• No results found

Functional perspectives on first language development

Functional/pragmatic perspectives on second

5.2 Functional perspectives on first language development

Researchers studying child language have been interested for many years in the meanings that children are trying to convey, the possible relationship between developments in children's messages and developments in the for-mal systems through which they are expressed.Table 5.1 is drawn from one of the best-known 1970s child language studies, already referred to in Chapter 1 (Brown, 1973); here, we see children's two-word utterances being interpreted as expressing a range of semantic relations. For example, in Brown's data the utterance 'Daddy hit' is interpreted not as an expres-sion of the formal syntactic relationship Subject + Verb, but as a combina-tion of semantic categories of 'Agent' (or 'doer') plus 'Accombina-tion'. As the examples show, the child's language at this point is lacking in function

words and overt morphological markers of case, tense, number, etc. This is one key reason why it has been suggested that formal categories devised to describe the mature adult system may not be useful at this developmental stage. Some researchers in this tradition have argued essentially 'that syn-tactic categories develop as prototypes based on semantic information' (Harley, 1995, p. 371). Others who believe that formal syntactic categories have an independent origin have nonetheless accepted that interactions between syntactic, semantic and pragmatic information are vital in driving forward first language acquisition (see survey by Ninio and Snow, 1999).

Budwig (1995) produced*a useful survey of broadly functionalist approaches to the study of child language development. She brought together a wide range of perspectives on the relationship between form and function in child language, and on development in this relationship over time. She has divided them into four main 'orientations' (Budwig, 1995, pp. 3-13): cognitive orientation, textual orientation, social orientation and multifunctional orientation.

Table 5.1 Eleven important early semantic relations and examples

Relation Example Attributive 'big house'

Agent-Action 'Daddy hit' Action-Object 'hit ball' Agent-Object 'Daddy ball' Nominative 'that ball' Demonstrative 'there ball' Recurrence 'more ball' Non-existence 'all-gone ball' Possessive 'Daddy chair' Entity + Locative 'book table' Action + Locative 'go store' Source: Brown, 1973

5.2.1 Cognitive orientation

Cognitive orientation can be exemplified by the work of Slobin (1985), which we have already referred to in Chapter 4. Slobin proposes the existence of a 'basic child grammar5, in which children construct their own form-function relationships to reflect a child's-eye view of the world. For example, Slobin suggests, on the basis of cross-linguistic comparisons regardless of the par-ticular target language that is being acquired, that 'one of the opening wedges

for grammar is the linguistic encoding of a scene in which an agent brings about a change of state in an object (Budwig, 1995, p. 10).

5.2.2 Textual orientation

As far as textual orientation is concerned, 'the issue of central importance is the extent to which particular linguistic devices are employed to help organize stretches of discourse both intrasententially and across broader stretches of text' (Budwig, 1995, p. 11). At the level of discourse, functional linguists are interested in'how both vocabulary and grammar (e.g. connec-tives such as and/but/whereas, deictic elements such as this/that, pronoun systems, etc.) are deployed to create textual cohesion across sequences of clauses and sentences (see Halliday 1985, Chapter 9). In child language studies, functionally oriented research concerned with textual matters has examined topics such as the systems used by older children to establish cohesion in narratives (Karmiloff-Smith, 1987). The following example is drawn from a study of children's gradual acquisition of the different dis-course functions of determiners:

Time 1 C: Isabelle gave a talk about her rabbit and Alexia will give a talk about the tortoise

E: About which tortoise?

C: . . . the tort.. . well, hers, and well . .. not only hers . . . well . .. the tortoises, about all the tortoises

Time 2 E: You remember that Isabelle gave a talk about her rabbit and Alexia gave one about the tortoise?

C: Yes

E: About which tortoise?

C: About the animal, the tortoise (shrugs shoulders as if it were quite obvious)

(Karmiloff-Smith, 1979, pp. 222-3, author's translation from original French) At Time 1, when child C is aged 7 years 9 months, she has difficulty distin-guishing the deictic and generic functions of the definite article; by Time 2, when child C is aged 9 years 2 months, generic functions are used without any difficulty.

5.2.3 Social orientation

Functionalist child language research with a social orientation is interested in relationships between the development of children's formal language system, and aspects of their social world. Some of this work examines the speech acts that children perform, and their relationships with lexical or

grammatical choices (see Ninio and Snow (1999, pp. 353-60) for a recent overview). For example, Deutsch and Budwig (1983) re-analysed some of the data gathered by Brown (1973), arguing that expressions involving first-person possessive determiners (my pencil) consistently expressed different speech acts from expressions involving the child's own name (Adam pencil) - the first group were indicative ('That's my pencil'), whereas the second group were volitional (CI, Adam, want a pencil').

