I. Introduction

IN CHAPTER ONE I developed a notion of intuition, and of the distinction beween matter and form in intuition, that will serve as the basis of my account of the application of concepts to intuitions. I provided some evidence that the interpretation accords with Kant’s intentions in the Aesthetic, but argued that the distinction is independently defensible. I also considered the claim that sensation is the matter of sensory intuition. I argued that this claim is compatible with the view that something else serves as matter in intuition as well. I suggested, but did not try to show, that, at least for Kant, something that we might call “imaginative anticipation and retention” will need to be what serves as such material. It will need to do so in order that it may thereby be available for forming into the concepts that one is said to “apply” to intuitions. Beginning with Chapter Three, I show that this suggestion is not only consonant with Kant’s claims in the Logic, but is required in order to make sense of what is otherwise most obscure in it. The present chapter offers additional clarification of the suggestion. Section II does this by relating the view to some issues of concern to philosophers of a more behavioristic or functionalistic stamp. Section III adds some additional points of reference by showing how the proposal relates to some notions developed by other philosophers in three very different philosophical traditions.

II. Conceptual Material and a Subject’s Dispositions

Whatever we make of Kant’s pronouncements concerning the role of imaginative anticipations and retentions in conception, everyone presumably agrees that, at least like a certain sort of anticipation and retention, some of a subject’s “dispositional” states are essential to its possession of conceptual abilities. Some readings of Kant go so far as to identify concepts with dispositions. As it stands, this runs afoul of Kant’s claim that concepts are the “predicates” of possible judgments (A69/B94). This seems to imply that they are not, for example, merely dispositions toward the making of judgments. In any event, the question would remain as to the nature of judgments themselves.

The irrelevance of the suggestion in this respect might suggest a variation on the dispositional approach, namely, the view that concepts are multiply instantiable aspects of (possible) representations; that it is in virtue of these aspects that the latter ever are (if intuitional in nature) judgmental with respect to objects (i.e, in virtue of which they ever are Erkenntnisse); and that in turn these aspects play their essentially “predicative” role only in virtue of being appropriately connected with a subject’s dispositions. According to this approach, specification of the relevant dispositions, and the relevant connection, would ipso facto clarify the nature of judgment itself. To a certain extent, this is my own proposal. But there are various ways in which the “connection” might be construed.

A purely functionalist approach might seek to establish the connection with a subject’s dispositions purely in terms of the “typical causes and effects” of the relevant predicative aspects of mental states.¹ We might propose, for example, that any feature tokenable or instantiable as a “predicate” in some state derives its conceptual status from its purely causal connections with other features tokenable or instantiable in states. Apart from the feeling that this ignores the crucial notion of consciousness (e.g., ignores the problem of the very consciousness of the connections in question), this may seem especially plausible as an elucidation of Kant’s insistence that concepts are predicates of possible judgments only to the extent that they serve a function of “rule-governed synthesis” among possible representations. The causal connections, in the proposed view, would simply define the rules of synthesis. The proposal may also seem plausible because capacities and dispositions may in their own turn seem eventually reducible in terms of the causal powers of internal states of a subject.

The suggestion would, of course, need refining. Any truly conceptual state of a subject—apart, perhaps, from the state of idly having some content “in mind”—always involves additional capacities and dispositions, over and above whatever may be involved in the constitution of a content as such in the first place. Otherwise, there would be no difference between the mere presence of a content in mind, like idle doodling, and what a subject actually happens to be thinking at any moment.² But this, it may be claimed, is no problem for a functionalist. The additional capacities and dispositions may be regarded as so many additional “causal powers” of a content, on an occasion of its presence in consciousness, over and above those involved in its constitution as an in principle abstractly presentable content in the first place.

As an interpretation of Kant, there would seem to be an obvious problem in the proposal. For one thing, functionalism appears to get things backward. This is because, for Kant, causes and effects, at least insofar as they are possible objects of knowledge, are essentially governed by systems of causal laws. So according to a functionalist approach, the very idea of conceptual content would have to presuppose the idea of a system of causal laws naturally obtaining. But in Kant’s own thinking, the very idea that a system of laws naturally obtains seems to be derivative from the idea of a being who is capable of representing determinate sorts of objects in the first place (B164).

To be sure, Kant does not maintain that pure understanding prescribes particular laws. Apparently, it merely prescribes the general form of lawfulness in nature; it is up to the empirical (“impure”) understanding to find what the laws are (B165). Nevertheless, it is difficult to see, in Kant’s view, how the very idea of a nature that is subject to particular laws can be anything other than the idea of a realm of appearances that is in some way the correlate of a subject capable of judging in a certain way, namely, of a subject who is capable of taking certain sequences of appearances as “necessary.” I argue in Chapter Four that the apprehension of a causal order is in fact based, for Kant, upon an apprehension of the intentional correlate of one’s own anticipations and retentions in experience. If this is so, then any account that attempts to explicate the notion of representational content in terms of causal roles will be circular. The relevant concept of causality would already contain an implicit reference to the possibility of representation.

A second consideration concerns Kant’s insistence on the “spontaneity” of acts of the understanding: “the mind’s power of producing representations from itself, the spontaneity of knowledge, should be called the understanding” (A51/B75; cf. B130, B162n). An act of spontaneity is one that must be regarded as expressing a causality other than according to causal laws. It must be regarded as expressing a “free” causality, wherein something is determined to occur, but not simply as a law-governed consequence of antecedently obtaining conditions (A444-7/B472-5, A533/B561). This is contrary to the kind of causality presumed to be in question in functionalist accounts of the mind, or at least in functionalist accounts of a purely causal sort.³

Kant develops his notion of spontaneous causality primarily with respect to the problem of intentionally performed actions. In that case, he specifically attributes acts of spontaneity to “reason” as distinct from understanding. But the only difference seems to be that acts of the latter are limited to the conceptualization of sensible appearances (and their form). Apart from that, understanding’s operations are as free as those of reason. They are, qua acts of the understanding, limited by nothing other than the nature of understanding itself:

As a pure spontaneous faculty [reason] is elevated even above understanding. For though the latter is also a spontaneous activity and does not, like sense, merely contain conceptions which arise only when one is affected by things, being passive, it nevertheless cannot produce by its activity any other concepts [emphasis added] than those which serve to bring the sensuous representations under rules, and thereby to unite them in one consciousness.⁴

It is difficult to be sure what Kant means by the spontaneity of understanding. It is particularly difficult in view of the fact that he concedes that all events are necessitated by causal laws. The only difference is, apparently, that we need to distinguish two points of view with respect to any event that expresses spontaneity: the same event that is subject to causal laws may also be regarded as expressive of spontaneous causality. It is not clear what the qualification is supposed to amount to. However, Kant seems to be distinguishing precisely between the conceptual and other aspects of conceptual states, and to be denying that the former could be determined by causal laws. In the passage just quoted, and as I emphasized in quoting it, he does not simply speak of conceptual states as products of spontaneity. It is the very concepts that are involved in such states that are products of spontaneity. Consider also the following passages:

Our cognition springs from two fundamental sources of the mind; the first is the capacity of receiving representations (receptivity of impressions), the second is the power of cognizing an object through these representations (spontaneity of concepts). (A50/B74)

Concepts are based on the spontaneity of thought, sensible intuitions on the receptivity of impressions. (A68/B93)

As we shall see in more detail in subsequent chapters, Kant is serious in proposing that a spontaneous faculty of understanding is responsible for the “production” of concepts themselves, not simply for the production of states or events that embody, token, or instantiate them. So far as I can see, this can only mean that the conceptual dimension of a subjective state cannot be a matter of that state’s involvement in some network of typical causes and effects.

This may be thought to read too much into Kant’s claims about spontaneity. Perhaps his point is that conceptual states are (or embody) actions. If conceptual states are actions, then, analogously to the case of ordinary intentional actions, it would follow that the occurrence of a conceptual state could never be adequately comprehended in terms of the working of causal laws. The point can be granted, it seems, without raising any question concerning the “production” of the very concepts involved in conceptual states. It only concerns their character as “actions,” not their specifically representational dimension. But consider the following passage:

Should one ask whether understanding itself is determined, either in itself or [emphasis added] with respect to how it acts or fails to act [an sich selbst so wohl als in Ansehung dessen was er thut oder unterlässt], then we must say: there is no possible experience that can show this, because that would always be mere appearance. Understanding is itself no object of sensory intuition.⁵

The point seems to be that the spontaneity of acts of understanding is not simply their spontaneity as actions of some kind. It is their spontaneity precisely as acts of understanding. It is difficult to see what this distinction is, if it is not between a spontaneity involved in the factual occurrence of conceptual states on occasions and one involved in the original constitution of their conceptual content. In any case, to the considerations so far adduced, I would simply add the following: If anything like the functionalist view really were Kant’s, then it is difficult to see why he found it so difficult to be clear. In particular, it is difficult to see why so much of his effort, and so much apparent confusion, concerned the relationship between “imagination” and conceptual representation. If all that Kant had in mind, in his notion of “spontaneity,” were explicable in causal terms, then it ought to have been much easier to have said so.

Apart from interpreting Kant, the approach is independently objectionable. It is not plausible to suppose that the dispositions one ascribes to a subject, in ascribing conceptual states to it, are without circularity reducible to powers for affecting and being affected possessed by those states themselves. A recent example of Stephen Stich’s seems to me to make this point, though Stich himself employs the example for a somewhat different purpose.⁶ The general point is this: that whether or not we are prepared to judge that the causal powers of a state of a person (or of some feature of a state of a person) have remained unaltered throughout an undeniable course of altering the subject’s condition depends solely on our judgment concerning the typical causes and effects that this state would continue to have, given the satisfaction of certain antecedent conditions. It does not hinge, in general, on a judgment as to whether or not the alteration has undercut the very possibility of those conditions being satisfied in the first place. For example, if a subject is no longer able to access certain stored memories, then that subject may be judged to lack a portion of whatever capacities and dispositions are normally associated with judging that McKinley was assassinated. But that is perfectly compatible with supposing that the “content” originally constitutive of judging that McKinley was assassinated continues to be such that, were the access in question regained, then the typical results would continue to follow from the tokening or instantiation of that content. And it is perfectly compatible with supposing that relevant additional states remain such that, were further conditions satisfied, then that content would in its turn arise as a typical effect of them. The point is that there is no reason in principle why our judgment concerning alterations in the “causal powers” of states (or of their features) needs to coincide with our judgment concerning alterations in the capacities and dispositions ascribable to the subject of those states.

The argument does not rest on supposing that a purely causal approach, in terms of internal states rather than subjects as wholes, must fail to capture some special sense of “agent causality.” Even if all causality is ultimately state- or eventcausality, the point is that there is no reason in principle why our judgments concerning a subject’s capacities and dispositions need to be secondary to judgments concerning the capacities and dispositions of a subject’s internal states. However we eventually analyze the latter notion, we should always be free to rule, in any particular case, that a subject’s capacities and dispositions have undergone impairment, but its internal states have the same “causal powers” as before. Nothing in the concept of a state’s causal powers rules that out.

One may object that I am adopting too narrow a conception of the causal powers attributable to the states of a subject. According to various standards, that may be. But there is only one way to make the objection work in the present context. We would have to insist that our judgments concerning the causal powers of states, at least in cases that involve the ascription of conceptual contents to states, have got to be brought into conformity with our ordinary ascriptions of certain sorts of capacities and dispositions to the subjects of those states. But that, in effect, would be to concede at least an important part of the point in dispute. The conclusion would then be unavoidable that, whatever else we may say about the conceptual states of a subject, they must be states in which a special sort of “material” is essentially ingredient. They must be states in which a sub-set of the very capacities and dispositions of the subject itself is ingredient, or at least in which their “categorical basis,” as such, is.

