Lecture 4 Social-Interactionist Approaches to Language

Document Sample
Lecture 4 Social-Interactionist Approaches to Language Powered By Docstoc
					Lecture 4: Social-Interactionist Approaches to Language Acquisition Language as a social-cognitive system v Language as an abstract symbolic system 1. The structure of human language has arisen due to the social-communicative functions language serves in human relations 2. The requisite experience for language acquisition is social interaction with other speakers Language development can only be properly understood by situating it within the communicative context in which it occurs 1. Continuity between pre-verbal and verbal communication: Extent to which linguistic developments such as word learning, and the acquisition of sentence and discourse structure are built upon pre-verbal communicative abilities Bruner (1975) points out that: a) Infants are able to engage in episodes of joint attention with their caregivers before word learning begins b) Pre-linguistic communication has a topic comment structure analogous to that of language onto which syntactic structure can be grafted 2. Environmental Support: Extent to which language is acquired in a communicative context that facilitates learning Bruner (1975) argues that the involvement of the child in interaction provides her with a Language Acquisition Support System (LASS) Child is seen as an apprentice in the acquisition of communicative competence Abilities are initially possessed by the dyad/infant-mother system and only later by the individual (Vygotsky, 1962) 1. Turn-taking: Young infants respond differently to animate and inanimate stimuli and show alternation of vocal behaviour with their caregivers well before recognisable productive language (Trevarthen, 1975) By the second year dyads show smooth turn-taking with few overlaps between speakers (Schaffer et al., 1977). Bruner (1975) claims that such interchanges can be seen as ‘protoconversations’ (i.e. non-verbal dialogues with a conversation-like structure) Schaffer et al. (1977) analyse early face-to-face interaction Early infant behaviour shows a rhythmic pattern in which patterns of intense excitement alternate with patterns of relative calm. Caregiver thus able to create a conversation-like exchange by weaving behaviour around the child’s natural activity patterns Initially caregiver alone is responsible for smoothness of face-to-face interaction. Speakerswitch pauses are initially much longer from caregiver to child than from child to caregiver


Kaye (1981) traces origins of turn-taking back to feeding situation Infants emit rhythmic bursts of sucking followed by pauses of indeterminate length that feeders tend to interpret as signals to ‘jiggle’ the baby ‘Jiggling’ actually suppresses sucking. However, this means that there is an increase in the chances of the child starting to suck again when jiggling stops creating a subjective contingency between ‘jiggling’ and sucking for the feeder Feeder is thus ‘conned’ by nature into ‘thinking’ that jiggling is effective and hence treating the child as an interactive partner from the beginning Joint attention: Starting around 4 months and with high frequency by 9 months, infant able to follow adult's line of regard (Scaife & Bruner, 1975) Collis & Schaffer (1975) showed that joint attention develops gradually and is initially maintained by constant parental monitoring of infant's line of regard Tomasello & Farrar (1986) showed that children’s rate of early vocabulary development was related to mothers’ skill in talking about the child’s current focus of attention Intersubjectivity and Cultural Learning: Tomasello (1992) argues that language learning is a particular kind of imitative learning which involves imitating both the words or phrases used by the adult model and the communicative purpose to which that word or phrase is put Snow describes her son Nathaniel using the phrase: “That’s a good idea” in the following sequence: ‘Mummy. Go outside now. That’s a good idea” in an attempt to persuade her to let him go outside and play rather than go to bed Topic-comment structure: Before 8-9 months infants tend to engage in manipulative play with objects or face to face interaction with people. However, ability to monitor and manage gaze direction means that by 8-9 months child is able to co-ordinate these skills and interact with adult over objects Children at this age will typically look up at critical junctures in their play to make eye contact with adult and hence share their emotional response to the object. Bruner (1975) argues that these kinds of interactions have a topic-comment-like structure onto which syntactic structure can be grafted Reciprocal role-taking: Participation of child in conventionalised games (e.g. give-andtake, peekaboo) demonstrates ability to reverse roles in rule-bound interactive sequences (Bruner, 1975) Ability to take different perspectives on the same event is a necessary prerequisite for the development of interpersonal deixis (I/You distinction). Such games also provide routinised formats which scaffold the child’s acquisition and use of deictic language Anne 8-10 months Learns to play exchange game


13 months 13.5 months 14 months 15 months

Picks up M’s ‘Thank you’ but uses it when giving and receiving Uses ‘Thank you’ only when receiving Uses ‘Look’ when giving ‘Look’ replaced by ‘There’

