Inflection systems and first language acquisition

Research project P 13681 Inflection systems and first language acquisition Wolfgang U. DRESSLER 28.06.1999 This project aims at investigating, in parallel and in mutual crossfertilization, from the perspectives of universal grammar, typological (crosslinguistic) comparison, and in respect to language-specific system adequacy: A) from the point of view of morphology theory: the inflection systems of 14 project-central languages (with particular emphasis on Austrian German), adding selectively further, typologically heterogeneous languages, B) in psycholinguistic longitudinal studies: the early phase of first language acquisition of inflection systems in (mostly) the same languages. Part A differs from previous research also by insisting on the concepts of grammatical productivity and potentiality, which will be sharply distinguished from normative acceptability, as well as from regularity, default status, morphotactic and morphosemantic transparency. Another decisive difference is also the inclusion of psycholinguistic evidence (primarily through integrating project part B). There is a systematic differentiation between a) the hierarchy of inflection systems starting with rule-derived inflection paradimgs vs. b) the stored morphological similarity relations (in terms of continua). These two different types of inflectional structuring - based on (especially productive) rules vs. lexical storage - will be compared with the corresponding differentiation in prototypical and non-prototypical derivational morphology of 4 languages. Also electronic text corpora will be analysed, and both off-line and on-line tests will be administered. Natural Morphology provides the theoretical framework, which will be compared with the explanatory power of other morphological frameworks. Part B is to continue, in systematic correlation with project part A, the international "Crosslinguistic Project on Pre- and Protomorphology in Language Acquisition" started in 1993 and coordinated by Dressler. Part B will deepen, accelerate and interprete its relevant parts, based on longitudinal spontaneous productions by children up to the fourth year of age, always comparing children`s linguistic input (produced by the caretakers). The theoretical framework of constructivists self-organisation of modularity (also to be systematically compared with other frameworks) will demand rigorous investigation of interindividual variation and especially of blind-alleys in language acquisition. Crosslinguistic comparison will demand weighing unviersal and typological criteria. The integration of both project parts implies, for both of them, an enlargement of problems to be solved, and thus - as can be already seen from preceding work - the prespective of a more complete description and explanation (both in number of languages as in detailed accounting of each of them). This should provide new standards for theoretical discussion and provide adequate material for a follow-up project which should contrast "normal" with "handicapped, disturbed" and delayed language acquistion.

The project "Inflection systems and first language acquisition" funded by the FWF has aimed to examine a) the inflection systems of 15 different languages (German, French, Italian, Spanish, Russian, Polish, Bulgarian, Georgian, Lithuanian, Latin, Modern Greek, Finnish, Hungarian, Hebrew, and the Sepecides` Romany dialect) and b) the early phases of children`s first language acquisition in about the same languages. The most important result of project part A was the exact assignment of verbs and nouns to inflectional classes according to the criteria of productivity (important in Naturalness Theory) and default inheritance. For the individual languages, results were presented in several publications one being comparative (at the International Morphology Meeting 2004). We expect changes in practical grammars to be based on our research. In project part B ("First language acquisition of morphology"), longitudinal studies of Viennese children (Jan, aged 1;3 - 6;0, Lena, aged 1;4 - 4;3) were carried out. The children were tape-recorded in everyday situations in interaction with their mothers. Transliteration and coding were done according to the standards of the international child language database CHILDES. Especially the acquisition of verb inflection as well as the acquisition of noun diminutives and plurals was examined - always in comparison with the maternal input. Also the results of other languages (especially French, Italian, Spanish, Croatian, Russian, Lithuanian, Greek, Hebrew, Arabic, Finnish, Turkish, Yucateco Maya) were compared. Delimitations between the "premorphological" phase (when children use precursors of morphological operations e. g. reduplications such as "wow-wow") and the "protomorphological" phase (when children start to use morphology in a creative way by making e. g. first plural overgeneralizations such as German "Jeepen" instead of "Jeeps") were stated more precisely; common characteristics and differences were documented. The phase "morphology proper" (following "protomorphology") was introduced: it is the time when the child`s inflectional system comes closer and closer to the adult`s system. The child Jan reached protomorphology at age 1;8 and morphology proper at 2;5. But there are considerable differences 1) between early and late talkers acquiring the same language and 2) between children acquiring different languages: Turkish-speaking children learn the uniform agglutinating morphology of Turkish earlier than German-speaking children German morphology, as it confronts them with a great number of ambiguous inflectional endings. These results are an important basis for future experimental studies and for studies on delayed and disturbed language acquisition.