Whamit!

The Weekly Newsletter of MIT Linguistics

Ling Lunch 9/11- Juliet Stanton

Speaker: Juliet Stanton (MIT)
Title: Learnability shapes typology: the case of the midpoint pathology
Date/Time:Thursday, September 11, 12:30-1:45pm
Location: 32-D461

The aim of this paper is to explore the idea that learnability shapes typology: that the range of linguistic variation we observe is delimited by constraints on the types of grammars that can be acquired accurately and reliably (cf. Boersma 2003, Alderete 2008, Staubs 2014). I focus specifically on the midpoint pathology (Eisner 1997, Hyde 2008, Kager 2012), a term used to characterize a type of unattested stress system in which stress is drawn to the middle of mid-length words, but not others. Kager (2012) argues that eliminating the pathology requires us to eliminate contextual lapse constraints (e.g. *ExtLapseR) from Con and adopt weakly layered feet; I argue instead that systems exhibiting the pathology are unattested because the necessary ranking is difficult to learn. I present modeling results that support the current proposal, and show that the absence of the midpoint pathology can in part be attributed to extra-phonological limitations on the learner’s input. I discuss some immediate challenges for the proposal, and show that its global predictions are borne out. I argue that, if we can explain the absence of all midpoint systems as the result of constraints on learnability, then there is no need to exclude them from the learner’s hypothesis space. In other words: to explain the absence of the midpoint pathology, we do not need to eliminate contextual lapse constraints, nor do we need to adopt weakly layered feet.