[lingtalks] Jon Sprouse to speak at USC October 29th

[Karma Dolma]

The Colloquium Committee is pleased to announce our first speaker for the 2007-2008 academic year: 


Jon Sprouse
University of California, Irvine
Monday, October 29th 
2:00pm to 3:00pm in GFS 118 
(Reception to follow)


Below please find the title and abstract for the talk.


We hope to see you there,
The Colloquium Committee




"Acceptability, grammaticality, and the role of experimental syntax"


The past 10 years have seen a significant increase in the number of studies using formal acceptability judgment methodologies, methodologies that are collectively known as experimental syntax. Many reasons have been offered for the adoption of experimental syntax, from obtaining more reliable data to being more sensitive to small differences in acceptability. But while informally collected judgments have historically been confined to investigations of the content of grammatical knowledge, the judgments collected using experimental syntax have recently been used to argue for theories about the nature of grammatical knowledge. In this talk, I argue that this approach to experimental syntax is correct in its goals, but a bit premature in its implementation. We absolutely should be using formal experiments to better understand the relationship between acceptability and grammaticality, but we must first develop a deeper understanding of the nature of our experimental techniques, 
the influence of experimental designs, and the factors that contribute to acceptability effects.


As case studies, I will highlight three claims about the nature of grammatical knowledge that have been supported using experimentally collected judgments, and that have been interpreted as evidence for very specific theories of language:

(i)    that grammatical knowledge is continuous (rather than categorical), and therefore best modeled with an Optimality Theoretic approach (e.g. Keller 2000, Keller 2003)

(ii)  that grammatical knowledge is probabilistic (rather than discrete), and therefore language acquisition is a type of statistical learning (e.g. Bresnan 2006, Jaeger 2006)

(iii) that acceptability effects can be explained by processing resource limitations, therefore grammatical knowledge is simple (rather than complex) (e.g. Kluender and Kutas 1993, Sag et al. 2007)

 

I will present a series of three experimental studies that suggest that these interpretations arise from our shallow understanding of judgment tasks, experimental design, and the factors influencing acceptability respectively. I argue that these results underscore the importance of further research into the relationship between acceptability, grammaticality, and the methodologies of experimental syntax.


Karma Dolma
Office Assistant
USC Department of Linguistics
213 740-2986