Contingent necessity versus logical necessity in categorisation

Pothos, E. M., Hahn, U. & Prat-Sala, M. (2010). Contingent necessity versus logical necessity in categorisation. Thinking and Reasoning, 16(1), pp. 45-65. doi: 10.1080/13546780903442383

[img]
Preview
PDF - Accepted Version
Download (462kB) | Preview

Abstract

Critical (necessary or sufficient) features in categorisation have a long history, but the empirical evidence makes their existence questionable. Nevertheless, there are some cases that suggest critical feature effects. The purpose of the present work is to offer some insight into why classification decisions might misleadingly appear as if they involve critical features. Utilising Tversky's (1977) contrast model of similarity, we suggest that when an object has a sparser representation, changing any of its features is more likely to lead to a change in identity than it would in objects that have richer representations. Experiment 1 provides a basic test of this suggestion with artificial stimuli, whereby objects with a rich or a sparse representation were transformed by changing one of their features. As expected, we observed more identity judgements in the former case. Experiment 2 further confirms our hypothesis, with realistic stimuli, by assuming that superordinate categories have sparser representations than subordinate ones. These results offer some insight into the way feature changes may or may not lead to identity changes in classification decisions.

Item Type: Article
Additional Information: This is an Accepted Manuscript of an article published by Taylor & Francis Group in Thinking and Reasoning on 04/12/2009, available online: http://www.tandfonline.com/10.1080/13546780903442383
Uncontrolled Keywords: Categorization, Critical features, Similarity
Subjects: B Philosophy. Psychology. Religion > BF Psychology
Divisions: School of Social Sciences > Department of Psychology
URI: http://openaccess.city.ac.uk/id/eprint/4685

Actions (login required)

View Item View Item

Downloads

Downloads per month over past year

View more statistics