2004, p. 5) further argues
that the presence of familiar sounds after birth plays an important role for the infant in
adjusting to the new experiences, and to direct the infant’s attention towards the voices
over any other auditory stimulation in the environment. The infant is primed to direct her
attention and to respond to speech, to social interactions, and to direct her listening to the
linguistic models.

Diary data collected for this thesis provide some evidence of intrauterine priming related to
a nursery rhyme the mother used to sing to her foetus frequently. Immediately after birth,

101
and for several months after, the mother was able to settle crying Ria with the familiar
sounds and rhythm of the nursery rhyme over any other rhyme.

5.2 Prestages
The next stage of linguistic development begins at birth. Traditionally in research literature
this is described as the first developmental stage, however, in this thesis it follows the
prenatal period as the second stage. The beginnings of language acquisition are termed the
prestages, since they subsume two developmental periods prior to linguistic productions:

 the earliest stages


 the babbling stage (Oksaar, 1983).

The prestages extend for a period of approximately 10 months, with individual differences
between children (Oksaar, 1983). During the pre-stages an infant is becoming familiar
with her environment - the important people and the languages they speak. She starts
learning how to interact with the familiar people through observation of the ways they
interact with her by means of their languages.

Case studies in BFLA tend not to focus on the preverbal stages of linguistic development,
but are concerned mostly with later developments such as unitary vs. separate language
systems (Genesee, 2007; Volterra & Taeschner, 2007), development of morpho-syntax (De
Houwer, 1990), language mixing (Lanza, 2004) and lexical development (Pearson, 1998).
On the other hand, main topics of experimental studies in the early pre-linguistic stages
focus on early language processing and discrimination (Bosch & Sebastián-Gallés, 2001;
Fernald, 2006).

Indeed in the first 10 months of the infant’s life, which is the period of the pre-stages, it is
difficult to directly observe any signs of bilingual language development. There is no
bilingual data available, since the production is limited to cooing, babbling, paralinguistic
means of communication and proto-words, which were invented by the infant herself, and
as such are not based on any languages.

Likewise, in the first year when phonetic development takes place it is not logical to talk
about the development of two phonological systems as yet, but rather the sounds a child is
producing (De Houwer, 2009a, p. 177). In first linguistic productions a child is not yet
producing individual sounds of the language.

Addressing language discrimination is not possible outside laboratory experimental


settings with measurements of indices of brain activity. Experimental studies using such

102
technologies to determine BFLA infants’ speech perception show that at least up to the age
of 11 months bilingual infants are not yet able to distinguish the sounds of their native
tongues from each other, or distinguish them from sounds of a different unfamiliar
language (Fernald, 2006, p. 22). Perceptual narrowing that allows bilingual infants to
discriminate their two languages occurs at a later stage (Garcia-Sierra et al., 2011, p. 556).

5.2.1 Bilingual sensitivity


During the pre-stages Ria accepted equally being addressed by the languages in her
environment, Slovak and English. When distressed she appeared to settle solely with the
familiar rhythm of the mother’s language and her voice rather than the individual sounds
of the language. However, the mother observed at a later stage (after the age of 1;0) that
Ria tended to become unsettled and was more likely to request the mother’s attention more
intensely if the mother spoke English around her to other English speaking interlocutors or
on the telephone. It was as if Ria had an awareness of the usual language in the dyad and
when the mother was speaking English, she was therefore not addressing her and not
paying attention to her. Thus at this later stage Ria was most likely able to discriminate the
sounds belonging to the language usually spoken to her by the mother.

Apart from this example there is no data that would provide any direct evidence of
language discrimination, thus this ability was observable at a later developmental stage.

5.2.2 Recognizing linguistic forms in the input


During the prestages the emphasis is on the development of comprehension of the first
familiar words and expressions, which is difficult to measure correctly. However, an
insight to the developing receptive bilingual skill was gained, and considered an indication
of the developing bilingualism.

According to Tomasello (2003, p. 20) an ability to recognize sound patterns in speech


sequences emerges around 4-5 months when an infant is able to recognize word-like sound
patterns and associate them with objects or events.

In this thesis recognition of sound patterns in the input first emerged at the age 0;4. Ria
showed signs that she remembered the familiar aspects of her daily routines and associated
certain places with particular routines, such as play time in the family lounge, feeding in
the bedroom, play-time on the change table. Ria responded to familiar sound patterns of
utterances the parents tended to repeat during such routines, as in the following examples:

103
- when the mother was strapping Ria in the car safety seat, she used to ask ‘Kde je
macík?’ “Where’s the teddy?” and Ria would reach out for the toy;

- when the mother asked Ria if she would like a feed ‘Dáme si hami-hami, áno, hami-
hami?’ “Should we have a feed, yes, a feed?”, Ria turned towards the mother and open
her mouth, wanting to attach;

- the mother often asked Ria about the father, ‘Čo robí daddy, kde je daddy, čo robí?’
“What’s daddy doing, where’s daddy, what is he doing?”, and Ria looked towards the
father.

In the above examples Ria was most likely responding to the familiar words ‘hami-hami’
“feed” and ‘daddy’. Likewise, favourite toys appeared to be known by the label. Thus
when Ria was responding to familiar, regularly occurring sound patterns, she started
linking them to a person, object or action.

At 0;5 months Ria associated linguistic sound sequences with familiar contexts. Initially
she reacted to utterances that were linked to a familiar situation. In anticipation of the
familiar activity she expressed happiness or excitement through paralinguistic means,
usually by kicking, jiggling, smiling, or laughing. This occurred in relation to daily
routines with relatively same order of events such as bath time or feeding. She also
responded to utterances regularly occurring in particular contexts, e.g. when the father
called out ‘Come to daddy!’, she looked at him with a happy expression; Ria responded by
leaning towards the mother when called with the command ‘Poď sem, poď! Ria, poď!’
“Come here, come’n! Ria come’n!”. At the end of the sixth month Ria also responded to
her own name by turning head towards the person who uttered it. Thus comprehension was
not limited to words alone, but it included longer utterances to which Ria was able to
respond appropriately.

From recognition of familiar sound patterns in certain situations Ria was gradually
learning to recognize individual words. With word comprehension the ability to use a
deictic pointing gesture emerged. A microanalysis of this development suggests
interpersonal interactions are crucial in linguistic development. The pointing gesture
emerged during joint attention in a specific context and in reference to one specific word
‘nose’– the parents were pointing at a toy, showing parts of the face and Ria imitated the
pointing gesture. In this context she also pointed in response to familiar questions asking
her to identify the referent. More importantly, Ria responded equally to the English
question ‘Where is Kaloo’s nose?’ and the Slovak TE involving a different toy, ‘Kde má

104
Piglet nos?’. Thus she recognized the equivalent linguistic forms in the two languages: the
meaning of the questions, as well as the TEs ‘nos’ and ‘nose’ as used by the parents. She
referred to the correct referent on the respective objects with a pointing gesture. At first
she was able to identify the referent on few favourite toys only, but soon extended the
category for the words and pointed at the referents on other toys, pictures in books and
people.

This microanalysis demonstrates several aspects of developing bilingualism. Ria


understood the TEs ‘nos’ and ‘nose’, as well as the general category of objects the words
referred to. She also understood the more complex request to locate the referent. Thus by
pointing she demonstrated that she was able to segment speech, extract familiar words,
map them onto the referents.

Since Ria’s input consisted of two languages it was necessary to consider whether she
responded equally to both. Diary records state that indeed most familiar sound patterns in
Ria’s comprehension appeared as TEs, for example:

- the Slovak request ‘Kde je daddy?’,

- the English request ‘Where is daddy?’.

In this example, the Slovak equivalent expression extended to various objects, while in
English it initially related only to the caregivers, and later extended to other objects and
persons as well. Thus comprehension of TEs did not necessarily emerge simultaneously.

The pace at which Ria was acquiring translation equivalents in comprehension was
assessed through observations of behavioural responses when Ria’s attention was directed
to the same referent by each parent (using their respective languages) on two separate
occasions. Comprehension of referents was established through elicitation. Responses such
as turning eye gaze towards the object, pointing at the object, or looking around for the
labelled object and smiling when located were considered as correct comprehension of the
word.

In the last month of the prestages (0;9) Ria’s comprehension repertoire included TE labels
for several familiar objects. She was able to identify the correct referents with a pointing
gesture in response to two TE questions:

‘Kde je xxx?’

‘Where is xxx?’.

105
Although the number of referents was rather limited at this stage, it increased at a steady
pace in both languages in the following months. Initially more Slovak words or
expressions were in comprehension, but their English TEs did not lag behind for more than
a few weeks. This slight difference can be explained by the imbalance in the input, as the
mother who addressed Ria in Slovak was the main carer. The exact extent of bilingualism
could not be determined, however, based on the analysis of the input during this period,
the general assumption was made that since Ria was exposed to the two languages
consistently from birth, she was considered a bilingual since her receptive abilities in the
two languages, which led to productive abilities, were developing in parallel.

The above developments marked the onset of intention-reading communication. According


to Tomasello (2003, p. 21) intention-reading skills emerge around 9-12 months as a new
set of social-cognitive skills that allow the infant to:

 engage in joint attention


 understand communicative intentions
 learn cultural aspects by imitation and role reversal.

The basis of intention-reading is in the infant’s ability to share attention to objects and
events with others, to follow attention of others as well as to direct attention of others, to
understand others’ communicative intentions, and to learn various cultural aspects
imitatively (termed imitative learning or role reversal imitation) (Tomasello, 2003, p. 3).
While prior to the social-cognitive development of intention-reading an infant was capable
of dyadic communication with a parent, the communication now became triadic, since the
child is capable of coordinating and sharing attention to a third entity (an object or an
event) with the adult. These skills enable symbolic communication and are crucial for
language development such as acquisition of linguistic symbols, expressions and
constructions. By 0;10 Ria was making the transition into the next developmental stage.

5.3 Transition to language


The transition period extended approximately from 0;10 until 1;4. This period has been
suggested as a separate developmental stage by several authors as based on their case
studies (Halliday, 1975, p. 139; Oksaar, 1983). It is viewed as a phase between the
babbling stage and the one-word stage when a child’s semantic abilities start developing
(Oksaar, 1983).

In this thesis the transition period was significant for two major developments: rapid
development of linguistic comprehension and intention-reading communication. It was
106
also marked by the use of proto-words and first adult words. In this stage the early months
of language familiarization were peaking and transformed into comprehension and
production of early linguistic forms.

Ria’s linguistic repertoire included three types of productions:

 Pure babbling
 Proto-words
 First adult words.

5.3.1 Pure babbling


During the transition period canonical babbling became prominent. Ria continued to
babble freely to play with sounds, such as [dada dadadada] [baba ba] [paːpaː paːpaː] [ma
mama] [hehehehe] [hahaha]. She was also stringing sounds into combinations to which she
was assigning particular rhythm and intonation much like in a stream of speech. Thus
babbling served as a practice of the speech organs, various sounds and rules of prosody.

In the early transition period babbling sounds and their combinations did not appear to be
language specific as some authors observed (Cruz-Ferreira, 2006, p. 63), but reflected
sounds and the prosody of the two languages towards the end of the transition period
(1;3.8). Such combinations differed from canonical babbling, since they were not
reduplications of the same syllables, but random combinations of sounds with intonation
and phonetic and syllabic rhythm characteristic of the two languages. The following
excerpt shows Ria’s babbling combinations with English like prosody used in a bilingual
context while interacting with the English speaking father:

%sit RIA is learning how to stand on a stepping stool in the dining, DAD is
helping

*RIA [ak]

%act RIA runs up to the stool

*MAM Can’t believe herself?

*DAD She can’t stop doing it. I’d like to put it away.

%act RIA is trying to step up

*RIA [au]

*DAD Careful! Very careful! Very careful. You didn’t have your second foot on
far enough.
107
*RIA [ja aɢa aː]

*DAD Yeah, you got it.

*RIA [au wa]

*DAD Yeah. You’re probably getting tired.

*RIA [aː]

*DAD You probably getting tired from doing it now. Ah, you’re gonna have a
sleep on there, couldn’t you? That’s a good idea.

%act RIA tries holding onto DAD’s hand

*DAD You cheat, you use me. Oup, that didn’t work, hey?

%act RIA slips off the stool

%gpx RIA shakes her head

*DAD She shakes her head. She slipped off and she’s like…no.

%gpx DAD mimics RIA and shakes his head

*DAD That didn’t work at all. Hihihi!

*RIA [agwɜːk aɲɪɟːa agaʁ wɜː]

*DAD Anger?

%com DAD assigned one babbling sequence to a similarly sounding English word
‘anger’

*DAD Good, that’s one good foot up there, now the other one. You had it. Ah, you
gonna do it different way? You try to do it without leaning down? That’s
the funniest! She’s got her hands on there and she puts her foot up on her
hand and then she can’t get her hand out.

*RIA [aut]

(1;3.5)

The second excerpt shows Ria interacting in a Slovak only environment and using
combinations of sounds that resemble Slovak syllables:

%sit RIA is learning to sort blocks through a shape sorter, MAM is passing her
shapes

*MAM A túto? No. No, aj tam sa zmestí, dobre! A túto? Túto kam dáme?
108
%eng And this one? Yeah. Yeah, it fits there too, well done! And this one? Where
will we put this one?

*RIA [tato]

*MAM Kam ju dáme?

%eng Where are we going to put it?

*RIA [joːgec aɟɪ bagɪ ɟɪkojɛ]

%gpx RIA is looking directly in MAM’s eyes and MAM is looking at her too

*MAM Ano?

%eng Yes?

*RIA [sɛm tɛːja ɛp ojc]

*MAM Kam ju dáme?

%eng Where are we going to put it?

*RIA [auɪ xɛɟoc cɛːbac dowa]

*MAM Ano, naozaj?

%eng Is that right?

*RIA [nɪmdau xuː ɟabau]

*MAM Naozaj? No, kam ich dáme?

%eng Really? Well, where will we put them?

*RIA [ɛpsɛc]

(1;3.21)

After the first year (1;0.7) babbling filled a gap where Ria had no alternative for vocal
expression, but needed the interactional exchange with an adult. Thus the purpose of
babbling address to others was social interaction.

5.3.2 Protolanguage
Halliday (1975) described prelinguistic development according to what a child was
achieving with his communicative intentions before language emerged. Other creative
means to communicate intentions were used – the protolanguage, kinetic and paralinguistic
means. According to Halliday (1975), proto-language emerges when a child’s expressions
become symbolic rather than involuntary (babbling) prior to the transition to the adult
109
language. An infant’s social world has only a limited number of contexts in which he can
use symbolic expressions to communicate. Protolanguage allows the child to communicate
his intentions within his social context (Painter, 1985, pp. 52-55).

Halliday termed the expressions protowords (also called babbling words, vocalizations or
idiomorphs). Protowords consist of random combinations of sound sequences that are both
phonetically and semantically regular, can be variations of a sound, or ‘a continuum of
intervening articulations’ (Painter, 1985, p. 56). Protowords are invented by the child and
are not based on any language. They are used repeatedly to refer to a particular object,
person or request, but always in contextually similar situations. They are accompanied by
paralinguistic elements such as gestures or kinesics.

In this thesis, symbolic productions that carried intentional meanings emerged during the
transition stage (0;9), but were more limited in comparison to Halliday’s (1975) data. Ria
used several short-lived vocalizations that could be classified as protowords, however, she
preferred one universal proto-word. It appears the vocalization emerged first as a
ritualization, since the rather agitated nature of the sound had the desired effect on the
parents – they always attended to Ria and acted on her request. Thus it became an effective
means of communication and Ria started using it frequently in various contexts. The proto-
word was used equally when interacting with both parents, thus it was not tied to a specific
language. It was used frequently, served several functions, and had the greatest scope of
meaning, which continued to extend over time.

The protoword emerged towards the end of the babbling stage as a sound produced in the
throat, a [hmː] sound, which resembled a consonant produced with a closed mouth with a
long rising tone giving it a nervous, urgent tone held for a longer period. It was produced
with varying intensity. The vocalization did not have one stable form but several variations
were recorded. As Ria turned 1;0, the initial sound changed into a rather vocal vowel [əː].
Later variations were [əːː] [eː] or [aːː], all long, stretched vowels. All variations showed a
level of urgency and resembled a straining sound one might produce when trying to reach
out for something. When the request was more urgent, it approximated a vowel, and when
Ria was more content, it approximated a [hm] sound. Ria also produced this vocalization
more often and more urgently when she was tired and easily upset. Around 1;0.13 the
vocalization assumed a more stable form and from this point it was recorded most often as
a reduplicated [əː əː] with a rising tone.

110
The vocalization was often used in combination with physical means of achieving an
outcome, such as pulling herself up or towards the direction she wanted to go, stretching
her arms up to be picked up, with a pointing gesture when requesting objects.

If Ria did not achieve the desired response from adults, she continued repeating the same
vocalization, with or without gestures, until the parents reacted according to the request,
knowing that it was an effective way that served to achieve the desired outcome. When
vocalizing repeatedly, its tone became gradually more urgent and frustrated, depending on
how quickly the parents reacted, and if they understood Ria’s request correctly, e.g. they
passed her the desired object.

However, variations in tone were not differentiated across specific functions, as shown by
the following example:

- two variations, [hmː] and [aː] were used within one situation while Ria was requesting
the same object - the family was at the dining table and Ria pointed at an object. The
mother did not immediately understand which object was requested and passed the
wrong one, after which Ria pointed again vocalizing, then repeating the request but
with a different variation of the vocalization, until the mother responded appropriately
(1;1.9).

This example indicates the variations were interchangeable, since there seemed to be no
change in the meaning, only an added emphasis.

There were two initial functions - request for objects, request for involvement or to be
picked up. It then extended to a request for actions and activities Ria would like to do, as
well as a request for a service from the parents. Ultimately the aim was to be engaged in an
interaction with either parent and /or their activity, to be directly involved in the activity,
observe, participate and imitate. The functions gradually expanded towards the end of the
first year and where most varied just after the first birthday. By the end of 1;0.26 the
vocalization became Ria’s most typical expression. The variations, functions and examples
within specific contexts are shown in APPENDIX B.

Thus while Halliday’s (1975) protolanguage analysis consisted of a rich repertoire of


idiosyncratic vocalizations, in this thesis proto-words formed only a small part of Ria’s
expressions of meanings. Ria used primarily kinetic and paralinguistic means which served
various functions, as discussed in the following section.

111
5.3.3 Using gestures to communicate intentions
The use of gestures in the early stages of linguistic development appears as an
idiosyncratic development, since each child develops her own ways to communicate
intentions (Goldin-Meadow, 1998, p. 30).

During the transition period Ria’s early communicative acts were based on gestures, body
language and other paralinguistic means of communication. Such gestures were more
symbolic, since they served not only declarative purpose as previously, but also imperative
purpose. Ria directed an adult’s attention to an object or an event in order to communicate
her intentions, to direct the parents to act on objects and requests. The most prominent
form of communication were various ritualized gestures (the deictic pointing, holding up
and passing objects) and kinetic means of requesting (body positioning, turning eye gaze).

Paralinguistic forms of communication were not tied to specific languages. Ria was able to
engage both parents equally, regardless of the language used to address her, as in the
following example:

The family was playing in Ria’s room. Ria passed a musical toy to the mother, who
turned it on and Ria started dancing. Simultaneously, the father picked up a doll and
pretended the doll was dancing. When the song finished the parents put down the toys.
Ria passed the toy to the mother again, then turned to the father and passed the doll to
him, requesting them to perform the same actions again. When they responded
appropriately Ria’s facial expression showed contentment. (0;11.11)

In interactions with several people simultaneously, Ria combined two or more gestures,
such as holding up an object and directing her eye gaze to a specific person, thus
requesting that person to perform the desired action with the object.

5.3.4 Emerging Word Comprehension


According to Pruden et al. (2006, p. 267) there are two competing theories on word
learning in the field of first language acquisition. One theory suggests that word learning
takes place through associative mapping of words onto referents from the child’s point of
view, in other words developing associations between a linguistic form and a referent
following repeated object-label pairings in her input. The opposing theory suggests that
word learning takes place through the exchange of social information and social cues that
allow the child to interpret social intent of her interlocutors, and thus map words onto
referents from the speaker’s point of view (Pruden, et al., 2006, pp. 266-268). However,
Pruden et al. found that both mechanisms appear to be employed in word learning. While
112
10 month old infants were learning words through associative learning not relying on
social cues, at 12 months they became sensitive to social cues but were not yet able to use
them for word learning. This ability developed around 18 months, and up to 2 years
children were relying on social cues to map labels to referents (Pruden, et al., 2006, p.
277). Thus it was suggested that word learning moves along a continuum from associative
learning to socially influenced word learning, and that the latter is made possible when
infants become able to read communicative intentions of others. Word learning rates could
also be explained by this developmental shift, since the initial associative learning requires
repeated pairings of a word with its referent, while socially guided word learning occurs at
a faster pace (Pruden, et al., 2006, p. 278).

Similar observations were made in this thesis both in regards to the mechanism, as well as
the rates of word learning over the age periods from 10 months to 2 years. The early word
learning was slow and related to salient items in the input, as will be analysed in the
following paragraphs. Later word learning occurred at a considerably faster rate and items
did not require repeated pairings in input to be learnt, as analysed in Chapter 6.

In the babbling stage early association of familiar sound patterns with familiar contexts
was said to emerge at the age of 5 months (See section 5.2.2). In the following months Ria
was able to associate increasingly more contexts with familiar sound patterns. However,
such associations now related to particular words or expressions rather than similar sound
patterns. Thus the comprehension was of adult words. As demonstrated through the diary
records, Ria was gradually able to respond to increasingly more questions, requests or even
non-direct requests addressed to her by the parents:

- during the bath time routine the mother said ‘A teraz si umyjeme vlásky!’ “And now
we’ll wash hair” and Ria spontaneously leaned backwards to wet her hair

- when the mother asked Ria to show endearment, such as ‘Urob daddymu mój daddy!’
“Do my daddy!” (inviting Ria to stroke her father) or ‘Daj mame pusu.’ “Give mummy
a kiss”, Ria responded with the appropriate actions (0;10.27).

During the transition period there was a considerable increase in the comprehension of
individual words, which emerged with several familiar words. When Ria made the
association between the phonetic form and the referent (including animate and inanimate
objects) she was able to recognize the word in a stream of speech, and her attention
focused on the object, as was shown through signs such as head-turn and eye gaze towards
the object. She was then able to simultaneously indicate the referent with paralinguistic

113
communication means, either gestural (eye gaze, pointing) or vocal (babbling or
protoword). This was evident firstly in elicitation by the parents, as they asked her to
identify a familiar object. However, Ria also spontaneously responded to familiar words in
a stream of speech by directing the parents’ attention towards the referent. Examples from
both languages were captured. The first example shows comprehension in Slovak context
with the mother:

%sit RIA and MAM are playing with a musical toy in RIA’s room, RIA is
pressing buttons, MAM provides commentary

*RIA [aba wabuː]

%glo xxx

*MAM Peek-a-boo!

%com Repeating sounds produced by the toy

%act RIA turns a butterfly shaped button

*RIA [hɛ aːda]

*MAM Dá sa otočiť motýlik, však? To je taký motýlik.

%eng You can turn the butterfly around, can’t you? It’s a butterfly.

%act RIA turns left, looks up and points at a butterfly decoration on the wall

%com comprehending the word butterfly RIA responds by directing the mother’s
attention to a familiar referent

*MAM Tam sú, ano, aj tam sú motýliky!

%eng There they are, yes, there are butterflies too!

*RIA [aɟa]

%act RIA looks up at the light shade with butterfly motive and points at it

*MAM Aj tam sú, ano, Rianka má veľa motýlikov.

%eng And there they are, yes, Ria has many butterflies.

%act RIA turns to her right and points at another butterfly wall decoration

*MAM Aj tam sú motýliky, ano. Aj tu je motýlik, aha. Aj tu je. Aj tu je. Aj tu je.


Vidíš, to je motýľ.

114
%eng There are butterflies too, yes. And here’s a butterfly, look. Here’s one too.
Here’s one too. Here’s one too. See, it’s a butterfly.

%act MAM points at the butterfly on the toy

(0;11.15)

The second example was taken from an English context while interacting with the father:

%sit DAD and RIA are playing in the room, RIA is pulling out various objects
from toy drawers and DAD provides commentary

%act RIA pulls out the whole drawer

*DAD Hoho!

*RIA [babaː]

%act RIA is looking inside another drawer; there is a balloon next to it, but RIA
does not notice it, and DAD comments

*DAD There’s a balloon there, a purple balloon.

%act RIA turns and looks at the balloon

%com RIA comprehends the word balloon

*DAD Yeah, balloon.

%act RIA starts playing with the balloon

*DAD Yay!

%act RIA is laughing

*DAD Yay!

*RIA [m m]

(0;10.31)

In both examples Ria responded to familiar words she segmented from the stream of
speech, and directed the parents’ attention to the usual referent she associated with the
word. Ria’s spontaneous pointing indicated a response to the familiar words, and initiated
interaction about the referents with the parents, thus communicated her intention of joint
attention.Therefore the nature of the interaction changed from being purely interpersonal
and focused on the two people to involvement of a third entity – object or event.
115
Until the end of the transition period Ria’s comprehension related to the immediate
situational contexts in which an interaction took place. However, she was developing the
ability to relate familiar words to experiences removed from the immediate context. This
was demonstrated in several examples recorded in the diary:

- While reading a picture book with the mother and focusing on a picture of a
strawberry Ria pointed in the direction of the refrigerator and laughed, thus pointing out
the usual place for strawberries (1;3.27).

- Ria responded to pictures of a cart and a stacking toy by pointing in the direction of
the garden, where she was normally accustomed to play with these objects (1;3.28).

- While reading a book about birds with the mother, Ria first looked in the direction of
the garden, then pointed in the direction of her room, and uttered a word that
approximated [taːtʃɪ] ‘vtáčik’ “birdie”. In an attempt to understand Ria’s intended
meaning the mother asked ‘Čo tam je?’ “What’s there?”, to which Ria responded by
walking to the window in her room and pointing at the roof of the neighbouring house,
where birds often perched and Ria was accustomed to watch them with the mother.

In these examples Ria demonstrated that word comprehension extended to all types of a
particular referent and not just a specific item, since both the objects and pictures of those
objects belonged under one conceptual meaning. Moreover, she related them to
experiences outside the context, and using gestural communication means directed the
mother’s attention to this connection.

5.3.4.1 Comprehension in bilingual environment


In studies of monolingual children there is a well-established observation that
comprehension precedes production (De Houwer, et al., 2006, p. 332). In this thesis
evidence for comprehension preceding production in a bilingual child was found. When
Ria began to speak, she already understood words in both languages. In fact, word
comprehension extended to both languages as soon as it emerged. Bilingual development
of Ria’s receptive vocabulary therefore preceded the development of expressive
vocabulary.

To date in BFLA research there is little empirical evidence on whether and how bilingual
children comprehend translation equivalents (TEs). One study found that children as
young as 13 months understood TEs and accepted cross-language synonyms (De Houwer,
et al., 2006, p. 344). Ria also understood meanings of TEs across the two languages, thus

116
the Mutual Exclusivity Bias was not supported in comprehension, which reflects the
findings in De Houwer et al. study.

There were also instances in which she understood only one member of a TE pair, yet was
unsure of the meaning of the other member. However, she generally picked up TEs with a
short time lag. Comprehension of TEs first appeared around 0;11, as in the following
example from the diary:

- The family was talking and playing in the lounge when the father addressed the
mother, with the word ‘pussycat’ contained in an otherwise English stretch of
utterances; when Ria recognized the word (responding to the morpheme ‘cat’ rather
than the compound), she immediately pointed at a wooden statue of a cat placed on a
bookshelf. Several minutes later the mother set up a small experiment: she uttered a
Slovak sentence containing the TE ‘mačka’, but pretended to address the father rather
than Ria directly; Ria recognized also the Slovak word and immediately pointed at the
statue again. This form of ‘probing’ for comprehension of TEs became a game for the
family, as the mother asked Ria to identify various objects in the room using the
question ‘Kde je xxx’ “Where is xxx?”. Ria responded by pointing appropriately at the
correct referents each time, even if she did not spot it immediately, she looked around
to locate it. The father then asked to identify the same referents in English. Ria was able
to identify most, with slight delay in reaction to the English TEs. She did not recognize
all English TEs, such as the referent for ‘shoes’, however, she was able to locate it
when the mother provided her with the Slovak translation ‘topánky’. (0;11.11)

It was difficult to quantify how many words Ria understood at any given time. She
understood more than individual words, but whole structures that carried the meanings of
various concepts, for example descriptions and commands.

However, when the parents attempted to test comprehension they did not ask about
embedded meanings, they asked questions which probed for single referents, and thus did
not reflect comprehension fully, for example when reading a picture book with pictures of
children, Ria correctly identified referents for concepts such as ‘ryšavé dievčatko’ “little
girl with red hair” or ‘blonďavý chlapček s modrými očkami’ “blond boy with blue eyes”.
The parents also probed for clear categories of objects with exact referents that Ria had
many different examples of, such as ‘oči’ “eyes”, ‘mrkva’ “carrot”, ‘medveď’ “bear”.

Moreover, in Slovak, a language with complex morphology and declinations, referents


have several forms of a single word linked to them. This is due to the fact that various

117
forms of a word carry different grammatical meaning, which is expressed with different
suffixes, whereas in English the forms are limited to singular and plural forms. Thus in
Slovak a child needs to map multiple linguistic forms onto one referent while decoding the
grammatical information relayed in that form. For example where in English the referent
“eye” has two forms ‘eye’ and ‘eyes’ in singular and plural respectively, it has the
following forms in Slovak:

Case Singular Plural

Nominative ‘oko’ ‘oči’

Genitive ‘oka’ ‘očí’

Dative ‘oku’ ‘očiam’

Accusative ‘oko’ ‘oči’

Locative ‘oku’ ‘očiach’

Instrumental ‘okom’ ‘očami’7.

There is also the added variation of various forms of diminutives used especially in CDS.
Diminutives are frequent in Slovak CDS, which is also reflected in child speech
(Stejskalová, 1998, p. 11). They are a reflection of positive emotional charge that children
evoke in adults, which is transferred to interaction between an adult and a child (Hlavatá,
1998, p. 18). Thus the emotional charge provides the linguistic model, and will necessarily
be echoed in the child’s comprehension and speech. In fact, in a study with a Brazilian
Portuguese-English BFLA child the author treated diminutive forms and counterpart
standard word forms in the child’s input as different types of words (Nicoladis, 2001, p.
139).

This approach suggests diminutives generally expand word comprehension, since the form
carries an additional meaning, for example in this thesis the Slovak forms ‘oko’, ‘očko’,
‘očičko’ all referred to a single referent, “eye”, and forms ‘ucho’, ‘uško’, ‘ušinko’ all
refered to a single referent, “ear”. However, the diminutive forms ‘očko’, ‘očičko’ and
‘uško’, ‘ušinko’ (meaning “little eye” and “little ear”) carried the additional information
which expressed the mother’s attitude to the referent showing affection.

7
There are 6 nominal cases in each singular and plural number in the Slovak language, and as such there are
actually 12 different grammatical forms a noun can have. However, some case endings are polysemic,
resulting in single forms with several grammatical meanings, which need to be inferred from the context.
118
Frequent use of Slovak diminutives was recorded in this thesis through both the diary and
video-recordings. They were not limited to the Slovak language, but also appeared in
English input. While there is no standard diminutive form in the English language,
diminutive forms are widely used by English speaking parents when addressing children.
In this thesis the father ‘softened’ already emotionally charged words, especially
appellatives and affectionate terms, with the use of suffixes –ie and –ies. As well as the
standard forms birdie ‘bird’, doggie ‘dog’, horsie ‘horse’, tootsies ‘toes’, sweetie ‘sweet’
other creative family forms appeared, such as lambie ‘lamb’, angelie ‘angel’, darlingie
‘darling’, munchkinie ‘munchkin’, housie ‘house’.

The difference between the languages in the use of diminutives was that in English,
diminutives were not as diverse, they were not used as frequently, fewer word types
occurred in diminutive form, and they were formed with two interchangeable suffixes. In
Slovak diverse word types occurred in diminutive form, the morphological process was
less regular with wide range of suffixes and changes in some stem forms.

In this thesis comprehension vocabulary in the two languages was not recorded
systematically, thus there is not a complete list of words in each language, but rather
sporadic records of word learning. What is available, however, are insights in the ways in
which Ria acquired early TEs in receptive vocabulary. Several mechanisms were
observed:

 Individually in a monolingual English/Slovak context


 By association in a bilingual context
 Simultaneously in a bilingual context

In a monolingual context Ria learnt an equivalent in one of the languages. This was the
case mainly with vocabulary from the stronger language, Slovak. Thus Ria needed to
acquire the corresponding English equivalent at a different time and in separate context
when interacting with the father, if similar context occurred, as in the following diary
example:

Ria was interacting with the mother and her Slovak aunt, the mother pointed several
times at the aunt’s dimple in the cheek and provided a label; she then asked Ria to
identify the referent herself ‘Kde má Robi jamku?’ “Where’s Robi’s dimple?”; Ria
responded by pointing appropriately (1;0.11).

In this situation the father was not engaged in the conversation, thus there was no
opportunity to acquire the English TE immediately.
119
Other typical monolingual contexts when words were acquired included book reading,
when the mother pointed at pictures and provided labels, then reversed the roles and asked
Ria to identify the referents on the pictures.

Instances of word acquisition in a monolingual English context were more limited due to
the exposure patterns, nevertheless they were recorded through reports by the father.
Reporting of new words by either parent created contexts in which the other parent had the
opportunity to introduce the TE with some delay, as in the following diary excerpt:

Ria and the father were playing with a ball, and Ria learnt to associate the verb in the
utterance ‘Kick the ball. Kick it!’ with the action; later that evening the mother invited
Ria in Slovak to perform the action ‘Kopni loptu, no, kopni!’ “Kick the ball, yeah, kick
it!”, but Ria did not respond until the father rephrased the utterance into English; thus
the Slovak TE was introduced by association with the already familiar English TE;
having acquired both TEs Ria was able to perform the action in response to a request in
either language (1;2.14).

In a bilingual context, the learning of TEs occurred through two possible learning
mechanisms. The first mechanism was by association with an existing equivalent, thus
through the same associative learning as above, but since the second TE was introduced
immediately after the first, without the time lag, as described in the following diary record:

The family was in the garden when Ria noticed the moon and pointed at it. The mother
provided the Slovak label ‘mesiačik’ “moon”, repeating it several times: ‘To je mesiac.
Mesiačik.’ “That’s a moon. The moon.”. Ria pointed at the moon several times, with
the mother providing commentary each time. Then the roles reversed and the mother
asked ‘Kde je mesiac?’ “Where is the moon?”, to which Ria responded by pointing.
Several moments later the father asked Ria in English ‘Where’s the moon, Ria?’ and
Ria responded by looking around, trying to locate the referent - she was not certain of
the referent, but understood the prompt to identify an object. The mother rephrased with
the familiar Slovak TE ‘mesiačik’ and Ria looked up and pointed at the referent
immediately. Subsequently, the father repeated his English commentary several times:
‘Moon, that’s the moon.’, then probed again: ‘Where’s the moon?’, Ria responded by
pointing at the referent immediately (0;11.14).

Around Ria’s first birthday, during the family’s overseas trip, the father was Ria’s only
source of English and she was exposed to more Slovak input from varied sources.
Naturally, her comprehension repertoire in Slovak increased at a greater rate in comparison

120
to English. The diary records state there were many occasions when the father asked Ria a
question in English, to which Ria did not respond. The father overcame this by requesting
the mother to rephrase his utterances into Slovak, and then he followed in with repetitions
of English TEs. Through this strategy Ria responded to the Slovak utterance immediately,
while she was introduced to the English TE. Thus she had an opportunity to learn both
equivalents (words or expressions) by association with only a short delay. Moreover, this
strategy reinforced bilingualism, since each parent used different language as a tool to
express the same meanings, falling back onto the other language when necessary.

Thus although comprehension extended to both languages, during the early comprehension
stage (0;11) Slovak appeared to be Ria’s stronger language, and her receptive vocabulary
was greater in Slovak than in English. The mother often helped Ria by supplying the
Slovak TE as an explanation, after which Ria was able to identify the referent. Thus she
was able to observe both equivalents simultaneously and acquire them directly in relation
to the referent. As a result more English equivalents were acquired by association with
Slovak TEs than vice versa, which further amplified during the family’s overseas trip
(1;0).

The second word-learning mechanism in bilingual contexts was simultaneous acquisition


of TEs within the same joint attentional frame with both parents. In triadic interactions
each parent labelled an object or provided a short commentary in their respective
languages. Thus on such occasions Ria learnt both TEs simultaneously. While to an
outsider this may seem rather chaotic, Ria was used to the parents addressing her in two
languages simultaneously from birth, making it the norm. Learning was taking place in a
systematic fashion, since within the joint attentional frame all three members were
focusing on the same object. Ria was able to use the parents’ eye gaze as a cue while each
parent was labelling the object in the usual language of address.

Sharing picture books with the parents offered an opportunity to probe for Ria’s
comprehension of TEs, especially since during this stage the parents often read by asking
Ria to identify objects on the pictures. Ria was able to identify referents appropriately in
both languages, suggesting that her comprehension of the contexts in familiar books was at
a similar level. Indeed there were many opportunities when Ria asked one of the parents to
read a book and then requested the other parent to read it again8. She was aware the

8
The mother read all picture books to Ria in Slovak, rephrasing written texts into Slovak. This was the usual
reading routine until Ria became aware of print, and requested all English books to be read in English.
121
parents read books differently, each using their respective language, thus showing early
metalinguistic awareness of two input languages.

There are important implications for BFLA infants that stem from the above observations.
A bilingual child may or may not use translation equivalents in production for a particular
concept even if she has both TEs in her receptive vocabulary. Indeed, differences in Ria’s
comprehension and production in the two languages were observed towards the end of the
transition period (1;4). While she understood both languages more or less equally, at this
stage she seemed to draw her expressive vocabulary mainly from Slovak. This disparity
can be explained by Ria’s language exposure patterns, since overall she received more
interpersonal input in Slovak than in English. Thus although production appeared to be
dominant in Slovak at this stage, Ria’s comprehension was relatively equal in both of her
languages. Indeed, De Houwer (1998, p. 257) argues it is not relevant to talk about
language dominance or even proficiency when we are concerned with a linguistic system
that is still developing. It is more appropriate to use the term stronger language, while the
strength of each language can change over time with changing circumstances and language
exposure.

5.3.5 First words


The onset of speech is a period rather than a specific point in time. In this thesis it started
with single word utterances mostly reproduced after adults. Ria was firstly observing
language in her environment and then reproduced it. This form of reproduction, called
imitative learning (Tomasello, 2003, p. 27), was prominent during this stage not only in
Ria’s speech but also in reproduction of the family’s activities and actions the parents
engaged in, such as pretending to feed a doll, weeding and watering plants in the garden,
reading books, as well as various actions performed with objects. Ria initiated such
reproduction of adults’ actions also spontaneously, without previous interaction within a
joint attentional frame which was focused on the object and its use. The same form of
spontaneous imitative learning was observed with first linguistic reproductions. During
joint interactions Ria participated not only by coordinating her attention accordingly, but
also through vocal response. She attempted to reproduce adult words by producing sound
approximations which did not have the conventional adult word forms yet. According to
Vihman & McCune’s (1994, p. 522) criteria, correct use of the early words was validated
through many situational contexts in which Ria used them. Thus by reproducing linguistic
and other behaviours Ria strived to connect with the people by participating in events that

122
took place in her environment. She did not absorb only those experiences that were
mediated by an adult’s attention, but learnt from independent observation as well.

5.3.5.1 The role of onomatopoeia


The first reproductions were not yet dictionary words, but included onomatopoeic words
which had a symbolic relationship with their referents, and were consistently used in
relevant contexts. Ria learnt first onomatopoeiae through ritualization, thus they were
related to referents iconically. However, onomatopoeiae were reproductions of sounds
produced by different animals which belong to baby-talk and are commonly used in both
Slovak and English languages in CDS, thus are culturally specific. Vihman and McCune
(1994, p. 521) also treated similar productions as first word candidates. Onomatopoeic
sounds initially form the basis of expressive repertoire of infants who are starting to use
adult language (Stejskalová, 1998, p. 10). According to Stejskalová onomatopoeiae
further form the basis of first combined utterances, as well as the basis for creative
morphology in children’s productions of nominals, verbs and adjectives derived from
them.

In the data in this thesis, the family often used sounds produced by various animals during
playtime. Ria knew which ones to expect for which referent and anticipated them with a
smile. Some onomatopoeic sounds appeared as early as in the babbling stage. Not all were
linguistic sounds, however, all were used consistently to refer to one entity and Ria was
able to produce them when elicited by the parents. During the transition stage they became
more symbolic, since they signified the referents as well as imitated the sounds associated
with the referents. Thus they held an important place as the first words. Ria’s
onomatopoeic productions were variations of baby-words used in the adult languages and
contained sounds that closely approximated them. She started by reproducing the sound
impression and gradually approximated it to an adult onomatopoeia, as shown in
APPENDIX C. Most importantly, onomatopoeiae appeared as constant productions in
regular contexts, and were used spontaneously (without parental elicitation) during joint
attentional frames. They were used extensively throughout the transition period as
spontaneous productions, for example when Ria saw pictures of the referents in a book. As
such they formed the greatest semantic group of words during this period. Ria used more
Slovak onomatopoeiae, since more forms are conventionalised in comparison to English.

123
5.3.5.2 Producing words
The first spontaneously used adult words appeared just before the first birthday and were
related to aspects of Ria’s life necessary for survival. As with onomatopoeia, early
productions gradually developed into forms that approximated the adult targets.

There were several early words that occurred frequently. The first was a Slovak baby-word
both parents used to refer to Ria’s breastfeeds, ‘hami’, ‘hamiki’ or ‘hami-hami’ meaning
‘eat’. The closest English equivalent often used by English speaking parents is ‘boobie’.
The mother usually asked ‘Prosíš/Dáš si hami-hami?’ “Would you like/Do you feel like
boobie?” Ria did not use this vocalization in other situations, only to request a feed.
Several variations emerged at first, [ham], [h h] or [am] (0;11.02). After the first birthday
the word was pronounced as [hama hama], thus approximated the adult word closer

(1;0.6), or [hamɪ] (1;0.4) in which case the production matched the target word. It was
often accompanied by a gesture that intended to achieve the result physically, as
demonstrated in the following video excerpt:

%sit RIA and MAM are playing on a rug, RIA is asking for a feed

*RIA [hama ham]

%act RIA is grizzling

*RIA [aːu aː] [ham ham ham]

*MAM Ham ham, dáš si hami, ano?

%eng Ham ham, would you like boobie, yes?

%act RIA is grizzling and reaching for MAM’s breast

*RIA [hamɪ]

(1;0.19)

For a short period Ria extended the meaning of the production to other food and eating in
general, as in the following examples:

- while the mother was preparing lunch and setting the table Ria was becoming
unsettled anticipating eating. She pointed at the food and leaned towards the food
uttering [hama hama] (1;1.20)

124
- as the family returned home after shopping Ria took a block of chocolate from a
shopping bag and unwrapped it, when the father asked ‘Ria what are you doing?’ she
responded [hamham] (1;3.15).

The reduplicated combination [ham ham] was also extended to a utensil used for eating:

- as Ria picked up a spoon during play she held it up and uttered [ham ham], with the
intended meaning being most likely ‘We eat with a spoon’.

