Search

The contemporary discussion of non-conceptual content inaugurated by Gareth Evans and John McDowell has generated a range of differing views as to Kant's position on the issues raised. I argue that for Kant perception is prior to thought and that it is as being prior that perception connects us to reality in outer intuition. I then argue that for Kant thought relates to perception by being the rule for perceptual procedures. This accounts for thought's extending in scope beyond what we actually perceive to all that is manifest in space and time. As against Merleau-Ponty this Kantian understanding of thought beyond perception does not distort the nature of reality which remains essentially that which can be engaged.

According to Kant's theory of thought or cognition, thoughts are rules for empirical reactions in the compass of spatial and temporal constructions. Theses rules function to represent our situation in relation to all the ways it is proper to interact with reality. After outlining Kant's theory, I present a modified version in which rules are identified with executive mechanisms for behavioural output. Following Kant, I show how such rules can pertain to the past in terms of mechanisms for being beyond or past stages of temporal constructions. This identification of rules with mechanisms allows for a real definition of the truth of thoughts as the active realizability of the mechanisms that thoughts are. I show how this modified version can encompass the full scope of even relativistic spatio-temporal reality, and indicate why this theory deserves consideration as against rival descriptive and causal theories of cognition.

In a 1772 letter to Marcus Herz1 Kant asks how intellectual representations can refer to an object when they do not produce the object. How can a thought in me represent something which is not affecting me or present to me, granting that the thought itself does not produce the object? This question is directed by Kant against his own view in the Innaugural Dissertation. There he had argued against Leibniz that spatio-temporal construction together with sensation did provide for adequate and precise representation as in geometry. However he still agreed with Leibniz that thought by itself was adequate for representation of objects (which objects constituted the intelligible world). The principle of this representation by thought was most likely some sort of isomorphism. The thought in its structure and in its component concepts was isomorphic to one and only one object and thus represented that object. Modern Leibnizians would be those for whom definite description understood attributively provides an adequate account of reference to objects. The failure of this account is presupposed and indeed sets the stage for the Critique of Pure Reason.

For Kant, geometry studies space which is a pure intuition and at the same time a form of empirical intuition. Geometrical statements somehow require or pertain to constructions or productive activity such as drawing, rotating, pointing, bisecting, etc. Space somehow provides the pure manifold which enables thought to pertain to objects. Still more, reality in space is neither finite nor infinite. These are all constraints from Kant’s text on what geometry might be for Kant. The 19th and 20th century have added further constraints on what a justifiable philosophical account of geometry would have to be. It would have to allow for non-Euclidean geometry, for the time parameter making a difference to the spatial geometry (more generally for geometry being space-time geometry) and for a connection between geometry and the distribution of matter. Putting these two classes of constraints together, we get a set of constraints on what a Kantian theory of geometry that is also in itself plausible would have to be. It is such a theory that I should like to begin to present in outline form.