Enoch Aboh (University of Amsterdam)

Grammaticalization: a population factor

One of the most researched questions in linguistics has to do with how to account for linguistic typological variation, and what this variation teaches us about the evolution of language. The question is old and has to do with the origins of mankind.
One way to approach this question has been to show that patterns of language use in human communicative settings may affect the language system in the long run in such a way that linguistic entities are subject to specific developments. This has been assumed to be the case in grammaticalization, a process whereby an open-class item (e.g., a content word), acquires a grammatical function thus becoming a closed-class element. Studies by Givón (1971), Heine (2003), Heine and Kuteva (2005), and related work have shown that the process follows a developmental path including semantic reduction as well as loss of categorial and morpho-phonological specifications. The developing form is commonly believed to compete with existing forms serving similar functions in the language. Put together, these descriptive generalizations give the impression that grammaticalization is (i) a unidirectional process, and (ii) an independent language-internal process (though Bruyn 1995 shows that it may be contact-induced as well). Grammaticalization therefore represents an explanatory theory of language change (but see Fischer 2009 for a critique).
In this talk, I show that traditional views of grammaticalization as an independent linguistic phenomenon are misleading. Instead, I argue that what is referred to as grammaticalization actually represents a sequence of distinct synchronic linguistic behaviours within a speech community at a certain point in time. Put together, as is often done in historical linguistics, this succession of independent and arguably unrelated linguistic group behaviours presents us with a neat picture of diachronic change, in a way similar to kaleidoscopic motion. I will show that grammaticalization as described in the literature does not involve core linguistic components but can be better understood as a population factor: the aggregation of individual changes in I-languages within a given speech community over a certain period. I argue that these changes at the level of I-languages can be accounted for in terms of recombination of syntactic features (Aboh 2015). In this framework, learning biases observed at the population level are explained in terms of UG-based constrains on recombination.

References
Aboh, Enoch O. 2015. The emergence of hybrid grammars: Contact and change. Cambridge: Cambridge University Press.
Bruyn, Adrienne. 1995. Grammaticalization in creoles: The development of determiners and relative clauses in Sranan. Studies in Language and Language Use 21. Amsterdam: IFOTT.
Fischer, Olga. 2009. Grammaticalisation: Unidirectional, non-reversable: The case of To before the infinitive in English. In Olga Fischer (ed.), Pathways of changes. Amsterdam: Benjamins.
Givón, Talmy. 1971. Historical syntax and synchronic morphology: An archaeologist’s field trip. CLS #7, University of Chicago: Chicago Linguistics Society.
Heine, Bernd. 2003. Grammaticalization. In Brian D. Joseph and Richard D. Janda (eds.) The Handbook of Historical Linguistics. Oxford: Blackwell.
Heine, Bernd and Tania Kuteva. 2005. Language Contact and Grammatical Change. Cambridge: Cambridge University Press.