Kant's third critique deduction

I will be concerned here only with the proximate goal of the deduction of taste in the Critique of Judgment. The larger, systematic question of how the transcendental deductions in all three Critiques cohere is vexed and relies on a great deal of textual analysis and reconciliation of apparent contradictions. We can, however, begin by saying something about the unifying thread in the larger project, which becomes most apparent in the third Critique: It must be the notion of purposiveness, since:

• structures in nature reflect purposiveness, absent the presumption of which we would be unable to conceive of objects as such,
• practical reason presupposes in us as moral agents a capacity to dictate and obey purposes not found in nature, effecting those purposes in the world through our choice of maxims—an act of judgment that determines will and produces pleasure—and
• the purposiveness inherent in our judgment of beauty makes possible a disinterested pleasure coupled with a claim to intersubjective universality.

The purposiveness involved in aesthetic judgment is pure, in the sense that it is unattached either to determinations of objects as falling under concepts, or to interest in them as objects of will or desire. Kant's general approach is to show how aesthetic judgments as described in the “Analytic of the Beautiful” can be accounted for using cognitive machinery already at hand in the critical philosophy. In particular, he must at least suggest how the beautiful can yield pleasure without interest, and he must reconcile subjectivity with an odd sort of objectivity.

To require a proof of the above, however, would be to ask too much of Kant. As in the first two Critiques, the notion of a transcendental deduction is predicated on an assumption of considerable agreement: that we do in fact make certain kinds of judgment; that in making them we claim universal assent, i.e., that we take our judgments to be about how the world really is, in some broad sense; and that the objectivity (or intersubjective universality) of such judgments presupposes a common mental apparatus. A transcendental deduction is not a deduction from axioms downward to particular conclusions but rather an ascent to the faculties and categories that would have to be in place in order for certain actual judgments to be possible.

The argumentative strategy from initial agreement is less apparent in Kant's critical philosophy than in his two “popularizing” works, the Prolegomena to Any Future Metaphysics [cite:@kant:prolegomena] and the Groundwork of the Metaphysic of Morals [cite:@kant:groundwork]. Sadly, there is no such work corresponding to the third Critique, but the “Analytic of the Beautiful” partly serves this purpose by interpolating examples within the architectonic, as when Kant distinguishes between our approval of the good and our enjoyment of the beautiful:

In order to consider something good…I must have a [determinate] concept of it. But I do not need this in order to find beauty in something. Flowers, free designs, lines aimlessly intertwined and called foliage: these have no significance, depend on no determinate concept, and yet we like them [cite:@kant:cj , Ak. 207, page 49].

A moment's reflection on one's own attitude toward “free designs” and “lines aimlessly intertwined” does seem to establish¹ that we sometimes discover an uninvested joy in otherwise useless forms and patterns.

The faculties presupposed in aesthetic judgment must satisfy certain conditions. First, they must explain how a judgment properly so-called can apply to a purely subjective experience. Second, they must offer an account of the pleasure we take in beautiful things, given our lack of interest in them as determinate objects. Kant's explanation addresses both requirements at once, since the interplay of the faculties involved in cognition turns out to give rise to pleasure just in its failed effort to determine an object.

Kant's theory of mental faculties recognizes their functional roles in constituting and interpreting the objective world. In the first Critique, the imagination appears to be given short shrift, since the “Transcendental Aesthetic” addresses sensation, while the “Transcendental Logic” (specifically, the “Transcendendal Deduction of the Pure Concepts of the Understanding”) deals with the understanding. But discussion of the imagination is taken up in the deduction, where it becomes clear that imagination plays an essential role in linking sensation to the understanding:

A pure imagination, which conditions all a priori knowledge, is thus one of the fundamental faculties of the human soul. By its means we bring the manifold of intuition on the one side, into connection with the condition of the necessary unity of pure apperception [the understanding] on the other [cite:@kant:cpr , A 124, page 146].

Conveniently for Kant's deduction in the third Critique, the imagination stands at the ready as partner to the understanding. The two faculties interact systematically to synthesize and resolve objects under concepts, rendering them not just coherent groupings of properties but bona fide objects of certain types. That this process should sometimes go on without resolution, to put matters in baldly strategic terms, gives Kant what he is looking for: an activity of mind that ranges freely over formal properties of objects, that remains uninterested in determinate concepts of those objects as such (because none are reached), yet that is capable of producing a judgment not about the object under consideration but about the cognitive process itself. The dynamism behind this account is made possible through a coherent view of the imagination as making a functional contribution to normal cognition.

Kant's theory of imagination is notoriously difficult to pin down. This is partly Kant's own fault, since he seems to insert the notion wherever it serves his immediate purpose and ends up characterizing imagination in sundry ways. Regarding its place in his scheme of faculties, Kant variously claims that the imagination is a part of sensation, part of the understanding, and a faculty in its own right. Regarding function, he ascribes to imagination a productive and reproductive aspect, the role of synthesizing the manifold presented in sensation, and the capacity of schematization in the production of concepts. It is left as an “exercise for the student” to integrate these differentiae within a coherent framework. I will only begin to do so here, but I hope to sketch a notion of imagination powerful enough to bridge the enormous gap between sensation and understanding and to participate on equal terms with the understanding in the “quickening” of faculties that Kant says is essential to aesthetic pleasure.

