Where do constructions come from? Synchrony and diachrony in a usage-based theory

Diachrony as part of linguistic theory Returning now to our discussion of language as a complex adaptive or self-organizing system, this chapter addresses directly the processes of change that create emergent structures, for these processes or mechanisms of change are the ultimate basis for the explanation of why language is the way it is. The importance of diachrony for understanding grammar, especially in a typological context, but also for understanding cognitive processes, has been emphasized by a number of linguists over the decades; for instance Greenberg 1963, 1969, 1978, Givón 1979 and elsewhere, Heine et al. 1991, Haiman 2002, as well as in my own work – Bybee 1985, 1988c and Bybee et al. 1994 and elsewhere. Language change is not just a peripheral phenomenon that can be tacked on to a synchronic theory; synchrony and diachrony have to be viewed as an integrated whole. Change is both a window into cognitive representations and a creator of linguistic patterns. Moreover, if we view language in the manner described in this book, as both variable and having gradient categories, then change becomes an integral part of the complete picture. This chapter and the next two deal directly with diachrony. We begin our discussion in sections 6.2–6.5 with the diachronic phenomenon of grammaticalization, which has been intensely studied over the last few decades. In my view, the empirical research into grammaticalization has contributed more to our understanding of grammar than any other empirical work during the same period.