The Weekly Newsletter of MIT Linguistics

LingLunch 10/28 - David Pesetsky (MIT)

Speaker: David Pesetsky (MIT)
Title: Clause Size Revisited: Kinyalolo’s Constraint as the engine behind Exfoliation phenomena
Time: Thursday, October 28th, 12:30pm–1:50pm

Abstract: In earlier work (Pesetsky 2019/2021), I argued that clauses of reduced size are derived from full and finite clauses in the course of the syntactic derivation, thus reviving a proposal from the first decade of research in generative syntax (abandoned on grounds that no longer hold water in contemporary syntactic models). I presented a set of derivational opacity arguments in favor of the derivational view: properties of reduced clauses including infinitivals that are difficult to explain if reduced clauses are born that way — but are natural and easy to explain as by-products of an early derivational stage in which they are full and finite. The overall approach also unified phenomena normally not viewed as related: grouping under one explanation the special behavior of subjects in nonfinite clauses with the so-called complementizer-trace effect observed under Ā-movement from finite clauses. I will review these results and take them to be sound.

At the same time, however, the specific proposal that I advanced to explain how clauses become reduced and why is a much less settled matter. In that earlier work, I proposed a rule of “Exfoliation” that applies whenever a movement-inducing probe contacts a goal across a CP clause boundary, and that goal does not occupy the edge of its clause. Exfoliation peels away the clausal layers between the goal and the CP edge as needed, thus permitting extraction. This proposal, however, required at least three kinds of additional innovations to do its job: (1) an anti-locality restriction that prevents the goal from simply moving to the desired edge; and (2) a toP distinct from TP, so that infinitives may be said to differ from finite clauses in whether TP had been stripped away, as well as a similar novel head between T and C, to explain the behavior of complementizer-trace effects in languages like French and Bùlì; and (3) an “Exposure Condition” that suppresses the pronunciation of these novel heads when Exfoliation does not take place.

This talk tentatively explores an alternative. I will ask whether subjects might not be obligatorily extracted by successive-cyclic movement through Spec,CP after all (so there is no anti-locality restriction — just the opposite!) — and whether it could be the configuration created by hyper-local movement itself that triggers clause reduction. On this view it is not movement across the clause boundary itself, but the necessary precursor, local movement to spec,CP, that mandates reductions of elements such as C and T in the clausal spine. As it happens, just such a proposal has been advanced in a partly (but not totally) different set of contexts by Kinyalolo (1991), Carstens (2003, 2005), Oxford (2020) and others: in configurations where two adjacent heads agree with the same element, one or the other of these heads characteristically reduces (“Kinyalolo’s Constraint”). I will explore a potential new narrative for clause reduction based on this idea. This alternative appears to eliminate all three of the extra assumptions listed above that are necessary on the Exfoliation approach — with additional dividends for the unification of anti-Agreement phenomena (Ouhalla 1993, 2005; Baier 2018; and others) with the broader picture.