First, the situation of the imagination within Kant's faculty theory is self-consistent. Imagination must be a faculty in its own right, because it performs functions beyond those of sensation yet falls short of grounding any of its presentations objectively, according to the categories—for that, the understanding is needed. Nonetheless, the imagination is connected to sensation, because its presentations retain their origin in sensibility—they are interpretations of intuitions, as I will claim. In order to produce a unified experience, the understanding requires a candidate experience to unify. By itself, the understanding anticipates nothing of particular experiences except how they might be made objective; it fulfils its purpose only by applying the categories to presentations produced by the imagination. From a functional standpoint, then, the imagination must interact intimately not only with sensation but also with the understanding. In the aesthetic scenario, as I will suggest, the imagination is perhaps best seen as a sort of “subroutine” to the understanding, with the latter seen as the “driver” or calling function.

Second, the aspect of imagination capable of participating in the “free play” of cognition is likely the productive imagination, with the reproductive imagination playing a supporting role. Only the productive imagination, Kant says, is capable of 'spontaneity' [cite:@kant:cpr , B 152, page 165], but this does not necessarily mean that reproductive imagination lies dormant during aesthetic contemplation. Indeed, I take it that reproductive imagination is always needed to “keep things straight”—to keep the lawlike, empirical aspects of objects distinct from the rangings of the productive imagination. This need seems to hold at both low and high levels of representational abstraction, since artistic beauty, at least, seems to arise out of both objectively “irrelevant” (or perhaps “cognitively impenetrable”) formal material in the sensible representations of things, and the forms of association between things (including people) which have been fully represented as objects in the work.

In many plays and stories, for instance, including several of Shakespeare's comedies, we follow the exploits of primary and secondary (and even tertiary, e.g., in “As You Like It”) love interests. The couples involved are fully identifiable as persons, yet the formal fact that they pursue their confused affections in parallel (and sometimes in permutations and exchanges of partners) elicits a particular delight. What beauty we glean from dramatic devices of this sort must of course emerge, over and above the conceptual content of the play. Such beauty takes as its conceptual point of departure a higher order of representation than does, say, the beauty of abstract figures painted on the actors' costumes, yet in both cases reproductive imagination presumably disciplines and sets in relief the free play of cognition by providing a backdrop or “ground” consisting of empirical representations.

In order to get a sense of imagination's role in aesthetic contemplation, let me say a bit more about its production in ordinary cognition: the nature of its presentations and how those are referred to the understanding. Kant says that the imagination is "the power of intuitions or exhibitions [cite:@kant:cj , Ak. 287]." It attempts to assemble and group the manifold of sensory inputs into entities, or wholes, and even to construe those entities in object-like ways. Kant refers to these roles in the acquisition of knowledge as “synthesis” and “schematization.” In view of both, I will suggest that a reading which limits Kant's notion of the imagination to the construction of mere images is too weak, because it prevents the imagination from supplying the understanding with everything needed to impart objectivity to empirical representations.

The understanding, after all, constitutes only the final stage of synthetic unification. It posits a presentation of the imagination as an object of possible experience for any subject, not just the 'I' having the experience at a given moment. The bulk of the synthesis of concrete properties and relations in the object has already been handled by the imagination. Even the particular lawlike empirical concepts with which the object is associated in the mind have been constructed in the act of schematization, a function of the reproductive imagination.²

Others have argued against attributing a (merely) Humean theory of imagination to Kant. J. Michael Young points out that  [cite:@jmyoung:kantimg page 140] if the imagination formed only images, it would be incapable of grounding mathematical judgments.

In the '$7 + 5 = 12$' example from the first Critique, the synthesis involved in the judgment as a whole goes beyond the conceptual bases of 7, 5, and 12 in intuition. As Kant notes, there is nothing in our concept of 7 or of 5 that suggests the concept of 12. Addition of numbers must be more than a concatenation of intuitions reproduced as series of images. For intuitions may be taken in different ways. I may bring to mind the image of seven swans a-swimming or of seven unidentified flying objects. Either will do as an intuition of a collection of seven things. Just as I have to consider the images in terms of swans or UFOs in order for each to be an image of something, so I must consider each in respect of its representing precisely seven items. And this is the more general point about imagination, whether applied to physical objects or to quantities: it cannot merely produce images; it also must propose or provisionally determine what they are images of. Without this additional step, the understanding would have no conceptual content to subsume under the categories.

