Morphology-syntax interactions in language change

Posted on November 11, 2016 by

Our seminar speaker for Tuesday 15th November will be Alexandra Simonenko (McGill/Ghent) on Structural and acquisition-based models of the morphology-syntax interactions in language change. The talk will be based on joint work with Benoit Crabbé (Alpage) and Sophie Prévost (LaTTiCe). It will take place at 4.15pm, as usual, in Samuel Alexander A113, followed by drinks and dinner with the speaker to which all are welcome.

Here’s the abstract:

In this talk we examine the nature of the relation between the availability of null subjects and the “richness” of verbal subject agreement, known as Taraldsen’s Generalisation (Taraldsen 1980, Rizzi 1986, Adams 1987), from the point of view of grammar change in Medieval French based on corpus data. We present the first corpus-based quantitative model of the syncretisation of verbal subject agreement in Medieval French and use it, together with a model of the loss of null subjects, to evaluate two hypotheses relating these phenomena. First, we test the predictions generated by the hypothesis that null subjects and non-syncretic agreement exponents are related at the structural level, both being surfaces manifestations of the same functional head. Modelling the diachronic variation between old non-syncretic and new syncretic endings on the one hand, and between old null subjects and new overt pronominal subject on the other as a competition between an old grammar with the relevant functional head and a new grammar without, we show that the two changes proceeded at the same rate. On the Constant Rate Hypothesis of Kroch (1989), this result is compatible with the hypothesis attributing both changes to the disappearance of the same syntactic element. The structural hypothesis, however, is also shown to generate a number of predictions that are not borne out. The second hypothesis we explore is based on Yang’s (2009) variational acquisition model whereby in the process of language acquisition syncretic endings create a bias against the null subject grammar, which eventually drives the latter to extinction. We argue that such an approach could be used to reconcile the intuition behind Taraldsen’s generalization with the fact that it has proven non-trivial to formulate the notion of agreement richness in a way which would unequivocally predict whether a language is pro-drop.