In this concluding chapter we consider our results in the larger frameworks of multilingualism and language change and show that these areas cohere. We consult an ongoing program of study of the acquisition of multilingualism and report results documenting both structural constraint on the multilingual course of acquisition of relativization and the integration of principles of language-specific grammar. The results reported in this book reflect general properties of language acquisition and are not limited to the child or to monolingual contexts alone. We relate our results regarding the nature of language change in acquisition to that of language change in historical contexts and argue that a gradual integration of linguistic knowledge culminates in the creation of a linguistic system over time in both contexts. Finally, we consult a more refined notion of “recursion,” and in keeping with recent theoretical developments in the implementation of recursion in Universal Grammar, we reject a claim that the developmental results in this book reflect an absence of recursion in early child language. We describe the implications of our results on the acquisition of relativization for our understanding of the Language Faculty in the human species.

In this chapter, we begin to refine a paradigm for the study of language acquisition which addresses the apparent paradox: evidence points to both continuous universal grammatical knowledge and development of language over time. We contrast our paradigm, “Grammatical Mapping,” to other models existent in the literature, briefly reviewing empirical data from studies supporting them. We argue that these data are in fact consistent with a paradigm of Grammatical Mapping and converge with results from our study of relativization across English, French, and Tulu. These results refute simple versions of both neural-based and construction-based approaches to the acquisition of relativization.

In this chapter we introduce a proposal from the linguistic theory of relatives, one which minimally underlies all forms of relative clause formation (Kayne 1994 and subsequent work). We adopt a theoretical paradigm which moves from the study of specific relative clause structures to general principles of relativization which underlie all relatives all typological variants. We formulate the leading hypothesis that headless relativization is a developmental primitive in terms of this paradigm. This representation of Universal Grammar (UG) offers a template which relates Determiner Phrases and Complementizer Phrases, which in turn allows the acquisition of relative clause in specific language grammars.

In this chapter, moving beyond observational studies of natural speech alone, we report the systematic design, method, and results of an experiment in English concerning our hypotheses that acquisition of relativization in a language-specific grammar is developmentally achieved through constraint by a UG template relating Determiner Phrases and Complementizer Phrases. Headless relatives are developmental precursors for headed relatives over the course of development of a specific language grammar. This experiment provides a prototype for the experimental work in the following two chapters, which focus on French and Tulu. Comparisons of production and comprehension in this English study and experimental design both allow this study to address the effects of syntax and semantics and their integration in the developmental course of relativization. Growth curve analyses across these tasks allow us to refine our understanding of the nature and timing of their integration in development.

In this chapter we present the complexity of the task of acquisition of relativization by providing a typological survey revealing the basic dimensions of variation in relative clause formation across languages. The presence or absence of headedness is one fundamental distinction in this cross-linguistic variation. Each and every child must be able to acquire any one or more of the relative clause variants.

In this chapter, a first cross-linguistic test of our leading hypothesis is reported. This chapter presents the design and results of an experiment testing the production of relativization in French, including both headed and headless forms, by children in Belgium. Consistent experimental design and methods allow comparison across French and English. Analyses of the grammatical variation in relativization across English and French allowed investigation of the interaction of language-specific factors in this next test of our leading hypothesis. Results are consistent with the hypothesized developmental primacy of headless relativization. They also reveal the nature of children’s mapping between principles of Universal Grammar and specific language grammar (Grammatical Mapping). The concept of ‘grammatical transparency’ is introduced.

In this chapter we introduce our research paradigm, relating it to a general theory of a Language Faculty in the human species, and anticipating our core argument that universal constraints guide acquisition while experience supports the child’s creation of a language-specific grammar. We present an overview of the structure of the book.

In this chapter, comparable designs and methods are applied to a language which is grammatically distinct in branching direction and word order from either English or French: Tulu, a Dravidian language spoken in Southern India, which allows several variants of relativization. The chapter introduces critical aspects of Tulu grammar, focusing on the pronominal feature system and its interaction with Determiners in the language. Despite variation in the expression of headed and headlessness, Tulu data confirm a primacy of headlessness in the development of relativization. Once again, the data reveal that the child is engaging in Grammatical Mapping between a universal template and linguistic properties of a language-specific grammar.

In this chapter, we summarize experimental results from the English, French and Tulu research reported in the previous chapters. We explicate commonalities and language-specific differences which have been discovered across these studies. We recognize the linguistic knowledge that is evident in the child across languages throughout development, but return to a leading question: Why are headless relatives precursors across languages in acquisition? Why are headed forms particularly delayed? We confront the essential problem: reconciling the strong linguistic knowledge continually present in the child with the evidence of developmental delays in relativization. We introduce the need to compose a more refined approach to a theory of language acquisition.