In the concluding section of his Carus lectures Sellars argues “this essay ... can be construed as a restatement and refinement of the argument of Philosophy and the Scientific Image of Man” (FMPP III 89). This may come as a surprise, given that his aim in the Carus lectures is to articulate a mono-categorial ontology of pure processes. The retrospective assessment also reveals that what in the 1960's is thought of as a clash of Images (PSIM), has been an issue about the primacy of categories all along. If this reading goes through, then the key theoretical notion for resolving the clash lies in Sellars' category theory.

In this talk I present an interpretation of Sellars' category theory according to which categories are meta-classificatory rules. And although Sellars has worked on categories at several points throughout his career, these attempts can be shown to be continuous for both Images. To think of Images as clashing is to think of their fundamental categorial components as explanatorily incompatible. Now, if a unified reading of categoricity can be given across Image, the real issue of concern is one about novel schematisations of categories. For if the ”conceptual structures of theoretical science give us new ways of schematizing categories” (SM II §49), what clashes are the criteria we develop for the (ultimate) categorial home of entities, qualities and persons.

A first step is to spell out how this classificatory reading of categories can serve as a framework for interpreting categoriality in both Images. This involves accounting for the relationship between Manifest, philosophical (transcendental) and scientific, empirical categories, and thus showing how categories work in (sub-)disciplines inside and outside of (transcendental) philosophy. To this end the continuity thesis is developed as a refinement of the meta-classification thesis.

Continuity Thesis: In as much as transcendental categories are rules for meta-classifying (features) of mental acts into types, ontological categories are rules for classifying mind-independent (features of) contents of representations according to their epistemic powers.

What invites the distinction between transcendental categories and ontological categories then also allows for their continuity. For if the former are won by way of reflection the features that mental acts necessarily have, the latter pertain to the features the contents of these acts have (KTE §25). And while it makes sense to think that the types of mental acts we deploy are final and dependent on us as finite thinkers, it likewise makes sense to consider the number of features of (possible) contents to be much larger, varied and subject to future research. This gives a perspective on ontological categories as meta-classificatory rules for organising features (in the broadest and most specific sense of the term) into contents we can think of and contents we can experience. As reliable and shareable rules they guide our synthesis of features we represent into coherent subject matters. While this is also true of empirical concepts, what qualifies something as ontologically basic is that it is irreducible and therein constitutive of further contents or units of concern.

A second step is to show how this allows for a novel reading of categorial terms in the Scientific Image, i.e. disciplines outside of philosophy; because each discipline has their own rules for the classification of ‘represented contents' into kinds. And each may claim these kinds to be ‘ontologically fundamental' for their field of research. On the reading developed here, ontological fundamentality, though, means first and foremost, that at the object-level of classifying something to be of a certain kind is irreducible in the order of coming into being. It is what regiments the genesis of a discipline's subject matter. In as much as such object-level classification happens within, or by way of engaging in judgment types which invoke transcendental categories, these ontological classifications rest on and presuppose the transcendental categories which the Kantian tradition is concerned with (IKTE §44).

A third step is to discuss the argument in favour of this reading of the relationship between the categories of Manifest Image and the categories of the Scientific Image. Although varying in degrees, scientific disciplines work on the assumption that the objects of their domain are available (for ordering and qualifying) prior to all methodological disputes about how to demonstrate the categorial home. This central assumption is what distinguishes scientific ontological categories, or acts of classifying entities in an object-level sense, from transcendental categories at the heart of the Manifest Image. These latter ones are meta-level rules for classifying mental acts generally, i.e., as being an act of a certain judgment type (e.g. that p causes q). Still, the cognitive operation underlying these two distinct domains about which categorial ordering is to be achieved is one and the same: classifying according to rules that takes an epistemically relevant feature to be salient in determining the power of the content thus represented. If this reading goes through, we arrive at a novel understanding of the clash of Images, it's prospects and its time-invariant features.

References

FMPP . (1981) “Foundations for a Metaphysics of Pure Process” (The Carus Lectures), The Monist 64: 3–90.

IKTE (1978) “The Role of Imagination in Kant's Theory of Experience,” in Categories: A Colloquium, H. W. Johnstone, Jr. (ed.) University Park, PA: Pennsylvania State University Press: 231–45.

KTE (1967) “Some Remarks on Kant's Theory of Experience,” Journal of Philosophy 64: 633–47.

PSIM (1962) “Philosophy and the Scientific Image of Man”. In Frontiers of Science and Philosophy, Robert Colodny (ed.), 35–78. Pittsburgh, PA: University of Pittsburgh Press.

SM (1968) Science and Metaphysics: Variations on Kantian Themes, The John Locke Lectures for 1965–66. London: Routledge & Kegan Paul.