Monday, 13 September 2010

Sellars and Kant on Intuition (III)

I have been going through the first chapter of Wilfrid Sellars' Science and Metaphysics recently with the last posting  on this topic discussing the second movement of the chapter and revealing how, in it, Sellars substantiated his conception of "sheer receptivity" and connected it to a revised account of "sense impressions". In this posting I will look at the third movement of the first chapter.

The third movement opens with what Sellars hopes is a relatively uncontentious claim to the effect that visual perception involves, in some sense, conceptual representations. If we do assume this to be so then it will follow that verbal reports made concerning what it is we are seeing will express a "conceptual episode" that contains the twin elements of perceptual capacity and a reference to how the general environment is causally affecting one. This point is then used to begin to explore the relationship between ordinary speech that refers to sense impressions and the philosophical analyses that have been undertaken of such. The contrast drawn is that, generally, people seem to have no compunction in referring to a general impression of a red book on a brown table but philosophical analysis of impressions rarely allows for the sense that what that phrase might report is a distinct sensory impression. The reason why philosophers appear to have been reluctant to analyse what is given in ordinary speech in a way that reflects the structure of that speech appears to have been that philosophical analysis of such impressions produces a thinner sense of what is given than ordinary speech appears to report.

However, there is a point at which philosophical analysis and ordinary speech get closer together and this is when ordinary speech reflects a problem in the speaker's response to the datum given. So, speakers become, for example, confused about what is given to them and then their speech reports become correspondingly thinner (and more like philosophical analysis). Sellars illustrates this with a progressive distinction between speakers concerning whether the datum they have given to them is "a red book over there" as opposed to "a rectangular physical object" or even it merely "looking like" there is a red book over there. The second claim would be a kind of minimal objective claim (much like philosophical analysis tends to give) whilst the last moves from a vocabulary that endorses something as given to one which places it in a merely "seeming" location. We can imagine the second claim relativised by also being put in a "seeming" manner.

If the second claim was so relativised then we have a kind of reflection on the statement in question. Now, in general situations the reason why this kind of reflection often appears required is, as first suggested, because of a kind of puzzlement about the datum given. It doesn't seem necessary to reflect otherwise. However, philosophical analysis has often operated as if the experience of the sense impression was typically like the one that ordinary speech only reflects in these cases of puzzlement. By contrast, what Sellars seems to aim at now is a distinction between the presence of minimal sense impressions and the necessity of minimal conceptual episodes like those given when we place the impression under the cover of a "seeming" notion.

However, rather than immediately directly arguing against the assumption that says  a kind of minimal conceptual episode is required if there exist minimal sense impressions, Sellars first looks at what would have to be given in the minimal conceptual episode, if it existed. There would have to be, Sellars suggests, three distinct aspects of the visual perception. These would be: a) a purely physical aspect, however described; b) a primary mental aspect sufficient to produce the minimal conceptual episode; c) a "rich" conceptual episode that allows for reference to such things as "cabbages and kings and pigs in barnyards".

Sellars does not next instantly reject the division of the experience into the three elements in question. Nor does he reject the distinction between the "receptivity" of the sense and the "guidedness" of the sense by conceptual representation. The second is, in fact, taken to be required by Kantian analysis. For example, Kant refers to productive imagination "taking up" (A120) the manifold which implies, to Sellars, that the manifold is something independent. (See also, which Sellars does not cite, B145.) 

However, whilst the latter point seems required for the Kantian analysis to work, this manifold should not be construed as belonging itself to the conceptual order but to be something non-conceptual. It is only if it is understood, in accord with the general pattern of Sellars' analysis, as so non-conceptual, that we have a notion that there is something "outside" the conceptual order, something that he takes to be a sense of "intuition", the sense at work in the uniqueness of perceptual judgment where the latter refers to some notion that there is something given, a particular this as such. This continued insistence on the understanding of intuition in this way is part of what makes Sellars' analysis so constantly reminiscent of that of Maimon whilst also distinguishing it from Maimon since the latter takes the consequence of this point to be that space and time are not really intuitions at all.

The general point that Sellars closes the third movement with is that the heterogeneity of the receptivity of sense (in the radical way required by the notion of "sheer receptivity") is what would be needed for Kant to escape the dialectic of German idealism that leads to Hegel. Without this heterogeneity we are well on the way to such idealism and the movement towards it seems, indeed, part of the effect of Maimon's analysis.

No comments: