Sobre la distinción innovador/consejador y los modelos secuenciales en la lingüística histórica.

In this paper, I propose a review of two metalinguistic categories: [+/- innovative] and [+/- grammaticalized]. On the basis of some analogies in theiruse by different frameworks of historical linguistics (a traditional one, the comparative-historical method, and a modern one, the grammaticalization theories), I argue that both of them can be described in terms of "sequential models" and I deal with the way this kind of model takes account of language change. Historical evidence is brought in from several well-known Romance structures, and in particular, from the development of present perfects in Spanish and French, in order to point out to some of the methodological limits of the models: firstly the existence of mismatches between the sequential structure and empirical data, due to a high level of generalization in hypothesis, and secondly, what I will call the "stage 0 fallacy". [ABSTRACT FROM AUTHOR]