Programme StuTS 68

Conférence

A Model-Theoretic Approach for Greenberg’s Universal 20?

Novembre 21, 2020 Odille Morison en Theoretical Linguistics

Although there are several ways to account for the typological data available concerning J.H. Greenberg’s Universal 20, in my MA thesis I tried to sketch a single broader frame to account for all the various insights given by previous researches. What I came up with was an explanation of the data that is ascribable to a few principles: first, the interaction within the nominal domain of two Final-Over-Final-Constraints (one being the original one posited by T. Biberauer, A. Holmberg, M. Sheehan and I. Roberts and the other one being postulated by me); second, a strict ban on two linear orderings: [Numeral Demonstrative Noun] and [Adjective Numeral Noun].
By pushing forward the inquiry on a metatheoretic syntactic level, there are two aspects that I find highly interesting: first, G.K. Pullum’s way of conceiving the term “formalization” overlaps with what K. Hengeveld proposed back in 1999. Second, the original Final-Over-Final syntactic principle is no longer regarded as a “Constraint” by its authors, but rather as a “Condition” applied to a Minimalist framework; hence, Minimalism is departing from its original Generative-Enumerative state, getting (again) towards a hybridization with a Model-Theoretic Approach (see works by S. Müller, G.K. Pullum and B.C. Scholz). However, instead of fostering this process, one could simply ask: is a pure Model-Theoretic Approach the best way to describe the Final-Over-Final-Constraint and to account for my way of explaining U20?

The talk focuses on a theoretical topic, but it unavoidably involves typological considerations as well.