Even if it is not the case that the capacities and dispositions of a subject, of the sort involved in the ascription of conceptual content, are only circularly reducible to the causal powers of states of a subject, the argument presented in the preceding chapter, against a purely “formal” approach, should continue to tell against a specifically functionalist view of content. That argument concerned the need to connect the conceptual content of a state with the object-directed character of that state. Even if we conceded that everything essential to conceptual content, considered just in itself, was capturable by reference to the powers of cognitive states (or of their features), the problem would remain of explaining why such content should ever be regarded as directed toward the objects presented in particular intuitions. According to a functionalist account of conceptual contents, as causally functional features of possible states, what could possibly give us anything more—when those features are in fact tokened by object-directed states—than the case of a text that merely happens to be inscribed on an arrow? The answer, I have argued, is that insofar as contents are directed toward objects, the states that apprehend those objects must do so precisely through whatever material is constitutive of the contents themselves. Even if it were the case that such material is describable in causal terms, the fact remains that it is at most the material out of which conceptions may be formed. Its mere tokening or instantiation in would be cognitive states is not, except in that limited sense, constitutive of the representational content of any state.

In any event, nothing stands in the way, on the alternative that I propose, of equating the basic “material” of conception with various of the dispositions of a subject (or with their categorical basis). Kant’s own way of seeing the point is, I have suggested, simply to recognize the role of “anticipations and retentions” as material for the forming of concepts. To this extent, the proposal leans more toward behaviorism than functionalism. But there is also a sense in which it leans more toward the latter. For the proposal also insists that what constitutes conceptual content must be “tokenable or instantiable,” not simply within the subject of intuitional states, but as an actual feature of those very states. In order to serve as material for predication in intuition (whether sensory or imaginative), it must also serve as material in intuition.

III. Some Comparisons

The extension of the Kantian framework of matter and form, which I shall argue is at least implicit in the Transcendental Deduction, yields a view that is both like and unlike some views about the intentionality of consciousness that have been offered by other philosophers. I begin with attention to a philosopher who would in fact reject some essential aspects of that framework. Despite this difference from Kant, the suggestion that I have put in terms of a distinction between matter and form in both intuition and conception, bears some similarity to Searle’s account of the co-determination of content against a non-intentional Background.⁷

Searle accepts the irreducibility of intentional content, as a feature of mental states. But he rejects, in effect, the notion of intuitional directedness that I espoused in Chapter One. Since it is irreducible, Searle does not try to define the notion of intentional content. Rather, he simply characterizes one central type of intentional content and suggests that all intentional states either possess an instance of it or else involve more complex structures, the basic components of which are states that possess such content. (Searle also distinguishes the “direction of fit” in which a content may be present and phenomenological “feelings” that may be characteristic of a state. But the latter do not serve as “material” in the way that sensations, for example, do for Kant.) The central case is that in which content “determines” a mental state’s “conditions of satisfaction.” In the case of belief, for example, it determines the conditions under which the belief would be true; for desire, the conditions under which the desire would be satisfied.

We may concern ourselves primarily with what Searle says about “experiences” (in the technical sense in which he employs that term). For him, experiences have an intrinsic intentional content. Here is what he says about visual experiences:

. . . the two forms of mental phenomena, belief and visual experience, are intrinsically Intentional. Internal to each phenomenon is an Intentional content that determines its conditions of satisfaction. The argument that visual experiences are intrinsically Intentional, in sum, is that they have conditions of satisfaction which are determined by the content of the experience in exactly the same sense that other Intentional states have conditions of satisfaction which are determined by the content of the states. (P. 40)

The intentional content of a visual experience is an internal and irreducible feature that “determines” the conditions that would need to be satisfied in order for that experience not to be hallucinatory. It is important to avoid regarding this notion as trivially obvious. It seems trivial to claim that there are conditions that need to be satisfied in order for an experience not to be hallucinatory, and that these conditions have something to do with the internal features of an experience. But Searle’s claim goes beyond this. It says, if I might so put it, that it makes a difference to an experience itself whether or not it is an hallucinatory one. It is not just that we possess a concept of success or failure, with regard to possible experiences, and that this concept concerns some connection between the internal features of an experience and the actual state of the world. It is rather that the experience itself “has to determine what counts as succeeding” (p. 38). This cannot, of course, mean that an experience actually lays claim to being a veridical experience. That would imply that a belief or judgment needs to be in question; furthermore, Searle maintains that one’s beliefs and judgments might in fact contradict the content of a visual experience (p. 55). But it seems to imply that, in some sense, an experience at least essentially contains a (non-linguistic [p. 49]) “description” of some possible state of affairs, and a representation (or, as Searle prefers, a “presentation”) of the world as fitting the description.⁸ However, the notion that an experience contains descriptions, in a certain “direction of fit,” cannot be a basic notion for Searle. It is itself to be explicated in terms of the fact that experiences contain contents “determining” conditions of satisfaction.

In some ways, Searle’s experiences may seem comparable to sensory intuitions in Kant. But there are differences. First of all, Searle rejects any attempt to elucidate the notion by appeal to a primitive notion of experiential “directedness.” To the contrary, the notion of directedness needs itself to be explicated in terms of that of a state’s conditions of satisfaction (p. 39). The same goes for any other attempt to make sense of the idea of apprehending (possibly non-existent) “intentional objects,” whether in the case of mere belief (pp. 16-17), for example, or in that of “imagining” objects (where the intentional content, for Searle, is in effect always reducible to that of merely imagining-that: p. 18). The point is that whatever it is that determines experiential “directedness” must be, for Searle, just the same sort of thing that determines the content of, say, a non-experiential belief or desire. It is this, in turn, that needs to be seen as an irreducible aspect of mental states.

According to the Kantian approach, recognition of primitive directedness in at least certain sorts of states—in those that are “intuitional” (i.e., sensory states and certain instances of imaginative awareness) — does not of itself preclude recognition of states that lack such directedness and yet still possess “content.” But if there are such states, then, in the view that I shall elaborate, their intentional contents would not be primitive features of those states. Rather, to whatever extent they are anything more than dispositional aspects of those states, they would have to be “constituted” features, that is, features that are in a sense formed by mental activity (though not necessarily by “action” in any purposive sense of that term) out of material internal to the states in question. In addition, such material must, at least potentially, be material for intuitionally directed states as well. In the Kantian view, that is, nonintuitional intentional “content,” if there really is such a thing,⁹ fails on two accounts to be primitive. First, it is essentially “constituted” content. Second, its possibility is parasitical upon that of intuitionally directed content. In any case, I have already rebutted the supposition—in fact, Searle’s own—that the recognition of primitive directedness entails commitment to unusual entities (p. 17).

If they are not quite intuitions, in Kant’s sense, then it may seem best to regard “experiences” as the upshot of some kind of embodiment, or some kind of “realization,”¹⁰ of concepts in the medium of something like Kant’s “sensations.” But Searle himself is unclear as to the connection between content and “concepts.” The most that he seems to say is that intentional content is usually connected with, or affected by, one’s possession of certain concepts (and, a fortiori, by beliefs and judgments that utilize those concepts). In certain cases (e.g., of “seeing-as”),

. . . one wants to say that a certain conceptual mastery is a precondition of having visual experience; and such cases suggest that the Intentionality of visual perception is tied up in all sorts of complicated ways with other forms of Intentionality such as belief and expectation, and also with our systems of representation, most notably language. Both the Network of Intentional states and the Background of non-representational mental capacities affect perception. (P. 54)

As we shall see more clearly presently, the relation between such states and capacities, on the one hand, and intentional content, on the other hand, is in an important sense an external one for Searle. To be sure, he speaks of such material as “reach[ing] inside the Intentional content to determine” aspects of a state’s conditions of satisfaction (P. 66). But all that this can presumably mean is that such material helps in many cases to determine a state’s conditions of satisfaction. (So even in the case of an actual influence from one’s “conceptual masteries,” it would not necessarily follow that the relevant intentional content is itself conceptual in nature. In fact, Searle does not commit himself as regards the need for some minimal extent of such influence.)¹¹ In any event, even when conceptual masteries help to determine conditions of satisfaction, it seems to remain the case that intentional content is an irreducible feature of the state whose content it is. So there could be no sense in which it is itself even partially “formed” out of such masteries in the first place. (In the view that I am ascribing to Kant, the possibility of an intuition that is lacking in conceptual content cannot be excluded either, though Kant himself has a tendency—to which he does not consistently adhere—to suppose that such states must at least be devoid of “consciousness.” As for questions concerning the “objects” of such depleted states, the proposal is in fact compatible with conceding a considerable indeterminacy. At most, it would seem we can say, any answer needs to take the form of the following: The state in question is directed toward some possible region of space/time that might, under such-and-such circumstances, be conceptualized in this way or that.)

Now despite Searle’s rejection of primitive object-directedness, his view of the role of Background material, in the determination of conditions of satisfaction, is comparable to my suggestion concerning the distinction between matter and form in the constitution of content. As we have seen, what Searle himself argues is that at least some contents determine conditions of satisfaction only “against” a Background of non-intentional material:

An Intentional state only determines its conditions of satisfaction—and thus only is the state that it is—given its position in a Network of other Intentional states and against a Background of practices and preintentional assumptions that are neither themselves Intentional states nor are they parts of the conditions of satisfaction of Intentional states. (P. 19; cf. pp. 54ff, 141ff)

The following are some examples: (1) the “preintentional stance toward the world” involved in the fact “that I recognize degrees of the hardness of things as part of ‘how things are’”—for example, that one can intend to peel an orange but not a car, but without having any kind of belief that one can do the former but not the latter (p. 144); (2) the kind of “understanding” that allows one to navigate among the various senses of “open” in “The chairman opened the meeting,” “The artillery opened fire,” and “Bill opened a restaurant,” and to reject as nonsense such sentences as “Bill opened the mountain,” “Sally opened the grass,” and “Sam opened the sun” (pp. 145-46); (3) the kinds of “associations” that allow one to distinguish metaphorical from non-metaphorical language, and to distinguish intentions in metaphor, but which are not formulable in terms of one’s grasp of linguistic rules (p. 149); (4) whatever it is by virtue of which the “expert skier is flexible and responds differently to different conditions of terrain and snow,” given that it could not possibly involve either conscious or unconscious consultation of rules or conscious or unconscious calculation of any sort (p. 150). An additional example, inspired by Searle, is the following: (5) the sense in which, at least ordinarily, one’s “habit” of using a desk-top to sit on precludes the very possibility of certain anticipations, upon entering one’s office, yet without there actually being a belief that these anticipations are justified.¹²

As Searle notes, it is difficult to describe Background material in a way that does not seem already to entail its structuring as intentional content. Searle himself speaks of the Background as comprising “assumptions,” “presuppositions,” and “presumptions” (p. 156), and even, as we have seen, “recognition” and “understanding.” In most of the examples, as well, it would seem in order to speak of a subject’s “expectations” or “anticipations.” Searle objects to this terminology (p. 157), though he does speak of “associations.” In any event, we need to avoid supposing that what is in question is something that a subject could formulate, even if asked to. And of course, what is in question cannot be “beliefs,” in Searle’s sense, though they might always be described as beliefs in some other sense. In addition, Searle also describes the Background material as “nonrepresentational” (p. 143). But, of course, Kant need not be limited in this way. For Kant, “representational” (e.g., sensations) states do not always possess intentional content. Searle himself offers no account of the general nature of Background material, although his inclination to describe it by reference to “capacities” and “practices” (p. 156), and in terms of the vocabulary of “knowing-how” (p. 143), may suggest a leaning toward some kind of dispositional approach.

In the view that concerns us, the anticipations and retentions that serve as material in the formation of intentional content, for Kant, must share at least two features with the material in Searle’s Background. First, they cannot, on pain of circularity, themselves possess content in the sense in question (though they might in some other). Second, they must nevertheless contribute to the content of certain intentional states. Unlike Searle, however, Kant will regard such content as wholly constituted out of such material. (At least, this must be the case for a basic set of contents. Once formed, these may in turn provide material for the formation of additional contents. I generally abstract from this complication.)

Apart from the fact that, according to Searle’s approach, intentional content is not originally “formed” from the material contained in the Background, the most important difference between the approaches is connected with the issue of primitive directedness. The distinction between matter and form is dependent upon that notion. The matter of an intuition, as we have been considering it, is the set of states through which an intuition is directed. That is, it is the set of states that, by virtue of being ingredient in an intuition, make some difference in the way in which that intuition is directed. (This is, of course, compatible with conceding that other factors, besides ingredient states, can make a difference in the way in which an intuition is directed. For example, it says nothing about the role of concepts, which are at most formed out of ingredient material.)

What the approaches have in common is their recognition that non-intentional states always influence the way in which intentional states are directed. But the problem with Searle’s approach, from a Kantian perspective, is that it leaves it entirely mysterious how this could possibly be the case. How are the states in the Background in fact able to “reach inside” an intentional content, in order to bear on its determination of conditions of satisfaction? Searle himself describes the material in question as constituting a part of one’s “stance” toward the world. But it would seem unable, in his view, ever to form an actual part of one’s intentional stance. If it could indeed have an influence upon one’s intentional stance, it seems never to be an actual constituent of that stance. According to the Kantian alternative, the contrary is the case. This is because, in that view, the material in question is the very material out of which intentional stances are formed in the first place.

While Searle does speak of Background material as “reach[ing] inside” of content, he also puts the point in different terms. Background material, he says, simply “provides a set of enabling conditions that make it possible for particular forms of Intentionality to function” (p. 157). Here, the conditions in question are explicitly causal conditions—though merely enabling, not causally determining, ones (p. 158). Searle’s view therefore seems to be this: in an entirely non-causal sense, intentional contents “determine” the conditions of satisfaction for the states whose contents they are; Background material, in turn, has at most a causal influence on which contents are able to be the contents of which states in the first place.

Apart from the apparently Cartesian difficulty as to how such material might have an influence on the instantiation of primitive properties of an entirely different order,¹³ the proposal involves an implausibly external approach to the relation between intentional content and non-intentional material. Surely, the manifold of one’s (non-intentional) “expectations” and “stances” is as much a constituent of one’s visual experience, say, in skiing, as are the “phenomenal qualities” in which those experiences are “realized” (p. 45). Though they are not reflectively present, nor even such that they might be articulated with effort, they form an integral part of the way in which the skier is seeing. In some special and immediately phenomenological sense, the intentional content of a visual experience determines, for Searle, what one sees things “as” (pp. 50-57). Therefore, the expectations and stances in question should be a part of the very content. But in Searle’s own view, they cannot be. At most, they could exercise an “enabling” influence on one’s ability to see things “as” something or other. The advantage of the Kantian view, in addition to its recognition of primitive intuitional directedness, is precisely that it helps overcome this apparently arbitrary, and troublingly Cartesian, dichotomy. But then, as we have already seen, these two points are really just aspects of one.

Despite these differences between Searle’s approach and the one I have said is Kant’s, both Searle and Kant take intentional content to involve an irreducibly mental function. In addition, both recognize the role of non-intentional material in the performance of that function. I now want to consider two philosophers whose thought is directly influenced by Kant himself. The first of them is Husserl.¹⁴ For our purpose, the most obvious difference between Husserl and Searle is that the former relies on a primitive notion of intentional directedness (i.e., directedness toward some “object” that may or may not be real). This is, of course, not equivalent to Searle’s notion of a mental state’s possession of intentional content. That is, it cannot be equated with the notion of an intrinsic character by virtue of which a state has “conditions of satisfaction.” In Husserl’s view, rather, appeal to primitive directedness would be needed in order to make any sense of Searle’s notion of content in the first place. For Searle, as we have seen, the reverse is the case: any useful notion of object-directedness must be explicated in terms of the notion of content.

In the “pregnant” sense of the term consciousness, Husserl tells us, the relevant notion of being object-directed is simply that of having “consciousness of something.”¹⁵ More explicitly:

Under intentionality we understand the own peculiarity of mental processes “to be consciousness of something”. . . . [For example,] a perceiving is a perceiving of something, perhaps a physical thing; a judging is a judging of a predicatively formed affair-complex. . . . In every actional cogito a radiating “regard” is directed from the pure Ego to the “object” of the consciousness-correlate in question. . . . ¹⁶

To be sure, Husserl draws some important distinctions concerning the “something” toward which consciousness is intentionally directed. Apart from the obvious distinction between directedness as such, and directedness toward really existing objects (and obtaining states of affairs), the most important is that between “actional” or “wakeful” directedness and directedness toward what is merely in the background of consciousness.¹⁷ It follows from the latter distinction that the (complete) “intentional object” toward which consciousness is directed cannot simply be an object (not even a possible object) in the usual sense of the term. It would seem rather to involve a structure that, in a more or less determinate way, contains a vast number of actual and possible objects. This may seem to constitute a difference from Kant as I have presented him. I shall not be concerned to settle this question, although I do think that my approach to Kant implies that, whether he did so or not, Kant was at least committed to recognizing the notion of an “object” of consciousness in a sense that is broader than that of an object of “intuition” as I have so far characterized that notion. That is, Kant needs to recognize a notion of intuitional directedness that involves more than that of directedness toward possible regions of space and regions of time (and even more than whatever holistic awareness of these latter had already been seen in the Aesthetic).¹⁸

Husserl introduces the term noema in order to indicate the full “object” with both of the qualifications that I have already noted—that is, (a) in abstraction from the question of the reality of the “object” in question, and (b) inclusive of aspects and features, and of whole realms of unexplored possibilities, extending well into the “background” of consciousness as usually understood.¹⁹ On some readings, the noema is not to be understood in such terms at all. Husserl’s view is sometimes taken to be more like Searle’s: attention to an act’s noema is simply attention to a quasi-Fregean meaning-content that determines its intentionality.²⁰ Perhaps, according to this approach, we might still say that attention to an act’s noema is attention to what determines its “object-directedness.” But if we do, then we should not regard this as clarificatory of the notion in question. That would get things backward: the very notion of intentional directedness, according to the Fregean approach to Husserl, needs to be explained in terms of that of intentional “content.” It is impossible to try to settle this issue here.²¹

The nearest similarity between Husserl’s and Kant’s approaches to intentionality may seem to concern the relation between intuitional directedness in Kant and that particular structure of the noema that Husserl refers to as its “central point” (or rather, as the central point of its noematic “core”). Husserl speaks of that central point as the “object” or the “determinable X” in the noema. He does this despite the fact that, as we have already seen, the noema as a whole is regarded by him as the “intentional object” of consciousness in its own right:

Each consciousness has its What and each means “its” objective something; it is evident that, in the case of each consciousness, we must, essentially speaking, be able to make such a noematic description of “its” objective something, “precisely as it is meant”; we acquire by explication and conceptual comprehension a closed set of formal or material, materially determined or “undetermined” (“emptily” meant) “predicates”. . . . The predicates are, however, predicates of “something,” and this “something” also belongs, and obviously inseparably, to the core in question; it is the central point of which we spoke above. It is the central point of connection or the “bearer” of the predicates. . . .²²

It may appear natural to identify the central X with the “referent” of a Kantian intuition (that is, with the “referent” of a Kantian intuitional state.)²³ Correspondingly, the act of reference to, or the apprehension of, the noematic X may seem most naturally equated with Kant’s notion of intuitional directedness as such. To put the suggestion differently, the distinction between directedness toward the X and directedness toward it precisely as the “bearer” of certain predicates may seem most naturally equated with Kant’s distinction between intuition and predication in intuition. However, it is not possible to draw the comparison in quite these terms. This is because Husserl seems to intend the distinction to have general application, therefore not to be limited to what I have taken be distinctively intuitional awareness—sensory awareness and the imagining of objects.

A further difficulty may appear to lie in the fact that, as seems clear from the continuation of the passage just quoted, directedness toward a noematic X is always the intentional correlate of directedness toward an in principle re-identifiable object. Of course, Husserl could not mean to suggest that any object that is so represented must in fact be a re-identifiable piece of reality. In the case of a sensory noema, it might be a mere hallucination, for example. In another sort of case, it might be a purely fictional character, at most re-identifiable in an appropriately imaginary context of discourse. But in either case, the X would seem to be, for Husserl, the intentional correlate of an instance of directedness toward what is at least represented as an in principle re-identifiable object. In apparent contrast with this, purely intuitional directedness, for Kant, is neutral with respect to the question whether the “object” in question is in fact so represented.²⁴

This difficulty need not stand in the way of equating the apprehension of a noematic X with intuitional directedness in Kant. In fact, Husserl himself does not regard the notion of apprehending a “re-identifiable object” as at all incompatible with that of apprehending what is, and even what one takes to be, purely hallucinatory. In the case of sensory apprehension, for example, Husserl distinguishes between “objects” as mere “spatial phantasms” (Raumphantome) and as concrete material things; he elaborates at considerable length on the various levels of apprehension wherein the former are finally apprehended as (or “constituted” as) the latter. Throughout, the point is that, at each of these levels, we may continue to speak of an appropriately correlative notion of “object-identity.”²⁵

Suppose one is hallucinating a pink elephant, and takes oneself to be doing so. Without having to regard the object in question as anything more than hallucinatory, we may still speak of its potential re-identifiability. For example, as one attends to the hallucinated animal, it is perfectly in order to suppose that one has observed the animal to move in various ways. In other words, we are entitled to say that it has moved, without having to abandon the supposition that it is not a real thing at all. The presence of an X in the corresponding noema is presumably the correlate of this possibility. A parallel point could be made in regard to the purely imaginative case.²⁶

It seems clear that Kant has little interest in this level of phenomenological description. The level at which he is concerned with the possibility of identifying and re-identifying objects of intuition is precisely that at which those objects are to be identified as real things. All the same, nothing in Kant’s view prevents equating the apprehension of the X in a Husserlian sensory (or imaginative) noema with Kant’s own notion of intuitional directedness as such (that is, with his notion of directedness, considered apart from the real thing-hood of the “object’’ in question). In fact, Kant’s own argument in the “Anticipations of Perception’’ (A166ff/B207ff) seems to rely on this notion. He tells us, first of all, that the purely “mathematical’’ categories—which are what is in question there—are concerned merely with the intuition of appearances, not with their existence (A160/B199). But the whole point of the argument then is that, despite this limitation, we can imagine any given appearance as remaining the same, except for an internal change that Kant takes to be correlative with the possibility of diminishing the presence of sensation in any intuition apprehending it. (At the point of diminution to zero, one would simply be merely imagining some possible region of space and no longer sensorily—not even hallucinatorily—apprehending one).

Appearances, as objects of perception, are not pure, merely formal intuitions, like space and time. . . . [T]hey contain, that is to say, the real of sensation as merely subjective representation, which gives us only the consciousness that the subject is affected, and which we relate to an object in general. Nor from empirical consciousness to pure consciousness a gradual transition is possible, the real in the former [n.b., not “reality” in the usual sense] completely vanishing and a merely formal a priori consciousness of the manifold in space and time remaining. (A166/B207-8)

Of course, this consideration could at most permit an equation of intuition, as such, with a special case of the apprehension of the X in a Husserlian noema, namely, with the case in which the noema in question is the correlate of sensory or imaginative apprehension. It remains unclear how to construe the X-structure in cases other than the sensory or imaginative. It is not sufficient to recall, for example, that we are even then supposed to be concerned with the intentional correlates of modes of consciousness, and that—from a phenomenological perspective—what one is conscious of is (as a noematic “whole”) always supposed to contain some difference correlative with every single difference in the intentional state itself. Given this, and assuming that there is indeed such a thing as a state of consciousness in which one merely thinks an object, but is not intuitively directed toward it (not even imaginatively)—the most we could say is that, for Husserl, there must be some noematic difference correlative with this fact. But that difference could at most be correlative with an element in the thought that there somewhere is (or at least might be or have been) an object of a certain sort. Surely, Kant’s intuitional directedness is not a special case of this. But then, by the same token, it seems impossible to regard the intuitional X-structure, in Husserl’s own view, as a special case of it either. In light of this difficulty, the more fruitful comparison might therefore be, not between intuitional form in Kant and (the noetic counterpart of) Husserl’s intuitional X-structure, but rather between the former and (the noetic counterpart of) the whole noema in Husserl. In that case, we would simply have to qualify the comparison by noting that Kant himself had a much narrower conception than Husserl did concerning the “objective” correlate of intuitional directedness.²⁷

We are finally ready to turn to the role of non-intentional elements in the determination of “intentional content.” Like Searle, Husserl is emphatic regarding the role of such elements. Unlike Searle, he formulates his account both in terms of the distinction between matter and form in consciousness and in terms of an act/ object correlation. Broadening the notion beyond Kant’s, Husserl speaks of the ultimately foundational matter as something purely “sensuous.” It should be obvious why it is at least similar to sensation as Kant construed it. It is like sensation in that, in whatever way it might come to be related to objects, it is of itself no more intentionally directed than Kantian sensations are. At most, it is material through which intentional acts are directed. Correspondingly, that material’s ingredience in intentional acts is necessarily reflected in that toward which those acts are intentionally directed. It is necessarily reflected in the noematic correlate in precisely the way in which Kant’s sensations are reflected in the “matter of appearances.”²⁸

It is undeniable, I think, that this twofold structure—of matter and form and of act/object correlation—is what Kant had already attempted to establish in the Transcendental Aesthetic. What is in question for us is the possibility of extending it beyond the Aesthetic’s account of the role of sensation (in Kant’s narrower usage), and of its reflection in the world of appearances correlative with the intuitions containing it. (As we shall see in Chapter Four, the representation of a world of appearances—the representation of a “nature”—rests on this extension.) What is in question is the extent to which Kant saw an additional “faculty” as providing a function for the incorporation of additional material—“sensory” in a way, yet both more than sensation and less than intuition—as a necessary condition for the conceptualization of appearances.

Husserl in any case extends Kant’s notion of “sensory” material. What is most relevant for our purpose is his inclusion of what he calls “drives” (Triebe):

. . . “sensation-contents” such as color-Data, touch-Data and tone-Data, and the like, which we shall no longer confuse with appearing moments of physical things—coloredness, roughness, etc.—which “present themselves” to mental processes by means of those “contents” [but are rather material through which those moments are apprehensible]. Likewise the sensuous pleasure, pain and tickle sensations, and so forth, and no doubt also sensuous moments belonging to the sphere of “drives.”²⁹

The notion of “drives” is crucial, because Husserl connects it with that of a special type of “motivation.³⁰ The latter concerns the fact that, in apprehending qualities through a body of sensational material, one is always “motivated” in the direction of more or less determinate progressions of (possible) apprehension. Now if one is already conceptualizing appearances in some way, then it is of course obvious that one has a “motive” for the anticipation of more or less specific modes of continuing perception (contingent upon the satisfaction of equally anticipable conditions). But what we are concerned with is not the kind of motivation that is a product of already constituted conceptualization. We must be concerned with something pre-conceptual.

Husserl attempts to elucidate this aspect of “sensuous” motivation by connecting it with a class of sensations that Kant in fact ignored, namely, kinaesthetic sensations:

Looking at an object, I have at the same time a consciousness of the position of my eyes and, in the form of a unique, systematic, unfulfilled horizon, a consciousness of the whole system of possible positions standing freely at my disposal. Further, what is seen in the given position of the eyes is so connected with the whole system, that I can say with evidence: were I to turn the eyes in this or that direction, then these or those visual appearances would accordingly stream by in a particular order. . . .³¹

Such appearances, Husserl says, are “kinaesthetically motivated.”³² The possibility of obtaining them is “carried” by kinaesthetic sensations.³³ Thus “intentional content” is, at least in part, constituted against a background of non-intentional material of precisely the sort that Searle has in mind. But the difference from Searle, and the affinity with Kant, is crucial: the intentional background is no mere Background in Searle’s sense. It is part of what intentional acts are made of.

As suggested, what is comparable to this is the Kantian notion of imagination in the Deduction. In the Aesthetic, we encounter imagination only as “pure intuition.” In the Deduction, imagination needs rather to be the source of those “anticipations and retentions” that are comparable to Husserl’s “sensuous” motivations. Unlike imaginative intuitions, these must serve as mere material for object-directed acts. (Whatever directedness they possess on their own—and they surely possess some—it is neither Husserl’s nor Kant’s interest to elucidate it.)³⁴ On the other hand, within the Deduction itself, imagination also appears in a twofold capacity. Sometimes Kant uses the term to stand, not for the pre-intuitional faculty of anticipation and retention, but rather for that of the embodiment of the latter, as material in intuition, precisely with respect to their eventual function as material for conception. The terminological shift, I shall argue, is all that there is to Kant’s apparent rethinking, between the two editions of the Deduction, of the respective roles of “reproductive” and “productive” imagination.

For the sake of a final, and perhaps surprising, comparison, we may consider a third philosopher, from a third philosophical tradition. Croce is commonly regarded as paradigmatically Hegelian as opposed to Kantian.³⁵ But in the respects that concern us, we come closer in Croce to the Kantian view than we did even with Husserl. In Husserl we explored, on the noetic side of consciousness, the structure of material vs. intentional forming in acts of apprehension (that is, in the apprehension of “objects” through the material in question). But Kant makes a twofold distinction on the level of intentional form. He distinguishes between the “forms” of intuition and conceptual understanding. In Husserl, by contrast—apart from some puzzles concerning the notion of the noematic X—there seems to be but a single formative function, involved in any instance of intentionality as such, whether it is purely intuitional or explicitly “conceptualized.” Any relevant difference, expressible in terms of the latter distinction, would seem merely to hinge on some difference in the material in question. In this respect, Croce is more Kantian than Husserl, and quite clearly un-Hegelian.³⁶ For he explicitly distinguishes irreducible forms of intentionality.

For our present purpose, the following points may be highlighted. First, both Croce and Kant regard intuition as a basic mode of “(re)presentation.” Specifically, they regard it as a mode whereby one is directed toward an at least potential object of judgment. Conceptualization, then, is that mode of mentality through which one is able to recognize, with more or less definiteness, the actual (or even possible) character of such objects. Importantly, both of these modes of mentality are indifferent with respect to the actual existence of the objects in question.³⁷

Second, in addition to insisting that intuitions are intrinsically object-directed, Croce also maintains, like Kant and Husserl, that their directional form can function only to the extent that it incorporates a special kind of material to serve as its vehicle. In early formulations, Croce sees mere “sensations” as providing at least a substantial part of such material. But he later rejects, unlike Kant and (apparently) Husserl, the very idea of a material that is intrinsically devoid of affective and dispositional content.³⁸

Third, and as noted, Croce does not simply extend the notion of intentionally formable material. He also distinguishes levels of intentional formation. On each of these levels, the upshot of forming on the preceding level is what provides the relevant material. In this, Croce is simply following Kant’s view that, for example, while sensations are the proper material for intuition, intuitions formed from sensations are the material for acts of understanding as such.

