Juggling Two Languages: How Bilingual Children Assert Themselves

in Family Conversation

Elisabeth DEMOURY N° 21002745

Présenté à l’Université Sorbonne Nouvelle – Paris III Département du Monde Anglophone

Mémoire de Master 2 Recherche Master Langues, Littératures et Civilisations Étrangères et Régionales (LLCER)

Études Anglophones

Mémoire dirigé par Madame Aliyah MORGENSTERN Professeur des Universités, Université Sorbonne Nouvelle – Paris III Membre du jury Monsieur Éric CORRE Professeur des Universités, Université Sorbonne Nouvelle – Paris III

Année universitaire 2014 - 2015

1

Juggling Two Languages: How Bilingual Children Assert Themselves

in Family Conversation

Elisabeth DEMOURY N° 21002745

Présenté à l’Université Sorbonne Nouvelle – Paris III

Département du Monde Anglophone

Mémoire de Master 2 Recherche

Master Langues, Littératures et Civilisations Étrangères et Régionales (LLCER)

Spécialité Études Anglophones

Année universitaire 2014 - 2015

2

3

ACKNOWLEDGEMENTS

I would first like to give my sincere acknowledgements to Thomas, Oliver and

James, whose liveliness and spontaneity made this study very pleasant. I would also like

to thank their parents for allowing me to enter their private lives and to undertake this

research. I am grateful to them for sharing their family history and experience in raising

their children bilingually. Were it not for their friendly cooperation, this study would have

never been possible.

I am extremely grateful to my teacher, Ms. Aliyah Morgenstern, for all her

guidance and kindness throughout these two years of academic studies. Her support and

her assistance have been a tremendous motivation for me to write this dissertation and

have always pushed me forward.

I would also like to thank Ms. Frédérique Atangana and Mr. Christophe Parisse

who so generously helped with the obscure functioning of Computerized Language

Analysis. Their availability and patience have been a valuable help for this study.

Finally, I wish to thank my friends for their unfailing confidence in me and in the

success of my project. I am especially grateful to Clara, Juliette and Alberic for their

unconditional support and for always believing in me.

4

ABSTRACT

The present dissertation finds its roots in a multicultural and multilingual context.

It deals with the way bilingual speakers manipulate two languages as part of the same

conversation. This research is based upon video recordings of family dinners and attempts

to examine the use of French and English by three bilingual children, and to clarify the

motivations behind the children’s language alternations. Using the method of

conversation analysis, this study focuses on the family’s interactive construction of

everyday interchanges. It thus presents language alternations as an intentional process and

argues that language mixing in bilingual children actually reflects an additional

communicative competence. After demonstrating that the three children are dominant in

French, the qualitative analysis of the data has revealed that the speakers tend to switch

languages when they want to either blend into conversation or mark a rupture to signal

their disagreement or request attention. We also underline a possible association of one

language with one interlocutor and suggest a potential link between the use of one

language and the emotional state of the speaker. Mixing two languages is therefore

considered a powerful and constructive device to assert oneself in family conversation.

Keywords: bilingualism, bilingual acquisition, language mixing, family dinners,

communicative competence.

5

RÉSUMÉ

Ancré dans un contexte multiculturel et multilingue, ce mémoire porte sur le

mélange des langues dans la conversation bilingue. L’analyse des interactions au cours de

repas familiaux a permis de s’intéresser aux productions langagières de trois enfants

bilingues (Français-Anglais) et d’éclaircir les motivations à l’origine de leur passage

d’une langue à l’autre dans la conversation. Cette étude présente donc le phénomène

d’alternance des langues comme un procédé volontaire et maîtrisé, témoin de la

compétence communicative des locuteurs bilingues. Après avoir déterminé une

dominance en Français chez les trois enfants, une étude qualitative a permis de dégager

plusieurs tendances prédominantes dans les alternances des locuteurs. Ainsi, changer de

langue s’avère un précieux outil aussi bien pour s’intégrer efficacement à une

conversation et faire entendre sa voix, que pour créer une rupture volontaire dans

l’interaction. Les analyses montrent que l’on peut aussi supposer une association tacite

entre un locuteur et sa langue maternelle, ainsi qu’une dimension affective liée à chacune

des langues. Jongler avec deux langues se révèle alors une pratique incontournable dans

la parole bilingue afin de s’imposer dans la conversation familiale.

Mots-clés: bilinguisme, acquisition bilingue, alternance des langues, diners familiaux,

compétence communicative.

6

TABLE OF CONTENTS

Acknowledgements ................................................................................................................ 3 Abstract .................................................................................................................................. 4 Résumé ................................................................................................................................... 5 Table of Contents ................................................................................................................... 6 List of Figures and Tables ...................................................................................................... 9

I. INTRODUCTION ............................................................................................................ 11 II. THEORETICAL BACKGROUND ................................................................................ 15 1. Bilingualism ..................................................................................................................... 15

1. 1. What is bilingualism? .............................................................................................. 15 1. 1. 1. Definition(s) ..................................................................................................... 15 1. 1. 2. The continuum of bilingualism ........................................................................ 16

1. 2. Different types of bilingualism ................................................................................ 17 1. 2. 1. Early and late bilinguals ................................................................................... 17 1. 2. 2. Balanced and dominant bilinguals ................................................................... 19 1. 2. 3. Compound, coordinate and subordinate bilinguals .......................................... 19 1. 2. 4. Folk and elite bilinguals ................................................................................... 20 1. 2. 5. Additive and subtractive bilinguals ................................................................. 21

1. 3. Code-switching, code-mixing or language mixing? ................................................ 22 1. 4. Concluding remarks ................................................................................................. 24

2. Bilingual Acquisition ....................................................................................................... 25 2. 1. Simultaneous acquisition of two languages ............................................................. 25

2. 1. 1. One system or two? .......................................................................................... 25 2. 1. 2. Issues of delay and proficiency ........................................................................ 26

2. 2. Sequential acquisition of two languages .................................................................. 27 2. 3. Communicative competence .................................................................................... 28 2. 4. Bilingual learners ..................................................................................................... 30 2. 5. Concluding remarks ................................................................................................. 30

3. Studies on Family Dinners ............................................................................................... 31 3. 1. Communication time ................................................................................................ 31 3. 2. The weight of culture: internalizing norms and rules .............................................. 32

4. The Conversation Analytical Approach ........................................................................... 34 4. 1. Why conversation analysis? ..................................................................................... 34 4. 2. Main characteristics ................................................................................................. 35

III. THE DATA .................................................................................................................... 39 1. Presentation of the corpus ............................................................................................ 39

1. 1. Family background and history ........................................................................... 39 1. 2. The children ......................................................................................................... 41

2. Recording ..................................................................................................................... 45 2. 1. Collection of the data ........................................................................................... 45 2. 2. Why family dinners? ............................................................................................ 46 2. 3. The videos ............................................................................................................ 46 2. 4. Technical difficulties ........................................................................................... 49

7

3. Coding ......................................................................................................................... 50 3. 1. Coding choices .................................................................................................... 50 3. 2. Transcription problems ....................................................................................... 52

4. Concluding remarks .................................................................................................... 54

IV. QUANTITATIVE RESULTS. ...................................................................................... 57 1. General linguistic overview ........................................................................................ 57 2. Lexical diversity (Type/Token Ratio) ......................................................................... 58 3. Mean Length of Utterance (MLU) .............................................................................. 61 4. Mean Length of Turn (MLT) ...................................................................................... 64 5. The issue of gestures ................................................................................................... 69 6. Concluding remarks .................................................................................................... 73

V. THE ALTERNATE USE OF TWO LANGUAGES ..................................................... 75 1. Family conversation: overview of a selected extract .................................................. 75 2. Switching languages to blend into conversation ......................................................... 82

2. 1. How children adapt and jump in conversation .................................................... 82 2. 2. How parents deal with language changes ........................................................... 91

3. Switching languages to create a break in conversation ............................................... 99 4. One interlocutor = one language? ............................................................................. 104 5. The language(s) of emotions ..................................................................................... 112 6. Concluding remarks .................................................................................................. 114

VI. GENERAL CONCLUSION ....................................................................................... 117

Bibliography ...................................................................................................................... 121

Appendix 1 – Filming Authorization ................................................................................ 124 Appendix 2 - Transcription Conventions .......................................................................... 127 Appendix 3 - Transcription Extract 01.31.2015 ................................................................ 128 Appendix 4 – Recap Charts: Language Alternations ........................................................ 131

8

!!

!

!

!

!

!

!

!

!

!

!

!

!

!

!

!

!

!

!

!

!

!

!

!

!

!

!

!

!

!

!

9

Figure 1: Compound and coordinate bilinguals ................................................................ 20

Figure 2: Distribution of linguistic items .......................................................................... 58

Figure 3: Comparison of the children’s MLU ................................................................... 63

Figure 4: Comparison of the children’s MLT ................................................................... 68

Figure 5: Distribution of the children’s gestures ............................................................... 72

Table 1: Presentation of the Recordings ............................................................................ 47

Table 2: Number of Linguistic Items (Tokens) ................................................................. 57

Table 3: Number of Different Words (Types) .................................................................. 58

Table 4: Lexical diversity in French (Carroll’s measure) ................................................. 59

Table 5: Lexical diversity in English (Carroll’s measure) ................................................ 60

Table 6: Global Mean Length of Utterance ...................................................................... 61

Table 7: Mean Length of Utterance in French .................................................................. 62

Table 8: Mean Length of Utterance in English ................................................................. 62

Table 9: Adults’ Mean Length of Utterance ..................................................................... 63

Table 10: Global Mean Length of Turn ............................................................................ 65

Table 11: Mean Length of Turn in French ........................................................................ 66

Table 12: Mean Length of Turn in English ....................................................................... 67

Table 13: Gestures per Language (%GPX) ....................................................................... 70

Table 14: Tokens/Gestures ................................................................................................. 71

Table 15: Overlaps Originated by Each Participant .......................................................... 75

Table 16: Projectabilities ................................................................................................... 78

Table 17: Repairs .............................................................................................................. 79

LIST OF FIGURES AND TABLES

10

11

I. INTRODUCTION

“You can never understand one language until you understand at least two”,

according to American philosopher John Rogers Searle. Surely, bilingualism and

multilingualism have nowadays become a worldwide phenomenon and bear testament to

people’s greater mobility across all parts of the globe. Individuals tend to be more and

more exposed to different cultures around the world along with different languages,

which consequently results in a higher number of cross-cultural and cross-linguistic

relationships. The present dissertation finds its roots in this multicultural and multilingual

context. It deals with the way bilingual speakers manipulate their two languages as part of

the same conversation. Indeed, when living in bilingual communities, speakers have a

choice to make when it comes to the way they wish to express themselves. This

phenomenon first struck me while participating in an academic exchange in Canada, an

officially bilingual country. I found it fascinating how two languages could be

interconnected and complementary; and how bilingual speakers may alternate from one

language to another without any discontinuation in their discourse. It became apparent

that having two languages at one’s disposal could be a powerful device for speakers when

conducting a conversation.

This dissertation examines the acquisition of English and French in three bilingual

children within a family context and attempts to clarify the motivations behind the

children’s language alternations. The present research is based upon video recordings of

family dinners. As it will be detailed in this dissertation, mealtimes indeed constitute a

privileged moment in the family’s daily routine when the children learn to coordinate

food consumption with conversational activity. When looking at bilingual children’s

behavior, it is interesting to study families’ everyday practices since they create a good

opportunity for the children to express their desires, claim their opinions and affirm their

choices. In this context, the purpose of this dissertation is to answer the following

research question: How do bilingual children assert themselves in family conversation? In

other words, how do they juggle their two languages in order to establish their place and

mark their individuality? In order to answer these questions, the present study uses the

method of conversation analysis and focuses on the family’s interactive construction of

everyday interchanges. The data exploitation also made it possible to answer a series of

questions inherent to the issue of bilingualism in family conversation: Are the children

12

dominant in one of their languages? Does each language have a specific function in

communication? To what extent are the children’s switches related to their interlocutor or

to the topic that is being addressed? How paramount is the role of the parents in their

children’s way of dealing with two languages? The present study tends to demonstrate

that children are masters of their own choices in bilingual communication and that

switching from one language to another may allow them to gradually assert their identity

within their family conversations.

Indeed, as the analyses revealed in the present research, language alternations

result from the children’s initiatives and are voluntarily used to various purposes in

conversation. They can allow the participants to join and adapt to an ongoing

conversation. But they can also be useful to create intentional breaks and mark a point of

rupture in an interaction. This study thus argues that the fact that bilingual children mix

and alternate their languages actually reflects an additional communicative competence.

The arguments developed in the present dissertation are supported by the study of

three bilingual children, respectively at the ages of 5;11, 4;6 and 2;3 at the time of the

recording. It was important to me to account for the children’s language practices in their

regular environment. Indeed, a number of early studies have already shown that

naturalistic settings offer rich insights into children’s linguistic patterns along with their

daily customs and habits. Nevertheless, it should be noted that the brevity of this research

prevents us from making definitive assertions about the linguistic development of the

children observed, as their competences vary depending on their specific environment and

evolve throughout their lives. However, this study aims at giving a small contribution to

the field of bilingualism by presenting language alternations as an intentional process

showing the children’s communicative competence and their uniqueness as bilingual

learners.

The first chapter in this dissertation presents an overview of the major issues in

bilingualism and deals with its multiple dimensions and classifications. It also accounts

for various theories on bilingual acquisition along with some questions on delay and

proficiency when learning two languages, either simultaneously or sequentially. This

section also puts a particular emphasis on the study of family dinners and details the

approach and characteristics of conversation analysis.

The second chapter outlines the methodology that was used for this study,

including the collection of the data and the transcribing of the recordings, along with the

13

difficulties encountered and the theoretical issues related to the gathering of naturalistic

data. It also provides a description of the family background and the children’s language

environment.

The third section presents a quantitative analysis of the children’s use of their two

languages. In this chapter, a particular attention is paid to the boys’ linguistic abilities,

their lexical diversity and the mean length of their utterances and turns. As part of the

multimodal dimension of conversation, the issue of gestures is also addressed in order to

determine whether the children use more gestures in one language to compensate a

potential linguistic dominance in their other language.

Finally, the fourth section presents a qualitative analysis of the corpus in terms of

language alternations. It first proposes an overview of how a family conversation

functions, based on the method of conversation analysis. It then offers a closer look at

some selected extracts that are interesting for they offer a window onto the children’s way

of dealing with two languages. This chapter shows that bilingual communication takes

place through various mechanisms and that all co-participants are masters of their

switches in languages in order to assert their place within family conversations.

14

15

II. THEORETICAL BACKGROUND Before introducing the case study, let us present an overview of the major issues

in bilingualism and bilingual acquisition.

1. Bilingualism 1. 1. What is bilingualism? 1. 1. 1. Definition(s) Although the concept of bilingualism does not seem to raise specific problems at

first sight, it actually reflects a complex reality. Bilingualism is commonly referred to as

the ability to understand and speak two languages. According to Webster’s dictionary

(1961), a bilingual person is defined as “having or using two languages especially as

spoken with the fluency characteristic of a native speaker”; and bilingualism as “the

constant oral use of two languages”.

Early studies were hence guided by the popular view that bilinguals had to be

equally fluent in both of their languages. This is the approach presented by Bloomfield,

who defines bilingualism as “the native-like control of two languages” (1933:56).

In contradistinction to this definition, which includes only “perfect bilinguals”,

Macnamara (1967a) proposes that a bilingual person is anyone who possesses a minimal

competence in only one of the four language skills: listening comprehension, speaking,

reading and writing, in a language other than his mother tongue.

Between these two extremes, later researchers began to acknowledge that a perfect

type of bilingualism is rarely attained and that the fundamental principle is the regular

practice of both languages rather than their perfect knowledge (Grosjean, 2008). Indeed,

according to Grosjean, equal fluency in both languages is rarely achieved simply because

“the needs and uses of the two languages are usually quite different” (2008:14). Elizabeth

Lanza is also in agreement with this claim in Language Mixing in Infant Bilingualism – A

sociolinguistic perspective. Indeed, she states that by learning two languages, bilinguals

are “not only acquiring two linguistic systems but (…) more generally acquiring

communicative competence, as he or she develops a social identity simultaneously with

the development of language” (Lanza, 1997:7).

Claude Hagège (2005) shares this point of view as well, yet he has an interesting

perspective on the definitions of bilingualism. He suggests that a perfectly bilingual

person is mostly considered as such in the declarations of monolinguals who are surprised

16

to hear someone speak in a foreign language that sounds inaccessible to them (Hagège,

2005:245). Hence bilingual people would be qualified as such from a biased point of view

since definitions result mainly from monolinguals.

All these definitions, which range from native-like competence in two languages

to a minimal proficiency in a second language, raise a number of theoretical and

methodological difficulties. Indeed, they lack precision and clarity, as they do not specify

what is meant by native-like competence, which varies considerably within a unilingual

population; nor by minimal proficiency in a second language. Can we exclude from the

definitions of bilingual someone who possesses a very high competence in a second

language without necessarily being perceived as a native speaker on account of a foreign

accent? Can a person who has a few notions in a foreign language but is not able to use it

in communicative situations legitimately be called bilingual? In order to help ranging

various levels of bilingualism, some researchers suggest using a continuum. The

following subsection discusses this notion.

1. 1. 2. The continuum of bilingualism

Bilingualism, as simply defined as the use of two languages by an individual, is

best regarded as occurring on a continuum (Baetens Beardsmore, 1986). At the one end of

the continuum is the monolingual speaker. At the other end, we find the individual who

has acquired both languages in naturalistic contexts in childhood, and who is described as

having equal and native-like fluency in both his first language (L1) and second language

(L2). Along this continuum are those individuals who illustrate greater or lesser degrees

of bilingualism, involving the bilinguals who speak both languages fluently and

proficiently but are more dominant in one than the other; and the adult second language

learners, with varying degrees of proficiency and mastery of the second language.

Furthermore, François Grosjean claims that bilinguals find themselves at various

points on a “language mode continuum” (Grosjean, 1994). At the one end of Grosjean’s

continuum, bilinguals are in a totally monolingual language mode in that they are

interacting with monolinguals of one or the other of the languages they know. One

language is active and the other is deactivated. At the other end of this continuum,

bilinguals are in a bilingual language mode in that they are communicating with other

bilinguals who share their two languages and with whom they know they can mix

languages. In this case, both languages are active. These are end points but bilinguals also

17

find themselves at intermediary points depending on such factors as interlocutor,

situation, content of discourse, topic, and function of the interaction.

Attempts have already been made to classify bilinguals according to their degree

of fluency and competence in the languages that are spoken, simply by age, context,

manner of acquisition of the languages, or by hypothesized language representations. The

subsequent section will discuss the different types of bilingualism in accordance with the

children on whom the analyses in this dissertation are based.

1. 2. Different types of bilingualism Bilingualism has multiple dimensions. There are various components and aspects

associated with the complexity of understanding bilingualism. Highlighting the multi-

dimensionality of bilingualism, researchers have suggested various classifications

determined by the different aspects they focused on. Using this overview, I will try to

clarify the type(s) of bilingualism into which the children who are part of this research

project may fit.

1. 2. 1. Early and late bilinguals

Bilinguals can be categorized into early and late bilinguals, according to the age of

exposure to two or more languages. Early bilingualism is defined as the acquisition of

more than one language in the pre-adolescent phase of life (Baetens Beardsmore,

1986:28). On the other hand, late bilingualism refers to the bilinguals who have learned

their second language after what Lenneberg called the “critical period”, which refers to a

stage after which language acquisition occurs with more difficulty (Lenneberg, 1967). For

example, this occurs when a second language is learned in adulthood or adolescence.

According to Lenneberg’s hypothesis, after that critical period, native-like fluency will

only be attained with difficulty.

At the time of the recording, all three children are under the age of six. Therefore,

they can be considered as being early bilinguals since they acquired both languages

before this critical period, basically before puberty. More recent researchers have

suggested that the limit period should be placed even earlier. De Houwer (2009), for

example, argues that bilingual acquisition includes the input of both languages within the

first week of life.

18

Early and late bilinguals are distinguished according to their attainment of

linguistic competence. Early bilinguals are mainly regarded as attaining native-like

competence in both languages, whereas most late bilinguals are regarded as not having

attained the complete competence of L2 (Baetens Beardsmore, 1986). As already pointed

out in the first part of this chapter, the notion of native-like competence is difficult to

define, especially since it varies considerably within a monolingual population.

Therefore, one cannot claim for certain that all three children will grow up to attain a

native-like competence in both languages, nor can it be said that they will not.

Furthermore, early bilingualism can also be classified into two types:

simultaneous vs. sequential early bilingualism. Simultaneous early bilingualism occurs in

situations when a child learns two languages at the same time, from birth. Sequential

early bilingualism occurs in situations when a child who has already partially acquired a

first language then learns a second language early in childhood. An example can be when

a child moves to another place where the dominant language is not his native language.

This is exactly the case for the first two boys who are part of this research. After living in

New Zealand for four years, in the case of Thomas, and two years and nine months in the

case of Oliver, the children moved to France where they had to learn French in order to

attend school and adapt to their new environment. The dominant language around them

was not their native language anymore. Nevertheless, it is still considered ‘early’ as none

of the children was in their pre-adolescent phase yet. The case of the third boy, James, is

slightly different. James was only 5 months old when the family moved to France. Can

we consider that his acquisition was sequential because he was not exposed to both

languages from birth? Or is it appropriate to say that he acquired both languages

simultaneously? We could even wonder whether his dominant language would differ

from his brothers’ since James did not have the time to fully acquire English before

moving to France.

In the case of late bilingualism – which can be considered as sequential – the

acquisition of a second language takes place after the full acquisition of a first language.

In this type of bilingualism, since the bilinguals have already acquired a first language,

they tend to use their experience to learn the L2. Because the children under study in this

dissertation were all under six at the time of the recording of their conversations, they can

hardly be considered as late bilingual individuals. Nevertheless, this does not exclude the

19

possibility that they might have begun to experience French through their knowledge of

English.

1. 2. 2. Balanced and dominant bilinguals

The distinction between balanced and dominant (or unbalanced) bilinguals (Peal

and Lambert, 1962) is based on the relationship between the fluency and proficiencies of

the respective languages which bilinguals master. Those who acquire similar degrees of

proficiency and mastery in both languages are defined as balanced bilinguals, while on

the contrary, dominant (or unbalanced) bilinguals are those whose proficiency in one

language is higher than in the other language. In other words, in dominant bilingualism,

the individual is more proficient and competent in one of the two languages; while

balanced bilingual people are more or less equally competent in the two languages that

they speak.

One of the purposes of this research project is to find out whether the children are

equally competent in their two languages or whether they have a dominant language. For

practical purposes, I started from the assumption that the children were dominant in

French. Indeed, the software that I used for coding and analyzing the data required the

participants to have a dominant language. Because the children have now been living in

France for two years, and because it is the language of their mother, I speculated that they

were dominant in French. On this matter, I was at variance with the mother’s opinion.

Indeed, the mother considers that her boys are all Anglophones. So if the children do have

a dominant language, is it the language they first started acquiring before emigrating? Or

is it the language that is dominant in the society in which they dwell today? In order to

answer these questions, I will present some quantifiable results later in this study,

looking, among other things, at the children’s mean length of utterances and lexical

diversity in each language.

1. 2. 3. Compound, coordinate and subordinate bilinguals

According to Weinreich (1953), compound, coordinate and subordinate

distinctions deal with the way two or more linguistic codes are organized and stored by

individuals. In compound bilinguals, two sets of linguistic codes (e.g. “Dog” and

“Chien”) are stored in one meaning unit. In other words, there is only one semantic

system for words, which is used for both L1 and L2. On the contrary, in coordinate

20

bilinguals, each linguistic code is stored and organized separately in two meaning units

and the bilinguals have two semantic systems for words. One system of meaning is for

words that the individuals know in their L1, and the other is for words they know in their

L2. The following figure (1) clarifies this point:

COMPOUND COORDINATE

Figure (1). Compound and Coordinate Bilinguals

Furthermore, in subordinate bilinguals, linguistic codes are organized in one single

meaning unit but the second language is assumed to be understood and interpreted

through the first language. In other words, there are two sets of linguistic codes but only

one meaning unit, which is accessible merely through the L1.

I will try to find some extracts in my data that could suggest whether the

children’s bilingualism is compound, coordinate or subordinate. As I have argued before

in this chapter, their bilingualism could be classified as ‘early sequential’. Therefore, we

can wonder whether each of their linguistic codes is stored and organized separately in

two meaning units; or if they are attached to only one semantic unit. In the latter case, are

the linguistic items part of a compound system, or does the child interpret his second

language through his first language? In other words, is there a hierarchy in the child’s

languages?

1. 2. 4. Folk and elite bilinguals

There are still several other classifications of bilinguals, which are dependent on

other variables, such as cultural identity and language usage. Along with the different

types of individual variables put forward earlier in this dissertation, bilinguals can also be

classified depending on diverse social variables.

Semantic

System Semantic

System 1

Semantic

System 2

English Français English Français

21

Concentrating on the social status of language, Fishman (1977) states that

bilinguals can be classified into “folk” and “elite” bilinguals. Folk bilinguals refer to

groups of individuals whose minority language does not have a high status in the

predominant language society in which they reside. In contrast, elite bilinguals refer to

those who speak a dominant language in a given society and who can also speak another

language which provides them with “additional value” within the society in which they

live (Fishman, 1977).

Since English is recognized as an indispensable and unavoidable competence

around the world, we can consider the family members as being ‘elite’ bilinguals. Indeed,

their minority language does have a high status in the French society. Moreover, it

definitely provides them with ‘additional value’ within France as well as within Europe.

However, acquiring a second language may sometimes come at the cost of losing some

proficiency in one’s first language. This issue is approached in the following subsection.

1. 2. 5. Additive and subtractive bilinguals

Depending on how one’s second language (L2) influences the retention of one’s

first language (L1), bilinguals can be classified into additive or subtractive bilinguals.

This classification emerges from Lambert (1974). According to Lambert, bilinguals who

can improve their L2 without losing their L1 proficiency are called additive bilinguals.

On the contrary, those whose L2 is acquired at the cost of losing their L1 can be called

subtractive bilinguals. In order to be additive bilinguals, both of the languages learned by

the individuals should be valued in the society in which they live. In other words, when

learning a second language does not interfere with the L1 learning, both languages can

develop; which is considered as an additive bilingualism. In contrast, subtractive

bilingualism occurs when learning a second language interferes with the L1 learning. In

this case, the mastery of L1 diminishes while the proficiency of L2 augments.

Consequently, the first language could ultimately be replaced by the second language

(Lambert, 1974).

Looking at quantifiable results, the lexical diversity and mean length of utterances

in each language should reveal whether the children have acquired French at the cost of

losing their proficiency in English. A priori, because both French and English are valued

in the French society and because the children are still regularly exposed to the two

languages, it seems the children in this dissertation may belong to the category of additive

22

bilinguals. This hypothesis is yet to be verified and does not exclude a partial loss of

proficiency or mastery in English.

Because bilingual speakers have two languages at their disposal when they

converse, they often tend to mix both languages. It can either be on a sentence level or it

can be within one single utterance. The following section discusses the terminology

linked to the various types of language alternations.

1. 3. Code-switching, code-mixing or language mixing?

The term code-switching is broadly discussed and thus burdened with the issue of

terminological confusion. Not all researchers agree on the territory covered by terms such

as code-switching, code-mixing, code-alternation or language mixing (Boztepe, 2003).

Much of the work labeled code-switching is interested in syntactic and morphosyntactic

constraints on language alternation (e.g. Poplack, 1980; Sankoff & Poplack, 1981; Joshi,

1985; Di Sciullo & Williams, 1987; Belazi et al., 1994; Halmari 1997 inter alia; as cited

in Nilep, 2006). Alternatively, studies of language acquisition use the term code-

switching to describe either bilingual speakers’ or language learner’s linguistic abilities

(e.g. Romaine, 1989; Cenoz & Genesee, 2001; Fotos 2001, inter alia; as cited in Nilep,

2006). The term code, on the other hand, seems to be a relatively neutral

conceptualization of a linguistic variety – either a language or a dialect (Romaine, 1995;

as cited in Boztepe, 2003).

According to Lanza’s definition, code-switching has become widely accepted as

“the use of two or more languages at the word or utterance level and across utterances in

a single conversation” (Lanza, 1997:1). Yet, definitions usually vary across studies.

Milroy and Muysken simply define code-switching as “the alternative use by bilinguals of

two or more languages in the same conversation” (1995:7; as cited in Boztepe, 2003:4).

For instance, in one of the recordings in this study, the first child Thomas answers

“because” – twice – to his mother who is asking him “pourquoi?”. This is an example of

a simple use of two languages in the same conversation. Some researchers (e.g. Auer,

1995) use the term code-alternation as a hyponym to replace code-switching, but it is

marginally used in that sense. Moreover, the influence of the one-system hypothesis (see

2.1.1.) has led to the use of code-mixing or language mixing as a reference to the

indiscriminate use of two languages, thus implying one single system. In the same vein,

Elizabeth Lanza uses language mixing as a generic term to describe any contact between

23

two or more languages, in which code-switching is treated as one type of language mixing

(Lanza, 1997). For example, during one dinner, the mother asks her second boy, Oliver,

“(do) you like fish?”, to which the child answers “j’aime bien attraper le poisson qui est

dans l’eau”. If we refer to Lanza’s definition, this is an example of language mixing since

the two languages interact in the same conversation.

Furthermore, Poplack (1980) distinguishes between intra-sentential and inter-

sentential code-switching. Intra-sentential code-switching involves a shift in languages

within one utterance. For example, when the first-born child, Thomas, says

“no it’s vendredi”; or when the second boy, Oliver, utters “we dit thank you”, these are

instances of intra-sentential code-switches. On the other hand, inter-sentential code-

switching refers to the use of two languages at the utterance, sentence or discourse level.

The data used in this dissertation does not provide a lot of intra-sentential code-switches.

However, as it deals with the way young bilingual children juggle two languages within a

family conversation, it corresponds to Poplack’s definition of inter-sentential code-

switching. For example, during one of the meals recorded for this study, the first child

Thomas suddenly choses to start a new utterance in French while he was telling a story in

English: “And after there was a message; et après il va voir la reine.” Additionally, Auer

points out that there is a strong correlation between language uses and speech activities

(Auer, 1995). His research is largely based on conversation analysis and tends to consider

code-switching as a mechanism to structure interactional exchanges.

To recap, speakers use codes in their attempts to communicate with other

language users. Bilingual speakers may sometimes switch their communicative codes in

order to signal a change in situation or in conversational settings. In other words,

switching codes is a means by which language users may contextualize communication.

This dissertation aims at showing that, indeed, code-switching may be a powerful tool in

discourse in order to signal changes or create breaks in the conversational context. This

context may be very local (the end of a turn at talk for example), very general, or

anywhere in between (Nilep, 2006). Furthermore, it is important to recognize that such a

switch in the form of communication is accomplished by the participants in a particular

conversation. In other words, code-switching is accomplished by parties in interaction,

and the meaning of their behavior emerges from the conversation itself. The approach

adopted here thus understands code-switching as the practice of individuals in particular

discourse settings.

24

1. 4. Concluding remarks

Bilingualism is a complex reality, multidimensional and perpetually changing and

developing. There exists no perfect definition of what it means to be bilingual. Several

approaches come into consideration and several features are to be taken into account.