Other work looks much more broadly at the social context within which children interact, and the types of speech events in which they are engaged, and seeks to link these wider influences to linguistic development. A strik-ing example is the work of Ochs (1988), on the acquisition of Samoan, where she argues for a link between children's acquisition of inflectional morphology, and socially patterned variation in adults' usage. The specific example analysed by Ochs concerns the acquisition of ergative case mark-ing.* In Samoan, ergative case marking is optional, and rare in women's domestic talk. Samoan children seem to acquire this feature much later than do children learning other ergative languages, such as Kaluli, for example. This social orientation on child language acquisition is revisited more fully in Chapter 8.

5.2.4 Multifunctional orientation

The functional approaches to child language studies that have been out-lined briefly pay attention, respectively, to the relations between grammat-ical development and prototype events; between grammar, pragmatics and text organization; and between grammar and the social world. Budwig (1995, p. 13) cites the work of Gee or Gerhardt as an example of work on child language that seeks to integrate the study of these different sets of relationships in a multifunctional orientation (Gee and Savasir, 1985;

Gerhardt, 1990). For example, Gerhardt studied the use of the forms will

*'Ergative' languages are those in which 'the subject of an intransitive verb [S] receives the same treatment (morphological and/or syntactic) as the object of a transitive verb [O], while the subject of a transitive verb [A] receives different treatment' (Van Valin, 1992, p. 16).

Take for example, pairs of sentences such as:

1 The boy [A] opened the door [O]

Subject + Transitive verb + Object 2 The door [SJ opened

Subject + Intransitive verb

In an ergative language, O and S (in our two sentences, 'the door') will be marked with the same case ('ergative'), while A ('the boy') will be marked with a different case ('absolutive').

This contrasts with 'accusative' languages (such as Russian, for example), where A and S are marked with the same 'nominative' case, and O is marked with a different case ('accusative').

and gonna by three-year-old children and argues that they are used in dif-ferent discourse contexts, to express difdif-ferent speech acts:

Gonna appears in discourse in which the children were planning and organis-ing; it implies a more distant intention to act in a particular way. In contrast, will appears in the context of ongoing cooperative peer play, and refers to an immediate intentional stance.

(Budwig, 1995, p. 13) In her own longitudinal research, Budwig (1995) examines the self-refer-ence forms (/, me, my, Own Name, etc.) used by a group of two-year-old children to express the semantic notions of agentivity and control, and also seeks to explain variability in usage in terms of the different pragmatic functions that are being expressed. For example, at 20 months, Megan used the three forms I, my and Meggie for self-reference; my was seen as express-ing high agentivity (my open that)) while Meggie expressed mid or low agen-tivity (Meggie swinging) and / was used typically for mental state verbs (/

wanna wear that).There were also differences in usage that could be related to pragmatic function, for example my typically appeared in disputes over control of objects: (my cups! said as Megan grabs cups from another child).

Over time, however, Megan extended the use of / to perform a wider range of functions and her use of my and Own Name became more target-like.

Budwig's 1995 study is typical of recent research on form-function rela-tionships in child language. It has a number of characteristics that are also found in much SLL research in the functional tradition:

• Her data comprises longitudinal case studies of a small number of indi-vidual children; her prime concern is to trace the evolving patterns of relationships between language form and function over time.

• That is, her research is interested in the evolving developmental process, rather than in end states; acquisition is viewed as a slow, incremental business, and researchers are especially interested in the first emer-gence of new forms.

• She is concerned to link different levels of analysis of learner language (e.g. paying attention to intonation as a signal of pragmatic function) and she is concerned to collect data from a variety of social settings, for example peer interactions as well as caretaker-child interactions, in the interest of accessing a wide range of pragmatic functions.

In conclusion, Budwig reviews possible factors that may drive children for-ward to continually reorganize their systems of form-function relationships along the documented developmental path: linguistic maturation; cognitive

development; encounters with target input; and communicative need. As yet, she argues, child language data do not offer definitive support to any one theoretical position: 'the specific mechanisms guiding the reorganiza-tion process are . . . quite vague' (Budwig, 1995, p. 197). We will review below the efforts of functionalist SLL researchers to address the same fun-damental problem.