This points up an interdependency between association and interpretation. When we talk about perception, we often speak of “pattern recognition,” and we tell a story about how, say, patches of color in our field of vision get grouped into larger patterns, and the visual centers of the brain are able to interpret these patterns as representing certain objects, or at least (in the case of novel objects) objects with certain determinate traits. What the story is in danger of leaving out is how the very patterns corresponding to determinate objects and attributes got selected in the first place. There may be many ways of combining elements of an intuition. One supposes that the perceptual heuristic is more efficient than the brute-force approach of looping through all the combinations, running each up the flagpole of the understanding. It's more likely to be an iterative and adaptive process in which pieces of the puzzle are provisionally grouped, then checked against lower-level concepts such as color and shape. The “promise” of each chunk would be evaluated as conforming to known concepts and exhibiting lawlike relations, or as “making no sense,” in which case the chunk would be discarded before much time had been wasted on carving it out.

Imagination as mere grouping would define bare images, without having any way of generalizing or characterizing them. Because the grouping process is inseparable in this way from an ongoing interpretation of content, I prefer to read Kant as arguing for a maximal notion of imagination, what I would call the “interpretive imagination.” I also find myself referring to the entities that the imagination proposes to the understanding as “proto-objects,” which may approximate what Wilfrid Sellars [cite:@sellars:kantimg] calls “image-models.” Sellars' gloss on Kant's notion of imagination at first seems imagistic, but I don't believe it is.

Sellars distinguishes between what we see and what we see of what we see in the case of a closed book:

(a) I see the book
(b) The book has pages inside.
(c) I see the book as having pages inside,
(d) I do not see the inside pages.
(e) The book has a back cover.
(f) I see the book as having a back cover.
(g) I do not see the back cover.

I believe this distinction clears up an equivocation about 'seeing'. The 'seeing' of (d) and (g) is about images; the 'seeing' of (c) and (f) would more aptly be called 'thinking'. I should add, with Young [cite:@jmyoung:kantimg page 142], that the imagination may well invoke images; it is just not limited to this activity.

Having laid out, roughly, the powers and responsibilities of the imagination in cognition, we can tell a story about the free play of cognition that leads to aesthetic judgment.

The imagination interacts with the understanding by “proposing” candidate objects to the understanding. The understanding then tries to resolve these “proto-objects” into bona fide objects by bringing their properties under pure concepts. The understanding may well arrive at a determinate concept of an object, e.g., as being a chair. But although this concept may afford pleasure in the perfection of the chair, it cannot serve as the basis of a disinterested appreciation of the chair's beauty. A chair will never be beautiful, we might say, qua chair. Beauty begins only when the chair is presented by the imagination as possibly representing something over and above a chair, and it is only because the understanding also fails to make good on this potential representation that we are left to reflect pleasurably on our own cognitive effort.

The imagination, operating freely, continually reshuffles attributes and redraws the boundaries of and within objects, proposing a range of interpretations. Its capacity for free play would seem to reside in its hyperactivity, which (as an evolutionary feature of cognition) would tend to maximize its success rate in appraising unfamiliar objects. The imagination might run wild (and perhaps does, in pathological cases) but for its subordination to the understanding. In the cognitive case, the interplay between imagination and understanding terminates when the understanding “approves” the imagination's presentation of an object by successfully resolving it under a concept. This process, too, may iterate before halting—e.g., I may think I am approaching someone I know before realizing that it is a stranger, or I may believe I see a bogeyman in my closet at night before I make out the features of a hockey mask I remember having stored there. The cognitive search may also branch into more than one successful path, as when I discover that I can interpret the same image as representing a rabbit or a lady, but not both in the same judgment. In purely aesthetic reflection, on the other hand, where no conclusive concept is available, the process goes into a loop. Each input from the imagination is rejected by the understanding, whereupon the imagination queues up another presentation.

Whether the understanding finds each proto-object sufficiently thing-like to interpret or, like the imagination, it just can't help itself, its circuit with the imagination just is the purposiveness we detect in aesthetic contemplation. It may be more precise to say that the purposiveness resides in the imagination's presentations, which would suggest that the presentations must meet minimal requirements, such as boundedness or coherence, to avoid being rejected out of hand by the understanding (and so leading, perhaps, to judgments of the sublime). The imagination tries again and again to construe the object as “this kind of thing” but gets trapped between sensibility and understanding, because its construals can't be brought under the categories. The presentation goes beyond mere sensation, and the pleasure we feel in it is more than just pleasure in a sensation. This intermediate pleasure is the reflection that our cognitive faculties are engaged in a free exercise of power. The form of the imagined object (or proto-object) strikes the understanding as suitable for at least an attempt at resolving it under concepts. In a cognitive judgment, some concept or other would be found to cause the object as such. In an aesthetic judgment, however, this final step never arrives, yet the purpose-granting engine of the understanding has been engaged, and we like this.

As for the deduction proper, if we buy Kant's account of the imagination and understanding, and if we agree that there are certain experiences that we enjoy in a contemplative, disinterested way, then his deduction is simply the recognition that the faculties of cognition operate similarly for everyone, albeit subjectively. This is not farfetched, if we accept as universal other faculties lying on either side of the subjective/objective divide which the imagination occupies, that is, if we accept reason, on the one hand, as granting knowledge based on universal and necessary concepts, and perception of, say, colors, on the other hand, as producing consistently subjective results from person to person.