Marina Terkourafi @ LEL Seminar

Posted on April 18, 2018 by

Next week, on Tuesday 24th April, we welcome Marina Terkourafi at the LEL Research Seminar. Marina specialises in pragmatics and sociolinguistics at the University of Leiden. Her primary research areas concern “issues of pragmatic variation and experimental pragmatics”. She has been conducting fieldwork in Cyprus, with “extensive research on the use of politeness markers, the interplay between local and standard codes, and processes of language change since the Middle Ages”.

Please see below for details of the seminar and for the title and abstract for Marina’s talk.

Date: Tues, 24th April 2018

Time: 16.15

Room: Samuel Alexander A101


Against a unified account of scalar terms
Marina Terkourafi
Leiden University Center for Linguistics, the Netherlands

Scalar implicatures are a type of inference in which the use of an informationally weaker term (e.g.,some) is taken to mean that an informationally stronger term (e.g., all) does not apply. An important factor affecting the derivation of scalar implicatures seem to be interpersonal dynamics such as face and affect; however, this relationship is not yet very well understood. In this talk, I present results from an experimental study of eight types of scalar terms in face-boosting and face-threatening contexts. Our results for two of the terms (‘or’ and ‘some’ in subject position) confirm those of previous research, in that these terms are more likely to receive a lower-bound interpretation (no scalar implicature derived) in facethreatening contexts. However, our results for six newly-tested terms point in the opposite direction: with these terms, participants are more likely to derive the implicature when facethreat is imminent compared to when they expect their face to be enhanced. These results challenge a unified treatment of scalars from two sides: on the one hand, participants handle different terms differently depending on how they position themselves interpersonally vis-à- vis each other; and on the other, some terms are more susceptible to interpersonal concerns than others. I build on these results to argue for a closer integration of interpersonal and cognitive factors than previously assumed, with interpersonal factors along with linguistic properties of the terms themselves biasing toward a particular mode of reasoning over another, effectively integrating interpersonal factors into the inferential process from the