The Weekly Newsletter of MIT Linguistics

Ling-Lunch 11/04 — Vincent Rouillard (MIT)

Speaker: Vincent Rouillard (MIT)
Title: Temporal in-Adverbials: A Study of Polarity Chameleons
Date: Thursday, 11/04
Time: 12:30 — 1:50pm
Location: 32-D461

Temporal in-adverbials include expressions like “in three days”, which can occupy at least two syntactic positions. Surprisingly, where such expressions occur in the syntax determines whether or not these are negative polarity items (NPIs). These expressions can modify a VP, restricting the set of events in its extensions to those of a duration specified by the adverbial. As shown in (1) and (2), in-adverbials show no polarity sensitivity in these configurations.

(1) The workers built the house in three days.
(2) The workers didn’t build the house in three days, (they built it in two).

Another available syntactic position for in-adverbials has them modifying the perfect. In such configurations, in-adverbials seem to measure the duration of the Perfect Time Span, an interval introduced by the perfect (Iatridou et al. 2002). What (4) states is that at least three days have passed since the workers last slept. As shown by the contrast between (3) and (4), in-adverbials behave like NPIs in these environments.

(3) *The workers have slept in three days.
(4) The workers haven’t slept in three days.

Where in-adverbials display polarity sensitivity can be captured if we make two assumptions. The first is that these expressions have alternatives and obligatorily fall under the scope of an exhaustification operator (Chierchia 2013). The second is that whenever a contradiction is guaranteed to be derived from the logical material in an utterance, this results in ungrammaticality (Gajewski 2009). I show that the only syntactic configuration among (1-4) where in-adverbials are predicated to be bad is when they modify the perfect in upward-entailing environments. This is because only there does exhaustification always lead to a contradiction. In-adverbials therefore represent a strong argument in favor of a general theory of NPIs, where the grammaticality of these expressions is dependent on the logical relations that exist between a sentence that includes NPIs and the sentence’s alternatives.

The presentation will be given in person, however, if you cannot make it to Stata on Thursday, you will be able to join via Zoom. Please contact Ling-Lunch organizer (kukhto@mit.edu or mortier@mit.edu) for zoom link.