Bilinguals can be classified into various types of bilingualism. Some

characteristics of bilingualism involve the degree of fluency and competence in the

languages spoken, the context and manner of acquisition of the languages, the age of

acquisition, and the various hypothesized processing mechanisms. Bilinguals can be

classified based on various facets, both at the individual level as well as at the social

level. Therefore, depending on the dimensions of their bilingual characteristics, the

speakers may be categorized into different types of bilinguals.

Nevertheless, it is important to bear in mind that bilingualism is multidimensional

and that most of these dimensions are usually interrelated. It could be argued that one

who is exposed to two languages from birth (simultaneous bilingual) is most likely to

grow up to be a balanced bilingual. These dimensions of bilingualism are therefore

continuous and not simply categorical constructs. One cannot draw clear boundaries

between types of bilingualism within a given dimension.

As a result of this overview, the children under study in this research project could

be considered as early bilinguals, conceivably sequential early bilinguals. Further

analyses are needed to demonstrate whether they are compound, coordinate or

subordinate bilinguals; additive or subtractive; balanced or dominant bilinguals. In the

latter case, it will be interesting to consider which of French or English is their dominant

language, and to determine whether each language has a specific use or function.

25

2. Bilingual Acquisition It is estimated that there may be as many children who grow up learning two

languages as one (Genesee, Nicoladis & Paradis, 1995). Yet, childhood bilingualism is

often poorly understood by many and regarded with skepticism by others. Because of a

lack of familiarity with or knowledge about early bilingualism, parents and educators may

express doubts about childhood bilingualism and might expect negative consequences to

result from children learning two languages during the early years of their lives. Such

concerns are common in communities where most children grow up monolingual. Indeed,

consequently, most adult members of the community come to view monolingualism as

normal and bilingualism as abnormal.

Nevertheless, researchers have been actively involved in studying bilingual

acquisition in order to shed some light on various aspects of bilingual development. This

section addresses issues related to bilingual acquisition and attempts to provide some

answers regarding the case of the three children studied in this project.

2. 1. Simultaneous acquisition of two languages 2. 1. 1. One system or two?

Much of the research on the simultaneous acquisition of two languages has been

motivated by the hypothesis according to which children who are exposed to two

languages from their birth go through an initial stage when the languages are not

differentiated at first (Volterra & Taeschner, 1978). This is called the ‘unitary language

system hypothesis’. The most explicit formulation of this hypothesis can be found in

Volterra and Taeschner (1978:312):

“In the first stage the child has one lexical system which includes words from both

languages (…), in this stage the language development of the bilingual child seems to be like the

language development of the monolingual child. (…) In the second stage, the child distinguishes

two different lexicons, but applies the same syntactic rules to both languages. In the third stage the

child speaks two languages differentiated both in lexicon and syntax (…)”.

This hypothesis thus proposes that the initial state of the developing bilingual

child is essentially monolingual. As a result, this approach has led researchers to compare

the development of bilingual children to that of monolingual children acquiring the same

languages. On the one hand, such a comparison can be interesting as it may reveal to what

extent bilingual first language acquisition actually differs from monolingual acquisition.

26

On the other hand, such comparisons may be inappropriate because they risk attributing

differences that bilingual children exhibit to deficits in their capacity to acquire language.

Alternatively, it can be argued that the linguistic competences of bilingual children – like

those of bilingual adults – should be examined and evaluated on their own merit

(Grosjean, 1994).

2. 1. 2. Issues of delay and proficiency

One common concern that is often expressed by parents and childcare

professionals about bilingual acquisition is that learning two languages in childhood is

difficult and can result in delays in language development. However, children who have

regular and rich exposure to both languages on a daily or weekly basis exhibit the same

milestones in language development as monolinguals roughly at the same ages (Genesee,

Nicoladis & Paradis, 1995). Furthermore, Genesee insists on the importance to remember

that there are large individual differences in language acquisition, as some children

acquire their first words and use complex utterances much earlier than others. In such

cases, delays in the emergence of these utterances simply mean that the child has taken

longer to reach this stage. The same types of differences are characteristic of bilingual

children.

Notwithstanding, according to Genesee, it is important that the parents of bilingual

children provide systematic exposure to both languages most of the time and that they

avoid radical changes to the language environment of the child. Such changes may disrupt

language development and create difficulties for the child. De Houwer (2009)

corroborates this idea in arguing that the input in both languages should be regular,

almost daily.

Another concern about early bilingual acquisition is that children may never

master either language fully or never become as proficient compared to monolingual

children. Again, Genesee contradicts this statement in claiming that bilinguals can acquire

the same proficiency in all aspects of two languages over time, as monolinguals; even

though they usually have less exposure to each language. Provided that the children are

regularly and substantially exposed to each language, they acquire the same proficiency

in the grammatical and phonological fields of their two languages.

Generally speaking, bilingual children’s overall proficiency in each language

reflects the amount of time they spend in each. Thus, a child who has just returned from a

27

visit to a grandparent where only one of the languages was used may prefer to use only

that language for a while. It may then seems that the child has lost some proficiency in the

other language; but this usually is a temporary shift in preference that is corrected once

the child is exposed to both languages again (Grosjean, 1994).

All three children in this research project are regularly exposed to both languages.

Indeed, the first two boys go to a French school and the last one spends the day with a

francophone Senegalese caretaker. Yet, in the evenings and on the weekends, either the

father or the American au-pair Christy is present, thus assuring an Anglophone presence.

Furthermore, although the mother is a native French speaker, she freely alternates

between French and English, also providing the children with regular input in both

languages.

In conclusion, Genesee (2008) argues that parents can ensure that their children

achieve full proficiency in both languages by providing them with rich experiences with

each. Moreover, it is also important for parents to maintain both languages because it is

part of the family culture and plays an important role in the child’s developing identity. It

helps them feel unique and connected to their family.

2. 2. Sequential acquisition of two languages In what is called sequential bilingualism, one language is acquired during the age

of early syntactic development, and the second language is acquired once the structural

foundations of the first language are in place. For example, if the child of immigrant

parents is exclusively exposed to the family language until (s)he goes to school and the

second language is introduced at a later age, then the child is considered a sequential

bilingual. Sequential bilingual children may also be immigrant children. That is to say,

they were born abroad and then they moved to the majority language speaking country

after the first years of their lives. They learned their first language at home in a

monolingual environment, then learned a second language in the new country. This is the

case for the three boys herein. Indeed, the three of them were born in New Zealand,

where the first two boys attended the equivalent of preschool. When the whole family

moved to France, all three children had to adapt and learn a new language, which differed

from what they had started to acquire as their native language hitherto.

28

Several theoretical issues arise in the study of the sequential acquisition of two

languages. These concern the nature of the L2 acquisition, the rate of acquisition, and

ultimate attainment (Montrul, 2008; as cited in Kail, 2015).

According to Haynes (2007), there are five stages in sequential language

acquisition. The first stage is Preproduction and is also referred to as “the silent period”

when learners gradually build up their vocabulary without speaking the language but

more by echoing the language. Then the second stage is that of Early production. At this

stage, learners have the capacity of constructing words in short phrases and memorize

short language forms, although not necessarily used correctly (Haynes, 2007). The third

stage is referred to as Speech emergence, when learners are able to speak simple phrases

and short sentences. By then, they are able to engage in conversation and ask simple

questions. They can also understand short stories if they are supported with pictures. The

fourth development stage by Haynes is Intermediate fluency. At this stage, the learners

have an active vocabulary and can form longer and more complex phrases, both spoken

and written. There might be some grammatical errors but they demonstrate excellent

comprehension. The last stage is called Advanced fluency, at which point the learners are

considered near-native (Haynes, 2007).

In the same way as monolinguals learn their first language, sequential bilingual

learners must also learn how to use their newly acquired language accurately and

appropriately. Although the process of language learning may be similar, there are also

considerable differences. Indeed, bilingual learners address the process of learning

another language already possessing knowledge of a linguistic system, its structures and

its rules. As they learn the new language, second language learners incorporate the new

linguistic input into their existing model of language. This was the case for the children in

this project. They experienced French through their already existing model of English.

2. 3. Communicative competence Just as monolingual children, bilingual children face a number of communication

challenges. They have to produce utterances that are comprehensible to their

interlocutors, use language in socially appropriate ways, and sometimes get their meaning

across when language acquisition is still incomplete. However, being bilingual entails an

understanding of interpersonal communication that exceeds the one required for

monolingual conversation, as the bilingual individual has the ability to communicate

29

appropriately and effectively in two languages (Genesee, 2008). Furthermore, the ways of

communicating in certain social situations or of expressing certain meanings can be

somewhat different from one language to another. Bilingual children thus also acquire the

social language skills that they need in order to express themselves in an adequate way.

This ability to make appropriate language choices with different interlocutors is

fundamental to bilingual communicative competence. As Genesee (2008) noted, bilingual

children are able to use their languages differentially and appropriately with their parents

who habitually speak different languages with them. Furthermore, they show a similar

sensitivity when interacting with strangers with whom they had no prior experience

(Genesee, 2008). Indeed, Genesee’s “Evidence from Montreal” aims at showing that the

children can adjust their language mixing to match “the mode of language” of their

interlocutors (referring to Grosjean’s continuum). This responsiveness to the linguistic

preferences of their interlocutors – whether they are familiar or not – indicates that

bilingual children’s ability to use their developing languages appropriately reflects true

communicative competence. In other words, children are able to make direct adjustments

to accommodate their interlocutors along with their language preferences and/or abilities.

If we refer to Grosjean’s continuum (1994) and his scale on “language mode” (see 1.1.2.),

children choose to be more or less bilingual in each situation of communication,

according to what the situation requires them to be.

The question that thus arises from this observation is: how do young bilingual

children know which language is appropriate and which is not? How do they choose their

mode of language (monolingual or bilingual)?

Elizabeth Lanza (1997) argues that bilingual children’s understanding of

appropriate language choices results from the same processes of language socialization in

the home that influence the development of communication in monolingual children

(Lanza, 1997:13). However, this hypothesis cannot explain children’s performance with

unfamiliar interlocutors where prior experience and knowledge are lacking. Yet, as

previously noted, bilingual children can exhibit appropriate language choices with

strangers and match their language mixing to that of newcomers.

30

2. 4. Bilingual learners

As children learn language, they also learn through language about relationships

and social values. They begin to learn about the culture into which they are born. It is

through language, in their everyday interactions with their family at home, and their peers

at school, that cultural structures are transmitted to children. In this research project,

cultural values are transmitted through family dinners, which is why I chose to look at

spontaneous conversations during mealtimes.

In this context, bilingual learners also develop a sense of cultural and social

identity; and language is central to this process. Over the years, bilingualism has been

associated with a greater awareness and sensitivity to linguistic structure, along with

cognitive and cultural advantages (Genesee, 2008). Children who have the opportunity to

speak two or more languages are now considered as having an advantage and are

encouraged to maintain both languages, so they can enjoy the benefits that may

accompany the bilingual status. They are pushed to cultivate their home language as well

as their majority language. This is now seen as an essential quality, not only for the

children’s future cognitive development, but also for their ability to establish and

maintain a strong cultural identity, to develop and sustain strong ties with their families,

and to thrive in a global, multilingual world.

2. 5. Concluding remarks

Throughout the first years of their lives, nearly all children become fully

competent in at least one language. Although this is a complex task, which requires much

effort, it is expected and considered normal. Parents and educators scarcely worry about it

even though it is the most complex accomplishment of early childhood. Even more

remarkable are the children who simultaneously – or early sequentially – acquire

proficiency in two or more languages during their preschool years. Within approximately

the same time frame as it takes monolingual children to learn one language, bilingual

children learn two languages and become capable to use them in socially diverse

situations and appropriate ways.

While simultaneous bilingual acquisition follows a path that is similar to

monolingual development, sequential second language acquisition occurs in a predictable

series of stages. But whether the acquisition of the two languages occurs simultaneously

or early sequentially in the child’s development, bilingual learners still develop

31

considerable competences in interpersonal communication. They are able to adapt to each

situation of communication and express themselves in an appropriate way, given that they

were provided with adequate and rich experience in both languages. Most research

concludes that there may be some general advantages and positive effects of bilingualism

on the linguistic, cognitive and social development of children.

Moreover, bilingual children also tend to develop a bicultural identity as they

maintain their heritage language, social practices and cultural values in the home, all the

while dwelling in the majority language and routines of the community in which they

live.

3. Studies on Family Dinners 3. 1. Communication time Family dinners are not only about eating and feeding one’s body; they are a time

for sharing and constitute a good opportunity for the participants to talk about their day.

Indeed, it is one of the rare moments when the whole family reunites and all co-

participants sit together around the same table. They allow researchers to observe

language practices in a more detailed manner and in a naturalistic environment.

Moreover, anthropologist Elinor Ochs also highlights mealtimes as being a

process of apprenticeship for the children. In this context, apprenticeship is defined as

“learning through active observation and direct participation in activities together with a

more knowledgeable participant” (Lave & Wenger, 1991; Rogoff, 1990, as cited in Ochs

& Shohet, 2006:35). Indeed, because the parents constantly solicit their children’s

attention and involvement during mealtimes, the latter gradually learn to position

themselves from observers and overhearers to more central participants. Indeed, they can

help prepare the meal, or serve it, but they are also consumers and communicators. It then

becomes apparent that family dinners are not just about eating. They constitute a

privileged moment in the family’s daily routine, when the children learn to coordinate

food consumption with conversational activity. Mealtimes allow the children to engage

conversation with more experienced participants, who can explicitly convey sociocultural

messages through speech activities. Many scholars have already underlined the

importance of studying families’ daily practices when looking at children’s behavior. A

number of studies have indeed been devoted to conversations during dinners (Sacks,

32

1992; Schegloff, 2009; as cited in Morgenstern, Beaupoil, Caët, Debras & le Mené, 2015)

in order to identify the relation(s) between food rituals and interactional settings (Keppler,

1994; Ochs, Pontecorvo & Fasulo, 1996; Blum-Kulka, 1997; Aronsson & Gottzén, 2011;

as cited in Morgenstern et al., 2015). Mealtimes are therefore a window onto everyday

socially established ways as much as they are a window onto everyday linguistic codes.

3. 2. The weight of culture: internalizing norms and rules Anthropologists and ethnologists have long considered ways in which mealtimes

promote moral beliefs and values along with social order. According to Elinor Ochs,

mealtimes are also cultural sites that ensure the socialization of people into competent

and appropriate members of a society (Ochs & Shohet, 2006:35). Indeed, family dinners

also serve as vehicles for cultural practices and representations. They are universal

occasions for members to “engage in the activities of feeding and eating but also to forge

relationships that reinforce or modify the social order” (Ochs & Shohet, 2006:36). Hence

family dinners allow the researchers to account for cultural practices and the acquisition

of norms, habits and family values as well. An ethnographic study conducted by Kremer-

Sadlik & Morgenstern et al. showed that French family dinners in particular constitute

privileged moments for children to appropriate cultural practices and preferences

(Kremer-Sadlik & Morgenstern et al., 2015). In this study, the researchers compared the

rituals during mealtimes in French versus American families. They demonstrated that the

French children generally tend to eat more fruit and vegetables than the American

children. Among other things, their analysis revealed that the children’s local practices

are intimately linked to cultural preferences along with the ritual organization of the

meals. It thus seems that eating behaviors strongly participate in the construction of the

identity of family members. Indeed, everyday these members share repeated practices and

construct common representations, which result in one common cultural identity.

Moreover, aside from cultural habits and norms, parents also teach their children the rules

that govern the rituals of family meals. The latter gradually learn to position themselves

vis-à-vis those rules. Indeed, as Morgenstern et al. pointed out, refusing or transgressing a

rule until (s)he is punished is also a way for the child to show that (s)he knows and

understands the norms and good manners of mealtime (Morgenstern et al., 2015).

Besides, the latter paper specifies that the transmission of rules is not necessarily vertical,

from parents to children. It can also be from children to parents; or it can be horizontal,

33

from child to child. Indeed, as the present research study features three children during

family dinners, we can notice that sometimes the boys correct each other when it comes

to good manners, showing that they have internalized the rules and now act as a relay in

their own way. It is thus deducible that food practices are intimately linked to the

transmission of norms and values.

Moreover, one of the essential dimensions when transmitting norms and rules to

children is leaving them with a choice. Indeed, the latter may deliberately choose to deny

or transgress the rules that are being taught to them. Therefore, mealtimes also constitute

an opportunity for the children to assert themselves and express their desires, ideas, and

make their own choices (Brougère & de la Ville, 2011; as cited in Morgenstern et al.,

2015). Family dinners thus become a perfect time and place for the parents to shape the

children’s cultural territory, but also for the children themselves to assert their identity

and their personality within their family context. As already pointed out, children

gradually leave their status of apprentice to become more and more central participants in

the family life. Besides, mealtimes also permit to reactivate and reinforce emotional ties

between parents and children by sharing the same food. Indeed, sharing food is an

essential social activity for family members. In ancient Greece, for example, the term

“oikos” (family) was stipulated as “those who feed together” (Lacey, 1968:15; as cited in

Ochs & Shohet, 2006:37). Once again, this underlines the paramount dimension of

mealtimes as a socializing time for the family. However, some cultures are more attached

to the tradition of eating together than others. As Kremer-Sadlik & Morgenstern et al.’s

study pointed out, dinners where everyone sits at a common table is a vanishing ideal for

many families in the United States. Besides, Ochs, Smith and Taylor (1989) found in their

study on American family dinnertimes that mothers and children sometimes dine before

the father returns home, or that children eat before their parents. This is the case for the

children in this study, for instance. As the father tends to come home rather late from

work, the three boys eat earlier in the evening, with their mother and their caretaker who

are present at the table but do not dine with them. The three adults eat together later, once

the children are put to bed. This observation corroborates Ochs’s idea that families’

practices differ from one culture to another, and is often confronted with the busy

schedules of working parents. This issue of working parents was indeed addressed by

Ochs, Shohet, Campos and Beck in “Coming Together at Dinner: a Study of Working

Families” (2007). This study showed that, despite long working hours, many families

34

wish to have dinner together because of the perceived and documented links between

family dinners and child well-being (Ferguson et al., 2006: as cited in Ochs et al., 2007).

Finally, family dinners play an essential role in the children’s socialization since

food may also be used as an influential tool during mealtimes. Indeed, food can be used

as a reward when trying to teach good manners at the table. Ochs showed that parents

tend to use dessert as a carrot or a stick when they want to get children to eat the main

portion of their meal. It is thus interesting to notice that meat and vegetables are

considered as food that the children must eat but generally do not want to, whereas

dessert is food that the children fancy and crave (Ochs & Shohet, 2006). Food can thus

also be linked to pleasure and reward, which creates additional motivation for the child to

want to participate and proclaim their own place within the family meals.

To conclude, ethnographic evidence from various parts of the world has supported

the notion that dinnertimes are filled with social import. More than just about eating and

feeding, family dinners are a socializing time where children become apprentices of their

own culture and identity. They gradually learn to assert themselves, express their desires

and affirm their choices. They also internalize a number of rules and norms that are being

taught to them through the sharing of food practices and the learning of good manners

and appropriate behavior around the table. Dinnertime is also an opportunity for the child

to learn how to coordinate the activities of eating and conversing. Each participant has a

communicative role in the family’s interactions at the table. Moreover, mealtimes are

definitely cultural sites where the members of different generations come to transform

each other’s way of interacting, acting or seeing the world. The ritual practices of

mealtimes orient children in their comportment. Their cultural preferences along with

their language uses are constructed and shaped during everyday mealtime interactions.

4. The Conversation Analytical Approach 4. 1. Why conversation analysis? Conversation analysis (CA) is a popular approach to the study of discourse. It is a

way of analyzing the pragmatics of ordinary conversation, focusing on the interactive and

practical construction of everyday interchanges. Levinson defines conversation as “that

familiar pre-dominant kind of talk in which two or more participants freely alternate in

speaking, which generally occurs outside specific institutional settings like religious

35

services, law courts, classrooms and the like” (1983:284). The purpose of conversation

analysis is to study spontaneous conversations without specific pre-construed notions of

the speakers’ identities. Therefore, it is based on naturalistic data – as opposed to elicited

data. CA is a reaction to sociolinguistic theories, as the movement arose in reaction to the

quantitative techniques, and the arbitrary imposition on the data of supposedly objective

categories (Levinson, 1983). In order to conduct a conversation analysis, two or more

participants must be engaged in speaking. It cannot be the study of monologues or

speeches, for instance. Some of the first researchers interested in conversation analysis

were ethnomethodologists. Their methodology came from sociological practices. In order

to collect naturalistic data, researchers started using audiotape recorders. Today,

videotape recorders enable linguists to register and encode conversations in a very precise

manner, also including non-verbal aspects of communication such as gaze, facial

expressions, pointing, etc. (Morgenstern, 2009). In any case, this approach centers around

the conversation itself and the emphasis is put on the videos and their transcriptions.

4. 2. Main characteristics Conversation analysis (CA) seeks to describe conversation in a way that builds

upon the choices of the people who are participating in it. It does this by paying attention

to the way that each utterance displays an interpretation of the previous utterance and is

formulated in a specific context, both verbal and non-verbal. CA is based on turn-taking,

that is to say the idea that when we converse we take turns in speaking (A-B-A-B-A).

A turn in conversation can be allocated; it can be ignored, or it can be masked by sounds

like uuum for example. The rules that govern turn-taking are referred to as the ‘local

management system’ (Sacks H., Schlegoff E.A., Jefferson G., 1974). Sometimes, if two or

more participants speak at the same time, a turn-taking can be violated. This phenomenon

is called an overlap and often occurs for a short duration that can be counted in

microseconds. Usually, participants take turns after a transition relevance place (TRP).

TRPs are junctures at which the turn at talk could legitimately pass from one speaker to

another. Basically, each point in the ongoing conversation at which a turn passes from

one participant to another usually occurs at a transition relevance place. It marks the end

of a tone constructional unit (Levinson, 1983). If a turn is consciously allocated, then

there is projectability. It can be verbally expressed (“over to you”, “what do you think?”,

“any complaints?”); or it can include a gaze, a glance, or a particular intonation. Besides,

36

the use of intonation and stress plays an important role in determining the speaker’s

intentions (rising/falling/pause). Those who are not actually active in the conversation can

use back-channelling (Mey, 1993). This way, they intend to show that they follow and

understand their interlocutor’s remarks, without interrupting their turn at talk.

A turn in conversation can also include what are called pre-sequences (Levinson,

1983). These pre-sequences are hedging devices whose goals mainly are to be polite and

to avoid embarrassment. Pre-sequences can either be for inquiring (pre-request), for

inviting (pre-invitation or pre-arrangement if more formal) or for revealing new

information (pre-announcement).

Usually, sequences work in adjacency pairs (Mey, 1993). This means that the

turns that are produced by different speakers are adjacent. They are composed of a first

part followed by a second part, which is specific to the first part. Adjacency pairs can fall

into patterns like ‘greeting/greeting’, ‘question/answer’ or ‘offer/acceptance’ (Mey,

1993). In the case of a question or an offer, the response can either be preferred or

dispreferred (Levinson, 1983:336). A preferred response corresponds to what is expected

in theory (agreeing, accepting a request); it is also referred to as unmarked. On the other

hand, a response can be dispreferred, that is to say that it is not the response that the co-

participant expected. In this case, it is marked (disagreement, refusal of a request, delay or

absence of an answer, etc.). If an adjacency pair is considered meaningless or superficial,

it can be referred to as a phatic communion (Levinson, 1983). This sequence helps

creating a tie between the co-participants. It is the case in the pattern ‘greeting/greeting’

for example. When the adjacency is broken and the speakers insert a sequence within

another sequence, this is called insertion sequence (Mey, 1993). In this case, instead of

having the specific second part of the sequence right away, there is another sequence that

takes place in between. Therefore, the pattern is no longer 1-2; but becomes 1-1’-2’-2.

Another device that is common in conversation analysis is the use of repair. A

repair is used for the correction of misunderstandings, mishearings or non-hearings. It can

be initiated and performed either by the speakers themselves or by another participant. As

a result, there are three different types of repair in the end: A) self-initiated-self repair,

B) other-initiated, self-repair, or C) other-initiated-other-repair (Levinson, 1983). A repair

mainly intends to fix a conversational breakdown and to restore mutual understanding in

the conversation. Since this study features family dinners involving parents with their

three children, we might expect a certain number of repairs throughout the conversations.

37

To sum up, when trying to understand a particular utterance or conversational

action, it is important to consider where and how that action is located in a sequence of

other conversational actions. When people speak in an ongoing conversation, they do so

in the light of what has just been said, and in anticipation of what might take place in a

future sequence. The participants understand the talk of other people and construct their

own speech accordingly. The meaning of an utterance, the way it was formulated and the

way it is interpreted by the co-participants, thus depends on its context, both verbal and

non-verbal.

This dissertation therefore attempts to look at the linguistic practices of bilingual

children in their everyday interactions, using the method of conversation analysis. It

shows that a conversation is not just a simple succession of turn-takings. Conversation

Analysis allows us to go further in analyzing people’s social interactions and attempts to

explore the various mechanisms of conversation. Indeed, one does not speak in a uniform

way that never changes; conversational settings are continually reconstructed and can

greatly differ from one person to another. There is no “correct” or “ideal” frame for a

conversation. People construct and articulate their exchanges, thus producing a social

reality in which they are fully engaged.

38

39

III. THE DATA 1. Presentation of the corpus

1. 1. Family background and history

I was given the opportunity to film a family that some might qualify as atypical,

for the mother is French and the father New Zealander. This family can be classified as

upper middle class, as both parents received a college education. The mother, Sandrine, is

currently a homemaker. She aspires to become an English teacher in France. The father is

a corporate and financial lawyer for the OECD (Organization for Economic Co-Operation

and Development). The mother is a native French speaker. She was born in France and

has worked in many places all around Europe after she completed a language degree in

France (Langues Étrangères Appliquées). She and her husband John met in Italy.

Although John was born in New Zealand, he was working in London at the time and went

to Italy on a weekend for a business trip. After their marriage, they both settled in New

Zealand to start their family.

The three children were therefore born in New Zealand, an English-speaking

country. Their mother confesses that, at the time, she spoke to them more in English than

in French. According to her, the reason might be that she did not want to confuse them

too much. Furthermore, she claimed she had the feeling that her children would

understand her better if she spoke the same language as their father, and the same

language everybody around them was speaking. Besides, English words are generally

shorter than French words and, thanks to the stress pattern, the meaning of a phrase is

usually a little bit quicker to access than in French, which has no stress pattern. This

might explain that, when faced with a danger or a predicament, Sandrine admits that her

first choice assuredly was – and still is actually – English rather than French.

Her attitude towards French changed after they moved to live in France at the end

of March 2013. The whole family has lived near Paris after John changed jobs.

Supposedly because they were henceforth in a French environment, Sandrine began to

talk more and more in French to her three children. But, according to her, their use of the

French language was practically non-existent before then. They were used to hearing the

mother talk in French from time to time, but not really to speaking the language. When

they arrived in Paris, the first two boys – Thomas and Oliver – began to attend school in

French. This was something completely new to them after having spent the first several

40

years of their lives in New Zealand. Additionally, the youngest child, James, now spends

his days with a French-speaking Senegalese child-minder.

The family has had an au-pair girl since late August 2014. Her name is Christy

and she is American. She came from Cincinnati, Ohio, to learn more about the French

culture and improve her linguistic skills. She is present mainly during weeknights to help

with the children while the father works late. Indeed, although the family moved to

France more than two years ago, the three children are still living a “New Zealander way

of life”. They tend to dine quite early in the evening and go to bed around 8 o’clock.

Hence they usually have dinner before both parents during the week. The mother and the

caretaker are present at the table but they don’t dine with the children. The three of them

eat before the rest of the family, then they are given a bath before they are put to bed. The

three adults, Sandrine, John and Christy, dine together later in the evening.

According to the mother, her three children are undoubtedly Anglophones. This

presupposition is due to the fact that they were born in an English-speaking country and

spent the first years of their lives over there. What concerns the mother the most is the

eventuality that her children would not attain a level of proficiency in French “as good

as” the one they have in English. Although she does not appear to be worried about their

language development, she is concerned about maintaining their dual linguistic abilities.

Indeed, both parents seem very positive towards their children’s bilingualism. They see

their multilingualism as an asset in this world that is perpetually changing and most likely

evolving to one big village. In the eventuality that they are bound to move to an English-

speaking country again, the mother definitely intends to keep speaking French to her

children in order to keep French language as an important part of her boy’s lives. She

considers her children’s bilingualism as quite significant now, along with their bilingual

and bicultural identity.

Nevertheless, both parents are very indulgent towards their children’s linguistic

choices. They do not seem to want to force any language on them for any type of situation

in particular. It seems that the parents’ strategy towards their children’s linguistic

development has resulted in the application of a one person – one language approach

while interacting with the three of them. Most of the time, the mother uses French when

addressing her children while the father uses English. Although Sandrine recognizes that

sometimes she switches to English when she addresses her husband. This may have an

effect on the other participants in the conversation. Indeed, the mother claims that when

41

the whole family is brought together (on the weekends or during the holidays) the

dominant language around the table is definitely English. Besides, as already evoked

earlier, she strongly believes that her children are definitely dominant in English because

they were born in New Zealand and spent the first years of their lives there.

Furthermore, she admits that she still uses English a lot. Especially when faced

with situations of danger. According to her, her children understand better and quicker if

she talks to them in English. Therefore, she declares that she is more likely to utter things

like “watch out!”, “careful!” or “car!” than « faîtes attention » or « il y a une voiture qui

arrive ». However, now that the children evolve in a French environment and go to a

French-speaking school, or have a French-speaking child-minder in the case of the

youngest, a closer look at the actual children’s use of language is needed to verify the

hypothesis that they are dominant in English, as in the mother’s opinion.

1. 2. The children

At the time of the recording, the three boys were respectively 5;11, 4;6 and 2;3.

This section briefly presents each of their individual background along with their

linguistic development.

Thomas

Thomas is the firstborn of the family. He was born in New Zealand on February

8th 2009. He is a delightful six-year old who speaks both English and French. Thomas did

not strike me as reserved with strangers or with anyone in general. He is a very social

child and seems eager to share things with the people around him. He appreciates contact

with anyone and seems to bond rather easily with the people he does not know. He is a

very lively and quite talkative child, with a good sense of humor. He usually is in

perpetual motion and is always trying to get the adults’ attention back to him. I noticed

that whenever a parallel conversation is going on with one of his brothers, Thomas insists

on either giving his opinion, repeating what is being said or showing that he knows better

than his brothers. He seems to enjoy having all the attention. He is also quite a singer.

Several of the recordings feature him singing at some point at the table. Again, singing

might be an occasion for Thomas to make himself heard, to assert his place in the

conversation. It could also be a result of what he is learning or practicing in school. One

42

of the recordings suggests that Thomas’s class forms a choir and practices singing once a

week.

Prior to their moving in France in 2013, Thomas went to preschool and school in

New Zealand. The caretakers and teachers spoke exclusively in English. The only

sporadic contact that Thomas had with French was from his mother. The only times when

she positively spoke French were when she was reading French stories to them or singing

French nursery rhymes. So when the family arrived in France, the mother confessed that

Thomas’s French was practically non-existent. Apart from hearing his mother from time

to time and repeating some of her expressions, he had never really been exposed to

French on a regular basis.

This all changed in 2013 when they moved to the Parisian area. Nowadays,

Thomas henceforth has more contact with French as their home is in a completely French

environment. Besides, he now goes to a French school and has many monolingual friends,

with whom he mainly interacts in French. According to his mother, it was difficult for

Thomas to adapt to a French school because he had not been exposed so much to the

French language before. Therefore, he had some trouble catching up with the other pupils.

Sandrine made a point of honor to help Thomas with his homework and more generally

with his abilities in French. She gave him intense lessons during the holidays and,

eventually, it paid off. By September 2013, less than five months after their arrival in

France, Thomas’s grades had greatly improved. Today, Thomas’s parents are very happy

with his results and consider that he is doing very well at school.

Nevertheless, Thomas is still in contact with the English language quite a lot.

Indeed, the American au-pair girl, Christy, is at home every night and helps the children

do their homework. She is also present when they are having dinner and then takes care

of giving them a bath before they are put to bed. As the recordings show, she scarcely

ever speaks French. As far as the weekends are concerned, English seems to be

dominantly present as the father is home and uses mainly the English language when he

addresses his three boys. So English is still very present at home. Moreover, every

Wednesday afternoon, Thomas goes to an English school. His mother claims that he has

very good results. However, it seems that Thomas actually prefers the French school.

According to Sandrine, oftentimes he does not wish to go to the English school because

he does not have as many friends there as he does at the French school. It could also be

that the schoolteacher is different and Thomas may prefer his French teacher.

43

Sandrine’s opinion is that Thomas still does not master French as well as a French

monolingual child. She still considers him ‘behind the average level of proficiency of

children his own age’. But Thomas is one step ahead of his classmates because he

skipped a grade while transferring from New Zealand to France. He currently is in Cours

Préparatoire when he should be in Grande Section de maternelle. So he is one year

ahead. A further analysis is needed to determine Thomas’s language development and

level of proficiency in his use of each language.

Oliver

Oliver is Sandrine and John’s second child. He was born in New Zealand as well,

on July 10th 2010. He is a lovely boy who is very social and pays a great deal of attention

to what is going on around him. Similarly to his brother Thomas, he also likes having all

the attention. Oliver is eager to show what he knows and what he can do. He requires his

parents’ attention in order to show them that he is also capable of doing what the others

are doing at the table. I noticed that whenever his mother or father reprehends one of his

brothers, Oliver is the first one to side with them and to repeat whatever the parents are

saying. It almost seems like he yearns to be the responsible one; and he is very proud of

himself when he is rewarded for it. He is also a very curious boy. He wants to know all

that is happening and what the adults are talking about. He strongly wants to share

information with his co-participant(s). The recordings also suggest that he is a fan of

superheroes. Oliver is at the age when he likes to disguise and pretend that he is someone

– or something – else. He might be wearing a cape during dinner or telling about how he

pretended to be a dog when asked about what he did during the day.

When the family arrived in France, Oliver did not go to school right away. He

stayed at home with his mother from April to September. At this stage, Sandrine was

already speaking more and more French to help Thomas with school. She started doing it

with all of her children, including with Oliver at home. However, she describes that

Oliver was not very talkative at that time. He used to stay rather silent whenever he was at

home with his mother. This reminds us of the silent period that Haynes qualifies as a

‘pre-production’ stage in second language acquisition (Haynes, 2007).

In September 2013, Oliver started French school in Petite Section de maternelle.

Today, he goes exclusively to a French school everyday. On Wednesdays though, the

school is closed so he stays with his little brother James and their francophone Senegalese

44

child-minder, Nafi. If they are at home, the American au-pair girl Christy is supposedly

present too. But she is the only Anglophone presence, as Nafi and the other children are

all Francophone speakers.

So for Oliver as well as for Thomas, French was practically a virgin territory

before the family moved to Paris. Today it seems that Oliver is rather at ease with French.

Like his brother, he is predominantly in a French-speaking environment. Moreover, the

fact that their own mother started speaking to them and answering them in French may

also have had quite an impact on their ritual conversations. Although Oliver is still

struggling with gender references, his French is quite fluent now and it is also very

spontaneous. Most of the time he addresses his mother in French, having integrated the

idea that it is now the appropriate language to communicate with her.

James

Last of the family but not least of the co-participants, James is the youngest boy

among his siblings. He was born on October 30th 2012, in New Zealand as well. The

family moved to France five months later. Therefore, James was not exposed to the

English language as much as his brothers.

He spends a great deal of time with Nafi, a Senegalese and Francophone caretaker,

and one other little girl of his age who is a French monolingual. So, as far as his time of

exposition to the French language is concerned, he is the only one among the siblings

who has been more exposed to a French-speaking environment. Indeed, the only source

of English for James is in the conversations spoken at home. Utterances in English may

certainly occur on the part of his father or by Christy, but also by his brothers.

It seems though that James has understood the functioning of the family’s

conversations as a bilingual one. Most of the time indeed, he refers to his father as

“daddy”, and this is how he begins the occurrences that are addressed to him; and he calls

his mother “maman” and identifies her as such. This use of vocatives in different

languages shows that he understands the functioning of his family’s conversations as

bilingual. James does his share of bilingual conversation with such pre-announcements. It

will be interesting to see whether James differentiates the languages in a similar manner

regarding his parents’ utterances. Is he bound to answer in French due to his larger

exposition time to the French language, or does he use both English and French for

purposes of efficiency in his communication with all the family members? In order to

45

answer these questions, a closer look at James’s utterances in the family’s conversations

is needed.

2. Recording

The methods to collect data have greatly evolved over time but the recording of

natural conversations remains the basic source of information to analyze language

acquisition and to perform conversation analysis. Before any machine could allow any

recording, researchers would write on diaries the progress of their own children, as did

the linguists Deville, Bloch or Guillaume, along with the psychologist Piaget or the

biologist Darwin; who regularly took notes on the productions of their offspring

(Morgenstern, 2009). Then, audio tape recorders allowed researchers to record

conversations and transcribe them. Today, video tape recorders enable linguists to record

natural conversations and to encode them in a very precise manner, reporting not only

verbal communication but also gazes, reactions, attitudes, facial expressions, pointing

gestures and other non-verbal communication elements (Morgenstern, 2009).

Nevertheless, these types of recording oblige researchers to face a new challenge

regarding the question of degree of authenticity of the data (Caronia and Caron, 2005).

2. 1. Collection of the data

Two options are offered to the researchers when collecting data for the purpose of

studying language acquisition. The first one is referred to as naturalistic data or natural

speech act data (Beebe and Cummings, 1996). This method involves the observation and

recording of children’s language use in interaction with their parents and/or their

caretakers in context. Samples of spontaneous conversations are collected in natural

settings. It is a real life situation that the researcher attempts to observe in a way that is as

minimally intrusive as possible. The second option is for the researcher to study elicited

data or questionnaire data (Beebe and Cummings, 1996). With this methodology, the

researcher uses written role-play questionnaires, which are designed in order to study

specific linguistic forms.

For this research project, I chose the first option, which is the naturalistic

approach, for purposes of conversation analysis. My original project was to observe the

linguistic practices of bilingual children in their everyday conversations. Therefore, the

first step to such an analysis was to record bilingual children’s daily use of language.

46

2. 2. Why family dinners?

This dissertation deals with a naturalistic approach and involves the recordings of

family dinners. The reason why I chose to study family dinners is because it is a very

special moment during the day to observe language practices. It is one of the rare

moments when the whole family reunites and all co-participants sit together around the

same table. Besides, dinnertime is a time for sharing and constitutes a good opportunity to

talk about their day and tell their family members what happened earlier for each

participant. Moreover, family dinners are also a cultural time. Indeed, “[m]ealtimes

facilitate the social construction of knowledge and moral perspectives through

communicative practices that characterizes these occasions” (Ochs & Shohet, 2006:36).

They allow the researchers to observe language practices, but also to account for cultural

practices and the acquisition of norms, habits and family values as well. Furthermore,

ethnographic evidence from various parts of the world has supported the argument that

food and eating are “saturated with social import” (Ochs & Shohet, 2006:46). Meals are

cultural times when every member of different generations and genders come to learn,

think, share, bargain, direct, provoke, transform each other’s way of acting, or interacting

with one another in any other way, all the while nourishing their body. It is a window

onto everyday socially established ways as much as it is a window onto everyday

linguistic codes.

2. 3. The videos

In order to record family dinners, I had to reach an agreement with Sandrine’s

family as to when and how I would come to the house and film mealtimes. I was able to

record several dinners in various family contexts.

Before recording anything, I went to meet the children for the first time, without

any recording devices. Indeed, it was important to me that the children knew who I was

before I started filming them. This way, they were less impressed by my presence as a

stranger in their house, if not at all. Besides, it allowed me to get in touch with the place

and to make contact with the three boys without any pressure or filming obligations.

The four dinners that I use for my work in this dissertation accumulate to a

recording time of 128 minutes and 54 seconds. They were all filmed in diverse situational

contexts and do not all share the exact same participants, aside from the three children

and the mother who are present in all recordings. In this section, I will present them and

47

their main characteristics following a chronological order. Table (1) synthesizes the

characteristics of each recording used in this study:

Table (1): Presentation of the recordings

Date Meal Duration Participants

November 30th 2014

Lunch

49 mn 00 sec

Mother, Father, 3 children

December 3rd 2014

Dinner

33 mn 43 sec

Mother, Au-pair girl, 3 children

December 10th 2014 Part 1

Dinner

04 mn 42 sec

Mother, Au-pair girl, 3 children

December 10th 2014 Part 2

Dinner

16 mn 03 sec

Mother, Au-pair girl, 3 children

January 31st 2015

Dinner

25 mn 26 sec

Mother, Father, 3 children

Total 128 mn 54 sec

Average Duration 32 mn 14 sec

On November 30th, I went to Sandrine and John’s home to record a family lunch

on a Sunday. The father was home, as it was during the weekend. I set up the equipment

and the camera around the table in the kitchen. I then decided to leave the room with the

aim of being as minimally intrusive as possible in the family’s habits. Once the

paraphernalia was in place, I stayed in the house in case any technical difficulty would

arise, but I went upstairs so as not to disturb the proper proceeding of the family meal.

The parents were asked to turn on the camera once they were ready to sit down to lunch,

and then to turn it off once all the children were finished and had left the table. This first

recording features both parents and the three children, and lasts for 49 minutes on the dot.

On December 3rd and December 10th, I had actually left the mother in charge of

the camera. Because we were having trouble fixing a date for filming, I had left the

equipment at the house after I had recorded the first meal on the Sunday before. After

agreeing on the best way to set the camera around the table, Sandrine was thus the one

who installed everything to film these two dinners. These two recordings feature the three

boys, the mother, and Christy, the American au-pair girl. As explained previously in this

paper, the children dine early on during the week, as they follow a “New Zealand way of

life”. The adults dine together later in the evening, while the children are already put to

bed. Therefore, the mother and Christy are present at the table but do not eat with the

48

boys. They are only there to supervise and to make sure that each of them eats as they

should.

The second recording, on December 3rd, lasts 33 minutes and 43 seconds. The

camera is set at the end of the table at a reasonable distance. Although this setting allows

the viewer to have a close shot of the participants around the table, it does not seem to

impress nor bother the children in their mealtime conversation.

The third recording, which took place on December 10th, lasts 20 minutes and 45

seconds in total but was originally filmed in two parts. There is a short ellipsis after 4

minutes and 42 seconds, and then the mother starts recording again. There is no way to

know exactly how long the camera was turned off. However, judging by the actions of the

co-participants, it should not be more than one minute, two at most. The peculiarity of

this footage is that the mother chose to film with a handheld camera. She later reported to

me that James was being a daredevil that night and that, for safety reasons, she thought it

was better to remove the tripod of the camera from the kitchen. This change of angle in

the recordings is interesting and allows the researcher to see another aspect of the dinner.

The viewer could almost have the feeling of being at the table with the participants.

However, one of the prominent problems with this technique is that it is impossible to see

the whole table at the same time. Therefore, all of the participants scarcely appear on

screen all together. The mother moves from one area of the kitchen to the other,

depending on who is having a conversation around the table.

The fourth and last recording features the three boys with both their parents. It was

recorded on January 31st 2015. Although it was a weeknight, the father was able to come

home early so the whole family had dinner together. Hence the main difference with the

other recordings that took place on a weeknight is that the parents are eating together with

their children. The au-pair girl, Christy, is absent from the recording. The camera is set at

the end of the table, closer than the times before. This might explain why the references to

the dinner being filmed are more numerous in this footage. The camera may have been

more intrusive for the children than the previous times. The recording lasts 25 minutes

and 26 seconds.

The whole corpus of data cumulated thus amounts to 128 minutes and 54 seconds,

with an average duration of 32 minutes and 13.5 seconds for each meal.

49

2. 4. Technical difficulties

Recording family dinners involving three children does not come without any

technical difficulties. The first complication encountered when filming a mealtime is to

manage having all the participants into the camera’s field of vision. The place of the

camera around the table is very strategic. Indeed, the equipment must be close enough to

have a good quality of image and sound. But, at the same time, it must be at a fair

distance in order to include everyone in the shot and to be as minimally intrusive as

possible.

Moreover, children generally behave differently than adults. They are not always

disciplined at the table. Indeed, they may sometimes get up and move around the table,

the kitchen, or even go from one room to another. Therefore, keeping every participant in

the camera’s eye at all times can be an issue in some recordings. Besides, the fact that

children are very lively and tend to react to almost everything that is happening around

the table, often leads to overlaps in conversation. The quality of the sound in the

recordings can therefore decline depending on how many co-participants speak at the

same time. Furthermore, children can be inclined to scream or yell to make themselves

heard; which does not improve the conditions for a better hearing of everyone’s

utterances.

In addition, the fact that the mother was in charge of turning on and off the camera

in all of the recordings does not allow the researcher to have control over when to start

and when to conclude the recording. Depending on each film, the footage starts more or

less early during the meal. This can call into question the average length of each dinner,

since not all of them start at the exact same moment.

Finally, one of the most difficult matters in recording an everyday life family

situation is what William Labov coined as “the observer’s paradox” (Labov, 1972:209).

Indeed, the latter states that:

“The aim of linguistic research in the community must be to find out how people talk when

they are not being systematically observed; yet we can only obtain this data by systematic

observation” (Labov, 1972:209).

Labov thus points out the fact that the observation of an event is inevitably

influenced by the presence of the observer. The ultimate goal of the linguist is to observe

the unobservable: this is the ‘observer’s paradox’ (Labov, 1972:209). This issue is by the

50

bye underlined by Letizia Caronia – Department of Education Sciences of the University

of Bologna – in one of her projects entitled “Évaluation d’une technologie immersive

dans le contexte d’un centre des sciences”, in partnership with André H. Caron –

Department of Communication of the University of Montreal (Caronia & Caron, 2005).

Gathering data on natural speech is therefore a challenge faced by all

sociolinguists while doing fieldwork. The authenticity of the data can be questioned

because it may not be representative of the speakers’ typical speech. The paradox lies in

the fact that the task of gathering natural data is undermined by the researcher’s presence

itself.

The advantage of observing children might be that the latter tend to be rather

spontaneous in their way of being. We can suppose that they forget about the camera after

a few minutes, but there can be no solid evidence on this matter. The researcher can never

really be certain whether the participants are fully aware that they are being filmed or not.

Except for when they refer to the presence of the camera, in which case the observer

knows that the participants are conscious of being recorded, most of the time there are no

means to be sure.

3. Coding

After recording several family dinners on videotape, the next step was to

transcribe the data. To that purpose, I used a specific software called ‘CLAN’, standing

for ‘Computerized Language Analysis’ (MacWhinney, 2000).

This program was developed by Brian MacWhinney and Leonid Spektor in the context of

the CHILDES and TalkBank projects. It is a free for download software that allows

transcription, coding, analysis and sharing of transcripts of conversations linked to either

audio or video media. Nevertheless, although I did use a tool specifically designed for the

coding and analysis of video data, transcribing remains a personal matter that is proper to

the researcher. Therefore, there is an undeniable part of subjectivity in transcribing a

conversation.

3. 1. Coding choices

One of the definitions of ‘Transcribe’ in the Webster dictionary (1961) is ‘to make

a copy of a dictated or recorded matter in longhand or on a machine’ (Merriam, G. &C.

1961). In other words, transcribing a conversation into a coded format involves

51

translating oral speech into a written text. However, a transcription is bound to be an

inexact reflection of what is said and meant by the participants, simply because it is done

by a third party who has their own interpretation of the situation. As Elinor Ochs puts it:

“transcription is a selective process reflecting theoretical goals and definitions” (Ochs,

1979:44). She also points out that the process of transcription has not been foregrounded

in empirical studies of verbal behavior, with the exception of conversation analysis

(Sacks, Schlegoff and Jefferson, 1974). Every researcher has their own personality and

their own singularities. As a result, every transcription most probably will differ from one

another. This proves that, no matter how objective the transcriber aspires to be, the

content of the transcription always has a part of subjectivity.

In order to minimize this part of subjectivity, it is necessary to set a range of rules

and norms in coding. These rules should make the transcription easy to read and provide

paramount information for the analysis of the data. Thanks to the software CLAN, this

information includes verbal and non-verbal elements of communication.

The following example illustrates different types of information that can be found

throughout the transcriptions:

Example 1

��������#$ �� >N�P3�0��1�5�7EF�-3@F3�E����������JB@F�� BA;@F7�>7�BG>>�67���,����������,�� -3@F3�K7E�4GF�-3@F3�(A@;53���������35F�� D793D67�EA@�BG>>�7@�>7�F;D3@F����������,�� @AF�-3@F3��>3GE���� �����9BJ�� E75AG7�>3�FSF7�

As we can see in this short extract, the names of the speakers appear on the left

and are preceded by an asterisk. Each name also refers to the status of each participant:

CHI for the children (CHI1-CH2-CHI3 ranked by age); MOT for the mother; FAT for the

father; CAR for the caretaker and OBS for the observer. Each participant has their

orthographic transcription, which enables the reader to visualize what was verbally said.

The use of slashes such as [/] indicates the repetition of a word or a phrase, in which case

the phrase is delimited by specific brackets < >. Double slashes [//] entail a repetition with

correction; and triple slashes [///] imply a repetition with reformulation.

The transcription also includes non-verbal elements of communication. These

elements are situated below the participants’ utterances, on the lines that start with the

percent sign ‘%’. The different types of non-verbal communication elements are

52

classified according to several categories. It can be a gesture (%gpx); pointing with the

index, or sideways palm (%xpnt); it can describe an action or an activity (%act) or the

contextual situation (%sit). A comment can also appear (%com) in order to provide

additional information if needed for full comprehension.

Regarding the issue of coding two languages, each participant was allocated with

what was assumed to be their dominant language. Whenever they speak in their dominant

language, no specific marking is used. However, whenever they switch languages and

produce an utterance in their second language, it is specified at the beginning of each

utterance that is coded. For example, if the father or Christy – whose dominant language

is English – produce an utterance in French, the transcription will begin with [- fra]. In the

same vein, if the mother or the children – whose dominant language was assumed to be

French for this project – switch to English, the coding will include [- eng]. Example 2

illustrates this point:

Example 2

��������#$ �� <7�535:7�F3�B:AFA����������9BJ�� F7@6�EA@�4D3E�6DA;F�7F�B3G?7�>3FRD3>7�67H3@F��#$���������B:A�� 535:7���9397��������!�.�� 0� 8D31�FG�9397E�67�B:AFA�����������!�.�� A:�;�A=����� ����!�.�� KAG��3�D7�4>A5=;@9�*>;�EA�:7�53@�F�47�E77@����������9BJ�� F7@6�EA@�4D3E�93G5:7�7F�B3G?7�>3FRD3>7�67H3@F��#$����������#$��� 0 �7@91�I:3F�0�1�����������35F�� E7�>QH7�7F�D793D67�67�>�3GFD7�5TFR�67E�?3;@E�67��#$ �7F�!�.���������#$��� 0 �7@91�I:3F�����������!�.�� I7��3�D7�4>A5=;@9�KAGD�8357�EA�KAG�53@�F�47�E77@�;@�8DA@F�A8�������� �������������F:7�53?7D3��

Such marking of differentiations for the two languages later enabled me to list and

identify the number of items, utterances and turns in each language for every participant.

A complete key to the transcription is provided in Appendix 2. It corresponds to the

transcription conventions of the software CLAN, which is the tool that was used to

transcribe and analyze the totality of the data collected for this research project.

3. 2. Transcription problems

One of the main issues regarding transcription is to choose what to transcribe and

what to leave aside. The choices of the researcher on this matter already reflect his/her

53

particular interests along with his/her hypotheses. However, as already pointed out in this

chapter, no transcription can be completely and undeniably neutral. It is bound to be an

inexact reflection of what is said and meant by the participants, simply because the

content of the transcription is based on what the researcher is looking for. Therefore, it

always has a part of subjectivity:

“Pour clarifier le codage, il faut préciser son but, la visée du projet, et l’utilisation qui est

envisagée des corpus. En effet, un codage ayant pour but une étude phonétique, phonologique,

syntaxique, sémantique, ou pragmatique n’aura pas du tout la même structure et ne sera pas basé

sur les mêmes présupposés” (Morgenstern & Parisse, 2007).

When properly starting to transcribe the videos, one of the first problems that

rapidly occurred was to define utterances for each participant. An utterance is different

from a turn in conversation. Turn-taking is at the basis of conversational analysis and

refers to the idea that when we converse we take turns in speaking (A-B-A-B-A). An

utterance, on the other hand, is basically a vocal statement. It can be only part of a turn.

The software CLAN, which I used for transcribing the totality of the data, uses a format

that offers a large number of lines in order to codify as many elements as needed for

analysis. It allowed me to transcribe both utterances and turns. Nevertheless, boundaries

are often not clear and it is difficult to know exactly when to segment a participant’s turn

into several utterances. My choice was to consider a new utterance after a pause that was

long enough for me to notice it and to react on it while transcribing. Usually, these were

equivalent to approximately two-second pauses within a turn.

Transcribing utterances is rather fastidious and takes a lot of time. But on the other

hand, it allows the researcher to notice some elements that would have gone unnoticed if

the conversation had been fragmented only in turns. For example, it becomes easier to

notice a gaze, a pointing gesture or an overlap if we deal with shorter portions of the

actions. It permits a sort of ‘zooming in’ on the situation and to further study the turns of

each participant.

Furthermore, the transcription process presents some difficulties regarding a

number of issues. For example, Ochs suggests that the transcription of children’s speech

should be different than for the adults because their cognitive capacity and

communicative skills do not match (Ochs, 1979:47). Besides, the simultaneity of several

54

utterances can also be an issue because there is no transcription convention that is precise

enough to report these overlaps. In the software CLAN, overlapping utterances are

written below one another. If the simultaneity were marked vertically, on the same line

but divided into two columns for the two participants, overlaps would be easier to

visualize and account for (Ochs, 1979). The same principle could also be applied for

verbal and non-verbal transcriptions. Instead of having a main line with the participant’s

verbal speech and a second line describing his/her gesture, there would only be one line

divided into two columns and presenting the verbal elements on one side and the non-

verbal on the other (Ochs, 1979).

Another difficulty in transcribing is to be certain of what has been exactly uttered

by each speaker. Sometimes the quality of the recording does not allow a clear

comprehension of an utterance. Sometimes an utterance is masked by an overlap (or

several overlaps), which makes it difficult to fully understand it. Every time an utterance

was unintelligible, I chose to transcribe it as ‘xxx’. This symbol entails that I was aware

that the participant had produced an utterance, but unfortunately I was unable to

distinguish it clearly. Occasionally, the problem could also come from the English accent.

I recognize that I found it difficult in the beginning to perfectly discern the father’s

utterances, due to his New Zealander accent. Nevertheless, I tried to be as accurate as

possible in my transcription and to remain as close as possible to the original utterance.

4. Concluding remarks

The collection and transcription of the data proved to be a very long process,

which was very interesting and certainly enriching. Being able to collect my own data and

to work on my own corpus has been thrilling. As long and tedious as it was, I definitely

learned a lot while recording and coding the videos. This project enabled me to proceed

with a quantitative and qualitative analysis on language practices of bilingual children.

One of the things that I like about this research project is that the data is the result of my

own personal work along with a tremendous cooperation on the part of Sandrine’s family.

I am pleased that I was able to conduct my own methodology for this project.

Collecting the data for such a project is thus a lingering yet dynamic process. It

evolves along with the researcher and changes according to his/her interests or

hypotheses. However, one has to keep in mind that any transcription has its limits, as it

reflects the researcher’s subjectivity. As objective and neutral as I tried to stay throughout

55

the collection of the data, I am aware that my work reflects my own interpretation, first

when transcribing the data, and then when interpreting it. My transcriptions were not

submitted to any judge – apart from the correctness of the software CLAN – or to any

observer. However, I acknowledge that the reliability of a research study is also subjected

to an external arbitration and agreement in order to be validated. Therefore, my

methodology is not totally impartial. It is a reflection of my own choices and my own

perception on my research project.

56

57

IV. QUANTITATIVE RESULTS.

1. General linguistic overview

For this research project, the children have been observed periodically during

family meals over a period of three months. The results of this study arise from the

totality of the data collected, which is a total of 128 minutes and 54 seconds of footage.

The recordings spread over four family meals; each session lasting approximately 32

minutes. The number of linguistic items – also referred to as tokens – correspond to the

total number of words produced by each participant. In total, Thomas produced 2 797

linguistic items across the four sessions. Oliver produced 2 545 items, and James 432.

The first observation that can be made is that Thomas and Oliver share approximately the

same amount of tokens. James, on the other hand, the youngest boy, produces almost six

times less items than his brothers during the recorded times. These numbers are illustrated

in the table below:

Table (2): Number of Linguistic Items (Tokens)

The second observation that can be made is that all three children have a superior

number of linguistic items in French than in English. This means that the children use a

higher number of words in French during their conversations. Indeed, Thomas uses 2,7

times more words in French than he does in English (2 046/751). For Oliver the rate is at

2,5 times (1 821/724), and for James approximately 2,7 times too (315/117). In other

words, 73% of Thomas’s and James’s items are French against 27% English. For Oliver,

the ratio is of 72% French tokens for 28% English. The following graphs in Figure (2)

illustrate those rates.

Thomas Oliver James

French 2 046 1 821 315

English 751 724 117

Total 2 797 2 545 432

58

Figure (2). Distribution of linguistic items.

It is interesting to note that, although the numbers of tokens greatly differ between

Thomas and Oliver on the one hand and James on the other hand, the ratio between

French and English items is fairly the same. Approximately two thirds of the children’s

items are in French and one third in English. These results may lead to think that the

children are more proficient in French. It seems to be the dominant language for the three

of them. Only for Oliver is the scale slightly reduced, which suggests that he might be

using English slightly more frequently than his brothers, but that French would still be his

dominant language. This hypothesis is to be verified by means of the study of the

children’s lexical diversity along with their utterances and turns. A closer look at their

interlocutors’ utterances is also needed in order to determine whether the adults speak

more in French or in English.

2. Lexical diversity (Type/Token Ratio) When looking at the children’s linguistic skills, it is interesting to compare the

number of different words (types) with the total number of words the children produce

(tokens). To that purpose, Type/Token Ratios (TTR) have been extensively used as an

index of lexical diversity (Richards, 1987). If we look at the number of different words

(types) in absolute terms for each child along with their mother, here are the numbers

obtained:

Table (3): Number of Different Words (Types)

Thomas Oliver James Mother

French 519 452 124 936

English 240 203 29 275

Total 759 655 153 1 211

73%!

27%!

Thomas's!tokens!

French'English' 72%!

28%!

Oliver's!tokens!

French'English' 73%!

27%!

James's!tokens!

French'English'

59

These first results lead to think that the children’s vocabulary is more diverse in

French than in English. In other words, their lexical variety seems to be higher in French.

However, these numbers are more accurate when compared to the total number of words

of each participant (tokens). But when dealing with TTRs, a problem occurs when the

speech samples differ from one participant to another. As Brian Richards pointed out:

“[t]he ratio […] falls away rapidly as the number of tokens increases” (Richards,

1987:205). Indeed, Richards showed in his study that TTRs can be misleading. Based on

two different speech samples, one of a two-year old on the one hand and one of a

professor of linguistics on the other hand, his graphs seemed to demonstrate that the

speech of the two-year old showed greater lexical diversity than the prose of the professor

(Richards, 1987:205). This was due to the fact that the higher the number of tokens raises,

the more the TTRs tend to fall.

As a consequence, it does not seem wise to compare the three children simply

based on their total number of words. As already noticed in the first section, the number

of tokens greatly differs between Thomas and Oliver on the one hand and James on the

other hand. Nevertheless, an alternative may be to use J. B. Carroll’s lexical diversity

measure. Indeed, according to Carroll:

“A measure of vocabulary diversity that is approximately independent of sample size is the

number of different words divided by the square root of twice the number of words in the sample”

(Carroll, 1964:54; as cited in Richards, 1987:208).

This is the measure that I used in order to account for the children’s lexical diversity. One

has to note that although TTR is regularly negative when related to the sample size, with

this calculation the index becomes positively related to the number of tokens. Tables (4)

and (5) summarize the children’s lexical diversity in each language. I compared their

diversity in French with their mother’s and their diversity in English with their father’s.

Table (4): Lexical diversity in French (Carroll’s measure)

Thomas Oliver James Mother

Types 519 452 124 936

Tokens 2 046 1 821 315 5 036

TTR Types___ √(2xTokens)

8.113

7.490

4.940

9.326

60

Table (5): Lexical diversity in English (Carroll’s measure)

Thomas Oliver James Father

Types 240 203 29 577

Tokens 751 724 117 2 717

TTR Types___ √(2xTokens)

6.193

5.335

1.896

7.827

These tables allow us to draw the conclusion that the children’s vocabulary is

more diverse in French than in English. Indeed, the Type/Token Ratio is higher in French

for all three children. This thus suggests that the children’s utterances in French are

composed of a greater number of different types of items. Hence their speech is more

varied in French.

The results also appear to be proportional to the children’s age. Thomas’s vocabulary is a

bit more varied than Oliver’s in both languages, although by less than 0.9 percentage

point. James on the other hand, is slightly further behind his brothers, which is in

complete adequacy with the other results, especially since he is the youngest boy.

Another interesting thing to notice is that both parents have a higher diversity of

vocabulary than their children, each in their native language. The mother’s TTR is a bit

higher than the father’s, implying that she has a slightly higher lexical variety.

Hence, the first results suggest that the children are more proficient in French than

in English. Their number of words is indeed higher in French and the diversity of their

vocabulary is more important in French as well. In other words, the three children

produce more words and within a greater variety in French than in English. It then seems

that French is the dominant language for the three of them. In order to confirm this

hypothesis, a closer look at their utterances in each language is needed. The following

section will deal with the children’s Mean Length of Utterance. But we have to keep in

mind that if the children are indeed dominant in French, this could also be the result of

their interlocutors’ speech. The next section will therefore account for the adults’

utterances as well, to determine whether they speak more in French or in English and thus

explain or contradict the finding that the children are more proficient in French.

61

3. Mean Length of Utterance (MLU)

The Mean Length of Utterance (MLU) refers to the average number of words per

utterance. As in the previous section, the results arise from the totality of the data

spreading over four family dinners. As already pointed out in the chapter on coding the

data, it was difficult to define where an utterance begins and ends for each participant.

Indeed, boundaries between utterances and turns are often not clear and it is difficult to

know exactly how to segment a participant’s turn into several utterances. For this

research, I considered that a new utterance began after a pause that was long enough to be

noticed, usually equivalent to approximately two-second pauses within a turn. Table (6)

summarizes the children’s number of utterances along with their Mean Length of

Utterance.

Table (6): Global Mean Length of Utterance

Thomas Oliver James

Utterances 798 826 314

Words 2 674 2 483 430

MLU 3.351 3.006 1.369

In total, Thomas produced 798 utterances; Oliver accumulated 826 utterances, and

James produced a total of 314 utterances across the four sessions. When related to their

number of words, these results imply that Thomas’s utterances consist of 3.35 words on

average, while Oliver’s utterances have a mean length of approximately 3 words, and

James’s of 1.37 words. Thomas and Oliver are thus fairly close to one another regarding

their MLU, suggesting that their language development is relatively comparable. James,

on the other hand, has yet to reach his brothers’ global proficiency. Another interesting

thing to notice is that, although Oliver produces fewer words than his older brother, his

number of utterances is higher. This means that Oliver segments his turns in a more

important manner than Thomas.

Interesting differences are noticeable when differentiating the MLU in the

children’s two languages. Tables (7) and (8) result from these calculations:

62

Table (7): Mean Length of Utterance in French

Thomas Oliver James

Utterances 577 591 221

Words 1 989 1 782 313

MLU 3.447 3.015 1.416

Table (8): Mean Length of Utterance in English

Thomas’s total of 798 utterances thus divides up into 577 in French and 221 in

English. Oliver accumulated 826 utterances, including 591 in French and 235 in English.

As for James, 221 of his utterances were in French and 93 in English, adding up to a total

of 314. These results corroborate the first hypothesis that the children are more proficient

in French than in English. Indeed, not only do the three boys use more words in French

than in English, but they also produce a higher number of utterances in French.

Comparatively, this leads to think that they express themselves more often in French than

in English. Besides, as we have shown, their lexical diversity is also higher in French.

Furthermore, if we compare their MLU in the two languages, all the children have

a higher Mean Length of Utterance in French. This means that their utterances in French

consist of more words than their utterances in English. In other words, they are longer and

slightly more elaborate. To visualize the differences in the children’s utterances regarding

each language, Figure (3) presents a comparison of their MLU. When looking at these

graphs, we notice that the children’s MLU is slightly higher in French than English in all

three cases. The most important gap between the two languages affects Thomas. His

French MLU is of 3.447 words per utterance, while it is of 3.1 words in English. While

this difference in proficiency is not gigantic, it is however the most sizable of the three

siblings.

Thomas Oliver James

Utterances 221 235 93

Words 685 701 117

MLU 3.100 2.983 1.258

63

Figure (3). Comparison of the children’s MLU

On the other hand, it is difficult not to notice that Oliver’s length of utterances is

closely the same from one language to another. Indeed, it is only 3.015 in French for

2.983 in English, leading to the smallest ratio among the children. This leads to believe

that Oliver may practically be as proficient in both languages. Although his utterances are

more numerous in French – as it is the case for his two siblings – his proportion to his

number of words is the only one to be that reduced between the two languages. This

entails that his utterances in English consist of nearly as many words as his utterances in

French. In other words, they are relatively equal in terms of length.

Now the reason why the children produce more utterances in French may come

from their interlocutors themselves. In order to clarify this point, Table (9) summarizes

the adults’ MLU:

Table (9): Adults’ Mean Length of Utterance

French English

Mother Father Caretaker Mother Father Caretaker

Words 5 002 177 - 763 2 247 791

Utterances 975 49 - 168 518 172

MLU 5.130 3.612 - 4.542 4.338 4.599

These results were obtained over the four recording sessions. The data for the

mother thus come from the four meals in which she participated. The data for the father

and for the caretaker represent the two dinners they were each part of. According to the

results, several conclusions can be drawn. First of all, the au-pair girl, Christy – referred

to as caretaker in the transcriptions and ergo in the table above – speaks exclusively in

0'0,5'1'

1,5'2'

2,5'3'

3,5'4'

Thomas' Oliver' James'

French'MLU'English'MLU'Global'MLU'

64

English. The numbers show that she does not utter one word in French over the two

dinners in which she participated. Secondly, the father speaks more English than French,

although he does switch to French now and then, contrarily to the caretaker. Still, he

produces more than ten times more utterances in English. The mother, on the other hand,

speaks a lot more French than English. Indeed, she produces almost six times more

utterances in her native language.

If we add up the total numbers of utterances produced by the adults, we obtain

1 024 (975+49) in French, for 858 (168+518+172) in English. It becomes very clear then

that the language in which the children are spoken to the most around the table is French.

This could explain that the three boys produce in their turn more utterances in French, as

it seems to be the dominant language among their interlocutors.

To sum up, the children’s utterances are longer in French than in English. In a

global manner, it can thus be deduced that the three boys are more proficient in French.

Yet some individual differences are noticeable. While Oliver’s MLU is very close from

one language to another, Thomas shows a clearer gap suggesting that French occupies a

more dominant place in his everyday conversations. However, one should note that his

MLU in English is still higher than Oliver’s all in all. The results presented above seem to

suggest a proportionality of the children’s language proficiency according to their age.

Furthermore, it can be observed that French is also the language that appears to be spoken

the most among the adults. This claim may explain why the children would produce more

utterances in French as well, as it is also dominant in their interlocutors’ speech.

Nevertheless, the fact remains that studying the Mean Length of Utterance can be a

delicate matter due to the difficulty of defining and demarcating utterances, as already

mentioned. For more complete results, it is useful to look at the Mean Length of Turn

(MLT), which question is treated in the next section.

4. Mean Length of Turn (MLT)

The Mean Length of Turn (MLT) refers to the average number of words per turn

at talk. Indeed, a conversation is based on turn-taking, that is to say the idea that when we

converse we take turns in speaking. The rules that govern turn-taking are referred to as

the ‘local management system’ (Sacks H., Schlegoff E.A., Jefferson G., 1974).

65

Sometimes, if two or more participants speak at the same time, a turn-taking can be

violated. This phenomenon is called an overlap and often occurs for a short duration that

can be counted in microseconds. Many overlaps are observable in the transcripts of the

footage on which this research is based. This section is a study of the children’s turns in

conversation, in terms of number of turns and Mean Length of each Turn. The purpose is

to confirm that the three siblings’ proficiency in French is dominant, as was suggested in

the previous sections after looking at their MLU and their lexical diversity.

It is important to note that, comparatively to the previous analysis, the numbers of

words and utterances slightly differ from the preceding tables. This disparity in the

numbers is due to the fact that the algorithm that is used by the software CLAN to

generate the results related to MLT is different than the one used to calculate the MLU.

Indeed, the main difference is that the utterances that are coded as ‘unintelligible’ in the

transcription (“xxx”) are excluded as utterances in the count of MLU whereas they are

included in the count of MLT. This explains why the numbers are not exactly equal; and

it involves that the number of utterances is higher in the study of MLT. Table (10) is a

general overview of the children’s turns and MLT.

Table (10): Global Mean Length of Turn

Thomas Oliver James

Utterances 861 855 335

Turns 707 698 313

Words 2 861 2 575 453

Words/Turn 4.047 3.689 1.447

Utterances/Turn 1.218 1.225 1.070

These data affect the four sessions of recording as they do for the whole

quantitative study. The results show that the number of turns is globally proportional to

the number of utterances and words. Even with a different calculation of the number of

utterances (the MLT including unintelligible utterances while the MLU excluded them), it

still is observable that Thomas produces more utterances than Oliver who himself

generates more utterances than James. When relating the number of turns and the number

of words, it is not surprising that the children’s MLT is slightly higher than the MLU. For

example, Thomas has a global MLU of 3.351 words, while his MLT shows that his turns

66

are composed of 4.047 words on average. Thomas uses more words per turn than per

utterance simply because a turn comprises more than one utterance on average. The same

pattern occurs with the other two siblings. While Oliver’s global MLU is of 3.006 words

and James’s of 1.369 words per utterance, the numbers are higher per turn as they raise up

to respectively 3.689 and 1.447 words per turn.

Interestingly, although Thomas has a higher MLT than Oliver – meaning that he

uses more words per turn than his little brother – Oliver actually has a higher rate when

dealing with the utterances per turn. In other words, Oliver’s turns comprise fewer words

but more utterances than his big brother’s. This means that Oliver’s utterances are shorter

than Thomas’s, which corroborates the results obtained when looking at the children’s

MLU. It also means that Oliver fragments his turns more than his older brother. So when

Oliver speaks, his turns in conversation are more divided than the other boys’. If we look

at James’s MLT for example, he speaks fewer words than his two brothers and produces

less utterances throughout the recordings. Yet his results are proportional when dividing

them per his number of turns. He produces 1.447 words per turn and 1.070 utterances per

turn on average. These calculations validate the previous observation that his number of

words per turn is higher than his number of words per utterances. They also entail that his

turns are the less fragmented of all siblings, as his ratio of utterances per turn is the lowest

among the three boys. Then again, as James produces fewer words than his brothers and

is situated further behind them in his linguistic development, these results are to be

considered with caution.

It will now be interesting to look at the potential differences in MLT regarding the

children’s two languages. The following charts, Tables (11) and (12), summarize these

calculations:

Table (11): Mean Length of Turn in French

Thomas Oliver James

Utterances 617 609 242

Turns 495 484 223

Words 2 086 1 840 336

Words/Turn 4.214 3.802 1.507

Utterances/Turn 1.246 1.258 1.085

67

Table (12): Mean Length of Turn in English

Thomas Oliver James

Utterances 244 246 93

Turns 212 214 90

Words 775 735 117

Words/Turn 3.656 3.435 1.300

Utterances/Turn 1.151 1.150 1.033

Similarly to the children’s MLU, their number of words and utterances divides up

with a larger proportion for the French language. Thomas’s total of 861 utterances thus

divides up into 617 in French and 244 in English. Oliver accumulated 855 utterances,

including 609 in French and 246 in English. As for James, 242 of his utterances were in

French and 93 in English, adding up to a total of 335. These results corroborate once

again the idea that the children are more proficient in French than in English, as they

produce a higher number of words, utterances and turns in French rather than in English.

This leads to think that they express themselves more often in French. Indeed, the

children produce approximately twice as many turns in French throughout the four

sessions: Thomas, for example, has a number of 495 turns in French against 212 in

English; Oliver accumulates 484 turns in French for 214 in English; and James 223 for 90

in English.

Again, this dominance of French in the turns of the children may be explained by

the fact that it is also the dominant language among the adults speaking at the table. What

is worth noticing, comparatively to what has been observed regarding the siblings’ global

MLT, is that all of the children’s turns seem to be less fragmented in English than in

French. Indeed, while Thomas’s turns comprise 1.246 utterances in French, they are only

composed of 1.151 utterances in English. The same patterns can be observed with Oliver:

1.258 utterances per turn in French for 1.150 in English; and with James: 1.085 utterances

per turn in French against 1.033 in English. It then seems that the children divide their

talk in a lesser way when speaking in English. This could lead to think that they are more

fluent in English; their speech flows better. Indeed, the fewer utterances there are per

turn, the less fragmented the turns are.

68

One other observation that can be made is that, contrarily to the boys’ global

MLT, Oliver’s ratio of utterances per turn is no longer the highest one. Indeed, Thomas

has – but for minor details – the highest ratio of utterances per turn when it comes to

English. In other words, this means that not only is Oliver’s speech more fluent in English

than in French, but it is even a little bit more fluent than his older brother’s. These results

may be more obvious when comparing the children’s Mean Length of Turn – in terms of

utterances per turn – as in Figure (4):

Figure (4). Comparison of the children’s MLT in terms of utterances

This figure illustrates some of the points that have been observed beforehand.

Oliver has the highest number of utterances per turn – globally and in French – meaning

that his turns are more fragmented than his brothers’. However, when it comes to English,

Oliver’s MLT is fairly equal to Thomas’s – even slightly lower – suggesting that Oliver’s

speech is a bit more fluent than his brother’s in English, as he produces fewer utterances

per turn. The three brothers’ Mean Lengths of Turn follow a rather similar pattern in that

their turns in French are longer than their turns in English. These results are proportional

to their number of words and utterances; and they corroborate the results obtained

regarding the Mean Length of their Utterances in the previous section.

As a concluding remark, the study of the children’s Mean Length of Turn seems to

validate the hypothesis proposed by the study of their Mean Length of Utterance. Indeed,

the boys’ MLT is higher in French than in English. It can then be concluded that the

children are more proficient in French. They use more words and produce longer turns in

the latter language. These results are consistent with their MLU being slightly higher in

0'0,2'0,4'0,6'0,8'1'

1,2'1,4'

Thomas' Oliver' James'

French'MLT'English'MLT'Global'MLT'

69

French as well. On the other hand, the children’s number of utterances per turn is also

higher in French than in English. It can thus be deduced that, although the children’s turns

in English consist of fewer words, they are globally less fragmented. In other words, there

are fewer noticeable pauses in their talk. This suggests that it is easier for the three of

them to produce a turn in English. Their speech is therefore more fluent. On the other

hand, some individual differences are always observable. While globally most results are

proportional to the children’s age and linguistic development, Oliver’s speech in English

seems to flow slightly better than his older brother Thomas.

5. The issue of gestures The question of language dominance may lead to investigations as to how a child

can compensate a lesser proficiency in one of his two languages. The idea in this section

is simply to look at the children’s number of gestures throughout the data. The statistics

regarding their gestures may be proportional to their linguistic abilities in the two

languages. Or, on the contrary, the results may show an opposite tendency, therefore

suggesting that the children might use gestures in order to counterbalance their linguistic

discrepancies.

First, regarding the counting of the gestures, one has to know that the

transcriptions of the data may not report every single gesture produced by every

participant. Gestures were transcribed when they were openly visible in the conversation.

Depending on the quality of the recording, or on the relative subjectivity of the transcriber

– a point already addressed in this dissertation – the number of gestures obtained in this

study is not indisputable. A few of them might have been missed. Indeed, some may

differ on the definition of what a gesture is. Therefore, several researchers transcribing

the same video may not necessarily agree on an identical number of gestures. For this

analysis, the videos have been transcribed and analyzed by only one researcher. The

statistics obtained must therefore be considered with caution, as they may not be an exact

reflection of the conversational situations. Nonetheless, the goal was to be as thorough as

possible and to transcribe every gesture that was noticeable and relevant in the

participants’ interactions.

In order for these results to make sense, the meaning of ‘gesture’ must be

delimited. What were considered gestures in this study were primarily head and hand

70

movements with a communicative function. For example, nodding and shaking the head

were counted as gestures. Concerning hand movements, any deliberate movement in

space of either one hand or both was considered a gesture. A hand or a finger raised in the

air, a sweeping movement with the hand, an open palm, a circular movement with the

hand or the wrist, are some examples of gestures produced by the participants. Raising the

shoulders may also have been considered a gesture. If several gestures were part of a

same position of ‘shrugging’, then they were counted as only one gesture. Indeed, as

Streeck (2009) pointed out about what he called the ‘shrug position’:

“A prototypical shrug position involves several body-parts, the eyebrows (which are being

raised), the hands (which are turned so the palms face up), the forearms (which may be lifted)

and the shoulders (which are also raised). In addition, the head may be tilted. Thus, shrugs are

compound enactments” (Streeck, 2009:189; as cited in Debras, 2013:109).

As a consequence, if the gestures were part of an overall ‘shrug position’, then only one

gesture was counted. However, taken separately or in a different context than a shrug,

each gesture was considered unique. Most of the gestures that are used for these results

were co-verbal. They accompanied the speakers’ utterances, in either of their languages.

A few gestures – mostly head movements – were not co-verbal but were simply a silent

response to an interlocutor: shaking the head or nodding, for instance. These particular

gestures – although they were not co-verbal – were coded in the language that the co-

participants were speaking at the time they occurred. If the speakers were exchanging in

French, a silent nod was then coded in French. If they were speaking English, a shake of

the head was coded as a gesture in English. Table (13) summarizes the numbers that arose

from the transcription of gestures in the data:

Table (13): Gestures per Language (%GPX)

Thomas Oliver James

French 137 84 31

English 29 24 3

Total 166 108 34

Several observations ensue from this table. First of all, the numbers of gestures

appear to be proportional to the children’s age. Indeed, the results imply that the oldest

boy, Thomas, produces more gestures than his younger brother Oliver, who himself uses

71

more gestures than James. The first two boys are relatively close to one another in their

use of gestures, although Oliver still uses fewer gestures than his older brother. James, on

the other hand, produces only a third of Oliver’s total amount of gestures, and almost a

fifth of Thomas’s. These results seem to be proportional to the children’s number of

linguistic items. These conclusions can be drawn from the absolute numbers of gestures.

Nevertheless, because the size of the samples of linguistic items differs for every child, it

is necessary to consider the ratio between the number of tokens (cf. Table (2)) and the

number of gestures (cf. Table (13)). The following table illustrates those ratios:

Table (14): Tokens/Gestures

Thomas Oliver James

French 14.9 21.7 10.2

English 25.9 30.2 39.0

Total 16.9 23.5 12.7

From the calculations above, we can thus deduce that Thomas globally makes a

gesture every 16.9 words, while Oliver every 23.5 words, and James every 12.7 words. It

thus seems that James actually uses more gestures than his older brothers. The results in

absolute terms are therefore misleading since James’s sample of linguistic items is

reduced compared to Thomas and Oliver. It seems that James is the one who uses more

gestures after all – if we consider his total number of gestures.

When comparing the children’s number of gestures in each of their languages, we

can clearly detect that the children’s utterances in French are accompanied by more

gestures than their utterances in English. It is the case for all three boys. According to

Table (14), Thomas uses a gesture in French every 14.9 words against one gesture every

25.9 words in English. Similarly, Oliver makes one gesture every 21.7 words in French

for one gesture every 30.2 words in English. Surprisingly, James is the one who uses the

more gestures in French (one gesture every 10.2 words) and the fewer gestures in English

(one gesture per 39 words). It thus seems inaccurate to claim that the use of gestures

increases in accordance with the child’s age. Furthermore, although all three children

seem to use more gestures accompanying their utterances in French than in English, some

clear individual differences arise when it comes to the distribution of their total number of

gestures. The following graphs in Figure (5) illustrate this point:

72

Figure (5). Distribution of the children’s gestures

While the children’s number of gestures follows the identical trend of being higher

in French, the statistics show that the total number of gestures is not allocated in an

identical manner for all the three boys. Indeed, Figure (5) above demonstrates that

Oliver’s proportion of gestures accompanying his utterances in English is higher than for

his two brothers. His gestures while speaking English represent 22% of his total number

of gestures. This proportion is smaller for Thomas, for whom it represents 17%, and it

only reaches 9% for James. As a result, James’s proportion of gestures accompanying his

utterances in French is the highest of the three children, with a distribution of 91%. On

the other hand, Oliver has the lowest rate of allocation when it comes to his gestures in

French with a proportion of 78%.

It is interesting to note that, contrarily to the distribution of the children’s

linguistic items, these graphs denote some clear individual differences among the

siblings. Indeed, although the number of tokens greatly differed between Thomas and

Oliver on the one hand and James on the other hand, the ratio between their French and

English items was fairly the same. Approximately two thirds of the children’s items were

in French and one third in English. Here the situation is different as the percentages

greatly vary from one child to another. The general tendency remains but to a different

degree for each boy.

To sum up, this section has shown that the three boys tend to use more gestures

when speaking French than when speaking English. It is also interesting to note that the

youngest boy seems to use gestures more often than his siblings, in French and in total.

The results therefore suggest that the children do not compensate their linguistic

dominance in French with more gestures in English. On the contrary, it shows that the

children tend to use more gestures in their dominant language. However, although this

83%'

17%'

Thomas's!gestures!

Accompanying'French'Accompanying'English'

78%'

22%'

Oliver's!gestures!

Accompanying'French'

Accompanying'English'

91%'

9%'

James's!gestures!

Accompanying'French'

Accompanying'English'

73

tendency is shared by all three children, their amount of gestures are not equally

distributed in each language.

6. Concluding remarks

This quantitative analysis of the children and adults’ language practices has led to

demonstrate that Thomas, Oliver and James are more proficient in French than in English.

Indeed, first we have shown that in total they all use more words in French throughout the

four mealtimes under study. Then, when looking at the diversity of their vocabulary, we

have demonstrated that there are fewer different types of items in English, thus suggesting

that the boys’ lexical diversity is higher in French. In other words, the three children

produce more words and within a greater variety in French than in English. We have

confirmed this hypothesis of a dominance in French with the calculation of the children’s

Mean Length of Utterances and Turns, which both show that they produce longer

utterances and turns in French. Yet, this finding may be moderated by the observation that

the children’s turns are less fragmented in English – as they globally consist of fewer

utterances – implying that their speech flows slightly better. Some individual differences

suggest that Oliver’s speech in English might be slightly less segmented than his older

brother’s, Thomas. Other than that, in a global manner, the children’s linguistic skills

follow a proportional pattern in relation to their age.

It has also been shown that French is likewise dominant among the children’s

interlocutors’ speech. Indeed, their main interlocutor is their mother, and her total number

of words shows that she speaks French more often than English. Even though there is

always one adult whose native language is English present at the table, the total number

of words and utterances produced in French is still higher. This may be the reason why

the boys speak more French themselves, as they adapt to their interlocutor(s).

Finally, the study of gestures related to the children’s use of their two languages,

tends to show that they use more gestures in their dominant language. As a result, we can

conclude that the children do not compensate their linguistic discrepancies with a higher

number of gestures. On the contrary, their number of gestures is proportional to their

amount of verbal communication in each language.

74

75

V. THE ALTERNATE USE OF TWO LANGUAGES The purpose of this chapter is to analyze examples that reflect how the children

assert their place in family conversation. The detailed study of the family interactions

revealed four main trends in the way the participants juggle their two languages. Indeed,

they may voluntarily create breaks in the conversation, or on the contrary, they may adapt

their speech to their interlocutor’s in order to request attention. The children’s switching

between languages may also result from their association of one language with one

interlocutor, regardless of which language they are spoken in. In some cases, it can also

be related to the participants’ emotional state. This analysis offers a closer look at some

selected extracts that are interesting for they offer a window on the children’s way of

dealing with their two languages.

1. Family conversation: overview of a selected extract This first section consists of a detailed study of a short extract that is taken from

the recording of the dinner on January 31st 2015. It deals with an overview of how the

family conversation functions, using different elements of conversation analysis. It adopts

the point of view that people cooperate together in the construction of a conversation

(Levinson, 1983). In this footage, both parents were present on a weeknight. It is an

approximately three-minute long extract of conversation. As a reminder, at the time of the

recording, the three boys were respectively 5;11, 4;6 and 2;3. The complete transcription

of this excerpt can be found in Appendix 3.

First of all, one dominant phenomenon that we can identify is the large number of

overlaps in conversation. In the case of an overlap, turn taking is violated in that two or

more participants speak at the same time, often for a short duration. According to the

video’s transcription, in which overlaps are marked using the symbol ‘+<’, there are 24

utterances that occur in violation of another turn-taking in the conversation. Most of them

originate from the children. Table (15) illustrates this point:

Table 15: Overlaps Originated by Each Participant CHI1 7 CHI2 6 CHI3 7 MOT 1 FAT 3

Total 24

76

As we can see in this table, the family’s conversation is filled with overlapping

utterances, and most of these overlaps result from the children. One interpretation could

be that the children are less aware of the rules of conversation than the adults. They may

not necessarily pay attention to turn-taking, nor easily detect a TRP: Transition Relevance

Place (Sacks H., Schlegoff E.A., Jefferson G., 1974). These are junctures at which the

turn at talk could legitimately pass from one speaker to another. It marks the end of a tone

constructional unit (Levinson, 1983). But the children may not necessarily pay attention

to these. As a result, they often begin a new turn before their co-participant’s turn is over.

Besides, many – if not all – children are often seeking after their parents’ attention, which

could explain why they would be impatient to be the one speaking. Furthermore, as there

are three children around the table, each of them more or less has to compete with the

others to get the next turn at talk and be heard. They can thus create breaks in the

conversation, as it happens in the following example (1):

Example (1). ������!�.�� KAG�:36�B3@53=7E�3F�E5:AA>�����������#$��� ����3:�;�EAGB7��0�1��������#$ �� 0�7@91�@A����������#$��� 0�7@91�@A�4753GE7�$� � ��������#$��� ���3:�;�EAGB7��� ������(*.�� AG;�5�7EF�>3�EAGB7��� � ����9BJ�� :T5:7�>3�FSF7��

�������#$��� 0�7@91�6366K��� ������!�.�� ??:�A��� �������#$��� 0�7@91�I:7@�0�1�� �������#$��� 0�7@91�I:7@�0�1� �������#$��� 0�7@91�I:7@�$��3�?�9A@@3�47��� � � � � � � � � � � � � � � � � � � � � � � � � � � �E7H7@FK�� ��

� In this example, the fact that James (CHI3) intervenes to talk about the soup while

the father is asking a question creates a break in the conversation. Oliver (CHI2) starts

answering with “No because I”, then he pauses because James interrupts him again. The

mother responds to James with a preferred response, acknowledging his utterance, and

then Oliver changes the topic of the conversation. He does not continue to answer his

father’s question but he begins talking about their age difference. This break in

conversation was thus originated by a repeated overlap on the part of James.

Nevertheless, the conversation between Oliver and his father keeps going, only the topic

has now changed.

Sometimes, several conversations actually overlap. There is a passage in this

video in which Oliver is answering his mother’s question: see example (2). Meanwhile,

James babbles something about his food being next to his plate and the father reacts to the

latter’s utterance, looks at what James is doing, responds to him and finally winks at him.

77

Then he goes back to the conversation happening between Oliver and his mother,

acknowledging what his son has just said.

Example (2). ������(*.�� CG�7EF�57�CG7�HAGE�3H7L�83;F�3H75��:D;EFK���� �������#$��� ���A@�0��1� �������#$��� 43�:�;�A@�0�1�A@�0�1��A@�3��0�1��A@�3�83;F�CG7>CG7�5:AE7��0���1�A@�3� ����� � �?;E�>7E�?3;@E�63@E�0�1�63@E�0�1�63@E�0�1�63@E�>7�E��8>7GD�E��CG;�7EF�4>3@5��� �������#$��� ���5T�FR�0��1��������35F�� D793D67�>3�F34>7�N�5TFR�67�EA@�3EE;7FF7�� �����!�.�� ���??:�A����������35F�� E7�FAGD@7�H7DE��#$����������#$��� ���N�5TFR���������!�.�� ���A:�;����������#$��� 7F�3BDQE����������35F�� !�.�83;F�G@�5>;@�6�A7;>�N��#$����������#$��� A@�3�83;F�G@�4A@:A??7�67�@7;97����������!�.�� A:�;�A=3K���� �� The fact that CHI3 starts a turn does not entail that CHI2 stops talking to his

mother, so the two interactions happen at the same time on the video. Maybe it is because

there are always two adults who are present – at least it is the case during all recordings –

that the children intend to make the most of the situation and start a turn at talk with either

one of them, regardless of any other conversation already going on. I have several other

examples in my overall data where two conversations occur in parallel for a certain

number of seconds.

In this short extract, I also noted the use of back-channelling in the conversation.

Back-channelling is used by those who are not actually active in the conversation, in

order to show that they follow and understand their interlocutor’s remarks. It includes

phrases like “yes”, “I see”, ”really” “aha”, “oh ok”, etc. I found 10 examples of back-

channelling in this extract, all of which are used by the father. In other words, the father

makes a point of honor to indicate that he is paying attention to Oliver’s talk and

following what he is saying.

Another recurrent phenomenon in this video, as already pointed out, is the

occurrence of transition relevance places (TRPs). I counted 25 TRPs obvious enough in

this conversation. In the same vein, I noted several examples of projectability. This refers

to the allocation of a turn. It can be verbally expressed (“over to you”, “what do you

think?”, “any complaints?”); or it can include a gaze, a glance, or a particular intonation. I

was able to identify 17 examples of projectability. Most of the projectabilities come from

the parents to invite a turn from the children; the rest of them come from the children to

78

request approval from the parents. The footage reveals that the most common means to

allocate a turn is through intonation, sometimes accompanied by gaze (Table 16).

Table 16: Projectabilities Intonation 11

Intonation + Gaze 6 Total 17

Total number of turns 73 Intonation is an important feature for projectability. Indeed, intonation is very

useful to allocate a turn, in which case the prosody is rising most of the time. On line 52

for example, the mother’s ‘oui?’ is pronounced with a rising intonation because she is

trying to elicit a turn on the part of Oliver. Her ‘oui?’ is not really a question since her

boy has already answered her beforehand. My interpretation of this ‘oui?’ is that the

mother is trying to bring out another turn from CHI2, to get him to talk more and to

elaborate his answer. That being said, we should note that the transcriptions for this study

are limited regarding intonation because it is not systematically indicated. It does appear,

but only if an utterance is pronounced in a marked way, with an intonation that could be

perceived as unusual. Most rising intonations are marked with an interrogation point, as

many can be interpreted as more or less interrogative utterances.

When trying to find some adjacency pairs, four were clearly identified, all of

which correspond to the pattern ‘question/answer’ (Levinson, 1983). Two pairs refer to

the father asking questions about the potential baking of pancakes during the day. Two

children respond, first positively on the part of Oliver but negatively on the part of

Thomas; then both children respond in a negative way further in the conversation. The

other two adjacency pairs refer to the mother inquiring about the soup being good, to

which the children respond in a preferred (or unmarked) way, by agreeing.

Nevertheless, some other pairs could be classified as ‘phatic communions’

(Levinson, 1983); that is to say they are pairs that are rather meaningless but help create a

tie between the co-participants. In this footage, there are three occurrences of pairs that

correspond to this pattern, mostly because one participant repeats what their interlocutor

has just stated in a previous turn. For example ‘oui?/oui’ or ‘soupe/oui c’est la soupe’.

Surprisingly, although the footage features a conversation between two parents

and their children, only a few repairs were counted. A repair is a device for the correction

of misunderstandings, mishearings or non-hearings. It can either be initiated or performed

79

by the speakers themselves or by another participant. Table (17) classifies the repairs that

I was able to identify.

Table 17: Repairs Self-initiated/Self-repair 0

Other-initiated/Self-repair 2 Other-initiated/Other-repair 3

Total 5 Total number of turns 73

Total number of “errors” 8 The low number of repairs in this extract can be quite surprising, considering that

this is a conversation between children and their parents. No self-initiated/self-repair was

identified but several other-initiated repairs appeared in the extract. One example of an

other-initiated/other repair is illustrated in example (3):

Example (3). �������#$��� 0�7@91��I:7@�$��3�?�9A@@3�47�E7H7@FK��0��1��� �������#$��� ���3:3�A�EAGB7�� ������!�.�� I:7@�KAG��3�D7�E7H7@FK��� �������#$��� ���0�7@91�K73:�� �����!�.�� K73:���� �����#$��� 0�7@91�I:7@�$��3�?�9A@@3�47�E7H7@FK�3@6�3�:3>8���� In this extract, the father repeats his son’s utterance while changing the verb to the

present tense, which is the correct grammatical rule for a future time clause in English

after “when”. The father thus does the repair himself. However, Oliver does not seem to

realize that his father has just corrected him and he goes on using the future tense again.

Moreover, we can notice that CHI2 repeats his utterance but with a correction, passing

from ‘when I (am) gonna be seventy’ to ‘when I (a)m gonna be seventy and a half’. This

correction may reveal that Oliver actually has detected a repair on the part of his father,

which is the reason why he attempts to change his message. He simply did not pick up on

the grammatical nature of the repair.

Another repair worth mentioning is the one illustrated in example (4):

Example (4). ������(*.�� P3�E�3BB7>>7�67�>3�83D;@7����������#$ �� ���5�7EF�>3�83D;@7��������!�.�� 0�8D31�5�7EF�83D;@7�5�7���@���7EF�B3E�8>7GD����������#$ �� ���83D;@7���This example also refers to an other-initiated/other repair; but its peculiarity is that it is

performed by several speakers at the same time. While the mother first does the repair,

Thomas (CHI1) corrects his brother too at the exact same moment – appearing as an

80

overlap on the transcription. Then the father corroborates the repair immediately after,

again with an overlap from Thomas. This is an interesting moment in the conversation.

Almost all speakers participate in the correction of Oliver’s utterance. This could almost

be phrased as a “collective repair”. I found one other example of such a phenomenon in

my data, where the mother and father make the exact same repair at the same time:

Example (4bis). ��������#$ �� >7�?3�DEBR3>�5�0���>;F1���������9BJ�� D793D67�(*.���� ��!�.�� 0� 8D31�?3DEGB;3>����������(*.�� ���?3DEGB;3>� ���������#$ �� ?3DEGB;3>�� In this short extract, Thomas is reading from a book in French about mammals in

Australia. When he mispronounces “marsupial”, both parents correct him at the exact

same moment. Thomas then repeats the word to show that he has assimilated his parents’

correction. This example may also be considered as a “collective repair” since it is a

repair made by more than one person at one same moment. We note that in both cases, in

order to be understood by the child, the parents correct him using the same language,

French. The father switches from his native language to communicate efficiently with his

children.

Furthermore, going back to the extract under study, the former repair actually

follows an interesting claim on the part of Oliver. Indeed, earlier in the conversation, the

boy tells the story about how they made a snowman with ‘the snow’ during the day. This

‘snow’ is then referred to in French as ‘les fleurs blanc(hes)’. But later in the

conversation, Thomas actually informs his parents that they used flour. This is an

intriguing declaration because it entails that Oliver may actually be thinking of ‘flour’ –

whose phonetic pronunciation sounds exactly like ‘flower’ – and so he eventually says

‘fleurs’. Instead of saying ‘farine’, he says ‘fleurs blanc(hes)’, indicating that he is

probably thinking of ‘flour/flower’ when speaking in French. This example leads me to

think that Oliver might be thinking in English, even when speaking in French to his

mother, moreover about a topic that has not been previously evocated in English. It is an

interesting window onto Oliver’s mindset and a clue about his mode of thinking. If we

relate it to the different types of bilingualism, it thus seems that Oliver’s bilingualism

might be subordinate. Indeed, as detailed in the first chapter of this dissertation,

81

subordinate bilingualism refers to a hierarchy in the child’s languages, in that the child

interprets his second language through his first. This example could thus corroborate the

idea that Oliver’s bilingualism might be subordinate. Once everyone realizes Oliver’s

mistake, they all collectively produce an other-initiated/other-repair. We note that Oliver

does not respond to their repair but stays silent.

We can also find a few examples of pre-announcements originating from the

children. It can be the simple use of a vocative like “daddy” or “maman”, or a more

elaborate phrase. Throughout this footage, six utterances could be considered as pre-

announcements on the part of the children when addressing their parents. One may also

note that, in the case of the pre-announcements emanating from the children, intonation

usually rises in order to disclose that the speaker is about to say something else. One

interesting element in particular is the way James uses pre-announcements. Indeed, it is

common for him to call out to his parents during family conversations in order to get their

attention. One interesting thing to notice is the clear language differentiation that is done

from one parent to another. James seems to have understood the functioning of the

family’s conversations as bilingual. When calling on this father, his vocative is in English

while he uses a French vocative to get his mother’s attention. And both parents answer in

their native language, comforting James in the idea that he addressed them in the correct

language. Yet, we can notice that these vocatives are frequently not followed by any

actual announcements. James calls his father ‘daddy’, to which his dad responds by a

‘yes?’; then he turns to his mother and calls ‘maman’, to which the mother replies ‘oui

mon coeur?’, but the child utters nothing more. This is interesting because the fact that

the parents respond to CHI3 with a rising intonation suggests that they are expecting more

and, in the same way, they are showing their child that he has their attention. Yet, this

attention seems to be quite enough for James, since he simply smiles and gets back to his

soup without any further statement.

To sum up, this section proposed an overview of how a family conversation

functions. It has been shown that the interactions between participants are not purely a

simple succession of turn-takings. Human communication takes place through various

mechanisms and elaborate techniques. There is no “correct” or “ideal” frame of

conversation. Some utterances overlap; some are to be repeated, corrected or

reformulated. Some turns can be misheard or misunderstood, and can lead to a repair on

82

the part of either the speaker or another participant. It is up to the members of a family to

construct and articulate their exchanges. Furthermore, one does not speak in a uniform

manner that never changes. Conversational settings are continually reconstructed and can

greatly differ from one person to another. Indeed, even though there are a number of

common parameters (see Grice, 1975; Levinson, 1983; Mey, 1993), there are still many

differences related to culture, family and even the individuals. This may be even more

accurate in a bilingual conversation, where all the participants constantly reevaluate the

appropriate language depending on various factors. The following sections offer a closer

look at the participants’ alternate use of two languages.

2. Switching languages to blend into conversation The purpose of my analysis is to show that juggling two languages can prove to be

a very helpful tool in conversation. The idea is to study how and why the participants

switch from one language to another between turns. It can be for a variety of purposes in

conversation. It can also be linked to the interlocutor, the topic that is being addressed, or

to the emotional state of the speaker. This section is based on numerous examples from

the collected data, which suggest that changing languages can be a way for the

participants to join a conversation. We can notice that this strategy is not just adopted by

the children to participate in an interaction, but it is also used by the parents to make

themselves heard and understood. In an attempt to reveal standardized – if not systematic

– switches in languages, all examples were classified beforehand. The recap charts in

Appendix 4 illustrate the language alternations that led to the qualitative analysis in this

dissertation.

2. 1. How children adapt and jump in conversation

Several extracts from the videos feature a change of language on the part of the

children in order to adapt to an already ongoing conversation or to their interlocutor’s

immediate speech. For instance, let us consider the following example:

Example (5) – date: 12/03. ��������#$ �� <�7���@���7@�H7GJ�B�>�GE����������(*.�� 3:�;�.A?�FG�8;@;E�FA@�6;@7D�E;@A@�F�G���@��3E�B3E�6�7��B7F;F7�5DA;J�>N����������#$��� ���?AG;>>7�5�0�1�?AG;>>7�5�0�1�?AG;>>7�5�0�1�?AG;>>7�5�0�1�?AG;>>7�5����������#$ �� 3DDSF7�FA;� ����������#$��� ���0�7@91�:G?�;���� �����#$��� 0�7@91��:D;EFK���

83

�������366�� E�36D7EE7�N���,��������(*.�� ���3:�;�43:�;�:7�A�����������,�� K7E����������(*.�� ���5�7EF�5A??7�P3�:7;@�;�.A?���������#$��� 0�7@91�:G?�;����������(*.�� ����<7��F7�BDA?7FE����������#$��� 0�7@91�I:K�KAG�3D7�3>>�I:;F7�����������JB@F�� BA;@F7�>7�H;E397�67���,����������,�� I:K�3?�$�3>>�I:;F7������ �����#$��� 0�7@91�K7E������������,�� 4753GE7�?K�83?;>K�5A?7E�8DA?�"7D?3@K����������(*.�� �����0���D;F1���������#$ �� 0�7@91�"7D?3@K�B8�A����������9BJ�� :3GEE7�>7E�RB3G>7E���������#$ �� 0�7@91�@A�KAGD�83?;>K�6A@�F�5A?7�8DA?�"7D?3@K���

In this extract, Thomas (CHI1) and his mother are conversing in French. He is telling his

mother that he does not want any more fish. The latter insists that he eats up everything

on his plate, otherwise he will not be rewarded with a little cross next to his name on the

board where they keep score of their good behavior. Then, Oliver (CHI2) addresses

Christy, the caretaker, to ask her why her face is white. He does so in English, and the au-

pair girl responds in English as well. At that moment, Thomas jumps in the conversation

to express his surprise and refute Christy’s statement. He switches to English to integrate

the ongoing conversation between his little brother and their caretaker.

During the same dinner, another interesting switch of languages occurs on the part

of Thomas. Similarly to the previous example, Oliver and Christy are conversing in

English. Oliver asked Christy to touch his forehead to determine whether he has a fever.

The latter responds that he is indeed a little warm but that she has cool hands herself.

Listening to them, Thomas chooses to address his mother in English as well, asking her if

she also has cool hands. The data shows that it is rather unusual for Thomas to

spontaneously address his mother in English. This choice may result from the fact that the

contextual conversation at the table was happening in English:

Example (6) – date: 12/03. ��������#$��� 0�7@91�$��3�?�E;5=�����������#$��� 0�7@91�4753GE7�>AA=����� �����#$��� 0�7@91�FAG5:�:7D7����������9BJ�� ?7F�?3;@�93G5:7�EGD�EA@�8DA@F����������,�� ??:�A����������9BJ�� FAG5:7�8DA@F�67��#$�����������,�� $F�6A7E�877>�I3D?�4GF�$�:3H7�H7DK�5A>6�:3@6E�EA�;F�BDA434>K�6A7E@�F�:7>B����������#$��� 0�7@91�@A�����������#$��� 0�7@91�@A�4753GE7�����������#$ �� ���0�7@91�6A�KAG�:3H7�5AA>�:3@6E����������366�� E�36D7EE7�N�(*.�

84

��������#$��� 0�7@91�F:3F�E�I:K����� ����(*.�� ���0�7@91�@A��������366�� E�36D7EE7�N��#$ ���������#$��� 0�7@91�F:3F�E�I:K�����������#$��� 0�7@91�$�0�1��$�53F5:�GB��0�1�$�53F5:�GB�FIA�����������#$ �� ���5�7EF�0 � 1 ���������#$ �� 5 �7EF�4DV>3@F� � ���������#$ �� AG�5�7EF�B3E�4DV>3@F�����������#$��� ���0�7@91�EA���������(*.�� 0�7@91�@A�;F�E�A=3K���So Thomas probably chooses to address his mother in English because it is the language

that Oliver and Christy are speaking at the time. Besides, Thomas's question is on the

same topic, which could also reinforce the need to use the same language as his brother.

However, although the mother answers him in English, Thomas switches again to

continue his questions in French, asking “c’est brûlant ou c’est pas brûlant ?”. He might

have realized that he had not addressed his mother in her supposedly preferred language.

As a consequence, Thomas keeps the conversation going in French. The mother, on the

other hand, still responds in English, assuring “no it’s okay”. This is interesting because it

suggests that – in this case – the language that is used prevails over the language of the

interlocutor. Indeed, although the mother is spoken to in French, she continues to answer

her son in English, which is predominant at the table since Oliver and Christy are still

conversing. Eventually, we note that the mother switches back to French as Thomas

keeps talking to her (6bis):

Example (6bis) – date: 12/03. � ������#$ �� E�;��F�G��3H3;E�6;F��CG7��5��R�F3;F�4DV>3�@F���� ������#$��� ���0�7@91�$�53F5:�GB�FIA��� �����(*.�� 5�RF3;F �BAGD�D ;D7�� Furthermore, it is interesting to note that, after this parenthesis in French with Thomas,

the mother returns to the conversation with Oliver and switches back to English. Once

again, she adapts her speech to her child’s in order to communicate with him. And

Thomas does the exact same thing. After finishing his aside conversation in French with

the mother, he also switches back to English to be a part of Christy and Oliver’s

conversation again. Indeed, a little further down in the transcription:

Example (6ter) – date: 12/03. ���������,�� EA�I:7@�KAG�I7D7�3�434K�KAG�9D7I�GB�FAA�CG;5=>K�����������#$��� 0�7@91�K7E������������,�� 3@6�@AI�KAG�:3H7�3�87H7D�����������#$��� 0�7@91�K7E������������,�� :3�@�;�A=3K���

85

���������,�� ��0���D;F1���������(*.�� �����0���D;F1��������(*.�� 0�7@91�3> > �D ;9:F� ���� ����(*.�� 0�7@91�3@6�I:�K��0 � 1 �I:K�6;6�KAG�9DAI�GB�EA�CG;5=>K������������#$��� 0�7@91�:G?�;����������(*.�� 0�7@91� ;E� ; F �4753GE7�KAG�0� 1 �KAG�3>I3KE����������(*.�� 0�7@91�3F7�3> > �KAGD�?73F������������#$��� 0�7@91�K7E����������(*.�� 0�7@91�I:7@�KAG�I7D7�3�434K�����������#$ �� ���0�7@91�@A���������#$ �� 0�7@91�@A�����������#$��� ���0�7@91�K7E���������#$ �� 0�7@91��:7�3F7�3> > �:;E��0� 1 �:7�3F7�3> > �:;E�BAA����� ����(*.�� 0�7@91�A�:�.A�?�����������#$��� ��0���D;F1����������#$ �� �����0���D;F1� Although the mother and Thomas were both just speaking French to each other, the two

of them feel the need to switch back to English to continue participating in Oliver’s

conversation. Indeed, they have to use the appropriate language when addressing the

other family members if they wish to be considered as co-participants in the interaction.

During the other dinner involving Christy – on December 10th – there are a few

more examples of how the children blend into conversation with the help of their two

languages. In the following example (7), Oliver jumps in a conversation with his brother

Thomas in English:

Example (7) – date: 12/10. �����(*.�� BAGDCGA;���� ������#$ �� 0�7@91�4753GE7��� �����(*.�� 43:�;�BAGDCGA;���� ������#$ �� 0�7@91�4753G�E7�0���5:G5:AF71�� �����9BJ�� ?7F�E7E�?3;@E�7@�DA@6�3GFAGD�67�E7E�K7GJ� ������E;F�� ��,�6A@@7�G@�K3AGDF�N�5:35G@� � �����#$��� 0�7@91�@A��$�I3@F�F:7��0�1 ��$�I3@F�F:7�4>G7�EBAA@��0�1 � �� ������35F�� R5:3@97�67�K3AGDF�3H75��#$ � �������#$��� 0�7@91�$ �I3@F�F:7�4>G7�EBAA@��� �������#$��� �����0���5D;71� �������#$��� 0�7@91�I7�5:3@97���We can see in this example that, although the mother addresses her first boy, Thomas, in

French asking him “pourquoi ?” twice, the latter timidly responds in English with

“because”. This is interesting because Thomas does not adapt his language to his

mother’s but he continues to answer her in English. Following this, Oliver jumps in to

claim that he wishes to have a spoon of a different color to eat his yoghurt. He says so in

English to his brother and declares “we change”. One interpretation could be that Oliver

addresses Thomas in English because the latter was just talking in this language. In this

86

case, Oliver adapts to Thomas’s momentary language in order to easily communicate

with him. Another possible interpretation could be related to the fact that it is the

caretaker who just gave the yoghurts to the children. Therefore, the reason why Oliver

expresses his dissatisfaction in English might be because the person who gave him the

spoon of the color he does not want is Anglophone. In this case, we consider that Oliver

chooses to speak English in reference to the situation that was caused by the caretaker.

In addition, some excerpts suggest that if the children do not adapt their language

to the ongoing conversation, their utterance might not be taken into consideration. The

following two examples, (8) and (8bis), illustrate this claim. While Thomas, Christy and

the mother are interacting in English, Oliver tries to catch his mother’s attention but he

does not obtain any answer on her part:

Example (8) – date: 12/10. ��������#$ �� 0�7@91�;F��;�E�EA?7F:;@9�7>E7�����������#$��� �����0���5D;71����������,�� 3�?AGF3;@�>;A@������������#$��� ���?3?3@�FG�0� 1 � FG�B7GJ�?7FFD7����������,�� F:3F�>AA=E�D73>>K�3�53F�����������#$ �� 0�7@91�K73:����� ������,�� $�E3;6�3�53F���

Example (8bis) – date: 12/10. � �����(*.�� �������0�7@91�3�?A@=7K��� ������#$ �� �������� �����9BJ�� ����E75AG7�>3�FSF7�� ������#$���� � � ����?3?3@������������,�� ����3@�3@F7>AB7�0�1����� �����#$���� � � ����FG�E3;E�����������#$ ���������0�7@91�I:3F������������,�� ����3@F7>AB7����

In these examples, Thomas is mimicking an animal and he is trying to make his

mother and Christy guess which one he is imitating. Oliver calls on to his mother in

French but she gives him no response. This might be because Oliver does not use the

same language as the other co-participants. Therefore, his utterances are not taken into

account within the exchange that is going on between the other family members. Oliver is

neither able to communicate with his mother nor to interrupt her interaction with Thomas

and Christy. This illustration corroborates the idea that a switch in languages is needed in

order to take part in a current conversation. Otherwise, the speaker might remain outside

the interaction, which is the case for Oliver in this example.

When looking at the other videos, involving the father, the same pattern can be

observed. The children feel the need to adapt their speech to the conversation happening

around the table in order to communicate with their parents. Example (9) is taken from

the video on November 30th and shows how Thomas has to switch to English to follow

his family’s lead and take his part in the interaction:

87

Example (9) – date: 11/30. �������(*.�� 0�7@91�?3K47�I7�E:AG>6�5A@FDA>�3�>;FF>7�4;F�?AD7�����������#$ �� �����0���D;F1��������!�.�� K73:����������(*.�� 0�7@91�I:3F�KAG��3�D7�I3F5:;@9�A@�FH�0���D;F1����������#$��� ���0�7@91�6366K����� ������#$��� 0�7@91�$�9AF�@A�=7F5:G�B����� ����!�.�� KAG�9AF�@A�=7F5:GB����� �����!�.�� F:3F��;�E�4753GE7�KAG�:3H7@�F�8;@;E:76�KAGD�5:;5=7@���� ������#$��� 0�7@91�@A������ ������#$��� 0�7@91�=7F5:GB�;E�0�1�;E�0��1��;E�8AD��0�1�;E�8AD�0�1�73F;@9�I;F:�F:7�5:;5=7@�� ��� ������#$ �� ����5�7EF��8DA;�6��� �����(*.�� ���0�7@91�KAG�I3@F�?AD7�5:;5=7@���� �����!�.�� K73:�$�=@AI���� ������#$ �� ���0�7@91� <GEF�3�4;F� ��Y���������(*.�� 0�7@91�KAG�I3@F�F:;E�A@7�:7D7�����������366�� E�36D7EE7�N��#$ ��������(*.�� 0�7@91�$�F:;@=�;F��;�E�3����������(*.�� 0�7@91�3�I;@9����� �����#$ �� 0�7@91�A=3K�� When looking at this extract, we see that the mother, Oliver and the father are all three

speaking in English. Thomas intervenes in French to tell his mother that his chicken is

cold. The mother responds to Thomas but she does so in English, asking if he wants more

chicken. As a consequence, Thomas readjusts his language choice and answers his mother

in English with “just a bit”. We can notice that they then both continue in the same

language while the mother serves him more chicken. Another thing that is striking when

looking at this example, is that Oliver uses English from the start of his intervention,

contrarily to his brother. He understands that it is the appropriate language to speak at that

point in the conversation, based on the previous utterances of his mother. Another

element that may be playing a role in Oliver’s choice of English at that moment, is the

fact that he addresses his father. Indeed, when talking to his father, Oliver’s vocatives are

mostly in English whereas his vocatives to his mother are mainly in French. In other

words, Oliver tends to start his utterances to his father with “daddy”, which then triggers

a turn in English; while he begins with “maman” when talking to his mother, which

triggers a turn in French.

Interestingly, a little later during the same lunch, a similar situation occurs but this

time Thomas switches to English right away, even though he addresses his mother:

Example (10) – date: 11/30. �������(*.�� A=�;>�83GF�3FF7@6D7�G@�B�7�F;F�B7G�*>;H7D����������!�.�� KAG�:3H7�FA�I3;F�3�>;FF>7�4;F�?3F7�E:AG>6�$�5GF�:;E�5:;BE�3E�I7>>�������� � �47�53GE7�F:3F�?3=7E�F:7?�>7EE�:AF���

88

�� ����(*.�� 0�7@91�K7E����������!�.�� �����������35F�� 5AGB7�>7E�8D;F7E�67��#$����������#$ �� 0�7@91�KAG�FAA�53@�KAG�5GF�?K�5:;BE�����������366�� E�36D7EE7�N�(*.��������(*.�� �����������35F�� 5AGB7�>7E�8D;F7E�67��#$ ���������#$ �� 0�7@91�A:�;�F:;E�A@7�FAA�� ��������JB@F�� BA;@F7�EA@�3EE;7FF7��������(*.�� 0�7@91�I:3F�6A�KAG�E3K�@AI����� �����#$ �� 0�7@91�F:3@=�KAG���The mother tells Oliver in French to wait a little bit before eating his chips because they

are hot. The father feels the need to tell him the same thing but in English. Then he asks

the mother whether he should cut Oliver’s chips. The latter answers her husband in

English, and when Thomas asks his mother to cut his chips as well, he uses English too.

Contrarily to the precedent example, Thomas does not try to address his mother in French

first. He has understood that English is predominant at that moment and that in order to

get his mother’s attention, he has to speak to her in English as well. We note that the

mother complies right away and cuts Thomas’s chips without a word. Then they continue

in English when the mother reminds her son to say thank you.

Most of the time, the children may change languages to match their interlocutor’s.

But if the latter suddenly switches and no longer uses the language in which the

conversation was happening, the child might show that he does not wish to follow their

co-speaker’s lead. As in example (11):

Example (11) – date: 11/30. ��������#$ �� 0�7@91�F:7�D344;F�:36�3�>A�F�A8�<A4E�FA�6A����������!�.�� A=3K����� ����(*.�� 0�7@91�I:3F�6;6�:7�:3H7�FA�6A�����������!�.�� $�F:;@=�F:7�D344;F��;�E�3�9;D>�@A������������#$ �� ���0�7@91�:7�@77676�FA���������#$ �� 0�7@91�9A�;@�F:7�EGB7D?3D=7F�����������#$ �� 0�7@91�:7�@77676�FA�9A�A@�F:7�4GE�:7�@77676�G:�;�����������#$ �� 0�7@91�I:3F�7>E7������������#$ �� 0�7@91�:7�@77676�FA�9A�G:�;�;@�F:7�?GE7G?����������9BJ�� E75AG7�>3�FSF7��������(*.�� BAGD�HA;D�CGA;�3G�?GER7������������#$ �� 0�7@91�@A����� �����#$ �� 0�7@91�9AF�FA��F:7��?GE7G?�4GF�0 � 1 �4GF�����������#$ �� 0�7@91�3@6�:7�0�1�:7�0�1�:7�@77676�FA�6A�F:7�F;5=7FE����������!�.�� A:�;�A=3K�����������#$��� ���??:�A���������(*.�� 0�7@91�A�:�;�A=3K�EA�:7�I3E@�F�9A;@9�FA�F:7�?GE7G?�FA�>AA=�3F�3@�7J:;4;F ;A@���������(*.�� 0�7@91�:7�I3E�9A;@9�F:7D7�FA�IAD=�����������#$ �� 0�7@91�G:�;��

89

What happens in this example is that in the middle of Thomas’s story, the mother asks

him a question in French: “pour voir quoi au musée ?”. The child’s reaction simply is to

answer “no” as to notify his mother that she is not using the right linguistic code and that

he will not switch. Therefore, he continues his story in English. The mother is thus forced

to go back to English if she wants to exchange information with her son. She does so at

the end of the extract (965-966); and Thomas shows that he takes her utterance into

consideration by uttering a “uh” as a sign of agreement. Again, this shows the importance

of using the right code at the right time when taking part in a bilingual conversation.

Otherwise, one might not obtain their interlocutor’s attention. The children – just like the

parents – need to adapt to each other’s language use if they want to get through to them.

This point is clearly illustrated in the next example (12):

Example (12) – date: 11/30. ��������#$ �� 7> >7�7J;EF7�7@5AD7������ ���(*.�� 0�7@91�@A��I:7D7��0�1�G:�;�I:7D7�I3E�;F�;@��AD6A9@7������������#$ �� ���'G5K����������(*.�� 0�7@91�6;6@�F�I7�E77�'G5K�����������(*.�� 0�7@91�EA?7I:7D7�D757@F>K�����������(*.�� �����������(*.�� 0�7@91�$�6A@�F�=@AI����������9BJ�� E75AG7�>3�FSF7��������!�.�� �����������(*.�� 0�7@91�F:7�JJJ�A8�EA?7F:;@9����� ����!�.�� ������ ����#$ �� 0�7@91�5A?7�A@������ ����9BJ�� F3B7�G@7�8A;E�EGD�EA@�>;HD7�3H75�?3;@�93G5:7 As we can see in example (12), Thomas is asking questions in French about Lucy – the

skeleton of the first woman ever found – to his mother. The mother is no longer sure

whether they saw that skeleton in a museum, therefore she asks the father if he

remembers, all the while sort of thinking out loud, all this in English. Thomas

understands that he will not get his mother’s attention back by speaking French so he

switches to English and utters a clear “come on” while hitting his book with his left hand.

This shows impatient he is to keep reading with his mother. We know that the book is in

French so the fact that Thomas switches to English is not related to the language of the

book. It clearly is a strategy on the part of Thomas to get his mother to hear him.

The same pattern occurs with the second boy, Oliver. Indeed, the latter also

understands the necessity to join a conversation in the appropriate language in order to be

taken into consideration by the other participants. In the following example (13), his older

90

brother Thomas and their mother are conversing in French. At the exact same moment –

resulting in an overlapping conversation on the other side of the table – the father is

exchanging in English with James, the youngest boy, who has gone behind the camera.

They are talking about who is present on the screen when Oliver jumps into the

conversation between his father and James.

Example (13) – date: 01/31. �������!�.�� I:3�F�0�1�I:�3F��0��1�I:3F�3D7�KAG�>AA=;@9�3F������������#$��� 0�7@91�3F�KAG������ �����#$ �� 7F�CG3@6�?A;�<7�H3;E�3HA;D�6;J�3@E�����������!�.�� ���3F�I:A�����������#$��� ���0�7@91�3@6�.:A?3E���������!�.�� ;E�.:A?3E�F:7D7������������#$��� 0�7@91�3@6�?A�?�����������!�.�� ;E�?A?�F:7D7�3E�I7>>������������#$��� 0�7@91�3@6�6366K����������(*.�� ���<�3GD3;�CG3D3@F7�E7BF�3@E���������!�.�� 3@6�6366K������������#$ �� CG3D3@F7�E7BF����� ����!�.�� �����0���D;F1���������#$��� 0�7@91�3�@6��F:3F�� ;�E�?7��0�1 � ����������#$ �� CG3@6�?A;�<7�H3;E�3HA;D�6;J�3@E�����������#$��� ���0�7@91�F:3F�� ;�E�?7���������JB@F�� BA;@F7�>3�53?RD3�7@�>3�D793D63@F��The fact that Oliver intervenes in English demonstrates his will to participate in his

father’s conversation with his little brother. Had he wanted to join the conversation

between his mother and Thomas, he might have chosen to speak in French and not in

English. This proves all the more the importance of choosing the right language when

wanting to integrate an ongoing interaction. Language choice thus plays an essential role

in the process of asserting oneself in conversation. The same phenomenon occurs again

later during the meal while the father jokes about cooking spiders for dinner. This time,

Oliver and also his mother, both jump into the conversation in English:

Example (14) – date: 01/31. �� ����!�.�� :AI�34AGF�EA?7�5D;EBK�EB;67DE������������#$��� 0�7@91�3:�;�@A����������!�.�� F3=7�EB;67DE����������9BJ�� 67GJ�?3;@E�7@�43FF7?7@FE�3>F7D@RE��������!�.�� 43@9��F:�7?�A@�F:7�:736����������9BJ�� 97EF7�43FF7?7@F�BA;@9�6DA;F�87D?R��������!�.�� 6DK�F:7?�AGF����������!�.�� 3@6�8DK��F:�7?�;@�F:7�B3@����������(*.�� ���0�7@91�I;F:�EA?7�A@;A@E���������!�.�� I;F:�A@;A@E����� ����(*.�� ���:7;@�;����������366�� E�36D7EE7�N��#$ �

91

�������(*.�� 0�7@91�I;F:�A@;A@E�����������#$��� 0�7@91�6366K���������!�.�� 5D;EBK�EB;67DE�A@;A@E������������#$��� 0�7@91�@A�0���D;F1���������9BJ�� E75AG7�>3�FSF7��������!�.�� A=�����������#$��� 0�7@91�AD�:AI�34AGF�IAD?E�����������!�.�� IAD?E�K73:����� ����!�.�� IAD?E�EB39:7FF;���As previously evoked, both the mother and Oliver address the father in English in order to

participate in his speech. The mother suggests adding some onions and the father repeats

her utterance in English, thus showing agreement. Later in the conversation, Oliver makes

a suggestion of his own by proposing to cook some worms. He does so in English, which

is the language that most participants switch to when addressing the father of the family.

It has thus been shown in this section that the children must adapt their speech if

they want it to be taken into account by their co-participants. A change in languages is

sometimes needed to obtain a response on the part of their parents or simply to catch their

attention. However, if the children deliberately choose not to switch languages and to

persist in addressing an interlocutor in a specific language regardless of the current

conversational settings, they might not be heard, and therefore might not reach their

communication goal. The children may thus have to eventually switch languages in order

to adapt to an already ongoing conversation, if they want to assert their place within that

conversation.

Symmetrically, if the parents wish to be understood and communicate effectively

with their children, they also have to deal with language changes in order to make

themselves heard. The following section offers a closer look at how the parents have to

juggle two languages as well.

2. 2. How parents deal with language changes

The children are not the only ones around the table who have to adapt to their

interlocutor(s). The same is true for the parents who may also switch languages if it

allows them to join a conversation with their children more easily. This claim implies that

the parents’ linguistic use of their two languages is not rigid. They do not always

necessarily follow the one parent–one language rule. They too have to handle two

languages in their everyday family conversations. Let us start by looking at how the

mother manages English in addition to her native language to communicate with her

92

children. In the following extract, for instance, the mother no longer uses her mother

tongue but she switches to English to address Oliver, who has himself switched to answer

the caretaker:

Example (15) – date: 12/10. �� �����#$��� :G?�;�7EF�57�CG7�F�G��3E�?3>�3G�6AE�����������JB@F�� BA;@F7�(*.��������(*.�� AG;�?A@�5A7GD�����������#$��� ?3;E��:D;EFK�83;F�B3E�4;7@�>7�43;�@����������(*.�� �����0���D;F1����������,�� �����0���D;F1����������,�� I:K�6A@�F�$�6A�;F�I7>>������������#$��� 0�7@91�4753GE7�0� 1 � �����������,�� 4753GE7������������#$��� 0�7@91�47�53GE7��0�1�47�53GE7��0��1�4753GE7�0�1����� ����(*.�� 0�7@91�4753GE7�I:3F����������(*.�� 0�7@91�KAG�53@�F� <GEF�E3K�4753GE7�����������#$��� 4753G�E7����Oliver and his mother are first speaking French at the beginning of this example. Oliver

mentions Christy, who intervenes in English, and the rest of the conversation goes on in

English. Oliver is the first to switch and then his mother switches as well. She has to if

she wants to be part of the interaction and communicate her claim to Oliver that “[he]

can’t just say because”. Another example of the mother jumping into a conversation in

English with her children occurs during the big family lunch with the father:

Example (16) – date: 11/30. �������!�.�� 3@6�E:7�;E�3�:7����������!�.�� -=K������������#$ �� ���0�7@91�:7���������!�.�� -=K�;E�3�?3@����� �����#$ �� 0�7@91�K73:�����������#$��� JJJ�0���5D;71���������(*.�� 0�7@91�-=K�;E�3�?3@��$��0� 1 � F:AG9:F�$� 0 � � 1 � $ � F:AG9:F�F:7K�I7D7�3�> > �������� � 9;D >E�F:AG9:�3D7�KAG�EGD7�F:3F�-=K�;E�3�?3@������������#$ �� 0�7@91�I7>>�$�=@AI�:7��;�E�3�?3@����������9BJ�� :T5:7�>3�FSF7���������#$ �� 0�7@91�4753GE7�0�1�4753GE7�0�1�4753GE7�-=K�0�1�-=K�35FG3>>K�:7�0�1�������� � :7�6;6@�F�5:3@97�;@FA�3�83;DK�����������#$��� ���?3?3@�0���DO>71���������#$ �� ������ ����9BJ�� E75AG7�>3�FSF7��������!�.�� -=K�6A7E@�F�FD3@8AD?�;@FA�3�83;DK����������(*.�� ���0�7@91�:7�6;6@�F�:3�;��������9BJ�� E75AG7�>3�FSF7���������#$ �� 0�7@91�:7�53@�F����������(*.�� 0�7@91�:7�FD3@E8AD?E�;@FA�I:3F�F:7@��� In this extract, Thomas is talking about the cartoon that he watched during the morning

called the Winx Club. He and his father are exchanging on whether all characters are girls

93

or if there are male fairies too. The mother intervenes in their conversation in English.

She chooses to use the language in which the discussion is already engaged. It is a

strategy on the part of the mother in order to exchange efficiently with her son and her

husband. We can also notice that at one point in the conversation, the youngest, James,

calls on to his mother by grumbling “maman”, but his appeal remains unanswered. This

illustrates the necessity to match languages when wanting to blend into a conversation.

Most times, the mother may join a conversation in English because the other

participants will simply keep talking in this language, regardless of her attempts to

intervene in French. In the following example (17), the mother begins asking a question

to Thomas but the latter starts a turn in English and he continues his utterance without

even noticing that his mother tried to communicate with him. The mother politely

suspends her speech to let Thomas continue. Then she has no other choice but to switch

to English to take part in the conversation:

Example (17) – date: 11/30. �������#$ �� 0�7@91��3�53@B3=7�5��0�1�:G?�;�3�53@B3=7�5��� ������(*.�� ���.A?�5�7EF�CGA;��� ������!�.�� 3�B3@53=7��� �������#$ �� 0�7@91�3�B3@53=7��;E��?367�8DA?�=;I;��� ������!�.�� @A��� �� ����#$ �� ��0���D;F1�� ������!�.�� ;F��;�E�?367�8DA?�8>AGD�3@6��� �������#$ �� ���0�7@91�3@6�3�=;I;��;�E�?367�8DA?��3��53@�B3=7��0��1�B3@53=7�� ������!�.�� 8DA?�EG93D��� ������!�.�� 3�@6��� �������#$ �� 0�7@91�?;>=���� ������!�.�� 8>AGD�EG93D�3@6��� ������(*.�� 0�7@91�3@6�I3F7D�@A�AD�?;>=�AD�EA?7F:;@9���� �������#$ �� ���0�7@91�?;>=��� ������!�.�� EA?7�43=;@9�0�1�43=;@9�BAI67D���� �� ���(*.�� ���0�7@91�799E��� ������(*.�� 0�7@91�@A���� ������!�.�� ���F:7D7��;�E�EA?7�799E�;@�:7D7�� ������(*.�� 0�7@91� ; F �� ;�E� > ;=7�5DSB7�E�:7;@�;��� ������9BJ�� :3GEE7�>7E�RB3G>7E� ������(*.�� 0�7@91�4GF� ; F �� ;�E�G:�;�F:;5=7D� �� The mother tries to ask Thomas something but then interrupts herself to let her son finish

his sentence. When the father keeps the conversation going in English, the mother

understands that she has to switch too if she wants to intervene in their exchange. She

does so to convey her thoughts on how to make pancakes. While Thomas had not noticed

her intervention in French earlier, this time he identifies her utterance as being relevant in

the conversation and allows her to continue by suspending his speech in his turn. A few

94

seconds later, when the debate on the recipe of pancakes comes to an end, the mother

immediately switches back to French, to ask Thomas what we can assume to be the

question that she previously intended to ask him:

Example (17bis) – date: 11/30. ������(*.�� 0�7@91�4GF�;F��;�E�G:�;�F:;5=7D��� ������!�.�� K73:�;F��;�E�>;=7�5DSB7�E��� ������(*.�� 0�7@91�I7>>�I7�BGF�477D�;@�5DSB7E�E�I7>>�@AF�7H7DK4A6K�4GF�$�6A��� �������#$ �� �����0���FAGEE71� ������(*.�� �5�7EF�CGA;�FA@�8DG;F �BDR8RDR�.A?��0�� � 1 ���� �� ����#$ �� CGA;���� ������(*.�� FA@�8DG;F �BDR8RDR�5�7EF�CGA;����The mother switches to French to ask Thomas what his favorite fruit is. It is possible to

deduce that it was the same question she intended to ask him when she was interrupted in

English earlier. What is interesting is that Thomas makes her repeat her question. It is as

if he did not understand her right away because she switched languages again. Thomas

may have been expecting an utterance in English, therefore he might be a bit destabilized

by his mother’s sudden change to French. This could explain why he does not answer

right away and it suggests that the child might sometimes need a second to readjust to the

conversational situation.

The mother can also choose to persist in using French in the middle of an English

conversation if she wishes to impose her native language within the exchange. In example

(18), taken from the family dinner on January 31st, Oliver and his father are speaking

English, and the mother manages to intervene while avoiding switching languages:

Example (18) – date: 01/31. �������#$��� 0�7@91�KAG��3�D7�9A@@3�47�7;9F:K��� ������E;F�� !�.�E�7EEG;7�>7�@7L� ������!�.�� @A��� ������(*.�� 0�7@91�??:�A�I7��I;� > > �47�3�> ; F F >7�47�A>67D�F:3@�F:3F� >AG>AG�E�� � ������5A?�� >AG>AG�B7F;F�@A?�7@�8D3@P3;E� ������!�.�� $��I;�>>�47�>;=7�3�:G@6D76��� � �����#$ �� 0�7@91�@A���� ������9BJ�� B7@5:7�>3�FSF7� ������(*.�� 0�7@91�E7H7@FK����� �������#$ �� 0�7@91�K73:��� ������(*.�� A@�3GD3��BD7ECG7�57@F�6;J�3@E��� ������9BJ�� :T5:7�>3�FSF7� ������!�.�� 0�8D31�AG3;E��� ������(*.�� ��0���D;F1�� �������#$ �� ���57@F�6;J�3@E��� �������#$��� ��0���FAGEE71�� � ����(*.�� A@�E7D3�FDQE�0�1�FDQE�H;7GJ�� �������#$��� �?3?3@�;�>��K�3�67E�:ACG7�FE��0���1��� ������(*.�� ����

95

������9BJ�� :T5:7�>3�FSF7� �������#$��� ;�>��K�3�CG3@6�?S?7�67E�:ACG7FE�?3?3@��� ������35F�� E7�8DAFF7�>7E�K7GJ� There are several interesting switches in this example. First, we notice that the mother

starts participating in the conversation in English. Then she switches to French, claiming

“on aura presque cent dix ans”. Contrarily to what could be expected, the mother is not

excluded from the exchange but the other participants do take her utterance into account.

Besides, not only do they consider her intervention but they also switch to French

themselves. The father acknowledges her claim with “ouais”. At the same time, Thomas

exclaims “cent dix ans!”, whereas he had just expressed his disbelief in English to his

father (171:”no:”). The most pertinent language for this conversation has thus switched to

French for all the interlocutors. The mother was able to impose her language in the

interaction. Even Oliver, at the end of the example, switches to French to tell his mother

that he has the hiccups. So the mother’s intervention in French created a shift in

conversation and led all participants to follow her lead and change languages as well.

The mother is not the only one to switch from her native language while

conversing with her children. The father also abandons English from time to time to blend

into his children’s interactions in French. The following example (19) illustrates this

point:

Example (19) – date: 11/30. ������(*.�� �7@�EADF3@F�6G�4GE��0���1�7@�67E57@63@F�6G�4GE�CG�7EF�57�CG�;>E�A@F�HG�� �������#$ �� �����0���FAGEE71� �������#$ �� G@�5:3F�73G��0��1�>7�5:3F73G��67��>3�D7;@7��� ������!�.�� 0� 8D31�3:�>7�5:3F73G�67�>3�D7;@7�A=�� � �������#$��� ��0���FAGEE71�� ������!�.�� 0� 8D31�B3D57�CG7�>3�D7;@7�5�7EF�G@7�9DAG;@7�5��63@E�>7��0� 1 �63@E� ������ � >7�?A@67�67�0�1 �67�+7BB3�+;9���� �������#$ �� ������ ������9BJ�� :T5:7�>3�FSF7� �������#$ �� 7F��7>>7�3�B3E�>7�6DA;F�3G��0���1��B3E�67�5:7H3>;7D��0���1�7>>7�3��� ������(*.�� ���0�7@91�EA�4D;F;E:�0���D;F1� �������#$ �� B3E�>�7��6DA;F�@;�5:7H3>;7D�@;�BD;@57EE7�@;��� ������!�.�� 0� 8D31�A�:�;�G@�5:7H3>;7D�G@7�BD;@57EE7�� �� ������!�.�� JJJ�0���5:G5:AF71�� ������!�.�� �6A�F:7K�:3H7�3�=;@9��0� � � 1 � ;E�F:7D7�3�CG77@�;@�!D3@57����In example (19), the father switches to French in order to participate in Thomas’s telling

of the story of Peppa Pig, a cartoon that he watched on TV in the morning. It is a strategy

on the part of the father to adapt to his son’s language choice. It allows him to be active in

96

the exchange and to show Thomas that he understands the story that the latter is telling.

What is interesting is that while he is switching to French, the mother addresses him in

English to express her opinion about Thomas’s story (“so british [=! rit]”� By the end of

the extract, we can see that the father tries to switch back to English by asking Thomas

whether there is a queen in France. The exchange that follows shows that Thomas did not

pick up on his father’s attempt and persists in talking in French. Eventually, the father has

to switch again, understanding that he will not be able to bring English back in the

conversation:

Example (19bis) – date: 11/30. ������!�.�� �6A�F:7K�:3H7�3�=;@9��0� � � 1 � ;E�F:7D7�3�CG77@�;@�!D3@57���� �� ����#$ �� ���� ������9BJ�� :T5:7�>3�FSF7� ������!�.�� ;E�F:7D7���� �������#$ �� ������ ������9BJ�� :T5:7�>3�FSF7� ������!�.�� I:A��;�E�F:7�CG77@�A8�!D3@57���� �������#$ �� <�7���@7��E3;E�B3E��� ������B:A�� 5:R�B3E� ������9BJ�� E75AG7�>3�FSF7�7@�D793D63@F�(*.� ������(*.�� ; > ��@��K�3�B3E�67�D7;@7�;5;�?A@�5A7GD��� �� ���9BJ�� E75AG7�>3�FSF7� �������#$ �� ��0���D;F1�� ������(*.�� ;5;�5�7EF�G@�BDRE;67@F��� ������E;F�� �#$��CG;FF7�>3�F34>7� �������#$ �� 7F�5 �7EF�CGA;�G@�BDRE;67@F���� �������#$��� ���?3?3@��;>�7EF�AU��0�1�� ������(*.�� G@�BDRE;67@F�6�7��>3�DRBG4>;CG7�5�7EF�7G:�;��� �������#$��� ����;>�7EF�AU��0�1�;>�7EF�AU� >;E347F:��� ������!�.�� 0� 8D31�5 �7EF�5A??7�>3�D7;@7�0� 1 � ��When the father asks Thomas if there is a queen in France, his son silently nods. But then

when he is asked the name of the queen, Thomas turns to his mother and says “je (ne)

sais pas”, to which the mother responds in French as well. There can be two

interpretations to the fact that Thomas uses French at that moment. It could either be

because the entire exchange so far had taken place in French and that Thomas does not

feel the necessity to switch; or it could be because his utterance is implicitly a question to

his mother. Indeed, while Thomas declares that he does not know, he shares a look with

his mother, as to ask for her help, her opinion. His mother answers him in French and

their interaction continues. In the end, the father has to switch again and declares “c’est

comme la reine”. English has not been reestablished within the conversation so he has to

give way to his co-speakers’ wish to keep the conversation going in French. Even Oliver,

97

when he jumps into their discussion to ask his mother where Elisabeth – the observer – is,

uses French as well.

Sometimes the father adopts the French language to notify the children that he

understands what they say. This results mainly in the use of back-channelling in French.

According to Mey (1993), back-channelling is used by those who are not actually active

in the conversation. This way, they intend to show that they follow and understand their

interlocutor’s remarks, without interrupting their turn at talk. This happens quite a few

times with the father, who seems to make a point of honor to show his boys that the fact

that they express themselves in French is not an issue for him. The following example

(20) regroups some extracts that illustrate this point:

Example (20) – date: 11/30. �������!�.�� A:�;�K73:������������#$ �� 3BDQE�;>�H3�HA;D�>3�D7;@7�����������#$ �� 7F�;>�6;F�<�7��B7GJ�B3E����������!�.�� 0� 8D31�AG3;E�����������#$ �� �7F�>N��0�1��7F�>N��0�1�7F�>N�A@�3�HG�3G�0�1�3G����Y��� ������#$ �� ?A6Q>7����� �����(*.�� ??:�A���� �����9BJ�� :T5:7�>3�FSF7�� �����!�.�� ���0� 8D31�G@�3GFD7�?A6Q>7�AG;� �� ������#$ �� ��0���D7EB;D71�67�0���>;F1��� ������#$ �� ?3�??;8QD7�0���>;F1���� �����!�.�� 0� 8D31�AG;� � �� ������#$ �� 7EF�0���>;F1��� � ����#$ �� �3�BB3DG���0��1�0���>;F1��Y��� ������#$ �� B>357@F3;D7E�0���D;F1���� �����(*.�� ����� �����9BJ�� :T5:7�>3�FSF7�� �����!�.�� �����0���D;F1�� ������#$ �� �5�7EF�G@��0�1��5�7EF�G@�6;88;5;>7�?AF��0��1���� �����!�.�� 0� 8D31�AG3;E�:7;@�;����� ������#$ �� 5�7EF�G@�?AF�6;88;5;>7���� �����(*.�� AG;�5�7EF�G@�?AF�6;88;5;>7�CG�A@��@��GF;>;E7�B3E�473G5AGB���� �����!�.�� 0� 8D31�@A@���In the first part of the example, Thomas is narrating what happened in the Peppa Pig

cartoon that he watched on TV in the morning. We can see that the father uses French to

show his agreement with his son’s speech. By saying “ouais”, he shows Thomas that he

understands the proceeding of his story and encourages him to keep going. In the second

part of the example, Thomas is reading from a book in French about the mammals in

Australia. Even though the participants are engaged in a reading activity and not a

spontaneous conversation, the father still indicates his approbation by acquiescing: “oui”,

98

or by repeating: “un autre modèle oui”. The fact that he keeps speaking French allows

him to create a tie with Thomas and might even give him legitimacy to correct or approve

of Thomas’s reading. In the third part of the example, the co-speakers are back in a

conversation, as Thomas expresses his surprise about the word he just read:

“placentaire”. Again, the father supports his claim with back-channelling markers like

“oui hein@i ?”. One other thing that is interesting to notice is that Thomas does a self-

initiated/self repair in the last short extract. He begins to say “c’est un difficile mot”, but

immediately corrects himself by reformulating: “c’est un mot difficile”. This is striking

because it suggests that Thomas is actually using the English paradigm, which consists in

placing adjectives before substantives. But rapidly he realizes his mistake and makes a

repair on his own. The mother confirms his repair by repeating the same phrase,

expanding it with an answer about the rarity of the word. The father, on the other hand,

shows agreement with Thomas just before the latter corrects himself. He may not have

detected his son’s mistake right away. As an Anglophone himself, it is possible to assume

that the structure of Thomas’s utterance did not immediately appear to him as erroneous.

We can also notice that the father may switch to French when supervising the

children. Indeed, if he wishes to call out to his boys while they are engaged in a

conversation in French, the father will most likely use the same language:

Example (21) – date: 11/30. ������(*.�� H;7@E�F�3EEA;D��� �� ���JB@F�� BA;@F7�>3�5:3;E7�67��#$�� �������#$��� @A�@�0�1�@A@�?3;E��� ������JB@F�� BA;@F7�>�7JFRD;7GD�67�>3�B;Q57�AU�;>�E7�FDAGH7� ������(*.�� ��0���D;F1�� ������!�.�� 0� 8D31�:7K�;�CG�7EF�57�CG7�FG���� ������366�� E�36D7EE7�N��#$�� Example (21bis) – date: 01/31. �������(*.�� 47@�;�FG��@7��H7GJ�B3E�G@�K3AGDF������������#$ �� �����������9BJ�� E75AG7�>3�FSF7��� ����(*.�� ?3@97�G@�B�7�F;F�B7G�B>GE�67�BA??7����������(*.�� 7G:�;�.A�?��0��1�*>;����������E;F�� �#$��FAGD@7�FAG<AGDE�>7�6AE��������!�.�� 0� 8D31�CG�7EF�57�CG7�FG�83;E�*>� ;H7D������������E;F�� �#$��E7�D7FAGD@7�7F�?3@97�EA@�4AGF�67�B3;@� When calling the children to order, the father may thus adapt to the conversational

situation and use French. The goal may be to catch the boys’ attention rapidly and

99

efficiently. Switching languages to use the current one in conversation might then be a

strategy on the part of the father to make himself heard.

So the parents also need to switch from time to time if they wish to be understood

and communicate effectively with their children. Ultimately, both parents and children

may switch languages at one point in order to adapt to an already ongoing conversation, if

they want to assert their place within it.

3. Switching languages to create a break in conversation

Blending into a conversation is not the only reason why the children may switch.

Sometimes, the goal behind a shift in language is to voluntarily create a break in

conversation. By using the other language, and not the one that is dominant at the time in

the interaction, the participants are able to mark their disagreement, request attention, or

simply cause the conversation to move on.

The following example (22) shows how Thomas deliberately intends to break off Oliver’s

speech. The latter is telling Christy that he was born brown and that he turned white later,

when he grew up. Thomas interrupts him in French and shouts that it is nonsense:

Example (22) – date: 12/03. ������(*.�� 0�7@91�KAG�I7D7�4DAI@�>;=7�)38;���� �������#$��� 0�7@91�K7E��� ������(*.�� 0�7@91�IAG3�A�F:3F��;�E�H7DK�4DAI@��� �������#$��� IAG3AG�A�BA?�A�BA?�A���� ������(*.�� 0�7@91�3@6��KAG�FGD@76��0�1��KAG�FGD@76��0�1�KAG�FGD@76�4>A@6�<GEF�>;=7�F:3F��� �������#$��� 0�7@91�K7E��� �� ����#$ �� ; > �6;F �@ � ;?BADF7�CGA;�0���5D;71��� �������#$��� 0�7@91�K7E�� ��������,�� <GEF�A@7�63K�KAG�IA=7�GB�3@6�KAG�I7D7�3>>�I:;F7���� ������9BJ�� B3G?7�>3FRD3>7�67�:3GF�7@�43E� �������#$ �� �; > �6;F��0� 1 �:G?�;�; > �6;F �@ � ;?�BADF7��CGA;�0���5D ;71��� ������(*.�� ���EE:�A�� ������(*.�� A=�;>�3�>7�6DA;F�6�3HA;D�6�7��>�;?39;@3F;A�@�.A�?��� While Oliver, the mother and the caretaker are all three exchanging in English, Thomas

voluntarily creates a break by choosing to intervene in French. Moreover, he shouts rather

loudly to insist on marking a rupture in his little brother’s story. His strategy is successful

since Oliver is interrupted in his narration; and Thomas catches the attention of his

mother who shushes him and switches back to French to defend Oliver.

Earlier during the same meal, Thomas had already created a break in conversation

by once again introducing French within an exchange in English. This time, he also marks

the rupture physically, by getting up of his chair.

100

Example (23) – date: 12/03. ���������,�� :3H7�KAG�E77@�F:3F�3F�3>>������ ����(*.�� �������������9BJ�� E75AG7�>3�FSF7����������,�� A:�BDA434>K�@AF��47�53GE7�;F�E�8AD�5D7E57@F�DA>>E�I:;5:�;E�F:7�������� � 3?7D;53@�5DA;EE3@F�E����������(*.�� 0�7@91�:3�@�;�K73:�����������#$ �� 3G�D7HA;D� ���������35F�� E7�>QH7�67�E3�5:3;E7��������(*.�� @3@�7:�;�0�1�7:�;�0�1�7:�;�0�1�7:�;����������JB@F�� BA;@F7�>3�5:3;E7�67��#$ �� �������,�� ���;@�F:7�5A??7D5;3>E��� ����(*.�� FG�H;7@E�;5;�FAGF�67�EG;F7���� �����JB@F�� BA;@F7�>3�5:3;E7�67��#$ ��In this extract, the mother and Christy are speaking English. Suddenly, Thomas gets up

and says “au revoir”, voluntarily interrupting the conversation between the two adults.

The mother is forced to call him to order and she does so by switching back to French.

Once again, Thomas was able to disturb the ongoing interaction thanks to his change in

languages, which is a powerful tool to assert his individuality among the family members.

Similarly, Oliver adopts the same approach when he wants to distinguish himself

and catch the attention of a family member. While Christy is speaking English to the

mother, Oliver jumps into the conversation in French to arouse his mother’s interest and

change the topic:

Example (24) – date: 12/03. ������(*.�� 0�7@91�EA��:D;EFK�;E�35FG3>>K�8D7@5:��� �������#$��� ���?3?3�@��� ��������,�� K7E�I7>>�?K�9D3@6?3��� � ����(*.�� �����0���D;F1� � �����,�� ;�9G7EE�?K�9D3@6B3�;E�F:;E�>;=7�3?7D;53@�DAG?3@;3@�JJJ��� � �����#$��� ���:G�?�>N�0� 1 � >N�3BDQE�?7D5D76;�� � ����(*.�� 0�7@91�K73:��� � �����#$��� ���5�7EF� <7G6;�� � ������,�� EA�0���D;F1�� � �����#$��� ���?3?3@��� � �����#$��� 3BDQE�?7D5D76;� < �3;�EE�A�5�7EF� <7G6;� �� � ����(*.�� AG;��

Oliver intervenes in his mother’s conversation with Christy and the fact that he chooses

French immediately contrasts with what the two adults are talking about. Just like

Thomas, he is able to break off the conversation and he has his mother switch to French

(“oui”). He managed to get her attention even though he engaged a completely different

topic.

Changing languages to change topics is actually quite common in the recordings.

It also is a form of rupture in the flow of interaction, implying that the participants move

101

on to another frame of conversation, or to another phase of the meal. In the following

example (25), Thomas suddenly switches his turns from French to English:

Example (25) – date: 12/10. ��������#$ �� <�7��83;E�CGA;�?3;@F7@3@F�?A;������������#$��� �����0���5:3@F71��������(*.�� 43:�FG��@7��H7GJ�B3E�G@�K3AGDF������������#$ �� AG;���� �����(*.�� A=�43:��:D;EFK�7>>7�H3�F�7��6A@@7D�G@�K3AGDF�����������#$��� �����0���5:3@F71���������#$ �� 0�7@91�I:K�3D7�KAG�8 ; >?;@9������������#$ �� 0�7@91�I;F:�F:3F� ���������JB@F�� BA;@F7�>3�53?RD3��������(*.�� 0�7@91�I7>>�KAG�=@AI�EA?7F;?7E�$��3�?�8;>?;@9����������(*.�� 0�7@91�<GEF�FA�:3H7�:G?�;����������(*.�� 0�7@91�EA?7�?7?AD;7E�� It is interesting to notice that, although Thomas keeps addressing the same interlocutor –

his mother – he swiftly goes from one language to another. The reason for this shift most

likely is the change of topic. While they were previously talking about dessert, Thomas

then asks his mother why she is filming him, as one of the dinners was recorded with a

handheld camera (cf. III.2.3.). We note that the mother follows her son’s lead and

answers his question in English. In another recording, the mother herself switches from

French to English in order to notify the end of the activity that she was performing with

her first-born son, Thomas:

Example (26) – date: 11/30. �������(*.�� FG�F�7��D3BB7>>7E������������#$ �� AG;����������(*.�� CG7�>7E�4R4RE�;�>E��0�1�;�>E��5A??7�>N����������JB@F�� BA;@F7�>7�>;HD7�67��#$���������(*.�� FG�HA;E��>3�?3?�3@���0�1�>3�?3?3@�>N�7>>7�7EF�7@57;@F7�����������#$ �� ����<7�E3;E��0�1���������#$ �� <7�E3;E������ ���(*.�� 0�7@91�A=�E:AG>6�I7�?3=7�3�5A8877����������!�.�� $�53@�?3=7�;F����������35F�� E7�>QH7�BAGD�83;D7�6G�538R��In this extract, the mother and Thomas are just finishing reading his book on the

mammals in Australia. The mother suddenly switches to English to mark the end of the

activity and declares that it is time for coffee. Her change of language serves as a clear

break in her interaction with her son. Furthermore, the father responds immediately to her

suggestion. We might assume that another reason why the mother switches to English is

because she addresses her husband. This would be related to the question of which

language is associated with each participant, which is dealt with in the next section of this

chapter. In any case, what is certain is that the mother’s switch clearly marks a breaking

102

point in her conversation and suggests that the family members are moving on to another

activity, another phase of the meal. The mother actually uses this strategy a number of

times throughout the data. She constantly uses French when she wants to obtain her

children’s attention, call them to order or emphasize her remarks. As in the following

examples:

Example (27) – date: 11/30. �������#$ �� 0�7@91�3�=;I;�53@�6D3I�3�=;I;��� ������!�.�� 3�=;I;�53@�6D3I�3�=;I;���� �������#$ �� ��0���D;F1�� ������!�.�� ���$�6A@�F�F:;@=�EA�� �� ���(*.�� A=�.A?�H;7@E�F �3EE7A;D�N�F34>7��� ������!�.�� ���JJJ�� ������(*.�� .A?���� ������(*.�� .:A?3E�H;7@E�F �3EE7A;D� ;5 ; �E � ; > � F7�B>3;F � �� ������JB@F�� BA;@F7�>3�B>357�67��#$ � Example (27bis) – date: 11/30. �������!�.�� �I:A��;�E�F:7D7���0�1������������#$��� 0�7@91�6366K�����������!�.�� I:A��;�E�F:7D7������������#$��� ���0�7@91�$�I3@@3�E77���������35F�� E7�>QH7�7F�5AGDF�H7DE�>3�53?RD3��������(*.�� A=�@3@�0� 1 �@3@�0� 1 �@3@�HAGE�D7H7@7L�HAGE�3EE7A;D�HAGE�@7�FAG5:7L�B3E�������� � N� >3�53?RD3����������(*.�� 7F�HAGE�D7H7@7L�HAGE�3EE7A;D�N�F34>7�E� ; > �HAGE�B>3;F � ���������(*.�� *>;H7D����Indeed, in these two examples, the mother interrupts her sons’ exchange with their father

to order them to go back to their seats. While Thomas and Oliver are both speaking

English with their father in each example, their mother’s call stands out from the

conversation. Her switch emphasizes the importance of her utterance and forces the

children to stop whatever they are doing and to follow their mother’s orders.

In the same logic, a warning on the part of the mother can also induce a shift in

languages. The purpose is similar, that is to create a break in the ongoing exchange in

order to catch her interlocutor’s attention faster. As in the following example (28):

Example (28) – date: 11/30. ��������#$ �� 0�7@91�38F7D�:7�:3E�3�5AEFG?7�EA�4>G7�FAF7�E35=����������9BJ�� 83;F�G@�97EF7�67�>3�FSF7���������#$��� 0�7@91�6366K�����������!�.�� ?:�A������������#$��� 0�7@91�KAG�=@AI�F:3F�F:7�9;D>E�F:3F����������(*.�� A=�HAGE�83;F7E�3FF7@F;A@�3GJ�8D ; F7E�B3D57�CG�7> >7E�EA@F�FDQE�FDQE�5:3G67E����������#$��� ���0�7@91�JJJ���� �����#$��� 0�7@91�6366K�����������(*.�� ; > � 83GF�EAG88 >7D� � ��������(*.�� 6�355AD6���

103

In this example, Thomas, Oliver and the father are exchanging in English. The mother

cuts the conversation short by warning them that the fries are hot. She interrupts them in

her native language to create a clear breach in her family’s interaction. Once again, this

shift emphasizes the importance of her speech and stands as a warning that she is

providing essential information.

Finally, one last reason why the children would switch languages in order to break

from a conversation is to indicate that they wish to leave the table. Therefore, their

change of language marks their wish to end the ritual of the meal and to retreat from the

family conversation. The following extracts illustrate this claim:

Example (29) – date: 01/31. �������!�.�� 6A�KAG�>;=7�;F������ �����#$��� 0�7@91�K7E����������!�.�� D73>>K������������#$��� 0�7@91�@A����������9BJ�� E75AG7�>3�FSF7���������#$ �� <�7��B7GJ�EADF ;D �6�7�� >3�F3�4>7������������#$ �� ?3?3@�E� ; > � F7�B>3;F � ���������(*.�� 47@�;�FG��@7��H7GJ�B3E�G@�K3AGDF����Example (29bis) – date: 01/31. ����!�.�� 5A?7�435=�:7D7��� ����(*.�� ���%3?7E�� �����#$��� 0�7@91�K7E��� �����!�.�� K73:��� � ����#$ �� <�7��B7GJ�EADF ;D �6�7�� >3�F34>7�E� ; > � F7�B>3;F ���� �����(*.�� 3FF7@6E�?A@�B�7�F;F�5A7GD�B3D57�CG�A@��@��3�B3E�8;@;���We can thus notice that each time Thomas asks his mother whether he can leave the table,

he does so in French even though the conversation going on at that moment is in English.

This is a way for him to simply break from the family interaction and indicate that he

wishes to withdraw from the table. Another interpretation is possible. Indeed, we could

also argue that Thomas may actually use French because he is asking the permission of

his mother. He never addresses his father or the caretaker to ask them if he is allowed to

leave. The person who is empowered to dismiss the children from the table seems to be

implicitly and exclusively the mother. This could also be a reason why Thomas asks so in

French.

This section has shown that deliberately switching languages in the middle of a

conversation may be a useful device to create a breaking point in interaction. When the

participants choose to speak the other language, and not the one that is dominant at the

104

time around the table, they are able to stand out from the other family members. It allows

them to mark their disagreement, to request attention, to emphasize their remarks and to

cause the conversation, or the meal, to move on.

