Symposium
Chair: Edith L. Bavin
Discussant: Ann Peters
The papers in the symposium aim to extend our knowledge about different strategies used by infants in the early stages of acquiring a language. The objective is to show that while infants may follow different paths in their gradual acquisition and mastery of structures or a grammatical category in the language, similar strategies are detectable across languages. Three of the four papers present longitudinal data from infants acquiring languages other than English. Such data is crucial in furthering our knowledge about how infants break into the language system and the factors that might influence their early language development.Understanding how infants segment the language they hear spoken in their social environment and how they determine the distribution for the grammatical markers of their language is of theoretical importance. A young child's use of a linguistic form is not evidence for productivity; productivity requires a child to draw upon knowledge of the language system to construct new utterances. For children to show that they have mastered patterns for combining words and grammatical markers in their language there must be evidence that they can form multi-word structures that they have not already heard, and that they can use the grammatical markers of their language in appropriate contexts. In order to reach this stage, there must be analysis of the input language on the part of the child. Since Brown's longitudinal study of three children acquiring American English, it has been clear that age is not the best predictor of grammatical development. In addition, several researchers (e.g., Bates, Lieven, Nelson, Peters) have argued that infants may follow different paths to productivity, both in the segmentation of the input language and in the combining and use of linguistic forms. In the first paper of the symposium, longitudinal data from an infant acquiring English and another acquiring German will be discussed in relation to whether the phrases first used by the two children represent partial imitations of structures heard in the input. Structural differences across the two languages provide a crucial comparison. Drawing on longitudinal data from three children acquiring Spanish, the second paper discusses the gradual development of a specific grammatical category, grammatical person with some variation across the children. The third paper illustrates how the forms of grammatical markers develop; children acquiring Swedish pass through three phases from segmenting the grammatical markers of their language to using them appropriately, with some differences across the three children studied. The fourth paper, drawing on longitudinal data from three infants acquiring Australian English, illustrates three paths to productivity. The discussant will draw out similarities across the strategies presented, focussing on whether cues in the input facilitate their use. The predicted outcome is a better understanding of how infants convert their input language into a productive language system.
Details of individual items:
paper
This paper is concerned with the ways in which children segment the speech that they hear and use these segmented chunks to start producing structured utterances of their own. When children first start producing structured multi-word utterances, they often 'omit' many of the non-content words and morphemes which would be present in an adult utterance with the same meaning, giving rise to so-called 'telegraphic speech' e.g. 'Daddy going' instead of 'Daddy's going'. There are a range of different theories which attempt to account for this phenomenon, including a number which suggest that children begin by picking up the words which are most salient, either in terms of their position (e.g. utterance-final) or in terms of stress. For instance the prosodic spotlight hypothesis (Peters, Stromqvist) suggests that the stress patterns in the speech that children hear tend to occur on content words and this results in children registering and producing these words before they produce the 'function' words that often come between them in adult speech. A new and contrasting hypothesis is that 'ungrammatical' utterances reflect the actual surface order patterns of the speech that the children hear. Thus when a child hears a question such as: 'Is Daddy going', s/he may initially segment off 'Daddy going' as a possible string in the adult language. Although these two hypotheses are not mutually exclusive, they should both be tested in terms of their implications for languages other than English, the language on which they were first developed and which, in terms of both its stress and word order patterns, appears to fit both hypotheses with at least initial plausibility. German is a language in which discontinuous but systematic word order with respect to auxiliaries and the main verb plays a central grammatical role. Thus the infinitive in German must always come at the end of the utterance ('Soll Mama das trinken?' 3D 'Should Mum this drink'/Should Mum drink this?') and material is frequently placed between the auxiliary and the main verb. This therefore contrasts interestingly with English in which discontinuous word-order dependencies are less central in early development.In this paper we report on data from a study of one English-speaking and one German-speaking child's development from the age of 2;0. Each child was tape recorded for five hours per week (one video and four audio-recordings) and in addition the children's mothers kept a daily diary of innovative utterances. The size of the database, its consistency and the close relatedness of English and German allow us to pinpoint and test this surface-order hypothesis in detail. All the grammatical and ungrammatical progressive utterances in the English child's corpus and all the grammatical and ungrammatical utterances-final infinitives in the German child's corpus were analysed. We discuss the relative patterning of the elements omitted in the children's utterances in relation to the patterning in each mother's speech.
