Brigitte Sassen. Journal of the History of Philosophy. Volume 35, Issue 3. July 1997.
The idealism debates between Kant and his contemporaries were protracted and vehement. Interestingly, all parties in the debates-Kant’s critics, defenders, and Kant himself-began with basically the same conception of (material/ empirical) idealism as a position that is either skeptical with regard to the independent existence of the external world (alternatively referred to, by Kant, as problematic, skeptical, or Cartesian idealism), or that denies the existence of material substance outright (dogmatic or Berkeleyan idealism, for Kant). Positions quickly diverge, however, when it comes to transcendental idealism. Kant’s critics argued consistently that transcendental idealism reduces to material idealism. Kant rejected these arguments, insisting, in turn, that transcendental idealism is the only escape from material idealism. Nonetheless, he was influenced by these criticisms. He repeatedly returns to the differentiation between transcendental and material idealism in the years following 1781, and he casts these differences in what appear to be increasingly realist terms. However, while it is clear that Kant was annoyed by the numerous idealism charges and his critics’ persistence in, as he saw matters, misunderstanding his thought on this and other issues, it remains unclear just how seriously he took the substance of their criticisms. I want to go some distance in answering this question by considering the development of Kant’s treatment of idealism not just text-internally, but by placing it in the context of the criticisms with which he was faced.
These criticisms come in basically two groups. The first group contains critiques of the A-edition of the first Critique and of the Prolegomena. Among them are the famous Gottingen or Feder/Garve review, Garve’s original version of the review, and various other materials that pertain less directly to the idealism problem. In the second group are specific criticisms of the B-edition, particularly its Refutation of Idealism. These include such diverse treatments as the discussion of idealism in Platner’s historical Aphorismen, Schulze’s criticisms in Aenesidemus, which is, in the first instance, a critique of Reinhold s Elementarphilosophie, and a review of the Critique by H. A. Pistorius.
The following sketch of the first set of criticisms will be a brief overview of those criticisms that did have an effect on the development of Kant’s refutation of idealism. They precipitated, as I shall show, Kant’s turns from attacking skeptical, Cartesian idealism in the A-edition Fourth Paralogism, to attacking dogmatic or Berkeleyan idealism in the Prolegomena, back to attacking skeptical idealism in the B-Refutation. My analysis of the second set of criticisms will involve a more detailed examination of Pistorius’s review. The review is important for two reasons. It serves to exemplify, more than other critical material written during this time, the general discontent with the argument of the Refutation, and with the Kantian approach as a whole, that was prevalent even after the publication of the B-edition. Moreover, Pistorius’s review is arguably one of the evaluations Kant did take issue with in some of his unpublished writings on idealism after 1787. At the very least the objection Pistorius raised, or an objection like his raised elsewhere, did influence the development of Kant’s engagement with idealism after the B-edition. My exploration of that development in the final section of this paper will, for reasons of brevity, be limited to two Reflections dated 1788-go (R 56535654) where Kant does, finally, take the realist turn his contemporaries had demanded all along.
From Descartes To Berkeley To Descartes
In the A-edition of the Critique, specifically the Fourth Paralogism, Kant’s argument against material idealism focusses on the skeptical idealism he associates with Descartes. The argument is a modest one, at least when it is compared with Kant’s later, much more radical refutations. At this early point Kant is content to undermine the central premise of skeptical idealism-that only inner experience is immediate and certain-and to argue that our knowledge of the existence of outer objects is not, as the idealist claims, a matter of an inference drawn on the basis of this ostensibly immediate and certain inner experience. He seeks to establish these points by uncoupling outer experience from inner experience, arguing, much to the dismay of early critics, that given the resources identified by transcendental idealism, both outer and inner experience afford immediate knowledge of things, states, and events, but that in both cases this is a knowledge of appearances only. Kant is aware of the price transcendental idealism exacts, but clearly thinks it a price worth paying since, in his view, it provides us with certain knowledge of the existence of the empirical world.
In this argument the dogmatic idealism Kant elsewhere attributes to Berkeley, plays a secondary role at most. Kant does refer to dogmatic idealism towards the end of the Fourth Paralogism, but states there that since the position is a function of a confused conception of matter, its resolution must be deferred to the next chapter, the Antinomies. The Berkeleyan variant of dogmatic idealism simply does not figure and is not the form of material idealism Kant is in the first instance concerned with. This is in stark contrast to the Prolegomena, where Berkeley has pride of place. This is obviously the case in the Appendix (“Test of a Judgment about the Critique that Precedes Its Examination”), Kant’s response to the much discussed Gottingen or Feder/Garve review. Kant was outraged by the review and severely criticized what he took to be its obvious lack of respect, its polemical tone, its constant misrepresentation and, most importantly, the reduction of transcendental to Berkeleyan idealism. But even in the earlier remarks following 13 at the end of the first part of the Prolegomena, the focus is primarily on what Kant takes to be dogmatic rather than skeptical idealism. Witness his initial definition of idealism as the “assertion that there are only thinking beings, that all other things that we believe to perceive in intuition are only representations in thinking beings to which in fact no object existing outside of us corresponds.” While skeptical idealism is mentioned at the end of these remarks as the received meaning of idealism, Kant is not concerned to deal with it at this point.
Now, while the external reason for the shift to dogmatic idealism is clear enough, the same cannot be said of the manner in which Kant deals with it. In the Prolegomena remarks as well as in an unpublished note on this issue Kant lets the different attitudes toward the existence of things taken by the material-likely Berkeleyan-idealist and the transcendental idealist respectively speak for themselves: the former is said to deny their actuality, to attribute them to dreams, the latter is said to deny only that they can be known as they are in themselves. And, as he also points out in this context, that does not have any implications for the existence of things independently of us, an existence, he adds, that he has never doubted. In the Prolegomena Appendix, however, there is a shift of emphasis from the primarily ontological terms of the initial treatments to an answer cast entirely in epistemological terms, centred around the notion of space. Kant claims that since Berkeleyan idealism takes space to be empirically derived, it does not have a criterion of truth. Transcendental idealism, by contrast, according to which space is an a priori and hence necessary form of intuition, has the resources to differentiate truth from semblance. Transcendental idealism, on this argument, can establish the objective reality of knowledge; Berkeleyan idealism cannot.
Quite apart from the question of whether this is a legitimate critique of Berkeley, the response is a surprising one. While Feder does think that Kant is unable to differentiate truth from semblance, and reality from dreams or semblance, his main complaint is that Kant, like other idealists, reduces everything to representations. Thus, on the Kantian philosophy “everything is merely representation and law of thought; … representations modified in us and ordered in accord with certain laws are just what we call objects and world.” Since he takes this reduction to be the essence of idealism, he does not see on what grounds Kant differentiates transcendental from other forms of idealism. Clearly, for Feder someone who claims not to be a (material) idealist, indeed, someone who claims to have refuted idealism, would have to be a transcendental realist and would have established the independent existence of things. But that is not an option available to Kant. After all, he has rejected transcendental realism, and he does “reduce if not things, at least everything of which we can have knowledge, to representations in both the Aedition and the Prolegomena.
