Journal of Philosophical Investigations

Document Type : Research Paper

Author

professor at the Federal University of Rio de Janeiro

Abstract

In this paper, I take for granted the view of a long tradition tracing back to Kant that the content of perception is nonconceptual, nonpropositional, and iconic. However, I challenge the idea that this content is either an existentially quantified proposition (the existential view), an object-involving proposition (the particularist view), both (the pluralist view), or still that there is no fact of the matter about the elusive content of perception (Block). Instead, I propose an alternative hybrid model as the most suitable for perception, namely a mix of the representation of properties (relativistic content) and acquaintance with whatever causes the relevant token experience. Although this format is iconic or map-like, the best semantic model for understanding this relativistic content of perception is an open sentence with predicates and free variables. Since this content is neither particular nor existential, it is incomplete (at least in the light of Fregean semantics). That is, it is neither accurate nor inaccurate per se. Perceptions do not represent particulars, let alone the causal relationship between particulars or environmental conditions and the token experience. In other words, neither particulars nor causal relations belong to the content of the experience. Instead, particulars and causal relations belong to the evaluative circumstances of the content (Lewis's context-index pairs). Perceptions represent "de re" properties as accurate or inaccurate attributions to what is causally responsible for the relevant token experience.

Keywords

Main Subjects

CAPTCHA Image