The acquisition of person and object reference in two-year-olds: morpho- syntactic forms for pragmatic functions from a cross-linguistic perspective

Person and object reference involves both morpho-syntax and pragmatics. In the acquisition of reference, therefore, children are faced with learning diverse skills. Firstly, they have to acquire the relevant morpho-syntactic forms (e.g. nouns, pronouns). Secondly, in pragmatics they need to learn amongst other things to take the listener’s perspective into account. There has been considerable work on the acquisition of morpho-syntactic forms per se (e.g. Chiat 1986). Previous research on the acquisition of reference, however, has mainly focussed on the cognitive-pragmatic aspects, usually using a narrative context (e.g. Hickmann 2003). The main findings show that children at 6 years of age still experience difficulties in estimating the listener’s knowledge; this leads to the use of too many referential devices that presuppose listener’s knowledge (e.g. pronouns, definite NP’s). This line of research has focused, for methodological reasons, on children over 4 years of age since younger children are less able to re-tell stories (Wigglesworth 1997). Hardly anything is known about how children in the early stages i.e between two and three years of age learn to refer or how they do this in spontaneous speech (Tomasello 2003). The present study focuses on this question. It also explores how the acquisition of the morpho-syntactic forms used for reference influences the pragmatic aspects of reference. By comparing the acquisition of reference across languages that are morpho-syntactically different the interaction between morpho-syntax and pragmatics in this area will become clearer. English, Dutch and French are investigated in this study. These languages are related, but some crucial aspects of the referential systems are different, most notably the determiner systems and the way gender / number can be used in pronominal reference. Longitudinal data from three English, three Dutch and three French children between 2;0 and 3;3 were taken from the CHILDES-database (MacWhinney 2000). The data from the different languages were compared on the basis of age. Firstly the timing of acquisition was analysed with respect to referential devices to third person entities in each of the three languages. Subsequently, the pragmatic functions for which the children use these morpho-syntactic forms were categorized. A division of pragmatic functions was made between non-specific and specific reference. Instances of specific reference were further categorized for referent introductions, referent maintenance, referent shift and labelling (or identifying) (following amongst others Hickmann 2003). Each instance was further specified as correct or incorrect. Here we will focus on the pragmatic errors. The results indicate that there are differences between the three languages in the acquisition of the morpho-syntactic devices, most strikingly in the determiner. With respect to the pragmatic functions – all are used in all three languages from 2;0 onwards. The acquisition of form-function combinations shows that there are universal as well as languagespecific influences on nominal and pronominal reference. Firstly, between 2;0 and 2;6 indefinite determiners are primarily used to label referents and to refer to non-specific entities in French and Dutch. In French, the indefinite determiner is also used to introduce referents, while in Dutch this use seems to be acquired a little later, at 2;6. In English, the indefinite determiner is used across all functions; this result is, however, hard to interpret, as in English true uses of indefinite determiners cannot be distinguished easily from fillers. Secondly, in all three languages definite, demonstrative and possessive determiners are used for the specificreferential functions: introduction, maintenance and shift. Differences between the languages are, however, also apparent. For example, the French children also use definite determiners to refer non-specifically as early as 2;0, whereas the Dutch and English children hardly do this at all in the period observed. The French and the English children also use definite and possessive forms more often to identify (or label) referents than the Dutch. Finally, the English children use the genitive noun plus noun e.g. Becky’s doll for introductions, maintenance and shift from 2;0 onwards and far more frequently than the Dutch and French children. For pronominal reference all children were seen to use pronouns to refer to specific entities for introductions, maintenance and shift. There is a difference, however, between demonstrative pronouns and personal pronouns. All children use demonstrative pronouns more often to introduce entities, whereas they use personal pronouns more often for referent maintenance and shift. In the personal pronouns there are differences between the three languages, for example the French and English children make more extensive use of personal pronouns compared to the Dutch preference for demonstratives. From the error analysis it appears that in all three languages the children tend to overgeneralize the definite noun and pronouns since they use these where there is no mutual knowledge. The indefinite forms are used far less but then correctly. These observed similarities and differences will be explored both on the basis of astructural comparison between the three languages, and also from a developmental-pragmaticpoint of view. ReferencesCHIAT, S. (1986). Personal pronouns. In: P. Fletcher, Garman, M. (ed), Language Acquisition.Cambridge: Cambridge University Press, 1986, pp. 339-355. MACWHINNEY, B. (2000). The CHILDES Project: Tools for Analyzing Talk. Hillsdale, N.J.:Lawrence Erlbaum.HICKMANN, M. (2003). Children's Discourse; Person, Space and Time across Languages.Cambridge: Cambridge University Press.WIGGLESWORTH, G. (1997). Children's individual approaches to the organization of narrative.Journal of Child Language, 24: 279-309.