Nevertheless, note that Kant’s epistemological answer to Feder in the Appendix is an answer to Feder, even if it is not the sort of answer Feder either wanted or was able to understand and appreciate. For Feder ends the review on a polemical note by asking an epistemological question:
when, to assume the most extreme position with the idealist, everything of which we can know or say something is merely representation and law of thought; when representations modified in us and ordered in accord with certain laws are just what we call objects and world, what then is the point of this fight against our commonly accepted language? What is the point and what is the source of the idealist distinction? While Feder probably intended this question ontologically, as a question about transcendentally real causes by which we can distinguish reality from illusion, and, by implication, differentiate Kant’s position from the material idealism he claims to refute, Kant chooses to answer it in the epistemological terms in which it is framed. And, as we know, Kant’s argument is that the source of the idealist distinction (presumably of his position from material idealism) is the apparatus of a priori forms and concepts supplied by transcendental idealism, which is said to provide us with the ground of the certainty of our knowledge (of the empirical world) and which the Berkeleyan idealist, ostensibly lacking a priori conditions of knowledge, cannot provide. Given any particular set of sensory experiences, the apparatus of a priori forms and principles is supposed to provide a set of constraints sufficiently rigorous to ensure that one and only one interpretation of the objects represented by those experiences is possible. This is the import of the following passages from Remark III to Part I of the Prolegomena:
… the difference between truth and dream is not determined by the nature of the representations which are referred to objects-for those are the same in both casesbut through their connection according to the rules which determine the coherence of representations in the concept of an object, and whether or not they can stand together in an experience.
… even if we did not think at all about the origin of our representations, and connect our sensuous intuitions-no matter what they contain-in space and time according to the rules of the coherence of all cognition in one experience, both deceptive semblance or truth can arise, depending on whether we are careless or careful. This is merely a function of the use of sensuous representations in understanding, and not of their origin.
Whether he is correct about this or not, this is the important advantage Kant takes transcendental idealism to have over its empirical or material counterpart, not, at this point, the ontological matter of the existence or nonexistence of things in themselves. That is here not the issue for Kant.
For all his emphasis on the certainty of our knowledge, however, Kant was not altogether unaware of his contemporaries’ overriding ontological concern. In addition to repeatedly proclaiming our certainty regarding the existence of material things in space-a claim that did not satisfy his contemporaries-he often argues that even for the transcendental idealist things in themselves must exist as affecting objects since sensations are given to us, not created by US.
It is a good thing that while Kant does repeatedly appeal to this argument by affection, as it is commonly called, it does not have much of a role to play in his refutations. His contemporaries were quick to point to its weaknesses. Jacobi, for instance, argued that it makes an illegitimate causal claim about things in themselves: if we cannot know anything about things in themselves, we cannot know that they cause sensations in us. More tellingly perhaps, Platner and Schulze claimed that this argument does not establish anything regarding the world as it in itself, certainly not the actuality of things existing independently of us. They conceded that we might have to be affected by something in order for sensations to be given to us, but argued that this claim goes the way of all arguments from effect to (unknown) cause. It establishes at most that there must be something distinct from ourselves. Both Platner and Schulze affirm that since Berkeley too believed that there is something distinct from ourselves that affects us in some way (God), this argument does not differentiate Kantian from Berkeleyan idealism in any way. For Kant cannot in the end say anything about the affecting object to contradict the view that we are in some way affected through an infinite spirit.
While of dubious value as an argument, however, the argument by affection does serve to point to the heart of the idealism debate. Prior to the Gottingen review Kant seems not to have been at all sensitive to the concerns animating the persistent idealism charge. The existence or nonexistence of things in themselves, or, for that matter, the whole issue of idealism, simply wasn’t an issue for him. There are few remarks on idealism in his writings prior to the 1780s, and the refutation of idealism implied by the argument of the A-edition Fourth Paralogism is of secondary importance; Kant is here primarily concerned with the rational psychology that is exemplified, in his view, by the skeptical idealism of Descartes. His surprise and outrage at the Gottingen review, furthermore, shows that it did not occur to him that anyone could think of equating his philosophy with either skeptical or dogmatic, or, for that matter, Leibnizian idealism. Of these, the reduction of transcendental to dogmatic idealism is entirely inappropriate, as Kant points out in one of the remarks added to Prolegomena, since dogmatic idealism is motivated by concerns over the impossibility of the existence of matter or its ability to affect us, whereas transcendental idealism is concerned only with limiting the knowledge we can legitimately have. As far as Kant is concerned, the latter has no bearing on the existence of things independently of us. For his contemporaries, of course, it is precisely that existence that is called into question by transcendental idealism.
Nor does Kant help matters with his first responses to the idealism charge. By emphasizing that the certain knowledge we do have is of empirical objects in space, that this is a knowledge of appearances, and worse yet, by not retreating from his earlier claim that space is only “in us,” Kant only confirms what his contemporaries had known all along: that transcendental idealism is idealism, stuck behind the veil of representations or appearances, and unable to say anything about the world independently of us. And when he does try to say something about that world by appealing to the objects that supposedly affect us, he adds insult to injury. For the argument by affection is just the sort of inferential argument he has himself criticized in his assessment of problematic idealism. It is perhaps no wonder that his contemporaries persisted with the idealism charge. But Kant failed to acknowledge the difficulties his contemporaries had in coming to terms with transcendental idealism. In R5642, having reaffirmed what he takes to be the essential difference between ordinary and transcendental idealism, Kant asks with some dismay “(just what would one demand that I should claim in order not to be an idealist?” The answer should have been and perhaps was obvious to him. His contemporaries wanted him to demonstrate the actual existence of external things corresponding to (or directly known through) our perceptions, not just the actual existence of appearances considered as representations in us. In fact, they wanted him to give up what they considered to be the root of his idealism and what Kant, by contrast, took to be the advance and contribution of his philosophy, viz., the distinction between appearance and thing in itself, and the consequent limitation of all our knowledge to appearances. Not surprisingly, he would not do that. Instead, he tends to emphasize the virtues of the transcendental philosophy: the possibility of synthetic a priori judgments that it establishes or, more importantly in this context, the certainty of the knowledge of the empirical world that it affords. And by doing so he fails to answer the point of the idealism objection, at least as his contemporaries understand it, just as they failed to understand the point of transcendental idealism.