paper
This talk presents data from three children acquiring Spanish as their native language. It specifically addresses what these children know about linguistic PERSON. Some accounts of the development of inflectional languages claim that children learning inflectional languages like Spanish and Italian have PERSON very early. These data challenge that claim. Two opposing positions can be found in the literature on children's early grammatical development. On the one hand, some researchers have argued that children's earliest steps in development are piecemeal, and that children's acquisition of grammatical constructs is a long, drawn-out process (e.g., Pine & Lieven 1993, 1997, Tomasello 1992). This includes developments in inflectional languages (Pizzuto & Caselli 1992, 1993, 1994). On the other hand, others have argued that children come equipped with innate knowledge that allows them to acquire grammatical forms accurately and quickly. Among those elements that are purported to be learned early and rapidly is PERSON (Hyams 1992, Grinstead in press).This talk examines data from 3 children learning Spanish to determine whether their acquisition of PERSON takes place early and quickly. Data from one child cover the period between 1;8 and 2;1, and from the other two children the period between 1;6 and 2;6. All verbal constructs produced by the three children were examined for (1) uses of PERSON marking on the inflectional suffixes of verbs, (2) uses of overt noun phrases and pronominal subjects, (3) uses of noun phrase objects, and (4) uses of PERSON in pronominal/clitic objects. The data from these three children indicate that PERSON emerges only gradually and in a piecemeal fashion in Spanish, and through parallel developments across a number of domains. One of the subjects only begins to show a differentiation of person in the verb forms by the end of the study (just after age two); the others seem to develop person through the confluence of three routes: (1) PERSON marking in inflectional affixes on verbs seeps in very gradually, on a verb-by-verb basis, and with no consistent ordering across tenses (Gathercole, SebastiE1n, & Soto 1999); (2) the use of object clitics also emerges only gradually, verb by verb; and (3) object clitics emerge only with the development of full object noun phrases, also a gradual, verb-by-verb process. The data from these three children are inconsistent with the nativist stance claiming the sudden, early, and accurate use of PERSON; they support, instead, the position that linguistic PERSON emerges through a gradual convergence of related constructs in the child's linguistic repertoire.REFERENCESGrinstead, J. (in press.) Tense, number and nominative case assignment in child Catalan and Spanish. Journal of Child Language.Hyams, N. M. (1992.) Morphosyntactic development in Italian and its relevance to parameter-setting models: Comments on the paper by Pizzuto and Caselli. Journal of Child Language, 19, 695-709.Pine, J. M., & Lieven, E. V. M. (1993.) Reanalysing rote-learned phrases: Individual differences in the transition to multi-word speech. Journal of Child Language, 20, 551-571.Pine, J. M., & Lieven, E. V. M. (1997.) Slot and frame patterns and the development of the determiner category. Applied Psycholinguistics, 18, 123-138.Pizzuto, E., & Caselli, M. C. (1992.) The acquisition of Italian morphology: Implications for models of language development. Journal of Child Language, 19, 491-557.Pizzuto, E., & Caselli, M. C. (1993.) The acquisition of Italian morphology: A reply to Hyams. Journal of Child Language, 20, 707-712.Pizzuto, E., & Caselli, M. C. (1994.) The acquisition of Italian verb morphology in a cross-linguistic perspective. In Y. Levy (Ed.), Other children, other languages: Issues in the theory of language acquisition (pp. 137-187). Hillsdale, N.J.: Erlbaum. Tomasello, M. (1992.) First verbs: A case study of early grammatical development. Cambridge: Cambridge University Press.
paper
no abstract
paper
Early developmental differences across infants' language may be attributed to initial segmentation strategies with some children segmenting individual words and others larger chunks, which must then be analysed into smaller linguistic units, that is, words and grammatical markers. Infants' grammatical knowledge is often assumed to develop only after their vocabulary level is large enough for words to be categorized. For example, Radford (1990) argues that infants acquiring English develop category knowledge about words at around the age of 20 months and start to develop other grammatical (syntactic) knowledge at around the age of 24 months. However establishing evidence for grammatical knowledge is problematic since forms may be restricted in use to a limited set of constructions when a child first uses them. This paper illustrates three strategies for segmenting and combining language units into longer utterances on the path to productivity. Data is from a longitudinal study of three children acquiring Australian English. The children were videotaped in their home environments over a two year period, one starting at age 16 months and two starting at age 20 months. The data from the three infants in the first year of data collection reveal three styles: the use of clearly articulated chunks taken from the input, which are later juxtaposed to create new utterances; the use of partially unintelligible chunks with fillers in grammatical morpheme slots; and the use of single words which are later joined into longer utterances on the basis of a limited set of frames and slots. Possible explanations for the differences observed across the three children are discussed in relation to structural and functional features of the adult language used in interactions with the children during the early taping sessions.