sellars on kantian intuitions

SELLARS ON KANTIAN INTUITIONS

MICHAEL WOODS

In an essay entitled "Sensibility and Understanding" Wilfred Sellars exploits an ambiguity "implicit in Kant's notion of intuition. 1 "Intuition", he points out, "is Janus faced" for, as he ex- plains, what Kant speaks of as "intuitions" might well be conceptual in a broader sense, so that they turn out to be representations of 'thises'. This paper is an attempt to explore this observation.

Inasmuch as Kant wavers in his insistence on the distinction between concepts, essentially mediate and discursive, 2 andintuitions, essentially immediate and singular, it is evident that the notion of a 'conceptual intuition' would be incongruous if not mistaken. And this would seem especially so in view of the fact that Kant strongly rejects the possibility of intellectual intuition for anyone but the "primordial being". 3

Sellars, of course, is aware of this. But to substantiate his view that there are intuitions which involve something over and above mere receptivity, that is, 'conceptual intuitions', he points out that the productive imagination is said to be the understanding function- ing in a special way and to be engaged in the representing of individ- uals involving receptivity. Textual evidence is found for this in the second edition Deduction 4 although as Sellars himself suggests, such an interpretation would not be so firm if reference were made to the first edition.S The point is nevertheless an interesting one.

Much of Sellars' argument depends on his assumption that an empirical intuitional representation can be linguistically represented. Hintikka has a similar view when he says that Kant's notion of intuition "is not very far from what we would call a singular term". 6 Now Sellars had already suggested that intuitions, in certain con- texts, have a form which could be said to be conceptual in virtue of their synthetic character. But it is difficult to s e e how this minimal, and as yet unclarified claim, should entitle him to assume that intuitions can be linguistically represented and, moreover, can be represented as the subject of sentences. He argues, however, as follows:

137

MICHAEL WOODS

For of intuitions those, at least, which are synthetized by the productive imagination would seem to have a form illustrated by

'this cube' which though not a judgment, is obviously closely connected with the judgment

`this is a cube'.

r cube' is then said to have a 'this such' form which although conceptually rich is 'incomplete' until it is attached to a predicative expression to form a singular proposition.

There are number of points to be made here about this linguistic representation of an intuition and the form it is said to enjoy:

(A) It may be said that, rather than "illustrate" the form of an empirical intuition in the broader sense, which he assumes, the expression

'this cube'

would more plausibly be said to express a concept the addition to which of an indexical expression could at best be said to limit its extension.

Inasmuch as the reference of this indexical expression will be indeterminate (unless a number of conditions are appended) the singular hyphenated expression could not even be said to refer to at most one object.

The argument that

'this cube'

expresses just a 'singular concept' would account, then, for its being 'conceptually rich', so that when Kant says that intuitions without concepts are blind he means, not that they must be conceptual to be expressible and therefore intelligible but that they cannot be intelligible without concepts or/n abstracto: this is quite another matter.

(B) Although Sellars does not explicitly say so, the demonstrative expression "this" as it occurs in the judgment

'this is a cube'

seems to represent an empirical intuition in virtue of its "obvious connection" with

'this cube'.

138


And it must be said that the demonstrative pronoun would seem to be a Plausible choice to represent an empirical intuition for it expresses singularity and possibly immediacy and we can be said to subsume it under a concept by attaching it to a predicate ("This is a red circle"). It seems simply to represent an indeterminate, or at least undetermined, object which has submitted to the conditions of sensibility and therefore to be an adequate representation of an empirical intuition.

There are important difficulties associated with this view, however, and the seemingly ingenuous demonstrative conceals more than might at first appear. A minimal pre-requisite of an expression which might represent an empirical intuition is, if it refers at all, that it refer to at least and at most one object, for Kant insists throughout the Critique on the singularity of intuitional representation. 7 And in order that this be achieved by a demonstrative expression, as I have suggested, certain conditions must be fulfilled. Hence, "this" may be said to refer to at least and at most one object if, it is added: "as uttered by the speaker P at place L and at time T". But the only condition which Kant places upon intuitional representation is that, in order to be such, it must have submitted to the forms of sensibility. The introduction of the coordinates of speaker, place and time is an introduction of conceptual determinations of the intuitional representation. The demonstrative pronoun must therefore be discounted as a possible candidate for the linguistic representation of an empirical intuition.

(C) It may be further argued against the possibility of representing an empirical intuition linguistically that an intuition, however empirically indeterminate, is at least categorially determinate; that is to say, is at least something; 8 if it is possible to linguistically represent an intuition, the intuition will necessarily be represented as an object and will be seen, therefore, not merely to have submitted to the conditions of sensibility, but to the concept of an object in general. The expression "this", for example, represents an object, albeit empirically indeterminate and cannot, therefore, represent an intuition for it already expresses a representation which has submitted to a minimal conceptual determination. Consider Sellars; however:

"The view that before we can have representations of the form 'X is a cube'

139

MICHAEL WOODS

we must have representations of the form 'this cube'

is a puzzling one."