The initial stage of the idealism debate played itself out around this point. After the Prolegomena Kant encountered Feder’s complaint that idealism has not been refuted since that requires the “demonstrated existence of bodies” in Garve’s original version of the review. This is hardly surprising since the Gottingen review had been fashioned from it. More sophisticated versions of the objection can also be found in texts that question the separation of things in themselves and appearances. One of these texts is a review of the Prolegomena written by Hermann Pistorius, one of the reviewers for the Allgemeine deutsche Bibliothek. We don’t know of course whether Kant did actually read this review, but he might well have. He respected Pistorius, and the journal was a major academic journal at the time. If he did read it, Kant should have been gratified. The review is for the most part an excellent summary of the Prolegomena, and, what is more, Pistorius recognizes that transcendental idealism is not to be reduced to Berkeleyan or Leibnizian idealism: not to the former because that “destroys the world of bodies,” not to the latter because that encompasses outer sensations. Yet even Pistorius ends by questioning the barrier between things in themselves and appearances, and wonders about the ground of appearances and about real or actual things-both those real or actual things to whom things appear and those that appear. He develops this question further in his review of Schultz’s Erlauterungen, where he claims that the subject, in particular, must be considered a real or actual thing, not, as he thinks transcendental idealism argues, an appearance or semblance only. Otherwise, he argues, we would have a situation in which semblances appear to semblances, not real things. There would, in other words, be no final appeal, only an infinite regress. A similar concern is also raised by Ulrich who argues for the applicability of the categories to things in themselves.
Even this sketchy summary of key criticisms should be sufficient to indicate that the overriding concern in the idealism debate had to do not so much with whether or not Kant was a Berkeleyan idealist, but with both the actual existence of Kantian things in themselves (Feder, Garve) and with our cognitive access to such things (Pistorius, Ulrich). Given the persistence of this charge and what must have appeared to Kant as a persistent failure on the part of his contemporaries to appreciate transcendental idealism, it is perhaps not surprising that Kant should attempt another version of the refutation in the Bedition and that he should do so with an aim that is at least at first sight not strikingly different from that of his original refutation in A, to demonstrate again the “existence of objects in space.” Indeed, when Kant returns to the idealism question in the B-edition of the Critique, he proceeds almost as if the Prolegomena had never been written. He returns here primarily to the refutation of skeptical or Cartesian idealism. Kant does mention dogmatic (Berkeleyan) idealism in the introductory paragraph, but dispatches it quickly by reference to the Transcendental Aesthetic.
There are, however, some radical changes between the argument of the Refutation and the A-edition treatment of skeptical idealism. While the overall tenor of the argument is the same as that of his earlier critique of skeptical idealism-transcendental idealism is not to be reduced to and serves in fact to refute material idealism-Kant now sets about establishing the point rather differently. The discussion is no longer consigned to the Dialectic as it is in the A-edition, or to appendices and remarks as it is in the Prolegomena, but appears as a corollary to the discussion of the Postulate of Actuality, a much more appropriate place for arguments of this sort. And unlike the argument in A, Kant is now no longer content to establish that our experience of outer objects is immediate, and not, as he thinks Descartes argued, a matter of inference based on the certainty of the immediate knowledge of ourselves. Kant is no longer content, in other words, to unhinge outer experience from a dependence on inner experience and to establish that both afford the same sort of immediate access to their “objects.” In the B-edition he goes one step further by recoupling them, this time however, as he proudly claims, by reversing Cartesian idealism. More specifically, he makes inner experience (the knowledge we have of ourselves as substances persisting through time) contingent on outer experience (the knowledge of the “object in space outside of” ourselves). Finally, and perhaps most significantly, Kant now denies that the “object in space outside of” ourselves is only a representation in us, an equation he had made in both the A-edition and the Prolegomena. He seems to want to say it really exists independently of us, without, however, equating it with things in themselves.
While there can be little speculation about what prompted Kant to return to a refutation of idealism in the B-edition-as indicated above, that would seem to be explained readily enough by the persistence of the idealism charges-it is a real question what prompted him to do so in a manner so strikingly at odds with his earlier arguments against idealism. Why has it become necessary to argue not just for our immediate, that is, noninferential, access to (empirical) objects, but to now make outer experience an integral part of inner experience? Or, to put this differently, why is it necessary to argue that the demonstration of the fact that we have experiences of outer things, and do notjust imagine them, is accomplished only when it is shown that “even our inner experience, which Descartes does not doubt, is possible only under the presupposition of outer experience”? Some have argued that the answer to this question can only be found text-internally, not by appeal to the criticisms of Kant’s contemporaries. These can at most serve to indicate what might have motivated Kant to rethink the argument, but they do little, or so the argument claims, to explain the substance of the new line of argumentation. Others, by contrast, do point to specific pieces of criticism to develop at least some part of the required explanation. Ulrich’s Institutiones logicae et metaphysicae and Schultz’s review of it are clear explanatory favorites. Erdmann, for instance, claims that Kant took his cue for the new proof from Ulrich’s objection that appearances or representations cannot be permanent because they are always changing. Brandt appeals to the same texts but focusses specifically on a question Schultz raises at the end of his review regarding the applicability of the categories to objects of experience. But Brandt’s appeal to Schultz’s review is a peculiar one. For unlike other idealism charges, Schultz’s question does not concern either the existence of or our access to things in themselves. He wonders, rather, whether Kantian objective reality is in fact a function of the categories or whether it may require an order preestablished by God. He makes this point by way of a nice appeal to Hume’s skeptical doubt regarding necessary connection, suggesting that it could well be possible that the apparently necessary order we observe among appearances is merely an artifact of the way in which God gives these appearances to us (as regularly or constantly conjoined) rather than a function of natural laws prescribed by the understanding. In this case, the lawfulness of nature would be derived a posteriori, not prescribed a priori. And as long as this cannot be ruled out, the a priori status of the laws, and, by implication, the categories cannot be asserted.
Quite apart from the fact that it is far from obvious just what this argument has to do with the new Refutation of Idealism, the appeals to Ulrich and Schultz or, for that matter, the text-internal accounts, cannot explain some of the more puzzling aspects of the text. Among these are Kant’s renewed emphasis on the actuality of things in space, on the existence of something outside of and distinct from ourselves, and his claim that neither space nor the permanent in space is something merely in us. These do make sense, however, in the broader context of an idealism critique that demands that the existence of outer objects be demonstrated and that these objects not be reduced merely to representations. This is the case also for the return to a focus on skeptical idealism. If the task is to demonstrate the indubitability of the existence of outer objects, then this is better shown in the Cartesian than in the Berkeleyan context. After all, the objections of the latter require that before the existence of things in space can be established, certain questions regarding the possibility of matter itself be settled. And that is a point that is not at issue between Kant and his critics (except perhaps Feder), whereas the actuality of things in space outside of us is. Given all of this, I am reluctant to accept any of the lines of speculation here indicated and propose instead that the Refutation is best viewed in the broader context of the concerns raised by Kant’s contemporaries.