The word was used in both language contexts, since it was a family word, and the father
also referred to Ria’s breastfeeds as ‘hami’. Thus it belonged to both languages. An
example from the English context occurred during play with dolls with the father, where it
was used in its extended meaning ‘food’ (Ria was not accustomed to a bottle with milk and
did not form this association):

- Ria passed a bottle to the father, who pretended to feed a doll, after which Ria uttered
[ham].

Other first words were appellatives for the parents. Ria used several variations to refer to
the mother [mama] (1;0.4), [mam] and [mamɪ] (1;2.7). Other variations appeared as well,
such as [mamajama] (1;2.3) or [mamaja] [maja], which were interpreted as ‘mama moja’
“my mummy”, a phrase the mother often used herself when picking up Ria. Later variant,
also used by the mother, was a diminutive ‘mamička’ “mummy”, produced most often as
[mamɪɟa], and occasionally with the consonant cluster [mamɪcka] (1;3.8). An appellative
for the father was at first produced much like the babbling combination [dada] or [dadada]
(1;2.7), however it was clearly distinguished by the situational context that the word was
referring to the father. Both appellatives assumed the adult forms [mamɪ] and [dedɪ]
(1;3.21) by the end of the transition stage.

Among appellatives Ria’s own name referred to self, e.g. when Ria spotted her reflection
in the mirror she pointed at it and uttered [ana] (1;1.17). Only days later the production
approximated the target word even closer [jana] (1;1.24), and it was also used when
pointing at a photograph (1;1.25). Her own name was also one of the first words used in
the form of a holophrase when requesting to do an action herself, replacing the usual
proto-word vocalization [əː], as shown in the following examples:

- Ria liked to watch the father while cooking, she requested to hold the spoon by
repeating her name until the parents interpreted the request appropriately.

125
- Ria uttered her name as she picked up a bottle of baby oil and applied it to her
stomach (1;2.21).

During this stage, however, reproduction of words after adults was more characteristic of
Ria’s speech. After the first birthday adult words emerged as consistent reproductions
during various routines and activities, following a new visual or auditory stimulation,
through sharing picture books with the parents as they were labelling objects in the
pictures, or after direct elicitation by parents, especially when appellatives of family
members were requested. Ria repeated appellatives slowly and clearly, such as [mama]
after a prompt ‘Povedz mama!’ “Say mama!” (1;0.27).

Thus word reproduction took place mostly within joint attentional frames. Ria was
attempting to produce either words she was familiar with or the key words from the
parent’s utterance. In a bilingual context she reproduced after the parent with whom she
was engaged in direct interaction, as in the following excerpt:

%sit RIA and MAM are playing with sorting blocks in the lounge, DAD comes
in and they talk about their day

*MAM And we made a snake, daddy!

*DAD Yes?

*RIA [snɛ]

%glo snake

*DAD You made a snake?

*MAM Ria, poď ukázať daddymu, akého hada sme urobili v knižnici, poď.

%eng Ria, come and show daddy the snake we made in the library, come on.

*DAD Mami said you made a snake.

*MAM Poď mu ukázať.

%eng Come and show him.

*DAD Where is it?

*RIA [x]

*MAM Because they’ve

*DAD Is that your snake? Cool!

126
*MAM changed the program, you know. We got there and yeah, they’ve changed
the program, they don’t have baby bounce anymore.

*RIA [xtə gov xɪc]

%glo xxx xxx xxx

%gpx RIA grabs the snake and carries it to DAD, holding up the snake

*DAD Snake! Ssss! Cool!

*RIA [ku]

%glo cool

*DAD You made it, did you? Ssss!

*RIA [ɪsss]

%glo ssss

*DAD You stuck decorations on? Cool!

(1;3.16)

Reproductions were the most typical form of production between 1;0-1;3, period during
which it was not possible to capture an exhaustive list. Examples were occurring in both
languages, Ria was imitating after the parent’s address in the respective language.

By age 1;2.30 Ria gained enough control of the articulatory system to be able to reproduce
many words within varied situational contexts with the mother and father. Occasionally
she was able to self-repair productions, as in the following examples:

uttering [ʒaʒa] followed by [aba] ‘žaba’ “frog”, (1;3.16);

uttering [baba] followed by [babo] ‘bábo’ “baby” (1;3.21),

thus attempting to reproduce the words correctly within the same situation.

Reproductions prevailed for approximately two months, after which Ria became confident
with the forms and reached readiness to produce spontaneous utterances. Many of the
spontaneous utterances were responses elicited by the adults in the form of open questions
about objects and people (‘Where is daddy?’ ‘What’s that?’ ‘Who spilt the water?’). But
most importantly, Ria also initiated interaction with words rather than gestures or proto-
words.

Spontaneous productions were preceded by Ria’s spontaneous choice to perform certain


actions (around 1;3.20). Rather than imitating actions she performed them spontaneously,
127
or after an adult’s request to do something, e.g. placing dirty clothes into the washing
machine, shopping items into shopping trolley, rubbish in the bin, replacing various
household objects where they belonged.

The first ten spontaneous words, shown in Table 5.1, were drawn from both languages:

Table 5.1 First 10 words

Age of
Phonetic form Target Source
first Meaning
and variations word language
use
Slovak
1. 0;11 [hama] [hamɪ] hami requesting feed
(Family word)
2. 1;0 [mama] mami mummy Either
3. 1;0 [ano] [aɲɔ] ano yes Slovak
[ana] [jana]
4. 1;1 Riana Riana Either
[nana] [ɲana]
5. 1;1.25 [am] tam there Slovak
6. 1;2 [dada] [dædɪ] daddy daddy Either
no (don’t do
7. 1;2 [ɲɛ ɲɛ] nie nie Slovak
that)
8. 1;3.20 [ɟɪs] this this English
9. 1;3.20 [je] yeah yes English
baby talk for
10. 1;3 [baː] bác Slovak
‘fall’

As shown in Table 5.1, these words were appearing one at a time, with a gap of
approximately one month inbetween, and the first milestone of 10 words was reached only
at the end of 1;3. Semantically, these words were mainly appellatives for the parents and
self, and function words. All of these words were used frequently by the parents.

On occasion, Ria used an utterance spontaneously in one instance, then stopped using it for
some time and resumed its use at a later stage:

[bu] ‘bus’ (1;2.1)

[kokoko] ‘kohút’ OR ‘kotkodák’ “rooster” OR “cluck-cluck” (1;3.25)

[jogu] ‘jogurt’ “yoghurt” (1;3.25).

There were also periods when a particular utterance was dominant over others, e.g. [baː]
‘bác’ “up-a-day” used often when Ria was learning to walk.

The end of the transition period was also marked by holophrastic use of words. One of the
first words used in this way was [mama] ‘mami’ “mum”. Its meaning was not limited to
asking for the mother, but to request services, direct the mother’s actions and indicate
128
ownership. The particular meaning was determined by the situational context, as shown in
the following examples:

- Ria pointed at toothbrushes uttering [mama], indicating agency. The parents’


immediate interpretation was that the mother was to brush Ria’s teeth, but Ria was not
satisfied until the mother interpreted the holophrase as brushing her own teeth (1;3.25);

- Pushing the stroller outside the room uttering [mama], requesting the mother to leave
a room where Ria did not want to be;

- Pointing at the vacuum cleaner uttering [mama], indicating agency;

- Pointing at the mother’s suitcase and uttering [mama], indicating ownership (1;3.26).

Towards the end of the transition period first attempts at combining words appeared:

[mama hama] ‘mama hama’ “mummy eat” - the mother and Ria were watching the
father prepare Ria’s favourite fruit, in anticipation Ria pointed at the mango (1;2.30)

[dada. dada ʃaʊ. dada. ʃaʊ. ʃaʊ.] ‘daddy. daddy shower. daddy. shower. shower.’ -
interacting with the father while he was providing a commentary on his actions
(1;2.30).

However, such combinations did not become typical of Ria’s speech until 1;7.

In summary, the transition period was a stage marked by rich communicative and
linguistic development. Firstly Ria’s social-cognitive development allowed her to
understand communicative intentions of others, and she was learning ways to direct their
attention. Interpersonal communication in joint attentional frames played an important role
in this development. Attempts at communication by means of protolanguage and
paralinguistic means served the achievement of needs and requests. The development of
comprehension as a precursor to speech was the prevalent aspect of this stage. Ria was
becoming familiar with adult words and phrases from both languages, including TEs,
related to her everyday contexts and physical environment. By the end of the period, she
responded appropriately to utterances in both languages addressed to her by the parents.

During this stage Ria also experienced imitative learning both in non-linguistic aspects of
development and in the use of language. She reproduced several adult words after the
parents, and gradually progressed into spontaneous productions of first words. First words
were used in a holophrastic sense. Ria also demonstrated awareness of her own
productions, gained better control of the articulatory system and consciously attempted to
approximate adult words.
129
6 Using words (1;4-1;6)
In Ria’s spontaneous productions several first words emerged in the transition period.
However, due to the small number of items that were used frequently (see section 5.3.5.2)
the decision was made to mark the onset of the one-word stage at the start of 1;4, when Ria
achieved the first milestone of 10 words in expressive vocabulary. Thus there is an overlap
of the stages, which emphasizes the gradual transition into adult language. While proto-
words and babbling were the main forms of expression in the transition stage, in the one-
word stage words became dominant.

The one-word stage ended towards the end of 1;6 when two and multiple-word
combinations emerged. This stage was marked by major growth in both comprehension
and production of adult words. Ria used words as the predominant form of
communication, and combined them with other forms, such as kinetic and gestural means,
often connected into sequences.

When Ria was unable to express her intentions with words she relied on gestures
combined with proto-words or pure babbling, for example the typical proto-word
vocalization [əːəːəː] used with a pointing gesture to request objects and services (1;4.19).
At times it appeared more gestures than in the previous stage were used. This could be
explained by the fact that her receptive vocabulary was growing rapidly, while the
expressive vocabulary did not reach a similar level yet, thus Ria aimed at communicating
and expressing her intentions in whatever way was available.

Babbling also served as a substitution strategy to fill gaps in Ria’s developing expressive
vocabulary. This was evident when she combined babbling with adult words, as in the
following example:

Ria was watching the father cook in the kitchen, while producing strings of babbling
combinations intensely adding her name [jana] ‘Riana’ at the end of the string, as if
explaining something about herself (1;6.6).

Babbling was also used in a ‘pretend’ way when Ria could not express herself using adult
language as yet. An example of this was while reading by herself, as in the following
excerpt:

%sit RIA is pretending to read MAM’s book

*MAM Čo tam je napísané?

130
%eng What does it say?

*RIA [ʃɛcɛtaːwado fɛf aue uwowo]

%glo xxx

*MAM Čo všetko tam je?

%eng What’s in there?

*RIA [bɛ ʃɛ cɛ ʃɛ]

%glo xxx

*MAM Je tam napísané?

%eng That’s what it says?

*RIA [ano]

%glo ano

%eng yes

*MAM Héj? No, prečítaj to mamičke.

%eng Really? Well, read it to mummy.

(1;6.23)

Babbling combinations were also used in play with sounds, such as an invented sound
combination [gogɪ gogɪ gogɪ] used playfully just for pleasure in different situations (1;5.1
to 1;5.22).

6.1 Early word development


According to Tomasello (2003, p. 87) there are two main aspects of the word learning
process. The first is the social environment that consists of numerous reoccurring social
interactions such as routines, activities and games that a child and her carers engage in. It
is these interactions that provide joint attentional frames and which lead to the learning of
cultural communicative conventions (Tomasello, 2003, p. 89). The second aspect is a
child’s own social-pragmatic capacity that allows her to participate in her social
environment (Tomasello, 2003, p. 89). She participates by coordinating her own attention
with the communicative intentions of other people, by learning to read such intentions, as
well as identifying specific utterances and interpreting their functional purpose within the
situation, and within the overall social context. How a child interprets linguistic symbols
contained in adult’s utterances is therefore dependent on the social context. She then learns
131
to produce the linguistic symbols in interactions in order to communicate her intentions to
others (Tomasello, 2003, p. 90). This process has been termed the interpersonal principle
by some authors who focus on the social context of such interactions (Halliday, 1975, p.
101; Painter, 2006). There seems to be a parallel with explanations that emphasize the act
of directing each other’s attention on the same referents, and thus creating joint attentional
frames around which interaction occurs (Tomasello, 2003; Tomasello & Farrar, 1986).

In this chapter the social-pragmatic word learning theory will be illustrated on the bilingual
data. In the following sections the underlying process of acquiring words in a bilingual
context will be described. Observations will be made on how Ria became familiar with
two words for one concept and learnt to produce them in appropriate contexts.

The data show first words were acquired through the process of socialization, through
daily interactions between Ria and the parents. Each parent-child dyad in the bilingual
family engaged in different contexts, guided by the cultural conventions of each language.
The data contain examples of interactions when Ria and either or both parents focused
their attention jointly onto an object or event. In these situations the parents directed Ria’s
attention and then verbalized the joint experience in the form of commentary whilst
repeating words, and thus allowing her to relate them directly to the concept represented
by a concrete physical object or event. Ria reproduced the words and received
acknowledgment of her attempts from the parents, who responded by repeating the
attempted word, giving further confirmation of the form. Ultimately Ria began to
spontaneously direct the parents’ attention by uttering the word for known referents while
pointing them out with a gestural or kinetic expression.

6.1.1 Rate of vocabulary growth


In the initial period there were two forms of word production:

 Words reproduced after an adult


 Spontaneously used words or responses to adult address.

Reproductions were direct repetitions of words used by adults immediately before. The
fact that a word is reproduced in a context does not show whether the child actually
understands the word, since it is difficult to determine whether a reproduction reflects her
ability to use a word productively. On the other hand, if a word is spontaneously produced
in an appropriate context, it is likely that the word is understood by the child (Pearson,
Fernández, & Oller, 1993, p. 113).

132
In the following analysis of vocabulary growth only spontaneous productions were
considered. This included Ria’s initiation of an interactional exchange as well as response
to adults’ address. Reproductions of words uttered by adults were not considered. While
they were most likely in Ria’s perceptive vocabulary, she also reproduced words which
she did not use productively until considerably later. However, they were considered in the
discussion of the phonetic shape of words in the following sections.

Conversely, spontaneous uses of onomatopoeiae and interjections were included in the


word count for several reasons. These words belong to baby-talk, nevertheless, they were
used by adults in the input and thus they were modelled to Ria as adult words. Moreover,
many of the Slovak onomatopoeiae were recorded and recognized as non-standard yet
conventionalised and frequently used words in Slovak CDS (Hlavatá, 1998, pp. 20-23).
Secondly, Ria’s productions approximated the adult forms consistently and they were
regularly used in appropriate contexts, thus fulfilling inclusion criteria suggested by
Vihman & McCune (1994, p. 522).

As mentioned in section 5.3.5.2, Ria used first adult words spontaneously in the transition
stage, starting around 0;11. However, rates of production were slow, and over the 5
months of the stage only several words were added to the vocabulary. The first milestone
in lexical development, the first 10 words was reached by the end of the transition period
at 1;4. Thus Ria entered the first-word stage using 10 spontaneous productions. Over a
span of 3 months during the one-word stage Ria reached other major word learning
milestones, and rapid vocabulary growth occurred:

 30 words reached within a period of one month, by the end of 1;4


 50 words reached by the first half of 1;5
 100 word milestone reached a month later at 1;6
 245 words reached towards the end of the first-word period at 1;6.

In the literature on language acquisition the 50 word milestone is followed by a vocabulary


explosion - a rapid exponential growth in a child’s vocabulary, as well as by early word
combinations (Cattell, 2007, pp. 5-6). This development was supported in the bilingual
data in this thesis as well. Overall, during the one-word stage vocabulary growth
progressed from 10 at the beginning of the stage to over 200 at the end of the stage within
3 months. The first 50 words used either in spontaneous productions or when responding
to an adults’ address during the one-word stage are reported in APPENDIX D. Bilingual

133
aspects of the vocabulary composition will be discussed in section 6.2. Emergence of
combinatorial speech will be addressed in section 7.3.

6.1.2 Reproduced words


Reproduction of words adults uttered in interactions with Ria emerged around 1;1 in the
transition stage, and included onomatopoeiae, appellatives for the important people, some
function words and other words related to immediate contexts. Initially Ria chose to
reproduce key words that were the focus of a joint attention of any given utterance within a
particular context. She reproduced words after both parents, in bilingual contexts often
both TEs simultaneously. This usage reflected the parents’ strategy to introduce both TEs
in triadic interactions, as described in sections 5.3.4.1 and 6.2.1.

Ria’s early forms were simplified and approximated the adult forms phonetically. Initially
only the first sound (in words with difficult clusters) or the first syllable was used. When
the stress fell on the second or last syllable Ria reproduced the stressed syllable:

[maː] ‘tomato’, natural stress;

[cɪ] ‘deti’ “children”, unusual stress, reproduced after the mother who stressed the
second syllable for emphasis (1;5.3).

However, it was possible to check correct association of such approximations with an


object or event, since they were observed on several occasions and consistently used in
relation to the same referents (Vihman & McCune, 1994, p. 522).

Reproductions were used extensively throughout the one-word stage and formed the
greatest part of Ria’s productions for several months. By 1;5.6 they became so frequent
that it was difficult to capture and record all instances. While spontaneous productions
were used as well, they became more prominent only gradually in the second half of the
stage. Around 1;6.18 the diary records state that spontaneous utterances were used as
frequently as reproduced utterances, and towards the end of the stage spontaneous
productions prevailed over reproductions. However, overall reproductions formed the
largest part of Ria’s productions during the one-word stage.

6.1.3 Spontaneous productions


Regular spontaneous productions emerged around 1;4 as holophrases on their own or
complemented with gestures, which emphasised their function. Thus the word as a
communicative act carried a more complex meaning than merely a label for a particular

134
referent. By combining the word with other modalities Ria was able to emphasise the
intended meaning of the utterance, as in these examples:

 Requesting service or object


- passing a book to the mother uttering [dɪs] ‘this’ –a request to read (1;4.2)
- pointing at her plate uttering [ɟʊɟʊ] ‘čučoriedku’ “blueberry” – a request to eat
(1;4.19)
- holding the father’s finger, physically trying to move him, uttering [mamɪ] ‘mami’
“mummy” – a request to follow the mother (1;4.8)
 Requesting a parent to act
- passing an object to the mother uttering [mama] ‘mama’ – specifying who is to
perform the action (1;4.19)
 Maintaining interaction
- [aɲo] ‘ano’ “yes” used with the nodding gesture - agreement
- pointing at a photograph of the father uttering [dada] ‘daddy’
- pointing at her own cheek uttering [jaɲa] ‘Riana’, meaning ‘this is Riana’
- cuddling the mother or favourite toys uttering [mɔː] ‘moja’ “my” (1;4.21)
 Expressing self and reacting to the environment
- uttering [ɲamɪ ɲamɪ ɲamɪ] ‘malilinký’ “tiny” while putting her two index fingers
together as if indicating the meaning in gestural modality as well - emphasizing the
word with a gesture (1;4.8)
- uttering [baba] ‘bábo’ “baby” and imitating baby’s cry after she had heard an infant’s
cry (but did not see the infant), thus expressing the actor
- uttering [ɲana] ‘Riana’ and pointing at red pants she was wearing, which were of the
same colour as a red object in a picture book, thus identifying her experiences with
something represented in a picture (1;4.19)
- pointing at herself and uttering her name as a response to a question ‘Kto to urobil?’
“Who did it?”, when something was spilt or messed up (1;4.26)
- pointing at her own head and uttering her name while looking at a picture of a hat -
indicating she was also wearing a hat
- when Ria achieved a new skill, such as manipulating shapes to fit through a sorting
block, she pointed at her face and uttered her name – indicating agency.

Thus the gesture which accompanied holophrases suggested the meaning, for example the
use of Ria’s own name in various contexts above could be glossed as statements or
exclamations:
135
“Riana is wearing red pants/a hat.”

“Riana did it!”

“Riana managed to do it!”

Some words were used spontaneously as a holophrase without a gesture and included
mainly Ria’s own name and appellatives of other family members. The function of
appellatives was regulatory, and it was focused on the person to perform the action:

- when Ria wished to perform a particular action herself, she uttered her name [jana]
‘Riana’, which could be glossed as “Riana will do it!”
- when Ria approached a hot object such as an oven or hot cup, and reached out to
touch it, she immediately retracted her hand and uttered [mamɪ # mamɪ] “mummy #
mummy” glossed as ‘Mummy will do it.’ (1;5.27).

Other holophrases conveyed Ria’s reactions to the environment:

- when feeding ducks Ria repeated the word [gaga] (1;4.2) ‘kačka’ “duck” several times

- the Slovak interjection [baː] for ‘bác’ was used spontaneously whenever an object fell,
as well as when Ria fell while walking.

Although in the above examples Ria produced the words spontaneously, they reflected the
usual comments provided by the parents in similar situations. Thus the items she had learnt
were observed in multiple object-label pairings and salient in the input, mirroring Pruden,
et al.’s (2006, p. 277) findings on associative mapping of words onto referents before 18
months.

6.1.4 Attention to language


According to De Boysson-Bardies (2001, p. 150) children arrive at language through
different approaches and different choices they make when using language as a tool for
communication. She states that “infants have not all noticed the same aspects of the
language during the course of the first year” (de Boysson-Bardies, 2001, p. 151). In the
initial stages of word learning this is reflected in different styles of language acquisition,
which were observed in several studies of French and American children. Using a cross-
linguistic comparative approach learning styles were determined by a child’s attention to
and choices in two aspects of language: strategies of production and the types of words
generally used. Two basic tendencies were identified (de Boysson-Bardies, 2001, pp. 151-
152):

136
 referential or analytic - typical of the use of mainly nominals in terms of word types,
while in production paying attention to phonetic structure in syllables, which results in
mainly monosyllabic words.
 expressive or holistic - typical of the use of more varied word types such as predicates
and function words, with frequent fixed expressions, while in productions focusing on
syllabic rhythm and intonation.

However, De Boysson-Bardies (2001, p. 175) emphasizes that children do not adopt a


single style, instead there are many variants depending on a child’s focus of interest, their
mother’s style, and the native language, since the structure of the language guides the
child’s choices. In a bilingual child the presence of two languages undoubtedly influences
her choices further.

Diary records of cumulative vocabulary showed Ria chose a style that used aspects of both
the expressive and referential tendencies. She tended to use the analytic style in
productions. She focused on the word as a basic unit and articulated productions clearly
from early on. First words were monosyllabic, however, initial sounds varied phonetically,
apart from a small number of cases. She tried to keep to the structure of the target word,
and managed to produce familiar words as closely as possible to the target. Where a sound
was difficult to pronounce in certain positions, she developed a pattern of regular
substitutions. Familiar words were used with high frequency.

However, Ria also used some fixed expressions, which are more characteristic of an
expressive style. On the other hand, the general trend in Ria’s choice of the first word
types suggested she preferred the expressive style, as discussed in section 6.1.4.2.

6.1.4.1 Strategies of production


First words produced by Ria did not have stable forms. Single word could be realized by
several variants, while each was approximating the target word differently, more or less
closely, until the form stabilized. Such gradual development was also observed by other
authors (de Boysson-Bardies, 2001, p. 145). In general, a word was first produced in a
form less similar to the adult word, and gradually developed into a form that more closely
approximated the target word, as in these examples:

Early Transient Stabilized Meaning

[mama] [mamɪ] (1;4.21) ‘mami’ “Mummy”

[dada] [dædɪ] (1;4.21) ‘daddy’

137
[aːʊ] [ɲaː ɲaː ɲaː] (1;4.12) [naːʊ] [mnaːʊ] (1;4.21) ‘mňáu’ and ‘meow’

[m m] [muː muː] (1;4.21) ‘múmú’ or ‘moo moo’

[ɲamɪɲamɪ] [maɲɪmaɲɪ] [mɪɲa mɪɲa] ‘malilinký’ “tiny”

(1;4.8) (1;4.21) (1;5.6)

Ria’s phonological development was evident mainly through reproductions, which were
the first to approximate adult words more closely in pronunciation. She attempted to
approximate the correct form as much as possible by repeating the word several times in
one instance:

[tado] and [lɛta] - reproducing two different forms of the Slovak word ‘lietadlo’
“aeroplane” after the mother (1;4.8);

[paː] [pawɪ] [papɪ] [paːɪ] [pabɪ] – attempting to reproduce the adult word ‘pavúk’
“spider” as closely as possible while seeing a spider for the first time (1;4.26);

[pa. pa. paʊː] attempting to reproduce ‘pavúk’ “spider” in a different instance (1;6.19);

[apu] and [æp] ‘apple’ reproduced after the father in an English context.

Thus she had the correct form in her perception and reproduced several possible variants in
an attempt to approximate to the target word. In some instances the reproduced and
spontaneous forms differed, as demonstrated by the following examples:

- the Slovak word ‘vlas’ “hair” was first reproduced as [vl] (1;5.21), then produced
spontaneously as [ac] (1;5.21), but in subsequent tokens the word was produced as [vaː]
(1;6.18), however, in the next developmental stage it changed momentarily to [wa]
(1;7.2) before it was pronounced as a closed syllable [vas] or [vasɪ] ‘vlasy’ in the plural
(1;7.28);

- the Slovak word ‘deti’ “children” was at first reproduced and elicited in a response as
[cɪː] (1;5.3), but when produced spontaneously, the form [ɟɛ] (1;5.21) replaced it.

However, the change in the second example possibly occurred due to the fact that in the
first instance the mother applied stress on the second syllable, momentarily resulting in a
greater salience of that syllable to Ria. By the time Ria produced the word spontaneously,
enough examples of the word appeared in the input, and thus Ria adjusted the form
accordingly.

Spontaneous productions, on the other hand, were undergoing natural phonological


processes. Before Ria reached the 100 word milestone she relied on various strategies of
138
production for a limited time (aged 1;4-1;6). In some words, and only at the beginning of
the one-word stage, the very first tokens of spontaneous productions approximated the
adult form closer in both the sounds produced and the syllabic structure, while in
subsequent tokens, the word was reduced to initial syllable or sound, fitting a typical
pattern of production. Thus the words were subject to a reduction process, as shown in the
following list:

Initial form Subsequent tokens Meaning

[paːlɪ] (1;4.9) [paː] (1;6.16) ‘páli’ “hot/burning”

[konɛno] (1;4.9) [kɔ] (1;6.10) ‘koleno’ “knee”

[kɛwɔ] (1;5.3) [kɛ] (1;5.18) ‘kvietok’ “flower”

[xɔwɔ] (1;5.10) [x] (1;5.26) ‘chlpy’ “body hair”

[ftaː] (1;5.12) [ʃaː] (1;5.21) ‘vtáčik’ “birdie”

[aʊɔ] (1;5.12) [aːʊ] 1;6.13 ‘auto’ “car”

[ɔkʊ] (1;5.22) [ɔ] (1;6.2) ‘okuliare’ “glasses”

[taːwa] (1;6.4) [taː] (1;7.1) ‘tráva’ “grass”

[sɛɟɛ] (1;6.10) [ʒɛ] (1;6.25) / [zɛ] (1;7.1) ‘zelená’ “green”

The initial tokens showed mainly omission errors with only few substitutions, while the
form was relatively close to the adult form. The later tokens reflected a specific
monosyllabic/initial sound pattern, which departed more from the adult form.

The initial tokens were in fact produced in a way similar to several most frequently used
words important for survival, which developed in the initial period of word acquisition.
Ria produced stable forms of these words which closely, if not entirely, approximated the
target words. These included appellatives for the parents, Ria’s name and request for feed
(‘mami’, ‘daddy’, ‘Riana’, ‘hami’), etc. These important words were also produced as
wholes, close to the adult form.

After the initial period, approximately after the first 50 words, productions seemed to
stabilize into a regular pattern. This process was evident with majority of the later words
produced spontaneously across both languages, a developmental pattern also noted by
Vihman (2002, p. 246). Vihman described a similar mechanism underlying early
productions. She termed the early words selected and later words adapted. Early words
reflected the adult forms much closer, and were produced as individual words. However,
139
once regular patterns were found in production, which occurred around the 50-100 word
milestones, words tended to be adapted and produced according to templates. This
development was taken to “signal that the child has begun the process of building a
phonological system” (Vihman, 2002, p. 250).

In Ria’s data several reduction processes were observed:

 Reduction to first syllable or the initial sound, e.g. [ʊ] ‘uhorka’ “cucumber”, [ɔ]
‘ostrý’ “sharp”
 Omission of the initial syllable resulting in a monosyllabic form of the stressed
syllable, mainly in some English words where the stress was on the second syllable,
e.g. [ma] ‘tomato’
 Omission of mid-word sounds in some longer words resulting in a monosyllable
formed by the initial and final sounds, e.g. [dɔ] for the Slovak ‘dievčatko’ “little girl”,
where the initial ‘d’ and final ‘o’ were retained to form a monosyllabic word.

Initial consonant clusters in words were simplified according to three possible patterns:

 Reduction of consonant clusters to the initial sound, e.g. the initial cluster hr reduced
to single sound h in [haː] ‘hrášok’ “sweet pea”, pr reduced to p in [paː] ‘prášok’

“washing powder”, gr reduced to g in [giː] ‘green’, mr reduced to m in [maː] ‘Mráz’


“Santa”
 Omission of unstressed word initial sound in a consonant cluster, e.g. [nʊ] ‘dnu’
“inside”, where the initial dental stop is omitted, since it is unstressed in normal
speech.
 Substitution of clusters (consisting of a fricative s and other sounds) by an initial ʃ,
e.g. sl substituted as [ʃiːp] in ‘sleep’, sp substituted as [ʃɪ] in ‘spi’ “sleep”, sv
substituted as [ʃɛ] in ‘svetielko’ “light”.

Thus the most common structure of early words was a monosyllabic CV or disyllabic
CVCV in familiar words. By the end of the stage, 1;6 - 1;7, Ria produced more complete
words:

 monosyllabic words were produced complete with a closed syllable structure CVC,
e.g. [sɛm] ‘sem’ “here”, [mɪʃ] ‘myš’ “mouse”, [aɪs] ‘ice’, [wɔm] ‘von’ “outside”,
[fam] ‘fun’
 polysyllabic words were reduced to a single closed syllable, e.g. [zɪp] ‘zipper’, [kɛm]
‘krémik’ “body cream” (1;7.1), thus retaining the first closed syllable, or omitting

140
middle syllables and retaining the word final sound, which created a monosyllabic
form
 Difficult sounds in the middle or final positions following a vowel were omitted or
substituted, mainly the lateral approximant l and fricative labiodental v, thus creating a
diphthong, e.g. [paːʊ] ‘pavúk’ “spider”, [baːʊ] ‘ball’, [dɔʊ] ‘dolu’ “down, [ʃɪʊ]
‘šiltovka’ “cap”, [faːʊ] ‘fall’.

Difficult sounds within words were simplified, such as reducing consonant clusters to a
single sound, usually the first sound of the cluster. Overall, words extended in length rather
than being reduced to one sound or initial syllable, and assumed more varied structures.

The same sound substitutions appeared in both languages, such as the final nasal alveolar n
in monosyllabic words with the nasal bilabial m, as in [wɔm] ‘von’ “outside” (1;6.22) and
[fam] ‘fun’ (1;7.1). Ria also used similar reduction processes in productions in both
languages. This led to the emergence of homophones within and across languages.
Homophones had the same form but different meanings, which could be ascertained by the
parents only within the specific situational context. The correct meaning was evident since
at this stage Ria used the pointing gesture when labelling objects. Table 6.1 demonstrates
monosyllabic words, some used for two or more concepts in one language, others for
concepts across the two languages. Meanings of the homophones could belong to different
word categories. However, the homophones reflected two or more different sound-
meaning pairings (Pearson, et al., 1993, p. 101), and the monosyllabic form was a result of
the reduction process.

Table 6.1 Monosyllabic homophones

Target words and language


Monosyllabic
Meaning 1 Meaning 2 Meaning 3 Meaning 4
form
vtáčik “birdie” outside hračky “toys”
[ʃaː]
Slovak English Slovak
mušlička
mucha “fly”
[mʊ] “seashell”
Slovak
Slovak
jahody
jablko “apple”
“strawberries”
[ja] Slovak
Slovak
(later [ja])
(later [va])
paprika paradajka
[pa] “capsicum” “tomato”
Slovak Slovak
stolička slon “elephant” shorts OR šoférovať “to
[ʃɔ]
“chair” Slovak šortky drive”
141
Target words and language
Monosyllabic
Meaning 1 Meaning 2 Meaning 3 Meaning 4
form
Slovak Either language Slovak
Piglet (toy
pretty piť “to drink” princess
[pɪ] character)
English Slovak English
Either language
korytnačka
koleno “knee”
[kɔ] “turtle”
Slovak
Slovak
spať “sleep” sleep slivka “plum”
[ʃɪ]
Slovak English Slovak
knižka “book” keksík “biscuit”
Slovak Slovak kíblik “bucket”
[kɪː]
Slovak
(later [kɪ]) (later (kɛ])
(Dedo) Mráz
má “has”
[maː] “Santa”
Slovak
Slovak
Meeka (story
meat
[miː] character)
English
Either language
semiačka sedí “is
červená “red” svetielko “light”
[ʃɛ] “seeds” sitting”
Slovak Slovak
Slovak Slovak
noha “leg OR
nos “nose” nôž “knife”
[nɔ] foot”
Slovak Slovak
Slovak
marhule
tomato maco “teddy”
[ma] “apricots”
English Slovak
Slovak
postaviť “to pohárik podprsenka
postieľka “bed”
[pɔ] stand up” “glass/cup” “bra”
Slovak
Slovak Slovak Slovak
bicykel
bib
[bɪ] “bicycle”
English
Either
uhorka
ucho “ear”
[ʊ] “cucumber”
Slovak
Slovak
ešte “more” elipsa “ellipsis”
[ɛ]
Slovak Slovak
čelo češe sa “is
[tʃɛ] “forehead” brushing hair”
Slovak Slovak

Since the words in the above list reflected multiple sound-meaning pairs, and their
appropriate and consistent use in context was confirmed for each, they were counted for
each meaning separately when measuring Ria’s expressive vocabulary.

142
6.1.4.2 Vocabulary composition
According to de Boysson-Bardies (2001, pp. 178-183) word learning is influenced by
cultural variations more than any other aspect, such as content of parental speech
addressed to children and children’s individual style or preferences. This is because such
aspects are influenced by the surrounding culture:

“…each child must speak the language of his or her environment and conform to certain
formal and social customs imposed on both the language and the culture. Children’s
vocabulary grows out of contact with the adult language, which, from the outset, dictates
the objects in the world that must be seen and learned, the manner in which these objects
are to be spoken of, and the modes of expression that allow children to be recognized as
speakers and understood.” (de Boysson-Bardies, 2001, p. 177)

Among the languages in the study (French, American English and Swedish) nouns were
the most represented word types in children’s vocabularies (p. 182). Certain semantic
categories were identified as constant: nominals related to familiar people, objects
necessary to survival, household objects, animals and games. Among other word types
social words were represented. Similar semantic categories were observed among Ria’s
first words as well. The following paragraphs describe Ria’s choices of new types of
words.

As mentioned in section 5.3.5.2 the order in which first words were acquired was guided
largely by Ria’s needs and interests - the important people, objects and events in the
environment and objects which captured her interest. New types of words appeared first as
reproduced productions. Initially, the dominant class were nouns, followed by
onomatopoeiae, verbs, adjectives and function words. Towards the end of the period
reproductions of more varied function words appeared. For example the first prepositions:

- the Slovak ‘v’ “in OR at”, was reproduced in a form of a blend with the following
word. Thus the blend had a form of the preposition ‘v’ + the initial sound of the
following word, [vɔ] ‘v robote’ “at work” (1;6.22); in this blend Ria omitted the initial

difficult liquid [r] and instead used the following vowel [ɔ], creating a monosyllabic
word;

- the Slovak [pɪː] ‘pri’ “by”, while the father and Ria were looking for her shoes, the
mother called out in Slovak where to look, uttering ‘pri kresle’ “by the sofa”, which Ria
reproduced (1;6.28). It appeared Ria chose to reproduce the preposition rather than the
noun it was related to because the location, as indicated by the preposition, was the key
143
information communicated, thus emphasising the location, rather than the noun, which
would have been more precise information.

The first 10 spontaneous productions reflected the most important concepts in Ria’s life:

 names for the important people, including her own name


 an item necessary for survival – requesting feed
 function words needed for referral to objects and events (‘tam’ “there”, ‘this’) and
words that facilitated correct identification of actions, events and objects for both Ria
and the parents (‘ano’ “yes”, ‘yeah’, ‘nie nie’ “no no”),
 baby-word related to the dominant skill of this stage - learning to walk (interjection
‘bác’ “up-a-day”).

Overall among the first 245 spontaneous productions in the one-word stage by far the most
common types of words were nouns and proper nouns, adding up to 141. Various semantic
categories were found among nominals:

 Proper nouns - appellatives, names of favourite story/toy characters


 People (boy, girl, etc.)
 Survival – objects such as food and drink and biological needs such as washing,
toileting
 Body parts
 Animals
 Outdoors and nature
 Toys
 Household objects
 Utensils and crockery
 Clothing
 Games
 Transport.

Nominals showed the greatest variation. Ria’s emerging awareness of self as a separate
being was manifested in the interest of body parts, items of clothing and games she was
able to play. Various objects related to daily activities and the physical environment of the
house and garden were also well represented, especially objects Ria was able to touch,
handle and manipulate, and plants and animals which were especially of interest. The food
category consisted of many words for seasonal fruits the family grew in the garden, which

144
were not used after the season finished. Lastly, objects seen generally outdoors on walks
and in parks were also represented.

Predicate forms (verbs and adjectives) were also used from early productive stages,
although much less than nouns. A total of 22 verbs were found, while only a few appeared
initially, and most emerged in the second half of the stage. Verbs included mainly words
which described actions Ria was able to perform (dám “I will give”, sedí “is sitting”,
vysávam “am vacuuming”), actions she wished to do (piť “to drink” , see, go, spi “sleep”)
or was requesting others to perform (otvoriť “open”, umyť “wash”, postaviť “to stand up”)
or observed others to perform (píše “is writing”, šoférovať “to drive”, cooking). States
were also among the verbs (gone, páli “it’s hot”, má “has”, chceš “you want”).

A total of 17 adjectives appeared throughout the stage, describing characteristics of people,


properties of objects (malilinké “tiny’, big, mokré “wet”, pretty, hot, ostrý “sharp”) and
several colours.

Onomatopoeiae were well represented especially at the beginning of the stage, and they
also belonged to the words with the greatest number of tokens per word in the early
months of the stage. Towards the end of the stage they were not used as often. Some
onomatopoeiae overextended to label the animals that produced the sound. Words for the
sounds produced by inanimate objects were also used (bell, train).

Ria used 7 interjections regularly during the stage and they were either related to games
(kukikuk “peek-a-boo”, šúch “weee”) or expressing Ria’s surprise (wow, jáj “oh”, oh oh)
as well as a baby-word used to announce something fell (bác “up-a-day”), often used in the
function of a verb.

Several examples of function words such as adverbials (7), locatives (6), deictics (8),
pronouns (2) and a conjunction also appeared among the first 245 words. Ria used varied
deictics in both languages (this, tam “there”, here, there, that, sem “here”, to “that”).
Locatives related to various places Ria wished to go (von “out”, dnu “inside”, outside),
requests to be picked up and placed back down (up, dole “down) and a request to get
dressed (on). Adverbials included negators (nie nie “no no”, nie “there isn’t”), affirmatives
(ok, ano “yes”), denoted quantity (veľa “lots”), as well as used to request more food (more,
ešte “more”). Among first pronouns was the associative dual ‘my’ “we” and interrogative
‘kde’ “where”. One conjunction ‘and’ was used. Lastly, 4 social words were regularly used
from early on (cool, please, pá/papa “bye”, ta).

145
Thus overall, Ria’s repertoire included mostly nominals, while other parts of speech, such
as predicates, function words and onomatopoeiae were also represented from the very
early stages.

6.1.4.3 Fixed expressions


Some early utterances were fixed expressions produced in one breath as a one-word
utterance. This was the case mainly with speech formulae Ria heard in her input
frequently, and were most likely learnt as one unit (de Boysson-Bardies, 2001, p. 199).
Only few of these were used regularly, and the first instances of such expressions were tied
to specific contexts. The early context-bound examples are shown in Table 6.2.

Table 6.2 Fixed expressions

Age Expression Meaning Context


looking inside a glass trying to
‘kde je?’
1;4.26 [ɟɛjɛ] find berries from a smoothie
“where is it?”
she had just finished
[cɔtɔ]
‘čo je to?/ čo to je?’ requesting labels for referents
1;5.6 [cɔjɛ]
“what is that?” while reading
[cɔjɛtɔ]
‘mami moja’
1;4.9 [mamajama] expressing affection
“my mummy”
1;5.14 [madædɪ] ‘my daddy’ expressing affection
1;5.20 [jamam] ‘ja mám’ “I have” reacting to environment
1;5.23 [pɛɪanʊ] ‘pre Rianu’ “for Riana” requesting object
1;6.4 [tʊɪɛ] ‘tu je’ “here it is” found an object
1;6.19 [wos] ‘what’s’ requesting label for object

Ria’s use of the Slovak wh-questions reflected the mother’s usual discourse - looking for
objects and probing for comprehension of various words while sharing books. The
intonation was rising, which is typical for a Slovak interrogative sentence. However, the
forms were not constant, since Ria attempted to pronounce them as close as possible to the
target form, producing several variants. Especially at a later stage, whilst trying to control
her articulation, Ria produced new variants: [ʃɛɟɛ] [ʃɛdɔ] [ʃɔdɔ] (1;5.12) and a shorter
[ʃɛdɪ] [ʃɛdɛ] (1;5.17). The latter variant appears to be a bilingual blend ‘Čo this/there?’
“What‘s that/there?” combining the two languages. Approximately a month later Ria also
used the English wh-question equivalent. By this stage the expression was dissociating
from the previous context of book reading, and Ria pointed at various objects in her
environment. As De Boysson-Bardies (2001, p. 148) also noted, the individual constituents
of the questions were not used on their own, only as parts of the fixed expressions.

146
Conversely, the affectionate expressions, although uttered in the form of a single unit,
contained a constituent which was used on its own regularly – parents’ appellatives. Since
these were affectionate terms, the expressions were not absolutely fixed, a phenomenon
which was also observed by De Boysson-Bardies (2001, p. 168). The other expressions
were used in single instance only during this stage.

6.2 Bilingual Repertoire


For a BFLA child the word acquisition process differs from a monolingual child of the
same age. Where a monolingual needs to learn one sound-meaning pairing, a bilingual
child’s lexical knowledge will spread across two languages, since she interacts with
speakers of two languages. To assess a BFLA child’s vocabulary inventory accurately,
lexical knowledge in both languages needs to be considered by assessing her lexical
representations in both languages. According to Pearson (1998, p. 350) there is no
consensus on a definition of lexical representation. It is generally referred to as a sound-
meaning pairing. Where a monolingual child acquires one concept and the word linked to
it, a BFLA child will have to learn two words for each concept. Thus she will learn a
Translation Equivalent (TE) pair.

It is likely that knowing a lexicalized concept in one language will facilitate learning of the
equivalent word in the other language, if the concept is shared across the languages.
However, this is dependent on the extent to which the two languages in contact share
concepts and have direct translation equivalents for words (Pearson, 1998, p. 361). Thus
when a bilingual child acquires a conceptual representation for an item in one language,
she may not necessarily be able to link a cross-linguistic synonym from the other language
with that representation. Instead she may need to acquire a different concept with a link to
another word altogether.

Similarly, there are considerable individual differences in the amount of TEs bilingual
children use productively (Pearson, 1998, p. 368). However, studies show that BFLA
children do in fact produce TEs as soon as they start producing adult words in their early
utterances (Holowka, et al., 2002, p. 241). The number of TEs in a child’s repertoire is
dependent on many factors, such as quantity and regularity of input in the two languages,
individual differences, language preference, the most frequently used language, language
of the social network, cultural differences between the languages and other.

