CLUL Talks with Roberto G. de Almeida

Associate Professor,
Department of Psychology,
Concordia University, Montreal, Canada

Sentence interpretation without lexical decomposition

• Date: Friday Oct 8, 2010
• Time: 11:00 - 13:00
• Location: Anfiteatro, Complexo Interdisciplinar, Universidade de Lisboa.
• Address: Av. Prof. Gama Pinto, 2,1649-003 Lisboa. A map can be found at this website
• For questions or more information, please contact Maria do Carmo Lourenço Gomes: mclgomes@fl.ul.pt .

Abstract

Much of the debate on the nature of lexical-semantic representation has centered on the representation of causative verbs. There have been numerous linguistic studies supporting the notion that lexical causatives such as kill or boil are represented by definitional, decompositional templates (e.g., Levin & Rappaport-Hovav, 2005). And there are also a few experimental studies supporting this view. A less prominent but nonetheless equally fertile ground for lexical-semantic decomposition theories comes from the so-called semantic coercion phenomenon. In semantic coercion, the alleged case of semantic decomposition of nominals comes from the interpretation of sentences such as The man began the book. The verb begin is said to 'coerce' the entity book (x) to be interpreted as an activity performed with x (e.g., Jackendoff, 1997). This coercion process is supposed to rely on the semantic analysis of the entity, which makes available information about what that entity is typically used for, thus licensing an interpretation of The man began the book as something like The man began reading the book (Pustejovsky, 1995, 2002). However, empirical evidence for this sort of semantic decomposition is slim, at best. And even if supporting evidence for this kind of process is real, it can be accounted for by a structural analysis of the predicates and with the aid of pragmatic inferences that make begin the book to be abductively interpreted as begin reading the book. I will discuss some recent behavioral and neuroimaging experiments from my laboratory suggesting that there is no clear supporting evidence for decompositional analyses of lexical-semantic representation. I will suggest instead an atomistic -- i.e., non-decompositional -- account of semantic representation for causatives as well as for coercion phenomena. I will argue that an atomistic theory of lexical-semantic representation is preferred for it is not committed to establishing an analytic/synthetic distinction for constituent features of concepts. I will further stress the role of linguistic-structural properties of causatives and coerced expressions, suggesting that semantic representation is to a large extent structurally determined and work as a trigger to pragmatic inferences supporting comprehension beyond linguistic representation.

Short Biography

Roberto G. de Almeida is currently an Associate Professor of Psychology at Concordia University (Montreal, Canada). He received a PhD in Psychology and Cognitive Science from Rutgers University (under the direction of Jerry Fodor) and was a post-doctoral fellow at the University of Alberta. His research focuses on the nature of morphological parsing, on the lexical-semantic representation and processing of verbs, and on sentence interpretation, including work with healthy subjects and Alzheimer's patients.

Website of Roberto G. de Almeida

For more information, please contact one of the CLUL Talks organizers, the CLUL post-docs FCT Ciência 2007-2008:

Previous CLUL Talks

• Amalia Arvaniti, May 28, 2010
• Roger Evans, March 5, 2010