But this is only puzzling when the expression 'this cube'

is said to represent an intuition, at least in the manner of referring to it, and I have been arguing against this. What ought to be said, a view to which Kant would subscribe, is that judgement is not possible without intuition and intuition cannot be represented apart from judgment. I t is doubtful, in any case, that Kant would have accepted Sellars' formulation:

"Before we can have judgement we must have intuition."

Talk of having judgements and intuitoins would certainly lead one to believe that we possess such "things". And the question here is whether an analytic device is being interpreted ontologically - that is, whether the method of distinguishing elements is being mistaken for an interpretation which demands separating "things".

In any case the point Sellars wishes to make here is that the expression

'this cube'

is incomplete unless it forms part of a judgement such as

'this cube is a die'.

It will be seen that this is a curious inversion of Frege's argument that the concept of propositional function

"X is a die'

for example, is "incomplete" or "unsaturated" unless a value is found for the variable. And the trouble is that, as far as Sellars is concerned, there is no argument presented as to why it should be that a singular expresssion is incomplete or, indeed, precisely what this means.

(D) But apart from urging the reader to accept that this is so ("Surely the presentations . . . are essentially incomplete") he does say this:

"While they (singular representations) can occur in a mental listing, they would be unable to play even this role unless one knew how to complete them ''9

140

$ELLARS ON KANTIAN INTUITIONS

Now Sellars does not give any reason for introducing a hyphen into the expression

(1) this-cube

or less tendentiously, perhaps, he does not explain why the expression is hyphenated. Indeed, it is used interchangeably with

(2) this cube.

But there is an important difference. The introduction of the hyphen in (1) could be meant to suggest that the expression is a n a m e

for an empirical intuitive representation. And this could be then said to be "completed" if attached to a predicate expression. Hence,

'this-cube is a die'

is the "completion" of the name sufficient to enable it to participate in a "mental listing". On the other hand if (1) were deemed a demonstrative singular expression then

'this-cube is a die'

would at the very least be a curious syntactical specimen. I would suggest, then, that (2) looks more like a demonstrative singular expression than (1) which is best regarded as a name. Both are species of linguistic representation, although I have argued that a demonstrative singular expression cannot represent an empirical intuition. Perhaps, nevertheless, a name can.

If (2) is a name then "this" as it occurs in it is no more a demonstrative that "that" as it occurs in

Thatcher.

Moreover,;

'this-cube'

could just as easily be the name for the leader of the Conservative Party as for a solid with six equal square sides although its being a name for the latter, given the number of objects which exist, it will be admitted, would be a remarkable coincidence. And this is because there can be no referential relation between a name and an object; rather, the relation is one of "standing for".

The question, then, is whether an empirical intuition can be linguistically represented by means of a name. Now it is required of such a name that what it stands for exists and is unique, and that the name can be reapplied to just this;applied, that is, to

141

MICHAE L WOODS

precisely that for which it stands on more than one occasion. The intuition, therefore, to which

'this-cube'

applies must be the same intuition to which it is reapplied assuming, indeed, that it is reapplied.

Of course it will be readily observed and admitted that chronic difficulties will attend any discussion about what constitutes "the same intuition". I propose, therefore, to take as short a cut as I can by making two points: firstly, intuitions, for Kant, are partic- ular representations which are single. 1~ This means both that an intuition is a representation o f one object and representation which has one occurrence. Particularity and unity are not intrinsic to an intuition but are attributions which are categorical, so that, secondly, any attempt to describe an intuition will involve the use of concepts. 11 It may therefore be concluded that intuition cannot be independently characterised and so, perforce, cannot be named.

(E) The only other possible interpretation of the expression

'this-cube'

is one in which it will be held as a truncated assertion, a shorthand disguise for

'this is a cube'

And if this is so, of course, it cannot be the linguistic form of an empirical intuition which, it must then be said, can have no expressible form at all.

In conclusion, it is clear that the attempt to linguistically represent intuitions rests on a mistaken interpretation that they enjoy a recognizably separate form or a form which is 'sui generis'. Undeni- ably, I would suggest, there is an ambiguity in Kant's term "intuition", an ambiguity which Sellars was at pains to exhibit. But intuitions however conceptual they may be said to be, are not linguistically representable and cannot have a form.

P.O.Box 171 MAPRIK

EAST SEPIC PROVINCE PAPUA, NEW GUINEA

142


NOTES

t Reprinted in his book Science and Metaphysics. Routledge and Kegan Paul, 1968, Chapter 1.

= Critique of Pure Reason, Kemp Smith translation, A320/B377. Herein- after, the first edition pagination is signified by 'A'; the second by 'B', in accurdanee with convention.

3 B72. 4 B151-153. s c.f. A l l 8 , and The Anthropology, Vol. III, B28: "The imagination is a

faculty of intuition without the presence of an object". J. Hintikka, "On Kant's Notion of an Intuit ion" from The First Critique: Reflections on Kant's Critique of Pure Reason, ed. T. Penelhum and J J . Macintosh, 1969, p. 43. c.f. A320, B376-377, A713/B741.

s A250: "All our representations, it is true, are referred by the understanding to some object."

* It should be noted that Sellars uses "representation" in the sense of "linguistic representation" or "representation of an intuitional representation".

t o A320/B375; A713/B741. tt A162/B202.

143

sellars on kantian intuitions

Documents