Our own study may reveal that, in this respect, a more Husserlian approach is preferable: conceptual form can be regarded as operating on intuitional states only by operating in respect to some body of material in those states, that is, by constituting the apprehension of something through that material. But does this then mean that, in the final analysis, we have no need for more than one basic “form” in regard to such material? Whatever our judgment on this, I hope it will be seen from Chapter Six that there must in any case be some distinction that is legitimately formulable in terms of a distinction between a lower and a higher level of intentional directedness. Croce himself proposes a scale with four such levels.³⁹

Croce is led, by these views, to some un-Kantian conclusions. I shall not attend to them in detail. Rather, I want to emphasize how all of the thinkers agree on a point of considerable significance. To put the point in the minimalistic terms that Searle favors: “intentional content” is co-determined by non-intentional factors extending well beyond the domain of mere Kantian “sensation.” In Husserlian terms, more is of course involved. This is because Husserl requires, in the place of a single notion of intentional content, the twofold structure of matter and form in intentional activity, on the one hand, and the correlation between (noematic) object and intentional act, on the other. The former structure will be our immediate concern in the next chapter, the latter in the one that follows it.

Unsurprisingly, the noesis/noema correlation does not appear in these terms in Croce. (Indeed, Croce’s attitude toward Husserl was markedly, even arrogantly, unfriendly.)⁴⁰ The point is present in terms that are un-Kantian as well. This is because, unlike Kant, Croce assumes not only that concepts function as form in regard to matter in intuition but also that they can only function in that role. This requires rejection of the analytic/synthetic distinction (or rather requires the conclusion, for Croce, that all judgments are really synthetic a priori) and, of course, a denial of the possibility of thoughts in regard to what is unimaginable (and indeed unimagined).⁴¹ These are points that need not concern us here. Nor need we be concerned with the apparently, but in fact only apparently, radical divergence from Kant that is embodied in Croce’s insistence that there really are no such things as empirical concepts in the first place—that so-called empirical concepts are merely “pseudo-concepts.”⁴²

Croce’s way of expressing the noesis/noema correlation is to be found, in effect, in his claim that, on each of the four levels of mentality (that is, on the four levels of lo spirito), intentional “form” does not simply operate with respect to a body of material. Rather, it actually informs that material: that is, it somehow transforms it. The upshot is what Croce calls a synthesis a priori. In it, the form in question is strictly inseparable from whatever “object” is in question. This, of course, is the basis of the rejection of purely analytic judgments. The conclusion rests precisely on the claim that the very object of judgment is always at least in part shaped by the judgment that one makes of it.

To make any sense of such claims, we need to presume two things. First, and most obvious, we need to presume that the “object” with which Croce is primarily concerned is the immediate object of intuition, considered apart from any question concerning its identification as a part of material nature—that is, considered apart from the sort of identification that is of primary interest to Kant. Material objects, in this sense, are something merely “constructed,” and so they are, in Crocean terms, not really perceptible at all, nor even imaginable.⁴³ This, so far as I can see, involves a merely terminological divergence from Kant. What Croce calls the construction of nature via mere “pseudo-concepts” is what Kant regards as an act of conceptualization proper. What remains the case is that, for both of them, such construction needs to rest on some kind of transformation of a given intuition, and of the material that an intuition embodies. It is simply that, for Croce, what is of primary importance is the upshot of that transformation with respect to the immediately given as such. (As we shall see in Chapter Four, Kant was not unaware of this level of reflection; its recognition is embodied in the obscure doctrine of Transcendental Affinity.)

For Kant, by contrast, what is of central concern is the possibility—indeed the necessity—of the very same (or an appropriately similar) transformation of a manifold of intuitions—all of them thereby “conceptualized” as intuitions of one “object.” This difference in point of focus is compatible with recognition by both that any relevant anticipation (or “retention”) of such manifolds, in connection with the conceptualization (or pseudo-conceptualization) of intuitions, must rest on a kind of anticipation that can only be effected from within any given intuition. (In Chapter Six we shall see why, strictly speaking, it is in any case not quite a question of a manifold of intuitions transformable in the very same way as a given one; at most it is a question of intuitions that are transformable in appropriately related ways. I would suggest that this distinction is in fact what lies beyond Croce’s view of the “concepts” in question as mere pseudo-concepts.)

This point should also make it clear that we need to make a second presumption with respect to Croce’s rejection of mere analytic judgments in Kant. In addition to presuming a primary concern with the (suitably transformed) given as such, we also need to presume that Croce has in fact shifted, in his concern with what is “given,” from the purely noetic to a correlative noematic standpoint. On the purely noetic level, we may speak of “forming” in the sense of an act whereby a certain body of material is embodied in a certain type of intentional state. At that level of reflection, we are simply not yet (explicitly) concerned with any kind of transformation with respect to intuitable objects. For the latter, we need to shift to a correlative noematic perspective. A number of passages in Croce seem to require a reading precisely in this sense.⁴⁴

It is arguable, in any case, that all of the thinkers whom I have chosen for comparison, agree on one point: that some kind of non-naturalistic “embodiment” of pre-conceptual material, within a truly conceptual state, is an essential part of the “formation” of any conception in the first place. But we have also noticed some differences. Unlike Husserl and Croce, for example, Searle could not put the point explicitly in terms of the notions of matter and form, nor does he combine his insight with a recognition of a noesis/noema correlation. Unlike Husserl and Searle, furthermore—but correspondingly more like Kant—Croce alone attempts to recognize a distinction between a genuinely conceptual and a sub-conceptual (although still truly intentional) level of mental “formation.” Despite the differences, some of which may in fact be merely terminological, I have introduced these thinkers in order to provide some diversity of perspective, independent of Kant himself, for assessing what might otherwise seem an implausible view of the nature of conception, and of its relation to Kantian intuition.