Part of the problem with identifying the ancestry of the specific argumentation in the B-Refutation is that it is far from clear just what Kant wants to accomplish here. Does he want to argue only for our certainty of the existence of objects in space, or does he want to do more than that? If the former, then it may appear mysterious why he thought it necessary to construct the reductio argument; after all, that existence could be taken as established on the basis of his A-edition arguments. However, note that, given other changes, the Aedition arguments are no longer appropriate to establish even our immediate certainty of empirical objects. For Kant here retreats from his earlier identification of appearances (objects in space) and representations, as well as from the claim that space is only in us. The noninferential knowledge that we are to have of the existence of such objects must, accordingly, be established in some other way than by pointing out that we are immediately aware of them. A new line of argumentation is required.
At the same time, it is difficult to believe that he could have hoped that the mere retreat from his previous equation of appearances and representations, and the attendant new demonstration of the “existence of objects in space” outside of us would satisfy those critics (Feder, Garve) who thought just such a demonstration necessary for a refutation of idealism. Nor did it. For the most part his critics thought he had missed the point. Platner, for instance, claims that rather than refute idealism, the Refutation actually endorses it. For, according to Platner, permanent objects in space are in the end “objects only in us,” which, he goes on to state, is “precisely what the problematic idealist. . seeks to prove.” However, if Kant wanted to make a stronger claim, viz. to demonstrate the existence of transcendental objects (and there are reasons to think that he might have wanted to do so that I will explore below), then this would have satisfied his contemporaries-if successful-but it would also have rendered the Refutation inconsistent with the remainder of critical idealism. For in that case Kant would be arguing that inner experience is contingent on the experience of things in themselves. That is, of course, precisely what Pistorius and Ulrich (and probably Feder and Garve as well) want him to claim, but if Kant were to do so then transcendental idealism would be in danger of collapsing into transcendental realism.
I will return to the question of what the Refutation was to establish at the end of the next section. Before doing so I want to consider how the argument was received by Kant’s contemporaries. The paradigmatic representative of that largely negative reception is Pistorius’s review of the B-edition, which focusses on the central problem of the Refutation, namely, the proper interpretation of the permanent in space.
The Pistorius Review
As a review Pistorius’s essay is far from comprehensive; he looks only at the Copernican Revolution and the Refutation. These topics are ideally chosen, however, to determine whether Kant can meaningfully and consistently establish that objects do exist independently of us, given the limits of his philosophy that are set out by the Copernican Revolution. Pistorius thinks not. To make his case he does not focus on the details of the argument of the Refutation; he is content to cite the argument in its entirety and leave it at that. Instead, he complains about Kant’s tendency to evade difficulties by using ambiguous expressions and asks only a single question: “are we to understand the existence of outer objects as an actually persisting existence of things or as a merely logical, apparent existence?” He rejects the first of these alternatives as inconsistent with Kant’s philosophy, and the second as ineffectual.
It might be immediately objected that Kant did not in fact leave the “existence of outer objects” unspecified, as Pistorius claims. In the B-Preface Kant offers a clarification of what the permanent objects in question amount to. He suggests that the sentence “[t]his permanent, however, cannot be something in me, since my existence in time can itself only be determined through this permanent” be replaced with the following passage: “[t]his permanent, however, cannot be an intuition in me. For all determining grounds of my existence which can be found in me are representations. As such they themselves require a permanent distinct from them with respect to which their change and hence my existence in the time in which they change, can be determined.” Pistorius does not pay attention to either this clarification or the longer explanation that follows, but had he done so he would likely not have been satisfied. For in the former Kant states only that the permanent in question is not to be equated with representations. He does not say just what the “permanent distinct from them” is to be equated with. In the subsequent clarification he does emphasize that it is “experience, not invention, sense, not imagination, which connects what is outside with my inner sense.” But since the status of the (outer) object of experience is precisely what Pistorius is concerned with, his demand that Kant further clarify the “existence of outer objects” remains justified.
Pistorius’s real reason for rejecting the first interpretation, viz. that we understand the “existence of outer objects as an actually persisting existence of things,” as Kantian transcendental objects or things in themselves, is the Copernican Revolution (more on that below). The actual evaluation of this option, however, appeals to two reasons. The first (and better) reason is one he mentions almost accidentally, and unfortunately does not further work out. He points out that it must be rejected because it violates the main critical law, a law he has already identified in terms of a possibly skewed version of the unknowability thesis, the unknowability of the “possibility, actuality, or necessity” of things in themselves. Now, Pistorius does not tell us why this interpretation violates the main critical law. He might, however, have thought as follows. The argument of the Refutation clearly requires that we perceive the permanent thing with respect to which we determine our existence in time. However, since the main critical law specifies that knowledge can be had only by means of the schematized categories, and that they are legitimately used only with reference of objects of sensory experience, we cannot know things in themselves. And so the argument either violates the unknowability thesis or the first interpretation must be rejected. Alternatively, he might have argued that the permanence of a thing, as well as our knowledge of it, is still something constructed by us. So the permanent thing infused with objective reality in the proper (critical) manner is an empirical object. As such it is for Kant’s contemporaries ultimately a representation only, that is, as Platner would claim an “object only in us.” And if that is the case the first interpretation must be rejected.
Of course, these are just speculations on what Pistorius might have said if he had thought this objection through. He does, however, have another argument. He rejects the first interpretation by appealing to a purported contradiction between the Refutation and the Transcendental Aesthetic, arguing that they posit contradictory foundations for the concept of time. While the first reading of the argument, which posits that the required permanent objects are transcendental objects, entails that time is grounded objectively, in actual external things, Pistorius believes that the Transcendental Aesthetic has already demonstrated “extensively and with many reasons” that it is only a “subjective form of sensibility,” not grounded in or abstracted from things in themselves. Pistorius repeats the same point with respect to the concepts of change and permanence. While, as he understands matters, according to the Transcendental Aesthetic “our concepts of succession in time, of alteration and… of the permanent are subjective only, [while they] spring from the essential constitution of the human representative power” and cannot be assumed to be properties of things in themselves, the Refutation claims “that there must be actual external things since we could not otherwise have a consciousness of the permanent in us.” There are, for Pistorius, two alternatives: either time is subjective, as argued in the Aesthetic, or it is objectively grounded in things in themselves, in accord with the first reading of the Refutation. But if the former, then the first proposed interpretation of the Refutation must be abandoned (and Kant has not in fact managed to refute idealism); if the latter, then the Kantian philosophy is inconsistent.
The notion of a form of intuition that lies at the ground of this alleged inconsistency is one of the points on which Pistorius does misunderstand Kant. Like many of Kant’s contemporaries, he thought that Kant had (re)introduced innate ideas. That is, he thought the subjectivity of time entailed not just a receptive capacity, but access to a determinate objective order. And if that were the proper way of viewing forms of intuition (and the categories), then Pistorius’s objection might in fact be valid. However, for Kant it is one thing to say that time is the form of intuition in accord with which all data must be given to us, and quite another to speak of the concept of time or, for that matter, of the perception of a determinate temporal order. Only the former is properly speaking subjective, perhaps even “given” or “innate,” but Kant is adamant that it must not be equated with innate ideas. Still, Pistorius has, perhaps inadvertently, put his finger on a central difficulty in the notion of a priori forms of intuition, namely, what it means for these forms to be given, a difficulty Kant eventually attempts to address with the notion of original acquisition. In a reply to a similar complaint made by Eberhard, he argues that the forms of intuition are original acquisitions. As such they are neither innate ideas nor derived from experience, but originated by the mind in virtue of its “particular receptivity,” or, as he puts it elsewhere, “the formal constitution of the senses,” on the occasion of experience. So the objective ordering relations are not given, but are built out of or on the basis of our constitution. On this account it is not contradictory to say that time is an a priori form of intuition and that it must be perceived, as an objective ordering relation, with respect to something permanent outside of ourselves, the permanent objects in space.
Whether or not Pistorius would have accepted this elaboration is another question, of course. I suspect that he probably would not have done so, indeed that he might have viewed this clarification as pure sophistry, an attempt on the part of Kant to introduce a priori forms of intuition that nevertheless arise in some mysterious and yet necessary way from experience. None of that matters here however. For as pointed out above, whether Pistorius’s second argument succeeds or fails, he could still argue that the interpretation conflicts with the implication of the Aesthetic, the unknowability and nonspatiotemporality of things in themselves.
Pistorius does not bother to determine whether the second interpretation of “the existence of outer objects” as “the logical existence of things, … an existence in appearance” is the correct one, that is, whether it is consistent with the Kantian philosophy. Nor does he discuss just what the “logical existence of things” amounts to. He seems to take both points as established, the latter as evidently a merely apparent existence, semblance, or existence only “in us,” and the former because that is the remaining of two alternatives. In his evaluation of the second interpretation, he focuses instead on the claims Kant makes for the Refutation, arguing that on this interpretation, and contrary to Kant’s assertions, Kant cannot in fact say anything about the external world proper.
To make this point Pistorius takes issue with two of the three remarks Kant adds to the Refutation. The first of these is Kant’s claim that transcendental idealism actually reverses idealism because instead of making outer experience contingent on inner experience, inner experience is here identified as contingent on outer experience. Pistorius claims that this cannot be maintained, at least not in a meaningful way. His disagreement with Kant has to do with what he thinks constitutes outer experience in Kant’s philosophy. If the existence of external objects is a “merely logical existence,” then outer experience reduces to inner experience. After all, it is something we construe ourselves, by applying our “concepts and rules to objects.” He asks twice how we could possibly conclude anything about the “actual existence of things” or the “objective reality of outer things” on the basis of constructions that are grounded in “the original constitution of our spirit” and asks, finally, whether Kant “does not know far more of external things than he could know of them if he wanted to remain consistent and true to his writings.”
With respect to Kant’s second remark, Pistorius repeats some of the points made earlier. Kant’s claim is that the argument of the refutation is confirmed by experience. It shows all time-determination to be contingent on constant change, which can itself be perceived only with respect to something permanent in space. Pistorius thinks this is nonsense. To support this claim he appeals again to the subjectivity of all time-determination that follows, in his opinion, from the Transcendental Aesthetic. The subjectivity of time (as an innate idea) ought to make appeals to permanent things superfluous. But that is not really the point he wants to make here. Rather, he wants to emphasize his earlier point regarding our inability to move from the subjectivity of the conditions of the possibility of knowledge to actual things. The subjectivity of time as a form of intuition may be responsible for the way in which we order intuitions, it may make the “representation of variability in intuited things necessary,” but this does not allow us to draw any conclusions about the permanence–a temporal characteristic–of actual things existing independently of us. Indeed, the subjectivity of time ensures that we must remain on the level of appearances.
We may wonder whether Pistorius has not missed the point of the Refutation. Kant is saying that if I am to make sense of my constantly changing, subjective experience, I must take it as axiomatic that some of the things I am experiencing endure over time when I am not looking at them. Otherwise, there would be nothing that could serve as a landmark or a clock and no way, therefore, to even begin making sense of my experience. So I am necessarily constrained to infer objects that are “behind” what I experience, not as things in themselves-nor, however, as mere representations in me.
It does not matter, however, that Pistorius misses this point. For his focus is on the nature of the peculiar permanent objects that we are, on Kant’s argument, constrained to infer. His discussion reveals that on this issue he parts company with Kant where other critics have done so as well: over the implications of a philosophy according to which the knowledge we can have is constructed in virtue of the “original constitution of the human spirit.” For Kant this approach ensures the objective reality of the empirical world, but for Pistorius it ensures that we cannot have knowledge of things in themselves, of objects truly outside of us and distinct from us. In tune with other positions defended earlier he similarly thinks that a successful refutation of idealism must be conducted in realist terms, that it must demonstrate, not just avow a belief in, the independent existence of (and our cognitive access to) things distinct from us, not of appearances or Kantian empirical objects, viz. of logically existing things. And whether or not Kant wanted to do so-a point Pistorius does not even attempt to settle-he thinks that given the resources of transcendental idealism Kant cannot in principle do so. The Refutation of Idealism was, for Pistorius, doomed in the Preface.
It is in fact quickly evident from Pistorius’s discussion of the Copernican Revolution that he has no sympathy at all with the a priori conditions of knowledge characteristic of transcendental idealism. He takes these conditions as a misconstrual of the degree to which knowledge is contingent on our cognitive and perceptual constitution. To be sure, he concedes that perceptual data are transformed in, as he puts it, the passage of particles of air through our organs of hearing, and the passage of light matter through our organs of sight. But he is quick to point out that this does not make, in this case, sound and color as passive as he thinks Kant takes them to be. It is not up to us, for instance, whether we hear a tone softly or loudly. That is a function of the speed with which particles of air hit our ear. And by the same token, we do not arbitrarily or willfully perceive objects to be of this or that color. That too is a function of the objects themselves and the manner in which light is refracted by them.
Note that Kant would not disagree with these initial points. After all, he too thinks that actual sensations are given to us. But he would object to the next step of the argument. Having established these uncontroversial matters, Pistorius goes on to make the same point with respect to the perception of space and time, and the determination of laws of nature. Here too Kant may think that it is an empirical matter not that sensations are located in a threedimensional space, but where they are located in that space. According to Pistorius, however, we would not be able to intuit things in space and time “unless there were properties and relations in them through which our determinate manner of intuition would be made possible.” And “we would not be able to prescribe the universal laws of nature if it was not itself ordered according to those laws, or at least according to laws that correspond to laws of our understanding.” Otherwise, he argues, nature would either be an undifferentiated mass, which is unlikely, or it would often contradict the laws we impose on it, which is not borne out by experience. Pistorius concludes, accordingly, that there is no reason to think-in fact, that there is every reason not to think-that nature conforms to our cognition as Kant intimates through the Copernican Revolution. And that, he thinks, leaves Kant with two choices: preferably, to abandon the Copernican Revolution, and with that transcendental idealism, or, less desirably, to accept a philosophy that is, at base, idealist in nature. Indeed, Pistorius might ultimately agree with Jacobi’s earlier and much stronger assessment regarding the necessarily idealist implications of transcendental idealism: “The transcendental idealist must accordingly have the courage to assert the most robust idealism that has ever been taught, and not to fear even the charge of speculative egoism, since he cannot possibly function in his system if he wants to do as much as reject the latter charge.”
Quite independently of what we think of the details of Pistorius’s argument-or, for that matter, of Jacobi’s pronouncement that transcendental idealism entails speculative egoism-Pistorius’s evaluation of the Refutation should lead us to consider, as it should have led, and perhaps did lead, Kant to reconsider just what the Refutation was to accomplish. Once that question is seriously posed, it is perhaps not difficult to see why Kant’s contemporaries were so confused by Kant’s argument. Kant’s references to “things outside of me,” to “the existence of actual things,” to “outer sense” and to the “immediate consciousness of the existence of other things outside of me” suggest that he does want to take the realist turn his contemporaries demanded. Yet, he also repeatedly refers to the “things outside of me” as “objects in space” or the “permanent in space” and since space is not, for Kant, a property of things in themselves, this suggests that the permanent in question does refer to the empirical objects which we constitute ourselves. And if that is the case, then Kant has changed his argument and manner of presentation, but he still adheres to the basic ideas defended in the A-edition and the Prolegomena. And that might mean, in turn, that his contemporaries’ objections are justified. For they argue that these are objects “only in us,” in spite of Kant’s claim that the permanent objects (in space) cannot be representations in me. But, they might ask, what else can Kantian objects in space ultimately be?
The same potential duality is also evident in the supposed clarification of the argument in the Preface. Here Kant himself identifies the possible objection that “I am immediately aware only of that which is in me, that is, of my representation of outer things,” not of the things themselves. Unfortunately, the reply he formulates there is not as clear as one might wish. He again refers to the relation of outer sense to something “actual outside of me,” the existence of which is demonstrated through its necessary connection to inner sense. But he leaves unexplained the status of the actual thing outside of me which I “experience, not invent, sense, not imagine.” He does, however, make a number of suggestions that do suggest a realist turn. He states, for instance, that “I am just as surely aware that there are things outside of me which relate to my sense, as I am certain that I myself exist as determined in time.” And a little later on, after having explained that it is a function of the rules identified by the critical philosophy to decide whether an intuition is produced by imagination or outer sense, he emphasizes that we can apply these rules in the first instance only because the conviction “that there actually is outer experience always lies at the ground.” Both references-to outer experience and to what relates to my sense-do suggest a realist view unless one wants to argue, as some admittedly do, that we are affected by the objects in space that we ourselves construct. But if Kant did want to take the realist turn, we have to wonder whether he could do so in a way that is consistent with the critical philosophy. Pistorius’s question still stands.
I suspect that Kant himself was not entirely sure just what the argument of the Refutation was supposed to accomplish. He was obviously irked by the persistent idealism charge and sought a line of reasoning that would force his contemporaries to concede that transcendental idealism does not in fact reduce to material/empirical idealism. But if so, he was fighting a losing battle. For his contemporaries would not be content with anything but a demonstration of the existence of transcendental objects or things in themselves. As Pistorius’s review makes clear, they regard the demonstration of the existence of empirical objects in space ostensibly “outside of us” as an obfuscation of the issue at hand. And Pistorius is right to question whether it is possible to provide the demonstration they deem necessary within the framework of transcendental idealism.
In spite of these problems, it must be acknowledged that considered on its own, that is, independently of any consideration of the argument as a response to the prevalent idealism critique, the argument is a successful one. Kantian empirical objects in space do meet the requirements of the task at hand. Consider: Kant emphasizes that the permanent must be distinct from a mere sequence or succession of representations or mental states. Since, according to Kant, empirical objects in space cannot be equated with sequences of actually or potentially changing representations, but have permanence and solidity in virtue of the fact that they are ascribed locations in an indefinitely extended space, the parts of which can all be postulated a priori to continue to exist independently of being perceived, they can indeed be considered sufficient for the task. As appearances they have the “solidity” and spatiality Kant consistently identifies with permanence. Of course, this still leaves a number of questions. Notably, where do sensations originate, just what is the status of the empirical objects that are neither representations in us nor objects “truly” outside of or independent of us, and what is the ontological status of the space in which they exist? If it is not “in us,” where is it to be located? These are questions to which we do not find ready answers in Kant’s texts. Indeed, these questions may require just the sort of excursion to Husserlian intentional objects recently explored by Aquila.
It is another matter to decide whether it is necessary to demonstrate the existence of things in themselves in addition to the epistemological objective reality that Kant defends. While in a post-Husserlian, indeed, postmodern time we may no longer think it necessary that the existence of objects independently of consciousness be demonstrated, Kant’s contemporaries, as this paper has amply shown, consistently demanded just that. And the interesting point here is that even if Kant did not think so in 1781, by 1787 he seems to have come to agree with them. It does matter to him, for instance, as his repeated appeals to the argument by affection indicate, where and how sensations originate. Nor can the ambiguity that characterizes the BRefutation be entirely accidental. And finally, he does return to the topic of a refutation of idealism repeatedly in the years following 1787, in ways that do suggest a clearer version of the realist turn only possibly indicated by the B-edition.
As already noted in the Introduction, I am going to limit my examination of Kant’s replies to two Reflections, R5653 and 5654, which Adickes dates 1788go. Adickes further suggests that Kant wrote these notes in connection with his reading of Eberhard’s idealism critique in the first volume of the Philosophisches Magazin. He bases this association on the similarity of writing and ink of these notes and Kant’s subsequent notes for the Entdeckung. However, if these Reflections are to be specifically associated with Kant’s reading of any critical material, there are, in my view, better reasons to support a connection with Pistorius’s 1788 review of the Critique discussed just above. These have to do, in the first instance, with the nature of Eberhard’s critique. Although Eberhard does refer to Kant’s supposed idealism in the first volume of the Philosophisches Magazin, that is by no means his major concern. Eberhard wants, instead, to vindicate the Leibnizian philosophy against what he takes to be the threat of the critical philosophy. Moreover, Eberhard merely reiterates the by then well-established Berkeley charge, and there are no references to Berkeleyan or dogmatic idealism in the Reflections at hand. Nor, for that matter, is there any evidence of the outrage that usually accompanies Kant’s rejection of the Berkeley charge. As well, in these notes Kant tries to come to terms with Pistorius’s central question about the ontological status of permanent Kantian objects.
In part, R5653 and 5654 make points already made elsewhere. Here, as in the A-edition Fourth Paralogism and the B-Refutation, Kant’s focus is on skeptical idealism. He variously reiterates his earlier arguments against skeptical idealism’s claim that we can know the existence of outer objects only inferentially and emphasizes, by contrast, the immediate and, as he calls it, originary access we have to such objects (as appearances). And, much as he has done earlier, Kant also responds repeatedly to the common dream/imagination objection. Although we do dream, of course, and imagine things in space, and may sometimes even confuse imaginary and real things, this does not, in his view, in any way undermine his point. For we can dream and imagine in the first place only because the materials for these fanciful operations of the mind were originally given through outer sense.
But Kant does more in these Reflections, particularly in R5653, than reiterate earlier arguments. As stated, he also attempts to respond to the central objection advanced by his contemporaries. Kant formulates it in this way: “I cannot be empirically conscious of the permanent outside of me, that is, [I cannot be conscious of it] as given in space, but only of my determination of their [its] representation insofar as I am affected by them [it] in accord with the form of space.” Although Kant does not mention any of his critics, the concession that we can only be aware of our “determination of the representation” of the permanent “in accord with the form of space,” not, one assumes, the thing itself as given in space, promises to address Pistorius’s complaint that Kant can consistently establish only the “logical existence of things,” or their existence as “appearance.” For Kant is far from granting the conclusion, uniformly defended by all of his early critics, that the actual existence of things must, on the presupposition of transcendental idealism, remain uncertain. Nor is he content to respond to the charge by appeal to his earlier arguments. While the basic argument of the B-Refutation remains intact-the existence of permanent objects in space is required for us to be conscious of our own existence in time-Kant takes the argument one step further.
What is at issue here is no longerjust what is required for inner sense, that is, for the awareness of myself as persisting through time. Kant now seeks to determine what is required for outer sense, that is, for the perception of the permanent objects in space outside of me that serve as the ground of my awareness of myself (as persisting through time). As his earlier arguments, including the much-repeated and maligned argument by affection, indicate, Kant had always been adamant that more is involved here than the consistent use of our cognitive processing operations. To perceive actual things in space sensations must also be given to us. In the text at hand, however, Kant no longer states this condition of outer sense by invoking the familiar argument by affection; he now makes it by appeal to what is best called the “determinability” argument. Rather than simply state that sensations must be given to us-from whatever source-Kant now claims that “[t]he intuition of a thing as outside of me presupposes the consciousness of a determinability of my subject in which I am not myself determining.” Contrasting it with the spontaneity proper to thought, he calls this “consciousness of a determination through other things” our “original passivity.”, That is said to be the ground or necessary condition that allows us to “represent things in space in intuition.”
Kant does not tell us why the intuition of an object in space requires not just that we are affected by something, but also the consciousness that we can be so affected or that we are passive in receiving sensations. Nor does he have a great deal to say about what the “consciousness of a determination through other things” amounts to. To understand Kant’s argument on this score, let me consider in somewhat greater detail what is required in order to recognize that something is permanent.
Kant tells us that in inner sense everything is in constant flux. His reasons for saying this can be made evident if we consider that he likely took the objects of inner sense to be a series of subjective states, like thoughts, feelings, and emotions, each of which exists only in time. Because they exist only in time, there is no criterion for ascribing identity to any of the states in this series. For example, a pain that I perceive now can no more be said to be the same pain that I perceived earlier than it can be said to be a different occurence of the same type of pain. There simply is no way of specifying, in a purely temporal sense, what would constitute the reappearance of the same object as distinguished from the emergence of a new but similar object. But if there is no criterion for the identity of objects. from one perception to the next, then everything must be in constant flux.
However, if I suppose that I exist and move around in space, then it becomes at least possible for me to forge a criterion for the identity of the same object over time, namely, continuity in existence. For objects in space can be supposed to have a position and to remain at that position (if at rest) or change it in a continuous and determinate fashion over time (if in motion). Consequently, I can take the identity of an object to be determined by, so to speak, tracing its world line back to its location at a previous time. In a purely temporal world I cannot do this. I cannot look back to where the pain was before and trace its continuity in existence on up to the present pain. Indeed, the very notion of continued unperceived existence of an object of inner sense is highly problematic. But if I suppose that I, as a perceiver, am located in a spatial environment, only a part of which I perceive, then the continued existence of unperceived locations is quite a coherent supposition. Thus, it is only with outer sense that identity, and with it unperceived existence, become notions that can even be modeled.
The implication of this account, that permanence entails repeatability of sensations, is central to the determinability argument. For insofar as I postulate that permanent objects exist, I must accept the counterfactual claim that if I were to move to a certain location in space, I would get a certain sensation. If I don’t think this, then I can’t think that there is anything permanent outside of me. Accepting that I would be affected in a certain way is just part of what is required for me to be able to think that there is something permanent. But insofar as I accept this, I must accept that I am passive and determined by something outside of myself. If I thought that I were active and cooking up my sensations myself (as when dreaming or imagining) then, again, I could not think that there is anything permanent outside of me. For, what I can think of myself spontaneously cooking up in imagination, I can think of myself refraining from cooking up in imagination, and if I can think that I could just as well not get the sensation of an object when I go to its location, then I am, in effect, denying that there is anything permanent there.
Permanence, therefore, requires a commitment to the essential passivity or determinability of myself as subject of experience. This is the new insight that carries Kant a step beyond the argument of the Refutation. For, he proceeds to argue, insofar as I am committed to this “original passivity,” I must think that there is something outside of or distinct from me that determines my state. I may not know anything about what this thing is like, but I can be sure that it exists outside of my representation.
The argument, in short, is a transcendental argument moving over the following stages:
(1) If I am to be aware of myself as an object of inner sense, I must be aware of objects outside of me in space (the argument of the B-Refutation).
(2) If I am to be aware of objects outside of me in space, I must accept that I am in fact passive or determined.
(3) Insofar as I think of myself as passive or determined, I must accept the existence of objects distinct from myself.
Note that it is only in virtue of the second step (not in virtue of the argument of the B-Refutation-step i) that Kant can affirm the existence of objects truly distinct from me, not just the existence of objects outside of me in space (appearances). It is only at this point, in other words, that he has managed to separate this previously confused distinction.
The argument plays itself out in the following way. The first thing to emphasize here is that it is based on a new insight about the subject, not about affecting objects. Consider a moment what the argument implies. It posits a new transcendental characteristic of the subject, a transcendental condition of sensation or receptivity that complements the one of thought. Just as the transcendental unity of apperception expresses a consciousness of myself as a single unified subject, a consciousness Kant rightly deems necessary for thought, so the consciousness of myself as a subject who can be affected or determined by things other than myself expresses another essential feature of the subject: that I stand in relation to a sphere of things distinct from myself. If I was not conscious of myself as standing in relation to distinct and independent permanent objects that affect me, I could not be conscious of myself as determinable or passive. This consciousness is as much a consciousness of the specific empirical objects in space (which I constitute and know) as it is a consciousnesss of things that are “truly” independent of and distinct from me. I may not know such things as they are in themselves-indeed, given the unknowability thesis (which Kant does not deny), I cannot know them as they are in themselves. Nevertheless, given the consciousness of my determinability, I am certain of their existence. For the consciousness of my determinability places me, considered as a single unified subject effecting a unity of apperception, in the context of and in a relation to a sphere of things that actually exist as distinct from myself and which I must think of as affecting me. Moreover, I represent these things in the only way in which I can do so, given my aesthetic/cognitive makeup-in space.
All of this adds up to a number of interesting facts. First, what I am directly aware of are the objects of my intentional acts, namely, empirical objects in space (appearances). But that awareness is at the same time a certainty of independently existing things that affect me. I represent these things as in space, the form of intuition that can capture relations of distinct things. Second, the argument shows, by implication, why Kant can legitimately say, as he does repeatedly, both that the objects that we constitute cannot be in us—even though, as constituted, such objects are strictly speaking “in us”—and that we “cannot think of space as in” us, even though, as a form of intuition, space is strictly speaking also “in us.” For space serves to represent the realm of things outside of ourselves, of which, given the determinability argument, we are aware, and that realm is independent of and distinct from ourselves. By the same token, empirical objects are, as it were, placeholders for transcendental objects that I now know, and do not just suppose, to exist but the particular natures of which must remain inaccessible. Hence they must similarly be represented as outside of us and distinct from us.
For Kant all of this adds up to a proof of dualism, but a dualism strikingly different from that defended in the A-edition Fourth Paralogism. There it is one of inner and outer representations, where outer representations are called outer because “they relate perceptions to space … [which] is in us.” In the later Reflection, the dualism is of two distinct senses, two distinct relations to the subject and, as a result, two distinct spheres of “things.” By inner sense we are related to ourselves (as appearances), our psychological events and states, in the successive ordering of time. Inner sense delivers representations only as related to ourselves; it cannot generate relations of ourselves to other things. Nor can it generate the intuition of space. Outer sense, by contrast, does just that. In virtue of the consciousness of ourselves as “immediately related to something that exists as distinct from the subject” we can have a “consciousness of the object as thing outside of” ourselves. Inner and outer sense are distinct senses, equally necessary for knowledge, yet also closely connected and necessary for each other. Combining the Refutation and the determinability argument Kant states that “the consciousness of myself … [is] at the same time a consciousness of an outer relation.”
Supposing that Kant is correct to posit this transcendental condition for outer sense and, with that, for the perception of space and things in space, that parallels the transcendental condition for thought, does it demonstrate the existence of a sphere of things independently of and distinct from us? And would Pistorius and company accept this argument? The latter question can most likely be answered in the negative, if only because by this time they were so convinced of Kant’s idealism that no argument would have sufficed to persuade them to change their minds. On the basis of what we have seen of their arguments in this paper, we can speculate that they might raise three objections. They might suggest, first, that outer sense could still reduce to inner sense, that outer objects reduce to things only “in me.” Kant has an answer to this objection, however. One facet of this answer is the by now familiar claim that fanciful operations of the mind are possible only because the material for them is originally given through the senses. A more interesting answer takes us back to the argument of the B-Refutation. Kant points out, repeatedly, that if outer sense were to reduce to inner sense, then inner sense itself, that is, the consciousness of ourselves as persisting through time, would be impossible since that is contingent on the permanent in space outside of us (viz., outer sense). For the reasons indicated above, inner sense cannot generate either space or outer relations. Outer sense cannot in fact reduce to inner sense. Harping on the established criticism, his contemporaries might argue secondly that, since knowledge remains restricted to appearances, we still cannot know the independent things that are ostensibly distinct from us. That would be true enough, and a point Kant would have to concede (given the unknowability of things in themselves). But the objection would surely miss the point. For unlike his earlier arguments, the “determinability” argument does not hide behind empirical objects. Nor does it need to posit a knowledge of the nature of things in themselves. For Kant’s purposes, it is enough that the argument establishes that we can nevertheless know them to exist. In other words, if Kant is correct to think that without the awareness posited by the determinability argument we could not perceive space, and, by implication, time, then he is justified in positing the existence not just of empirical objects in space (appearances), but also of wholly independent or transcendental objects. Third, Kant’s contemporaries would likely raise the causal objection by suggesting that this argument makes an illegitimate causal claim about things in themselves. It does maintain, after all, that they affect us. Again, this is a point Kant would have to concede. He could, however, point out that there is this important difference between the causal argument by affection and the determinability argument. Unlike the former, this argument does not attempt to establish the existence of things in themselves by applying the causal principle. It makes its point, rather, by analysis of the claim of my determinability. It is an argument, in other words, about the subject, specifically, about what the subject’s determinability commits us to. If Kant is right to claim that I cannot spontaneously generate space and things in space, and that the perceptions of outer sense are contingent on the consciousness of my “original passivity,” then he is committed, indeed forced, to think that distinct things do exist independently of us. Of course, he must also think that they affect us, but the argument is carried by the new realization about the subject, not by the fact that these objects affect us. The new realization about the subject leads us equally to independently existing outer objects and affection. Now, whether the determinability argument should be accepted is, of course, another question, and one that requires further investigation of the remaining Reflections and discussions of idealism.
Once the development of Kant’s idealism arguments is considered in the context of the criticisms with which Kant was faced, it is readily apparent that he did eventually come to be influenced by the substance of these criticisms. While the early (and direct) responses, particularly those of the Prolegomena, were more a function of Kant’s outrage at the Gottingen review and generated surface changes only, by the time he wrote the B-edition Kant had come to seriously reconsider the issue of idealism. He was careful here to present matters in such a way that he could not be immediately accused of empirical idealism. And once he made the requisite (and possibly empty) concession that neither the empirical object in space nor space itself are “in us,” he was similarly forced to rethink the argument of the Refutation. But it was not until after the B-edition that he did, finally, seriously respond to his contemporaries’ overwhelming concern regarding the actual existence of things. He seems to have taken Pistorius’s criticism that the existence of things in themselves cannot be demonstrated within the framework of transcendental idealism seriously. And once he did so, he also rose to the challenge with the “determinability” argument. That is an argument that takes a strongly realist turn while remaining in the context of transcendental idealism. It may be that we can ultimately know only empirical objects, yet, if the “determinability” argument is accepted, we can similarly be immediately, that is, noninferentially, certain that transcendental objects or things in themselves exist. Nor does Kant need to deny any of the features of the critical philosophy or to make illegitimate causal claims about things in themselves. This argument merely specifies a further transcendental condition of the possibility of experience, one, furthermore, that is nicely symmetrical with the condition of thought and as such appears legitimate.