Motherese/Child-Directed Speech: Parents tend to talk to children using slower, shorter, more stressed utterances, delivered in higher pitch and containing more part or whole repetitions (Snow, 1972) CDS tend to be tied to the here and now and to contain relatively few complex sentences, false starts, hesitations or grammatical errors Bruner sees CDS as a teaching language that paces input to child's linguistic level and hence supports mastery of linguistic structure. Mothers are providing children with graded language lessons, ‘upping the ante’ each time the child makes a step forward Advantages of Social Interactionist Approaches 1. There is more to language-learning than grammar learning and social-interactionist studies are a rich source of information about human communicative abilities and how they develop Even nativist models of syntactic development assume that children acquire rules by working backwards from contextually derived interpretations of the sentences that they hear. Social-Interactionist approaches can thus be seen as a necessary complement to nativist linguistic approaches to language learning 2. Focus on the language learning environment is valuable if only because it tells us what the input to the language learning child is really like Some nativist theorists (e.g. Radford, 1990) make much of the fact that English-speaking children initially use ‘Mummy’ and their own name in the early stages of grammatical development rather than ‘I’ and ‘You’. However, one obvious explanation is that English-speaking mothers model this kind of usage in their input (presumably because they feel that proper names are easier for their children to understand than deictic pronouns) 3. By focussing on communicative functions of language rather than the complexity of the language system itself, social interactionist approaches put language learning back into a developmental context. In some cases this may obviate or at least reduce the need to appeal to innate linguistic knowledge. Tomasello (1992) points out that the problem of the indeterminacy of meaning results from the way word learning is seen as the process of pairing words with their referents in the environment. If reference is conceptualised as a triadic rather than a dyadic process, word learning becomes a process of inferring the meaning of particular symbols on the basis of one’s understanding of speaker meaning


Tomasello & Kruger (1992) show that children learn many early verbs in contexts where event to which verb refers has yet to occur M: Eat your chips. M: Why don’t you kick it? Hard to explain if you assume that children acquire word meanings by pairing words with objects and events in the environment. Easy to explain if you assume that children acquire word meanings by decoding speakers’ communicative intentions and then using this information to infer what the words themselves mean (e.g. “It’s obvious what she wants me to do with these chips so ‘eat’ must mean eat”) Disadvantages of Social Interactionist Approaches 1. Although many of phenomena identified look like necessary prerequisites for development of a communicative system, not clear that they explain much about the development of the linguistic system itself Turn-taking may be an important conversational skill, but it tells us nothing about the growing sophistication of the child’s contributions to the conversations in which she is participating 2. Even when there is a relation between communicative structure and linguistic structure, relation is typically very indirect Although topic-comment structure is correlated with syntactic structure (e.g. topics more likely to be encoded by subject than direct object), this correlation is by no means perfect M: What fell? C: The truck fell M: What are you doing? C: I dancing 3. Idealises Mother-Infant system How high is facilitating language acquisition on most parents’ list of priorities? How are parents supposed to know what to do in order to facilitate language acquisition? Newport, Gleitman & Gleitman (1977) point out that, given the opacity of grammatical relations, notion that mothers ‘know’ how best to teach syntax replaces a strong nativist claim about the child with a strong nativist claim about the mother. 4. Are caregivers attempting to facilitate acquisition or to facilitate interaction? a) Cross (1977) showed that the simplicity of CDS was tuned not to child’s current level of productive language, but to child’s current comprehension ability b) Newport et al. (1977) point out that CDS is in some ways more complex than adultdirected speech. Declaratives (e.g. You kicked the ball) less common and Questions (e.g. ‘What did you do?’, ‘Did you kick the ball?’) more common Durkin (1987) points out that


inconsistent substitution of proper names for deictic pronouns makes input more confusing (e.g. ‘Mummy’s just having a cup of coffee, aren’t I?’) 5. Many cultures do not subscribe to Western Middle Class child-centred ideology Ochs (1985) Samoans reject notion of 'child-centredness' believing that low-status individuals should adapt to high-status individuals. Hence no speech modifications (at least of child-centred variety) Schieffelin (1985) Kaluli reject infant vocalisations as 'bird talk' from which child must be weaned. Adopt explicit teaching style involving use of elicited imitation Heath (1983) Trackton mothers see no point in talking to infants. No CDS (though child is immersed in socially meaningful situations) 6. Facilitation is only meaningful if one has a model of the process that is being facilitated Environment may support dyadic behaviour but how are the skills possessed by the motherinfant system internalised? C’s ability to share attention may reduce need for innate constraints on word learning, but how does child use understanding of mothers’ intentions to infer meaning of words? Simplicity of CDS may make it easier for child to interpret mothers’ utterances, but how does child extract the grammatical rules that underlie them?


Shared By:
Description: Lecture 4 Social-Interactionist Approaches to Language