Please use this identifier to cite or link to this item:
https://www.um.edu.mt/library/oar/handle/123456789/22274
Title: | Viewing time affects overspecification : evidence for two strategies of attribute selection during reference production |
Authors: | Koolen, Ruud Gatt, Albert van Gompel, Roger P. G. Krahmer, Emiel van Deemter, Kees |
Keywords: | Natural language processing (Computer science) Corpora (Linguistics) Linguistic analysis (Linguistics) |
Issue Date: | 2016 |
Publisher: | Cognitive Science Society |
Citation: | Koolen, R., Gatt, A., van Gompel, R. P., Krahmer, E., & van Deemter, K. (2016). Viewing time affects overspecification: evidence for two strategies of attribute selection during reference production. 38th Annual Meeting of the Cognitive Science Society, Philadelphia. 1-6. |
Abstract: | overspecified: they tend to include more attributes than necessary to distinguish the target referent. The current paper investigates how the occurrence of overspecification is affected by viewing time. We conducted an experiment in which speakers were asked to refer to target objects in visual domains. Half of the speakers had unlimited time to inspect the domains, while viewing time was limited (1000 ms) for the other half. The results reveal that limited viewing time induces the occurrence of overspecification. We conjecture that limited viewing time caused speakers to rely heavily on quick heuristics during attribute selection, which urge them to select attributes that are perceptually salient. In the case of unlimited inspection time, speakers seem to rely on a combination of heuristic and more deliberate selection strategies. |
URI: | https://www.um.edu.mt/library/oar//handle/123456789/22274 |
ISBN: | 9781510832985 |
Appears in Collections: | Scholarly Works - InsLin |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
cogsci2016-reference-under-load.pdf | 453.73 kB | Adobe PDF | View/Open |
Items in OAR@UM are protected by copyright, with all rights reserved, unless otherwise indicated.