CENTER FOR COGNITIVE SCIENCE COLLOQUIUM

ROBERTO CASATI

CNRS-CREA, Ecole Polytechnique, Paris, and
Department of Philosophy, SUNY Buffalo
casati@idf.ext.jussieu.fr

"Shadow Cognition"

Wednesday, October 27, 1999
280 Park Hall
2:00-3:30 p.m.
North Campus

I shall discuss the following aspects and hypotheses about shadow cognition.

1. Shadows are dealt with by the visual system primarily in a non-attentive mode. Algorithms for recovering shape from shading and cast shadows operate at the sub-personal level.

2. Shadows are conceptualized only as a result of attention's being paid to the visual scene, where they may, in appropriate conditions, appear as salient. Shadows are discovered in early cognition; their concept is not part of an innate mental lexicon.

3. Our pre-theoretical concept of a shadow is relatively underdeterminate, as a consequence of (1) and (2). (a) Babies, (b) untutored adults, and even (c) professional painters are or have been confused by shadows. The concept of a shadow is a mixed bag of Gestalt, causal and spatial features. It might be interesting to use shadows for studying the mechanisms that construct concepts.
4. Shadows are used consciously, at the personal level, as inferential objects. As the astronomer Kepler said, "all discoveries in astronomy are made through light and shadow." I shall present a number of images from art history and the history of science to substantiate some of these points.