To measure a bilingual infant’s vocabulary Pearson et al. (1993, p. 102) suggest multiple
measures consisting of two Single measures in each language, and two double measures

147
that asses the infant’s abilities in both languages taken together, while accounting for
shared abilities between the languages (Total Vocabulary and Total Conceptual
Vocabulary). The Total Vocabulary sums all the words in one language and all the words
in the other language, and subtracts phonetically similar doublets. The Total Conceptual
Vocabulary counts the number of concepts known by the child across the two languages,
thus when she knows a different word in each language for one concept, the concept is
counted once only.

However, these measures were based on words coded on the MacArthur Communicative
Development Inventory (CDI) and as such do not fully reflect true language use and a
child’s complete lexical ability. They do not take into account words that could belong to
either language, such as proper nouns, or words in which source cannot be easily
determined. Thus inadvertently they underestimate a child’s lexical ability. In assessing a
bilingual child’s vocabulary more attention needs to be paid to the shared aspects as well
as differences between the languages. Moreover, the people who model the languages in a
bilingual child’s environment influence the acquisition of the languages, which is reflected
in a child’s productions. Thus a child’s vocabulary needs to be assessed in light of what is
modelled as appropriate language use in the input. Therefore, in this thesis the following 6
measures are suggested to describe the bilingual child’s lexical development adequately:

 Combined total vocabulary including all lexical forms used


 Total conceptual vocabulary that accounts for the concepts known across languages
(each concept is counted once)
 Language neutral lexical items that belong to both languages
 Indeterminate lexical items
 Two separate measures for each language.

In section 5.3.4.1 evidence was presented that Ria’s comprehension of TEs preceded
production of first words, and it was evident as soon as word comprehension emerged
around 0;10. The same was observed in production. From 1;4 – 1;6, during the one word
stage, Ria’s productive combined total vocabulary was 245 lexical items. The total number
of TE pairings used was 25. This means that of the 245 spontaneously produced words
(lexical items) Ria’s total conceptual vocabulary consisted of 220 representations. Of the
25 TE pairs, 21 shared conceptual representations consisted of two lexical items – one for
each language – for a total of 42 items. The 42 doublets, i.e. members of TE pairs are
shown in Table 6.3.

148
Table 6.3 Translation equivalents in expressive vocabulary

Age Slovak member Age English member


[hh] [ww ww]
1 1;4.9 1;4.21
‘hav hav’ ‘woof woof’
[baːba] [bebɪ]
2 1;4.9 1;5.14
‘bábo’ ‘baby’
[anɔ] [jeː]
3 1;4.9 1;4.9
‘ano’ ‘yeah’
[paːlɪ] [ho]
4 1;4.23 1;6.16
‘páli’ ‘hot’
[konɛno] [niː]
5 1;4.23 1;6.16
‘koleno’ ‘knee’
[am] [de]
6 1;4.23 1;6.16
‘tam’ ‘there’
[aʊɔ] [kaː]
7 1;5.10 1;5.10
‘auto’ ‘car’
[kʊː] [kiː]
8 1;6.19 1;5.10
‘kľúče’ ‘key’
[ma] [eːn] [eːnɪ]
9 1;5.14 1;6.25
‘mravce’ ‘ant, ants’
[vaː] [teɪn]
10 1;6.29 1;5.26
‘vláčik’ ‘train’
[ʒʊ] [tɪ]
11 1;6.16 1;5.14
‘zuby’ ‘teeth’
[ja] [apu] [æp]
12 1;6.4 1;5.14
‘jablko’ ‘apple’
[ɛ] [moː]
13 1;6.16 1;5.18
‘ešte’ ‘more’
[vɛjɪ] [vɛː] [biː]
14 1;6.16 1;5.26
‘veľký’ ‘big’
[tɔ] [da]
15 1;6.16 1;6.4
‘to’ ‘that’
[sɛɟɛ] [giː]
16 1;6.10 1;6.25
‘zelená’ ‘green’
[a] [aɪʃ]
17 1;6.16 1;6.31
‘ľad’ ‘ice’
[pa] [ma]
18 1;6.16 1;6.31
‘paradajka’ ‘tomato’
[ma] [beː]
19 1;6.16 1;6.16
‘maco’ ‘bear’
[nɔ] 1;6.22 [na]
20 1;6.16
‘nôž’ ‘knife’
[wam] [ʃaː]
21 1;6.22 1;6.31
‘von’ ‘outside’

In addition, in 4 TE doublets one English form was paired with two meanings in the
Slovak language. With these items the meaning is expressed with different words in the

149
Slovak language, while in English the same word carries both meanings, therefore each
pairing was counted as a separate concept, as shown in Table 6.4.

Table 6.4 Translation equivalents in expressive vocabulary - continued

Age Slovak member Age English member


[kaga] 1;5.6 [puː]
1;4.30
‘kakala’ ‘did a poo’, verb
[hɔbw] [puː]
1;5.10 1;5.6
‘hovienko’ ‘poo’, noun
[tʊ] [hi]
1;6.4 1;5.27
‘tu’ (nonstandard from ‘tuto’) ‘here, at this place’
[ʃɛm] 1;5.27 [hi]
1;6.16
‘sem’ ‘here, to this place’

The acquisition of TEs varied considerably. For each pair it ranged anywhere from
doublets being used on the same day in different contexts, up to a time lag of 2 months
between spontaneous productions of the doublets in a pair. However, even during the lag
periods Ria showed comprehension of the TEs, and sometimes even reproduced the word
before she used it spontaneously.

There were 14 language neutral lexical items. This category subsumes words which could
belong to either language since they were used by speakers of both languages in the input,
although they actually originated in one (e.g. English ‘Piglet’ used by both parents). It
includes all proper nouns, appellatives, names of toy/story characters used by both parents,
and brand names used by the family to refer to specific items, as shown in Table 6.5.

Table 6.5 Language neutral words

Age Word Meaning


1;4.9 [mamɪ] mami “mummy”
1;4.9 [hama] [hamɪ] hami “boobie”
1;4.19 [jana] [nana] [ɲana] Riana
1;4.20 [dædɪ] daddy
1;4.23 [oke] o.k.
1;5.18 [ɔmɪ] Romi
1;5.27 [gede] grandad
1;6.3 [miː] Meeka (cartoon character)
1;6.10 [ne] Nela
1;6.16 [mɪ] Miša
1;6.28 [pɪ] Piglet (toy & story character)
1;6.28 [wɪ] Winnie (toy & story character)
1;6.28 [tɪ] Tiger (story character)
1;6.31 [ʃe] sketchers (brand name of shoes)

150
The indeterminate category comprised of 9 words for which the target language was not
easily determined. This was the case with onomatopoeiae which were homophones and
words that were close cognates in the two languages. The category also subsumed several
words that were different in the two languages yet shared some sounds. Due to
phonological processes under which Ria produced words the resulting simplified form
(usually initial sound/syllable) made it impossible to determine the target language. The
possible interpretations were two options, one for each language. Use of indeterminate
words is typical of very young bilingual speakers, and it was identified by several authors
(De Houwer, 2009a, p. 41; Deuchar & Quay, 2000, p. 54; Lanza, 2004, p. 128). Table 6.6
lists indeterminate words used during the one-word stage.

Table 6.6 Indeterminate words

Meaning
Age Word variations
Slovak English
1;4.9 [ɲaː] [ɲaːɲaː] mňáu meow
1;4.21 [mumu] [muːmuː] mú moo
1;5.26 [ʃoː] šortky shorts
1;5.27 [penɪ] peračník pencil case
kačka duck
1;6.3 [gaga] / /
čajka (overextension) seagull (overext.)
me (1sg object
1;6.4 [mɪ] my (1pl ‘we’)
pronoun)
1;6.4 [wɔ] [vɔ] voda water
1;6.16 [ho] horúco hot
1;6.16 [gɪ] gaťky ‘undies’ give (me)

The two separate measures for each language account for the number of words known in
each language. They were determined by summing the number of singlets and doublets
with shared concepts in each language separately. Of the singlets 130 belonged to the
Slovak language and 34 to the English language. Thus the Slovak vocabulary consisted of
155 words (130+21+4) and the English vocabulary consisted of 57 words (34+21+2). This
difference in the number of words known by Ria in each of her languages is consistent
with other studies, which found that bilingual children’s lexical learning does not occur in
parallel fashion, but vocabulary growth in one language is followed by a growth in the
other language, rather than growth occurring simultaneously (De Houwer, 2009a, p. 227;
Pearson, 1998, p. 360).

The most important factor influencing the learning rate for each language was Ria’s
emerging language preference. Due to the division of carer duties in the family, Ria spent
151
most of her waking hours with the mother as the primary carer, who addressed her in
Slovak. Thus they engaged in more varied contexts, which provided greater variation in
the input and more joint attentional frames. This is consistent with previous studies finding
that even very young bilingual speakers start showing preference to the most frequently
used language, which becomes the language of the primary sociolinguistic group
(Holowka, et al., 2002, p. 243). The primary sociolinguistic group was defined as “the
person or group with which the child had both the strongest bond and the most constant
contact” (p. 243). It is emphasized that this language can change overtime with changing
circumstances. Thus during the one-word stage the sociolinguistic environment had the
greatest influence on Ria’s rates of word acquisition.

6.2.1 Using TEs


The theoretical grounds of the Principle of Contrast, claiming that young children accept
only one label for a category at the earliest stages of lexical acquisition, have been
criticized in monolingual development (Clark, 1988, p. 327). Likewise, the Contrast was
not supported in the literature on bilingual development. Deuchar & Quay’s (2000, p. 47)
review of BFLA studies found no evidence for the Principle of Contrast, and on the
contrary found that bilingual children use cross-linguistic synonyms from early stages of
lexical development. This was also supported in their own study, and evidence was
presented that the child used TEs from the onset of speech at 0;10. The authors suggested a
close link between TEs and the child’s input, and concluded that the use of TEs did not
indicate appropriate language choice. Indeed, it is not yet clear from previous research
whether presence of TEs in a child’s repertoire is linked with appropriate language choice,
and it is suggested that different BFLA children follow different paths (De Houwer, 2009a,
p. 240).

The data in this thesis also suggest language choice is a pragmatic ability a bilingual child
needs to acquire so that she is able to judge which language is appropriate in which
language context. TEs were observed as soon as Ria was using the first 10 words (aged
1;3). However, TE acquisition was not always occurring simultaneously. Since first words
were learnt through frames of joint attention with the parents, the types of words acquired
depended on the contexts in daily interactions. Each parent and child dyad engaged in
different talk and thus different concepts were learnt through each language at a different
time. TEs were acquired in the expressive vocabulary through the same mechanisms as in
comprehension vocabulary:

152
 In monolingual contexts of each language
 By association in bilingual contexts
 Simultaneously in bilingual contexts.

To determine how Ria came to use TEs appropriately the process of acquisition and use of
cross-language synonyms for one concept will be demonstrated through microanalysis of
several examples.

In the initial one-word stage TEs were not always used in appropriate language contexts.
For example the first doublet members [aɲɔ] ‘ano’ “yes” and [je] ‘yeah’ (1;3) were used
interchangeably with the father. On some occasions Ria used the Slovak TE in response to
his address in English:

*DAD So you are ok sweetie apart from being tired?’

*RIA [ano] ‘ano’ “yes”.

(1;4.21)

However, in different instances the father reported that Ria was responding appropriately
using the English TE [jeː]. Similarly, the negation [ɲɛɲɛɲɛ] ‘nie nie nie’ “no no no” was
used in Slovak and bilingual contexts alike.

Most often a word was acquired in interactions with only one parent. The source for each
word in either language was the social context or set of contexts most often engaged in
with a particular parent who introduced it, or through a situation that made an impression
on Ria. Thus the production of each member of the TE pair took place at a different time.
An example demonstrating this was the word [pɪ] ‘pretty’ acquired as the first equivalent
from the father who often complemented Ria after the mother dressed her in the morning.
First she reproduced the word, and later used it spontaneously in both language contexts.
Even when the father used a synonym, such as ‘beautiful’, or when the mother used the
Slovak TE ‘krásne’, Ria’s preference was for the English equivalent.

Similarly to the development of comprehension vocabulary (section 5.3.4.1), emergence of


new words in Ria’s repertoire created new opportunities for introduction of TEs. The
family often engaged in a discourse in which they discussed new developments and the
parents reported new words to each other in Ria’s presence, including her in the
conversations. Thus they had an opportunity to introduce the missing words by relating the
newly acquired word in one language to the equivalent in the other language, and
emphasize the link between the two forms for one concept. Such discourse allowed the

153
word to enter Ria’s vocabulary in both languages within a relatively short period, as
demonstrated in the following example:

The word ‘tea’ was first acquired in the English context, since Ria often got up with the
father in the morning and they prepared breakfast tea together. The father was providing
commentary emphasizing the word with many repetitions. At first Ria reproduced [tiː]
(1;4.23), and later used it in spontaneous interactions (1;4.30). Thus the concept was
acquired in English, and Ria needed to acquire only the Slovak translation equivalent
‘čaj’. The mother introduced it as soon as she noticed Ria using the English equivalent.
At first Ria reproduced [caː] (1;5.1).

Thus there was a short lag of only several days behind English from when Ria acquired the
TE in her receptive vocabulary. However, in productive usage Ria initially developed a
preference for the English equivalent, even in a Slovak context, and sometimes extended
the meaning to water in a tea cup. It appeared that having both TEs in her receptive
vocabulary was not sufficient to start using them productively. Rather, the equivalent
acquired through the situational context as first was preferred in production.

However, certain awareness of the need for appropriate use of TEs was evident. Ria was
able to repair her language choice when the mother responded with the repetition strategy,
and used the word in Slovak. Likewise, she sometimes used both TEs simultaneously,
which suggested that Ria was aware of a bilingual context and the need to use two
different words for the same referent, as shown in the following video-recording excerpt:

%sit The family is having breakfast, RIA is excited about her cup of tea

*MAM Pijeme čaj všetci. Čo máme? Čo pijeme? Čo pijeme?

%eng All of us are drinking tea. What do we have? What are we drinking? What
are we drinking?

%add RIA

%act RIA has a sip of her tea

*MAM She’s, she got really good with that little one.

%add DAD

*DAD Yeah. Nice little handle.

*MAM With a little handle.

*DAD She likes the tea.


154
*MAM Čo to pijeme?

%eng What are we drinking?

%add RIA

*RIA [kɛɲo]

%glo yyy

*MAM Čaj. Čo to je?

%eng Tea. What is it?

*RIA [tʃa ti]

%glo čaj tea

%eng tea tea

*MAM Čaj, ano.

%eng Tea, yes.

(1;5.5)

In bilingual contexts TEs were acquired simultaneously as two words for one concept.
When the whole family was interacting together, they engaged in a shared discourse,
directing each other’s attention to an object. In these attentional frames both parents
modelled a word in their respective languages, as in the following video-recording excerpt:

%sit The family is in RIAs room, RIA is building blocks, DAD spots a pigeon
on the roof next door

*DAD Pigeon’s on the roof.

%act RIA looks towards the roof

*DAD You see him? She smiled when she saw him.

*MAM Holúbky tam sú, že?

%eng There are pigeons, aren’t there?

*RIA [holuː]

%glo holúbky

%eng pigeons

*MAM Holúbky, ano.

155
%eng Pigeons, yes.

*DAD She looked up there and then she smiled. Can she see ‘em from there?

*MAM Yeah. She wouldn’t smile, otherwise.

%add DAD

*MAM Holúbky tam sú.

%eng There are pigeons.

%add RIA

*RIA [holu]

%glo holúbky

%eng pigeons

*MAM Holúbky, ano.

%eng Pigeons, yes.

%add RIA

%act DAD makes a sound

*MAM What?

%add DAD

*DAD Ah, it’s gone.

*RIA [goː]

%glo gone

*MAM Ano, už sú preč.

%eng Yes, they’re gone.

*RIA [hi]

%glo here

*MAM Preč sú, ano.

%eng They’re gone, yes.

*RIA [pɛ]

%glo preč

156
%eng gone (lit. away)

*MAM Mhm. Už sú preč, holúbky.

%eng Mhm. They’re gone, the pigeons.

(1;5.10)

In this example the TEs ‘pigeons’ and ‘holúbky’ were used directly in reference to one
concept, and Ria showed comprehension of both, although she was not able to use them
productively yet. She chose to reproduce the equivalent that was more salient in the
situation, since it was modelled several times by the mother. In this excerpt Ria also
observed the use of another TE pair, ‘gone’ and ‘preč’. The English equivalent was already
in Ria’s expressive vocabulary and it was used consistently in both language contexts. In
Slovak, however, the concept is expressed in different ways, depending on the action or
state it refers to. This could explain why Ria preferred the English equivalent which was
easy to produce, applied to varied situations and consequently more salient in the input.

However, as demonstrated in the examples of simultaneous interaction in the two


languages within the family, it is unlikely Ria failed to accept two forms for one concept,
especially since the parents addressed her in their respective languages consistently from
birth. Indeed in Table 5.1, Table 6.3 and Table 6.4 it was shown that Ria used the first TE
pair at 1;3.20, and the number of TE pairs expanded rapidly from the beginning of the one-
word stage at 1;4, reaching 25 pairs by 1;6.22. Thus the Principle of Contrast was not
supported in the data. This observation reflects claims by De Houwer et al. (2006, p. 344)
who suggest that due to regular variation in input young bilingual children can develop
flexible strategies to word learning and accept cross-linguistic synonyms.

Ria’s use of newly acquired TEs reflected emerging language choice ability from
approximately 1;6. This can be demonstrated on the example of TE pair ‘keys’ and
‘kľúče’. The equivalents were used by both parents regularly in similar contextual
situations, thus Ria was able to learn them at approximately the same time without
simultaneous exposure. The first equivalent in Ria’s vocabulary was the English word [kɪː]
‘key’(1;6.2) and reproduced after the father when the family was preparing for a bike ride.
Spontaneous use of the TEs was observed few days later on two separate occasions
occurring on two consecutive days:

Ria was watching the mother lock the front door and unlock the garage and
spontaneously commented [ku] ‘kľúče’ “keys” (1;6.7). The next day the father was

157
unlocking the front door and as Ria was saying good-bye from the mother’s arms, she
turned towards the father and commented [kiː] ‘keys’ (1;6.8).

It is possible that in the Slovak context Ria mispronounced the English word ‘key’, while
the mother had interpreted it as Slovak equivalent. However, prior to that occasion Ria did
not produce any other English variants, which was the case with some other words. Thus it
was rather convincingly an example of both TEs used in addressing the parents with their
respective languages. The TEs were used appropriately in the two situations - the Slovak
equivalent in a Slovak context, and the English TE when addressing the father in a
bilingual context. Thus Ria chose contextually appropriate TEs to label the same referent.

It was not always the case that Ria had two direct TEs in her repertoire. Instead, she was
able to use culturally appropriate variation for each language. Thus she expressed the same
concept with two different linguistic means, pragmatically appropriate for each language,
as demonstrated with the following example from bilingual context, where the addressee
was not easily determined:

The family was at the breakfast table, and as they finished eating Ria pointed at empty
bowls uttering [ɲεɲɪ] ‘neni’ “isn’t” (1;6).

As noted in the diary, previously Ria preferred the English equivalent [go] ‘gone’ to
indicate that food or drink were finished (1;5.23). It is difficult to determine why in this
instance Ria chose the Slovak equivalent. What is significant, however, is the use of
appropriate linguistic means for the concept: in English the past participle ‘gone’, while in
Slovak a VP ‘nie je.’, meaning “(there) isn’t (anymore)”. Such distinction was learnt from
the input, and it can be interpreted as example of emerging language differentiation.

Some words in Ria’s early productions assumed a form which made it difficult to
determine the target language, as was the case with homophones in Ria’s productions of
words originating from different languages. While in some cases the target language could
be determined from the context, occasionally even the context did not provide a solution.
This was the case with close cognates with a similar form in both languages, or with words
that had the same or similar initial syllable, and therefore the same form in Ria’s
production, such as the TE pair [wo] for ‘voda’ or ‘water’ (1;5.22). This word was used
with both parents, its form suggested it could belong to either language, and contexts of
use did not clarify the target word, since the referent was the same. However, the series of
learning contexts in which Ria became familiar with the concept suggest the target
language was in fact the Slovak equivalent ‘voda’. When Ria was requesting a drink, the

158
mother focused on the requested referent, making it explicit with questions such as ‘Prosíš
si vodu?’ “Would you like water?” or ‘Čo si prosíš?’ “What would you like?”, while the
father focused on teaching the social word Ria should be using when making a request, as
shown in the following excerpt:

%sit The family is at the dinner table

*MAM No. Už je prázdny pohárik.

%eng Yeah. The cup is already empty.

%gpx RIA is holding up her cup to DAD

*RIA [go ə]

%glo gone yyy

%com RIA uses protoword [ə] to request more water

*DAD Where is Riana’s gone?

%act DAD understood RIA is requesting more water

*DAD What do you say?

*RIA [ə]

*DAD What do you say to daddy? What’s the word?

%gpx RIA is pointing at her empty cup

*MAM Čo si prosíš?

%eng What would you like?

*RIA [pɪ]

%glo please

%add DAD

*DAD Please! Please, isn’t it?

*RIA [ə]

%gpx RIA turns her body and eye gaze towards the kitchen

*DAD Water please!

*RIA [wo]

%glo water
159
*DAD Water please!

*RIA [pɪ]

%glo please

*DAD Yeah!

*MAM Výborne. Ty si anjelik. Rianka je anjel.

%eng Excellent. You’re an angel. Riana’s an angel.

(1;5.23).

In this situation the father modelled both words ‘please’ and ‘water’ which Ria
reproduced. However, in several repetitions he emphasized the social word associated with
the request over the referent, which resulted in Ria receiving positive feedback when she
used the social word herself.

In summary, at the end of the one-word stage Ria’s total production vocabulary (245
words) was greater than the conceptual vocabulary (220 words), thus she knew more
words than concepts. The total number of Slovak words (155) was greater than the total
number of English words (57), showing an emerging preference for the Slovak language.
However, Ria used words from both languages as well as translation equivalents as soon as
she started producing adult words. The shared component across the two languages was 25
TEs. The presence of TEs was not directly linked to appropriate language choice yet. In
productions Ria showed preference for the equivalent which was most salient in input even
though she had both TEs in perceptive vocabulary. However, the skill of pragmatic
language choice was developing gradually, which mirrors findings from previous research
that there is a considerable age variation in the emergence of pragmatic differentiation in
bilingual children (Nicoladis & Genesee, 1996, p. 460).

160
7 Using structures (1;7-2;0)
The progression from holophrastic utterances to expression of meanings with linguistic
structures was gradual. Again there was an overlap of stages. The onset of combinatorial
speech occurred at 1;7, however, holophrases still prevailed. After 1;7 two/multiple-word
combinations became more frequent and remained dominant until the end of the study at
2;0. Thus during this stage Ria was communicating with rich repertoire of utterances:

 Holophrases
 Two-word combinations
 Multiple-word combinations.

Previous research found that for combinatorial speech to emerge as a milestone in


linguistic development children need to learn a certain number of words (Tomasello, 2003,
p. 51). This finding was reflected in this thesis, since Ria’s development progressed into
combinatorial speech when she reached a threshold of 200 words in combined expressive
vocabulary.

According to Tomasello (2003, p. 92), learning of words is also interrelated with learning
of grammatical constructions. There is a correlation between the number of words children
understand and their understanding of syntactic constructions, while in turn, understanding
more constructions facilitates learning of more words. This link was also reflected in Ria’s
productions. When she started producing word combinations, morphological markers
emerged in all types of utterances (holophrases and word-combinations) and across both
languages. The first grammatical markers appeared around 1;7, and by 2;0 Ria produced
words with a range of markers. At the end of the study her utterances reflected incomplete,
yet correct grammar of both languages. Thus overall, this stage was typical of acquisition
of linguistic structures.

A BFLA child growing up with two morphologically different languages needs to learn
different structures for each language and how to distinguish between correct uses in each
language. Differences between Ria’s languages were significant. While English relies on
word order and function words to express grammatical relationships, Slovak is an
inflective language with morphological changes in the form of suffixes as well as word
internal changes in word stems. Grammar is expressed with complex inflection and
declination systems. Verb suffixes mark for person, number, tense, and gender, and
suffixes marking case, gender and number are attached to nouns, adjectives, pronouns and

161
some numerals. These morphological differences will be further highlighted in the
following sections which discuss the developments in word production, combinatorial
speech and lexico-grammar in detail.

Towards the end of 1;8 changes were noted also in Ria’s ability to hold a short dialogue
with the parents. This was mainly in the form of one-word or two-word utterances used in
response to the parents’ questions, with the dialogue extending to 3-4 turns. Previously,
Ria did not take more than one turn in dialogic exchanges, even though the parents
responded to all of her attempts to communicate. She did not answer further questions,
thus the parents provided answers, or simply repeated Ria’s utterance to acknowledge her
attempt to communicate with positive feedback.

As a new development, Ria not only initiated an exchange to engage the parents in
interaction, but when they responded by asking a further question, Ria took another turn,
providing new information, as shown in the following excerpts from Slovak and English
contexts:

*RIA [zaɲɪm. zaɲɪm.]

%glo za ním

%eng after him

*MAM Za kým?

%eng After who?

*RIA [ɔmɪm]

%glo Romim

%eng after Romi

*RIA [pɔʃac]

%glo poštár

%eng postman

*MAM Ano, poštár.

%eng Yes, postman.

*RIA [dɔɲɛsɔʊ]

%glo doniesol
162
%eng brought

*MAM Čo doniesol?

%eng What did he bring?

*RIA [pɔʃʊ]

%glo poštu

%eng mail

*RIA [gon. gon.]

%glo gone. gone.

*DAD What’s gone?

*RIA [dɪs. biːə.]

%glo this. beer.

Thus Ria’s communicative competence was developing further and she was now able to
engage in a short dialogue.

7.1 Changing words


During this stage Ria’s word productions continued to evolve in:

 the structure of the produced word forms,


 the growth of expressive vocabulary.

During 1;7 changes occurred in Ria’s productions mainly in the length of words, which
were previously truncated to a monosyllabic form. Monosyllabic words from both
languages were produced complete. If the target adult word was polysyllabic, Ria
produced longer forms consisting either of 2 syllables (CVCV) or a single closed syllable
(CVC). Several familiar polysyllabic words were produced much like the adult forms. By
the beginning of 1;8 Ria’s productions of polysyllabic words consisted of 2-3 syllables,
thus reflected the structure of the adult words. By 1;9 most of Ria’s productions were
attempts at complete adult words.

Despite the lengthening of words some regular patterns emerged in the form of
simplifications of consonant clusters and substitutions of difficult sounds. For example all
consonant clusters with the sound [s] at the onset of the cluster were reduced to [ʃ], as in
[ɔʃɪtac] ‘ostrihať’ “to cut hair” (1;9.8). Lateral sounds [l] and [r] were substituted with a
163
[j] in a vowel environment in the initial and mid-word positions, as in [jɪʃʊ] ‘ryžu’ “rice”
(1;8.9), [kʊja] ‘kura’ “hen” (1;7.28) and [ɔkɔjɔ] ‘okolo’ “around” (1;9.8). The Slovak
postalveolar retroflex [r] is the last sound Slovak speaking children acquire and is not
mastered typically until age 6 (Štefánik, 2000, p. 82). Ria still produced English words
with primary stress on the second syllable with the first unstressed syllable omitted, such
as [cɪfəkec] ‘certificate’ (1;11.29).

Thus over a period of 5 months, words which were in the one-word stage (from 1;4)
produced in the form of an initial sound or monosyllable, were more complete. From
approximately 1;9 Ria’s productions became longer, clearer, reflected the number of
syllables in the target word and thus reflected adult forms. In addition, intonation patterns
of the productions assumed the rhythm and stress of the individual languages.

7.2 New words


Ria’s production vocabulary continued to grow rapidly both in terms of vocabulary size
and types of different parts of speech in both languages, such as:

 Prepositions - First instances of prepositions (around 1;8) were not produced as


separate words, but pronounced as one unit with the noun under single intonation
contour (See section 7.3.2), e.g. [pɔdsɔjɔm] ‘pod stolom’ “under the table” (1;9.26).
Prepositions appeared as individual words in both languages from the end of 1;9:
[andə] ‘under’ (1;9.26)
[ɪn] ‘in/inside’ (1;10.6)
[dɔ sanʊ] ‘do stanu’ “into tent” (1;11.18)
[ʊ mamɪ] ‘u mamy’ “by mum” (1;11.18)
[ʒɔ ʃɪɲɛ] ‘zo skrine’ “from the cupboard” (1;11.18)
 Conjunction – a TE pair, English [en] or [e] ‘and’ and Slovak [a] ‘a’, was used to
name multiple coordinated subjects or objects, which were regularly referred to
collectively by the parents:
[baba, ɟɛtɔ a jɔbɪ] ‘babka, dedko a Robi’ “grandma, grandpa and Robi” (1;10.10)
[bɪgan e jɪtujan] ‘big one and little one’ (1;10.6)
[maʃums en paːsa] ‘mushrooms and pasta’ (1;10.6).
 Numerals in both languages were used in playful attempts to count:
[wan, tu], [eben sɪks] ‘seven six’ (1;10.15)
[jɛdɛn, da, tɪ, ʃɪjɪ, pec] ‘jeden, dva, tri, štyri, päť’ “one, two, three, four, five” (1;11.1).
Ria was also attempting to use numbers randomly, as she perceived their use by people

164
in the environment, especially in English:
[nambə sɪks] ‘number six’, [sɪksɪ naɪn] ‘sixty nine’ (1;11.1).
 Subject pronouns emerged around 1;10. The most salient was the Slovak 1sg [ja] ‘ja’
“I”, which Ria used to refer to self in multiple word utterances in various contexts:
[tɔ ja maːm! dʒʊʃɪt sɛmatʃɪk] ‘to ja mám! džúsik zemiačik’ “that I have! juice potato”
(1;10.16)
[dɔnʊːta, ja daja] ‘dovnútra, ja dala’ “inside, I put” (1;11.3)
[sɛjɛɟ ja pɔsɪm] ‘zeler ja prosím’ “celery I would like” (1;11.9)
[ʊmɪjɛm ja taɪtʃʊ] ‘umyjem ja tváričku’ “will wash I face” (1;11.10).
Only one other isolated Slovak pronoun appeared when referring to an inanimate object
with masculine grammatical gender:

[ɔn vɪsava] ‘on vysáva’ “ he (=vacuum cleaner) does the vacuuming” (1;11.6).

Amongst English pronouns the 2sg personal pronoun appeared:

[hajəʊ dedi. hajəʊ dedi. aː juː hoʊm?] ‘hallo daddy. hallo daddy, are you home?’
(1;11.18)
[tenk juː] ‘thank you’ (1;11.28).

However, it occurred only in formulaic expressions which Ria learnt as a whole from the
father.

7.2.1 Overextension
Overextension of meanings is a phenomenon typical of child speech during stages of
lexical development (De Houwer, 2009a, p. 216). In Ria’s productions only several cases
of regular overextension were found.

The Slovak verb [paːɪ] ‘páli’ “it is hot” was initially associated with hot food, drinks and
kitchen appliances. Ria overextended the meaning to denote any temperature different to
room temperature, such as extreme weather temperatures, overheated car in the summer,
hot sun ray coming into a room through the window, hot, warm and cold water, cold
freezer and fridge items. When Ria labelled cold items as hot, the mother responded by
recasting her utterance, ‘To je studené.’ “It’s cold”, which Ria reproduced as [sʊː]
‘studené’ “cold” (1;7.7), thus voluntarily repairing the error.

Ria also overextended the Slovak verb ‘večerať’ to eat dinner’ as a label to all mealtimes,
as in a 2-word combination [dæduʃ vɛtʃɛja] ‘daddush večeria’ “daddy is eating dinner”
(1;10.19) during breakfast.

165
An overextension was introduced also by the parents since they regularly used the brand
name to label the vacuum cleaner, thus Ria overextended the word [daɪson] ‘Dyson’,
meaning “vacuum cleaner” (1;9.12), to all objects of the type.

Overextension also occurred in one language only or in production only. Ria understood
the different meanings of Slovak words ‘mucha/ muška’ “fly/little fly” and ‘včielka’ “little
bee” in her comprehension vocabulary and responded to them appropriately, however, in
production she used only [mʊʃa] ‘muška’ (1;11.29) to refer to both objects, most likely due
to easier pronunciation. Conversely, the English TE [faːɪ] (1;7.1) was observed only as a
reproduced word, while the word [biː] ‘bee’ (1;11.29) was used correctly.

7.3 Combining words


Tomasello (2003, p. 114) defines word combinations as consisting of two words or two
holophrases, which are relevant to a specific situation, but do not yet show any
grammatical categories. First word combinations string words into longer utterances which
express varied communicative intentions. The order in which words are combined can
reflect the ordering patterns of adult speech, however, they are not organised according to
specific syntax just yet (Tomasello, 2003, p. 114).

In Ria’s productions spontaneous two-word and multiple-word combinations appeared as


early as the one-word stage (1;4), however, until 1;5 there were only isolated examples,
averaging 1-2 utterances a week. These consisted of words frequently occurring in Ria’s
expressive vocabulary, and combined nouns and verbs, social word, negation, adjective
and a demonstrative. They did not show regular syntax, but consisted of juxtaposed words
relevant to a particular situation or of coordinated multiple subjects, as shown in the
following example:

Ria and the mother were feeding ducks as the mother directed Ria’s attention to a
mother duck with ducklings. Ria responded by pointing and uttering [mama dada gaga]
‘mami, daddy, kačiatka’ “mummy, daddy, ducklings”. (1;4.8).

The utterance was produced in one sequence of words, thus a combination. Early
combinations provided Ria’s expression of her experiences of the context. These
utterances are reported in APPENDIX F.

Word combinations became a typical form of expression at approximately 1;6 when Ria
used them on multiple occasions daily to express varied intentions. From a structural point
of view, Ria was combining words in two ways, which was apparent from the intonations
used with each type of combination:
166
 Fluid combinations - utterances produced without prolonged pauses between words,
with a sentence-like intonation showing they belonged together:
[tɪ dædɪ] ‘teeth daddy (is brushing)’ (1;5;15)
[jana ʃiː] ‘Riana spí’ ”Riana is sleeping” (1;6.22)
 Telegraphic combinations - haltingly produced utterances with words that belonged
together and expressed a meaningful whole, but were uttered in blocks, as if two or
more holophrases with similar intonation and a pause between:
[bɪ. ʃoː] ‘big shorts’ (1;7.10)
[ʃaː. wom.] ‘vtáčik. vonku.’ “birdie. outside.” (1;7.21).

These two forms occurred in Ria’s productions concurrently for approximately 3 months
(1;4-1;6). From around 1;6 pauses between words were shorter, and around 1;10
telegraphic combinations were gradually replaced by fluid combinations consisting of two
or more words. Towards 1;11 they were almost phased out, while only very short pauses
appeared between words in some combinations, however without the falling intonation
contour.

Similar structural differences were found in multiple-word combinations which appeared


more frequently from 1;9 onwards. Three-word combinations consisted of:

 Three one-word blocks such as


[papʊtʃɪ. jana. ɔbʊc.] ‘papučky. Riana. obuť.’ “slippers. Riana. put on.” (1;9.26)
[dan. dædɪ. jana.] ‘done. daddy. Riana.’ (1;8.9)
 Two blocks: a fluid two-word block and one-word block, such as
[atɔ, gɪn. naɟɛmɛ.]9 ‘auto, green. nájdeme.’ “car, green. we will find.” (1;9.26)
[aːɪs, jukɪn. jana.] ‘eyes, looking. (at) Riana.’ (1;9.12).
 Fluid multiple-word utterance
[tɔtɔ, ʃʊpa, dɔkɔʃa] ‘toto, šupa, do koša’ “this, peel, in the bin” (1;9.13)
[buk, dædɪ, nadujan] ‘book, daddy, another one’ (1;11.5).

Thus the more complex meaning was expressed through three, two or 1 part of a whole
utterance unit.

The diary states that at the beginning of 1;11 Ria used multiple-word combinations
predominantly in Slovak, while she mixed some English words for which she had
preference into otherwise Slovak combinations (see 8.1.2). Overall, combined English

9
The comma denotes that although the words formed one unit, they were separated by a very short pause, as
described in (Vihman, 1985, p. 302)
167
utterances were shorter and used less often, and Ria relied mainly on holophrases to create
meanings. During 1;11 three/four-word utterances were still used predominantly in Slovak,
but they appeared in English as well.

7.3.1 Word order


According to Tomasello (2003, p. 112) majority of utterances children hear in input initiate
with item-based frames consisting of two words or morphemes, such as It’s…, Can you…,
Here’s…, Look at…,etc., which are used frequently. The nature of item-based frames is
reflected in children’s productions as well, since they learn and use them as whole units in
their speech. Tomasello (2003, p. 115) calls words organized into consistent ordering
patterns pivot schemas - constructions without syntax, which are reproduced from adult
speech in the input. One word in the utterance determines its function, and remains
constant, while the other word is variable. Tomasello (2012, p. 77) also argues that when
morphology emerges in productions, children do not use morphological markers
productively yet, but they organize words into item-based constructions which are based
around specific verbs in forms children hear in the input. Thus each verb is learnt in a one-
by-one basis.

Ria’s two-word combinations had a variable word order. Changes in word order did not
reflect changes in meaning. Different meanings were expressed by the types of words that
were combined. With each utterance Ria focused on different aspects of the situation.

Function English

Transitivity [dædɪ. loːn] ‘daddy (is watering the) lawn’ (1;7.10)

Agency [dædɪ, kiːnɪn] ‘daddy (is) cleaning’ (1;9.12)

Possession [teɪn. jaɲa.] ‘train Riana(‘s)’ (1;5.24)

Slovak

Transitivity [jana! xɛbik.] ‘Riana chlebík’ “Riana (wants) bread” (1;8.21)

Agency [ʃɛ jana] ‘sedí Riana’ “is sitting Riana” (1;6.22)

Possession [mamɪ taɪ] ‘mami čaj’ “mummy(‘s) tea” (1;7.28)

Word combinations consisted of various constituents, and included morphological


markers:

 coordinated multiple subjects or objects:


[mamɪ dædɪ jana] ‘mummy, daddy, Riana’ (1;4.19)
168
 subjects + objects/indirect objects:
[zabɪʃ dædɪ] ‘rubbish daddy (is emptying)’ (1;11.4)
[jana. babus!] ‘Riana. (will do) bubbles.’ (1;8.21)
[jaɲɛ mamɪ] ‘Riane mami’ “(to) Riana+DAT mummy (will give)” (1;8.24)
 subjects + adjectives:
[dædɪ, bɪk] ‘daddy big’ (1;8.31)
 subjects + adverbs/ determiners/ prepositional phrases:
[tɔtɔ, bʊʃɔ] ‘toto bruško’ “this tummy” (1;10.6)
 objects + indirect objects:
[pʊsʊ. dædɪ.] ‘pusu. daddy(mu)’ “(I’ll give a) kiss (to) daddy” (1;8.11)
 utterances containing direct address of the interlocutor and thus not true word
combinations:
[mamɪː, jɪxɔː!] ‘mami, rýchlo!’ “mummy, quick!” (1;10.18).
However, some regular ordering patterns were evident, which could be categorised as
pivot schemas. Schemas were used across both languages and expressed several meanings.
Other schemas were typical of one language only and reflected fixed constructions heard
in input:

English

[moː, pɪtsa?] ‘more, pizza’ (1;11.7)

[epu moː] ‘apple more’ (1;11.18)

[tubaʃ, gon] ‘toothbrush gone’ (1;8.31)

[miːt, gon] ‘meat gone’ (1;9.19)

[dan dædɪ] ‘done daddy’ (1;10.6)

[dan. dædɪ. jana] ‘done daddy Riana’ (1;8.9)

Slovak

[ɛʃɛ papɪka] ‘ešte paprika’ “more capsicum” (1;11.5)

[ɛʃɛ maːtoʊ] ‘ešte (to)mato’ “more tomato” (1;11.3)

[ja sama] ‘ja sama’ “I myself” (1;10.7)

[jana sama] ‘Riana sama’ “Riana self” (1;8.9)

[za ɲɪm. baːbom.] ‘za ním. bábom.’ “after him. baby.” (Dative) (1;8.8)

169
[zaɲɪm. dædɪm.] ‘za ním. daddym.’ “after him. daddy.” (Dative) (1;8.9)

Word order in Ria’s combinations remained variable also when verb morphology appeared
around age 1;7. Morphological markers marked the syntactical relationships between
words, but they were still organized around item-based constructions. In Slovak utterances
more morphological markers were observed, which resulted in more varied utterances,
while in English meanings continued to be expressed mainly through juxtaposition of
words as before. Two-word utterances consisted of various constituents, such as:

 subject+verb:
[ɪɈɛ kɔkɪdɔl] ‘ide krokodíl’ “crocodile is going” (1;10.7)
 object+verb:
[bukɪnʊ ɲɛʃɛm] ‘book-inu10 nesiem’ “book-inu I am carrying” (1;11.3)
 subject + infinitive:
[dædɪ, hac] ‘daddy, hrať’ “daddy, (I want to) play” (1;8.21)
 verbs + prepositional phrases, preposition omitted:
[sɪt. tʃeːə.] ‘sit. (on the) chair’ (1;7.21)
 verbs+ adverbials:
[dɔma maːm] ‘doma mám’ “at home I have”(1;10.6)
 verb constructions (V+INF):
[ɪɟɛm, pɛsɛc] ‘idem (sa) prezliecť’ “I’m going to get changed” (1;10.16)
 negated 2-word utterances:
[haːdʒac ɲɪɛ] ‘hádzať nie’ “to throw not” (1;10.21)
 Wh-questions:
[ɟɛjɛ jana?] ‘kde je Riana?’ “where is Riana?” (1;10.7)

Such combinations of constituents appeared across both languages.

With increased utterance length approximations to the typical word order in the languages
appeared. English has a fixed SVO order, while Slovak inflectional grammar allows
variable word order, and the subject does not have to be expressed explicitly, since it is
implicit in the verb inflectional ending. Ria’s multiple-word utterances in both languages
had variable word order of the three constituents:

Slovak

SVO [dædɪ. vajɪː. meso.] ‘daddy. varí. mäso.’ “daddy. is cooking. meat.” (1;8.21)

10
Book-inu is a bilingual blend discussed in section 8.1.2.
170
OSV [nɛxɪ. mamɪː. # os. os.] ‘nechty. mami. ostrihať. ostrihať.’ “fingernails.
mummy. cut. cut.” (1;8.21)

OVS [pɪʃamɔ. # sɛjɛnɛː ɲɛmaːm ja.] ‘pyžamo. # zelené nemám ja.’ “pyjamas. #
green don’t have I.” (1;10.8)

VOS [ɔbabɪc, kɪmʊ, dædɪ] ‘opraviť, klímu, daddy’ “to fix, aircon, daddy”
(1;10.31)

VSO [pɔdaː mamɪ, bukɪna] ‘podá mami, book-ina11’“will pass mummy. book”
(mummy will pass the book) (1;11.3)

SOV [ʃɪmak. vɔdʊ maː] ‘slimák. vodu má’ “snail. water has.” (1;11.6)

English

SVO [aːɪs, jukɪn. jana.] ‘eyes. looking. Riana.’ “eyes are looking at Riana”
(1;9.12)

OVS [miːt iːt, dædɪ] ‘meat eat, daddy’ (1;11.27) “daddy eats meat”

SOV [mamɪ, epu, iːcɪn] ‘mummy, apple, eating’ (1;11.10) “mummy is eating
apple”

However, the typical SVO order became more prominent in both languages during 1;11.
Moreover, some Slovak utterances appeared with implicit subject:

[dɔma mam vatʃɪk] ‘doma mám vláčik’ “at home (I) have train” (1;11.6)

[tʊ mam mɔdʊ sɔɪtʃʊ] ‘tu mám modrú stoličku’ “here (I) have blue chair” (1;11.8).

Thus differentiation of syntactical relationships between the two languages was emerging
in the last month of the second year.

7.3.2 Fixed expressions


In Ria’s productions pivot schemas and item-based frames appeared most typically in the
form of fixed expressions. In the one-word stage fixed expressions were un-analysed,
learnt as a whole, produced as single units and consisted of words which were otherwise
not used separately (see section 6.1.4.3). As Ria entered combinatorial speech they were
still produced as one unit, as suggested by the intonation contours, but they became less
fixed. They consisted of one fixed element and one alternating element which carried the

11
There is a rare error in case ending, as the Accusative should be an –inu suffix.
171
syntactical relationship. The individual components were used in combinations with other
words as well. Thus the structure was of a fixed expression, but not a static unit anymore.

Initially fixed expressions were directly reproduced after the parents, such as [ʃoju] ‘show
you’ and [ʃomɪ] ‘show me’ (1;7.10), [mabaːbɔ] ‘my bábo’ “my baby (doll)” (1;7.7). They
became spontaneous and more frequent around 1;8-1;9. Several types of schemas and
item-based frames with verbs were observed and they occurred in both languages,
although considerably more in English. In English 63 tokens were recorded (44 addressed
to father, 6 to mother, 13 to both), in Slovak 29 tokens (25 addressed to mother, 4 to both),
and 1 mixed variant addressed to the father:

[adujana] ‘other one+a’ (feminine pronominal suffix reflecting grammatical gender of


the referent) [1;11.4].

English types of fixed expressions are shown in Table 7.1 and Table 7.2.

Table 7.1 English fixed pivot schemas

English schema Examples


ADJ+one/ones [jɛdam] ‘red one’ 1;8.31
[ɲuːans] ‘new ones’ 1;9.26
[jɪtuɪan] ‘little one’ 1;9.19
DEMONSTRATIVE+one/ones [dɪsams] ‘these ones’ 1;9.12
NOUN+on [ʃokon] ‘sock on’ 1;9.12
[jaɪtson] ‘lights on’ 1;9.12
NOUN+time [naɪtam] ‘night time’ 1;9.19
[basam] ‘bath time’ 1;9.19
NOUN+up [baɪnap] ‘blind up’ 1;10.14
ALL+past participle [oʊgon] ‘all gone’ 1;9.20
[oːʊdesd] ‘all dressed’ 1;11.3

Table 7.2 English fixed item-based frames

English item-based frame Examples


DEMONSTRATIVE+is [dɪsɪs] ‘this is’ 1;9.12
[detɪs] ‘that is’ 1;10.6
[deə:ʃi:s] ‘there she is!’ 1;7.14
[deːəɪs] ‘there is!’ 1;8.21
INTERROGATIVE+is [weəʃɪs] ‘where she’s’ 1;10.27
VERB+it [koʊʃɪt] ‘close it’ 1;9.12
[iːtɪt] ‘eat it’ 1;9.12
VERB+this [hoʊdɪs] ‘hold this’ 1;10.22
VERB+me [ʃoʊmi:] ‘show me’ 1;8.9
VERB+you [mɪsuː] ‘miss you’ 1;10.19
[baɪtʃuː] ‘bite you’ 1;8.9
VERB+him [kætʃɪm] ‘catch him’ 1;9.19
VERB+some [hepsam] ‘have some’ 1;9.19
172
English item-based frame Examples
[pɪtsam] ‘pick some’ 1;9.26
VERB+up [pekap] ‘pack up’ 1;11.12
[geɪap] ‘get up’ 1;10.26

The construction ‘ADJ one’, ‘DEMON one’ and transitive ‘VERB it’ in English were
particularly productive. The types of schemas and item-based frames used in Slovak are
shown in Table 7.3 and Table 7.4.

Table 7.3 Slovak fixed pivot schemas

Slovak Schema Meaning Examples


pre+NOUN for [pɛjanʊ] ‘pre Rianu’ “for Riana” 1;5.23
after/ [zaɲɪm] ‘za ním’ “towards him” 1;8.21
za+PRONOUN towards [zaɲou] ‘za ňou (loptou)‘ “after it (the ball)”
(direction) 1;9.8
to/
ku+NOUN [kʊmamɛ] ‘ku mame’ “to mummy”1;8.21
towards
[dɔkʊpɛɲɛ] ‘do kúpeľne “to the bathroom”
do+NOUN to/into
1;9.12
[vaʊcɛ] ‘v aute’ “in the car” 1;9.12
v/v+ NOUN in/at
[vɔbɔcɛ] ‘v robote’ “at work” 1;11.18
na+NOUN on [nasɛm] ‘na zem’ “on the floor” 1;9.26
[pɪpɔtatʃɪ] ‘pri počítači’ “by the computer”
pri+NOUN by (place)
1;10.27
[pɔsɔjɔm] ‘pod stolom’ “under the table”
pod+NOUN under
1;9.26

Table 7.4 Slovak fixed item-based frames

Slovak item-
Meaning Examples
based frame
[ɟɛjɛ] ‘kde je’ “where is she”1;4.26
kde+VERB where
[tʃesuː] 1;8.9
[tʊjɛ] ‘tu je’ “here she is”1;6.4
tu+VERB here
[tusuː] ‘tu sú’ “here they are”1;8.9
[ʃɔdɔmaʃ] ‘čo to máš?’ “what have you got?”
čo to+VERB what’s 1;8.21
[tɔtɔjɛ] ‘čo to je?’ “what is that?” 1;10.7

In Slovak fixed expressions accounted for prepositional phrases, some locatives and
predicate constructions. Moreover, the Slovak items did not produce as many tokens as in
English, suggesting that item-based learning was typical mainly in English. A full list is
reported in APPENDIX E.

173
Ria used fixed expressions productively and at times attempted to combine items
ungrammatically, as in the following examples recorded in English contexts:

[bukan] ‘book one’ (1;9.12) while choosing a book to read

[dʒɪpan] ‘jeep one’ (1;9.12) while reaching out to pick up a toy car.

Ria also maintained several social expressions which she learnt as one unit and used as
unanalysed wholes. They were frequently heard in the input and included:

[aɪtɔaɪtɔ] ‘aj to aj to’ “that and that” (1;9.20)

[hoɪdon] ‘hold on’ (1;9.26)

[tɛnkʊː] ‘thank you’ (1;10.11)

[kamam] ‘come on’ (1;10.24)

[ʃuːsːun] ‘see you soon’ (1;10.31).

Another English construction initially learnt as one unit were tag questions. Ria was able
to use tag questions in response to the father’s address in which he used a tag question.
Initially she used it in the form of reproduction:

*RIA [puʃɪː]

%glo prší

%eng It’s raining.

*DAD It’s raining, isn’t it?

*RIA [ɪsnɪt?]

%glo Isn’t it?

Although tag questions were produced as a constant unit, Ria was also able to
appropriately invert the word order:

*DAD Where does it go? In your hair? Does it?

*RIA [ɪdas]

%glo It does.

Tag questions were not yet used spontaneously.

174
7.3.3 Expressing negation
With combinatorial speech negation appeared as well. The two languages express negation
differently: in Slovak to negate a verb ne- is prefixed to a verb, while English uses a
particle not, which can be placed in several different positions depending on the tense and
structure of the verb phrase. According to Meisel (2001, p. 20) cross-linguistic studies
show that children initially negate the whole utterance, and when finite verb forms appear,
negator is placed internally within the utterance.

These two stages of negation appeared also in Ria’s data. At the outset Ria rendered
negation in both languages with one structure. Slovak negator was juxtaposed to different
word types. Initially a non-standard form derived from the negated verb to be ‘neni’ “isn’t”
was used (standard form is ‘nie je’). Ria perceived this form as a single unit, since that is
how it was used by the mother in utterances addressed to her. The entire holophrase was
negated, as in the following examples:

[ɲɛɲɪ ʃaː] ‘Neni vtáčik.’ “isn’t birdie.” (1;6.4)

[ɲɛɲɪ. ɲɛɲɪ. hɪ. hɪ. ɲɛɲɪ] ‘neni. neni. here. here. neni.’ “isn’t. isn’t. here. here. isn’t.”
referring to food not being on plate (1.6.19)

Negation was rendered with three possible forms of the Slovak negator until 1;8:

 negator ‘nie’ “not”


 standard form ‘nie je’ “is not” of the inflected verb nebyť “not to be”
 non-standard form ‘neni’ “isn’t”.

All three structures were used interchangeably in both languages, juxtaposed before or
after the word they were negating, such as adjectives, nouns, verbs:

[punɛː. ɲɪɛ.] ‘plné. nie.’ “full. not.” (1;8.9)

[ɲɪɛjɛ. paːsɪ.] ‘nie je parsley’ “(there) is not parsley” (1;8.9)

[ɲɛɲɪ ɲɪtʃ] ‘neni nič.’ “isn’t anything.” (1;8.21)

[wək. ɲɛ.] ‘work. nie.’ “work. no.” (1;8.31)

[ɲɪjɛ badə] ‘nie butter’ “no butter” (1;8.31)

[xʊcɪː ɲɛ] ‘chutí nie’ “I like not” (1;9.4)

[xɔɟɪː ɲɪɛ] ‘chodí nie’ “goes not” (1;10.4).

175
The negator could be used as a holophrase, as in the following example, which prompted
the mother to request clarification:

%sit Ria and MAM are eating lunch

*RIA [ɲɪɛ]

%glo nie

%eng no

*MAM Čo nie?

%eng What no?

*RIA [xʊcɪː]

%glo chutí

%eng I like

*RIA [janɪma]

%glo Rianina

%eng Riana’s

*MAM Rianina ryža, ano.

%eng Riana’s rice, yes.

*RIA [ɦɪːbɪk. ɲɛɲɪ.]

%glo hríbik. neni

%eng mushroom. isn’t.

(1;8.24).

The first negative verbs reflecting adult forms appeared in Ria’s Slovak utterances
approximately 3 months later, around 1;10. They were formed by the negative prefix
attached to the verb and used productively in inflected verbs:

[ɲɛmɔːʒɛʃ] ‘nemôžeš’ “you can’t” referring to self (1;9.4)

[ɲɛɪɟɛ] ‘nejde’ “isn’t coming” referring to a duck approaching Ria (1;9.12)

[ɲɛmaːmɛ] ‘nemáme’ “we don’t have” (1;9.19)

176
[ɲɛjʊːbɪm] ‘neľúbim’ “I don’t like” (1;9.26).

Unilingual English utterances with negation emerged around the same time as Slovak
negated verbs. However, Ria still used the first stage negation, negating the whole
holophrase with the juxtaposed English negator no. Only few examples were recorded
before 2;0:

[siːds, noʊ, siːds] ‘seeds, no, seeds’ (1;9.19)

[noʊ jaɪts on] ‘no lights on’ (1;10.6)

[noʊ wək] ‘no work’ (1;11.3)

[noʊ boʊ go!] ‘no ball go’ (1;11.7)

Thus it appears that by 1;6 Ria was ready to express negation across both languages. In the
first stage she applied the Slovak non-standard negator across utterances in both languages
in an apparently undifferentiated pattern, where the Slovak negator was used to fill the gap
in English. Language specific negation appeared in Slovak utterances at 1;9. At this stage
English utterances were still negated through juxtaposition of the negator to an utterance,
however, the English negator replaced the temporary Slovak equivalent. A similar
sequence of stages with separate progression to the adult structure was found in Vihman
(1985, p. 312) and Meisel (1989, p. 19), thus it can be explained as a universal acquisition
feature rather than a sign of lack of differentiation.

7.4 Inflectional morphology


In Ria’s productions morphology emerged in both languages simultaneously. However,
different aspects relevant to each language appeared, which provided evidence for separate
development of grammars of the two languages in the one-word and early two-word stages
(Sinka & Schelletter, 1998, p. 324). Morphological markers were used in both holophrastic
and multiple-word utterances. Initially words with morphological markers appeared in
forms which were used frequently by the parents in the input, and no generalizations
across word classes were made, which is a developmental pattern observed by other
authors as well (Tomasello, 2003, p. 118).

First morphological markers emerged around 1;7 in the form of reproduced words across
both languages but especially Slovak, such as Slovak inflected verb forms (e.g. [maʃ]
‘máš’ “you have”, [tsɛʃ] ‘chceš’ “you want”, [vɛʃ] ‘vieš’ “you know” in the 2sg), English

continuous forms (e.g. [kamɪ] ‘coming’); Slovak noun declinations (e.g. [mamɛ] ‘mame’
“to mummy”, [mamɪn] ‘mamin’ “mummy’s”), and English possessive suffix (e.g. [mamɪs]
177
‘mami’s’ “mummy’s”). By the first half of 1;8 majority of the Slovak reproductions were
complete words with appropriate suffixes. English reproductions were marked with
necessary grammatical aspects (number, possessive, 3sg). Thus Ria distinguished between
the different morphology in operation in the two languages in her input, and demonstrated
this with correct reproductions.

In spontaneous productions morpho-syntax in both languages emerged at the beginning of


1;7, as soon as Ria ceased to truncate words, and produced longer and more complete adult
forms, as discussed in the following sections.

7.4.1 Nominal morphology


First grammatical category in nominal morphology was the possessive marker emerging
simultaneously in both languages, but only with few familiar appellatives referring to the
parents, self and favourite toys. Holophrases consisting of a person’s name with the
possessive marker were used while pointing at an object, thus Ria was denoting ownership
of the object rather than labelling it.

Possessives were language specific. In Slovak utterances Genitive suffixes were attached
to possessive adjectives related to familiar people, showing agreement with the referent’s
grammatical gender:

[mamɪm] ‘mamin (zápisník)’ “mummy’s (notepad)” (1;7.8) pointing at the mother’s


note pad, masculine grammatical gender

[mamɪna] ‘mamina (kniha)’ “mummy’s (book)” (1;8.9), feminine grammatical gender.

Neuter forms of the Genitive suffixes were used when referring to various objects that
belonged to people, without mentioning the object label, which could be glossed as ‘to je
X-suffix’ meaning “that’s X’s”:

[dædɪxɔ] ‘daddyho’ “(that’s) daddy’s” (1;9.4)

[mamɪnɛ] ‘mamine’ “(that’s) mummy’s’ (1;9.8)

[jaɲɪnɛ] ‘Rianine’ “(that’s) Riana’s” (1;9.8).

In English utterances the possessive marker ‘s was attached to appellatives:

[dædɪs] ‘daddy’s’ labelling objects belonging to the father (1;7.10),

[mamɪs] ‘mami’s’ “mummy’s” labelling objects belonging to the mother (1;7.10),

178
[babos] ‘bábo’s’ “baby doll’s” (1;7.21) labelling items belonging to a doll12,

[janas] [janɪs] ‘Riana’s/Riani’s’ (1;7.28) referring to objects belonging to self.

The next morphological marker to emerge in both languages with language specific
morphemes was the regular plural form of nouns. At first two masculine plural Slovak
nouns were used:

[vasɪ] ‘vlasy’ “hair (plural)” referring to the mother’s hair and distinguished from the
singular form used on the same day in a different context, when a hair got into Ria’s
mouth and she requested to remove it uttering [vas] ‘vlas’ “a hair” (1;7.17)

[xʊp] and [xʊpɪ] ‘chlp / chlpy’ “body hair(s)” used in both singular and plural forms
appropriately (1;7.28).

With a lag of one month plurals were generalized across nouns belonging to various
nominal classes, including feminine and neuter genders.

Masculine

[banɪnɪ] ‘banány’ “bananas” (1;8.21)

[nɔsɪ] ‘nosy’ “noses” (1;8.21)

[manʊːʃɛ] ‘vankúše’ “cushions” (1;8.24)

Feminine

[kɪʃɪ] ‘knižky’ “books” (1;8.21)

[vɛtsɪ] ‘veci’ “things” (1;8.21)

[jasɪnɪ] ‘rastliny’ “plants” (1;8.24)

Neuter

[ɟɛcɪ] ‘deti’ “children” (1;8.21).

Nominal plurals of all classes were established by the second half of 1;9.

Simultaneously, in English utterances the regular plural suffix –s was observed with nouns
that were most often used in plural in the input, referring to several items of the type:

[hæts] ‘hats’ used by the father when getting ready to go out (1;7.21)

[hæːns] ‘hands’ (1;7.21)

[tʃeːəs] ‘chairs’ (1;8.9)

12
The family adopted the Slovak word ‘bábo’ “baby” when talking about the particular doll.
179
[babus] ‘bubbles’ (1;8.21).

Occasionally the English regular plural was overextended to mass nouns which do not
require plural suffix, such as referring to two bank notes with [manɪs] ‘money-s’ (1;8.21).
Such overgeneralizations are typical in English speaking children (Štefánik, 2000, p. 50).

Thus possessive and plural suffixes appeared simultaneously in both languages with
language specific markers. Meisel (1989, p. 21) argues that if children express similar
functions in the two languages through different grammatical means, this indicates
differentiation of the two grammatical systems. Indeed, further morphological
development occurred separately in each language, as discussed in the following sections.

7.4.1.1 Slovak nominal morphology


Slovak nominal morphology is complex in comparison with English. There are 6 cases
with morphological markers expressing different grammatical relationships, as shown in
Table 7.5.

Table 7.5 Slovak nominal cases

Case Slovak meaning Corresponding English structure


Nominative subject of a verb word order
possessive relationship or other
Genitive possessive ‘s marker
close semantic relation
Dative indirect object prepositions to and for
Accusative object of a verb word order
prepositional case expressing Locative prepositions in, at, on, and
Locative
location and other relationships prepositions about, after
Instrumental Instrumentality Prepositions with and by

Slovak nouns are divided into three classes - masculine, feminine and neuter - each with 4
inflectional subclasses. Masculine subclasses are further divided into animate and
inanimate. Each subclass uses distinct suffixes for grammatical cases in the singular and
plural forms which are determined by a prototype for that subclass. A suffix is attached to
the noun stem and expresses the grammatical category of case, gender and number of the
noun. Moreover, nominal suffixes are polysemic and a single suffix can express several
grammatical meanings, as shown in the following inflected prototype word žena ‘woman’
for a subclass of feminine nouns:

Case Singular Plural

Nominative žen-a žen-y

Genitive žen-y žien


180
Dative žen-e žen-ám

Accusative žen-u žen-y

Locative žen-e žen-ách

Instrumental žen-ou žen-ami.

In this prototype word the suffixes -y and -e have 3 and 2 possible grammatical meanings
respectively. In some feminine and masculine subclasses the Nominative case is the stem
form without a suffix.

The Nominative was the first form of nouns Ria was able to produce in the one-word
stage. The remaining nominal cases appeared in Ria’s productions during the two-word
stage, all within a period of three months (end of 1;7-1;9). Ria initially learnt inflectional
endings with familiar words frequently used by the mother, before she extended them to
other types of words. Thus familiar appellatives were among the first nouns with Slovak
inflectional suffixes when used in a holophrastic sense. They emerged in most grammatical
cases around the same time, as the following examples demonstrate:

 Accusative
[mamʊ] ‘(chytím) mamu’ “(I will hold) mummy” (1;7.8)
[dedʊ] ‘(chytím) dedu13’ (chytím daddyho) “(I will hold) daddy” (1;7.8)
[janʊ] ‘Rianu’ “Riana+ACC” (1;8.21)
 Instrumental
[dædɪm] or [dædɪms] ‘(pojdem s) daddym’ “(I will go) with daddy” (1;7.18)
[babɔm] ‘(s) bábom’ “(with) baby” (1;8.21)
[ɔmɪm] ‘(za) Romim’ “after Romi”
 Dative
[ɔmɪʊ] ‘(dám) Romimu’ “(I will give) to Romi” (1;7.18)
[dædɪmʊ] ‘(ukážem) daddymu’ “(I will show) to daddy” (1;7.18)
 Locative
[paʒɪ] ‘(na) pláži’ “at the beach” (1;7.18).

Approximately a month later inflectional case endings were extended to other noun types.
Among the first was the Accusative case used in various functions, such as when talking
about objects in the environment:

13
English proper nouns and appellatives in the mother’s speech were morphologically integrated in Slovak
utterances, using appropriate suffixes, thus this usage was modelled to Ria in the input. However, in this
example Ria made an occasional error in suffix and used what appears to be the feminine suffix –u instead of
masculine -ho.
181
[kaba] ‘(videli sme) kraba’ “(we saw a) crab” (1;7.28)

[pavʊːka] ‘pavúka’ “spider” (1;8.24).

Accusative was also used to request objects or food:

[jɪʃʊ] ‘ryžu’ “rice” (1;8.9)

[mahʊjʊː] ‘marhuľu’ “apricot” (1;8.21)

Accusative was further used to talk about objects in Ria’s environment which she was able
to manipulate herself:

[pʊsʊ] ‘(umývam ti) pusu’ “(I’m washing) your mouth” (1;7.28)

[metʊ] ‘mätu’ “(I’m picking) mint” (1;8.21),

or objects that were manipulated by others:

[ʃɪmʊ] ‘špinu’ “dirt/mess” referring to the result of a doll’s pretend actions (1;8.21)

[nɛxɪ] ‘nechty’ “finger nails” referring to the object mother was to manipulate (1;8.21).

Ria was aware of the need to use the form appropriate for a particular situation in the
Slovak language, since she started self-repairing her utterances in some cases, for example:

when requesting almonds Ria uttered [mana. manʊ] ‘mandľa. mandľu.’ “almond. an
almond.” (1;8.9),

where the Nominative suffix was self-repaired to the Accusative, since the word appeared
in the grammatical position of an object.

Other nominal cases extended to all noun types by the end of 1;9:

Genitive used to express semantic relations other than possession

[sɪdʊ] ‘syru’ “of cheese” (1;8.21)

Ria was requesting “(a little bit) of cheese” during meal time and correctly applied the
Genitive suffix, which expressed the notion of ‘a part of’. Accusative case could be used if
requesting the object without specifying the amount.

Prepositional cases which may or may not use a preposition in Slovak also appeared
around the same time, however, Ria often omitted prepositions:

Locative in function of a prepositional phrase

[tɪtʊ] ‘(na) tričku’ “(on the) t-shirt” (1;8.21)

[kabɪtsɪ] ‘(v) krabici’ “in a box” (1;9.1)


182
Instrumental used to request manipulation or use of objects

[nɔʃɪkɔm. mamɪnɪ.] ‘nožíkom. mamini(m).’ “with knife. mummy’s.” (1;9.1)

[kɪʃu. sɛbow.] ‘knižku. (so) sebou.’ “book. with you (=me)” (1;9.20)

[mʊmɔʊ] ‘(s) múmou (=kravkou)’ “(with a) moo (=cow)” (1;9.20).

Thus nominal morphology was established by the second half of 1;9. Ria used English and
Slovak noun forms appropriate in specific situations across all classes, reflecting correct
grammars of the two languages. While nominal morphology was developing
simultaneously across languages, it was taking place separately within each language,
showing separate development of the grammars.

7.4.2 Gender and number agreement


In Slovak inflectional morphology modifiers such as adjectives or pronouns, as well as
pronouns substituting a noun, must show gender and number agreement with the nouns
they modify. There are three grammatical genders - masculine, feminine and neuter.

In Ria’s productions correct forms showing gender and number agreement with the
grammatical gender of referents emerged during 1;7 with the possessive Slovak suffixes in
Genitive (section 7.4.1). Same agreement was also applied to holophrastic pronominal
modifiers in the Accusative or Nominative, which Ria used to request a different item of a
kind:

Masculine

[ɪnɪː] ‘iný (keksík)’ “different (biscuit)” (1;7.21)

Feminine

[ɪnʊː] ‘inú (knižku / gumičku) “different (book / head band)” (1;7.21)

[dʊa] ‘druhá (papuča)’ “the other (slipper)” (1;7.28)

Neuter

[dʊɛ] ‘druhé (hami)’ “the other (breast)” (1;7.28).

Similarly, if adjectives were used as holophrases correct gender and number agreement
with the referent was evident, as in the following examples:

Accusative feminine singular

[tʃɪsʊ] ‘čistú (plienočku)’ “clean (nappy)” (1;8.9),

183
Accusative feminine plural

[ʒʊtɛː, mɔnɛː, jʊʃanɛː # ɔbtʃɛtʃɛ] ‘žlté, modré, ružové (plienočky) # (si) oblečie.’
“yellow, blue, pink (nappies) # will put on” (1;9)

Nominative masculine singular

[dʊhɪː] ‘dlhý (nechet)’ “long (fingernail)” (1;8.21)

Nominative masculine plural

[dʊhɛː] ‘dlhé (nechty)’ “long (finger nails)”.

Gender and number agreement was also maintained in two-word combinations:

[jʊkɪ. pʊnɛː] ‘ruky. plné.’ “hands. full.” (1;8.24).

Agreement between a pronoun and its referent appeared at the same time. However, Ria
used the first pronouns only as a variable element combined with fixed preposition ‘za’
“after” (denoting movement towards) within specific pivot schemas. Thus the pronouns
were used in the Instrumental case in a specific gender, and were associated with specific
situations, requesting to go after/towards a person(s):

[zaɲɪm] ‘za ním’ ”to him’

[zaɲɪm. dædɪm.] ‘za ním. daddym.’ i.e. “(Let’s go) to him. (To) daddy.”

[zaɲɪmɪ] ‘za nimi’ “after them”, i.e. “(Let’s follow) after the rubbish removalists.”
(1;8.9)

[zaɲɔʊ] ‘za ňou’ “to her”, i.e. “(Let’s go) to the duck” (1;8.21).

A pronoun in the grammatical gender of the referent was used in a holophrase referring to
a familiar object:

Neuter

[naʃɛ] ‘naše (auto)’ “our (car) (1;8.21).

Gender agreement was also evident in the first numeral Ria used in the Slovak language,
[da] and [dɛ] ‘dva/dve’ “two”, which was produced with the correct masculine or feminine
gender respectively, depending on the referent’s grammatical gender.

Thus the development of appropriate gender agreement took place hand-in-hand with the
development of nominal morphology, suggesting Ria had awareness of the category of
gender from early on. Early uses of appropriate gender in Slovak pronominal morphology
also emerged, however, related only to specific referents at this stage.
184
7.4.2.1 Comparatives
The grammatical category of degree, namely the comparative, appeared in isolated
instances in both languages:

[mekʃɪː] ‘mäkší’ “softer” (1;9.19)

[koʊsǝ] ‘closer’ (1;10.8)

[betǝ] ‘better’ (1;10.15).

The Slovak comparative reflected appropriate gender of the referent. However, in these
instances comparatives were used as context specific items, and were not generalized
across all adjectives yet.

7.4.3 Slovak verb morphology


Slovak finite verb forms are inflected and consist of the verb stem and a suffix. In addition,
changes can occur in the verb stem. There are 7 irregular verbs, and 14 regular verb
prototypes, each using different suffixes marking the following grammatical categories:
person, number, tense, mood and gender (Mistrík, 1988, p. 135). In Ria’s early speech the
indicative mood was most evident, with only several imperative forms and no conditional
forms. In the indicative the grammatical category of tense is expressed with different
suffixes for present, past and future, while gender (masculine, feminine or neuter) is
marked in the past tense only. The inflectional paradigm can be illustrated on the verb
prototype chytať ‘to touch’ inflected in the present, past and future tenses of the indicative:

Present Past

Singular Plural Singular Plural

1st chytá-m chytá-me chyta-l/-la/-lo chyta-li

2nd chytá-š chytá-te chyta-l/-la/-lo chyta-li

3rd chytá-ø chyta-j-ú chyta-l/-la/-lo chyta-li

Future

Singular Plural

1st chytí-m chytí-me

2nd chytí-š chytí-te

3rd chytí-ø chyt-ia

185
There are three imperative forms derived from the 3pl present form chytajú by omitting the
person suffix –ú:

2sg chytaj

1pl chytaj-me

2pl chytaj-te.

In Ria’s speech verb morphology appeared in the Slovak language slightly earlier (1;7)
than in English (1;9). The main observation in verb morphology was the fact that first
inflected forms were context specific and related to activities, actions or events linked to
the familiar people and to self. Initially verbs were used in those forms which frequently
appeared in the input, thus the whole verb paradigm was not available with each type of
verb. Similarly, each type of verb emerged in different tense and mood form.

Over 1;7-1;9 the first spontaneous uses of inflected verb forms were in the singular in all
three tenses, past, present and future. They reflected CDS used by the parents, whereby the
parents referred to Ria either in the 2sg or 3sg forms, while referring to self either with the
1sg or 3sg (e.g. ‘Daddy do it.’ used by the father). This usage was mirrored in Ria’s
language, and she was referring to self mainly with 2sg and occasionally 3sg:

2sg

Present

[jʊːbɪʃ] ‘ľúbiš’ “you love” (1;8.9)

[padaːʃ] ‘padáš’ “you are falling” (1;8.24)

[sɛɟɪːʃ] ‘sedíš’ “you are sitting” (1;9.12)

Future

[ʊkaːʒɛʃ] ‘ukážeš’ “you will show” (1;8.21)

[ɔbʃɛʃ] ‘oblečieš sa’ “you will get dressed” (1;8.21)

3sg

Present

[pɔsɪː] ‘prosí’ “she would like” (1;8.21)

[vɪkʊːka] ‘vykúka’ “is peeking out” (1;9.12).

186
Isolated examples of 1sg appeared as well, in both present and future tenses
simultaneously:

Present

[ʒɔbɪm] ‘robím’ “I am doing” (1;7.28)

Future

*MAM ‘Donesieš to daddymu?’

%eng “Will you take that to daddy?”

*RIA [dɔɲɛsɛm]

%glo donesiem

%eng I will bring (1;8.9).

In her response Ria used the correct 1sg after the mother’s address in 2sg. Thus it was a
productive use of inflected verb with appropriate person suffix, suggesting awareness of
different meanings expressed by different forms.

The 1sg was occasionally overextended to refer to others during pretend play or when
requesting a service from the mother, as in the following examples:

Present

[sɛɟɪːm] ‘sedím’ “I’m sitting” referring to a doll (1;9.4)

Future

[ɔbkʊːʃam] ‘obkúšam (=obkúšem)’14 “I will bite around” referring to the mother


(1;9.26).

From the second half of 1;9 verbs were inflected productively, and Ria was referring to
self with 1sg in present and future tenses. She used 3sg only when referring to self with her
name (1;10.6), thus talking about self as a different person. Varied verbs from all five
Slovak verb classes were used in 1sg:

Present

[ɪɟɛm] ‘idem’ “I’m going/coming” (1;9.12)

[maːm] ‘mám’ “I have” (1;9.20)

[patsʊjɛm] ‘pracujem’ “I’m working” (1;9.26)

14
An occasional error in suffix, applying the –am suffix from a different class of verbs instead of –em.
187
[vɪsavam] ‘vysávam’ “I’m vacuuming” (1;9.26)

[tsɪtʃɪːm] ‘cvičím’ “I am exercising” (1;9.26)

Future

[satam] ‘sadkám (si)’ “I will sit down” (1;9.26)

[ʒɔbɛjɛm] ‘zoberiem’ “I will bring” (1;9.26)

[ɔɟɔʃɪm] ‘odložím’ “I will put away” (1;11.18).

Productive use of verb inflections was evident especially through incorrect vowel
substitution in some future suffixes:

[ɔpam] ‘opam = op(rav)ím’ “I will fix” (1;9.26)

[vɪbam] ‘vybam = vyb(er)iem’ ‘vyberiem’ “I will choose” (1;10.6)

[zapam] ‘zapam = zap(n)em’ “I will turn on” (1;10.6)

[naɟɛm a ju zɔbam] ‘nadem a ju zobam = nájdem a ju zob(er)iem’ “I will find and it I


will take” (1;10.29).

In these examples Ria omitted parts of the word stems and treated the suffix -am from one
verb class as a regular suffix and applied it to different verb classes where the correct
suffixes would be –em, -iem and –ím.

When referring to familiar people and objects (father, mother, doll) Ria used 3sg forms in
verb tenses relevant for the context:

Present

[vaji] ‘varí’ “he is cooking” (1;7.19)

[kaːja] ‘krája’ “is cutting up” (1;7.19)

[kɔpɛ] ‘kope’ “is digging” (1;8.21)

[sɛɟɪː] ‘sedí’ “she is sitting” (1;8.24)

Future

[hɔɟɪː] ‘hodí’ “he will throw” (1;8.21)

[ʊtɛ] ‘utrie’ “she will wipe” (1;8.21)

[pɪːɟɛ] ‘príde’ “he will come” (1;8.21)

[xɔvaː] ‘schová’ “will hide” (1;11.18)

188
Past

[daja] ‘dala’ “she put” (1;7.19)

[bɔʊ] ‘bol’ “he was” (1;7.21)

[ɔdɪʃɛʊ] ‘odišiel’ “he left” (1;9.19)

[kʊːpɪʊ] ‘kúpil’ “he bought” (1;9.26).

When an impersonal verb was related to a different object, but had an impact on Ria, the
3sg was correctly used to describe what Ria felt or liked:

[pɪxaː] ‘pichá’ “it prickles” referring to an object but having an impact on Ria (1;8.21)

[ʃɪːpɛ] ‘štípe’ “it is stinging” referring to what Ria feels (1;8.21)

[xʊcɪː] ‘chutí’ “it tastes= I like” referring to Ria’s likes (1;8.24)

[bɔjɪː] ‘bolí’ “it hurts” referring to Ria’s feelings (1;8.24).

Thus verb morphology also began with inflected forms of frequently used verbs in the
input, which were related to familiar people and self. Inflected verbs referring to states or
actions performed by inanimate or animate entities that were not familiar to Ria appeared
approximately at the end of 1;8, however, they were much less frequent:

Present

[sʊʃɪː] ‘(bielizeň) suší (sa)’ “(washing) is drying” (1;8.21)

[tɔtʃɪː] ‘(šnúra) točí (sa)’ “(washing line) is turning”(1;8.21)

[jɔbɪː] ‘robí’ “is doing” (1;8.23)

[ɪɟɛ] ‘(slimák) ide’ “(snail) is going” (1;8.24)

[ʒavadʒa] ‘zavadzia’ “is in the way” (1;11.7)

[pəʃɪː] ‘prší’ “it’s raining” (1;8.21) - impersonal verb.

Forms inflected in the past tense showed relevant grammatical gender agreement with the
referent:

Past

[bɔjɔ] ‘bolo’ “it was” (1;8.16)

[ɔdɪʃa] ‘odišla’ “she left” (1;9.12)

[padɔʊ] ‘padol’ “he fell” (1;9.26).

189
The fact that learning of verb forms was related to context can also be demonstrated on the
plural forms in Slovak verb morphology. Early instances appeared sporadically from 1;8 in
specific contexts describing actions done by familiar people or objects, and reflecting
forms used regularly by the mother, in 3pl:

Present

[hajʊː] ‘hrajú’ “they are playing” (1;8.16)

[ɟɛcɪ kɪtʃʊː]15 ‘deti kričia’ “children are screaming” (1;9.12)

[havɪnkɔvɪa tʃɪnkajʊː] ‘havinkovia spinkajú’ “doggies are sleeping” (1;10.31)

[jana mamɪ ʃɪnkajʊː] ‘Riana mami spinkajú’ “Riana mummy are sleeping” (1;11.6).

Future

[pɪːdʊ] ‘prídu’ “they will come” referring to the mother’s family (1;8.24).

Past

[tʊkalɪ] ‘tulkali’ “were cuddling” referring to animals seen at the zoo (1;10.6).

The 1pl appeared at the beginning of 1;9 mirroring the mother’s parenting style, whereby
she frequently used the associative ‘we’ in 1pl, as a directive to teach expected behaviour,
as in the following example:

%act RIA throws a book on the ground

*MAM Knižky nehádžeme!

%eng We don’t throw books!

*RIA [tʃɪːtamɛ]

%glo čítame

%eng we read (them)

(1;9.8).

This directive was increasingly more frequent in the CDS as Ria was becoming
independent in exploring her surroundings, thus offering many types of verbs. The
associative ‘we’ appeared in Ria’s talk about activities she was engaged in or about to do,
thus reflecting the mother’s typical usage, for example:

15
In this particular example Ria overgeneralised the 3rd person plural suffix -ajú from one class of verbs and
applied it to a prototype verb of a class which normally takes an -ia suffix in 3pl. This suffix substitution
remained in Ria’s speech until 3;11.
190
[jɛʃɪːmɛ] ‘ležíme’ “we are lying down” referring to mother and self (1;9.19)

[ɪɟɛmɛ] ‘ideme’ “we are going” referring to the family (1;9.12)

[pʊcɪːmɛ] ‘pustíme’ “we will turn on” referring to the mother and self (1;9.26)

[ʃabamɛ] ‘vstávame’ “we’re getting up” referring to the family (1;10.16).

The last Slovak verb form to be acquired was the 2pl, which appeared in an isolated
example in the imperative at the beginning of 1;11 when Ria addressed both parents
simultaneously:

[pɔcɛ sabac] ‘poďte stavať’ “come build” (1;11.6).

Overall Ria used Slovak inflected verbs appropriate in contexts by the second half of 1;9.
It is not possible to talk about the development of verbal inflections in each tense
separately, since Ria did not learn individual tenses in a progression (e.g. present, past,
future), but rather learnt verbal forms appropriate in the contexts in whatever tense or
mood, in ways that reflected the use of verbs by the mother in the input. This could be
explained by the fact that (as the diary states) by the time inflected forms appeared Ria had
an understanding of past and near future events as differing from events taking place in the
present.

7.4.3.1 Other verb structures


Other verb structures emerged during this stage, however, they were learnt as unanalysed,
context dependent wholes. Among the first were Slovak infinitives which appeared as
holophrases from 1;8. A Slovak verb is in the infinitive form if a suffix –ť is attached to
the stem form, as in chyta-ť ‘to touch’. In Ria’s productions infinitives substituted a
phrasal construction and where used to request an activity or as a commentary when Ria
was about to start an activity, as shown in these examples:

‘chcem (sa) + INF’ “I want to infinitive’

[tʃɪːtac] ‘čítať’ “to read” (1;8.21)

[tatʃɪc] ‘tlačiť’ “to push” (1;9.12)

[kɛsɪc] ‘kresliť’ “to draw” (1;10.6)

‘poď (sa) INF’ “come and infinitive”

[hac] ‘(poď sa) hrať’ “(come and) play” (1;9.12).

Infinitives could also have the function of an imperative, as in the following excerpt:

191
%sit Ria is getting dressed

*RIA [ponoʃɪ]

%glo ponožky

%eng socks

*MAM No, vyber si nejaké. Aké si vyberieš? Vyber si zo skrine nejaké.

%eng Yeah, pick some. Which ones will you pick? Pick some from the cupboard.

*RIA [ʒo ʃɪɲɛ. pɔmɔts. pɔmɔts.]

%glo zo skrine. pomôcť. pomôcť.

%eng from cupboard. to help. to help.

*MAM Mám ti pomôcť?

%eng Should I help you?

*RIA [anoː]

%glo ano

%eng yes

(1;11.7).

Infinitives were not a direct command but they mirrored the mother’s usual response to
Ria’s requests for help, a construction consisting of a modal verb and an infinitive, as
shown in bold above. The function of a command was typically expressed with the
inflected 2sg imperative form of several verbs in specific contexts, again reflecting the
mother’s talk:

[kʊʃ] ‘(po)kúš’ “chew!” (1;7.18)

[pɔsɪ] ‘pozri’ “look!” (1;10.29)

[ʊkaʃ] ‘ukáž’ “show (me)!” (1;11. 18).

It appeared Ria derived infinitive forms from inflected verbs rather than from stem forms.
This was evident with verbs where in Slovak changes occur in the stem of the verb in some
inflections, while the infinitive form includes the basic stem with infinitive suffix. Such
inflectional overgeneralisation emerged with the reflexive verb báť sa, where bá- is the
stem and -ť the infinitive suffix (sa is a reflexive pronoun). The inflected present tense

192
forms bojím sa “I’m scared”, bojíš sa “you’re scared”, etc., have a -j- inserted between the
stem and inflectional endings16. Ria overgeneralized this rule to the infinitive form as well:

[bajac] ‘báť (sa)’ “to be scared” (1;9.26).

This suggests that infinitives were a later production. Ria acquired inflected forms relevant
to a particular situation first and derived the infinitive from them.

There are other grammatical verb categories expressed through morphology in the Slovak
language. The category of Aspect denotes the contrast between completion and lack of
completion of an action. It is marked with morphological markers in the form of prefixes.
Such prefixes were also subject to item-based learning and depended on the mother’s use
of the language. The first isolated instance in Ria’s speech was:

[pɛtʃɪːtac] ‘prečítať’ “to read through” (1;8.24)

This holophrase has the above function of the ‘I want to infinitive’ construction and could
be glossed as ‘I want to read this through.’, while the verb itself has an additional
syntactical marker, the Aspect prefix pre-, which denotes the fact that the speaker wishes
the action to continue until completed. This particular verb construction was frequently
used by the mother when referring to one of Ria’s favourite activities - reading with adults.
Prior to the use of the additional grammatical aspect Ria requested reading books with the
infinitive form [tʃɪːtac] ‘čítať’ “to read’ (1;8.21).

Later instances of Aspect appeared with frequently used verbs in the input:

[vɪxaɟɛnɛː] ‘vychladené’ “cooled down” (1;10.26) – a past participle of the verb


vychladiť, in which the prefix vy- marks the resulting state

[mamɪ dɔpapaʊja] ‘mami dopapala’ “mummy finished eating” (1;11.1) – past tense
with a prefix do- marking completion of an action

[napʊɲɪc] ‘naplniť’ “to fill up” (1;11.3) – infinitive with a prefix na- marking result of
an action

[mamɪ vɪpapaja] ‘mami (ja) vypapala’ “mummy (I) ate out” (1;11.3) – past tense verb
with vy- prefix marking completion of the action

[pɔsɪnkac ɲɪɛ] ‘pospinkať nie’ “sleep no” – infinitive verb with a po- prefix marking
duration of action.

16
There is also a vowel change from a to o in the inflected forms
193
Lastly, Slovak reflexive verb structures appeared, which consist of a verb and reflexive
pronouns sa or si. The mother used such phrases frequently when addressing Ria, who
learnt them in several fixed expressions linked to specific contexts:

[ɲɛbɔɪsa] ‘neboj sa’ “don’t be scared” (1;9;19), referring to self

[hacsa] ‘(poď) sa hrať’ “(come and) play” (1;9.20)

[vɪdʊ, pɔsɪmsɪ] ‘vidličku, prosím si’ “fork, I would like” (1;10.27).

These structures were produced under single intonation contour as one unit formed by
inflected verb + reflexive pronoun throughout the two-word stage.

7.4.4 The verbs ‘byť’ and ‘to be’


The irregular verbs byť and to be have the following forms in the verb paradigms in the
two languages:

Singular Plural

Slovak English Slovak English

1st (ja) som I am (my) sme we are

2nd (ty) si you are (vy) ste you are

3rd (on/ona/ono) je he/she/it is (oni/ony) sú they are

Inflected forms of the two equivalents emerged around the same time in unanalysed item-
based frames which Ria learnt as wholes. In Slovak Ria was using the constructions
kde+verb “where+verb” and tu+verb “here+verb” as one unit. Four verb forms were used
while Ria was not yet inflecting the verb productively:

2sg [ɟɛsiː?] ‘kde si’ (1;7.17)

3pl [ɟɛsʊː] ‘kde sú’


[tʊsʊː] ‘tu sú’ (1;8.9)

3sg [tʊjɛ] ‘tu je’ (1;8.9)

[ɟɛjɛ] ‘kde je’ (1;8.21)

1sg [ɟɛsɔm?] ‘kde som’ (1;9.4)

[tʊ jɛm!] ‘tu jem (= tu som)’ “here I am!” (1;10.10)

As shown in the last example, instead of using the correct irregular form ‘som’ Ria
regularised the irregular 3sg ‘je’ and applied it to 1sg by attaching a regular verb suffix,

194
producing [jem]. Thus Ria briefly used the verb productively as a regular verb. The 3sg
form [jɛ] ‘je’ “is” was in fact the only productive form used in several utterances,
indicating that the verb paradigm was not yet firmly established:

[mamɪ jɛ ʃaʃɪnka] ‘mami je šašinka’ “mummy is silly billy” (1;11.2)

[tʊ jɛ jɪbɪtʃa] ‘tu je rybička’ “here is the fish” (1;11.3)

[tɔtɔ jɛ ʒɪjafa] ‘toto je žirafa’ “this is giraffe” (1;11.6)

[tɔtɔ jɛ vɪxaɟɛnɛ] ‘toto je vychladené.’ “this has cooled down” (1;11.7)

[tam jɛ pɪgɛt] ‘tam je Piglet’ “there is Piglet” (1;11.18).

Likewise, in English the first forms of the verb paradigm appeared in item-based frames,
however, only in the 3sg:

[deəːʃiːs] ‘there she is’(1;7.14)

[deːəɪs] ‘there is’ (1;8.21)

[dɪsɪs] ‘this is’ (1;9.12)

[detɪs] ‘that is’ (1;10.6).

One other form appeared in a formulaic unit, which Ria learnt as a whole while talking to
the father on the phone:

[hajəʊ dedi. hajəʊ dedi. aː juː hoʊm?] ‘hallo daddy. hallo daddy, are you home?’

Other English forms were not observed productively before 2;0.

7.4.5 English verb morphology


English verb morphology emerged with an apparent time lag of approximately a month
after Slovak verb morphology. First English verbs were used in the stem form in
descriptions of actions or in the function of imperatives:

[luk. babo. babo.] ‘look. baby. baby. (=baby is looking)’ (1;8.9)

[sɪt] ‘sit!’ (1;8.9).

The first English inflected form was the progressive –ing, which appeared at the beginning
of 1;9, thus around the same time as Ria was using Slovak 1sg to refer to self and 3sg to
refer to others. In English Ria also referred to self in holophrastic descriptions of actions or
states:

[kamɪŋ] ‘coming’ (1;9.4)

195
[kiːnɪŋ] ‘cleaning’ (1;9.12)

[haːɪdɪŋ] ‘hiding’ (1;9.19)

[peɪŋ] ‘playing’ (1;9.20);

The progressive form also referred to others in multiple-word combinations:

[aːɪs, jukɪŋ. jana.] ‘eyes, looking. (at) Riana.’ (1;9.12)

The second inflected form to emerge was the English 3sg suffix –s in isolated holophrases:

[kams] ‘comes’ (1;9.4)

[sɪts] ‘sits’ referring to a spider (1;10.6).

Although irregular past tense forms of some verbs also emerged during this stage, Ria used
them unanalysed and only in specific contexts:

[gon] ‘gone’ (1;7.7)

[dan] ‘done’ (1;8.9)

[boʊk] ‘broke’ (1;11.23).

English phrasal verbs were also used during this stage as units within pivot schemas. Ria
produced them as wholes consisting of the verb as a changing element, while the other
element was constant. Such schemas expressed various structures, such as the transitive
verb construction, as described in section 7.3.2.

In summary, from 1;8 onwards Ria used context relevant inflected forms which reflected
the morphology of the two languages appropriately. She did not ignore the need for varied
morphological markers in the Slovak language, nor did she attempt to use morphological
markers with English productions where they were not required by English grammar.
Moreover, words that were shared by the two languages, such as appellatives, were used
with appropriate morphological markers in Slovak language contexts, as modelled in Ria’s
input, and without markers in English. Relevant Slovak noun, adjective and pronoun
declinations were evident across utterances and used productively. Likewise, in English
plural and possessive forms were used when necessary.

Use of inflectional verb morphology was related to situational contexts. Ria used forms
required for specific contexts first, and filled the verb paradigm as a later development.
Thus the correct use of the verb form, subsuming the grammatical functions of the person,
tense, mood and aspect was embedded in the situational context, and reflected verb forms
modelled by the parents in the input. Ria gradually built up an understanding of what was
196
appropriate in an item based manner, and eventually grasped regular patterns and applied
them across verb types. This finding is in line with previous research on early verb
learning, suggesting that early syntactic marking used by young children is verb-specific
and learnt item-by-item (Serratrice, 2001, p. 47; Tomasello, 2003, p. 119).

In verb morphology a lead-lag pattern was observed in the two languages. In the early two-
word stage, Ria was using inflected verb forms in Slovak as soon as she was able to
produce the whole word. In English, on the other hand, the bare forms were used for
several months, and the 3sg marker appeared slightly later. There are several possible
explanations for the lead-lag pattern.

One explanation could lie in the fact that while at the time morphology emerged (1;7)
Ria’s combined total vocabulary was more than 250, her vocabulary in each language
differed, with Slovak being the stronger language at this stage. Moreover, Ria spent
majority of wake-time in a Slovak context of the home. This explanation might imply
developmental delay in verb morphology in English, Ria’s less dominant language.

The lead-lag pattern could also be a reflection of the morphological differences between
the languages, rather than developmental delay. Inflectional suffixes are more salient in
Slovak than in English. Indeed, the pattern mirrors previous research on verb morphology
development in bilingual children with English and another morphologically richer
language (Serratrice, 2001, p. 69). Thus the finding in this thesis might imply a typical
lead-lag developmental pattern in a language combination consisting of English and a
Slavic language, a combination that has not been examined from this aspect so far
(Serratrice, 2001, p. 43).

However, when the verb types were examined, it was found that Ria did not acquire the
whole verb paradigm simultaneously (which would then be applied to all verb types).
Rather, she was initially producing verb forms which appeared as most salient in her input,
and the specific forms in each verb type reflected how the verbs were used in contexts.
Thus the lead-lag pattern was only apparent as a surface structure.

Moreover, Döpke (2001, p. 84) compared bilingual children acquiring English and
German, also a morphologically different language, with monolinguals acquiring the
respective languages, and found a similar lead-lag pattern in all groups. Thus the pattern in
development of finiteness in bilingual children paralleled the typical later realisation of
finiteness in English monolinguals in comparison with German monolinguals. These
findings indicate that morphologically richer languages show faster realisation of verb

197
finiteness. More studies on verb finiteness in Slovak-English BFLA, as well as
monolingual language acquisition in the respective languages, are needed to confirm the
pattern in this language combination.

Overall, grammatical development was taking place in a separate yet side-by-side fashion,
providing evidence for grammatical differentiation in the early combinatorial stage.

198
8 Emerging Bilingualism
In this chapter Ria’s developing bilingualism will be discussed. It will consider
developments from previous chapters in light of initial bilingual awareness, language
differentiation, pragmatic language choice and language mixing as interrelated building
blocks of emerging bilingualism. Language mixing and language choice is considered in
the one-word and two-word developmental stages, which have not been as well researched
as later developmental stages (Nicoladis & Genesee, 1996, p. 441; Quay, 1995, p. 371).
This chapter attempts to deepen the understanding of this phenomenon.

Language choice and code-switching (CS) are the results of language interaction. In adult
bilinguals language choice, in other words a decision which language is appropriate when
and with whom, is determined by different domains of language use, which are constructed
by varied factors such as participants, place and topic (Hoffmann, 1991, p. 178). When
bilinguals interact together they rely on both of their languages in the same conversation.
This has been termed code-switching (CS), and is broadly defined as alternate use of a
bilingual’s two languages (Hoffmann, 1991, p. 110), although exact definitions of CS and
terminology used tend to vary across studies. In adult bilingual speech CS is considered a
natural aspect which becomes a permanent feature of a bilingual speech community
(Hoffmann, 1991, p. 75). It can be observed in older children as well. Grosjean (1998, p.
137) further distinguishes between borrowing and switching. Switching does not have to
be integrated morphologically or phonologically, it is simply a switch in the language and
often occurs on the inter-sentential level. In contrast, borrowing is integrated and often
occurs on intra-sentential level.

In study of young children all forms of language interaction in the speech of the
developing bilinguals are termed mixing, and as Hoffman (1991, p. 75) explains, it seems
to be of a more transitory nature. However, bilingual children also need to develop a
pragmatic ability to alternate languages when it is appropriate to do so – the ability of
language choice. It is difficult to distinguish between the terms language choice and
mixing in the research literature, since no clear definitions are provided. It appears that the
same phenomenon is being studied from different perspectives. The term mixing focuses
on structural aspects of alternate use of a developing bilingual’s two languages, while
language choice focuses on pragmatic aspects. Language differentiation, on the other hand,
describes the process in which a child develops an awareness of two languages in her input
and in her productions, and as such it subsumes both mixing and language choice.
199
In the early literature developing bilinguals were criticized for mixing (Grosjean, 2008, p.
12). It was suggested that mixing was a result of poor development in one or both
languages, the development of fused linguistic system and lack of differentiation. While
ultimately it may not be possible to attain equal mastery of both languages, mixing can
actually show that a child is using pragmatic knowledge and has an awareness of two input
languages. Even when words are mixed from the other language, the concepts expressed
by the mixed word are correct, and they are used in pragmatically appropriate situations. It
is the linguistic form that is in the ‘wrong’ language. This can be regarded as evidence that
bilingual children who mix are proficient users of their two languages, and they are
substituting one language for another only on a temporary basis to maintain the continuity
of interactions (Cruz-Ferreira, 2006, p. 69). This approach to mixing provides an
alternative analysis, especially in small bilingual children. According to Cruz-Ferreira
(2006, p. 28), bilingual children learn which words belong to which language through a
process in which they attempt different language choices, until they eventually arrive at a
pattern consistent with patterns used by experienced speakers in their environment, i.e. the
target languages.

While language choice and language mixing are interrelated, in this thesis these two
phenomena are discussed separately. Discussion of language mixing in section 8.1 refers
to the combining of constituents from Ria’s languages-in-acquisition in individual
utterances, thus to structural properties of mixing at utterance level. Unit of analysis was
the mixed utterance. The amount of mixing occurring in each language context, structural
properties of mixing, and reasons for mixing are addressed. Ria’s use of mixing is also
considered in terms of monolingual and bilingual language processing and Ria’s language
mode on a continuum ranging from monolingual to bilingual modes (Grosjean, 1998, p.
136).

In this thesis borrowing refers to borrowing at a family level, when particular words were
‘borrowed’ from one of the languages and adapted by the whole family. For example the
father borrowed the Slovak word ‘bábo’ and used it instead of the English equivalent
‘baby’ when interacting with Ria, often in plural and possessive forms ‘babos’/‘babo’s’.

Language choice in section 8.2 refers to Ria’s developing pragmatic ability to use
contextually appropriate language in one-word and two-word utterances. Since language
choice assumes a level of pragmatic ability, following Lanza (2004, p. 198), qualitative
analysis at discourse level will be discussed, with unit of analysis being a conversational
turn at talk.
200
8.1 Ria’s language mixing
To fully assess Ria’s mixing the addressee and language context was recorded for each
utterance produced by Ria. Utterances were then categorized according to how words from
the two languages were used and combined. Based on previous research five utterance
categories were established (Deuchar & Quay, 2000, p. 70; Lanza, 2004, p. 123):

 Slovak unilingual
 English unilingual
 Indeterminate - utterances with word(s) that could belong to either language
 Mixed – lexical morpheme(s) from one language combined with bound morpheme
from the other language, or lexical morphemes, each from a different language,
combined into single utterance
 Bilingual – TEs combined into single utterance used by Ria when addressing both
parents simultaneously, repairing language choice or clarifying a new word, these were
pragmatic utterances which were not truly multiple-word combinations.

Moreover, the context of each utterance was coded. In bilingual contexts especially,
utterances had to be carefully coded according to addressee as well as the type of context,
since even in triadic interactions Ria often made her addressee explicit. The following
categories were found:

 addressing mother in Slovak context


 addressing father in English context
 addressing mother in bilingual context
 addressing father in bilingual context
 addressing both simultaneously in bilingual contexts - triadic interactions.

Overall, mixing rates were low. During the one-word stage (1;4-1;6) analysis of mixing
considered whether Ria differentiated between her parents’ preferred languages in her
language use. The unit of analysis was a holophrastic utterance consisting of a single word.
In total 517 tokens of one-word utterances were recorded. Ria was Slovak dominant, as
shown by the overall Slovak utterances used in comparison to English utterances.
However, this was due to uneven distribution of language contexts, as explained in section
4.3.1, rather than lack of differentiation. Ria still used more of the appropriate language
with each parent, and used any language when addressing both parents simultaneously in
triadic interactions, as shown in Figure 2.

201
Figure 2 Use of one-word utterances within contexts

250

200

150 Slovak utterances


English utterances
Language neutral
100

50

0
Addressing mother Addressing father Triadic interactions

In Figure 2 the distribution of utterances addressed to each parent per context was
collapsed. However, even in individual address of parents in bilingual contexts Ria
differentiated between each parent’s preferred language appropriately. When addressing
both parents simultaneously, Ria’s language use was distributed relatively equally among
Slovak and English utterances, thus showing sensitivity to the bilingual context.

Moreover, Ria mixed more in her more dominant language, Slovak, with her bilingual
mother than she mixed with her monolingual father. Of the total 517 tokens recorded only
54 tokens were categorized as mixed utterances, since they represented inappropriate
language choice within context. They consisted mostly of English items used in
interactions with the mother in either Slovak or bilingual contexts, as illustrated in table
Table 8.1. Only several tokens were uses of Slovak utterances in addressing the father,
which occurred mostly in bilingual contexts with the mother present. One token was an
address of a different English speaker with a Slovak word.

Table 8.1 Use of mixed one-word utterances in different language contexts (inappropriate language choice)

Addressee
Language context English to Slovak to
mother father
Slovak 30 -
English - 4
Bilingual 11 8
Totals 41 12

202
With the onset of combinatorial speech a unit of analysis became an utterance consisting
of two-word or multiple-word combinations. Utterances recorded in the diary and video-
recordings were categorized as above. Since combining utterances can potentially increase
language mixing, approximate proportions of each type of utterance in Ria’s productions
were determined for each month, as shown in Table 8.2.

Table 8.2 Use of different types of combined utterances (in proportion to total combined utterances)

Total Slovak English


Age Mixed Indeterminate Bilingual
utterances unilingual unilingual
1;4 5 2 1 0 2 0
1;5 12 5 3 2 0 2
1;6 39 25 5 6 3 0
1;7 50 28 8 6 5 3
1;8 113 74 20 12 3 4
1;9 98 61 22 12 0 3
1;10 288 215 40 28 4 1
1;11 444 304 87 45 2 6
1049 714 186 111 19 19
Totals
100% 68.06% 17.73% 10.58% 1.81% 1.81%

Number of mixed utterances indeed increased with the total number of utterances,
however, mixed two-/multiple-word utterances still accounted only for a small proportion
of total production in each month. List of mixed utterances is reported in APPENDIX H
and will be discussed below.

Ria’s dominance in Slovak as function of the primary sociolinguistic group was evident in
combinatorial speech as well. Nevertheless, she used predominantly unilingual Slovak and
unilingual English utterances when addressing her parents respectively. Thus she used
more of the appropriate language with each parent, as illustrated in Figure 3. In triadic
interactions, Ria used either language, showing sensitivity to the bilingual context.

203
Figure 3 Use of two/multiple-word utterances within contexts

600

500

400
Slovak utterances
300 English utterances
Mixed utterances
200

100

0
Addressing mother Addressing father Triadic interactions

Mixing occurred mainly when addressing the bilingual mother, followed by bilingual
contexts in triadic interactions. As shown in Table 8.3, Ria rarely addressed the
monolingual father with mixed utterances. The fact that mixing pattern was different with
each parent indicates Ria differentiated parental language proficiency (Quay, 2008, p. 20).

Table 8.3 Use of mixed combined utterances in different language contexts

Language Addressee
context Mother Father Both Self
Slovak 51 - -
English - 3 - 1
Bilingual 37 7 29
Totals 88 10 29 1

Occasionally Ria addressed her parents with a unilingual utterance in a language which the
parent did not normally use to address her, as shown in Table 8.4.

Table 8.4 Use of unilingual combined utterances in inappropriate contexts

Addressee
Language Unilingual Unilingual
context Slovak to English to
father mother
Slovak - 21
English 6 -
Bilingual 1 3
Totals 7 24

204
These utterances were considered mixes, since they were inappropriate language choices.
However, such uses were not frequent.

In childhood bilingualism literature several reasons for mixing were identified (Goodz,
1989; Hoffmann, 1991, p. 107; Lanza, 2004; Serratrice, 2005):

 mixing in parental input


 language dominance
 lexical gap in the vocabulary of one of the languages, since BFLA children acquire
items in one language at a time
 temporary unavailability of a word (reasons for such unavailability are often not
discussed)
 complexity of an item in one of the languages
 language preference.

If mixing in input could provide a possible interpretation for Ria’s mixing in her
productions, parental mixing pattern would be reflected in her speech. She would be
expected to use mainly unilingual utterances, while mixes would be limited to bilingual
and occasionally Slovak contexts, but avoided in English contexts. Indeed, the parental
mixing pattern discussed in 4.5.3 was reflected in the data. Ria mixed more productively
into Slovak with the bilingual mother, while mixing with the monolingual father was
limited. However, some differences from the pattern were evident. The parents rarely
mixed when addressing Ria directly, and Ria mixed in both languages more than she
observed in her input. Moreover, Ria was exposed to overheard bilingual code-mixing in
situations when the mother was interacting with her sister’s family, who were also
bilingual in Slovak and English. Thus it appears Ria learnt the ‘rules of mixing’ from
overheard speech rather than direct parental input. This suggests overheard speech plays an
important role in acquisition of pragmatic aspects and can affect production. Thus mixing
in parental input did not provide an exhaustive explanation.

Language dominance was another determinant of mixing found in some bilingual studies,
where children mixed more in their less proficient language (Genesee, et al., 1995;
Nicoladis & Secco, 2000). However, Lanza (2004, p. 136) examined directionality of
mixing in different contexts and found that the bilingual children used the grammatical
framework of their more dominant language with lexical morphemes from the less
dominant language, but not vice versa. Similarly, Quay (2008, p. 22) found that the
trilingual child in her study was not mixing more in her least proficient language.

205
In this thesis mixing also occurred mostly in the dominant language. As shown in 4.3.1 Ria
was dominant in Slovak because her relative input was greater in Slovak, and the mother
engaged her in child-centred activities during which she followed Ria’s lead and fed into
her interests. Slovak was the language of the closest emotional bond with the mother, who
was the main caregiver. Analysis of utterance types and approximate proportions of each
type also indicated dominance in Slovak. Contextual analysis of mixed utterances showed
that mixing occurred primarily as English lexical and grammatical morphemes in Slovak
grammatical framework, thus in Ria’s dominant language. Weaker proficiency therefore
cannot explain why most mixing occurred in monolingual Slovak and bilingual contexts.

In bilingual contexts bilingual language mode offers an interpretation, since Ria had both
languages activated and used the one she recalled faster in a particular instance. In
monolingual contexts mixing with the mother can be interpreted by Ria’s awareness of the
mother’s bilingualism, which meant that Ria had the other language, English, slightly less
active, but never deactivated completely. Mixing with the father in English or bilingual
contexts was rare, suggesting Ria was aware of the father’s monolingualism. Moreover,
since she used mostly lexical mixing with the father, and was more dominant in Slovak,
mixing can be explained by pragmatic reasons such as lexical gaps and temporary
unavailability of items, or psycholinguistic reasons such as preference for specific words
and learning context.

Lexical gap as a determinant of mixing was most prominent in the early one-word stage. If
Ria did not have both equivalents in a TE pair available in expressive vocabulary, since
learning of a couplet was tied to a specific context, she chose to use the available context
appropriate item from the other language, mixed into the base language, in order to avoid
communication breakdown, as shown in the following excerpt:

*RIA [pɪxaː. tɔtɔ.]

%glo pichá. toto.

%eng prickles. this.

*MAM Čo pichá?

%eng What prickles?

*RIA [gas]

%glo grass

(1;8)
206
Mixing due to lexical gaps did not cause a breakdown in communication for the bilingual
mother. It was a useful strategy to maintain the communication flow also with the
monolingual father, since on many occasions the meaning of the borrowed word could be
determined from the situational context:

*RIA [pʉʃɪː]

%glo prší

%eng it’s raining

*DAD It’s raining, isn’t it?

*RIA [ɪznɪt?]

%glo Isn’t it?

(1;9.19)

In the two-word stage lexical gaps in English occurred mainly because inflected present
and past tense forms of a modal verb were available in Slovak only, for example two
inflections of the verb mať “to have”:

[maː] ‘má’ “has” (1;7.3)

[maːm] ‘mám’“(I) have” (1;8.24).

However, since Ria tended to acquire translation equivalents relatively quickly, other
reasons were more relevant with increasing vocabulary. Among these temporary
unavailability of a previously used item was the most likely reason to mix. The most
plausible explanation for a temporary unavailability is that Ria was in a bilingual mode
with both languages active, and accessed the lexical item faster in the other language.
Failure to retrieve a word resulting in mixing was found in other studies as well (Lanvers,
2001, p. 449; Lanza, 2004, p. 165; Serratrice, 2005, p. 169).

Another major reason for mixing evident in the data was Ria’s preference for specific
words. In fact more detailed analysis of mixed utterances found that many mixed items
were tokens of the same preferred words used in different types of utterances. Occasional
short-term preference for a newly acquired item was found, but Ria also formed personal
preference for certain words which she maintained long-term, as will be discussed in 8.1.2.
Many mixed items were family words borrowed by both parents, thus reinforced as
belonging to both languages. Lastly, many mixed two-word utterances were English
lexical items negated with the Slovak negator. This usage was a developmental

207
phenomenon, as discussed in 7.3.3, rather than pure mixing. Complexity of items was not
found to be a determinant of Ria’s mixing.

Thus lexical mixing was explained by sociolinguistic reasons such as parental and
overheard input, lexical gaps or temporary unavailability, as well as by psycholinguistic
reasons such as bilingual mode, word preference and word learning contexts.

8.1.1 Intra-sentential mixing


Lanza (2004, p. 215) notes that in childhood bilingualism literature different types of
mixing over time are not examined, and suggests an analytical distinction between lexical
and grammatical mixing. The children in the study brought lexical morphemes from one
language into the base language, and combined them with grammatical morphemes from
the base language. Thus lexical mixing could also contain grammatical mixing. However,
Lanza (2004, p. 216) also found purely grammatical mixing, when grammatical
morphemes from one language were combined with otherwise base language lexical items.
Lanza (2004, p. 126) also points out the difficulties in categorizing word types as lexical
and grammatical when considering child speech from a sociolinguistic perspective,
especially in cases of bilingual development where two morphologically different
languages are acquired. Instead, Lanza distinguishes between open class (free morphemes)
and closed class (bound morphemes and function words) items. This distinction was used
also in this thesis.

In Ria’s data lexical mixing with or without grammatical mixing was found as well. From
structural point of view, during the one-word stage and the first three months of the two-
word stage (until 1;9), before grammatical morphology was established, Ria mixed by
bringing lexical or grammatical free morphemes from one language into an otherwise base
language utterance without integrating them morphologically. Overall, this type of mixing
prevailed until 2;0.

In the one-word stage, holophrastic utterances were considered mixed if they were in the
inappropriate language. Analysis in section 8.1 showed that Ria’s mixing rate was
relatively low. More detailed analysis of mixed utterances found 42 tokens of 30 word
types, of which 22 were English words used with the mother and 8 Slovak words used
with the father. When the types of words are considered, Ria mixed both open class and
closed class morphemes. However, overall more open class than closed class items were
mixed, as shown in Table 8.5. Grammatical mixing in the form of free closed class
morphemes occurred mainly within Slovak contexts with items drawn from English. None

208
of the mixed words contained bound grammatical morphemes, thus purely grammatical
mixing was not found.

Table 8.5 Mixed word types across languages (one-word stage)

Slovak English
Open class items 7 19
Closed class items 3 13
Totals 10 32

A list of mixed words from the one-word stage is reported in APPENDIX G. When
individual items were considered, Ria appeared to have several preferred words which she
used more often, although she had TEs available in comprehension and production. Since
availability of TEs was verified from the cumulative vocabulary list, many mixed items
could be categorized as isolated instances of momentary lapse or retrieval difficulty while
in bilingual language mode. Indeed, four open class items were couplets from TE pairs
teeth-zuby and key-kľúč, which reinforces interpretation of lexical mixing as result of
difficulty of item retrieval. Some items were mixed as novel words for which Ria did not
yet have the TEs.

During the 2-word stage, after some morphology in both languages was established (from
1;9), mixing that had the form of adult intra-sentential code-mixing also emerged. This
form of grammatical mixing was infrequent, but it was observed in both holophrases as
well as mixed utterances. In total 111 tokens of mixed utterances were found in the sample
of two/multiple-word utterances. When types of mixing were considered, mostly lexical
mixing occurred. Only 27 tokens of 18 word types contained grammatical mixes consisting
of lexical morphemes integrated morphologically and phonologically into an otherwise
base language utterance. None of these items contained purely grammatical mixing. Some
lexical items were indeed found as mixes in multiple instances with different inflectional
endings, as will be shown below. Morphologically integrated family words (used as
borrowings) accounted for 5 of the 27 tokens. Blends were also found and categorized as
specific type of grammatical as well as inflectional mixing, as will be discussed in section
8.1.2.

Structural analysis showed the earliest instances of grammatical mixing could not be
classified as such on Ria’s part, since they were nouns which the parents regularly

209
borrowed from the other language. In Slovak the following mixed forms where learnt
directly from input:

[tʃɪpsɪ] ‘čips-i’ “chips” (1;8.21)

[tʃɪpa] ‘čips-a’ “chip” (1;9.12)

[sosɪtʃe] ‘sausage-e’ “sausages” (1;9.12).

The English borrowing ‘čipsy’ is widely used in the Slovak non-standard language17,
however, it refers to dried chips. The singular form is not used in the Slovak language. The
mother often used the nonstandard form to refer to hot chips, thus using a lexical
transference when interacting with Ria. The mother also occasionally mixed the English
word ‘sausage’18 into Slovak base language and integrated it morphologically.

Likewise, a Slovak word was used with English plural suffix in a case where the father
adopted a Slovak word and used it regularly:

[babos] ‘bábos’ “babies’ (1;8.24).

Therefore, while in the above examples borrowing into the base languages occurred, due to
the modelling of these words in the input, Ria perceived them as belonging to either
language and used them accordingly.

In order to avoid communication breakdown when a word was not readily available Ria
used intra-sentential mixing in a productive way that was not observed in the input.
Moreover, there was a difference in the type of mixing used in different language contexts,
as discussed in the following sections.

8.1.1.1 Mixing with the mother


From 1;9 till 2;0 Ria produced several mixed utterances consisting of English lexical
morphemes integrated phonologically and morphologically with Slovak bound morphemes
into an otherwise Slovak base utterance. This form of mixing occurred with English
adjectives, nouns and verbs used in the Slovak context with a Slovak suffix, which marked
the relevant grammatical relationships and gender agreement.

English adjectives are usually not integrated morphologically when mixed by adult
bilingual speakers, and such mixed structures were not used by any bilinguals in Ria’s
input (i.e. mother and her family). The bilingual mother judged them as non-conventional

17
The standard Slovak word is ‘lupienky’.
18
The Slovak equivalent ‘klobása’ is conceptually slightly different to the English word ‘sausage’, which
refers to various kinds of smallgoods, while in Slovak different types are referred to with specific labels.
210
types of language-mixing. Therefore, without a model for mixing adjectives into Slovak,
these occurrences cannot be explained by mixed input, but were used by Ria productively.

Ria formed adjectival mixes by combining English lexical morphemes with Slovak bound
grammatical morphemes (Instrumental, Accusative and Nominative suffixes), maintaining
Slovak grammatical gender agreement with the modified referent:

[giːnɔʊ] ‘green-ou’ “with green (marker)” Instrumental feminine (1;9.12)

[gɪonʊ] ‘green-u’ “a green (marker)” Accusative feminine (1;9.12)

[gɪna] ‘green-a (farba)’ “green (colour)” Nominative feminine (1;9.19)

[gɪunʊ] ‘green-u (bodku)’ “green (dot)” Accusative feminine (1;9.19)

[gɪanɛː] ‘green-é (jablko)’ “green (apple)” Nominative neuter (1;9.19)

[gɪanɪː. zɛjɛnɪː] ‘green-aný. zelený’ “green. green” Nominative masculine (1;9.26)

[bɪga] ‘big-a (miešačka)’ “big (cement mixer)” Nominative feminine (1;9.19)

[adujuɲʊ] ‘other one-u’ “other one” Accusative feminine (1;11.4)

[mamɪː, nadujana!] ‘mami, another one-a!’ “mummy, another one!” Nominative


feminine (1;11.5)

These instances of mixing cannot be explained through vocabulary gaps either, since Ria
produced all TEs previously, for example the Slovak word [ʃɛɟɛ] ‘zelená’ “green” (1;6.10),
was used as the first TE (produced as [zɛjɛnaː] by 1;9.1), while the English equivalent [giː]
appeared at 1;6.25 (produced as [gɪn] by 1;7.28). Thus temporary preference or
unavailability for the English items is a plausible explanation.

Moreover, Štefánik (2000, p. 79) also found that after the age 2;5 the BFLA child in his
study, Natália, productively integrated English adjectives with Slovak grammatical
morphemes. This finding suggests such mixed structures could be developmental, and
needs to be confirmed in further studies with adult and childhood Slovak-English
bilinguals.

Conventional types of mixing found in Slovak-English bilingual speech also appeared in


Ria’s productions. These were English nouns integrated through Slovak nominal case
endings, which reflected the grammatical gender of the Slovak equivalent noun:

[zabɪʃana] ‘(do) rubbish-a’ “into the rubbish” Genitive suffix drawn from masculine
noun ‘do koša’ (1;10.7)

211
[ɲɛjɛ, tubaʃa] ‘nie je toothbrush-a’ “isn’t toothbrush” Nominative suffix drawn from
feminine noun ‘kefka’ (1;10.7).

Mixing of nominals was found typical in the speech of adult bilinguals, however,
according to Ripka (1992a, p. 348) English nouns ending in a consonant were considered
masculine. In contrast, Ria maintained the grammatical gender of the Slovak equivalent by
affixing a feminine -a suffix. This usage was similar to the mixed productions of Natália in
Štefánik (2000, p. 80), who found that the gender of English borrowed nouns and
adjectives was determined mainly by the Slovak grammatical gender of the replaced
constituents. Štefánik (2000, p. 80) interprets this as the child’s general awareness of the
grammatical category of gender. Maintenance of the grammatical gender by both children
is also an indication that mixing occurred due to momentary lapse of the Slovak TEs, since
both Ria and Natália were aware of the grammatical gender of the missing Slovak
constituents.

Unusual form of mixing occurred also with the following English plural nouns:

[bɪkɪsɛ] ‘bikkies-e’ “bikkies” (1;9.12),

[ʃobsɪ] ‘sobs-i’ “soursobs” (1;10.27)

[mamɪ semsɪ] ‘mami stamps-i’ “mummy, stamps” (1;11.1)

[ɪnɛː sɔksɪ] ‘iné socks-i’ “different socks” (1;11.1)

[fɪʃɛ] [fɪʃɪ] ‘fish-e’ ‘fish-i’ “fish” (1;11.3)

[cɛtɔ paːntsɪ] ‘tieto plants-i’ “these plants” (1;11.18).

To integrate these nouns into the base Slovak language Ria applied a redundant Slovak
plural suffix to an already pluralized English noun. Although this form of mixing did not
occur regularly in the input, according to the diary records the mother occasionally used
the English words ‘bikkies’ and ‘biscuits’ and even a mixed plural form ‘biscuit-i’ (with
redundant Slovak plural suffix). Thus Ria was possibly trying to reproduce this form,
overextending it to other plural nouns. Moreover, Ripka (1992b, p. 212) found adult
speakers of second generation Slovak migrants in USA also used English plural noun
forms as singular and attached the redundant Slovak plural suffix –i.

The Slovak equivalents of the above nouns were available in Ria’s productive vocabulary,
thus the most likely explanation for mixing was temporary unavailability, such as
difficulty of retrieval. However, the fact that Ria integrated these words morphologically

212
into the Slovak base language in a Slovak context suggests that she was aware of
borrowing items from English, and of the need to assimilate them into the base language.

Ripka (1992a, p. 349) further found English verbs in adult mixed speech were inflected
according to the Slovak verb type kupovať (to buy), thus formed with the inflectional
suffix -ovať. Ria also used this suffix, however only in the present tense:

[dɪgʊjɛm] ‘dig-ujem’ “I’m digging” (1;9.19).

More frequent in Ria’s productions were verbs with the inflectional suffix –nuť, which
appeared in the past forms -ol/-la/-lo, thus maintaining gender agreement with the
grammatical gender of the agent:

[piːpɔ] ‘beep-lo’ “(car) beeped” neuter (1;10.6

[fɪnɪʃɔʊ. dædɪ.] ‘finish-ol. daddy’ “finished. daddy” masculine (1;10.8)

[fɪnɪʃa?] ‘finish-la?’ “finished?” feminine (1;11.1)

[dʒampɔʊ] ‘jump-ol’ “jumped” masculine (1;11.7).

Štefánik (2000, p. 75) also found that Natália integrated English verbs with inflectional
suffixes of several Slovak verb types and did not limit them to kupovať.

Thus Ria’s mixing and morphological integration of English lexical morphemes into
Slovak was comparable with other bilingual speakers. While nouns and verbs were
brought into Slovak language in mixed speech of both children and adult bilinguals
(Ripka, 1992a, p. 348), mixing of adjectives was found in other children (Štefánik, 2000,
p. 79). Moreover, the mixing phenomena found in Ria’s speech up to the age 2;0 were
comparable with Natália’s mixing after 2;5. Some of the differences between adult and
childhood mixing could be explained by the fact that Ripka (1992a, 1992b) examined adult
speakers of Slovak dialects from a specific migrant speech community in USA, while Ria
and Natália provide data from different sociolinguistic backgrounds.

What was further significant about Ria’s mixing was the fact that the particular mixed
lexical items were not observed as mixes in her input, thus they were Ria’s productive uses
of mixing. Indeed, Serratrice (2005, p. 162) argues young bilinguals use their languages
creatively, including mixing, and do not simply imitate input.

8.1.1.2 Mixing with the father


Table 8.3 shows that Ria addressed only a very small number of mixed utterances to the
father. Detailed analysis of the utterances revealed that Slovak lexical items in their

213
context appropriate morphological shape (with Slovak grammatical bound morphemes)
were brought into an English utterance, thus were not integrated morphologically into
English, as illustrated in the following examples:

[jɔsasa dædɪ] ‘lososa daddy’ “(I’d like) salmon daddy” (1;11.6)

[moː ɔjɛʃɔk dædɪ?] ‘more oriešok daddy?’ “(would you like) more nut(s) daddy?”
(1;11.28).

In these examples Ria used Slovak nouns in the Accusative singular as would be
appropriate in the situation for Slovak, but no grammatical suffix was required in English.
This mixing pattern into English without morphological integration was prevalent. A rare
example of grammatical mixing into English was found, which was categorized as a blend
rather than a mix, since Ria combined two lexical items, each from different language into
one word:

[dɪsanalav. dæduʃ] ‘this+one-a+lav(ička). daddush19.’ “this one bench. daddy.”


(1;10.5).

In this example Ria blended the English determiner fixed expression this+one with the
Slovak noun, however, she also regressively integrated the determiner with Slovak
grammatical gender morpheme -a in agreement with the Slovak noun. Such blending
occurred only in this example.

8.1.1.3 Mixing in triadic interactions


Mixing occurred also when Ria was addressing both parents in triadic interactions.
However, since in those instances both parents were addressed simultaneously, it is not
possible to determine the target language of all utterances according to addressee. Nor is it
possible to determine the base/guest language in two-word and all multiple-word
utterances on structural grounds (Lanza, 2004, p. 172), as illustrated by this example:

[jubɪʃ dædɪ. jubɪʃ mamɪ. jubɪʃ babos.] ‘ľúbiš daddyho. ľúbiš mami. ľúbiš babo-s.’ “(you)
love daddy. (you) love mummy. (you) love babies.” (1;8.24).

In this utterance the Slovak inflected verb suggests Slovak base language, yet the use of
other language neutral items renders it indeterminate, including a borrowing that became a
family word, often used with English grammatical morphemes.

In some multiple-word mixed utterances the likely target language could be deduced
according to the higher number of words belonging to a language within the utterance.
19
Diminutive appellative form discussed in section 8.1.2
214
Both Slovak and English target utterances were found. Analysis of these showed a pattern
of mixing where lexical morphemes from each language, shown in italics in the following
examples, were combined within a single utterance:

English utterance:

[dædɪ # maː ʌndə] ‘daddy # má under’ “daddy has (chin) under (the beard)” (1;10.6)

Slovak utterance:

[maːtoʊ ja maːm] ‘(to)mato ja mám’ “(to)mato I have” (1;11.7).

In both base language utterances the mixed lexical items were found in their appropriate
morphological shapes. Thus the mixing pattern in triadic interactions was the same as that
found in English contexts with the father - without morphological integration.

In summary, the mixing patterns with both parents were appropriate pragmatic choices.
Ria integrated context appropriate English lexical items into the grammatical framework of
base Slovak utterances. Mixing into English was lexical, since Ria carried over Slovak
lexical items in their appropriate morphological shapes (with respective Slovak
grammatical morphemes) and without morphological integration into the base English
utterance. The same form of lexical mixing was found in triadic contexts.

Thus overall, from 1;9 Ria’s mixing reflected language specific behaviour appropriate for
the language context. The unidirectional form of intra-sentential mixing mirrored Ria’s
input. Mixing of English words into Slovak was presented mostly through intra-sentential
CS in overheard speech. On the other hand, English was modelled to Ria in monolingual
contexts only. Thus when she encountered a lexical gap or temporary unavailability of an
English item, she resolved it pragmatically by relying on the Slovak equivalent in its
appropriate morphological shape.

Mixing could not be explained by dominance, because Ria mixed more into her dominant
language, Slovak. Lanza (2004, p. 136) also found the bilingual children in her study
showed unidirectional mixing in their dominant language, since mixed utterances consisted
of English lexical morphemes with Norwegian grammatical morphemes but not vice versa.
A question arises whether the directionality of mixing was only apparent due to the
contrast in the morphological structures of the two languages. When a Slovak lexical item
was mixed into English base utterance, one would expect English grammatical morphemes
only if they would be appropriate in the particular instance. Since morphology is less
salient in English, mostly bare stem forms would occur. Instead, Ria carried over the items

215
in their complete Slovak morphological shape. Thus mixing did not merely imitate input,
but it was used creatively.

8.1.2 Bilingual blends


During the two-word stage Ria used morphology creatively. She was inventing new word
forms that were frequently emotionally charged. The most productive were Slovak
diminutive suffixes, which were modelled on the mother’s frequent use of diminutives.

Ria used standard diminutive forms of the masculine and feminine appellatives in the
Slovak language, which were also modelled by the mother:

[dædɪnkɔ] ‘daddy-nko’ “daddy”

[mamɪtʃa] ‘mamička’ “mummy”

[janka] ‘Rianka’ (1;9.26).

However, the mother also applied a playful suffix -uš to Ria’s name [rɪanʊʃ] ‘Rianuš’, and
thus invented a non-standard diminutive. Occasionally Ria adopted this suffix with
appellatives for the parents:

[mamʊʃ] ‘mamuš’

[dæduʃ] ‘daddush’ (1;9.20).

Ria often wished to express affection with inanimate objects, which is regular usage in
Slovak. However, if the word she used was English, instead of relying on the English
diminutive suffix, she created blends. Hoffmann (1991, p. 59) defines blends as forms
made up of two languages, with one phonetic shape used in either language and often
dissimilar to the adult forms. Ria blended English nouns with several variants of the
Slovak diminutive feminine suffix -ina in the following declinations:

Singular Plural

Nominative -ina/-uňa -ini

Accusative -inu/-unu/-oňu

Most blends were English nouns that have a Slovak TE in the feminine grammatical
category:

[paɪkɪnʊ] ‘pikel-inu’ “pikelet” (1;9.19) – enjoying pikelets, requesting more

[baʃɪna] ‘brush-ina’ “brush” (1;9.20) pretending a small cleaning brush was a small
hairbrush

216
[hætʊnʊ] ‘hat-unu’ “hat” (1;9.26) – requesting own hat

[bukɪnʊ] ‘book-inu’ “book” (1;10.6) – requesting to read a book

[gasɪnʊ] ‘grass-inu’ “grass” (1;10.6) – requesting to go on the grass in the garden

[kɪmʊnʊ] ‘cream-unu’ “sour cream” (1;10.7) – requesting favourite food

[sokɪnɪ] ‘sock-ini’ “socks” (1;10.8) – small socks

[beɪʃʊnʊ] ‘berry-unu’ “berry” (1;10.18) – favourite food

[boʊɲa] ‘ball-ňa’ “ball” (1;11.7) – small ball

Exceptions were the following English nouns, one with a masculine Slovak TE:

[peɪcunʊ] ‘plate-unu’ “plate” (1;11.6) – small plate,

and two with neuter Slovak TEs:

[pokɪnʊ] ‘pock-inu’ “pocket” (1;10.17) – small pocket

[sɪdʊɲʊ] ‘seed-uňu’ “seed” (1;9.20) requesting more of favourite seeds

[siːconʊ] ‘seed-oňu’ “seed” (1;11.22) – requesting favourite seeds.

Similar usage of varied Slovak diminutive suffixes with English nouns was found in
Natália’s speech, amongst which suffix -ina occurred in one example: ‘camel-ina’ “little
camel” (Štefánik, 2000, p. 78). Ria limited her use to one suffix and its variants, except for
two other invented suffixes, one similar to the Slovak forms:

[babʊɲɪ] ‘bub-uni’ “bubbie” - Ria’s doll (1;9.19),

[vedɪtʃo] ‘ready-čo’ “ready-set-go line” invented while racing in the garden (1;10.6).

These invented forms, however, did not appear with any other words.

Some blends, such as [bukɪna] ‘book-ina’ “book”, were used productively, and several
forms with varied inflectional endings were found:

Nominative singular [bukɪna] ‘book-ina’ “book”

Accusative [bukɪnʊ] ‘book-inu’ “a book”

Locative [bukɪnɛ] ‘book-ine’ “(in the) book”

Nominative plural [bukɪnɪ] ‘book-ini’ “books”.

Thus Ria’s blends were English lexical morphemes with Slovak diminutive suffixes. They
were words that carried an emotional charge. It appeared that when Ria wished to refer to

217
familiar or favourite objects with an emotional charge, but preferred the English equivalent
for a concept, or was unable to recall the Slovak equivalent due to momentary lapse, she
used the English equivalent. However, since in the English language diminutive can be
expressed with limited suffixes, Ria fell back onto the Slovak diminutives instead. This
was the case especially with objects which the mother normally referred to with a
diminutive in the Slovak language. Since this type of speech was usually related to the
mother’s CDS, it was found mainly in Ria’s utterances used in interactions with the
mother.

Some blends, such as [babʊɲɪ] ‘bub-uni’ “bubbie” and [bukɪna] ‘book-ina’ “book”, were
reproduced and adopted by the parents as family words when addressing Ria, and kept
alive for a longer period (Hoffmann, 1991, p. 61). Thus it was the parental usage, the fact
that they adopted the words in their speech, which made these items a regular and accepted
part of the family’s repertoire.

8.2 The developing sense of appropriate language choice


A bilingual child needs to acquire pragmatic ability of appropriate language choice as part
of her linguistic development. She needs to learn which language is appropriate in which
situation, with which interlocutor, and how the choices are governed. Although some
psycholinguistic awareness of language differentiation is present from earliest stages
(Genesee, 2007, p. 320), it takes time to develop the functional language choice ability
further. Findings of several studies indicate language choice occurs in children younger
than 2 years (Deuchar & Quay, 2000; Genesee, et al., 1995; Nicoladis & Genesee, 1996;
Nicoladis & Secco, 2000). Some studies suggest the ability to use languages differentially
can be observed as early as the one-word and two-word stages of development (Genesee,
2006, p. 62; Paradis, Genesee, & Crago, 2011, p. 72).

Deuchar and Quay (2000, p. 108) examined the process of developing language choice
before 2 years, and found that although the bilingual child in their study was not always
able to make an appropriate choice, she was developing a sense of what an appropriate
choice was. Statistically, it was significant that the child was more likely to use an
appropriate source language of a word in a specific context. Moreover, she was more
likely to choose appropriately when she had an option to make a choice between two
translation equivalents in her lexicon. The authors conclude language choice can be
observed only when the child has sufficient lexicon of 100 or more words in her expressive

218
vocabulary (Deuchar & Quay, 2000, pp. 98-100), and it needs to be followed by sufficient
language choice practice, inclusive of errors or inappropriate choices.

In the literature relatively little analysis was undertaken on qualitative aspects of BFLA
children’s language choice. According to De Houwer (2009a, p. 238) BFLA children can
adjust their language choice to their interlocutor between the age of one and two, thus the
language of address from birth plays an important role in developing this pragmatic ability.
Determinants of language choice in children identified in most studies were (Hoffmann,
1991, pp. 89-93):

 person and location


 lexical gaps/availability
 function or purpose of interaction
 preferred language
 topic (at a later age).

Hoffmann (1991, p. 93) points out that the above determinants are interrelated to varying
degrees, and in young children are affected by social and emotional factors.

The gradual process of Ria’s developing sense of appropriate language choice will be
described in line with Deuchar & Quay’s (2000, p. 108) findings. Ria’s language choice
was analysed within specific language contexts. Language context is defined by the
language used by the speakers in the situation. It is the base language into which words or
phrases from the other language can be brought. The language of the context affects the
speakers’ ability to activate one or both of their languages, bringing them to a particular
language mode on the continuum.

Ria’s language choice was observed as soon as she started producing adult words.
Reproduced utterances were direct repetitions of the words heard in input, and as such did
not show pragmatic language choice. However, they had the function of reinforcing each
parents’ preferred language. If Ria had reproduced words in the ‘wrong’ language, it could
mean she did not show sensitivity to the interlocutor’s preferred language, or that she used
the only language available for a particular referent, regardless of the language of address.
This form of language choice was not observed in the data. Thus only spontaneous
utterances and responses to address were analysed for language choice.

During the one-word stage (1;3 - 1;7), when lexical gaps occurred Ria preferred to
maintain continuity of interaction and used the equivalent which was available. This usage
reflects previous research which suggests young BFLA children in the one-word stage go
219
through a natural phase of mixing which is caused by lexical gaps, as they rely on their
other language to supply the missing items (De Houwer, 2009a, p. 240).

If Ria had the choice between two TEs, she used the equivalent that was most readily
available in her expressive vocabulary. This does not mean she was not differentiating
pragmatically between the two languages, instead, the data point to greater importance of
social and emotional factors, since the choice between the TEs appeared more specifically
related to contexts of interaction. Qualitative analysis in the early one-word stage found
the following factors influencing language choice:

 the language context within which a word was first learnt


 Ria’s word preferences
 activity (often suggested to emerge only later in the development).

Detailed analysis of individual words in contexts indicated that social and emotional
factors determined which equivalent was available and preferred by Ria. The language for
each word was determined by the context with the most salient uses of the word in input.
The preferred equivalent was in the language of the parent who introduced the concept,
engaged in the interaction through which the word was acquired, and used it most often
subsequently. Thus some words in Ria’s productive repertoire seemed to be language
specific, even though she understood both TEs, and was able to reproduce them. Yet, she
seemed to form a preference for production of particular words based on the learning
context, as demonstrated in the following diary records:

- Ria chose the English equivalent [tiː] when referring to ‘tea’ whether she was
interacting with the parents separately or simultaneously. She understood the Slovak TE
‘čaj’, and could produce it as [ca] when elicited by the mother, or in reproductions.
However, her preference in spontaneous uses appeared to be the English member of the
TE pair.

- Ria used the Slovak TE [baːbo] for ‘baby’, even though she understood both
equivalents, and produced the TE [bebɪ] spontaneously only later.

Analysis of learning contexts of the above examples showed that ‘tea’ was introduced by
the father, since he and Ria had a morning routine of making the breakfast tea together.
The later example ‘bábo’ was introduced by the mother and used mainly in a context of a
playgroup, where other babies were present and the mother and Ria talked about them.

Thus in the early one-word stage, before Ria acquired sufficient number of words in her
repertoire, the learning context was the major determinant of language choice. Learning
220
context subsumed the person introducing the concept with a novel word, as well as the
language in which the person normally addressed her. Ria developed such preference for
several words, and used them in the preferred language although she comprehended both
TEs and responded to them appropriately. Language preference for a particular word could
change overtime as Ria acquired the TE in expressive vocabulary. Reasons for such shift
in preference are unclear. By 1;11 Ria mixed only several words for which she maintained
preference in a particular language, although she had both members of the TE pair in
comprehension, e.g. [maːtoʊ ja maːm] ‘(to)mato ja mám’ “(to)mato I have” (1;11.7),
[bejiːs] ‘berries’ and [ʊhɔdʊ] ‘uhorku’ “cucumber”. Bilingual child’s language preference
at a relatively early stage has been shown in other studies (Lanvers, 2001, p. 449).

Thus there was a short adjustment period during the initial one-word stage, until the
second half of 1;6. Towards the end of the second year learning context became less
important20 and Ria’s pragmatic ability to make an appropriate language choice was
becoming fine-tuned. The question remains how and when the shift took place from a
learning context determined language choice to a pragmatic language choice.

According to Nicoladis (1998), presence of TEs during the one word stage is not a
sufficient evidence for language differentiation. What matters is whether a child uses TEs
in contextually sensitive ways, and her choices are pragmatically appropriate according to
interlocutors.

In Ria’s productions first evidence of two languages was the presence of TEs as early as
the 50 word milestone (1;5). However, signs of emerging pragmatic language choice were
not observed until 1;7, when Ria reached approximately 250 words in expressive
vocabulary at the beginning of the combinatorial speech. Her pragmatic language choice
ability was evident from several strategies:

 Alternating languages according to addressee


 Travelling along the language mode continuum
 Bilingual utterances used in bilingual contexts - supplying words in both languages in
triadic interactions, occasionally pointing or looking at the parent to whom each word
was addressed
 Self-repair of inappropriate language choice
 Pragmatic use of TEs for clarification, and achieving desired result
 Referring to one of the languages.

20
Although preference for some words was maintained until the third year.
221
Ria was able to alternate between languages according to the addressee, thus she was able
to adjust her language use to the interlocutor’s linguistic preferences. Language choice
became more explicit in the use of TEs with the appropriate person, as in the following
examples:

Father Mother

[jeː] ‘yeah’ [anɔ] ‘ano’

[peɪ] ‘play’ [hac] ‘hrať’ (1;8.21).

Ria was also able to differentiate close cognates. She used the correct intonation and
pronunciation typical of the respective language, such as:

Slovak English Meaning

[dɛɔdɔjant] [dɪodojen] deodorant

[lɛmʊc] [liːmə] lemur

[jɔgʊt] [jougəc] yogurt

[sojɪ] [sowɪ] sorry21.

Similar observations of language specific word stress were made by Cruz-Ferreira (2006,
p. 64).

Language specific intonation, as well as grammatical differentiation, were also found in


close cognates, and shown through inflectional morpheme attached to the Slovak TE, as in
the following utterances used within the same bilingual context but each addressed to
different parent:

Utterance Addressee Grammatical morphemes

[dædɪ dɔkʊɪ] ‘daddy (k) doktorovi’ Mother -ovi Dative suffix

[dædɪ dokǝ] ‘daddy (to the) doctor’ Father -ø.

By 1;10 Ria mostly made an appropriate language choice depending on her interlocutor.
She used Slovak with the mother and English with the father, thus adjusting her language
choice according to the interlocutor’s preference. The diary records noted that although
Ria used words from both languages with both parents, she used mostly Slovak words
throughout the day when spending time with the mother alone, and considerably more
English words when the father returned from work and the language context changed from
21
Although sorry is not a Slovak word it is widely used among speakers of Slovak with Slovak
pronunciation [sɔrɪ].
222
monolingual Slovak to bilingual. Likewise, during weekends when both languages were
used in the same setting, but by different people, Ria used more English. Thus she was also
travelling along the language mode continuum from a monolingual Slovak mode to a
bilingual Slovak-English mode.

Differential use of TEs was evident in language alternation when it was necessary to
alternate addressees from the father to the mother, yet activity remained constant. Ria was
able to select the appropriate TE for each parent, as demonstrated with the following
examples:

- The father and Ria were interacting in an English context while focusing their
attention at a spider. Ria then turned to the mother and spontaneously uttered [paʊː]
‘pavúk’ in Slovak, using the appropriate Slovak TE without modelling of the word by
the mother immediately before (1;7.28);

- Ria was playing outdoors with the mother, interacting in Slovak. When the father
joined them Ria acknowledged his presence with a spontaneous holophrase in English
describing the action: [peɪŋ] ‘playing’, thus her language choice of English directly and
appropriately engaged the father (1;9.20).

Ria also showed the ability to alternate between languages on demand according to her
interlocutor. This was evident in situations when one of the parents requested her to
address the other parent with a specific request, however, without an explicit prompt for a
particular language. In response, Ria alternated the languages appropriately, as in these
examples:

*MAM Choď sa hrať s daddym.

%eng Go and play with daddy.

%act RIA walks up to DAD

*RIA [peɪ?]

%glo play?

*MAM Riana, povedz daddymu, čo doktor povedal a čo ti kontroloval!

%eng Riana, tell daddy what the doctor said and what he was checking!

*act RIA turns to DAD

*RIA [iːəs, maʊf]


223
%glo ears, mouth

(1;10)

As shown, Ria was capable of situational switching in order to address the father in the
appropriate language – English, even though the mother’s request was in Slovak and did
not specify the language to be used in addressing the father.

An important aspect of bilingual life in a mixed-lingual family is the use of languages


when the whole family is interacting together. If the family’s approach is to continue to
separate languages according to a person, these situations can become a complex series of
language choices made by all interlocutors. This aspect has not received much interest in
the research literature to date (Lanza, 2004, p. 292). Quay (2008, p. 29) addressed these
issues in a trilingual environment where the language choices are even more complex.

The audio-video data in this thesis recorded in bilingual contexts provide insights into
situations, where the use of both languages separated by person was appropriate, and
deemed more important than the use of the monolingual’s language to be all-inclusive.
Typically, such recordings comprised of triadic interactions during meal or play times.
Simultaneous interactions with both parents demonstrated negotiations of the language of
the context. The following language choices were observed in the data:

 Ria alternated her languages according to the parent she was addressing directly - the
father in English and the mother in Slovak
 Ria addressed both parents simultaneously in either language
 the mother alternated languages according to interlocutor – the father in English and
Ria in Slovak
 the mother addressed Ria and the father simultaneously in English
 the father used English.

The following example demonstrates Ria’s ability to adjust languages as appropriate


according to the addressee’s preferred language in a triadic interaction:

*MAM Aj tebe mám pokrájať?

%eng Should I cut up yours, too?

*RIA [ɲɪɛ, dædɪ, pɔkajaʊ]

%glo nie, daddy, pokrájal

%eng no, daddy, has cut (it) up

224
%add MAM

%com RIA used past tense instead of future tense

*RIA [dædɪ!]

%glo daddy!

%add DAD

*DAD Yeah, making mine (lunch).

*RIA [jaɲɪnɛ. janas]

%glo (cut) Rianine. Riana’s.

%eng (cut) Riana’s. Riana’s.

%add DAD

%com self-repair of language choice (the possessive form)

*RIA [xʊmbabɛ]

%glo chrumkavé

%eng crunchy

%add MAM

*MAM Chrumkavé by si si prosila?

%eng Would you like a crunchy one?

*RIA [mam, paɪ, tɔtɔ. xʊmbabɛ mam.]

%glo mám, páli, toto. chrumkavé mám

%eng I have, it’s hot, this. I have crunchy one

(1;11.7)

Ria was learning to interact with both bilingual and monolingual speakers in one context,
and how to adjust her language choices according to the interlocutors’ language
preferences. Triadic interactions often produced utterances with cross-linguistic synonyms
through which the parents were addressed individually, as in the following examples:

*RIA [pɪkǝʃ]

%glo pictures

%add DAD
225
*RIA [ɔbaːsɪ]

%glo obrázky

%eng pictures

%Add MAM

(1;10.8)

*RIA [mamɪ jɛtadɔ]

%glo mami lietadlo

%eng mami aeroplane

%add MAM

%act RIA turns towards the father

*RIA [ejopeɪn dædɪ]

%glo aeroplane daddy

%add DAD

(1;11.28).

Further evidence that TEs were used as synonyms were bilingual utterances containing the
TE couplet, used when Ria chose to address both parents simultaneously in triadic
interactions, as shown in the examples below:

- the father was cooking while the mother and Ria were watching, Ria made a comment
about the actions in Slovak, [vajɪː] ‘varí’ “is cooking”, and immediately after in English
[kʊ] ‘cooking’. (1;7.8)

- [mɪ mɪ mɪ mesɔ. mesɔ] ‘meat meat meat mäso. mäso.’ “meat (repeated). meat
(repeated).” (1;8.21)

- [wotə. vɔdʊ] ‘water. vodu.’ “water. water” (1;8.21)

In these situations, Ria was either simultaneously addressing both parents in their
respective languages, or using the TEs to clarify her utterances to both, thus explicitly
showing an awareness of the need to use both languages. Ria’s bilingual utterances
combined the TEs into a single unit. Their usage reflected how the parents used the

226
languages in triadic interactions when talking about an object or introducing novel words -
they provided the words simultaneously in their respective languages.

Lexical duplication of TEs in bilingual children during combinatorial speech was observed
by other authors (Štefánik, 2000, p. 56; Wanner, 1996, p. 83). Wanner (1996, p. 83)
examined the language contexts in which the English-Japanese bilingual child (aged 1;9-
1;10) used such bilingual constructions and suggested several explanations:

 The child is addressing all interlocutors in one utterance


 The child is addressing a bilingual interlocutor
 The child is exploring the alternatives to determine the appropriate language choice
 The child alternates languages to address all interlocutors in the appropriate language.

Since this child used bilingual utterances only when addressing the bilingual father
together with other interlocutors, the author concludes that he demonstrated awareness of
the two languages, he used the TE items as synonyms, and associated the bilingual father
with both languages.

Although Ria also used the TEs as synonyms, the bilingual utterances occurred only when
she needed to address both parents in triadic interactions. This indicates Ria’s bilingual
awareness was well established by the two-word stage (1;8). Moreover, at a later stage Ria
explicitly alternated turning towards the parent she was addressing within the bilingual
utterance, as in the following examples:

*DAD You have cold hands.

*RIA [dʒampə]

%glo jumper

%add DAD

%act RIA turns towards MAM

*RIA [sɛtɪːk]

%glo svetrík

%eng jumper

%add MAM

(1;10.28)

227
%act RIA turns to DAD

*RIA [kɪs]

%glo kiss

%act RIA turns to MAM

*RIA [pʊsʊ]

%glo pusu

%eng kiss

(1;11.7).

Similar to the bilingual utterances were other playful uses of language. Ria explicitly
practiced certain conceptual and symbolic items she acquired as cross-linguistic synonyms.
For example when she was able to count to 2 in both languages, she playfully practiced
this skill, as in the following example:

Ria counted to two on her fingers verbalizing the numbers [jɛdɛn # da] ‘jeden # dva’
and after a short pause repeated the same actions with the English TEs [wan # tu] ‘one #
two’, smiling at the mother (1;9.12).

Thus she was aware of the two languages available as language tools, aware of her
knowledge of the TEs, and aware of the option to select from the TE pair.

8.2.1 Repairing language choice


The developing sense of appropriate language choice was also evident from the end of the
one-word stage in Ria’s ability to interpret communication breakdowns and parents’
requests for expansion or clarification of utterances which they did not understand. A cue
that a communication breakdown occurred was that the parents either did not respond
according to Ria’s expectations, or they explicitly let her know they did not understand.
Ria was able to interpret such cues as errors in language choice when indeed an error
occurred, and was able to self-repair. This pragmatic ability emerged in both language
contexts simultaneously, and included instances of explicit self-repair, as in the following
example:

Ria uttered [wɔm] ‘von’ “outside” while she was in the father’s arms (mother also
present) as they were looking outside through glass door, but the father did not respond
to her request to go outside. Ria self-repaired her language choice uttering a novel TE

228
[ʃaː] ‘outside’, to which the father responded. She acknowledged his response with a
smile (1;6.31).

In this instance the father did not react to Ria’s attempt at communication. She interpreted
the breakdown in communication as inappropriate language choice and self-repaired with
the English TE. The repair gained the father’s attention, and thus the strategy was
reinforced as effective.

Ria was able to self-repair language choice before the parents indicated a ‘wrong’ choice,
which sometimes resulted in an utterance with cross-linguistic synonyms, yet different to
the bilingual utterances discussed above:

Slovak context

- the mother picked up Ria and asked where she wished to go, to which Ria responded
with a pointing gesture indicating the direction and uttering [hiː] ‘here’ and immediately
self-repairing to the Slovak TE [sɛm] (1;7.2)
- addressing the mother Ria self-repaired a mix within the same utterance
[gɪanɪː. zɛjɛnɪː] ‘green-aný22. zelený’ “green. green” (1;9.26).

English context

- Ria was playing with the father in the garden, while carrying around a doll. She
directed the father’s attention to the doll uttering [baːbɔ] ‘bábo’ “baby” and
immediately self-repaired to English TE [beɪbɪ] ‘baby’. (1;7.1)

- addressing the father with a self-repaired utterance [ʊmʊ. woʃ.] ‘umyť. wash.’ “wash.
wash.” (1;8.21)

- [dædɪ. tʃaɪ! tʃaɪ tʃaɪ tiː.] ‘daddy. čaj! čaj čaj tea.’ “daddy. tea! tea tea tea.” (1;9.4).

At a later stage such self-repairs occurred mid-word, suggesting that bilingual language
awareness was well established:

Ria addressed the father with utterance [manɟɛ, ma aːmns] ‘mandle, ma(ndle) almonds’
“almonds, al(monds) almonds”, in which she realized error in language choice mid-
word, paused and self-repaired to English. (1;10.6).

Ria was also able to repair errors in language choice if one of the parents requested a
clarification because her original utterance was in the ‘wrong’ language. Thus Ria

22
A mix discussed in section 8.1.1.1.
229
understood the breakdown in communication which triggered the request as signifying a
pragmatic error in language choice, rather than a linguistic error.

During the period of combinatorial speech many novel words emerged and at times were
not immediately recognized by the parents, thus a communication breakdown occurred. In
these situations Ria’s repair strategies to clarify utterances reflected her pragmatic
development. Initially she pointed at the referent, later attempted to repeat the word several
times, making it explicit that the repetition was for the benefit of the interlocutors, and
lastly intentionally switched into the other language to clarify the meaning:

 Repetition of the novel word – sometimes accompanied by a hand gesture with the
index finger raised upwards and the hand waving lightly. If the parents understood, Ria
smiled and often repeated the word again, or simply affirmed with [anɔ] ‘ano’ “yes”,
e.g. Ria uttered [daːmɛ kabɪtʃʊ] ‘dáme kávičku’ “we’ll have a coffee”, but the mother
misinterpreted the word as [kabɪtsʊ] ‘krabicu’ “a box”, thus Ria continued repeating
until the mother mirrored the correct word (1;10.6).
 Use of TE - falling back onto a TE from the other language to explain, aware that this
strategy would allow the communication flowing:
e.g. Ria uttered [ʃatɪc] ‘vtáčik’ (a new way of pronouncing an otherwise well-
established word [ʃaː]) but since the mother did not understand, Ria used the English
TE [bədiː] ‘birdie’ to clarify (1;10.6).
 Attempt to use associations – creating the same meaning with different words,
e.g. when uttering [piːpɔ] ‘beep-lo23’ “beeped” and the mother did not understand after
several repetitions, Ria turned her gaze to a car and uttered [atɔ] ‘auto’ “car”, which the
mother interpreted as ‘the car beeped’ (1;10.6).

Thus pragmatic language choice was evident also when the parents did not respond to Ria
as desired, because they did not recognize a novel word in her production. To
communicate her intentions, Ria often opted to borrow the established TE in order to
clarify her request. She was able to fall back on an equivalent from the other language,
thus used the TEs pragmatically:

Ria uttered the English holophrase [be] ‘bread’ to request more bread, but since the
parents did not respond immediately, she switched to the Slovak [xɛ] ‘chlieb’ “bread”
(1;7.8), an equivalent that was already established, and gained the desired response.

23
Slovak past tense 3sg suffix attached to English verb
230
TEs were available also in instances when the parents requested a clarification of a novel
word in Ria’s production. This occurred especially in the Slovak context. Ria used the
English TE to clarify the meanings, as in the following excerpts:

%sit During bed time Ria was naming familiar people and animals

*RIA [mamɪ. sɪp. # dædɪ. sɪp.# jana. sɪp.# mɪʃa. sɪp.# ɔmɪ. sɪp. # naːʊ. sɪp. # ɦavo.
sɪp. # ɦajaː.]

%glo mami. sleep # daddy. sleep. # Riana. sleep. #Miša. sleep. # Romi. sleep. #
mňáu. sleep. # havo. sleep. # hajá.

%eng mummy sleeping. Daddy sleeping. Riana sleeping. Miša sleeping. Romi
sleeping. Meow sleeping. Doggie sleeping. sleeping.

*MAM Čo je hajá?

%eng What’s hajá?

*RIA [hajaː. sɪp.]

%glo hajá. sleep.

%eng sleeping.

(1;7.17)

*RIA [bɔjɪː]

%glo bolí

%eng it hurts

*MAM Čo ťa bolí?

%eng What is hurting you?

*RIA [ʊcɪca]

%glo ručička

%eng hand (diminutive)

*MAM Čo to je?

%eng What is it?

*RIA [hæn]

231
%glo hand

(1;8.24)

*MAM Ako sa budeš hrať s daddym?

%eng What will you play with daddy?

*RIA [jɔːpɪ]

%glo lopty

%eng balls

%com MAM does not understand

*MAM Čo to je?

%eng What’s that?

*RIA [boʊs]

%glo balls

(1;9.19)

These instances show Ria’s awareness of the mother’s bilingualism, since she used an
English TE to clarify the meaning, expecting the mother to understand. Clarification of a
novel word with an already established TE occurred in the other direction as well:

%act MAM is cleaning up after dinner

*RIA [peɪc jaɲɪn mamɪ]

%glo plate Rianin mami

%eng plate Riana’s mummy

*MAM Preč Rianin?

%eng Away Riana’s?

%act RIA repeats her utterance several times but MAM does not understand

*RIA [mɪsʊ jaɲɪnʊ]

%glo misku Rianinu

%eng bowl Riana’s

232
(1;11.28)

Ria’s use of a novel English word was not interpreted correctly by the mother and
prompted first a repetition of the utterance, then change of language. However to clarify
her initial utterance, Ria used an established Slovak word with a slightly different
meaning.

Moreover, Ria was able to realize a language choice error in the Slovak context even
though the mother did not. In some instances the mother was not aware that Ria used an
English utterance and instead judged it as a novel Slovak word requesting a clarification:

*RIA [vaʃɪm]

%glo washing

*MAM Nerozumiem ti anjelik, čo to hovoríš?

%eng I can’t understand you angel, what are you saying?

*RIA [vaʃɪm vaʃɪm pac!]

%glo washing washing prať!

%eng washing washing to do the washing!

(1;8.31)

Such strategies reflected communication patterns used by the mother. The mother
attempted to interpret Ria’s intended meanings explicitly, mirroring her utterances, so as to
show that Ria’s attempts at making meaning were not only acknowledged, but her
meanings were understood. The mother tried responding with what she perceived to be the
same word Ria had used, tried alternate words, or explicitly claimed she did not
understand. When Ria was capable of achieving the same goal herself, she used similar
strategies to express the intended meaning - repetition, TEs, and associations. Thus if the
parents’ responses suggested a communication breakdown Ria was trying to help them
understand her utterances.

Unless Ria used a novel word, the mother understood English borrowings and her
strategies did not require a request for clarification. She did not pretend not to understand,
but allowed Ria to perceive her bilingualism, whilst modelling her preferred language with
bilingual strategies:

 moving on and continuing in the interaction


 accepting Ria’s English utterance and recasting it into Slovak.
233
These move-on and recasting strategies are shown in the following video excerpts:

%sit RIA and MAM are looking at objects in the kitchen

*RIA [noʃ]

%glo nôž

%eng knife

*MAM Nôž, ano.

*RIA [dædɪ]

%glo daddy

*MAM Daddy s ním krája, však?

%gpx RIA points at another knife

*RIA [dɪs?]

%glo this?

*MAM Aj to je nôž.

%eng That’s also a knife.

(1;7.19)

%sit RIA hides behind play tent

*RIA [hajdɪŋ]

%glo hiding

*MAM Schováš sa?

%eng Are you going to hide?

*RIA [xobaːbam. xobabam]

%glo schovávam. schovávam.

%eng am hiding. am hiding.

*MAM Schovávaš?

%eng You hiding?

(1;11.18)

234
The monolingual father’s strategies differed slightly. He often made language choice error
explicit, since he only understood a limited number of Slovak words. Ria was able to
interpret the father’s strategies through which he let her know that a breakdown in
communication occurred:

 querying the meaning of a specific word,


 probing for what he would say himself, such as ‘What does daddy say?’ or ‘In
English!’.

The father’s strategies are demonstrated in the following examples:

*RIA [kɛsɪc]

%glo kresliť

%eng (I want) to draw

*DAD What’s that mean?

*RIA [doːwɪ]

%glo drawing

%gpx RIA turns to the mother

*RIA [kɛsɪm]

%glo kreslím

%eng I’m drawing

(1;9.19)

*RIA [jabʊkaː]

%glo jablká

%eng apples

%add DAD

*DAD What are they?

*RIA [epus.]

%glo apples

(1;11.4)

235
*RIA [ɔtɔɪc]

%glo otvoriť

%eng to open

*DAD What does that mean?

*RIA [oʊpən]

%glo open

(1;11.6)

The following example demonstrates the father’s request for a specific language and Ria’s
subsequent repair:

*DAD What’s this?

*RIA [dakiː]

%glo duckie

*DAD What colour is it?

*RIA [ʒʊtɛː]

%glo žlté

%eng yellow

%com Ria used the correct neuter suffix agreeing with the neuter gender of the
Slovak equivalent ‘kačiatko’

*DAD In English?

*RIA [jejoʊ]

%glo yellow

(1;10.17)

One of the father’s strategies in dealing with a language he did not speak but which was
regularly used in the family environment, was to attempt to infer meanings from the
context and respond to Slovak utterances in English. Occasionally his inferences were not
correct, which implicitly indicated to Ria that an error in language choice occurred, and
reinforced his identity as a monolingual English speaker. Ria was also able to self-repair,
as well as to repair the father’s interpretation errors, as in the following examples:
236
*RIA [tʃɪːtac]

%glo čítať

%eng (I want to) read

%add DAD

*DAD Do you need to do a wee?

%com DAD misinterpreted the utterance as [tsɪkac] ‘cikať’ “to wee”

*RIA [jiːd]

%glo read

(1;8.21)

*MAM Idem ti zobrať ponožky.

%eng I’m going to bring your socks.

*DAD Mami’s gonna bring your pants.

%com DAD infers the meaning incorrectly

*RIA [ʃoːks]

%glo socks

*DAD Shorts?

*RIA [ʃoks]

%glo socks

*DAD Daddy doesn’t understand, does he?

*RIA [noʊ]

%glo no

(1;10.8)

There were also instances when a communication breakdown did not occur even though
Ria made an inappropriate language choice in interaction with the father, as was the case
with some Slovak words in Ria’s productions that he became familiar with. Since there
was no need to request a clarification, the father used the ‘move on’ strategy, as in the
following example:
237
%sit RIA and DAD are reading books

*DAD You read daddy that book.

*RIA [dædɪ]

%glo daddy

*DAD Yep, tell me.

*RIA [ʃɪ. baːbɪ?]

%glo see. barbie (BBQ).

*DAD What’s there?

*RIA [baːbɪ!]

%glo barbie.

*DAD You looking for the barbie, yeah?

*RIA [dædɪ! dædɪ!]

%glo daddy! daddy!

*DAD Where is it?

*RIA [baːbɪ. ɲɛɲɪ. hm.]

%glo barbie. nie je. hm.

%eng barbie. isn’t. hm.

*DAD Yes it is. You just have to turn the page slowly. Slow. That, that’s too
many. Better. Any barbies?

%com Ria used a mixed utterance containing Slovak word to which the father
responded in English

*RIA [na]

%glo nah

*DAD What about here, no barbie? Nah. Nah.

(1;8.9)

The father also adopted and regularly used several Slovak words which Ria preferred, thus
making them family words, as in the following excerpt:

%sit DAD
238
*RIA [ʃi. ababo. ʃi.]

%glo see. a bábo. see.

%eng see. a baby. see.

*DAD Uhm.

%act RIA picks up the doll and puts it in her lap, looking in its face

*RIA [luk. babo. babo]

%glo look. bábo. bábo.

%eng look. baby. baby.

*DAD Bábo’s looking at you.

%eng Baby’s looking at you.

%com DAD repeats the Slovak word in his response

*RIA [je]

%glo yeah

*DAD Hello mami Riana!

%com DAD uses a high pitched voice, impersonating the doll

(1;8.9)

Thus by accepting and using some Slovak words the father allowed Ria to perceive his
acceptance and support for Slovak. However, by clearly marking instances of language
choice error that caused a communication breakdown, the father emphasized his
monolingual identity.

The discourse strategies used by Ria reflected the use of the two languages in the family.
In using these strategies Ria was behaving like a bilingual in the same way she observed
the mother to model bilingual behaviour. She was providing translations into English for
the monolingual father when inappropriate language choice occurred. Since the mother
accepted utterances in both languages, and requested an explanation only for novel words,
Ria did not need to use translations to clarify meanings. However, even in these situations
she fell back onto a known and established TE in English to provide explanation to the
mother, acknowledging the mother’s bilingualism. More importantly, Ria differentiated
between these strategies, and understood the parental request as either miscommunication
due to inappropriate language or miscommunication due to a novel word.
239
In summary, the above analysis indicates that the bilingual skill of pragmatic language
choice gradually emerged from an internal awareness of two input languages. Ria was
sensitive to her interlocutors’ preferred languages early on, yet in the one-word stage
language choice was influenced by Ria’s linguistic abilities and preferences. The
pragmatic ability of appropriate language choice became evident at the beginning of the
two-word stage (1;7), approximately 5 months before 2;0. TEs were not used as isolated
items, but Ria showed understanding of their synonymy, and made mostly appropriate
language choice with individual interlocutors. Holophrases and word combinations were
becoming differentiated according to the interlocutor’s preferred language. Use of
unilingual utterances was established. Thus Ria’s awareness of the choice between two
languages was evident on contextual basis. When errors in language choice occurred, she
used cues of what constituted appropriate language choice from parental discourse
strategies. Therefore parental implicit or explicit feedback on Ria’s choices gradually
shaped her pragmatic ability to make an appropriate language choice.

8.3 The developing sense of bilingualism


The question of language differentiation, the child’s ability to separate the two languages-
in-acquisition, has been fundamental in the childhood bilingualism literature. Much of the
discussion has centred on the issue of unitary language development hypothesis versus the
separate development hypothesis. Latest research found empirical evidence in favour of
early language differentiation in BFLA children, whose languages develop as two separate
linguistic systems. As Hoffmann (1991, p. 79) argues, if we assume this view of bilingual
development as a starting point, the concepts of separation and differentiation become
irrelevant, since a BFLA child operates with two separate systems from birth. However,
one aspect remains unexplained – pragmatic differentiation (Nicoladis, 1998, p. 114).
Nicoladis (1998) points out we need to gain understanding on how children develop
sensitivity to two input languages before second birthday, an approximate age when
language differentiation emerges and bilingual children start using their two languages in
monolingual fashion. Pragmatic differentiation is defined as social rather than neurological
ability to differentiate the two languages-in-acquisition (Nicoladis & Genesee, 1996, p.
440).

Data in this thesis suggested that language differentiation is not a point in development,
but rather a developmental process which starts at birth. A bilingual child first develops
awareness of two input languages. Advancements in neurolinguistics, which provide an
understanding of newborns’ and young infants’ perception skills, demonstrated that
240
linguistic development starts in-utero, and human babies are born equipped with the skills
necessary not only to recognize their native tongue, but to distinguish it from other
languages (de Boysson-Bardies, 2001, pp. 22-26). These skills have also been shown in
bilingual newborns, suggesting that BFLA infants possess representation of two input
languages as early as birth (Petitto, et al., 2001, p. 491). Other studies indicate that
bilingual infants are able to distinguish the presence of two input languages as early as 4
months (Bosch & Sebastián-Gallés, 2001, p. 45). Taken together, these findings show that
awareness of bilingual input starts developing at birth, and infants have the capacity to
differentiate languages perceptually. Words and speech have different inherent sound
characteristics in each language, which allow infants to distinguish between them on the
basis of prosody from early stages. What infants need to learn is how to use the languages
differentially in production. Thus the question should not be whether a BFLA child is able
to differentiate her languages, but how does perceptual differentiation on the basis of
prosody lead to an eventual differentiation in the child’s speech.

A child learns how languages are used from the linguistic models. Regular parental use of
both input languages plays an important role in establishing bilingual awareness during the
first 9 months. During this period, prelinguistic infants can already show signs of
sensitivity to two input languages. For example Cruz-Ferreira (2006, p. 62) attributed
strange behaviour of 4 month old infants when their mother spoke a different language
than the language of usual address to early signs of bilingual awareness. Similar
observations were made in this thesis. Around 0;4-0;5 Ria responded with fussiness when
the mother was addressing other interlocutors in English, suggesting that the mother’s
switch of languages triggered strange behaviour, as Ria was aware that the mother’s
attention was focused on others.

Further in development, when Ria progressed into intention reading communication (0;9 –
1;1), her comprehension of translation equivalents was observed in appropriate responses
to verbal address by each parent. Direct contextual evidence of learning of couplets in a
TE pair was also found. Use of TEs was also extended into production in the early one-
word stage (1;4) and indicated the developing sense of bilingual awareness.

Bilingual awareness emerged explicitly during combinatorial speech when Ria used
mostly unilingual utterances in the two languages. Further evidence in production was the
differential use of language specific morphology with words from the two languages. Ria
rarely used Slovak morphology with English words and vice versa, except for specific
examples belonging to distinct categories of mixing: learning context, lexical gaps, word
241
preference, temporary unavailability of items and creative uses of the languages in the
family. Thus mixing was not seen as a sign of lack of differentiation. On the contrary, it
was a reflection of Ria’s pragmatic competence. Mixing mirrored language use in the input
and was used productively. If Ria faced a situation in which mixing was not normally
modelled, she mixed creatively to avoid communication breakdowns.

Bilingual awareness was evident also in development of pragmatic language choice. In the
second half of the second year Ria was able to appropriately adjust language choice, as
well as spontaneously choose the appropriate language when interacting with speakers of
the two languages. She was able to repair language choice errors both spontaneously, as
well as following parental cues that an error occurred. Thus the course to pragmatic
differentiation for Ria hinged on grasping the principle that the family functioned with two
languages, and becoming attuned to the language preferences of her interlocutors.

Language differentiation therefore cannot be separated from linguistic environment. The


amount and type of exposure to the two input languages provided not only the models of
the two linguistic systems, but also models of how the two languages interacted. Such
models were family specific and provided Ria with the building blocks for appropriate use.
Ria based her productions in the two languages on observations of bilingualism in the
input. Thus the way the parents used the languages, and how they modelled what was
perceived appropriate, was important.

Several studies suggested that bilingual infants, when exposed to regular input in both
languages, come to associate them with the people who normally address them in those
languages, and that it is this association that makes bilingualism possible (Barron-
Hauwaert, 2004, pp. 5-7). However, the data in this thesis indicate a reversed language
association pattern for the process in which bilingual awareness and differentiation came
to Ria. She appeared to have sensitivity to the existence of two input languages early on
and became accustomed to different communication styles with each parent, much like she
became accustomed to the parents’ different parenting styles. From this early sensitivity
she developed a more specific awareness that the difference came from parental preference
to different input languages.

Signs of emerging language differentiation can be assessed also on the basis of phonetic
development. Cruz-Ferreira (2006, p. 63) observed clear signs of emerging language
differentiation in her subjects in the prelinguistic period (aged from 0;9 to 1;1). The
children appeared to use ‘language-specific-connected-speech routines’, which replicated

242
phonetic, rhythmical and intonational patterns of Swedish and Portuguese, and which were
used in their babbling directed to a particular speaker.

In Ria’s case, phonetic differentiation was not evident in early vocal development (cooing
and canonical babbling), nor did protolanguage show any specific features of either of the
languages. Language specific babbling was observed only when Ria became linguistically
productive. Along with first words she continued to use babbling which resembled
intonational patterns of the languages. This variance from Cruz-Ferreira’s (2006, p. 63)
findings suggests that whether a BFLA child shows early signs of phonetic differentiation
is idiosyncratic, and possibly dependent on the differences in the prosody of the two
languages.

Overt signs of the developing phonetic differentiation were observed during the one word
stage (from 1;4) in the form of language specific word stress and intonation patterns.
Words such as [dædɪ] ‘daddy’ and [kuːl] ‘cool’ were pronounced with English intonation,
and [mamɪ] ‘mami’ “mummy” with Slovak intonation (1;4.28). Such patterns were
observed both in spontaneous productions and reproduced words which Ria attempted to
pronounce, e.g. Ria reproduced English word ‘cards’ after the father as [kaː] with a rising
intonation, while moments later she reproduced the Slovak word ‘kabátik’ “coat” as [ka]
after the mother with an even intonation typical of the Slovak language (1;4.29). Thus
although structurally the forms [kaː] and [ka] appear rather similar, it was not only the
difference in the vowel length, but mainly the intonational patterns that set them apart.
This difference in the pronunciation of words showed Ria’s awareness of the need to
differentiate between the two languages on phonetic level, as well as semantically.

Language differentiation was also observed in metalinguistic behaviour from


approximately 1;10. It was evident through differential use of TEs, use of bilingual
utterances and various strategies in making choices, such as self-repair of inappropriate
language choices, appropriate interpretation of errors of language choice and repair.
Another sign was the emerging ability to respond to language labels appropriately and to
explicitly talk about the two languages in her input, such as labelling or commenting on
them. This discourse was learnt from the father, who labelled the languages when making
his lack of understanding explicit, requesting a clarification, and in contexts when Ria
requested to read Slovak books. The father often responded by explaining he did not
understand Slovak, referring to the language with its name, which Ria mirrored in her
speech:

%act RIA passes a Slovak book to DAD, requesting him to read


243
*DAD Daddy can’t read this book.

*RIA [ʃɔbɛk]

%glo Slovak

(1;10.23).

Thus it was the father’s discourse which triggered expression of metalinguistic awareness
before 2;0. Clyne (1987, p. 103) also found that metalinguistic awareness emerged in a
bilingual child around second birthday.

In summary, the above signs of emerging bilingualism suggest that having received
consistent input in two languages from birth, Ria was attuned to the preferred use of the
two languages in her environment. She developed sensitivity to bilingual input and
sensitivity to bilingual language choice early on, before the age of 2;0. The course of
differentiation was not considered a milestone or target that Ria had to achieve in order to
become a proficient speaker of the two languages. Instead, language differentiation was a
part of a package of linguistic abilities that were necessary during the course of linguistic
development. Throughout this course linguistic environment in the family provided the
model of appropriate language use. Language differentiation thus developed from Ria’s
early bilingual awareness and instinctive perception of two input languages. Ria was able
to make the transition from perceptive differentiation to explicit differentiation in
production, emerging in all areas of linguistic development, such as lexical, grammatical,
phonetic, pragmatic and metalinguistic by 1;7.

244
9 Conclusion
This case study portrayed a child’s emerging bilingualism from birth to two years in a
mixed-lingual family, taking into account the major linguistic developments from a
sociolinguistic perspective. Bilingual first language acquisition of two morphologically
different languages, Slovak and English, was examined through systematic analyses of
cumulative vocabulary, combinatorial speech, early morpho-syntax, mixing and
developing pragmatic skills of language choice, as well as an assessment of the overall
language learning environment and parental discourse strategies.

The child in this study, Ria, received regular consistent exposure to Slovak and English
from birth. The two languages were presented in the one parent-one language (1P/1L)
approach. The maternal Slovak language was the minority language, however, for Ria it
was the language of the primary sociolinguistic group – the language of the closest
emotional bond and most frequent exposure. The paternal majority English language was
spoken by the wider community. The father was a monolingual English speaker, while the
mother was a bilingual Slovak-English speaker. To date this language combination in
BFLA was relatively unexplored with only one other study on simultaneous bilingualism
carried out in a Slovak majority environment, where English was the minority language
(Štefánik, 2000).

The following research questions were examined:

1. How did the language learning environment influence Ria’s bilingual linguistic
development?
2. What was the course of bilingual linguistic development?
3. What was Ria’s lexical and morpho-syntactic development in the two languages?
4. When and how did Ria start using her languages in contextually appropriate ways?
In response to the first question, the data provide evidence that early establishment of the
linguistic bond, and thus the ‘default language mode’, was crucial in maintaining
consistency of separate language input and language choice in interactions. As a result of
the person-language bond, Ria came to expect a particular language from a particular
person (Slovak from the mother, English from the father). This was especially relevant for
the minority mother, since regular monolingual exposure to the minority language was
limited mainly to her. The monolingual majority father, on the other hand, could afford to
allow both languages in interactions with Ria, since there were multiple sources of
monolingual exposure to his language.
245
Although Ria was growing up as a bilingual in an English majority society, the minority
Slovak language, the language of the closest emotional bond, and thus the language of the
primary sociolinguistic group, became the dominant language in the first two years of life.
These findings indicate that establishment of bilingualism on family level was the result of
a combination of several factors, including the person-language bond, parental language
attitudes and impact belief, discourse strategies which negotiated use of preferred parental
language in a manner which encouraged maintenance of verbal interaction, and child-
centered conversational strategies of the minority parent.

In answer to the second question, Ria’s developmental patterns in the two languages were
appearing on the same trajectory as is typical for monolingual children speaking the
respective languages. Word comprehension preceded production. In the one-word stage,
the 10, 30 and 50 word milestone progressions were evident. Ria comprehended and
produced lexical items from both languages and used translation equivalents. Once she
reached approximately 200 words in the combined total expressive vocabulary, two-word
and multiple-word combinations as well as morpho-syntax appeared. Thus no delays or
differences in developmental stages were found.

The third question sought to examine the course of lexical and morpho-syntactic
developments. In the assessment of bilingual vocabulary a redevelopment of Pearson’s
(1998, p. 358) measurement was suggested. All productions across both languages were
considered, and homonyms and appellatives were not excluded, thus Ria’s competence
was not downplayed and possible phonological differentiation of otherwise similar items
was not neglected.

This thesis also provided insights into the learning and use of TEs in early BFLA. To date,
little evidence is available on comprehension and production of early TEs in the literature.
Analysis of lexical development described how Ria learnt TEs in comprehension and
production and several mechanisms were identified:

 Learning each equivalent separately in different contexts


 Learning TEs in association with a known equivalent
 Learning TEs simultaneously in a bilingual context.

These mechanisms explained why some items appeared in the vocabulary as TE couplets,
and how lexical gaps were filled. Initial vocabulary development was context based and
depended on the type of situations each parent engaged in with Ria. If the parents noticed
that Ria had used a new word in one of the languages, they introduced the equivalent. Thus

246
Ria was able to learn the TEs with only short time lags, and often had both TEs in
comprehension. However, she also formed a short term preference for some words in
production which was determined by the initial learning context.

As soon as Ria was able to produce full words in the early two-word stage, she used
morphological markers appropriately. Words in Slovak, the morphologically richer
language, showed the necessary markers (verb inflections, noun, pronoun and adjective
declinations, subject-verb agreement). English morphological markers also appeared
according to patterns found in English monolingual children, with plural and possessive
nominal markers first, followed by verb inflections. More importantly, Ria used words
with markers appropriate for the contexts.

In verb morphology items appeared first in the form which was used by the parents most
often, while other forms were filled in later. This pattern was found in verb tense, mood
and aspect equally. Thus learning of verb morphology was item-based, and Ria was slowly
filling in the verb paradigm according to usage in the input. Slovak verbs were used
productively in the singular towards the end of the stage, with subject-verb agreement in
the inflected forms. Several tokens of plural forms were also used but tied to specific
contexts. In English, continuous –ing form was used when appropriate, and subject-verb
agreement in 3sg was emerging as well. Verb constructions were first learnt as wholes and
produced as a single unit.

The last question addressed the issue of developing bilingualism. Sensitivity to two input
languages was observed during infancy from Ria’s emotional reactions when overhearing
the mother using the ‘wrong’ language, as well as through the calming effect of the
familiar rhythms and prosody of the language. The data showed lexical differentiation in
comprehension and production through presence of TEs. Ria understood and used TEs
differentially as cross-linguistic synonyms, she did not reject them, and the Principle of
Contrast was not supported. Thus previous findings that bilingual children accept cross-
linguistic synonyms in receptive vocabulary (De Houwer, et al., 2006, p. 344), as well as
in expressive vocabulary (Deuchar & Quay, 2000, p. 57) were reflected in the data.
Grammatical language differentiation was also evident towards the end of the one-word
stage through differential use of morphological markers necessary in the two languages,
which was taken as evidence for separate development of the grammars.

While the two languages were developing autonomously, Ria was not developing as two
monolinguals in one. Language contact in bilingual interactions produced language
mixing, which was occasionally observed in the mother’s speech, flagged in the father’s
247
speech, and transferred into Ria’s productions as well. Initial mixing in the one-word stage
was lexical and it was explained by specific psycholinguistic and social factors. Ria
developed preference for certain words in one language and used that equivalent despite
the availability of both TEs in her repertoire. Her preference for certain words was
determined by the learning context. Other explanations for mixing were in line with
previous research, such as pragmatic strategy to fill lexical gaps to avoid communication
breakdowns, and momentary lapse.

However, as soon as combinatorial speech emerged, instances of intra-sentential mixing


were also observed in Ria’s productions. Mixing in the two languages appeared to be
differentiated, showing Ria’s sensitivity to the grammar of the base language. The
grammar was not violated. Ria did not mix Slovak lexical morphemes with English bound
morphemes, but mixed English lexical morphemes with Slovak bound morphemes in a
way that reflected adult code-mixing, integrating them morphologically and phonetically
into the base language. Intrasentential mixing was productive, much like code-mixing
typical of older children and adult bilinguals. Ria was able to produce mixed word
structures creatively in ways not observed in her input. She created bilingual blends which
she used productively. Moreover, mixing was used in contextually appropriate ways, with
more mixing found in Slovak and bilingual contexts when interacting with the bilingual
mother, and rare uses of mixing in English context with the monolingual father.

Emerging pragmatic differentiation was demonstrated through the emerging skill of


appropriate language choice towards the end of the one-word stage, since Ria addressed
each parent mostly in their preferred language. Thus Ria’s early pragmatic language choice
skills were identified in line with previous research as following:

 use the two languages in contextually appropriate ways - Slovak with the mother and
English with the father
 use mixing in triadic interactions with both parents and in interactions with the
bilingual mother, but rarely with the monolingual father
 use mixing as a pragmatic strategy to fill in a word unavailable due to lexical gaps,
temporary lapse, preference, or to clarify a novel word
 use mixing creatively.

Overall, Ria used more of the appropriate language with each parent. She took clues on
appropriate language use from her input. Although the general input pattern was based on
the 1P/1L approach, careful examination of sociolinguistic aspects found ample language

248
contact in the bilingual family. Ria observed the mother as a bilingual who needed to use
two languages in everyday interactions, and the father as a monolingual needing one
language. Thus the parents were modelling bilingual and monolingual behaviour
respectively. This was evident especially in situations when the entire family was
interacting together. Parental language choices in triadic interactions were providing Ria
with the model of appropriate language use as expected in the family.

The family’s expectations were communicated to Ria through a continuum of parental


discourse strategies. Each parent developed her/his own strategies. The mother modelled
monolingual uses of Slovak, which was aimed at providing as much monolingual input to
Ria as possible. However, she modelled bilingual linguistic behaviour through language
strategies that accepted English items, and through language alternation when interacting
with the father and Ria in triadic interactions. The monolingual father’s strategies were not
by default monolingual. He was willing to accept bilingual mixing in his interactions with
Ria, however, his language proficiency modelled monolingual uses of English. Ria
interpreted parental cues for inappropriate language choice appropriately and self-repaired.
Likewise, when communication breakdowns occurred due to a novel word, Ria inferred
the cause of the breakdown appropriately and provided a clarification, sometimes
pragmatically falling back onto a known equivalent in the other language.

The findings in this thesis on parental discourse strategies are parallel with previous
research, namely that strategies that negotiated preferred language played an important
role in teaching appropriate language choice. Lanza (2001a, p. 225) concludes parental
discourse strategies are one of the contributing factors in a child’s language separation
according to interlocutor. In this thesis, I showed that such strategies indeed enforced the
preferred language of communication within each parent-child dyad. However, the success
of the strategies was based on the nature of the existent linguistic bond, which set the
normal language mode for each dyad.

This study contributed to the understanding of BFLA, as well-as to the area of cross-
linguistic developmental research by comparing two morphologically different languages.
It described specific bilingual phenomena that resulted from the language combination and
bilingual language socialization in the family:

 signs of sensitivity to different input languages appeared in infancy

249
 canonical babbling showed no language specificity, however, at the same time as first
words were emerging, babbling changed and showed prosody characteristic of the two
input languages, used freely and not tied to a specific context
 translation equivalents were used as soon as the 10 word milestone was reached, and
their number increased with growing expressive vocabulary
 learning of new word types was embedded in the situational contexts engaged in with
each parent, providing an explanation for differences in vocabulary in the two
languages and for lexical gaps
 in verb morphology an initial lead-lag pattern emerged whereby the proportion of
inflected verb forms in Slovak contrasted with the bare verb forms in English,
demonstrating separate developmental patterns in the two morphologically different
languages
 language dominance appeared as a result of the different proportions of exposure to
each language, and it was most evident in the vocabulary size and utterance length in
each language
 the child’s two languages were developing separately, yet in a side-by-side fashion,
while mutual interaction of the languages was present throughout all stages of early
development
 consistent, regular exposure to the languages-in-acquisition was crucial for the
development of both languages
 language differentiation was a process that started early in development, it took place
as lexical, grammatical, phonetic as well as pragmatic differentiation, and was evident
as early as the end of the one-word stage (1;7).

These findings are comparable with previous BFLA research referenced in the literature
review.

In summary, interpersonal first principle was at the heart of BFLA, embedded in the
establishment of person-language bond between the child and each parent, and in the
interpersonal exchanges they engaged in. It manifested itself in various aspects of
linguistic production. It was apparent in differential grammatical development of the two
languages through item-based productions of salient linguistic structures. Ria first learnt
lexical items salient in the contexts of the two languages. Since the contexts were
relatively limited, significant overlaps across the two languages were found. These
overlaps allowed for translation equivalents in Ria’s productions. However, the order in
which Ria learnt TEs was determined by routines and situations in which each parent
250
typically engaged. The contextual specificity of early language acquisition was observed in
the learning of grammatical structures as well. The most salient structures in the input were
the first to appear in Ria’s productions, initially as whole items relevant to specific
situations. As Ria’s vocabulary increased and more patterns were discovered, she applied
the structures as regular patterns to more word types. Meanwhile, she was able to
differentiate which grammatical patterns were needed for which language. In pragmatic
development, the interpersonal first principle was observed through the developing sense
of appropriate language choice and pragmatic uses of mixing, which mirrored preferred
parental language use.

9.1 Study Limitations and future research


The diary records and video-recordings collected for this thesis provided rich data on
simultaneous bilingual development. For the purpose of this thesis general analysis of
development in the first two years was carried out, assessing previous research findings on
lexical, morpho-syntactic and pragmatic language choice development, and determining
their validity with the present data. More detailed analyses of phonetic and phonological
development in the bilingual child, early language comprehension in the bilingual
environment and individual structural linguistic phenomena in development of each
language were outside the scope of this thesis. However, the available data captured such
aspects, and their analysis would provide further insights into BFLA.

The study design of this thesis in itself posed certain limitations. Being a longitudinal case
study, no generalizations that would be applicable to larger populations could be drawn
(e.g. groups of children of Slovak immigrants living in Australia). Similarly, since this was
a qualitative study with minimal statistical evidence relevant to one subject, no conclusions
could be drawn about general trends in bilingual language development. It was also
difficult to make appropriate comparisons with norms for initial linguistic development.
Monolingual norms are not adequate for bilingual children, and as Pearson (1998, p. 364)
argues, they should not be applied to bilinguals. Moreover, developmental norms for
bilingual children are not available.

The sociolinguistic analysis of the language learning environment in the initial stages of
BFLA in this study laid a foundation for a long term study in bilingual development. The
child continues on her journey through life as a simultaneous bilingual. However, the
specific language dominance and language use patterns that were established in the early
years may change over time with changes in the child’s social experiences. Furthermore, a

251
follow up study as the child enters school age would allow exploration of later linguistic
and literacy developments as well as bilingual language maintenance.

To improve general understanding of BFLA it is necessary to examine and describe


different kinds of data involving various language combinations with varied exposure
patterns. Future research may consider differences in BFLA development within a bigger
group of bilingual infants in different family types growing up with Slovak and English in
Australia, and across different families with bilingual siblings. Outcomes in early bilingual
development are dependent on individual differences in socialising environments. A
comparison of several case studies would allow conclusions on what types of socializing
environments result in bilingualism for life. Further research would also examine changes
in language use and dominance in bilingual children on a long-term scale, and possible
changes in the type of language mixing over time, following through into adulthood, thus
allowing continuity from studies on childhood to adult bilingualism.

Another aspect of BFLA that would be of interest to future research is a bilingual child’s
private speech and language choice, which has not yet been examined in detail. For
example Saunders (1984, p. 65) found that children’s choice of language in such situations
was sometimes determined by a topic or activities, since the child would tend to use the
language that was usually associated with it, such as a game most often played with one
parent or activity usually engaged in with another parent. Lastly, future research may
examine language choice and mixing in bilingual children in interactions with bilingual
peers, especially bilingual siblings, in early childhood when the family environment plays
a crucial role in bilingual socialization, thus bridging the gap in BFLA research.

252
APPENDIX A Transcription conventions
Phonetic transcription of Slovak sounds in IPA and their equivalents in Slovak
orthography:

IPA symbol Slovak orthography


a a
aː á
æ ä
ɛ e
ɛː é
ɪ i
ɪː í
ɔ o
ɔː ó
ʊ u
ʊː ú
ɪa ia
ɪɛ ie
ɪʊ iu
ʊɔ ô
p p
b b
t t
d d
c ť
ɟ ď
k k
g g
m m
n n
ɲ ň
ʎ ľ
f f
v v
s s
z z
ʃ š
ʒ ž
x ch
ɦ h
j j
r r
rː ŕ
l l
lː ĺ
ts c
dz dz
tʃ č
253
IPA symbol Slovak orthography
dʒ dž

254
Summary of speech lines, non-speech lines and abbreviations used in the transcriptions:

Main tiers

Speech lines indicated by asterisk and a three letter code identifying each speaker:

*RIA Riana

*MAM Mother (bilingual)

*DAD Father

*NEL Nela, Slovak aunt (bilingual)

*MIS Miša, cousin (bilingual)

*ROM Romi, cousin (bilingual)

*BAB Babka, Slovak grandmother

*DED Dedko, Slovak grandfather

*ROB Robi, Slovak aunt

*VAN Vanda, English aunt

Dependent tiers

Non-speech lines preceded by the symbol % (shown in order of use):

%sit Situation of the recording

%glo Target language gloss of nonstandard child forms

%eng Translation into English

%add Addressee (whom the speaker is addressing)

%act Activities or actions in which speakers are engaged

%gpx Facial gesture, body language, proxemic information

%com Comments and explanations

255
Utterance markers from the CHAT manual used in the transcriptions:

xxx unintelligible speech, not treated as a word

xx unintelligible speech, treated as a word

yyy unintelligible speech transcribed on %pho line, not treated as a word

yy unintelligible speech transcribed on %pho line, treated as a word

www untranscribed material

0 actions without speech

& phonological fragment

[?] best guess

text(text)text noncompletion of a word

Utterance terminators from the CHAT manual used in the transcriptions:

. period

? question

! exclamation

Other symbols from the CHAT manual used in the transcriptions:

text^text pause between syllable

: lengthened syllable

# pause between words

## long pause between words

### extra long pause between words.

256
Transcribed audio-video data:

Rec. no. Filename Date of recording Ria’s age Language context Situational context
1 MOV003 13/11/2007 8:19am 0;4.12 Bilingual Family is relaxing
2 MOV02E 25/12/2007 1:03pm 0;5.24 Bilingual Christmas with extended family
3 MOV030 26/12/2007 11:02am 0;5.25 Bilingual Christmas with extended family
4 MOV037 14/01/2008 3:48pm 0;6.13 Slovak Tummy time and changing
5 MOV03D 17/01/2008 8:07am 0;6.16 Bilingual Family is waking up
6 MOV046 2/02/2008 9:39am 0;7.1 Bilingual Snack and free play
7 MOV047 3/02/2008 3:49pm 0;7.2 Bilingual Playing with toys
8 MOV04E 23/02/2008 8:51am 0;7.22 Bilingual Free play
9 MOV054 5/03/2008 4:40pm 0;8.4 Slovak Free play and social games
10 MOV055 12/03/2008 9:45am 0;8.11 Slovak Free play and reading
11 MOV05D 2/04/2008 2:15pm 0;9.1 Slovak Playing in Ria’s new room
12 MOV062 19/04/2008 10:57am 0;9.18 English Free play in Ria’s room
13 MOV068 25/04/2008 8:06pm 0;9.24 Bilingual Bath time
14 MOV06C 11/05/2008 10:14am 0;10.10 Bilingual Breakfast time
15 MOV06D 11/05/2008 10:22am 0;10.10 Bilingual Free play
16 MOV06E 15/05/2008 10:37am 0;10.14 Slovak Play with toys
17 MOV06F 15/5/2008 1:50pm 0;10.14 Slovak Play with objects
18 MOV076 1/06/2008 8:11pm 0;10.31 English Play with toys
19 MOV078 6/06/2008 12:57pm 0;11.5 Slovak (ROM used some English) Play time with cousin
20 MOV07C 16/06/2008 7:36pm 0;11.15 Slovak (DAD used English in Play with toys
background)
21 MOV07E 16/06/2008 8:11pm 0;11.15 English Free play with objects
22 MOV086 26/06/2008 11:54pm 0;11.25 Slovak Play with extended family

257
Rec. no. Filename Date of recording Ria’s age Language context Situational context
23 MOV097 13/07/2008 5:46pm 1;0.12 Slovak Play with extended family
24 MOV098 14/07/2008 7:20am 1;0.13 Bilingual Play with extended family
25 MOV0A2 20/07/2008 3:21am 1;0.19 Slovak Outdoor play with extended family
26 MOV0A3 20/07/2008 3:22am 1;0.19 Slovak Outdoor play with extended family
27 MOV0AF 30/07/2008 6:47pm 1;0.29 Slovak Looking at objects
28 MOV0B1 7/08/2008 8:31pm 1;1.6 Bilingual to Slovak Play with toys and learning to walk
29 MOV0B5 13/08/2008 4:20pm 1;1.12 Slovak Reading books
30 MOV0BA 21/08/2008 2:39pm 1;1.20 Slovak Hanging the washing
31 MOV0C7 18/9/2008 3:01pm 1;2.17 Slovak Reading and playing
32 MOV0C8 21/9/2008 8:24am 1;2.20 Bilingual Helping in the kitchen
33 MOV0C9 22/9/2008 2:08pm 1;2.21 Slovak Playing with toys
34 MOV0CE 6/10/2008 6:00pm 1;3.5 Bilingual Learning to stand on a stool
35 MOV0D0 11/10/2008 5:40pm 1;3.10 Bilingual Play with extended family
36 MOV0D2 17/10/2008 3:24pm 1;3.16 Bilingual Playing and parents talking
37 MOV0D8 22/10/2008 5:27pm 1;3.21 Slovak Reading and playing
38 MOV0DA 29/10/2008 1:43pm 1;3.28 Slovak Play with toys
39 MOV0DF 12/11/2008 1:57pm 1;4.11 Slovak Play with puppets and social games
40 MOV0E0 20/11/2008 1:49pm 1;4.19 Slovak Playing and reading books
41 MOV0E5 3/12/2008 2:43pm 1;5.2 Slovak Playing and reading books
42 MOV0E8 6/12/2008 7:08am 1;5.5 Bilingual Looking at St Nicholas surprise
43 MOV0E9 6/12/2008 7:45am 1;5.5 Bilingual Breakfast time
44 MOV0EB 11/12/2008 5:07pm 1;5.10 Bilingual Play with blocks and talking
45 MOV0ED 21/12/2008 8:43am 1;5.20 Slovak Getting dressed
46 MOV0EE 24/12/2008 5:01pm 1;5.23 Bilingual Christmas Eve dinner
47 MOV0F1 25/12/2008 7:59am 1;5.24 Bilingual Opening Christmas presents
258
Rec. no. Filename Date of recording Ria’s age Language context Situational context
48 MOV0F2 5/1/2009 9:31am 1;6.4 Slovak Play with trains
49 MOV0F8 13/01/2009 4:37pm 1;6.12 Slovak Reading a book
50 MOV0F9 20/01/2009 3:24pm 1;6.19 Slovak Reading a book
51 MOV0FA 24/01/2009 1:42pm 1;6.23 Slovak Ria reading
52 MOV0FB 24/01/2009 2:49pm 1;6.23 Bilingual Reading a book
53 MOV0FC 30/01/2009 3:36pm 1;6.29 Slovak Reading, dancing, drawing
54 MOV0FF 10/2/2009 1:42pm 1;7.9 Slovak (Ria used some English) Doing the washing
55 MOV100 20/2/2009 1:50pm 1;7.19 Slovak (Ria used some English) Looking at objects in the kitchen
56 MOV103 10/3/2009 1:00pm 1;8.9 English (Ria used some Slovak) Reading books
57 MOV104 10/3/2009 1:18pm 1;8.9 Bilingual to English (Ria used Reading books
some Slovak)
58 MOV10A 24/3/2009 4:01pm 1;8.23 Slovak (Ria used some English) Reading books
59 MOV10B 24/3/2009 7:01pm 1;8.23 Bilingual to Slovak (Ria used some Talking and pretending to be on the
English) phone
60 MOV10C 15/4/2009 8:00pm 1;9.14 Bilingual Doing puzzles
61 MOV10E 17/4/2009 6:29pm 1;9.16 Bilingual Social games and rhymes
62 MOV114 28/4/2009 9:49am 1;9.27 Slovak Reading books
63 MOV116 6/5/2009 11:52am 1;10.5 English to bilingual Working in the garden
64 MOV117 6/5/2009 12:27pm 1;10.5 Bilingual Working in the garden
65 MOV11B 30/5/2009 3:18pm 1;10.29 Bilingual Reading a book
66 MOV11D 8/6/2009 9:47am 1;11.7 Slovak Getting dressed
67 MOV120 19/6/2009 3:57pm 1;11.18 Slovak (Ria used some English) Playing with toys

259
APPENDIX B Universal proto-word
Functions of Ria’s ‘universal’ proto-word:

Age Variation Function Examples in context / Comments


0;9.21 [hmː] Instrumental e.g. the mother put the child on the
-a request for desired floor when it appeared she
objects, especially when finished eating, but the child
out of reach and needing crawled back up to the mother and
an adult to manipulate requested more while vocalizing
them to pass them to her [hmː hmː]
[hmː] Instrumental e.g. the father was cooking and as
-a request to be involved the child crawled in, wishing to
in an activity, or to be see, at first she stayed next to him,
picked up looking up, asking to be picked up
by vocalizing [hmː hmː], but since
the father was not responding by
picking her up, only addressing
her, the child knelt and stretched
her arms up
0;10 [hm hm] Interactional- e.g. whilst the mother was on the
-requesting to be engaged telephone, the child was climbing
in the mother’s activity on furniture next to the mother,
fussing and vocalizing, then
climbing onto the mother, wishing
to use the telephone handset
0;10.13 [hm hm] Interactional/ e.g. the mother and the child were
Instrumental interacting with a group of people,
-request for attention and a everyone directing their attention
person’s verbal address at the child; as a new person
walked in and addressed others,
not paying attention to the child,
she started vocalizing in a
frustrated tone, wiggling and
pushing objects off the table, she
calmed down when the person
directed their attention to her
0;11.2 [mmmː] Personal- e.g. a small container not opening,
a -expressing frustration a door not opening when she was
forceful after objects did not in the way, unable to pick up a
sound behave in a desired towel on which she was sitting
manner
0;11.15 [əː] Instrumental/ Other way to express discontent
a vowel Interactional was a specific angry form of cry
produced -expressing discontent
with an with the parents’ actions
open
mouth

260
Age Variation Function Examples in context / Comments
1;0.4 [eː] + Instrumental - used especially if the parents did
pointing -requesting objects or not respond immediately
services
1;0.4 [əː] Instrumental/Interactional - often expressing discontent with
- expressing discontent the parents’ everyday household
with location or activity activities
1;0.13 [eː] + Instrumental e.g. the child requested to go to a
pointing -requesting location + playground when the family was
in a service (to be carried to a walking past; while there she
direction location) requested to move around
individual structures, pointing and
vocalizing each time she wished to
change from monkey bars to slides
or swings, etc., when not wanting
to go to a particular place, she
tensed up her body to express
unwillingness
1;0.26 [eː] + Instrumental
pointing -requesting objects, both
familiar and new objects
she wished to explore
1;0.26 [eː] + Instrumental
pointing -requesting food

261
APPENDIX C Onomatopoeia
Conventionalised and family onomatopoeiae used as first words:

Sound (age of first production)


‘Standard Language
Referent
onomatopoeia’ (if language
Approximated/ /source
Initial form “Meaning” specific)
alternate form
‘mňáu!’ OR
[əːh] (1;0.28) [aːʊ] (1;1.4) cat Either
“meow!”
[mu] or [u]
[m m m m] (1;1.15) ‘mú!’ OR ‘moo!’ cow Either
(1;1.4)
‘kač kač kač’
[dadada]
[dadadat] (1;2.7) “quack quack duck Slovak
(1;1.9)
quack”
‘hav hav hav’
[h h h]
“woof woof dog Slovak
(1;1.12)
woof”
[s s] ‘tsss’
snake Slovak
(1;2.1)
English
click sound
rabbit (imitated
(1;2.1)
after father)
[x x] ‘kroch kroch’ &
pig Slovak
(1;2.3) snorting
[ha ha ha]
[hu hu hu] (1;2.25) monkey English
(1;2.11)
[bɛː bɛː bɛː] ‘bé bé bé’
sheep Slovak
(1;2.21) “baa baa baa”
[huː huː huː]
‘hú hú hú’ owl Slovak
(1;2.25)
[pi pi]
‘pi pi’ chicken Slovak
(1;2.30)
[kakaka] ‘kvak kvak kvak’
frog Slovak
(1;3.21) “ribbit”
[kokoko]
‘kotkodák’ chickens Slovak
(1;3.25)
[ssss] [ʃʃʃʃ] (1;2.25) ‘čč čč’
train Slovak
(1;2.17) [tʃ tʃ tʃ] (1;2.30) “choo choo”
[tutuː] [tu] ‘tutút’
car horn Slovak
(1;3.8) (1;3.15) “toot”

262
APPENDIX D First 50 words
Cumulative vocabulary in the one-word stage:

Age
when Phonetic form and Source
Target word “meaning”
first variations Language
used
Slovak
1 0;11 [hama] [ham] [hamɪ] hami “boobie” (family
word)
2 1;0 [mamaː] [mamɪ] mami “mummy” Slovak
3 1;0 [aɲo] [ano] ano “yes” Slovak
4 1;1 [jana] Riana Either
5 1;1.25 [am] [taː] tam “there” Slovak
6 1;2 [dada] daddy English
7 1;2 [ɲɛɲɛː] nie je or neni “isn’t” Slovak
8 1;3.20 [Ɉɪs] [dɪs] this English
9 1;3.20 [je] yeah English
10 1;3 [baː] bác “up-a-day” (baby word) Slovak
11 1;4 [kaka] [kakkak] kvák kvák “ribbit ribbit” Slovak
[kaka] [kaː] [kakaː] kačkač “quack quack” OR
12 1;4 Slovak
[gaga] gaga “goose sound”
13 1;4 [kokkok] kotkodák “hen sound” Slovak
[ɲamɪ ɲamɪ ɲamɪ]
14 1;4 malilinké “tiny” Slovak
[mɪɲa mɪɲa]
15 1;4 [momo] moja “mine” Slovak
16 1;4 [babaː] bim bam “ding dong” Slovak
17 1;4 [xx] kroch kroch “oink oink” Slovak
18 1;4 [mm] [mu] [muː] mú “moo” Either
19 1;4 [hh] hav hav “woof woof” Slovak
20 1;4 [ɲaː] [ɲaːɲaː] [mɲaːʊ] mňáu “meow” Either
21 1;4 [habw] havo “doggie” Slovak
22 1;4 [huhú] hu-hú “toowit-towoo” Slovak
23 1;4 [mamajamaː] mama moja “my mummy” Slovak
24 1;4 [baːba] bábo “baby” Slovak
25 1;4 [huhu] huhuhuhu “monkey sound” Either
26 1;4 [s] [∫] ts (held longer) “snake hiss” Slovak
27 1;4 [ww ww] woof woof Slovak
28 1;4 [kuku kuku] kukikuk “peep-o” Slovak
29 1;4 [paːlɪ] páli “hot, burning” Slovak
30 1;4 [konɛno] koleno “knee” Slovak
31 1;4 [kuː] cool English
32 1;4 [ɟɛjɛ] kde je? “where is?” Slovak
33 1;4 [ɔke] ok Either
34 1;4 [aɲɛ] [aɲɪ] anjel “angel” Slovak
35 1;4 [maɲa maɲa] malina “raspberry” Slovak
36 1;4 [tiː] tea English
37 1;4 [kaga] kakala “did a poo” Slovak
38 1;4 [goː] gone English
39 1;5 [babɪ] bábika “doll” Slovak
263
Age
when Phonetic form and Source
Target word “meaning”
first variations Language
used
40 1;5 [ɲam] mňam “yum” Slovak
[b+vibration b+vib brum brum “bear sound
41 1;5 Slovak
b+vib b+vib] (meaning bear)”
42 1;5 [baː] (creaky) bé bé “ba ba” Either
kikirikí “cock-a-doodle-doo
43 1;5 [kuːkuː] Slovak
meaning rooster”
44 1;5 [∫∫∫] šš šš “train sound” Slovak
45 1;5 [caba] table English
46 1;5 [cɪː] deti “children” Slovak
47 1;5 [kɛwɔ] kvietok “flower” Slovak
48 1;5 [cɔtɔ?] [cɔjɛ?] [cɔjɛtɔ?] čo je to? “what is it?” Slovak
49 1;5 [puː] poo English
50 1;5 [piː] please English

264
APPENDIX E Fixed pivot schemas and item-based frames
Fixed expressions produced as a single unit, with one fixed and one alternating constituent:

Schema ‘Target form’


Age
Language, addressee(s) “Meaning”
[deəːʃiːs]
1;7.14 ‘there she is!’
English, Father
[ɟɛsiː?] ‘kde si?’
1;7.17
Slovak, Mother “where are you?”
[tʃesuː] ‘chase you!’
1;8.9
English, Father “chase you (=me)!”
[ɟɛsuː] ‘kde sú?’
1;8.9
Slovak, either “where are they?”
[tusuː] ‘tu sú!’
1;8.9
Slovak, either “here they are!”
[tujɛ] ‘tu je!’
1;8.9
Slovak, either “here it is!”
[ʃoʊmiː]
1;8.9 ‘show me!’
English, both parents
[bajtʃuː]
1;8.9 ‘bite you’
English, Father
[zaɲɪm] ‘za ním’
1;8.11
Slovak, either “(go) to him.”
[ʃodomaʃ] ‘čo to máš?’
1;8.21
Slovak, Mother “what have you got?”
[ɟɛjɛ] ‘kde je?’
1;8.21
Slovak, Mother “where is?”
[kʊmamɛ] ‘ku mame’
1;8.21
Slovak, Mother “to mummy”
[ʃɔdɔ] ‘čo to je?’
1;8.21
Slovak, Mother “what’s that”
[deːəɪs]
1;8.21 ‘there (he) is’
English, both parents
[dɪsam]
1;8.31 ‘this one’
English, Father
[jɛdam]
1;8.31 ‘red one’
English, Father
[kʊjaɲɛ] ‘ku Riane’
1;8.31
Slovak, Mother “(come) to Ria.”
[ɟɛsɔm?] ‘kde som?’
1;9.4
Slovak, Mother “where am I?”
[tʊjɛ] ‘tu je!’
1;9.4
Slovak, Mother “Here it is!”
[tʊsʊː] ‘tu sú’
1;9.4
Slovak Mother “here they are”
[bɪgam]
1;9.4 ‘big one’
English, Father
[zaɲou] ‘za ňou’
1;9.8
Slovak, Mother “(go) to it (=the ball).”
1;9.12 [koʊʃɪt] ‘close it’
265
Schema ‘Target form’
Age
Language, addressee(s) “Meaning”
English, Father
[iːtɪt]
1;9.12 ‘eat it’
English, Father
[jaɪkət]
1;9.12 ‘like it’
English, Father
[buːwan]
1;9.12 ‘blue one’
English, Father
[jejowan]
1;9.12 ‘yellow one’
English, Father
[bekan]
1;9.12 ‘black one’
English, Father
[dɪsam] [dɪʃaŋ]
1;9.12 ‘this one’
English, Father
[gudam]
1;9.12 ‘good one’
English, Father
[dɪsams]
1;9.12 ‘these ones’
English, Father
[ʃokon]
1;9.12 ‘sock on’
English, Father
[jajtson]
1;9.12 ‘lights on’
English, Father
[kʊnam] ‘ku nám’
1;9.12
Slovak, Mother “towards us”
[dɔkʊpɛɲɛ] ‘do kúpeľne’
1;9.12
Slovak, Mother “to the bathroom”
[vaʊcɛ] ‘v aute’
1;9.12
Slovak, Mother “in the car”
[dɪsɪs]
1;9.12 ‘this is’
English, Father
[naɪtam]
1;9.19 ‘night time’
English, Father
[basam]
1;9.19 ‘bath time’
English, Father
[kejɪɪc]
1;9.19 ‘carry it’
English, Father
[kætʃɪm]
1;9.19 ‘catch him’
English, Father
[jɪtujan]
1;9.19 ‘little one’
English, Father
[dɔʃɔjɪ] ‘do školy’
1;9.19
Slovak, Mother “to school”
[dɔkɔʃa] ‘do koša’
1;9.19
Slovak, Mother “in the bin”
[hepsam]
1;9.19 ‘have some’
English, Father
[oʊgon]
1;9.20 ‘all gone’
English, Mother
[pɔsɔjɔm] [pɔdsɔjɔm] ‘pod stolom’
1;9.26
Slovak, Mother “uder the table”
266
Schema ‘Target form’
Age
Language, addressee(s) “Meaning”
[voʃɪt]
1;9.26 ‘wash it’
English, Father
[bekɪn]
1;9.26 ‘back in (foot under the table)’
English, both parents
[dʊɪt]
1;9.26 ‘do it’
English, Father
[wendeə?]
1;9.26 ‘went there?’
English, Father
[pɪtsam]
1;9.26 ‘pick some’
English, Father
[ɲuːans]
1;9.26 ‘new ones’
English, Mother
[nasɛm] ‘na zem’
1;9.26
Slovak, Mother “on the floor”
[dɔʃɔʊkɪ] ‘do škôlky’
1;10.4
Slovak, Mother “to childcare”
[ʃuːʃon]
1;10.6 ‘shoes on’
English, Father
[hætson]
1;10.6 ‘hats on’
English, Father
[nadujan]
1;10.6 ‘another one’
English, Father
detɪs
1;10.6 ‘that is’
English, Mother
[tɔtɔjɛ] ‘čo to je?’
1;10.7
Slovak, Mother “what’s that?”
[dɔsakada] ‘do záchoda’
1;10.7
Slovak, Mother “in the toilet”
[nasanʊ] ‘na stranu’
1;10.7
Slovak, Mother “on the side”
[jɪtujans]
1;10.10 ‘little ones’
English, Mother
[tʊjɛm!] ‘tu som!’
1;10.10
Slovak, Mother “here I am!”
[baɪnap]
1;10.14 ‘blind up’
English, Mother
[mɪsuː]
1;10.19 ‘miss you’
English, Father
[hoʊdɪs]
1;10.22 ‘hold this’
English, Father
[adujan]
1;10.23 ‘other one’
English, Father
[gejap!]
1;10.26 ‘get up!’
English, Father
[pɪpɔtatʃɪ] ‘pri počítači’
1;10.27
Slovak, Mother “by the computer”
[weəʃɪs?]
1;10.27 ‘where she’s?’
English, Father
1;10.27 [tʊsɪː!] ‘tu si!’ (referring to self)
267
Schema ‘Target form’
Age
Language, addressee(s) “Meaning”
Slovak, Mother “here you are!”
1;11.3 [oːʊdesd] ‘all dressed’
[teɪsɪ]
1;11.4 ‘taste it’
English, Father
[adujuɲʊ]
1;11.4 ‘other one-u’ (Slovak feminine suffix)
Mixed, Father
[bɪgaɲ]
1;11.4 ‘big one’
English, Father
[wontsa?]
1;11.4 ‘want some?’
English, Father
[oʊpeɲɪc]
1;11.6 ‘open it’
English, Father
[woʊdɪt]
1;11.6 ‘rolled it’
English, both parents
[hebɪt]
1;11.7 ‘have it’
English, both parents
[hesam]
1;11.7 ‘has some’
English, both parents
[habɪt]
1;11.11 ‘have it’
English, both parents
[pekap]
1;11.12 ‘pack up’
English, both parents
[jaɪksam]
1;11.14 ‘like some’
English, both parents
[fɪksɪt]
1;11.18 ‘fix it’
English, both parents
[gejap]
1;11.19 ‘get up’
English, both parents
[gedɪt]
1;11.21 ‘get it’
English, Father
[dɪdec]
1;11.23 ‘did it’
English, Father
[jiːbɪc]
1;11.27 ‘leave it’
English, both parents

268
APPENDIX F Early word combinations
Language choice indicates one of five utterance types: unilingual Slovak or English
utterance, bilingual utterance, mixed utterance, or indeterminate if the language could not
be clearly determined and the utterance could belong to either language. The utterance
addressee is indicated first, followed by other interlocutors that were present.

Age Language choice ‘Target form’ Utterance Function


Addressee (others “meaning” length
present)
1;4.8 [mama dada gaga] ‘mami, daddy, 3-word Multiple
Indeterminate kačiatka’ subject -
Mother “mummy, daddy, Expressing
ducklings” presence of
objects
1;4.9 [mama momo] ‘mama moja’ 2-word Social
Slovak “my mummy” expression of
Mother feelings
1;4.19 [mamɪ dædɪ jana] ‘mami, daddy, Ria’ 3-word Multiple
Indeterminate “Mami, daddy, subject -
Both parents Ria, (we’re all Expressing
here).” presence of
people
1;4.19 [mamɪ jana] ‘mami, Riana’ 2-word Expressing
Indeterminate “mami and Riana presence of
Mother (are here.)” people
1;4.29 [dædɪ en mamɪ] ‘daddy and mami’ 3-word Expressing
English “daddy and mami presence of
Both parents (are here).” people
1;5.5 [det. de to mama.] ‘that. that to 2-word Expressing
Mixed mama.’ (to is the telegraphic recipient
Both Slovak TE of that)
“that. that that
mummy.”
1;5.5 [nanaː jɛ] (pointing at 2-word Specifying
Slovak herself) ‘Riana je’
Both “Riana is”
1;5.5 [tʃa ti] ‘čaj tea’ 2-word Providing
Bilingual “tea tea” clarification
Both
1;5.6 [muː # mɪɲa mɪɲa] ‘moo # malilinká’ 2-word Property of
Slovak “tiny cow” an object
Both
1;5.13 [mamɪ. ## dædɪ ɲɛ.] ‘mami. daddy nie.’ 2-word Action
Slovak “mami (is description
Mother changing my
nappy). daddy
not.”
1;5.14 [bebɪ. # babɔ] ‘baby. bábo.’ 2-word Providing
Bilingual “baby. baby.” clarification
269
Age Language choice ‘Target form’ Utterance Function
Addressee (others “meaning” length
present)
Father
1;5.14 [dam piː] ‘dám please’ 2-word Request
Mixed “give (me water)
Father (Mother) please.”
1;5.15 [dædɪ. tiː. go.] ‘daddy. tea. go.’ 3-word Action
English “daddy (will drink) telegraphic description
Both parents tea (and) go.”
1;5.15 [tɪ dædɪ] ‘teeth daddy’ 2-word Person acting
English “daddy (is on object
Father brushing his)
teeth.”
1;5.23 [pɛjanʊ. mamɪ!] ‘pre Rianu. mami!’ 2-word Indicating
Slovak “for Riana. mami!” telegraphic recipient
Father (Both parents) (mami gave a
spoon for Ria to
DAD)
1;5.24 [teɪn. jaɲa.] ‘train. Riana.’ 2-word Possession
English “train. Riana(‘s)” telegraphic
Both parents
1;5.26 [po. dædɪ.] ‘pošta. (pre 2-word Possession
Slovak daddy(ho).’ telegraphic
Mother “mail. (for)
daddy.”

270
APPENDIX G Ria’s Mixed Utterances
One-word English utterances mixed in interactions with the mother:

Word ‘target word’ Tokens Type of morpheme/word class


[taː] ‘ta’ (thanks) 2 Function word/closed class
[hi] ‘here’ 2 Function word/closed class
[dɪs] ‘this’ 2 Function word/closed class
[mo] ‘more’ 2 Function word/closed class
[ap] ‘up’ 1 Function word/closed class
[on] ‘(put) on’ 1 Function word/closed class
[piː] ‘please’ 2 Social word/closed class
[wos] ‘what’s that’ (fixed expression) 1 Function word/closed class
[goː] ‘gone’ 5 Lexical/open class
[kuː] ‘cool’ 3 Lexical/open class
[go] ‘go’ 4 Lexical/open class
[ho] ‘hot’ 2 Lexical/open class
[caba] ‘table’ 1 Lexical/open class
[tiː] ‘tea’ 1 Lexical/open class
[puː] ‘poo’ 1 Lexical/open class
[kiː] ‘key-(ring)’ 1 Lexical/open class
[bɪ] ‘bib’ 1 Lexical/open class
[bo] ‘ball’ 1 Lexical/open class
[tiː] ‘teeth’ 1 Lexical/open class
[beː] ‘bear’ 1 Lexical/open class
[foː] ‘fork’ 1 Lexical/open class
[pɪ] ‘pretty’ 1 Lexical/open class
[ʃiː] ‘see’ 1 Lexical/open class
[nɪ] ‘knee’ 1 Lexical/open class
[pu] ‘poo’ 1 Lexical/open class
[bɪ] ‘big’ 1 Lexical/open class

271
One-word Slovak utterances mixed in interactions with the father:

Word ‘target word’ “meaning” Tokens Type of morpheme/word class


[aɲɔ] ‘ano’ “yes” 3 Function word/closed class
[dʊ] ‘druhú’ “the other one” 1 Function word/closed class
[moj] ‘mój’ “my (daddy)” 1 Function word/closed class
[ma] ‘mravec’ “ant” 3 Lexical/open class
[paː] ‘pavúk’ “spider” 1 Lexical/open class
[taːwa] ‘tráva’ “grass” 1 Lexical/open class
[ʒʊ] ‘zuby’ “teeth” 1 Lexical/open class
[kʊː] ‘kľúče’ “keys” 1 Lexical/open class
[b+vibration] ‘brum brum’ 1 Lexical/open class
(onomatopoeia for bear sound)

272
Mixed two-word and multiple-word utterances

Language of utterance indicates the base language which was determined by the addressee
(Slovak or English), and into which a mix was introduced. If the language could not be
clearly determined and the utterance could belong to either language, as in some two-word
utterances or in bilingual contexts, the utterance was marked as Indeterminate.

Mixing type refers to lexical or grammatical mixing, blends or borrowing. Lexical mixing
was categorized as mixing where an item from the other language was used in an
unchanged form in an otherwise base language utterance. If the item was integrated
morphologically into the base language, using grammatical morphemes from the base
language, mixing was categorized as grammatical. Borrowing refers to items that were
used on family level or introduced by parents, and thus not identified as true mixing.
Blends were words consisting of items from both languages, which could also be
considered grammatical mix if containing grammatical morphemes.

The mixed items are shown in italics in the ‘Target form’ column. The determinant of
mixing refers to this item. Where two items were mixed in one utterance, two mixing types
and two determinants are shown, respectively.

The following abbreviations are used in notation of grammatical mixing:

ENI – English Noun Inflection

SNI – Slovak Noun Inflection

SVI – Slovak Verb Inflection

SG – Slovak Gender morpheme

273
Mixed two/multiple-word utterances addressed to the mother:

Language of
Utterance ‘Target form’ Mixing type
Age utterance Reasons for mixing
“Meaning”
(Others present)
[duː. da. jana. hiː.] ‘druhý. dá. Riana. here.’
1;6.18 Slovak Lexical Word preference
“Riana will put another (peg) here.”
[mamɪ. go. go. go. ka.
‘mami. go. go. go. kam? (za) daddym.’
1;6.19 dædɪ.] Slovak Lexical Lexical gap
„Mami let’s go. Where? To daddy.”
‘poo. vtáčik(ove).’
1;6.19 [puː. ʃaː. ʃaː. ʃaː.] Slovak Lexical Word preference
‘Birdie’s poo.’
‘nie je. here. nie je.’ Word preference /
1;6.19 [ɲɛɲɪ. ɲɛɲɪ. hɪ. hɪ. ɲɛɲɪ] Slovak Lexical
‘There isn’t any (food). Here isn’t.’ negator
‘that. want. Riana.’ Word preference
1;6.22 [det. det. det. sɛʃ. jana.] Slovak Lexical
“Riana wants that.”
páli. this.
1;7.10 [paːɪ.dɪs.] Slovak Lexical Lexical gap
“This is hot.”
‘sem. knižku. ta (for) mami.’ Lexical gap – learning
1;7.14 [sɛm. kɪ. taː mamɪ.] Slovak Lexical
“Put the book here. Ta for mami.” context
Lexical gap - learning
‘work. nie.’
1;8.31 [wək. ɲɛ.] Slovak Lexical context
“No work (for daddy).”
Negator
‘nie butter’ Lexical gap/
1;8.31 [ɲijɛ badə] Slovak Lexical
“I don’t want butter.” Negator
‘mäso.’
[mesɔ] (M: Čo mäso?)
Slovak
1;9.19 (M: Čo mäso?) ‘Riana, little.’ Lexical Lexical gap
(Father)
[jana, jɪtu] “meat.”
(M: What meat?)

274
Language of
Utterance ‘Target form’ Mixing type
Age utterance Reasons for mixing
“Meaning”
(Others present)
“Riana, little.”
‘hami. bye-bye. pusu ## telefón. auto.
ball.’ Borrowing introduced
[hamɪ. baɪbaɪ. pʊsʊ ## Borrowing
1;9.26 Slovak “We’ll have boobie. Say bye-bye. by mother /
tɛjɛsɔːn. atɔ. bɔːʊ] Lexical
Give kiss ## Play with telephone, car, Word preference
ball.”
‘garáž. go.’
1;9.26 [gajaːʃ. goʊ.] Slovak Lexical Lexical gap
“garage. Go.”
‘auto, green. nájdeme.’
1;9.26 [atɔ, gɪn. naɟɛmɛ.] Slovak Lexical Word preference
“Car, green. We will find.”
‘new ones. kúpil.’
1;9.26 [ɲuːans. kʊːpɪʊ.] Slovak Lexical Lexical gap
“New ones. He bought.“
‘oh-oh.’
(M: Čo sa stalo?)
[ooʊ]
‘toto. na zem. puzzles.’ Borrowing introduced
1;9.26 (M: Čo sa stalo?) Slovak Borrowing
“oh-oh.” by mother
[tɔtɔ. nasɛm. paʒʊs.]
(M: What happened?)
“This. On the floor. Puzzles.”
‘that is. dadd-ush ukázal.’
1;10.6 [detɪs. dædʊʃ ʊkaːzaʊ] Slovak Lexical Word preference
“that is. (what) daddy showed”
‘nie je toothbrush-a’ Grammatical: Lexical gap - learning
1;10.7 [ɲɛjɛ, tubaʃa] Slovak
“isn’t toothbrush” toothbrush+SNI context
‘kde je mami? toothbrush-a’ Grammatical: Lexical gap -learning
1;10.7 [ɟɛjɛ mamɪ? tubaʃa] Slovak
“where’s mami? toothbrush” toothbrush+SNI context
‘excuse me mami. sadká (si) Riana’
1;10.8 [suʃɪ mamɪ. sada, jana.] Slovak Lexical Lexical gap
“excuse me mummy. will sit Riana”
1;10.8 [zʊbɪ, mamɪ. basum.] Slovak ‘zuby mami. bathroom’ Lexical Lexical gap

275
Language of
Utterance ‘Target form’ Mixing type
Age utterance Reasons for mixing
“Meaning”
(Others present)
“teeth mummy. bathroom”
‘bum bolí’
1;10.8 [bam bɔjɪː] Slovak Lexical Momentary lapse
“bum hurts”
‘finish-ol. daddy.’ Grammatical: Lexical gap/word
1;10.8 [fɪnɪʃɔʊ. dædɪ.] Slovak
“finished. Daddy” finish+SVI preference
‘gaťky, Riana, podám. # little ones’
[gacɪ, jana, podaːm. #
1;10.10 Slovak “undies, Riana, I will pass. # little Lexical Momentary lapse
jɪtʊjans]
ones”
1;10.11 [tɛnkʊː mamɪ] Indeterminate ‘thank you mami’ Lexical Word preference
‘hot. chytať nie.’
1;10.13 [hot. kɪtac ɲɪɛ.] Slovak Lexical Momentary lapse
“hot. touching not”
[ʊjɔbɪːmɛ dædɪ Slovak ‘urobíme daddy cammomile’ Lexical Lexical gap - learning
1;10.24
kemomaʊ] (Father) “we’ll make daddy camomile (tea)” context
‘birdie spadol do vody’ Momentary lapse -
1;10.29 [bədiː sadɔʊ dɔ vɔdɪ] Slovak Lexical
“birdie fell in the water” newly acquired word
Slovak ‘to. goose. goose. goose.’ Previous context/topic
1;10.29 [tɔ gus. gus. gus.] Lexical
(Father) “that. Goose. Goose. Goose.” was in English
‘a to je guinny pigs. guinny guinny
[adɔ jɛ sɪnu pɪgs. gmɪnɪ Slovak New word - learning
1;10.29 pigs.’ Lexical
gmɪn pɪgs] (Father) context
“and that is guinny pigs”
Bilingual context,
Slovak ‘bunny. ano.’
1;10.29 [banɪ. anɔ.] Lexical previous context/topic
(Father) “bunny. Yes”
was in English
Bilingual context,
Slovak ‘toto. chicken.’
1;10.29 [tɔtɔ. tʃɪkən] Lexical previous context/topic
(Father) “This. Chicken”
was in English
1;10.29 [bədi. ʒʊtɪ. tɔtɔjɛ?] Slovak ‘birdie. žltý. čo to je?’ Lexical Momentary lapse -

276
Language of
Utterance ‘Target form’ Mixing type
Age utterance Reasons for mixing
“Meaning”
(Others present)
(Father) “birdie. Yellow. What’s that?” newly acquired word
Slovak ‘ešte this one krab’
1;10.29 [ɛʃɛ dɪswan kab] Lexical Word preference
(Father) “also this one crab”
Slovak ‘this one chrobák’
1;10.29 [dɪsaŋ. kɔbak.] Lexical Word preference
(Father) “this one beetle”
Previous context was in
Lexical /
Slovak/ English /
‘toucans. book-ine’ Blend –
1;10.29 [tʊkanʃ. bukɪnɛ] Mother Family word (used by
“toucans. (in) book-ina” grammatical:
(Father) the mother with Slovak
book+SNI
inflections)
[tɔtɔ sɛjɛnaː. paːtʃɪ ‘toto zelená. páči zelená. thank you’
1;10.31 Slovak Lexical Lexical gap
sɛjɛnaː. cenkuː] “this green like green. Thnak you.”
Lexical gap –
‘mami stamps-y’ Grammatical:
1;11.1 [mamɪ semsɪ] Slovak morphologically
“mummy stamps” stamps+SNI
integrated into Slovak
Momentary lapse –
‘iné socks-y’ Grammatical:
1;11.1 [ɪnɛː sɔksɪ] Slovak morphologically
“other socks” socks+SNI
integrated into Slovak
[nɔːʊ mamɪ. kʊpɪtsa ‘no mami. krupica s makom nie.’
1;11.11 Slovak Lexical Momentary lapse
makɔm ɲɪɛ.] “no mummy. goodness me no.”
[havɪnkɔː. damp. ‘havinko. jump. skákal.’ Momentary lapse –
1;11.3 Slovak Lexical
ʃakaʊ.] “doggie. Jump. He was jumping.” self-repair
‘vstávam, pozri. look.’
[sabam, pɔsɪ. juk.] (M: čo robíš?)
1;11.3 (M: čo robíš?) Slovak ‘postavím (sa).’ Lexical Emphasis
[pɔsabɪːm] “I’m getting up, look. look.’
(M: What are you doing?)

277
Language of
Utterance ‘Target form’ Mixing type
Age utterance Reasons for mixing
“Meaning”
(Others present)
“I will get up”
‘skákať! jump!’
1;11.3 [saːkac! dʒamp!] Slovak Lexical Emphasis
“to jump! Jump!”
‘mami, Cibo’
(M: čo sme tam robili?)
[mamɪ, ciːbɔʊ]
‘babycino # icecream’ TE not available
1;11.3 (M: čo sme tam robili?) Slovak Lexical
“mummy, Cibo” Lexical gap
[bɛbɪtʃiːno # aɪsiːm]
(M: What did we do there?) “babycino
# icecream”
Slovak ‘ešte (to)mato’
1;11.3 [ɛʃɛ maːtoʊ] Lexical Word preference
(other family) “more tomato”
‘Martin, dal, present’ Learning context -new
1;11.3 [maːcɪn, daʊ, pesent] Slovak Lexical
“Martin, gave, present” word
‘srnka. baby spinká’ Lexical gap (Slovak TE
1;11.3 [ʃənkaː. beɪbɪ ʃɪnkaː] Slovak Lexical
“doe. Baby is sleeping.” ‘mláďa’)
‘knižka cold’
1;11.4 [kɪʃa kɔʊd] Slovak Lexical Word preference
“book cold”
Lexical / Word preference/
‘cold, book-iny’ Blend- Family word (used with
1;11.4 [koʊt, bukɪnɪ] Slovak
“cold, book.” grammatical: Slovak inflections by
book+SNI mother)
Slovak ‘pencil. pencil zoberiem’
1;11.4 [pentsu. pentsu zɔbam] Lexical Word preference
(Father) “pencil. Pencil I will take.”
Blend- Family word (used with
Slovak ‘book-inu mám’
1;11.5 [bukɪnʊ maːm] grammatical: Slovak inflections by
(Father) “book I have”
book+SNI the mother)
1;11.5 [paːtʃɪ, bukɪna] Slovak ‘páči, book-ina’ Blend- Family word (used with

278
Language of
Utterance ‘Target form’ Mixing type
Age utterance Reasons for mixing
“Meaning”
(Others present)
(Father) “there you go, book” grammatical: Slovak inflections by
book+SNI the mother)
Slovak ‘ja mám fork’ Momentary lapse –
1;11.5 [ja maːm fok] Lexical
(Father) “I have fork” bilingual context
Slovak ‘nemám fork’ Momentary lapse –
1;11.5 [ɲɛmaːm fok] Lexical
(Father) “I don’t have a fork” bilingual context
Word preference,
‘mami, another one-a’ Grammatical:
1;11.5 [mamɪː, nadujana!] Slovak integrated
“mummy, another one!” another one+SG
morphologically
‘veľký too much’
1;11.6 [vɛɪkɪː tuː matʃ] Slovak Lexical Lexical gap
“big too much”
Slovak ‘mami, píš frying pan!’
1;11.6 [mamɪ, pɪːʃ faɪfen!] Lexical Reported speech
(Father) “mummy, write frying pan!”
Slovak ‘ja mám toast, mami’ Lexical gap – learning
1;11.13 [ja mam toʊs, mamɪ] Lexical
(Father) “I have toast, mummy” context
‘čo to je? tablet.’ Lexical gap – learning
1;11.13 [tɔtɔjɛ? tebec] Slovak Lexical
“what is it? tablet” context
‘another one kamienok’
1;11.16 [enadujan kamɛjɔc] Slovak Lexical Word preference
“another one rock”
Momentary lapse – new
‘tieto plants-y’ Grammatical: word preference,
1;11.18 [cɛtɔ paːntsɪ] Slovak
“these plants” plants+SNI morphologically
integrated
‘ešte cinnamon’ Lexical gap – learning
1;11.18 [ɛʃɛ ʃɪnamən] Slovak Lexical
“more cinnamon” context
[ʃɪɲɪtʃ. tʃɛvɛnɛː ʃɪɲɪtʃ, Slovak ‘spinach. červené spinach, mami.’ Lexical gap – learning
1;11.27 Lexical
mamɪ] (Father) “spinach. Red spinach, mummy” context

279
Language of
Utterance ‘Target form’ Mixing type
Age utterance Reasons for mixing
“Meaning”
(Others present)
[dædɪ # zaɲɪm # fəːs ‘daddy # zaním # first Riana’
1;11.28 Slovak Lexical Lexical gap
jana] “daddy # after him # first Riana”
‘thank you babka’
1;11.28 [tenk juː babka] Slovak Lexical Word preference
“thank you grandma”
‘plate Rianin mami’
1;11.28 [peɪc jaɲɪn mamɪ] Slovak Lexical Lexical gap
“Plate Riana’s mummy”
[tʃɪpsɔʊ ɲɪɛ. tʃɪps Slovak ‘čipsou nie. chips nemám.’ Borrowing: Borrowing introduced
1;11.30
ɲɛmaːm.] (Father) “with chip not. chips I don’t have.’ chips+SNI by mother

280
Mixed two/multiple-word utterances addressed to the father:

Language of
Utterance ‘Target form’ Mixing type
Age utterance Reasons for mixing
“Meaning”
(Others present)
[dam piː] English ‘dám please.’ Lexical
1;5.14 Lexical gap
(Mother) “Give (me water) please.”
English ‘shorts. dolu.’
1;7.14 [ʃoːts. dou.] Lexical Lexical gap
“Take the shorts off.”
Blend: Lexical gap – gender
English ‘this one-a lav(ička). daddush.’
1;10.5 [dɪsanalav. dæduʃ] this one + agreement with mixed
(Mother) “this one bench. Daddy.”
SG+lav(ička) Slovak noun
English ‘poo and vycikala # look’
1;11.3 [puː en vɪtsɪkaja # juk!] Lexical Momentary lapse
(Mother) “poo(d) and peed # look’
Indeterminate ‘lososa daddy’ Momentary lapse –
1;11.6 [jɔsasa dædɪ] Lexical
(Mother) “salmon daddy” bilingual context
‘ešte pomaranč. ešte daddy.’
(F: More orange?)
[ɛʃɛ pɔmajantʃ. ɛʃɛ dædɪ.]
Slovak ‘orange.’ Momentary lapse
1;11.7 (F: more orange?) Lexical
(Mother) “more orange. more daddy.” New word – lexical gap
[ojintʃ.]
(F: More orange?)
“orange.”
English ‘daddy má cross’
1;11.12 [dædɪ maː kos] Lexical Lexical gap
“daddy has cross”
Reported speech of Ria’s
[jabʊkɔ, ɛʃɛ, dædɪ. koːs!] Slovak & English and father’s utterances
‘jablko ešte daddy. of course!’
1;11.18 Addressed to self Lexical (Ria’s original utterance:
“apple more daddy. of course!”
(Father) Momentary lapse
/Momentary lapse)
1;11.28 [tuː ɔjɛʃɔk?] English ‘too oriešok?’ Lexical Lexical gap – learning

281
Language of
Utterance ‘Target form’ Mixing type
Age utterance Reasons for mixing
“Meaning”
(Others present)
(Mother) “too, (would you like) nut?” context
English ‘more oriešok daddy?’ Lexical gap – learning
1;11.28 [moː ɔjɛʃɔk dædɪ?] Lexical
(Mother) “more nut daddy?” context

282
Mixed two/multiple-word utterances addressed to both parents in triadic interactions:

Language of
Utterance ‘Target form’ Mixing type
Age utterance Reasons for mixing
“Meaning”
(Others present)
[det. deto mama.] Indeterminate ‘that. That+to mama.’ Word preference followed
1;5.5 Lexical
“That. That that mummy.” by self-repair
Indeterminate ‘vtáčikove poo’
1;6.23 [ʃatatwɔ pu] Lexical Word preference
“Birdie’s poo”
‘(on) bikes. daddy. Riana. ihrisko. fun.’
Indeterminate
1;7.15 [bɪ. dædɪ. jana. i. fam.] “We’re going on bikes, daddy with Lexical Lexical gaps
Riana, to the playground, it will be fun.”
Indeterminate ‘lawn-mower. sem.’ Word preference – newly
1;7.28 [lono. sɛm.] Lexical
“Put the lawnmower here.” acquired word
Indeterminate ‘nie je. parsley.’
1;8.9 [ɲjɛjɛ. paːsɪ.] Lexical Lexical gap - negator
“I don’t want parsley on my cucumber.”
Bilingual context –
difficulty retrieving, since
‘von. daddy. Riana. mami. chairs.
both items (von “outside”
[wom. dædɪ. jana. Indeterminate Riana.’
1;8.9 Lexical and chairs “stoličky”) were
mamɪ. # tʃɛːəs. jana.] “Let’s go outside, Daddy, Riana, mami,
available in productive
let’s put the chairs out, Riana’s too.”
vocabulary

‘veľký. cut.’ Word preference


Indeterminate
1;8.21 [vɛːjɪ. kat.] “Cut this big piece.” OR “This is a big Lexical Lexical gap
piece. Cut it.”
Indeterminate ‘green. svetielko.’ Word preference/
1;8.21 [giːn. ʃɛ.] Lexical
“The light is green.” Lexical gap
Indeterminate ‘meat meat meat meat. bolo.’ Lexical gap – learning
1;8,21 [mɪ mɪ mɪ mɪ. bojo.] Lexical
“There was some meat.” context

283
Language of
Utterance ‘Target form’ Mixing type
Age utterance Reasons for mixing
“Meaning”
(Others present)
Lexical gap – learning
Indeterminate ‘varí daddy meat.’
1;8.21 [vajɪː dædɪ mɪ mɪ mɪ] Lexical context
“Daddy is cooking meat.”
‘ľúbiš daddyho. ľúbiš mami. ľúbiš Family word (adopted by
Borrowing -
[jubɪʃ dædɪ. jubɪʃ mamɪ. Indeterminate babos.’ father and used with
1;8.24 Grammatical
jubɪʃ babos.] “You love daddy. You love mami. You English grammatical
babo+ENI
love babos.” morphemes)
Indeterminate ‘tu je barbie’
1;8.31 [tʊjɛ baːbiː] Lexical No equivalent available
“Here’s the barbie! (BBQ)”
[kat. kat. # nɔʃɪkɔm. Slovak ‘cut. nožíkom. maminym.’
1;9.1 Lexical Word preference
mamɪnɪ.] “Cut it with mami’s knife.”
‘this one, green. green and žlté.’
[dɪsam, gɪn. gɪn gɪn gɪn English “This one, the green one and yellow Momentary
1;9.12 Lexical
gɪn ɛ ʒʊtɛː] one.” (pointing at several different lapse/difficulty retrieving
objects)
‘daddy’s, Riana’s gone. vidličku,
[dædɪs, janas gon. vɪdʊ, English
1;9.19 vidličku!’ Lexical Momentary lapse
vɪdʊ!]
“Daddy’s, Riana’s gone. Fork, fork!
Slovak ‘cut. nožíkom! daddyho!’
1;9.20 [kat. nɔʃɪːkɔm! dædɪxɔ!] Lexical Word preference
“cut. With a knife! Daddy’s (knife)!”
Indeterminate ‘daddy, má under’
1;10.6 [dædɪ, maː ʌndə] Lexical Lexical gap
“daddy, has (chin) under (beard)”
Indeterminate ‘mami pichá. forhead’ Word preference –newly
1;10.6 [mamɪ pɪxaː. foːfet] Lexical
“mummy prickles. (on the) forhead” acquired word
[paːjɪ, hɪbɪk. maʃʊm. Indeterminate ‘páli hríbik. mushroom páli.’ Momentary lapse -
1;10.7 Lexical
paːjɪ] “hot mushroom. mushroom hot.” Clarification for father
1;10.7 [tɔtɔ. maːtoʊ] Indeterminate ‘toto. (to)mato’ Lexical Word preference

284
Language of
Utterance ‘Target form’ Mixing type
Age utterance Reasons for mixing
“Meaning”
(Others present)
“this. tomato”
Lexical gap
Indeterminate ‘nemáme towel’
1;10.18 [ɲɛmamɛ tawǝ] Lexical Previous context/topic was
“we don’t have towel”
in English
Indeterminate ‘mama má cheaks’
1;10.24 [mama maː tʃiːks] Lexical Bilingual context
“mummy has cheaks”
Indeterminate ‘ihrisko, play’
1;11.4 [ɪsɪjɔ, peɪ] Lexical Bilingual context
“playground, play”
Indeterminate ‘vypapám. # gone’
1;11.4 [vɪpabam. # gon] Lexical Word preference
“I will eat up. # gone”
Indeterminate ‘(to)mato, páli’
1;11.7 [maːtoʊ, paːɪ] Lexical Word preference
“tomato is hot”
Slovak ‘(to)mato ja mám’
1;11.7 [maːtoʊ ja maːm] Lexical Word preference
“tomato I have”
Indeterminate ‘little one čajík’
1;11.13 [jɪtujan tʃajɪːk] Lexical Word preference
“little one tea”
Slovak ‘ja mám tea’
1;11.14 [ja mam tiː] Lexical Word preference
“I have tea”
Indeterminate ‘medicine nie’
1;11.23 [mɛdɪsɪn ɲɪɛ] Lexical Bilingual context
“medicine no”
Indeterminate ‘daddy má chin’
1;11.27 [dædɪ maː tʃɪn] Lexical Bilingual context
“daddy has chin”
Lexical gap - self-repair of
[van ɔjɛxɔ. ɔjɛxɪ. van English ‘one orech-o. orechy. one orech’ grammatical morpheme,
1;11.27 Lexical
ɔjɛx.] “one nut. nuts. one nut” searching for the
Nominative singular form

285
Bibliography

Arnberg, L. (1987). Raising children bilingually: The pre-school years. Clevendon


Philadelphia: Multilingual Matters.
Baker, C. (2001). Foundations of Bilingual Education and Bilingualism (3rd ed.).
Clevedon, Buffalo, Toronto, Sydney: Multilingual Matters.
Baker, C. (2007). A Parents' and Teachers' Guide to Bilingualism (3rd ed.). Clevedon,
Buffalo, Toronto: Multilingual Matters.
Barron-Hauwaert, S. (2004). Language Strategies for Bilingual Families. Clevendon
Buffalo Toronto: Multilingual Matters LTD.
Bosch, L., & Sebastián-Gallés, N. (2001). Evidence of Early Language Discrimination
Abilities in Infants from Bilingual Environments. Infancy, 2(1), 29-49.
Burridge, K., & Kortmann, B. (Eds.). (2008). Varieties of English 3 The Pacific and
Australasia. Berlin New York: Mouton de Gruyter.
Cattell, R. (2007). Children's Language Consensus and Controversy (Revised Edition ed.).
London, New York: Continuum.
Chin, N. B., & Wigglesworth, G. (2007). Bilingualism: An Advanced Resource Book.
London & New York: Routledge.
Clark, E. V. (1987). The Principle of Contrast: A constraint on Language Acquisition. In
B. MacWhinney (Ed.), Mechanisms of Language Acquisition (pp. 1-33). Hillsdale,
NJ: Lawrence Erlbaum Associates.
Clark, E. V. (1988). On the logic of contrast. Journal of Child Language, 15(2), 317-335.
Clyne, M. G. (1987). Metalinguistic awareness in a bilingual child. Language Topics:
Essays in Honour of Michael Halliday
Clyne, M. G. (2005). Australia's Language Potential. Sydney: UNSW Press.
Comeau, L., Genesee, F., & Lapaquette, L. (2003). The Modelling Hypothesis and child
bilingual codemixing. International Journal of Bilingualism, 7, 113-126.
Comeau, L., Genesee, F., & Mendelson, M. (2007). Bilingual children's repairs of
breakdowns in communication. Journal of Child Language, 34, 159-174.
Conboy, B. T., & Thal, D. J. (2006). Ties between the lexicon and grammar: cross-
sectional and longitudinal studies of bilingual toddlers. Child Development, 77(3),
712-735.
Cruz-Ferreira, M. (2006). Three Is a Crowd? Acquiring Portuguese In a Trilingual
Environment. Clevendon, Buffalo, Toronto: Multilingual Matters LTD.
Cunningham-Andersson, U., & Andersson, S. (1999). Growing up with two langauges.
London & New York: Routledge.
de Boysson-Bardies, B. (2001). How Language Comes to Children. From Birth to Two
Years. (M. B. DeBevoise, Trans.). Cambridge, London: The MIT Press.
de Boysson-Bardies, B., & Vihman, M. M. (1991). Adaptation to Language: Evidence
from babbling and First Words in Four Languages. Language, 67(2), 297-319.
De Houwer, A. (1990). The acquisition of two languages from birth: a case study.
Cambridge: Cambridge University Press.
De Houwer, A. (1998). By way of introduction:Methods in studies of bilingual first
language acquisition. International Journal of Bilingualism, 2(3 special issue).
De Houwer, A. (2009a). Bilingual First Language Acquisition. Bristol Buffalo Toronto:
Multilingual Matters.

286
De Houwer, A. (2009b). An Introduction to Bilingual Development. Bristol Buffalo
Toronto: Multilingual Matters.
De Houwer, A., Bornstein, M. H., & De Coster, S. (2006). Early understanding of two
words for the same thing: A CDI study of lexical comprehension in infant
bilinguals. International Journal of Bilingualism, 10(3), 331-347.
Deuchar, M. (1999). Are function words non-language-specific in early bilingual two-
word utterances? Bilingualism: Language and Cognition, 2(1), 23-34.
Deuchar, M., & Quay, S. (1999). Language choice in earliest utterances: a case study with
methodological implications. Journal of Child Language, 26, 461-475.
Deuchar, M., & Quay, S. (2000). Bilingual Acquisition Theoretical Implications of a Case
Study. Oxford New York: Oxford University Press.
Döpke, S. (1986). Discourse structures in bilingual families. Journal of Multilingual and
Multicultural Development, 7(6), 493-507.
Döpke, S. (1988). The role of parental teaching techniques in bilingual German-English
families. International Journal of the Sociology of Language, 72, 101-112.
Döpke, S. (2001). The Interplay Between Language-Specific Development and Cross-
Linguistic Influence. In S. Döpke (Ed.), Cross-Linguistic Structures in
Simultaneous Bilinguals (pp. 79-103). Philadelphia: John Benjamins Publishing
Company.
Eilers, R. E., Pearson, B. Z., & Cobo-Lewis, A. B. (2006). Social factors in bilingual
development: The Miami experience. In P. McCardle & E. Hoff (Eds.), Childhood
Bilingualism (pp. 68-90). Clevendon, BUffalo, Toronto: MUltilingual Matters.
Ervin-Tripp, S., & Reyes, I. (2005). Child codeswitchin and adult content contrasts.
International Journal of Bilingualism, 9(1), 85-102.
Fernald, A. (2006). When Infants Hear Two Languages: Interpreting Research on Early
Speech Perception by Bilingual Children. In P. McCardle & E. Hoff (Eds.),
Childhood Bilingualism (pp. 19-29). Clevendon, Buffalo, Toronto: Multilingual
Matters.
Finch, G. (2005). Key Concepts in Language and Linguistics. London: Palgrave
MacMillan.
Garcia-Sierra, A., Rivera-Gaxiola, M., Percaccio, C. R., Conboy, B. T., Romo, H.,
Klarman, L., . . . Kuhl, P. K. (2011). Bilingual language learning: An ERP study
relating early brain responses to speech, language input, and later word production.
Journal of Phonetics, 39(4), 546-557.
Genesee, F. (2006). Bilingual First Language Acquisition in Perspective. In P. McCardle
& E. Hoff (Eds.), Childhood Bilingualism (pp. 45-67). Clevendon, Buffalo,
Toronto: Multilingual Matters.
Genesee, F. (2007). Early bilingual language development: one language or two? In L.
Wei (Ed.), The Bilingualism Reader (pp. 320-335). London and New York:
Routledge.
Genesee, F., & Cenoz, J. (2001). First Words. In J. Cenoz & F. Genesee (Eds.), Trends in
Bilingual Acquisition (Vol. 1, pp. 1-9). Amsterdam/Philadelphia: John Benjamins
Publishing Company.
Genesee, F., Nicoladis, E., & Paradis, J. (1995). Language differentiation in early bilingual
development. Journal of Child Language, 22, 611-631.
Goldin-Meadow, S. (1998). The Development of Gesture and Speech as an Integrated
System. New Directions for Child Developement(79), 29-42.
Goodwyn, S. W., Acredolo, L. P., & Brown, C. A. (2000). Impact of symbolic gesturing
on early language development. Journal of Nonverbal Behaviour, 24(2), 81-103.
Goodz, N. S. (1989). Parental Language Mixing in BIlingual Families. Infant Mental
Health Journal, 10(1), 25-44.
Grille, R. (2008). Heart to Heart Parenting. Sydney: ABC Books.
287
Grosjean, F. (1982). Life with Two Languages. Cambridge, Massachusetts, London:
Harvard University Press.
Grosjean, F. (1998). Studying bilinguals:Methodological and conceptual issues.
Bilingualism: Language and Cognition, 1, 131-149.
Grosjean, F. (2008). Studying Bilinguals. Oxford: Oxford University Press.
Grosjean, F. (2010). Bilingual. Cambridge, Massachusetts, London: Harvard University
Press.
Halliday, M. A. K. (1975). Learning how to mean - Explorations in the development of
language. London: Edward Arnold (publishers) Ltd.
Hiroko, K. (1998). Determinants of Language Choice in Bilingual Children. International
Journal of Bilingualism, 2(3 (special volume)).
Hlavatá, Z. (1998). Štatút detského slova v slovenčine. Paper presented at the VIII.
kolokvium mladých jazykovedcov, Modra-Piesok.
Hoffmann, C. (1991). An Introduction to Bilingualism. London & New York: Longman.
Holowka, S., Brosseau-Lapré, F., & Petitto, L. A. (2002). Semantic and Conceptual
Knowledge Underlying Bilingual Babies' First Signs and Words. Language
Learning, 52(2), 205-262.
Horňáková, K., Kapalková, S., & Mikulajová, M. (2005). Kniha o detskej reci. Bratislava:
Slniecko.
Hua, Z., & David, A. (2008). Study Design: Cross-sectional, Longitudinal, Case, and
Group. In L. Wei & M. G. Moyer (Eds.), Research Methods in Bilingualism and
Multilingualism (pp. 88-107): Blackwell Publishing.
Huotilainen, M. (2004). Foetal learning - a bridge over birth. Paper presented at the The
culture created by children and children's participation - Discussions on children's
peer relations, agency, play and narratives, National Research and Development
Centre for Welfare and Health (STAKES), Helsinki, Finland.
http://www.edu.helsinki.fi/lapsetkertovat/lapset/In_English/seminar.htm
Huttenlocher, J., Haight, W., Bryk, A., Seltzer, M., & Lyons, T. (1991). Early vocabulary
growth: Relation to Language Input and Gender. Developmental Psychology,
27(2), 236-248.
Ivanecký, J., & Nábělková, M. (2002). Fonetická transkripcia SAMPA a slovenčina.
Jazykovedný časopis, 53(2), 81-95.
Juan-Garau, M., & Pérez-Vidal, C. (2001). Mixing and pragmatic parental strategies in
early bilingual acquisition. Journal of Child Language, 28, 59-86.
King, K., & Fogle, L. (2006). Bilingual Parenting as Good Parenting: Parent's Perspectives
on Family Language Policy for Additive Bilingualism. The International Journal
of Bilingual Education and Bilingualism, 9(6), 695-712.
Kráľ, Á. (1984). Pravidlá slovenskej výslovnosti. Bratislava: Slovenské pedagogické
nakladateľstvo.
Kroll, J. F., Gerfen, C., & Dussias, P. E. (2008). Laboratory Designs and Paradigmes:
Words, Sounds, and Sentences. In L. Wei & M. G. Moyer (Eds.), Research
Methods in Bilingualism and Multilingualism (pp. 108-131): Blackwell Publishing.
Kuhl, P. K., Andruski, J. E., Chistovich, I. A., Chistovich, L. A., Kozhevnikova, E. V.,
Ryskina, V. L., . . . Lacerda, F. (1997). Cross-Language Analysis of Phonetic Units
in Language Addressed to Infants. Science, 277, 684-686.
Kwan-Terry, A. (1992). Code-switching and code-mixing: The case of a child learning
English and Chinese simultaneously. Journal of Multilingual and Multicultural
Development, 13(3), 243-259.
Lanvers, U. (2001). Language alternation in infant bilinguals: A developmental approach
to code-switching. International Journal of Bilingualism, 5, 437-464.
Lanza, E. (1992). Can bilingual two-year-olds code-switch? Journal of Child Language,
19, 633-658.
288
Lanza, E. (2001a). Bilingual first language acquisition: A discourse perspective on
language contact in parent-child interaction. In J. Cenoz & F. Genesee (Eds.),
Trends in Bilingual Acquisition (pp. 201-229). Amsterdam/Philadelphia: John
Benjamins Publishing Company.
Lanza, E. (2001b). Language Contact - A Dilemma for the Bilingual Child or for the
Linguist? In S. Döpke (Ed.), Cross-Linguistic Structures in Simultaneous
Bilinguals (pp. 227-245). Philadelphia: John Benjamins.
Lanza, E. (2004). Language Mixing in Infant Bilingualism A Sociolinguistic Perspective.
Oxford: Oxford University Press.
Lanza, E. (2008). Selecting Individuals, Groups, and Sites. In L. Wei & M. G. Moyer
(Eds.), Research Methods in Bilingualism and Multilingualism (pp. 73-87):
Blackwell Publishing.
Leopold, W. F. (1939). Speech Development of a Bilingual Child A Linguist's Record.
Evanston Chicago: Northwestern University.
MacWhinney, B. (2001). Last words. In J. Cenoz & F. Genesee (Eds.), Trends in Bilingual
Acquisition. Amsterdam/Philadelphia: John Benjamins Publishing Company.
MacWhinney, B. (2008). The CHILDES Project: Tools for Analysing Talk - Electronic
Edition. Retrieved 23/11/2008 http://childes.psy.cmu.edu/
Matychuk, P. (2005). The role of child-directed speech in language acquisition: a case
study. Language Sciences(27), 301-379.
McCardle, P., & Hoff, E. (Eds.). (2006). Childhood Bilingualism Research on Infancy
through School Age. Clevendon, Buffalo, Toronto: Multilingual Matters LTD.
Meisel, J. M. (1989). Early differentiation of languages in bilingual children. In K.
Hyltenstam & L. K. Obler (Eds.), Bilingualism across the lifespan (pp. 13-40).
Cambridge: Cambridge University Press.
Meisel, J. M. (2001). The simultaneous acquisition of two first languages: Early
differentiation and subsequent development of grammars. In J. Cenoz & F.
Genesee (Eds.), Trends in Bilingual Acquisition (Vol. 1, pp. 11-41).
Amsterdam/Philadelphia: John Benjamins Publishing Company.
Meisel, J. M. (2006). The Bilingual Child. In T. K. Bhatia & W. C. Ritchie (Eds.), The
Handbook of Bilingualism (pp. 91-113). Malden, Oxford & Carlton: Blackwell
Publishing.
Menyuk, P., & Brisk, M. E. (2005). Language Development and Education. Hampshire &
New York: Palgrave Macmillan.
Mistrík, J. (1988). Moderná slovenčina. Bratislava: Slovenské pedagogické
nakladateľstvo.
Nábělková, M. (2002, 22/2/2002). Medzi pasívnym a aktívnym bilingvizmom. Paper
presented at the Bilingvizmus: Minulosť, prítomnosť a budúcnosť, Bratislava.
Nicoladis, E. (1998). First clues to the existence of two input languages: Pragmatic and
lexical differentiation in a bilingual child. Bilingualism: Language and Cognition,
1, 105-116.
Nicoladis, E. (2001). Finding first words in the input. In J. Cenoz & F. Genesee (Eds.),
Trends in Bilingual Acquisition (pp. 131-147). Amsterdam/Philadelphia: John
Benjamins Publishing Company.
Nicoladis, E., & Genesee, F. (1996). A Longitudinal Study of Pragmatic Differentiation in
Young Bilingual Children. Language Learning, 46(3), 439-464.
Nicoladis, E., & Secco, G. (2000). The role of a child's productive vocabulary in the
language choice of a bilingual family. First Language, 20, 003-028.
Noguchi, M. G. (1996). The Bilingual Parent as Model for the Bilingual Child. Policy
Science, Mar 1996, 245- 261.

289
Ochs, E., & Schieffelin, B. (1995). The impact of language socialization on grammatical
development. In P. Fletcher & B. MacWhinney (Eds.), The Handbook of Child
Language (pp. 786). Oxford: Blackwell.
Oksaar, E. (1983). Language acquisition in the early years (K. Turfler, Trans.). London,
New York: Batsford Academic and Educational St. Martin's Press.
Oller, D. K., Eilers, R. E., Urbano, R., & Cobo-Lewis, A. B. (1997). Development of
precursors to speech in infants exposed to two languages. Journal of Child
Language, 24, 407-425.
Painter, C. (1985). Learning the mother tongue. Melbourne: Deakin University Press.
Painter, C. (1996). The development of language as a resource for thinking: a linguistic
view of learning. In R. Hasan & G. Williams (Eds.), Literacy in Society (pp. 50-
85). London and New York: Longman.
Painter, C. (2006). The 'interpersonal first' principle in child language development. In G.
Williams & A. Lukin (Eds.), The Development of Language (pp. 137). London &
New York: Continuum.
Paradis, J., Genesee, F., & Crago, M. (2011). Dual Language Development and Disorders:
a Handbook on Bilingualism and Second Language Learning.
Baltimore/London/Sydney: Paul H. Brookes Publishing Co.
Pearson, B. Z. (1998). Assessing lexical development in bilingual babies and toddlers.
International Journal of Bilingualism, 2(3).
Pearson, B. Z. (2008). Raising a Bilingual Child. New York: Living Language.
Pearson, B. Z., Fernández, S. C., & Oller, D. K. (1993). Lexical Development in Bilingual
Infants and Toddlers: Comparison to Monolingual Norms. Language Learning,
43(1), 93-120.
Petitto, L. A., Katerelos, M., Levy, B. G., Gauna, K., Tetreault, K., & Ferraro, V. (2001).
Bilingual signed and spoken language acquisition from birth: implications for the
mechanisms underlying early bilingual language acquisition. Journal of Child
Language, 28, 453-496.
Petitto, L. A., & Marentette, P. F. (1991). Babbling in the Manual Mode: Evidence for the
Ontogeny of Language. Science, New Series, 251(5000), 1439-1496.
Pruden, S. M., Hirsh-Pasek, K., & Golinkoff, R. M. (2006). The Birth of Words: Ten-
Month-Olds Learn Words Through Perceptual Salience. Child Development, 77(2),
266-280.
Quay, S. (1992). Explaining Language Choice in Early Infant Bilingualism. Paper
presented at the Sociolinguistics Symposium (9th), University of Reading,
England, United Kingdom.
Quay, S. (1995). The bilingual lexicon: implications for studies of language choice.
Journal of Child Language, 22, 369-387.
Quay, S. (2001). Managing linguistic boundaries in early trilingual development. In J.
Cenoz & F. Genesee (Eds.), Trends in Bilingual Acquisition.
Amsterdam/Philadelphia: John Benjamins Publishing Company.
Quay, S. (2008). Dinner conversations with a trilingual two-year-old: Language
socialization in a multilingual context. First Language, 28(5), 5-33.
Redlinger, W. E., & Park, T.-Z. (1979). Language mixing in young bilinguals. Journal of
Child Language, 7, 337-352.
Ripka, I. (1992a). Lexikálne prevzatia v nárečových prejavoch amerických Slovákov.
Slovenská reč, 57(6), 346-351.
Ripka, I. (1992b). Príspevok k štúdiu nárečových prejavov amerických Slovákov.
Slovenská reč, 57(4), 207-213.
Romaine, S. (1995a). The Bilingual Child Bilingualism (pp. 181-240). Oxford &
Cambridge: Blackwell.
Romaine, S. (1995b). Bilingualism. Oxford & Cambridge: Blackwell.
290
Saunders, G. (1984). Bilingual children: guidance for the family. Clevendon: Multilingual
Matters.
Serratrice, L. (2001). The emergence of verbal morphology and the lead-lag pattern issue
in bilingual acquisition. In J. Cenoz & F. Genesee (Eds.), Trends in Bilingual
Acquisition (pp. 43-70). Amsterdam/Philadelphia: John Benjamins Publishing
Company.
Serratrice, L. (2005). Language mixing and learning strategy. International Journal of
Bilingualism, 9(2), 159-177.
Sgall, P. (1999). Prague School of Typology. In M. Shibatani & T. Bynon (Eds.),
Approaches to Language Typology (pp. 49-84). Oxford: Oxford University Press.
Sinka, I., & Schelletter, C. (1998). Morphosyntactic development in bilingual children.
International Journal of Bilingualism, 2(3).
Štefánik, J. (2000). Bilingvizmus na pozadí dvoch morfologicky odlišných typov jazykov.
Bratislava: Vydavateľstvo Univerzity Komenského.
Stejskalová, M. (1998). Otázky kódovania v detskej reči. Paper presented at the VIII.
kolokvium mladých jazykovedcov, Modra-Piesok.
Sunderland, M. (2006). The Science of Parenting. New York: DK Publishing.
Swain, M. (1976). Bilingual first-language acquisition. In W. von Raffler-Engel & Y.
Lebrun (Eds.), Baby Talk and Infant Speech. Amsterdam: Swets & Zeitlinger.
Tamis-Lemonda, C. S., Bornstein, M. H., & Baumwell, L. (2001). Maternal
Responsiveness and Children's Achievement of Language Milestones. Child
Development, 72(3), 748-767.
Tamis-Lemonda, C. S., Bornstein, M. H., Kahana-Kalman, R., Baumwell, L., & Cyphers,
L. (1998). Predicting variation in the timing of language milestones in the second
year: an events history approach. Journal of Child Language, 25, 675-700.
Tokuhama-Espinosa, T. (2001). Raising Multilingual Children. Westport, London: Bergin
& Garvey.
Tomasello, M. (2003). Constructing a Language. A Usage-Based Theory of Language
Acquisition. Cambridge, Massachusetts and London: Harvard University Press.
Tomasello, M. (2012). The usage-based theory of language acquisition. In E. L. Bavin
(Ed.), The Cambridge Handbook of Child Language (pp. 69-87). Cambridge:
Cambridge University Press.
Tomasello, M., & Farrar, M. J. (1986). Joint Attention and Early Language. Child
Development, 57, 1454-1463.
Tough, J. (1977). The Development of Meaning. London: George Allen & Unwin LTD.
Tracy, R. (2001). Language Mixing as a Challenge for Linguistics. In S. Döpke (Ed.),
Cross-Linguistic Structures in Simultaneous Bilinguals (pp. 11-36). Philadelphia:
John Benjamins Publishing Company.
Vihman, M. M. (1985). Language differentiation by the bilingual infant. Journal of Child
Language, 12, 297-324.
Vihman, M. M. (2002). Getting started without a system: From phonetics to phonology in
bilingual development. The International Journal of Bilingualism, 6(3), 239-254.
Vihman, M. M., & McCune, L. (1994). When is a word a word? Journal of Child
Language, 21, 517-542.
Volterra, V., & Taeschner, T. (2007). The acquisition and development of language by
bilingual children. In L. Wei (Ed.), The Bilingualism Reader (pp. 304-319).
London and New York: Routledge.
Wanner, P. J. (1996). A Study of the Initial Codeswitching Stage in the Linguistic
Development of an English-Japanese Bilingual Child. Japan Journal of
Multilingualism and Multiculturalism, 2(1), 72-96.

291
Yow, W. G., & Markman, E. (2007). Monolingual and bilingual children's use of mutual
exclusivity assumption and pragmatic cues in word learning. Paper presented at the
Biennal Meeting of the Society for Research on Child Development, Boston, MA.

292

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy