Language Socialization in Bilingual Families: Functional Distribution of Languages in Baby-directed Talk

This paper analyzes language socialization practices in relation to revitalization of Kazakh and maintenance of Russianin urban bilingual families. It draws on audio data of self-recorded baby-directed talk. The analysis reveals that language practices re-produce the social order in which Russian is a more valuable and powerful linguistic resource than Kazakh. This implicit ideology of privileging Russian may explain how and why youngchildren in urban Kazakh families are turning out to be Russianspeaking despite quantitatively dominant input in Kazakh.


Introduction
This paper analyzes language socialization practices in relation to Kazakh revitalization and maintenance of Russian by analyzing baby talk. Baby talk (also referred to as childdirected talk, infant-directed talk, caregiver language, 'motherese', 'parentese') is a specific register of language directed to young children (infants and toddlers) in their preverbal stage [1][2][3].
Baby talk was not something I intended to study when I began my fieldwork. During initial observations, my primary focus was on the interaction between adults and older children so little attention was paid to adult-baby interactions. Only later, when I started transcribing audio-recordings and analyzing my notes, I became aware of babies' 'omnipresence'. Since babies were confined to home, most recordings included them as participants. More importantly, there was a notable difference in bilingual adult's language choices when addressing babies and the older kids. Babies were not only more likely to be addressed in Kazakh but as children they were also a principle target for talking Kazakh. Naturally, this observation raised a number of questions: -What is the use of Kazakh and Russian with and around babies? -What types of baby-centered activities can be identified? -What is the use of languages for these activities? How often and how consistently? -What is the pattern of code-switching in baby-directed talk? -What language ideologies contribute to maintenance of Russian?
The analysis of child-adult interaction is based on the approach developed within language socialization tradition (see [4] for program paper and [5] for the edited volume of current research) that has examined child-directed talk "within its socio-cultural and historical context, as a practice influenced by a community's ideologies concerning care giving, childhood, development, learning, competence, and communication" [3. P. 552]. Language socialization approach recognizes that cultural reproduction and social change is a dynamic process of a novice becoming an expert through participation in recurrent communicative practices and interaction with more knowledgeable practitioners. Central for language socialization enquiry is a focus on language and other semiotic resources as both a means and endpoint of the socialization process [6][7][8][9].
This dynamic view of language recognizes that language is not only "a tool to express, represents, or transmit ideas from one speaker to another", but also a powerful tool of socialization and "production of experience" [10. P. 295]: Viewing language as a socially organized system allows for the recognition that meaning it self can come to be determined based on regular co-occurrences of forms and functions. Specific linguistic forms come to be regularly and conventionally associated with particular situational features in such a way that when the form is used it evokes the larger contextual frame.
Ochs and Schieffelin argue that such (an) indexicality based model allows the researchers to explain not only why children acquire particular linguistic forms but also why they do not acquire the form despite the presence of the form in the input from experts. "Children's nonuse of grammatical forms may be a reflection of their indexical sensitivities and not a reflection of their lack of grammatical competence or awareness" [7. P. 169]. Analysis of language socialization practices also helps to explain why and how children may not be learning the languages of their bilingual or multilingual communities. In multilingual communities, different languages may be valued differently; adults' practices of language choice and code-switching may signal the values attached to each language. Language socialization studies of bilingual and multilingual communities show that coexistence of two or more languages: is rarely neutral or unproblematic state of affairs, it tends to be a focal point of cultural elaboration and social conflict with complex linkages to other, equally contested issues. Language differences (either real or perceived) may map onto and index, or may be used to constitute and reinforce, the boundaries of other social categories and divisions based on such notions as ethnicity, nationality, race, class, gender, religiosity, and generation [9. P. 350].
Through participation in social activities involving code selection children acquire values associated with each code and this cultural knowledge impacts the acquisition of codes [7. P. 188]. In this paper I will show how knowledge of Kazakh and Russian are transmitted to children along with their social values in such a way that leads to a language shift to Russian.

Findings
The chapter draws on 20 hours of self-recorded audio data of natural interaction between adults and children at home. All interactional audio data were transcribed broadly. After that, I have selected episodes of baby-directed talk containing code-switching between Kazakh and Russian (around 200 minutes). Finally, I transcribed in greater detail and analyzed episodes of talk in the most frequently occurring child-centered activities. For analytical purposes, I divided activities into three broad categories: nurturing (e.g., feeding, changing, soothing, and minding -merely being present to ensure the child's safety), pedagogizing (e.g., labeling, prompting and questioning) and entertaining (e.g., play, games, singing songs, listening to music, playing on game consoles and watching TV). Total time of these episodes is 34 minutes 45 seconds. Play activities are the most frequently occurring material in the data and the selection below reflects that fact. There are 12:55 minutes of play activities. This is followed by feeding activities (10:20 minutes), minding activities (6:35 minutes), and dressing/changing activities (6:00 minutes).
The quantitative data clearly points to differences between child-directed talk in Kazakh and Russian. While caretakers use more Kazakh than Russian when speaking with the infant, the bulk of the things they say in Kazakh are affectives and interrogatives. Together they make 55,3 percent of the tone units said in Kazakh while only 12,6 percent of the Russian sample are endearments and questions. In contrast, half of the utterances in Russian are directives in comparison to only 9 percent of the Kazakh sample. The example below shows contrastive use of Kazakh and Russian for play and real work tasks. This interaction starts with Grandfather expressing his endearment while hugging and kissing his grandchild (lines 1-2). Then Grandfather invites the child to play with him the pointing and labeling game by enquiring what the child wants "ne kerek?" ('what do you want') in lines 4-6. However, as soon as he lets the baby move off from his lap (line 7), the child either grabs or tries to reach for some object (line 8) which the adult finds unsuitable for boy. He tells the child that he does not need it 'you don't need this, it's kaka' (lines 9-12), gets up to put it away (line 13-15), and returns to asking what the child wants (lines 16-17).
( Code-switching from Kazakh to Russian marks a shift in footing from play to serious authority-laden activity. It happens at the crucial moment when the caretaker prevents a potentially risky situation by stopping the child 'you don't need this' (line 9); downshifting in volume accompanying shift to Russian also seem to mark the seriousness of the situation. When the problem is resolved, the granddad re-invites the child to play. At the same time he shifts back to the baby talk register, switches to Kazakh, and speaking at a normal volume (line 16). Such contrastive use of codes used to accomplish different functions (play vs. real adult work) signals the seriousness of the speaker's intent. I would also like to point to endearments "ah-ah-ah" in line 7 and "zhanym menim akem menim" ('my dear, my father') in line 16. These Kazakh endearments appear at sequence boundaries, while the main/serious talk in between is in Russian.
This exchange demonstrates several points. Firstly, it gives an example of the contrastive use of Russian and Kazakh for different activity types. Kazakh co-occurring with other 'baby talk' features is reserved for play activity while Russian is used for 'non-play' real work. Phonological shift and code-switching signal the moments of shift in interaction from one activity to other. Secondly, the sequence is yet another illustration of the division of labor between Kazakh and Russian as discovered by the quantitative analysis. As in the previous example, affectives like 'where is my golden?' and interrogatives 'what do you want?' are in Kazakh, while warning 'it's kaka', directing 'look up there' and informing 'we will put it (here)' are in Russian. Thirdly, the exchange further exemplifies the cooccurrence of Kazakh with other features of the baby talk register such as phonological modifications, repetitions, use of stereotypical endearments and rhetorical questions. The frequent use of the same formulaic expressions in Kazakh leads to a situation when the input in Russian the baby receives from his grandparents is much richer lexically and grammatically. Fourthly, the structure of this and the previous interaction sequences is LANGUAGES [11. P. 20] argues that use of turn-passing devices such as questions are "directed towards keeping the conversation going". She continues by stating that because very often these turnpassing units are not followed by any behavior which could be interpreted as communicative, the speakers are "forced into conversational repair procedures such as repetition or taking the baby's turn" [P. 20]. All in all, talk in Kazakh is a pretend conversation and quite often it sounds as a self-talk. It resonates well with the first point: Kazakh and Russian are used contrastively for real and play activities, for real and pretend talk.
Thus through repeated practices using Russian for real, serious and authority-laden activities and Kazakh for pretend, play and scripted activities, and for marking talk sequence boundaries, the adults seem to re-produce the social order in which Russian is a more valuable and powerful linguistic resource than Kazakh. This implicit ideology of privileging Russian prepares grounds for eventual shift to Russian as the child grows older. There are many ways the caregivers systematically construct the infant as Russian-speaking in their everyday language practices. They habitually insert focal words in Russian in what seems to be a talk in Kazakh, conduct verbal tutoring in Russian, and teach Kazakh through Russian. In the end, all these practices ensure that the child grows up speaking Russian as a dominant language.
The table below shows all the nouns used by the grandparents in the sample. The number in parenthesis shows the number of the noun tokens in the sample of babydirected talk. Table 2 List of nouns and their frequency in baby-directed talk 1

Kazakh nouns
Russian nouns Special baby talk terms Bivalent nouns ӑkem 'my father ' (24) mashina 'car' (4) bibi (24) ata 'grandfather' (23) zhanym 'my dear ' (33) miachik 'little-ball' (4) kaka (4) mama 'mom' (1) qulynym 'my fawn ' (24) palchik 'little-finger' (3) papa 'dad' (2) balapanym 'my chick' (6) nozhki 'little-feet' (1) azhe 'grandmother' (1) botam 'my baby camel' (1) molodets 'good boy' (2) noski 'socks' (6) kŏkem 'my dear' (3) vermishel' 'vermicelli' (1) televizor 'TV set' (5) tamaq 'food' (1) lozhka 'spoon' (2) radio 'radio' (3) ajaq 'foot' (9) miaso 'meat' (1) auyz 'mouth' (3) glaz The data summarized in the table shows that the number of tokens of Kazakh nouns is much higher than the occurrence of Russian nouns (104 vs. 24). However, a close inspection reveals that the list of Kazakh nouns consists mostly of nouns used as apart of endearment formulas. The first six nouns appear in the caregivers' talk only in a possessive form. That leaves only three last nouns on the list that were used to label the objects around the child in Kazakh, or 13 tokens. On the other hand, the list of Russian nouns is four times longer (3 Kazakh nouns, which are not part of endearment formulas, vs. 12 Russian nouns) and their number of tokens is almost twice higher (13 tokens vs. 24 tokens). The quantitative data clearly shows that the adults tend to label the objects around the baby in Russian. By doing so, they systematically enrich the child's vocabulary with the Russian lexical items while the child receives little input of the new Kazakh lexicon. The existence of such language practices implies that the adult projects the child as Russian-speaking.
The strongest evidence that the adults see the baby as Russian-speaking comes from observing caregivers as they engage the child in the verbal game of labeling, in Kazakh. Grandfather was pointing at the body parts and naming them in Kazakh: "ajaq" ('foot'), "bas" ('head'), "kŏz" ('eye'), "qulaq" ('ear'). However, then he prompted the child to repeat by saying "Skazhi ajaq" ('Say foot') which reveals that the adult perceives Kazakh only as a second language for the child. Thus, this attempt to teach the child some Kazakh words only underscores the fact that the caregiver regards the child as Russian-speaking who needs to be taught Kazakh through Russian.

Conclusion
The paper demonstrates that a study of caregiver's language socialization practices could provide us with insights into the way ideologies, beliefs, values, normative expectations, and indexical meanings organize the process of a young child becoming a competent member of society. Our data shows how in the process of language socialization children are socialized to linguistic order which privileges Russian and devaluates Kazakh. While Kazakh is preferred for expressing emotions and linked to ritualized, highly scripted activities such as plays and traditional verbal games, it is Russian that caretakers rely on when deal with serious tasks at hand and more crucial, real-life speech activities. Consequently utterances in Kazakh are more likely to convey old or no information in comparison to utterances in Russian, and input on Russian is lexically richer than input in Kazakh. At the same time, language practices seems to suggest that adultsproject the child as Russian-speaking. Adults systematically engage children in explicit teaching activities bylabeling the world around in Russian while equivalent activities in Kazakh are notably absent. To conclude, language socialization of pre-verbal children merits our serious attention as it shapes children's language development and it is consequential for language revival.