4. One interlocutor = one language? Sometimes the children may switch languages for different purposes than the ones

evoked beforehand. It might not be to blend into a conversation, nor to deliberately create

a breach and stand out from a family discussion, but it could realistically have to do with

the association of one language with one interlocutor. In order to shed some light on this

hypothesis, let us look at some examples in the data where the children switch languages,

regardless of the topic that is being addressed or the language in which they are being

spoken to. We will start with the assumption that the children might associate their

mother with the French language:

Example (30) – date: 12/03. ���������,�� ���:7K�*>;�6;6�KAG�F7>>�KAGD�?A?�F:3F�����������,�� KAG�8AG@6�KAGD�9>AH7�3F�E5:AA>����� ������#$��� :3�;����� ����9BJ�� AGHD7�9D3@6�>3�4AG5:7�� ������#$��� < �3 ;� FDAGHR�?A@�0�1 ��� ������#$��� ?A@�93@F�N� > �R5A>7� ��� �����(*.�� FG�>�3E�D7FDAGHR�5�7EF�EGB7D�P3�� In this first example, the caretaker reminds Oliver to tell his mom that he found his glove

at school. In order to do so, Oliver repeats what Christy has just said but he switches to

French while addressing his mother. It seems that, in his mind, a switch is needed in order

to communicate efficiently with his parent. The mother seems to confirm Oliver’s instinct

since she keeps the conversation going in French. This pattern is actually quite recurrent

with Oliver, as we can notice in the following examples:

Example (31) – date 12/03. �������(*.�� 0�7@91�I3E�;F�9AA6������������#$��� �����������9BJ�� :T5:7�>3�FSF7��������(*.�� 0�7@91�K73:������ ����(*.�� 0�7@91�KAG�> ;=7�8 ;E:������������#$��� �����������9BJ�� :T5:7�>3�FSF7���������#$��� :G?�;�:G?�;�����������#$��� < �3 ;?7�4;7@�3FFD3B7D� >7�BA;EEA@�CG;�7EF�63@E�> �73G��

105

Similarly to the previous example, Oliver addresses his mother in French even though the

latter is talking to him in English. In these extracts, the language used by the child’s

interlocutor does not seem to matter. What Oliver takes into account is the fact that his

mother’s native language is French. This is what determines his linguistic choice:

Example (32) – date 12/10. �������(*.�� 0�7@91�3>>�D;9:F�9GKE����� �����(*.�� 0�7@91�:3H7�KAG�8 ;@;E:76�KAGD�6;@@7D����� �����#$ �� ���0�7@91��$��3�?��0�1�$��3�?�9A@@3�6A�3@AF:7D�A@7��� ������#$ �� 0�7@91�38F7D�$�8;@;E:���� ������#$��� ���AG;�0 � 1 �AG;�0 � 1 �� ������#$��� AG;�?3;E�A@�3�8 ;@;�@AF�D7��6;@7D�?3;E�CG�7EF�57�CG�A@�H3�83;D7�?3;@F�7�@3@F������ �����(*.�� 7:�;�47@�;��A@�H3�3>>7D�?7FF�D�7�>7���0���1�A@�H3�3>>7D�BD7@6D7�G@�43;@��� �Even though Oliver switches languages to address his mother, this does not mean that he

wishes to create a break in the conversation. He manages to deal with the same topic and

to keep the exchange going, all the while switching to French. Besides, he is not the only

one to do so among his siblings. Thomas and James may also use a similar strategy with

their mother:

Example (33) – date: 12/10. �������(*.�� 0�7@91�A=3K�9GKE�KAG��3�D7�3>?AEF�A@�:A>;63K�����������#$ �� �������������9BJ�� :T5:7�>3�FSF7���������#$��� ��0���5:3@F71���� ����9BJ�� >QH7�>7E�67GJ�?3;@E�>7@F7?7@F�B3G?7E�AGH7DF7E���������#$ �� �; > �@AGE�?3@CG7��0�1 � ����������#$ �� G@�0� � 1 � ; > �@AGE�?3@CG7�67GJ�H7@6D76;E����������9BJ�� F7@6�BAG57�7F�;@67J��������(*.�� AG;� ����������#$ �� 67GJ�<7G6;E����������9BJ�� F7@6�5;@C�6A;9FE�67�E3�?3;@�93G5:7��������(*.�� AG;� � ���������#$ �� G@�>G@6;� � ��������9BJ�� F7@6�BAG57��� ����(*.�� AG;� ����������#$ �� 7F�E3?76;�6;?3@5:7� Although the mother starts asking the boys about the upcoming holidays in English,

Thomas does not follow her lead and answers her in French. This demonstrates his strong

association of French with his mother and shows that, once again, his interlocutor’s native

language prevails over the language that is being used or the language related to a certain

topic. Finally, the same thing can be observed with James:

Example (34) – date 01/31. ������!�.�� KAG�I3@F�5:77E7���� �������#$��� ���G:�;�� ������JB@F�� BA;@F7�>7�8DA?397�

106

������!�.�� I:3F�6A�KAG�E3K���� �������#$��� 0�7@91�6366K�� � � ������!�.�� ??:�A��� ������(*.�� FG�H7GJ�6G�8DA?397���� �� ���35F�� 5AGB7�6G�8DA?397� �������#$��� AG;� �� ������35F�� ?7F�>3�?3;@�63@E�>7�B3B;7D�67�8DA?397� ������!�.�� ���@A�0�1�@A�0�1�@A�0�1�@A�0�1�@A�� ������(*.�� CG�7EF�57�CG�A@�6;F���� �������#$��� �?�7D5;�� ��This extract is a good example of how James clearly differentiates his father’s language

as opposed to his mother’s. When addressing his father, he usually uses “daddy” or

responds in English, whereas he answers his mother in French and calls her “maman”.

Sometimes, the fact that the children address their mother in French allows them to begin

a conversation with her on the side, independently of an already ongoing interaction. For

instance, let us consider the following example:

Example (35) – date: 11/30. �������!�.�� ;E�;F�>;=7�3�?3>7�83;DK������������#$ �� 0�7@91�@A����������!�.�� A:�;�A=3K�����������#$��� ���?3?3�@����������35F�� EAG88>7�EGD�E3�8AGD5:7FF7���������#$ �� 0�7@91�FDG7�83;DK����� �����#$��� ���5�7EF����������#$��� 5 �7EF�FDAB�5:3G6����������E;F�� (*.�6A@@7�6G�=7F5:GB�N��#$���������(*.�� AG;�5 �7EF�0 � � � 1 � 83GF�0 � � 1 ��; > � 83GF��0� 1 �3FF7@6D7�G@�B7F; F �B7G�; > � 83GF�EAG88 >7D� ���� � In this example, Thomas and the father are not interrupted, while Oliver and his mother

can have a conversation of their own, on the side, in French. This pattern can sometimes

lead to overlapping conversations, happening at the same time, in two different languages

and on two different sides of the table. As in the following examples:

Example (36) – date: 01/31. �������!�.�� 6A�KAG�I3@F�EA?7�KA9:GDF������������#$��� 0�7@91�@A�����������#$ �� ?3?3�@����� ����E;F�� �#$��EADF�67�>3�B;Q57��������!�.�� �I:7D7�3D7�KAG�9A;@9��0�1�����������!�.�� I:3F�����������!�.�� I:7D7�3D7�KAG�9A;@9������������#$ �� ���?3;E�5A??7@F�0�1�����������#$ �� 5A??7@F�A@�7EF�63@E������������#$��� 0�7@91�:7��;�E�9A;@9�FA�B>3K����� ����!�.�� @A�:7�53@�F�B>3K�5A?7�A@�5A?7�435=����������366�� E�36D7EE7�N��#$��BG;E��#$���������9BJ�� FAGD@7�>3�FSF7�H7DE��#$����������#$ �� <7�5DA;E�����������#$��� 0�7@91�@A����

107

��������#$ �� CG7�0�1�CG7�CG3@6�A@�7EF�?ADF�����������#$ �� �A@�H3��0�1�A@�H3�63@E�>�7��5;7>����������9BJ�� >QH7�>3�?3;@�6DA;F7�7@�>�3;D���������#$ �� 7�F����������9BJ�� 43>3;7�>3FRD3>7?7@F�?3;@�6DA;F7�B3G?7�H7DE�>7�43E��� �����#$ �� �7F�BG;E��0�1�7G:�;�����������#$��� ���0�7@91�5A?7�3@6�73F���������366�� E�36D7EE7�N��#$����������#$ �� 7F�BG;E�A@�7EF�63@E�>�7EB357�� In this example, the father, Oliver and James are conversing in English. The father is

trying to get James to come back to the table after the latter stormed out of the kitchen

declaring that did not want a yoghurt. Oliver intervenes to tell his dad that James went to

play in the other room and then he takes his father’s side, turning to James and telling him

to come and eat. In the meantime, Thomas is telling his mother about what happens when

people die according to him. So while Oliver and James are exchanging with their father

in English one the one hand, Thomas and his mother are conversing in French on the

other hand. This happens quite a few times during the family dinners:

Example (37) – date: 01/31. ��������#$ �� 7F�5�7EF�CG3@6�CG7�FG�H3E�?AGD;D������������#$��� ���0�7@91�53@�$�:3H7�EA?7�4D736�3@6�4GFF7D�����������#$��� 0�7@91�D;9:F�@AI�4753GE7��$�6A@�F�I3@F�3@K?AD7��0�1���� �����366�� E�36D7EE7�N�!�.�� ����(*.�� ���5�7EF�63@E�FDQE�FDQE�>A@9F7?BE��� ������#$��� 0�7@91�$�6A@�F�I3@F�3@K?AD7���� ������#$ �� FA;�FG�H3E�3HA;D�CG7>�O97����� ������#$��� ���0�7@91�JJJ��� �����9BJ�� F3B7�>7�B3;@�EGD�>3�F34>7�� �����(*.�� H7DE�7G:�;�57@F�3@E���� ������#$ �� 57@F�3@E���� ������#$ �� AG�EE�A�57@F�7F�CG3D3@F7�3@E���� �����(*.�� @3�@�H7DE�57@F�3@E�����������#$��� ���0�7@91�$��3�?�@AF�8;@;E:76�4GF�0�1���� �����#$ �� 47@�;�?A;�<�7��H3;E�?AGD;D�CG3@6�<�7��H3;E�3HA;D�?;>>7�3@E�����������#$��� ���0�7@91�4GF�$�6A@�F�I3@F�3@K?AD7����������#$��� ���0�7@91�$�I3@F�3�KA9:GDF���������!�.�� KAG�I:3F��;�E�F:3F�����������(*.�� P3�5�7���@��7EF�B3E�BAEE;4>7�����������#$ �� BAGDCGA;�����������!�.�� KAG�6A@�F�I3@F�3�KA9:GDF�����������(*.�� ���B3D57�CG7�>7�0��1��������(*.�� >�:A??7�>7�B>GE�H;7GJ�EGD�F7DD7�����������#$��� ���0�7@91�K7E��$�I3@F�3�KA9:GDF��0�1���� �����#$��� ���0�7@91�@A����������#$��� 0�7@91��$��I3@F�3�KA9:GDF����������(*.�� ;>�6A;F�3HA;�D�7G:�;�<7�5DA;E�CG�7��5�7EF�57@F�CG3FADL7�3@E�����������#$��� ���0�7@91�6366K���������(*.�� >7�D75AD6���

108

This example follows a similar pattern. Again, while Oliver, his father and James are

conversing in English on one side of the table, Thomas and the mother are having a

discussion in French. Differentiating languages can thus allow the children to delimitate

interactions and to establish a frame with their interlocutors. According to the data, the

boys tend to choose their interlocutor’s native language in order to efficiently

communicate with them. And then, as we have already pointed out in a previous section,

it is up to the other participants to adapt to the chosen language if they wish to blend into

the ongoing exchange.

While most of the time the mother is associated with French, we can still find

some counter-examples when the child keeps answering her in English. As in the

following extract:

Example (38) – date: 12/10. ������(*.�� 0�7@91�I:3F�6A�KAG�F:;@=�6A�KAG�>;=7�FA�47�8;>?76���� �������#$ �� ������ ������9BJ�� 83;F�97EF7�67�43FF7?7@F�>7E�67GJ�;@67J�BA;@FRE�7@�>�3;D� �������#$��� ���JJJ�� �������#$ �� 0�7@91�@A��� �����(*.�� 0�7@91�I:K�@AF���� �����#$ �� 0�7@91�4753GE7�:G?�;�$�0�1�$�6A@�F�I3@F�FA�>AA=�;@�0��1�3F�F:7�53?7D3�� �����(*.�� BAGDCGA;���� ������#$ �� 0�7@91�4753GE7��� �����(*.�� 43:�;�BAGDCGA;���� ������#$ �� 0�7@91�4753G�E7�0���5:G5:AF71� � �����9BJ�� ?7F�E7E�?3;@E�7@�DA@6�3GFAGD�67�E7E�K7GJ� This extract stands as a counter-example of the one interlocutor–one language hypothesis

because, although the mother has switched to French in the middle of her conversation

with Thomas, the latter keeps answering her in English. In this case, the language used

prevails over the interlocutor’s mother tongue. Therefore, claiming that the children

address their mother in a systematized way would be inaccurate. It is difficult to predict

whether they will choose to switch while interacting with her. However, it seems that

using French when conversing with the mother still remains the dominant trend.

Similarly, we can notice that the children more easily tend to associate their father with

the English language:

Example (39) – date: 01/31. �������(*.�� 3H75�CGA;�HAGE�3H7L�83;F�>7�4A@:A??7�67�@7;97������������#$ �� ���A@�3���������9BJ�� F7@6�>3�?3;@E�H7DE�CG7>CG7�5:AE7�67DD;QD7�(*.�� ������#$��� 47�@�A@�0�1�A@�3�83;F����������!�.�� ���0� 8D31�3H75�67��0 � 1 �6�7�� >3�83D;@7�@3@�����������#$��� 0�7@91�@A�I;�F:��0 � 1 �I;�F:��0 � � 1 ��I;F:�F:7��0�1 �I;F:�F:7�E@AI���

109

In this extract, the mother and the first two boys are conversing in French. After the

children told her that they made a snowman, she asks them what they made it with. The

father intervenes in the conversation in French, in order to be integrated into the

interaction. However, Oliver immediately switches to English to answer his father. He

seems to strongly associate him with the English language and thus adapts his speech in

order to communicate with his father. This happens several times during the recordings.

Oliver seems to have no problem switching to English for his father, and then going back

to French to continue his exchange with his mother. Let us consider the following

example, taken from another family meal:

Example (40) – date: 11/30. �������#$��� >;E347F:�;>�0�1�;>�D7EF7�;5;�?3?3@���� �������#$��� BAGDCGA;� >;E347F:�; > �D7EF7�;5;���� ������!�.�� :G?�;���� �������#$��� 0�7@91�$ ��3�?�F3>=;@9�FA�?A�?��� ������!�.�� I:K�3D7�KAG�F3>=;@9�FA�?A?�I:K�3D7@�F�KAG�F3>=;@9�FA�?7���� �� ���!�.�� KAG�@7H7D�F3>=�FA�?7�3@K?AD7��� �������#$��� ?3?3@���� ������(*.�� ����CG�7EF�57�CG�;�>��K�3��0�1�>AG>AG��� ������(*.�� CG�7EF�57�CG�;�>��K�3���� ������!�.�� ����I;>>�KAG�F3>=�FA�?7��0�1��� �������#$��� BAGDCGA;� >;�E347F:��0 � � 1 � > ;E347F:���� ������!�.�� �I;>>�KAG�F3>=�FA�?7��0�1�I;>>�KAG�F3>=�FA�?7�� �������#$��� 0�7@91�@A��� � This example is particularly interesting because it clearly shows how Oliver distinguishes

his parents’ two languages. While he is asking a question in French to his mother, the

father tries to obtain his attention. Oliver turns to his father and swiftly changes to English

to state loud and clear “I (a)m talking to mo:m”. Switching for his father is not an issue

and actually seems to be the best way for him to be heard by his father; and then he

returns to his question to his mother in French. The father playfully complains that the

child never talks to him anymore. He keeps asking Oliver if he will talk to him instead of

his mother, to which the child responds with a firm “no:” then goes back to his dialogue

in French with the mother. This may confirm the hypothesis that the children associate

their father with the English language. Besides, earlier during the same meal, something

intriguing happens with Thomas while he is interacting with his father:

Example (41) – date: 11/30. �������#$ �� 0�7@91��3@6�38F7D�F:3F��0�1��� � �����#$ �� 0�7@91�3@6�38F7D�F:3F��� � ����#$ �� 0�7@91�F:7K�9A�AGFE;67��� � �����#$��� ����5��7EF�5:3G6�� � ����!�.�� :G?�;��� � ����!�.�� A88�F:7�4GE�F:7K��3�D7�9A;@9�AGF�A8�F:7�4GE���

110

� �����#$��� ���0�7@91�FA�F:7�93D67@�� � �����#$ �� 0�7@91�K73:��� � ����!�.�� ���0�8D31�AG3;E�� � ����!�.�� 0� 8D31� ; > �EADF�6G�4GE�� � � �����#$ �� 0�7@91� ; > �EADF�6G�4GE���� ������B:A�� BDA@A@5;3F;A@�N�>X3@9>3;E7� �� ���(*.�� @3�@��� ������(*.�� �7@�EADF3@F�6G�4GE��0���1�7@�67E57@63@F�6G�4GE�CG�7EF�57�CG�;>E�A@F�HG��� �������#$ �� G@�5:3F�73G��0��1�>7�5:3F73G��67��>3�D7;@7��� ������!�.�� 0�8D31�3:�>7�5:3F73G�67�>3�D7;@7�A=���In this example, the father suddenly decides to switch to French in the middle of a

conversation in English with Thomas. Even though the child masters both languages, he

does not understand what his father has just said. He repeats his father’s utterance with an

English pronunciation, suggesting that he actually believes that his dad is speaking

English. This most probably is the reason why he is unable to understand his father’s

speech. Indeed, because he strongly associates the latter with English, he does not realize

that his father has in fact switched languages. His mother intervenes to reformulate the

father’s utterance and rekindle the conversation. Thomas is able to continue his story and

the father shows that he has understood by repeating in French “ah le chateau de la reine

ok”, as a sign of agreement.� Another interesting link that the participants establish

between the father and English is shown through the following interaction between the

mother and the children:

Example (42) – date: 12/10. �������(*.�� 7F�?A;�5A??7@F�<7�?�3BB7>>7������������#$��� -3@6D;�@7�����������#$��� �����0���FAGEE71��������(*.�� :3��A����������(*.�� 0�7@91�3@6�636�����������(*.�� 0�7@91�I:3F�� ;�E�636�E�@3?7������������#$��� %A�:@�0���5D;71��In this example, the mother and the children are conversing in French until the topic

moves on to the father. When asking her boys what the name of their father is, the mother

switches to English. It is assumable that the family members associate the father with

English so much that, even when he is not present at the table, the simple fact that they

are talking about him leads to a switch in English. The children keep the conversation

going in the same language, until Oliver readdresses his mother directly and switches

back to French to mark the end of his reflection:

Example (43) – date: 12/10. �������(*.�� 0�7@91�I:3F��;�E�636�E�@3?7������ �����#$��� %A�:@�0���5D;71���������(*.�� ??:�A���

111

��������#$ �� 0�7@91��6A��KAG�=@AI�:AI�FA�ID;F7�%A:@�����������366�� E�36D7EE7�N��#$����������#$��� 0�7@91�@A�����������#$ �� 0�7@91�<�>�A�>�:�>�@�>����������(*.�� 0�7@91�F:3F��;�E�D;9:F�����������#$��� ���0�7@91�<�>�A�>����������#$��� 0�7@91�:�>�@�>����������(*.�� ??:�A����� �����#$��� ?A;� <7�E3;E�R5D;D7�%A:@� ���������35F�� D793D67�(*.��������(*.�� 3:�4A@�5A??7@F�P3�E�R5D;F������������#$��� 3>ADE���Y���After the mother switches to English when talking about the father, Thomas continues in

the same language and asks his little brother if he can spell “John” in English. Once

Oliver repeats what his older brother has taught him, he turns to his mother and declares

in French “moi je sais écrire John”. His language here becomes linked to his interlocutor

again, while it was previously linked to the topic: the father.

One last example that is worth evoking is when the mother declares that it is time for the

boys to go put their pajamas on. She says so in French, which triggers an interaction in

the same language with Thomas. But on the other hand, Oliver asks his father if he can go

put his pajamas as well, and to that purpose he switches to English:

Example (44) – date: 01/31. ������(*.�� A=�3BDQE�6;@7D�A@�H3�3> >7D�?7FFD7� >7�BK<3?3��� ������(*.�� 6�355AD6���� �������#$ �� ?3?3�@��� ������(*.�� AG;���� �������#$ �� <�7��6A;E�83;D7�CGA;��0���?3D?A@@71��� �� ���(*.�� ??:�A���� ������9BJ�� B3EE7�E3�?3;@�63@E�>7E�5:7H7GJ�67��#$ � �������#$ �� ��0���5:3@F71�� ������(*.�� 7EF�57�CG7�F�G��3E�83;F�F7E�67HA;DE�FA;�7G:�;�7@�3@9>3;E�5�7�F�3BDQE�?;6;�FA;��� ������9BJ�� 53D7EE7�>3�FSF7�67��#$ � �������#$ �� CGA;���� ������(*.�� 7EF�57�CG7�F�G��0�1�FG�3E�FD3H3;>>R�F7E�67HA;DE�7@�3@9>3;E���� ������JB@F�� BA;@F7��#$ � �������#$ �� AG;��� �� ���(*.�� AG;�F�G��3E�FAGF�8;@;���� ������(*.�� B3D57�CG7�F�G��3H3;E�473G5AGB�67�5:AE7E�N�83;D7�:7;@�;��� �������#$��� 0�7@91�6366K��� �������#$ �� ���??�A�<7��@7��� �������#$��� 0�7@91�53@�$�9A���� �������#$ �� ���5DA;E�B3E�HD3;?7@F�� �������#$��� 0�7@91�BGF�?K�BK<3?3E�A@���� ������(*.�� ���F�G���@��3E�B3E�FAGF�8;@;��� �� ���9BJ�� E75AG7�>7�FSF7� �������#$ �� ���?3;E�A@�3�83;F�� ������!�.�� K7E�KAG�53@���

112

Even though the mother addresses the topic in French, the fact that Oliver asks his father

is enough to make him switch languages. Therefore, in this case, the language of his

interlocutor prevails over the language of the topic. This illustrates and corroborates the

claim that the children may associate one dominant language with each of their parents.

5. The language(s) of emotions This section suggests that sometimes the languages the children choose to use may

be related to their emotional state. I noticed that when they ask for a kiss or a hug, they

almost always address their mother in French. On the other hand, when the latter

reprimands them or becomes irritated, she switches to English. For instance, in the

following examples, the children go to their mother in French while feeling rather cuddly:

Example (45) – date: 12/03. �������(*.�� 3>>7L�%3?7E�H;7@E�F�3EE7A;D�H;7@E�?3@97D����������366�� E�36D7EE7�N��#$���������JB@F�� BA;@F7�>�3EE;7FF7�67��#$����������#$��� G@�5O>;@� ���������B:A�� G@�F;�>;@��������(*.�� FG�H7GJ�G@�5O>;@������������#$��� AG;���Y�� ������(*.�� 6�:34;FG67�FG�?3@97E�4;7@��� �������#$��� �G@�5O>;@��0� 1 � � �� ������E;F�� E7�>QH7�7F�E7�6;D;97�H7DE�(*.� �������#$��� G@�5O>;@�� �� ������(*.�� AG;�?A@�5A7GD�3FF7@6E�B3E�?3;@F7@3@F�B3D57�CG7�>N�BAGD�>�;@EF3@F�<�3;�%3?7E���Y�� ������#$ �� :3@�;�<�7���@��7@�H7GJ�B3E��67��57>>7�>N��� ������#$��� ��0���5D;71� �����E;F�� E7�6;D;97�H7DE�(*.� � ����#$��� ?3?3@�FG�B7GJ��?7��83;D7�G@�4;EAG�EGD� >�7��B;76�< �3;�?3>�N� ���� � In these extracts, James and Oliver both ask their mother for a hug or a kiss in French.

The data does not provide examples where they do so in English, which suggests that they

tend to speak French to their mother when looking for comfort. On the other hand, when

the mother is upset and reprimands the boys, most of the time she switches to English.

The following extracts illustrate this point:

Example (46) – date: 12/03. �������(*.�� ���F�G��7E�G@�;?4R5;>7����������#$ �� <7��@7��EG;E�B3E�G@�;?4R5;>7����������(*.�� 0�7@91�KAG��3�D7�47;@9�E; > >K� ���������(*.�� 0�7@91�.A?�0�1 �.A?�$��3�?�EAD7�A@�?K�435=�KAG�=@AI�F:3F�JJJ�EA�B>73E7��� � .:A?3E��

113

Example (46bis) – date: 12/10. ������(*.�� ���A:�;��%3?7E�3DDSF7��0� 1 � � ������(*.�� %3?7E�3DDSF7��� �������#$��� ��0���B>7GD71�� ������(*.�� 0�7@91�A=��%3?7E�9D77@�53DB7F��0� � � 1 � � � � ������(*.�� %3?7E�KAG�9A�FA�F:7�9D77@�53DB7F��� � �����#$��� �����0���B>7GD71� ������(*.�� 0�7@91��9D77@�53DB7F��0� 1 �B>73E7��9D77@�53DB7F��0� 1 � �� �������#$��� �����0���B>7GD71� ������(*.�� 0�7@91�9D77@�53DB7F��� �������#$��� �����0���B>7GD71� ������(*.�� 0�7@91�$ �6A@�F�I3@F�3@K�I:;@;@9�;@�F:;E�:AGE7��� �������#$��� @3�@�0���B>7GD71�� ������(*.�� 0�7@91�KAG�9A�FA�F:7�9D77@�53DB7F�@AI��� Example (46ter) – date: 11/30. �������#$��� �����0���F;D7�4DGK3??7@F�>3�>3@9G71� ������(*.�� 3FF�7@6E��*>;H7D�EFAB�E���� ������(*.�� 0�7@91�EFAB�AD�$ ��I;� > > � F3=7�KAGD�KA9:GDF�AGF�� �� ������JB@F�� BA;@F7��#$��� � Example (46quater) – date: 01/31. ������(*.�� @3@�0�1�@3@�0�1�@3@�.A?�0�7@91��F:3F�� ;�E�K3=;��0� 1 ��F:3F�� ;�E� ������ � D73> >K�K3=;��0� � � 1 � $ �6A@�F�I3@F�FA�E77�F:;E� ; F �� ;�E� ;@5D76;4>K�K3=;� � ������9BJ�� 97EF7�>3FRD3>�67�>3�?3;@�6DA;F7� �� ����#$ �� ��0���D;F1����In all these short extracts, the mother switches to English when reprimanding her boys.

She usually starts in French then switches and continues in English. Her switch marks a

break in the ongoing exchange and serves as evidence of her discontentment. Her use of

English as the language for admonition is more obvious than in the case of the father.

Indeed, as it is the latter’s native language, it is difficult to assume whether he also

chooses to reprimand his children in English, or if he simply stays in his dominant

language.

Example (47) – date: 11/30. ������!�.�� A=3K�.A?�0�1�.A?�I:3F�3D7�KAG�6A;@9�4G6�6K���� ������!�.�� 3D7�KAG�8;@;E:76�FA�73F���� ������!�.�� @A�KAG��3�D7�@AF�8;@;E:76��47�53GE7�KAG�:3H7@�F�9AF�KAGD�=;I;8DG;F�� �� ��� �� EA�KAG�97F�GB�3@6�9A�FA�KAGD�B>357�� ������!�.�� .A?���� ������!�.�� E;F �3F�F:7�F34>7�5A?7�A@��� Example (48) – date: 01/31. � ����#$��� ���� � ����35F�� E3GF7�EGD�E3�5:3;E7� � ����!�.�� @A�;8 �KAG�6A�F:3F�KAG�I;> > � 83> > �6AI@�393;�@�� � � ����!�.�� IA@�F�KAG��� � �����#$��� ��0���D;F1�� � ����!�.�� EFAB�F:3F��� � ����(*.�� 3>>7L�LAG�A��

114

As we can see in these two extracts, the father stays in English while reprimanding

his sons, while the conversation taking place was not necessarily in the same language.

Indeed, we note that the mother is speaking French in the second example (“allez

zou@o”). It thus seems that both parents use English for expressing dissatisfaction and

reprehending the children.

Language can therefore be related to an emotional state with the participants. The

recordings all suggest that French is more likely to be used in a cuddly situation when the

boys are looking for comfort; while English tends to be associated with reprehension and

the raising of one’s voice.

6. Concluding remarks To conclude, this qualitative analysis has shown that a family conversation is not

purely a simple succession of turn-takings and that human communication takes place

through various mechanisms. While some parameters are common, there is no “correct”

or “ideal” frame of conversation. It is up to the members of a family to construct and

articulate their exchanges. Conversational settings are continually reconstructed,

especially within a bilingual family. When looking at the participants’ alternate use of

their two languages, the analysis has revealed four predominant trends for their switch:

First of all, both children and parents may change languages to adapt to an already

ongoing conversation at the table. A switch is sometimes needed on the part of the boys

to obtain a response from their parents or to be taken into account within an ongoing

exchange. And similarly, if the parents wish to be understood and communicate

effectively with their children, they also have to deal with language changes in order to

make themselves heard. The second main trend when switching languages is to

voluntarily create breaks in the conversation. Indeed, a deliberate switch in the middle of

an interaction may be a useful device to create a breaking point. It allows the individuals

to stand out from the other family members. That way, the participants can easily mark

their disagreement, request attention, emphasize their remarks, or cause the conversation,

or even the meal, to move on. Thirdly, a switch can sometimes result from the association

of one parent with one language. Indeed, occasionally, the children may choose to address

their mother of father in their native language, regardless of which language they are

spoken in. Finally, language switches may also be related to emotions. As it has been

shown, French is more likely to be used for comforting and cuddling, while English is

115

allegedly associated with admonition on the part of both parents. So ultimately, switching

languages is an essential tool that is used by all the participants in order to assert their

place within the family conversations.

116

117

VI. GENERAL CONCLUSION

In conclusion, this dissertation aimed at showing that the language alternations of

bilingual children are an intentional process that actually reflects an additional

communicative competence. Having two languages at one’s disposal in conversation is

indeed a powerful tool to distinguish and assert oneself among other family members.

This study has accounted for the multidimensionality of bilingualism and has shown that

the children in this research project could be considered as sequential early bilinguals. As

bilingual learners, they are able to adapt to each situation of communication and express

themselves in an appropriate way, given that they are regularly provided with adequate

and rich experience in both languages. Moreover, most research tends to conclude that

bilingualism is linked to general advantages and positive effects on the linguistic,

cognitive and social development of children. This assertion corroborates the idea that

language mixing is foremost a competence and an asset rather than a linguistic deficiency.

Furthermore, we also have accounted for the interest of studying dinnertimes

when looking at children’s linguistic behavior, as mealtimes constitute an opportunity for

the latter to learn how to coordinate the activities of eating and conversing. They

gradually learn to assert themselves, express their desires and affirm their choices. They

also internalize a number of rules and norms that are being taught to them through the

sharing of food practices and the learning of good manners and appropriate behavior

around the table. Dinnertimes are thus cultural sites where the members of different

generations come to transform each other’s way of interacting, acting or seeing the world;

and where the children’s language uses are being constructed and shaped everyday

through family interactions.

It has also been shown that when people speak in an ongoing conversation, they

do so in the light of what has just been said, and in anticipation of what might take place

in a future sequence. Therefore the participants understand the talk of other people and

construct their own speech accordingly. So when trying to understand a particular

utterance or conversational action, we have demonstrated the importance of considering

where and how that action is located in a sequence of other conversational actions. The

meaning of an utterance, the way it was formulated and the way it is interpreted by the

co-participants, thus depends on its context, both verbal and non-verbal. It is up to the

118

members of a family to construct and articulate their exchanges. In this way,

conversational settings are continually revisited, especially within a bilingual family.

When looking at the children’s language alternations, we were wondering whether

their bilingualism was balanced or dominant. The quantitative analysis conducted in this

dissertation has led to demonstrate that Thomas, Oliver and James are more proficient in

French than in English. Indeed, the three children produce more words and within a

greater variety in French. This result is quite surprising since the three boys were born in

an English-speaking country. It thus seems that, even though they had started to acquire

English as a first language, their current environment and the increased use of French by

their mother have ultimately led them to be dominant in French. Contrarily to the

mother’s opinion that her boys are Anglophones, this study seems to demonstrate that

French might actually have supplanted English in their everyday environment. Moreover,

this hypothesis has been confirmed with the calculations of the children’s Mean Length of

Utterances and Turns, which have shown that the latter are longer and composed of a

larger number of words in French. Yet, it has also been observed that the children’s turns

are less fragmented in English – as they globally consist of fewer utterances – implying

that their speech flows slightly better in English.

We have also established that French is likewise dominant in the speech of the

children’s interlocutors. Indeed, their main interlocutor is their mother, and even though

there is always one adult whose native language is English present at the table, the total

number of words and utterances produced in French is still higher. The linguistic choices

of the parents thus play an undeniable role in the way the children chose to manage their

two languages. This may be the reason why the boys speak more French themselves, as

they may adapt to their interlocutor(s). Furthermore, the study of gestures related to the

children’s use of their two languages, tends to show that they use more gestures in their

dominant language. As a result, we can conclude that the children do not compensate

their linguistic discrepancies with a higher number of gestures; but on the contrary, their

gestures are proportional to their verbal communication in each language.

Finally, as we were wondering whether each language had a specific function in

communication, the qualitative analysis has revealed that the children’s switches could

actually be classified into a few predominant trends. First of all, both children and parents

may change languages to adapt to an already ongoing conversation at the table. In order

to be taken into account within an ongoing exchange, or simply to be understood and

119

communicate effectively with one another, all family members have to deal with

language changes at one point, in order to make themselves heard. The second main trend

when switching languages is to voluntarily create breaks in the conversation. Indeed, a

deliberate switch in the middle of an interaction allows the individuals to stand out from

the other family members. That way, the participants can easily mark their disagreement,

request attention, emphasize their remarks, or cause the conversation, or even the meal, to

move on. Thirdly, the analysis seems to suggest that a switch can sometimes result from

the association of one interlocutor with one language. Indeed, occasionally, the children

may choose to address their mother of father in their native language, regardless of which

language they are spoken in. Finally, we also have underlined the possible link between

the children’s language switches and their emotional state. As the study of this family

proposes, French is more likely to be used for comforting and cuddling, while English is

allegedly associated with admonition on the part of both parents.

We have thus shown that switching languages is an essential tool that is used by

all participants in order to assert their place within the family conversations. Language

alternations result from the children’s initiatives and are voluntarily used to various

purposes in conversation. As an intentional process, language mixing ultimately shows

the children’s communicative competence and their uniqueness as bilingual learners.

While family dinners constitute privileged moments for all children to assert themselves

and affirm their choices, bilingual children experience the advantage of having two

languages at their disposal to do so. Juggling their two languages is therefore a

constructive device to assert their individuality in conversation and distinguish

themselves from the other members of their family.

The pattern presented in this dissertation is by no means exhaustive and all of the

preceding conclusions must be confronted with further analyses. Indeed, the brevity of

this research prevents us from making any definitive assertions. As the family meals were

recorded over an episodic period of three months for this study, further analyses would

have to be performed on a more extended period of time in order to provide a more

detailed account of the ways in which conversation functions within this bilingual family.

It would thus be interesting to conduct a longitudinal study to determine whether the

children are well and truly dominant in French, and to see whether their proficiency

increases with time in either of their languages. It would also be interesting to compare

and contrast the way the children in this family manage their two languages with other

120

bilingual families. Indeed, we can wonder whether the patterns observed in this

dissertation are specific to this particular family or whether they are recurrent in several

households. In the same vein, it could also be interesting to confront several families

speaking different languages to confirm or disprove the same tendencies in other cross-

cultural and cross-linguistic families. Hence, different analyses may lead to more accurate

and more reliable results, which would consequently enrich this research and help further

contributing to the fields of bilingualism and bilingual acquisition.

121

BIBLIOGRAPHY

Auer, P. (1995). “The Pragmatics of Code-switching: A Sequential Approach”, in Lesley Milroy and Pieter Muysken (eds.). One Speaker, Two Languages: Cross- disciplinary Perspectives on Code- switching, Cambridge University Press, pp. 115-135. Auer, P. (1998). Code-Switching in Conversation: Language, Interaction and Identity, Routledge, London. Baetens Beardsmore, H. (1986). Bilingualism : Basic principles, 2nd edition. Clevedon, England: Multilingual matters. Beebe, L. M., & Cummings, M. C. (1996). “Natural speech act data versus written questionnaire data: How data collection method affects speech act performance”, in S. M. Gass & J. Neu (Hg.), Speech acts across cultures. Berlin: Mouton de Gruyter, pp. 65-86 (Original version 1985). Bloomfield, L. (1933). Language. London: George Allen & Unwin. p. 56. Boztepe, E. (2003). “Issues in Code- Switching: Competing Theories and Models”. Teachers College, Columbia University. Vol 3, n°2, Caron, A. H. et Caronia, L. (2005). Culture mobile : les nouvelles pratiques de communication, Montréal : Presses universitaires de Montréal, Paramètres. De Houwer, A. (2009). Bilingual First Language Acquisition, Multilingual Matters: Textbooks.

Debras, C. (2013). L'expression multimodale du positionnement interactionnel : Étude d'un corpus oral vidéo de discussions sur l'environnement en anglais britannique (Doctoral Thesis). Université Paris 3- Sorbonne Nouvelle.

Fishman, J.A. (1977). The Social Science Perspective, in Bilingual Education : Current Perspectives, Social Science, Arlington (VA), Center for Applied Linguistics. Genesee, F. (2008). “Bilingual First Language Acquisiton : Evidence from Montreal”, Diversité Urbaine, pp. 9-26 Genesee, F. & Nicoladis, E. (2005). “Bilingual First Language Acquisition”, Journal of Child language, 22, 611-631.

Genesee, F., Nicoladis, E. & Paradis, J. (1995). “Language differentiation in early bilingual development”, Journal of Child Language, Vol 22, pp. 611- 631. Grice, H. P. (1975). Logic and conversation. In P. Cole and J.L. Morgan (eds.), pp. 41- 58.

122

Grosjean, F. (2008). Studying Bilinguals, Oxford: Oxford University Press. Grosjean, F. (1994). Individual Bilingualism, in The Encyclopedia of Language and Linguistics, Oxford, Pergamon Press. Grosjean, F. (1982). Life with Two Languages: An Introduction to Bilingualism. Cambridge, Mass: Harvard University Press. Hagège, C. (2005). L’enfant aux deux langues, Odile Jacob. Haynes, J. (2007). Getting started with English language learners: How educators can meet the challenge. Alexandria, Virginia: Association for Supervision and Curriculum Development. Kail, M. (2015). L’acquisition de plusieurs langues. “Que Sais-Je”, n° 4005. Presses universitaires de France. Kremer-Sadlik, T. & Morgenstern, A., Beaupoil P., Caët S., Debras C. & le Mené, M. (2015). “Eating fruits and vegetables. An Ethnographic Study of American and French Family Dinners”. Appetite, vol 89, pp. 84-92. Labov, W. (1972). Sociolingustic Patterns. Philadelphia: University of Pennsylvania. Lambert, W. E. (1974). Culture and Language as Factors in Learning and Education, in Cultural Factors and Learning in Education, R.D. Meade (eds.), Bellingham, Western Washington State University. Lanza, E. (1997). Language Mixing in Infant Bilingualism – A sociolinguistic perspective, Oxford: Oxford University Press. Lenneberg, E.H. (1967). Biological foundations of language, New York: John Wiley and Sons. Levinson S.C. (1983). Pragmatics. Cambridge : Cambridge University Press. Macnamara, J. (1967a). “The Bilingual’s Linguistic Performance”, Journal of Social Issues. Vol 23, pp. 58-77. Merriam, G. &C. (1961). Webster's Third New International Dictionary of the English Language, Unabridged. Riverside Press in Cambridge, Massachussetts. Mey J. (1993). Pragmatics: An Introduction. Oxford: Blackwell Morgenstern A. (2009). L’enfant dans la langue. Presses Sorbonne Nouvelle.

Morgenstern, A., Parisse, C. (2007). “Codage et interprétation du langage spontané d'enfants de 1 à 3 ans”. Corpus n°6, “Interprétation, contextes, codage”, pp. 55-78.

123

Nilep, C. (2006). “Code-Switching in Sociocultural Linguistics”. Colorado Research in Linguistics. Boulder: University of Colorado. Vol 19, pp. 1-22

Ochs, E. (1979). “Transcription as theory”. In E. Ochs, and Schieffelin, B. (Ed.), Developmental Pragmatics. New York, NY: Academic Press.

Ochs, E., Smith, R., & Taylor, C. (1989) “Dinner narratives as detective stories”. Cultural Dynamics n°2, pp. 238-57.

Ochs, E., Shohet, M. (2006). “The Cultural Structuring of Mealtime Socialization”. New Directions for Child and Adolescent development, n°111. Wiley Interscience.

Ochs, E., Shohet, M., Campos B., Beck, M. (2007). “Coming Together at Dinner: A Study of Working Families”. UCLA Sloan Center on Everyday Lives of Families. Working Paper n°70, pp. 57-70.

Peal, E., Lambert, W. E. (1962). The Relation of Bilingualism to Intelligence, Psychological Monographs, 76, pp 1-23. Poplack, S. (1980,). “Sometimes I’ll Start a Sentence in Spanish y Termino en Espanol: Toward a typology of code-switching”, Linguistics, 18, pp. 581-618. Richards, B. (1987). “Type/Token Ratios: what do they really tell us?”, Journal of Child Language, vol. 14, pp 201-209

Sacks H., Schlegoff E.A., Jefferson G. (1974). “A Simplest Systematics for the Organization of Turn-Taking for Conversation”, Language, Vol. 50, N°4, Part 1, pp. 696-735.

Weinreich, U. (1953). Languages in Contact, La Hague, Mouton.

124

Appendix 1 – Filming Authorization

PRATIQUES LANGAGIERES CHEZ LES ENFANTS BILINGUES

AUTORISATION PARENTALE

I. DESCRIPTION DU PROJET DE RECHERCHE.

Le présent projet de recherche s’inscrit dans le cadre d’un mémoire de Master 2 effectué par Mademoiselle Elisabeth Demoury, à la Sorbonne Nouvelle Paris III et sous la direction de Madame Aliyah Morgenstern. Il a pour objectif principal d’analyser les pratiques langagières chez les enfants bilingues. Pour mener à bien mon projet, je souhaite procéder à des enregistrements audio et/ou vidéo d’enfants bilingues anglais/français en situation d’interaction spontanée dans la sphère familiale. Ces enregistrements seront effectués au domicile des enfants, et peuvent avoir lieu lors d’activités quotidiennes comme la prise de repas, et/ou des moments de jeux et de lecture.

Les enregistrements audio et vidéo seront conservés par moi-même et je serai seule responsable des transcriptions et analyses en collaboration avec ma directrice de mémoire, Madame Aliyah Morgenstern. Des extraits des vidéos et de leurs transcriptions pourront être présentés dans des publications scientifiques, des colloques et des cours d’Université. Il est également possible que d’autres chercheurs de la communauté scientifique française, voire internationale, exploitent ces données dans le cadre de leur recherche sur l’acquisition du langage. Les données audio et/ou vidéo pourront ainsi être mises à disposition de la communauté des chercheurs via une plate-forme de mutualisation des ressources en ligne, notamment sur CHILDES (http/::childes.psy.cmu.edu/) qui est un système en ligne d’échanges de données portant sur le langage de l’enfant.

Il n’existe pas à ma connaissance de risques encourus par les participants à ce projet de recherche.

Je vous remercie de votre contribution à cette recherche scientifique et je m’engage à la plus grande vigilance quant aux aspects éthiques de l’exploitation des données recueillies.

125

126

127

Appendix 2 – Transcription Conventions

�� overlapping utterance

: lengthened syllable

. end of an utterance (≠ turn)

? interrogation sentence

! exclamation sentence

( ) unpronounced

< > delimitation of the phrase that is repeated / corrected / reformulated

xxx unintelligible utterance

0 . silence

[/] simple repetition

[//] repetition with correction

[///] repetition with reformulation

[- fra] utterance in French while speaker dominant in English

[- eng] utterance in English while speaker dominant in French

@o onomatopoeia

@i interjection

@c lexical creation

@s switch: borrowing from another language

%gpx description of a gesture

%xpnt description of a pointing gesture

%act description of an action

%sit description of the contextual situation

%com comment

��0���D;F1� extralinguistic activity: laughing�

��0���B>7GD71� extralinguistic activity: crying

��0���FAGEE71� extralinguistic activity: coughing�

��0���5D;71� extralinguistic activity: shouting�

��0���5:3@F71� extralinguistic activity: singing

��0���>;F1� extralinguistic activity: reading

��0���DO>71� extralinguistic activity: grumbling

128

Appendix 3 – Transcription Extract 01.31.2015

@Languages: fra, eng @Participants: MOT Sandrine Mother, FAT John Father, CHI1 Thomas Child, CHI2 Oliver Child, CHI3 James Child @ID: fra|Biling-Corpus_Demoury|MOT||female|||Mother||| @ID: fra|Biling-Corpus_Demoury|CHI1||male|||Child||| @ID: fra|Biling-Corpus_Demoury|CHI2||male|||Child||| @ID: fra|Biling-Corpus_Demoury|CHI3||male|||Child||| @ID: eng|Biling-Corpus_Demoury|FAT||male|||Father||| @Birth of CHI1: 08-FEB-2009 @Birth of CHI2: 10-JUL-2010 @Birth of CHI3: 30-OCT-2012 @Date: 31-JAN-2014 @Time Duration: 00:27-03:12 48 *MOT: alors elle est bonne la soupe ? 49 *CHI2: +< quoi ? 50 %act: regarde la caméra 51 *CHI2: oui: <il> est bonne . 52 *MOT: oui ? 53 *CHI2: oui . 54 *CHI2: ça fait bonne [/] bonne [/] bonne [/] et bonne . 55 %act: regarde la caméra 56 *CHI2: 0 . 57 %sit: tout le monde mange 58 *MOT: alors v(ous) vous êtes bien amusés aujourd'hui <avec Christy 59 qu'est ce que vous avez fait> [///] ? 60 *CHI1: quoi ? 61 *CHI2: +< ba:h@i . 62 *MOT: qu'est ce que vous avez fait avec Christy ? 63 *CHI2: +< on [//]. 64 *CHI2: ba:h@i on [/] on [/] <on a> [/] <on a fait quelque chose> [///] 65 on a mis les mains dans [/] dans [/] dans [/] dans le(s) 66 fleur(s) qui est blanc . 67 *CHI3: +< cô:té [//]. 68 %act: regarde la table à côté de son assiette 69 *FAT: +< mmh@o ? 70 %act: se tourne vers CHI3 71 *CHI3: +< à côté . 72 *FAT: +< oh@i . 73 *CHI2: et après . 74 %sit: FAT fait un clin d'oeil à CHI3 75 *CHI2: on a fait un bonhomme de neige . 76 *FAT: oh@i okay . 77 *MOT: avec quoi vous avez fait le bonhomme de neige ? 78 *CHI1: +< on a . 79 %gpx: tend la main vers quelque chose derrière MOT 80 *CHI2: be:n on [/] on a fait . 81 *FAT: +< [- fra] avec de: [/] d(e) la farine nan ? 82 *CHI2: [- eng] no wi(th) [/] wi(th) [//] <with the> [/] with the snow . 83 *CHI1: +< ouais . 84 *CHI2: [- eng] the snow was in the pot and we put our hands in it . 85 %gpx: deux mains en l'air poings crispés 86 *CHI1: +< [- eng] with playdough . 87 *FAT: mmh@o . 88 *CHI2: [- eng] and [/] and it made a snowman . 89 *FAT: really ? 90 *CHI1: [- eng] with flour . 91 %add: s'adresse à CHI2 92 *FAT: mmh@o .

129

93 %gpx: hôche la tête 94 *CHI1: +< [- eng] <with flour> [///] we put in flour . 95 *FAT: ok with flour . 96 *MOT: ha:n@i c'est ça quand il a dit les fleurs blanches . 97 %add: s'adresse à FAT 98 *FAT: 0 . 99 %gpx: hôche la tête 100 *MOT: +< 0 [=! rit]. 101 *MOT: ça s'appelle de la farine . 102 *CHI1: +< c'est la farine . 103 *FAT: [- fra] c'est farine c(e) (n') est pas fleur . 104 *CHI1: +< farine . 105 %add: s'adresse à CHI2 106 *CHI2: 0 . 107 *CHI3: [- eng] dada . 108 *FAT: [- fra] oui . 109 *CHI3: 0 . 110 %act: mange une bouchée 111 *CHI3: maman . 112 *MOT: oui mon p(e)tit coeur ? 113 *CHI1: d(e) la farine et du xxx . 114 *CHI3: +< xxx . 115 *MOT: <c'est bon> [/] ? 116 %add: s'adresse à CHI3 117 *MOT: c'est bon la soupe ? 118 *CHI3: oui . 119 *MOT: oui . 120 %gpx: hôche la tête 121 *CHI2: [- eng] daddy to make pancakes . 122 *FAT: to make pancakes yeah ? 123 *CHI2: +< [- eng] we [/]. 124 *CHI2: [- eng] we [/] we [/] we [/] we first put flour in the bowl . 125 *FAT: mmh@o . 126 *CHI2: [- eng] and [//] <and then>[/] we <and then> [/] and then we put 127 an egg in there . 128 *FAT: mmh@o . 129 *CHI2: +< [- eng] and [/]. 130 *CHI2: [- eng] and then it [/] it makes pancakes . 131 *CHI2: [- eng] we give it a s(tir) [//] a stir . 132 %gpx: touille avec sa main droite 133 *FAT: okay . 134 *CHI2: [- eng] and [/] and then it makes a pancake . 135 *FAT: did you make pancakes today ? 136 *CHI2: [- eng] yes . 137 *CHI1: [- eng] no . 138 *FAT: okay . 139 *CHI3: 0 [=! tousse]. 140 *CHI1: [- eng] we haven't made pancakes . 141 *CHI3: +< 0 [=! tousse]. 142 *FAT: you haven't had pancakes . 143 *FAT: you had pancakes at school ? 144 *CHI3: +< ah@i soupe . 145 *CHI1: [- eng] no . 146 *CHI2: [- eng] no because I . 147 *CHI3: +< ah@i soupe ! 148 *MOT: oui c'est la soupe . 149 %gpx: hôche la tête 150 *CHI2: [- eng] daddy . 151 *FAT: mmh@o . 152 *CHI2: [- eng] when [/]. 153 *CHI2: [- eng] when [/]. 154 *CHI2: [- eng] when I (a)m gonna be seventy . 155 *CHI3: +< aha@o soupe . 156 *FAT: when you (a)re seventy .

130

157 *CHI2: +< [- eng] yeah . 158 *FAT: yeah ? 159 *CHI2: [- eng] when I (a)m gonna be seventy and a half . 160 *FAT: seventy and a half yes ? 161 %act: se tourne pour voir ce que fait CHI3 162 %sit: CHI3 lève son assiette vers FAT 163 *CHI2: [- eng] <you (a)re gonna be eight(y) [//] eighty> [/] . 164 *CHI2: [- eng] you (a)re gonna be eigthy . 165 %sit: FAT s'essuie le nez 166 *FAT: no . 167 *MOT: [- eng] mmh@o we (wi)ll be a little be older than that loulou@s 168 %com: loulou=petit nom en français 169 *FAT: I (wi)ll be like a hundred . 170 *CHI1: [- eng] no: . 171 %gpx: penche la tête 172 *MOT: [- eng] seventy: ? 173 *CHI1: [- eng] yeah . 174 *MOT: on aura: presque cent dix ans . 175 %gpx: hôche la tête 176 *FAT: [- fra] ouais . 177 *MOT: 0 [=! rit]. 178 *CHI1: +< cent dix ans ! 179 *CHI3: 0 [=! tousse]. 180 *MOT: on sera très [/] très vieux . � �

131

Appendix 4 – Recap Charts: Language Alternations

�The following charts recapitulate most language alternations that led to the qualitative

analysis in this dissertation. In an attempt to reveal standardized – if not systematic –

switches in languages, all examples were classified beforehand.

�� �' )�$)"�$)�� Line Who

speaks To whom Language Response

in Switch?

to Topic

���� �� � ���������������� ����� ���� ���� ���� ���� ���� ���� ������� �� ���������� ��

�#$ �(*.�(*.��#$ ��#$ �!�.��#$ �!�.��#$��(*.�(*.�(*.�!�.�!�.�!�.�!�.��#$ �

(*.��#$ �!�.�(*.�(*.��#$ �!�.��#$��!�.�!�.��#$ ��#$ ��#$ ��#$ ��#$ ��#$ �(*.�

!D7@5:� @9>;E:�!D7@5:� @9>;E:� @9>;E:� @9>;E:�!D7@5:� @9>;E:�!D7@5:�!D7@5:�!D7@5:�!D7@5:�!D7@5:� @9>;E:�!D7@5:� @9>;E:�!D7@5:�

@9>;E:� @9>;E:� @9>;E:� @9>;E:�!D7@5:�!D7@5:� @9>;E:�!D7@5:�!D7@5:� @9>;E:� @9>;E:�!D7@5:�!D7@5:�!D7@5:�!D7@5:�!D7@5:� @9>;E:�

@9>;E:��

@9>;E:���

!D7@5:� @9>;E:�!D7@5:�

� @9>;E:� @9>;E:�

��

!D7@5:��

!D7@5:� @9>;E:�

?3;@F3;@76�?3;@F3;@76�?3;@F3;@76�?3;@F3;@76�?3;@F3;@76�?3;@F3;@76�?3;@F3;@76�?3;@F3;@76�?3;@F3;@76�?3;@F3;@76�E:;8F76�E:;8F76�

?3;@F3;@76�?3;@F3;@76�?3;@F3;@76�?3;@F3;@76�E:;8F76�

����� �� ����� ������ ���� ����

�#$ �(*.��#$ �(*.��#$��

(*.��#$ ���,��#$ ���,�

@9>;E:� @9>;E:�!D7@5:� @9>;E:�!D7@5:�

@9>;E:�!D7@5:� @9>;E:�!D7@5:� @9>;E:�

!D7@5:�!D7@5:� @9>;E:�!D7@5:� @9>;E:�

?3;@F3;@76�?3;@F3;@76�?3;@F3;@76�E:;8F76�E:;8F76�

� ��� ��+3DF� �

���

��

��

��

��

��

��

� ��� ��+3DF���

� ���������� ����������

�#$��(*.��#$���#$���#$��(*.�

�#$ ��#$ �(*.�(*.���,��#$��

@9>;E:� @9>;E:� @9>;E:� @9>;E:�!D7@5:�!D7@5:�

@9>;E:�!D7@5:�

��

@9>;E:� @9>;E:�

�!D7@5:�

��

@9>;E:� @9>;E:�

?3;@F3;@76�E:;8F76�E:;8F76�E:;8F76�

?3;@F3;@76�?3;@F3;@76�

� � �� �� ����� ������������������

�#$��!�.�!�.�(*.��#$��!�.�

!�.��#$ ��#$ �!�.�!�.��#$��

!D7@5:� @9>;E:� @9>;E:� @9>;E:� @9>;E:� @9>;E:�

@9>;E:�!D7@5:�!D7@5:� @9>;E:� @9>;E:�!D7@5:�

@9>;E:�!D7@5:�!D7@5:�

��

!D7@5:�

?3;@F3;@76�?3;@F3;@76�?3;@F3;@76���?3;@F3;@76��?3;@F3;@76�E:;8F76�

��

132

� (�,&$)"����, �&�� Line Who

speaks Language Interlocutors

speak in Response

in Switch?

to Topic

���� �� �� ����� ���� ������� �����������

(*.�!�.�(*.�(*.�(*.��#$ ��#$ �

!D7@5:�!D7@5:�!D7@5:�!D7@5:� @9>;E:� @9>;E:�!D7@5:�

@9>;E:� @9>;E:� @9>;E:� @9>;E:�!D7@5:�!D7@5:� @9>;E:�

!D7@5:� @9>;E:�!D7@5:� @9>;E:� @9>;E:�!D7@5:�!D7@5:�

!D7@5:��

!D7@5:��

@9>;E:��

!D7@5:�

E:;8F76�?3;@F3;@76�E:;8F76�E:;8F76�E:;8F76�

?3;@F3;@76�E:;8F76�

����� �� ���������� � �� �� � ���

�#$ ��#$ �(*.��#$���#$ ���,�

!D7@5:�!D7@5:� @9>;E:�!D7@5:�!D7@5:� @9>;E:�

@9>;E:� @9>;E:�!D7@5:� @9>;E:� @9>;E:�!D7@5:�

!D7@5:�!D7@5:�!D7@5:�!D7@5:�!D7@5:� @9>;E:�

!D7@5:�!D7@5:�

�!D7@5:�!D7@5:� @9>;E:�

E:;8F76�E:;8F76�

?3;@F3;@76�E:;8F76�

?3;@F3;@76�E:;8F76�

� ��� ��+3DF� �

���

��

��

��

��

��

��

� ��� ��+3DF���

�������������������

�#$ �(*.�(*.���,��#$��

@9>;E:�!D7@5:� @9>;E:� @9>;E:� @9>;E:�

!D7@5:� @9>;E:�!D7@5:�!D7@5:� @9>;E:�

@9>;E:� @9>;E:�!D7@5:� @9>;E:� @9>;E:�

@9>;E:� @9>;E:�!D7@5:� @9>;E:�

�

E:;8F76�?3;@F3;@76�E:;8F76�

?3;@F3;@76�?3;@F3;@76�

� � �� �� ��������������� � � ����

�#$���#$��(*.��#$ ��#$ �!�.�

@9>;E:� @9>;E:�!D7@5:�!D7@5:�!D7@5:� @9>;E:�

!D7@5:�!D7@5:� @9>;E:� @9>;E:� @9>;E:�!D7@5:�

@9>;E:� @9>;E:�!D7@5:�!D7@5:�!D7@5:� @9>;E:�

� @9>;E:�!D7@5:�!D7@5:�!D7@5:� @9>;E:�

E:;8F76�E:;8F76�E:;8F76�E:;8F76�E:;8F76�

?3;@F3;@76����� *) �$). ,'*�/.*,��� �'�)"/�" ��� Line Who

speaks To whom Language Spoken

in Interlocutor

switches? Topic

���� �� ������������ ���� � �� ���� ���� ���� � �� ������������������������

�#$���#$��(*.��#$ ��#$ ��#$ ��#$���#$���#$��(*.��#$���#$��(*.�(*.�

(*.�(*.��#$ �!�.�!�.�(*.�(*.�!�.�!�.�!�.�(*.�!�.�!�.�!�.�

!D7@5:�!D7@5:�!D7@5:� @9>;E:� @9>;E:�!D7@5:�!D7@5:� @9>;E:� @9>;E:� @9>;E:�!D7@5:� @9>;E:� @9>;E:� @9>;E:�

@9>;E:� @9>;E:� @9>;E:� @9>;E:�!D7@5:� @9>;E:� @9>;E:� @9>;E:� @9>;E:�!D7@5:�

��

!D7@5:�!D7@5:�

�!D7@5:�

��

!D7@5:�!D7@5:�!D7@5:�

�����

@9>;E:� @9>;E:�

E:;8F76�E:;8F76�

?3;@F3;@76�?3;@F3;@76�?3;@F3;@76�?3;@F3;@76�E:;8F76�

?3;@F3;@76�?3;@F3;@76�?3;@F3;@76�E:;8F76�E:;8F76�

?3;@F3;@76�E:;8F76�

����� �� � �� �#$�� (*.� !D7@5:� @9>;E:� !D7@5:� ?3;@F3;@76�

133

������������

�#$���#$���#$��

(*.��#$��(*.�

!D7@5:� @9>;E:�!D7@5:�

@9>;E:�!D7@5:� @9>;E:�

!D7@5:��

!D7@5:�

E:;8F76�E:;8F76�

?3;@F3;@76� � ��� ��

+3DF� ����

��

��

��

��

��

��

� ��� ��+3DF���

�� �������������������� ��

�#$ ��#$��(*.��#$ ��#$ ��#$ ��#$��

(*.���,��#$���#$����,�(*.�(*.�

@9>;E:� @9>;E:� @9>;E:� @9>;E:� @9>;E:�!D7@5:�!D7@5:�

!D7@5:� @9>;E:�!D7@5:� @9>;E:�

� @9>;E:� @9>;E:�

��

@9>;E:� @9>;E:� @9>;E:�!D7@5:�!D7@5:�

?3;@F3;@76�?3;@F3;@76�E:;8F76�

?3;@F3;@76�E:;8F76�

?3;@F3;@76�?3;@F3;@76�

� � �� �� ������������������� ���� ���� �� � ����

�#$ ��#$ ��#$���#$���#$���#$���#$���#$���#$��

!�.�(*.�!�.�(*.�!�.�!�.�(*.�!�.�!�.�

!D7@5:�!D7@5:� @9>;E:� @9>;E:� @9>;E:� @9>;E:�!D7@5:� @9>;E:� @9>;E:�

@9>;E:� @9>;E:�!D7@5:�!D7@5:�!D7@5:� @9>;E:�!D7@5:�!D7@5:�!D7@5:�

�!D7@5:� @9>;E:�

� @9>;E:�

��

@9>;E:� @9>;E:�

?3;@F3;@76�E:;8F76�E:;8F76�

?3;@F3;@76�?3;@F3;@76�?3;@F3;@76�?3;@F3;@76�?3;@F3;@76�?3;@F3;@76�

���� (*.$*)�'�-.�. �� Line Who

speaks Language Interlocutors

speak in Response

in Switch?

to Topic

���� �� ���� ���� ����

�#$��(*.�!�.�

!D7@5:� @9>;E:� @9>;E:�

@9>;E:�!D7@5:�!D7@5:�

!D7@5:� @9>;E:�!D7@5:�

!D7@5:� @9>;E:�

�

?3;@F3;@76�?3;@F3;@76�E:;8F76�

����� �� �������� ���� � �

�#$��(*.��#$���#$��

!D7@5:� @9>;E:�!D7@5:�!D7@5:�

!D7@5:�!D7@5:�!D7@5:�!D7@5:�

!D7@5:� @9>;E:�!D7@5:�!D7@5:�

� @9>;E:�

��

E:;8F76�?3;@F3;@76�E:;8F76�E:;8F76�

� ��� ��+3DF� �

��� ��� ���

�#$���#$��(*.�

!D7@5:�!D7@5:� @9>;E:�

!D7@5:�!D7@5:�!D7@5:�

!D7@5:�!D7@5:�!D7@5:�

��

@9>;E:�

E:;8F76�?3;@F3;@76�?3;@F3;@76�

� ��� ��+3DF���

� �� (*.� !D7@5:� !D7@5:� !D7@5:� � ?3@F3;@76�

� � �� �� � �� ���� ���� �� �

!�.�(*.�(*.��#$ �

@9>;E:�!D7@5:� @9>;E:�!D7@5:�

!D7@5:� @9>;E:�!D7@5:� @9>;E:�

!D7@5:�!D7@5:�!D7@5:� @9>;E:�

�!D7@5:�

��

E:;8F76�?3;@F3;@76�?3;@F3;@76�E:;8F76�

