Material Dependence and Kant’s Refutation of Idealism

The paper argues that in the Critique of Pure Reason Kant develops two anti-sceptical strategies. In the Fourth Paralogism ( CPR A) he believes himself able to refute the sceptic by demonstrating that external perception is immediate. This strategy is rather unconvincing. In the Refutation of Idealism ( CPR B) Kant promotes the material dependence of inner sense on outer sense . I show that Kant’s argument for material dependence has been widely overlooked, even though it is the strongest argument against external world scepticism he develops, since it anticipates mental content externalism while preserving transcendental idealism and empirical realism.


Introduction
Kant did not conceive of his philosophical project as an antisceptical mission. From the outset of his critical work, he aimed rather to rebut claims to metaphysical knowledge and to restrict the possibility of cognition to the transcendental conditions of experience. Yet the claim that human cognition is essentially restricted is not itself a sceptical claim, since Kant allows for a good deal of cognition, such as the transcendental conditions a priori of the possibility of cognition, that he takes to withstand sceptical attacks because their objective validity can be demonstrated or justified. Nonetheless, Kant has been reproached with scepticism ever since the publication of the Critique of Pure Reason. This allegation didn't come as a surprise. For contrary to the determination of the transcendental conditions of cognition, Kant's transcendental idealism seems to suggest that we can know nothing over and above mental representations of spatio-temporal objects, and that objects which we represent do not exist independently of the mind. As Kant puts it: "everything intuited in space or in time, hence all objects of an experience possible for us, are nothing but appearances, i.e., mere representations, which, as they are represented, as extended beings or series of alterations, have outside our thoughts no existence grounded in itself." . 1 In what follows, I will not offer yet another interpretation of the distinction between appearance and thing-in-itself. Rather, I argue that this distinction does not have sceptical implications and that transcendental idealism in fact offers an intriguing anti-sceptical strategy. More specifically, Kant's strategy is to demonstrate what can be called the material dependence of inner sense on outer sense, which anticipates what in more recent times has been termed mental content externalism. 2 Mental content externalism is the view that mental representations depend on extra-mental content and the way in which cognizers relate to that content. The Kantian version of mental content externalism is the material dependence of inner sense on outer sense, i.e., the view that mental representations in inner sense depend on the content given in outer sense such that we can only have inner mental representations if we have external representations of some sort. Kant firmly believes that material dependence resists any sceptical misinterpretation of the distinction between appearance and thing-in-itself and more generally scepticism about the external world. He develops 1 3 this strategy in the wake of the sceptical hypothesis. To show this, I set out in Sect. 2 by identifying key features of the sceptical hypothesis and correlating them with Kant's distinction between appearance and thing-in-itself. In Sect. 3 I apply these features to Kant's first explicit attempt to refute the sceptic, the Fourth Paralogism of the first edition of the Critique of Pure Reason. There Kant portrays a Cartesian sceptical scenario. As we will see, Kant is fully aware of the destructive force of the sceptical hypothesis and the need of an anti-sceptical strategy to overcome doubts about the reality of the external world. The strategy he offers in the Fourth Paralogism consists in the claim that external perception is as immediate, and thus as certain, as inner perception of one's own existence. I argue that this strategy fails in refuting the sceptic. Although the Fourth Paralogism addresses the question of epistemic access to the content of external representations, it does not implement the material dependence of inner sense on outer sense as such. This changes in the second edition of the Critique of Pure Reason. In Sect. 4 I show that Kant's strategy in the Refutation of Idealism is based on material dependence and that material dependence can be read as a variant of mental content externalism. Kant demonstrates that content-wise, perceptual beliefs or representations in inner sense depend on what is given in passive outer sense, such that the world is ontologically mind-independent and cannot be reduced to mere representations as the sceptic claims. Finally, in Sect. 5 I sketch why Kant does not simply presuppose material dependence to make a sufficiently strong case against the sceptic. Kant's anti-sceptical strategy rather originates in the undeniable finitude of discursive human thought that grounds transcendental idealism.

The Sceptical Hypothesis and the Distinction Between Appearance and Thing-In-Itself
The aim of this section is to explicate why transcendental idealism should not be conceived as a sceptical doctrine. For Kant does not claim that we can only know representations of objects and not true reality, i.e., objects as they are in themselves. Ruling out sceptical interpretations of the distinction between appearance and thing-in-itself is a prerequisite for appreciating Kant's anti-sceptical strategy because, as we will see below, the success of this strategy depends on a non-sceptical reading of transcendental idealism. This is no trivial observation, since it has been argued that Kant's anti-sceptical strategy is promising but by no means hinges on transcendental idealism. 3 By contrast, I argue that Kant's anti-sceptical strategy is an integral part of transcendental idealism.
The strongest sceptical interpretation of transcendental idealism is modelled after the classical sceptical hypothesis. A sceptical hypothesis (sh) is an abstract description of a non-excludable radically sceptical possibility. The somewhat standard version goes as follows: (1) I do not know that not-sh.
(2) If I do not know that not-sh, I do not know that p.
Conclusion: I do not know that p. To get this argument going, sceptical hypotheses such as the dream-and the evil demon-argument or its modern brain in a vat-versions are forced to make at least three assumptions: (1) We cannot know with certainty that or how reality is different from what we believe.
(2) Truth is independent of our beliefs, i.e., our beliefs do not constitute reality.
(3) Our beliefs might be brought about through something that is in itself epistemically inaccessible to us. (1) Captures the very idea of a sceptical hypothesis in that the sceptic wants us to accept that cognizers do not have the epistemic means to know whether or not they are ultimately mistaken in what they believe, e.g., because they are dreaming or deceived. This presupposes (2) that mental or cognitive states such as believing do not have an impact on the truth or objectivity of our beliefs, for if they did, it would be possible that we could-contrary to (1)-know how reality is different from what we believe. That reality might be different from what we believe, and that objective reality does not depend on beliefs we entertain, presupposes that (3) our beliefs might have causes that are cognitiontranscendent, for we cannot be sure about what the origin of our beliefs is.-I take these three assumptions of sceptical hypotheses to be uncontroversial and will therefore not engage in a nonetheless worthwhile discussion of further epistemic and/or modal aspects that sceptical hypotheses involve. Instead I will show how these three assumptions have been tacitly incorporated in sceptical interpretations of transcendental idealism, and that sceptical interpretations of the distinction between appearance and thing-in-itself must be ruled out in order to lay out Kant's anti-sceptical strategy.
In the immediate aftermath of the publication of the Critique of Pure Reason Kant faced the objection that the distinction between appearance and thing-in-itself has unavoidable sceptical consequences. 4 These sceptical interpretations of transcendental idealism are implicitly or explicitly modelled after the sceptical hypothesis sketched above. Accordingly, Kant supposedly claims: (1*) We cannot know what reality truly is since we can only cognize appearances. (2*) Things-in-themselves represent true, non-subjective reality, and since things-in-themselves are independent of our cognitive capacities, Kant is a realist about truth. (3*) Appearances are caused by things-in-themselves, and things-in-themselves are epistemically inaccessible to us.
Prima facie, these three claims correspond to the classical sceptical hypothesis, whereby transcendental idealism willingly or unwillingly supports sceptical doubts about external reality. For transcendental idealists seem to hold that cognition is restricted to appearances, that truth or true reality goes beyond what can be known, and that, as a matter of fact, the true cause of what we believe cannot be known: (1**) Epistemic restriction: We can only cognize appearances [ (1)  These core claims attributed to transcendental idealism supposedly reveal the correspondence between the sceptical hypothesis and transcendental idealism. For transcendental idealists must, under this interpretation, endorse the claim that cognizers as a matter of principle cannot rule out that mental representations are misguided: (A) We cannot cognize things-in-themselves. (B) If we cannot cognize things-in-themselves, we cannot know whether or not appearances correctly represent external reality. (C) We cannot know whether appearances correctly represent external reality.
The alleged scepticism of transcendental idealism is that cognizers only have epistemic access to appearances, that appearances depend on things-in-themselves as their true causes but that things-in-themselves are cognition-transcendent. Now if they are, we cannot know whether appearances do in fact represent reality as it is (things-in-themselves) such that, as in the sceptical hypothesis, we cannot rule out the possibility that our beliefs are brought about through something other than things-in-themselves, e.g., an evil demon. Much of this sceptical take on transcendental idealism hinges on the interpretation of things-in-themselves as cognition-transcendent causes of appearances as their mental effects. As already stated, here I will not engage in the complicated question of the correct interpretation of the distinction between appearance and thing-in-itself. 5 In what follows, I rather argue that in the Fourth Paralogism Kant himself considers the kind of sceptical hypothesis that critics have imposed on transcendental idealism. Kant's aim is to show that the distinction between appearance and thing-initself has sceptical consequences only under the assumption of the three main presuppositions of a sceptical hypothesis, which however transcendental idealists do not accept. Kant's strategy is in fact to rule out the sceptical hypothesis by rejecting the model of indirect causal perception of external objects in favour of the model of direct causal perception of external objects.

The Anti-sceptical Strategy of the Fourth Paralogism: Direct Causal Perception of External Objects
The critical Kant details his response to scepticism for the first time in the Fourth Paralogism of the first edition of the Critique of Pure Reason (A 366 ff). The anti-sceptical strategy developed in 1781 is significantly different from the one he puts forward some six years later in the Refutation of Idealism. Although the anti-sceptical argument of the Fourth Paralogism leaves several problems unsolved (see below), it is nevertheless worthy of consideration since it reveals Kant's acumen and diagnostic abilities concerning sceptical scenarios in general and the sceptical hypothesis in particular. In this section, I first briefly present the Fourth Paralogism and then show that the way Kant sets it up evidently corresponds to the structure of the sceptical hypothesis. Next, I discuss his strategy to rebut the sceptical conclusion of the Fourth Paralogism, and subsequently point to two major weaknesses of his anti-sceptical strategy in the first edition of the Critique of Pure Reason. Kant depicts the "Fourth Paralogism of the ideality (of outer relation)" by means of a Cartesian scenario and maintains that Cartesians typically argue as follows: (1) "That whose existence can be inferred only as a cause of given perceptions has only a doubtful existence".
(2) "Now all outer appearances are of this kind: their existence cannot be immediately perceived, but can be inferred only as the cause of given perceptions". C: "Thus the existence of all objects of outer sense is doubtful" (CPR A 366-367).
For the purpose of this paper, it is not relevant whether or not Kant's reconstruction of this inference is historically accurate. Although Kant is aware that Descartes is not a selfproclaimed sceptic, he is mainly interested in the sceptical consequences that he thinks arise from Cartesian ontological dualism. Accordingly, Cartesians argue that our knowledge of the existence of external objects is mediated through the causal inference from the perceptual representation of an external object to the external object itself outside us. However, causal inferences from a mental representation in me to an external object outside me are epistemically uncertain, since we cannot know whether our beliefs about the external world are true, and ultimately whether the external cause of my representation of an object outside me really exists: the "cause of given perceptions" (CPR, A 366), Kant argues, could in principle be produced by the perceiver herself and be "a mere play of our inner sense" (CPR, A 368). As a matter of principle, we therefore cannot know with certainty that there are external objects, since we cannot be sure about the actual causes of our perceptual mental states: "for here, even with our best consciousness of our representation of these things, it is obviously far from certain that if the representation exists, then the object corresponding to it would also exist" (CPR, A 371). Alluding to Descartes' cogitoargument, Kant points out that only the object of inner selfperception, "I myself with all my representations" (CPR, A 368), is indubitable since I am immediately aware of myself and the representations I have.
Given the set-up of this reasoning, all previously identified features of the sceptical hypothesis apply to the Fourth Paralogism. First, knowledge is restricted to the representations which cognizers have, second, reality (or truth) goes beyond what we can know, for we cannot know whether our external world beliefs are true, and third, the true cause of our representations or beliefs is epistemically inaccessible. Thus, as in the Cartesian dream-or evil demon-argument, the cause of our representations or beliefs is cognition-transcendent. Even with the best scientific means in our hands we would not be able to discover the truth of our claims about the external world.
From a systematic point of view, Kant's diagnosis of the Fourth Paralogism is far-reaching. The blind spot of the argument is, as he clearly sees, the claim that perception of external objects is only indirect, that is, causally mediated. Kant agrees with two points of the Cartesian reasoning: first, that the existence of what is immediately perceived is indubitable, and second, that the relation between the perceiver and the perceived external object is a causal relation. 6 What Kant denies, though, is the model of indirect causal perception of external objects as presupposed in the sceptical hypothesis of the Fourth Paralogism. That perceptions and beliefs about the external world depend on causal import cannot be doubted. The question, however, is whether this causal import must be conceived in terms of a causal inference in such a manner that the cause of perceptions and beliefs we hold about the external world would be beyond the reach of our cognitive capacities, as are physical objects in Kant's version of Cartesian ontological dualism. In this case the perceiver would infer from the mere awareness of an external representation that this representation must have a cause in the external world, while not being able to epistemically access the cause itself. What makes the causal perception of external objects indirect is the accessory cognitive operation the perceiver must perform with respect to the existence of the perceptual cause. The external cause is not immediately present in the perceiver's perception but only represented by a mental image or an idea in the perceiver's perceptual state. The cause itself is cognition-transcendent because of the unbreachable gap between the inner and the outer in Cartesian dualism.
This reasoning suggests that for Kant external perception is not indirect but direct, and that external perception is not to be conceived of in terms of mental imagery, i.e., representation of external objects mediated via internal mental placeholders. This is indeed the case. The perceptual model that Kant proposes as an alternative is the model of direct causal perception of external objects. The anti-sceptical aim of this model is to show that external perception is immediate, and therefore no extra inference from the inner to the outer is required to assure us of the existence of the external cause of outer perception. This is so because in external perception, we are related directly to objects in space outside us. 7 -What are the reasons for holding such a model? Kant starts from the idea that inner as well as outer perception is immediate and that immediate perceptual reference to external objects shows that the objects of outer sense in fact exist. For they do not exist "independently of us and our sensibility" (CPR, A 369). What Kant means is not that for external objects to exist they must be sensibly perceived by a 6 Since Kant does not give up his first edition strategy in the second edition, this seems to speak against Chignell's (2010Chignell's ( , 2011) recent critique of causal readings of the Refutation (see below). 7 From the fact that for Kant external perception is direct, it does not follow that he is a relationalist with respect to external perception, i.e., a naïve or direct realist. For relationalists typically argue that intuition is object-dependent in the strict sense such that if a cognizer intuits object o, then o really exists. If the cognizer claims to intuit o but o does not exist, then the cognizer is hallucinating or under an illusion (on Kant's alleged relationalism cf. Allais 2015). Further below I show that Kant does not hold object-dependence of intuition. perceiver. This would be a view almost identical with Berkeley's esse est percipi-principle, a view he clearly rejects. 8 That external objects do not exist "independently of us and our sensibility" rather means that there are transcendental conditions of cognition imposed on objects of perception, i.e., space, time, and the categories, such that what can possibly be an object of external perception for us stands under those conditions. We therefore directly refer to those objects in perception since, unlike in Cartesian dualism, they are not excluded from the domain of what we can cognize directly. Most importantly, space (as well as time) is not a thing-initself or a property of a thing-in-itself ("something given in themselves", CPR, A 369) but a pure form of our sensibility. This is why outer perception is, for Kant, immediate and does not require any causal inference from sense-perception to external objects as its causes. Objects in space can be perceived directly and there is no need to infer their existence after perceiving them.
Kant claims that transcendental idealism substantiates this. Objects considered as things-in-themselves independent of our sensibility are not (empirically) real, because for the transcendental idealist "matter" "is only a species of representations (intuition), which are called external, not as if they related to objects that are external in themselves but because they relate perceptions to space, where all things are external to one another, but that space itself is in us." (CPR, A 370). The term "a species of representations" is rather unfortunate since it seems to insinuate that external objects ("matter") are nothing more than representations. But this is not Kant's claim, for what he means is that, although external objects are ontologically independent given that their existence does not depend on their being perceived, they do not constitute a domain that transcends the field of possible cognition such as the Cartesian extended substance transcends the thinking substance. For the transcendental idealist, external objects are not objects that transcend our substance, since they are spatial, and, as Kant says, "space itself is in us", i.e., is the form a priori of outer sense.
On the face of it, this claim seems hard to defend. But Kant at least implicitly reminds us of the "Transcendental Aesthetic" where he argues against the claim that space is an ontologically independent (property of an) object of outer perception. By contrast, in transcendental idealism space is a formal precondition for any external perception and as such is imposed on objects we perceive outside us. 9 Only in this sense is space in us as a transcendental condition of experience whereby it safeguards us from the sceptical consequences Cartesians cannot avoid. However, even if we accept Kant's view that the world is ontologically independent, yet epistemically dependent in the sense that it cannot go beyond what we can possibly know, one may not be inclined to follow his anti-sceptical strategy to justify perceptual immediacy simply by relying on the claim that "space is in us". There are two fundamental difficulties with this strategy: First, the argument against the sceptical conclusion of the Fourth Paralogism is only indirect or apagogic. Kant opposes indirect and direct external perception and then demonstrates that external perception is in fact direct, so that sceptical doubts about external reality cannot arise. However, he does not give a direct proof of why, on his account, external perception is held to be immediate, nor of why the fact of immediate perception of external objects safeguards from sceptical doubts. In this regard his reliance on space being in us is unsatisfactory, for one would have expected Kant to, at the very least, elaborate on how space being a form of sensible intuition has the consequence that external perception is immediate. The second difficulty stands in close connection with the first. Even if it is the case that external perception is direct, and that we therefore do not infer the existence of external objects from what we perceive in us, it does not follow that external objects really exist. For it might nevertheless be the case that we are dreaming or deceived in our perceptual beliefs. Although Kant would insist that he has dissolved the ground of the claim that spatial objects transcend beyond cognition, the constitutive ground the sceptical hypothesis of the Fourth Paralogism must invoke, he would probably acknowledge that he is unable to provide a robust criterion for the distinction of waking-from dream-states such that we can know that our external perceptions do not result from "a mere play of our inner sense". In this respect his anti-sceptical strategy of the model of direct causal perception of external objects is unconvincing. But as we will see in the following section, Kant is rather unambitious with regards to providing a knock-down argument against the sceptic. Instead, the aim of his anti-sceptical strategy is to convince us that we have very good reasons not to follow the sceptic. 10 1 3

The Anti-sceptical Strategy of the Refutation of Idealism: Material Dependence and Mental Content Externalism
As in his moral philosophy, Kant, in his theoretical philosophy, is not primarily aiming to overcome scepticism. In the Groundwork of the Metaphysics of Morals, for instance, he aims above all at demonstrating "on what grounds the moral law is binding" (Kant 1999b, p. 97; AA IV 450) rather than persuading the moral sceptic to concede that morality in general cannot be put into doubt. Kant is convinced that those who doubt the common sense belief in morality and ask the sceptical question Why be moral at all? are beyond help. 11 Likewise, the predominant aim of the Critique of Pure Reason is not to refute the sceptic but to lay out the transcendental conditions of the possibility of cognition, which show that scepticism is ungrounded only subsequently. On the other hand, in the second edition of the Critique of Pure Reason (1787) Kant saw himself forced to react to quite some extent to criticisms of the first edition as a whole, and particularly of transcendental idealism, criticisms he already struggled with in the Prolegomena to any Future Metaphysics (1783) and with which he remained occupied in the second edition. What specifically upset him was the objection of Feder and Garve in their review of the first edition of the Critique, according to which transcendental idealism ultimately promotes external world scepticism. 12 Kant's new response to scepticism in the second edition of the first Critique is the Refutation of Idealism. In this section he proposes an anti-sceptical strategy different from the one in the Fourth Paralogism of the first edition. At first glance, the Refutation looks like a self-standing proof that is somewhat disconnected from the overall argument for transcendental idealism. This is, however, not the case. As I argue, Kant's anti-sceptical strategy in the Refutation hinges on transcendental idealism and the theory of the faculties of cognition that underlies it. In what follows, I first specify the target and reconstruct the argument of the Refutation. I then classify relevant readings of the Refutation that have been proposed and focus on a more recent criticism of socalled causal interpretations of the Refutation. According to this criticism, contrary to what many commentators hold, it is not the case that external objects are necessary for inner experience such that given undeniable inner experience, the sceptic must concede the existence of external objects. Contrary to this interpretation, I argue for the material dependence of inner sense on outer sense to show that the anti-sceptical strategy of the Refutation anticipates what nowadays is termed mental content externalism. To motivate the material dependence of inner sense on outer sense, I elaborate on Kant's theory of inner sense and discuss his anti-sceptical claim that (originally) the content that is given in outer sense is necessary for the possibility of content of inner sense. The main reason for this Kantian version of mental content externalism is, as I conclude, the discursive nature of finite human cognition.
Although Kant terms his proof a Refutation of Idealism, what is on his mind is actually a Refutation of scepticism about external objects. This is clear from his definition of "idealism": "Idealism (I mean material idealism) is the theory that declares the existence of objects in space outside us to be either merely doubtful and indemonstrable, or else false and impossible" (CPR B 274). The Refutation of Idealism targets the first kind of idealism, i.e., the "problematic idealism of Descartes", and not the latter, i.e., the "dogmatic idealism of Berkeley" (CPR B 274), according to which unperceived external objects do not exist. 13 Cartesian problematic idealism (scepticism) claims that cognizers do not have the epistemic capacities to prove the existence of external objects outside us by means of "immediate experience". To show that the experience of objects outside us is not illusory, the Refutation must therefore demonstrate "that even our inner experience, undoubted by Descartes, is possible only under the presupposition of outer experience" (CPR B 275-276). As we will see in more detail below, the claim that inner experience is conditional on outer experience is at the heart of the material dependence of inner sense on outer sense and Kant's version of mental content externalism.
Since the Cartesian sceptic takes the existence of objects outside us to be "doubtful" and "indemonstrable" because our cognitive capacities are not suited to give a sufficient proof, the Refutation of Idealism, like the Fourth Paralogism, must be construed as a sceptical hypothesis. Like in the Fourth Paralogism, the main charge is that the causal inference from "given effects", i.e., mental representations of external objects, to "determinate causes" is "only unreliable, since the cause of the representations that we perhaps falsely ascribe to outer things can also lie in us" (CPR B 276). What is immediately perceived, however, is indubitable, which, for the Cartesian, is merely the immediate inner experience "I am". Thus, in the Refutation, too, knowledge is restricted to the representations which cognizers have, reality or truth is beyond what cognizers can know, and the true cause of representations or beliefs is epistemically inaccessible for cognizers.
Yet, unlike in the Fourth Paralogism, Kant's anti-sceptical strategy in the Refutation is not indirect but direct. Kant is not drawing on the proof that both external and internal perception are immediate, although he does not give up this view. Rather, his key aim is to demonstrate that what the sceptic admits, i.e., "the empirically determined, consciousness of my own existence" ("I am"), is only possible under the condition of "the existence of objects in space outside me" (CPR B 275; cf. AA 18:610). The proof of the Refutation runs as follows: (1) "I am conscious of my existence as determined in time." (2) "All time-determination presupposes something persistent in perception." (3) The persistent cannot be an intuition in me.
(4) Therefore, the persistent must be "a thing outside me" rather than a mere representation of an external thing outside me. (5) Therefore: "The mere, but empirically determined, consciousness of my own existence proves the existence of objects in space outside me." (CPR B 275).
Kant reformulates this conclusion thus: "the consciousness of my own existence is at the same time an immediate consciousness of the existence of other things outside me" (CPR B 275). This move is legitimate in view of the claim he makes about time-determination. The critical move of the argument, however, is the ontological status of the persistent [ (3) and (4)]. The determination of my existence in time, Kant argues, presupposes a persistent thing outside me in relation to which my existence can be temporally determined, but this persistent thing cannot be a representation in me. For representations are not persistent in time as are permanent objects in space. Persisting things in space outside me therefore serve as the permanent (das Beharrliche) for time-determination in relation to which I can have knowledge of my persistent existence in time in spite of the fact that my representations undergo constant temporal changes. Since I have the temporally determined awareness that I exist now, and since time-determination is possible only through persistent objects in space outside me, the consciousness "I am" is an implicit, yet direct consciousness of the existence of objects outside me. The difficulty with this argument is that it seems to be doubtful or at least unclear why the persistent outside us in space cannot be a hallucination or a mere illusion caused in us, e.g., by an evil demon or even by ourselves through cognitive self-deception. To simply reiterate the fact that, as the first First Analogy (CPR A 182/B 224-A189-B 232) demonstrates, time-determination requires a permanent thing (object), which cannot be found in inner sense since all representations in inner sense are non-persistent, i.e., fleeting and fading in time, so that the thing can only exist in space, merely begs the question. Kant simply does not provide an explicit argument for why the persistent necessary for time-determination cannot merely be a hallucination or an illusion.
One way to respond to this concern that appears promising prima facie is to interpret Kant's anti-sceptical strategy in the Refutation in terms of perceptual relationalism. As indicated above (fn. 7), relationalism is the view that in perception, cognizers are immediately in touch with the objects of perception such that there is no mediation between perceived objects and our awareness of them through intramental placeholders or proxies. On the relationalist interpretation, then, Kantian intuition is object-dependent such that if we intuit an external object, the object is really there. Consequently, hallucinations or mental illusions cannot count as intuitions, for by definition their objects do not exist, at least not the way we take them to exist. 14 As we have seen in Sect. 3, Kant indeed seems to hold such a view, at least with respect to perception. For in the Fourth Paralogism, he argues that in external perception (intuition) we are directly or immediately related to objects in space because space is the form of outer intuition. Likewise in the Refutation he claims that "here it is proved that outer experience is really immediate" (CPR B 276). But does this mean that Kant is a perceptual relationalist such that his strategy in the Refutation is to point to the fact that the determination of our existence in time requires a persistent thing outside us and that this persistent thing is intuited in space and therefore not a hallucination or an illusion? Whatever perceptual relationalism amounts to in the end, I do not think that Kant holds object-dependence. For object-dependence stipulates that hallucinations or illusions cannot count as intuitions. But this is not the case for Kant, as the "stepladder" of representation (CPR A 219-220/B 276-277) shows: in the hierarchy of the organization of our mind (conscious) "representation" is the genus which divides into subjective and objective representation. The former is "sensation (sensatio)", the latter is "cognition (cognitio)". Now "cognition" as objective representation is either "intuition" or "concept" ("intuitus vel conceptus"). A third kind of objective representation is not possible since intuition is repraesentatio singularis and concept is repraesentatio generalis, i.e., an objective representation is either singular or universal. Where, then, would hallucinations or illusions fit in? To classify them as sensations is not possible since we do not sense or feel them but refer to them (although unsuccessfully) as objects of cognition. And since hallucinations or illusions cannot be classified as concepts because they are not general, i.e., abstract representations, the only possibility we are left with is that they are intuitions. For otherwise we would need to introduce a third mental class of mental entities besides intuition and concept. But such a third mental class is not available in Kant. What are the consequences of this for the Refutation of Idealism? The immediate consequence is that the intuited, perceived persistent in space that makes the undeniable determination of my existence in time possible is not, qua intuition, necessarily an actually existing thing but could in principle be a hallucination or illusion. For "[t]he difference between truth and dream, […] is not decided through the quality of the representations that are referred to objects" (Prolegomena, Kant 2002;IV:290). To accept Kant's claims about the determination of my existence in time and the persistent thing as its precondition therefore would not convince the sceptic to accept Kant's proof. The reconstruction of Kant's strategy in the Refutation must take another direction. In the remainder of the paper, I will focus on what I think is the more promising argument that Kant puts forth for the claim that the persistent thing cannot be in me but must exist in space outside me. To start with, I will briefly address some prominent interpretations of the Refutation and argue that they fail to take into account the main reason for why Kant thinks that the persistent must be outside me. I then propose an alternative reading, according to which the Refutation argues for the material dependence of inner sense on outer sense and connect this view with Kantian mental content externalism.
The quantity and quality of literature that has been produced on the Refutation of Idealism is breath-taking and cannot be considered here to the full extent. An examination of some standard readings that have been offered, however, is in order to allow for the contextualization of the proposal that I am going to make. We can classify these standard readings as follows: -Metaphysical reading: On this reading, the aim of the Refutation is to prove the existence of things-in-themselves as the underlying, persistent cause of appearances. 15 -Semantic reading: On the semantic reading the Refutation must be semantically transformed, e.g., along the lines of Wittgenstein's private language-argument, and/ or be construed as a transcendental proof. 16 -Psychological reading: The psychological reading understands the Refutation's temporally determined consciousness as consciousness of one's own biographical data as well as memory consciousness. 17 All of these readings are instructive in so far as they advance the debate on the Refutation, and some of them might even be complementary. As will become clear in what follows, however, none of these readings addresses what I take to be the obvious key point of the Refutation, i.e., the material dependence of inner sense on outer sense. Of course, as all commentators notice, the issue is to prove that the persistent necessary for time-determination is in outer rather than inner sense. The question, however, of why inner sense is incapable of providing what is necessary for time-determination is more or less ignored. At this point, it would be useful to add a fourth reading to the classification: the non-causal reading of the Refutation. This reading differs from other readings in that it attacks what most of the other readings implicitly agree on, i.e., the causal nature of the relation between what we are aware of in inner sense and what is given in outer sense. The non-causal reading, defended by Chignell (2010), is not proposing a new positive reading. Chignell rather attempts to clarify the debate by demonstrating that what since Guyer (1988) have been conceived as causal readings are unsuccessful all the way down. 18 On Chignell's critique of causal readings, there is no causal correlation between the succession of inner states and the succession of outer states, and thus external objects are not required for experience: "why could not my empirical self be the enduring item some of whose other states are perceived as severally simultaneous with a given internal succession?" (Chignell 2010, 496). 19 What Chignell has in mind here, which reflects his overall strategy, is that the attempt to causally connect inner and outer sense in such a way that particulars in outer sense are preconditions for experience in inner sense is hopeless: "it is very hard to find plausible principles that can take us from shared features of our inner experience to the world of external objects in the way causal refutations of idealism suggest." (Chignell 2011, p. 186). Chignell thus believes that the Refutation is not a proof on the basis of empirical time-determination of our mental states: "Kant's anti-sceptical argument is meant to be a priori" (Chignell 2011, p. 185). If this were so, then Kant would have overcome the sceptical hypothesis of the Refutation. This, however, doesn't seem to be the case since in concluding he explicitly confesses that he is unable to rule out once and for all that our external world beliefs may be caused by "dreams" and "delusions" (CPR B 279).
Although Chignell is as unconvinced as I am by the (causal) readings classified above, my reasons are different than his. Whereas Chignell questions the causal approach to the Refutation as such, I do not deny that an inner awareness of temporally determined mental states requires what is given in outer sense. On my material dependence reading, inner sense must be conceived as a cognitive capacity that is blank and passive with respect to its content which depends on what is given in passive outer sense for the realization of empirical consciousness. In his critique of causal readings, Chignell, too, neglects this crucial point, although Kant, in a supplementary footnote to the preface to the second edition of the Critique of Pure Reason, highlights by way of explicitly specifying the argument of the Refutation that it is outer sense and its objects "from which we after all get the whole matter for our cognitions, even for our inner sense" (CPR B xxxix, n.). Therefore, the cognitive function of the persistent necessary for time-determination cannot be fully elucidated independently of the theory of inner sense that justifies the Refutation's main claim that we cannot be aware of our empirically determined existence in time independent of objects in space.
Kant's theory of inner and outer sense is the key to the Refutation of Idealism. The theory considered in its entirety is quite complex, because it concerns not only inner and outer sense as such but also the farther-reaching doctrine of (transcendental) self-consciousness. Here I only consider those elements of the theory which are relevant for the argument of the Refutation. Kant modifies the relation between inner and outer sense throughout his critical philosophy. Whereas he, in the first edition of the Critique of Pure Reason, identifies inner and outer sense as quasi-equivalent faculties of representation, in the second edition he determines their relationship anew. After introducing outer sense broadly in the Transcendental Aesthetic as a "property of our mind", through which "we represent to ourselves objects as outside us, and all as in space", and characterizing inner sense as the faculty "by means of which the mind intuits itself, or its inner state" (CPR A 23/B 37), he further clarifies the relation and difference between inner and outer intuition in the "Critique of the Fourth Paralogism": "the representation of myself, as the thinking subject, is related merely to inner sense, but the representations that designate extended beings are also related to outer sense." (CPR A 371, cf. A 385) A specific articulation of the relationship of inner and outer sense is absent in the first edition of the Critique, but there Kant already understands inner and outer sense as distinct ways of representing what belongs to inner and outer intuition: whereas through inner sense we have representations of our "inner states" (CPR A 23), i.e., we are for example conscious of our "desires", "thoughts, feelings, inclinations or decisions" (CPR A 357f.) and perceive ourselves in this way internally, which is identical with empirical apperception (CPR A 107), in outer sense we intuit objects in space on the basis of outer affection. With respect to these determinations of inner and outer sense, nothing essentially changes in the second edition of the Critique. However, a modification, or rather specification, of the conception of their relationship is introduced in 1787.
Kant now argues explicitly for the material dependence of inner sense on outer sense. He shows that passive inner sense is dependent, in material terms, on passive outer sense. The connection between inner and outer sense has always been a deeply revealing but controversial subject in Kant scholarship. One can sum up the debate by drawing on the influential older study by Reininger: Kant's Doctrine of Inner Sense and Theory of Experience (Kants Lehre vom inneren Sinn und seine Theorie der Erfahrung). 20 Reininger distinguishes between two fundamental views concerning the relationship between inner and outer sense: (1) The coordination (or parallelism) of inner and outer sense, and (2) the subordination of outer sense to inner sense.-Against both theses outlined by Reininger, I will argue for material dependence of inner sense on outer sense.
According to the coordination thesis, Kant conceives of inner and outer sense primarily as relatively equivalent sources of cognition. For Reininger, both are essentially similar insofar as they are receptive faculties, but they are distinguished through their form, that is, time as the form of inner sense and space as the form of outer sense, as well as by their content, namely, inner objects for the former, and representations of material things for the latter. Because it is inconceivable how, through this parallel ordering and coordination of inner and outer sense, external objects can also be in time, Kant was forced, according to Reininger, to abandon the coordination-thesis in favor of the subordination-thesis, the view that inner sense is the all-encompassing sensible faculty with outer sense included in it as a partial sphere. 21 20 Reininger (1900). 21 Cf. Reininger (1900, p. 13, 56 f). On the shift from the coordination to the subordination thesis, which Reininger, in terms of the history of its evolution in Kant, pursues only sparsely, see, ibid., pp. 9-62. In this part of his study, Reininger also considers the historical influences on Kant's theory of sensibility; Kant, he claims, probably oriented himself along the lines of Locke's distinction between inner and outer sense. Reininger points out that, in the Anthropology, Kant differentiates between the sensus internus (inner sense) as a faculty of By contrast, on the material dependence-view, the interconnectedness of inner and outer sense is presented ultimately as a relationship of dependence of the inner sense on the outer sense, insofar as the former cannot produce any content without the content provided by the latter. As the theory of self-affection shows, inner sense is reliant on the representations given in outer sense, which the transcendental apperception posits in time. Through this process of temporally positing representations in inner sense, empirical self-consciousness is brought about, i.e., the kind of inner experience the Cartesian sceptic admits. In this way the faculty of cognition of the pure "I think", as well as those of inner and outer sense, are brought into relation with one another within the subject, where inner sense is affected through the determination in time of the representations given in outer sense, and empirical self-consciousness is thereby made possible. Further elucidation of this theory will show that Kant holds neither that inner and outer sense are merely coordinated, nor that outer sense is subordinated to inner sense.
Kant sketches the cognitive function of inner and outer sense and their dependence relation, insofar as it is relevant for his anti-sceptical strategy, in connection with the theory of self-affection of the transcendental deduction of the categories (CPR B). According to the theory of self-affection, empirical self-consciousness is constituted through the performance of synthesis, which the pure "I think" carries out in inner sense by affecting it under the name of the transcendental power of the imagination (synthesis speciosa). Two principal forms of self-affection must be distinguished, both of which clarify the possibility of empirical consciousness, i.e., the empirically determined consciousness of my existence in time discussed in the Refutation: first, self-affection on the basis of the synthesis of a pure, given manifold, and second, on the basis of the synthesis of an empirical manifold. In both cases, the pure apperception determines and triggers the empirical apperception in inner sense. In the transcendental deduction of the categories (B) Kant explains self-affection by means of the synthesis of a pure manifold: the understanding can "determine the inner sense through the manifold of given representations in accord with the synthetic unity of apperception, and thus think a priori the synthetic unity of apperception of the manifold of sensible intuition" (CPR B 150). Since inner sense, as the Transcendental Aesthetic reveals, contains a pure manifold of time (the pure flow of time), the understanding is able, by means of such synthetic determinations of inner sense, to think pure determinations of time such as the pure temporal modi persistence, succession and simultaneity. In the Transcendental Deduction, Kant clarifies how the empirical apperception is constituted with a description of our manner of representing time "under the image of a line" (CPR B 156): Since time is the form of inner sense, we cannot sensibly perceive time itself but only represent it through the drawing of a one-dimensional line. I therefore direct myself to myself while constructing (drawing) the "time-line" by successively positing the represented "parts" of time and, in this way, I affect myself in accordance with the order of time and determine myself as an empirical object of inner perception. Kant conceives of this process of self-affection as an "act of attention" (B 156 f. n.) of the understanding which combines the manifold in inner sense. In the essay What real progress has metaphysics made in Germany since the time of Leibniz and Wolff?, he explains self-affection slightly differently than in the Critique: self-affection is described there as brought about through the synthetic determination of not a pure manifold, but rather an empirical one. Anyone can detect and identify the phenomenon of self-affection in inner sense by means of "psychological observation": "for this we are required to affect the inner sense, in part also doubtless to the point of fatigue, by means of attention (for thoughts, as factual determinations of the power of representation, also belong to the empirical representation of our state)" (Kant 2002, p. 362;20:270). The result of this affection "mediated by the attention" on one's mental states is then the same as in the Critique, i.e., the consciousness of the "empirical I" or the I-object that "I am", which is determined in time by the intellectual, logical I or the I-subject in the apprehension of perceptions. Thus, for Kant empirical self-consciousness is the consciousness that arises through self-affection in inner sense via the inward observation of the synthetic determination of representations in accordance with the order of time. What Kant means is, for example, that when reading a philosophy paper that seems hard to follow, I can force myself to concentrate on the different steps of the argument being proposed when halfway through the text by recalling what has been said in section one, trying to combine it with what has been said in section two, and so forth, in order to discover whether I have understood the author's line of thought. When I recall the logical order of the paper, I attend to myself. In this way I determine my existence in time and become empirically aware of myself in terms of inner experience. Again, what self-affection in inner sense depends on is the content that is given in outer sense. Note that the dependence relation of inner sense on outer sense is therefore a doctrine that Kant does not create in the Refutation itself, but rather Footnote 21 (continued) perception and the sensus interior (internal sense), to which belong feelings like desire and aversion (ibid., p. 34ff). one which he takes up there from the second edition Transcendental Deduction of the Critique of Pure Reason. 22 The Cartesian sceptic (problematic idealist) does not doubt self-consciousness as expressed through the proposition "I am". The sceptic might even grant that empirical self-consciousness involves the cognitive procedure of time-determination in inner sense through self-affection that requires a persisting object as the permanent for the awareness of change in inner sense. But why should the sceptic concede that the persisting for temporal change of representations in inner sense can only be objects in space as perceived through outer sense? There are two replies to this question: first, inner sense cannot supply persistent objects because representations in inner sense are in constant flux (temporal succession) in accordance with the form of time, which does not allow for persistence. If the persistent is indispensable for the determination of my existence in time, and if inner sense cannot provide it, it must originate in outer sense, which allows for persisting objects because its form is space. As we have seen, the sceptic would remain unconvinced and insist that the persistent thing could still be a product of the imagination. It is for this reason that a second reply is necessary: Persisting objects cannot be mere products of the imagination because of the material dependence of inner sense on outer sense. The crucial point for the relationship between inner and outer sense and for the antisceptical strategy of the Refutation of Idealism as a whole is that the inner sense is reliant on the representations of outer sense. Therefore, inner sense, as an originally passive, blank capacity of our sensibility does not generate representations by itself in any 'productive' manner. As Kant notes on many occasions, representations and the "alterable" (CPR B 156) of outer sense are its "proper material" (CPR B 67). From the objects of outer sense, we get "the whole matter for our cognitions, even for our inner sense" (CPR B xxxix, n.). Perceptions in outer sense are the only representations that can serve as content of our inner experience, and therefore make the temporal determination of our consciousness in inner sense possible. 23 Inner sense is, therefore, materially dependent on outer sense insofar as the content of its representations are ultimately traced back to what is given in outer sense. However, outer sense is equally reliant on inner sense, because outer representations are nothing for me if I am not conscious of them in inner sense via the self-affection through which I become empirically conscious of myself. The relation of inner to outer sense must therefore be portrayed not as a mere coordination or subordination but as a material dependence. This notwithstanding, the inner sense should not be understood as an organ of inner perception. 24 The concept of inner sense has no physiological meaning in transcendental philosophy but rather designates the consciousness of my existence in time. I generate or have this empirical consciousness of myself when I spontaneously determine representations that I am conscious of in time. Representations arise and disappear in my consciousness after shorter or longer amounts of time and in this way signify an alteration of my inner states (of consciousness). Therefore, I am at the same time able to focus on specific representations, to direct myself to them. Kant attributes this sort of concentration on a representation as empirical self-consciousness in inner sense. However, inner and outer sense do not make possible any modes of consciousness which would stand independently of each other, enabling me either to perceive myself in inner sense independently from outer sense or in parallel with it, or to perceive the objects of outer sense independently from inner sense. According to Kant, "for an experience in general to be possible, the reality of outer sense is necessarily bound up with that of inner sense" (CPR BXLI, n.). In this way, the concept of outer sense stands for the grounding function of sensibility, on the basis of which I can be affected by outer, material or physical conditions. By contrast, the concept of inner sense designates the particular consciousness or cognition of outer representations, on the basis of which I become conscious of my existence temporally-empirically.
The material dependence of inner sense on outer sense should not be seen as a stand-alone view that serves exclusively to further the Refutation's anti-sceptical strategy. The argument has much broader implications for the interpretation of Kant and beyond, since it anticipates what in more recent times has become known as mental content externalism. Kant's case for the material dependence of inner sense on outer sense can be read as a case for mental content externalism. In more general terms, mental content externalism is the view that the content of mental representations is necessarily determined by non-mental objects that stand in causal relation to the appropriate sensory capacities of the cognizer who entertains these representations. That is, the impact of external objects on a cognizer's mental states is a necessary condition for representing mental content and its 22 In the later Reflections on Idealism Kant points out again that the material dependence of inner sense on outer sense is grounded in my attending to myself through the synthesis of representations in intuition and thereby determining my existence in time. Without the material of outer sense, self-affection and empirical self-consciousness would be impossible. Cf., e.g., AA 18:306-310 and 610 f. See Heidemann (2021). 23 Düsing (1980). 24 Cf. Reininger (1900, pp. 175-232). changes. 25 The material dependence of inner sense on outer sense is the Kantian ancestor of mental content externalism. Kant's theory of mental representation, although not employing contemporary terminology, argues along the lines of this theory. Kant thinks that inner and outer sense are unable to generate mental content on their own independently of causal stimulus on our sensory apparatus, for they are passive, blank cognitive capacities. Once outer sense receives and is filled in with outer sensory material, inner sense receives it as mental content such that mental representation and perception of the world becomes possible.
Although the transcendental conditions of the possibility of cognition form the set of truth and/or accuracy conditions for worldly beliefs that cognizers can hold, it is the content of outer sense as brought to consciousness, where it is cognitively transformed into mental representations in inner sense, that is at the very foundation of Kant's mental content externalism. Already in the first edition of the Critique of Pure Reason Kant states that "[t]his material or real entity, however, this Something that is to be intuited in space, necessarily presupposes perception, and it cannot be invented by any power of imagination or produced independently of perception, which indicates the reality of something in space.
[…] Whether we take sensations, pleasure and pain, or even external sensations, such as colours, warmth, etc., it is certain beyond doubt that it is perception through which the material must first be given for thinking objects of sensible intuition. This perception thus represents (staying for now only with outer intuitions) something real in space." (CPR A 273-274).
From this Kant concludes that "[e]very outer perception therefore immediately proves something real in space" (CPR A 274) for "in this space the real, or the material of all objects of outer intuition is nevertheless really given, independently of all invention" (CPR A 275). Kant insists that we cannot go any further than this, even if the sceptic asks us to show that mental representations really represent the external objects they are supposed to: "the strictest idealist cannot demand that one prove that the object outside us (in the strict sense) corresponds to our perception." (CPR A 275; cf. B xli n., 278 f) Kant's modest claim is that the material dependence of inner sense on outer sense makes such a proof superfluous. This claim is not question begging. It is a logical consequence of mental content externalism. For the question his argument poses is what the condition under which the sceptic could continue to insist that the correspondence between mental representations and the external world is doubtful really is. This gets us back to where we started, i.e., the sceptical hypothesis.
Recall that the force of the sceptical hypothesis depends on the cognition-transcendence of the cause of our representations. To deny the fact of such transcendence, (Kantian) mental content externalism must allow for a form of realism that integrates both the ontological independence and epistemic dependence of external objects. This combination is made possible through empirical realism (CPR A 371). Whereas transcendental realism (CPR A 371), like empirical realism, puts forth their ontological independence, it also argues for the epistemic independence of external objects; that is, in transcendental or metaphysical realism, reality can go beyond what we can cognize, because on this theory there are no transcendental conditions of the possibility of cognition. In empirical realism, however, this is not the case, since the real cannot reach beyond the transcendental conditions of the possibility of cognition as established by transcendental idealism. For this reason, as Kant argues, there is no place for the sceptical hypothesis in empirical realism and the theory that grounds it, transcendental idealism. For, in empirical realism, external objects are accessible to cognition as the real external causes of our perceptual beliefs about the world and cannot go beyond what is possible for cognizers to know. Therefore, the material dependence of inner sense on outer sense, the key argument of Kant's antisceptical strategy in the Refutation of Idealism, is not only supported by the argument that empirical consciousness of my existence requires determination in time, which in turn presupposes a persistent real object in space that neither passive inner nor passive outer sense is able to produce by itself. Its ultimate foundation is transcendental idealism and empirical realism.

Material Dependence, Discursivity, and the Finitude of Cognition
The paper started off with a defense of Kant's distinction between appearance and thing-in-itself against sceptical attacks. As it turned out, to conceive of this distinction in terms of the sceptical hypothesis does not do justice to the Kantian discussion of idealism and realism in the Fourth Paralogism (A). For in the Critique of Pure Reason (A), Kant demonstrates that transcendental idealism provides arguments against sceptical doubts about the existence of the external world by replacing the model of indirect causal perception with the model of direct causal perception of external objects. The idea is that direct perception of external objects rules out scepticism about external reality. The success of this anti-sceptical strategy of the Fourth Paralogism is, however, questionable. The anti-sceptical strategy of the Refutation of Idealism seems to be more promising. Here Kant argues for the material dependence of inner sense on outer sense. Since the sceptic cannot deny her empirical consciousness of her existence in time, she is therefore forced to accept, as the condition of the possibility of the awareness "I am", the existence of the external world. Furthermore, this thesis forms a constitutive part of transcendental idealism and empirical realism, which, taken together, demonstrate how the sceptical hypothesis can be ruled out on the basis of the ontological independence and epistemic dependence of external objects. Kant, however, does not claim that this strategy provides the ultimate refutation. Kant is more cautious, as we saw, because he does not deny that ultimately, we may still be misled about our cognitive claims despite the argument he puts forth against the sceptical hypothesis. His anti-sceptical strategy should therefore best be understood as a wellgrounded, plausible argument that takes the sceptic seriously and insists that while there may be good reasons for sceptical doubts, there are better reasons against them.
On the other hand, the argument from the material dependence of inner sense on outer sense might be stronger than even Kant saw. Sure, even if the sceptic concedes everything Kant puts forth in favour of material dependence, she can still claim in the end that inner sense is only a subjective presupposition, a cognitive capacity that we cannot rely on given that we have no reason to believe that human cognition could not function without it. At first glance this appears to be a powerful challenge, since Kant does not explicitly work out a justification of the reality of inner and outer sense as cognitive capacities, other than in a few sparse remarks in the Transcendental Aesthetic. Would he then ultimately be forced to give up on the project of the Refutation and its anti-sceptical strategy of material dependence? I don't think so. For the Refutation is systematically linked to the overall Kantian project which is built up from his fundamental insight of the finitude of human cognition. The fact of the finitude of human cognition does not follow simply from the transcendental conditions of cognition, which set limits on what we can possibly know, because finitude is not defined by these conditions alone. For that human cognition is finite essentially means that it cannot produce its object by itself; it is necessarily dependent on factors that are beyond its control. This is the case with Kant's dual-stem theory of cognition, what we might term his "cognitive dualism". Kant's cognitive dualism is the view that human cognition is discursive, meaning that it depends on two sources of cognition, sensibility and understanding, each of which produces a necessary component of cognition: intuition and concept. It is only through their cooperation that cognition is possible. Yet Kant does not merely presuppose that their cooperation is necessary for cognition; rather, it follows from the fact that the only concepts of which human beings are capable are discursive concepts. For Kant, a concept is a repraesentatio generalis, that is, a general, or universal, representation that is essentially abstract. As a matter of fact, the only concepts the human mind can operate with are discursive, abstract concepts. An abstract concept is an analytic representation, which, as Kant says, must be conceived "as a representation that is contained in an infinite set of different possible representations (as their common mark), which thus contains these under itself; but no concept, as such, can be thought as if it contained an infinite set of representations within itself." (CPR B 40). For instance, the concept 'tree' is the general, and thus abstract, representation that "contains" under it those marks (Merkmale) that are characteristic of particular trees such as oaks, beeches, or pines. Since all (empirical) concepts are abstracted from perception, all concepts depend on what is given to the senses. This means that it is originally impossible to form concepts independent of perception, and consequently that the content of human concepts, as abstract representations, cannot be determined merely by the understanding. The possibility of discursive, abstract concepts, without which there could be no cognition, requires the sensible given delivered in intuition, the repraesentatio singularis, which are themselves to be subsumed under concepts. The concept 'tree' contains under it all abstracted marks that are representative for trees in general but not any singular representation of an oak, beech, or pine. To conceptually represent the oak in my garden, for example, my intuition of it, a singular representation of a sensible given, must be subsumed under the concept 'tree'. The discursive general representation, i.e., the concept 'tree' cannot provide that content by itself.
Discursivity is at the heart of Kant's conception of the finitude of human cognition, and as such is relevant for his position of the material dependence of inner sense on outer sense. For to temporally determine my existence in inner sense, and therefore for my representation of myself under the concept "I am", the sensible given of outer sense is required. As Kant writes, outer sense and its sensible given would not be a necessary condition for my determination of my existence in time if human cognition were not discursive but intellectual: "If I could combine a determination of my existence through intellectual intuition simultaneously with the intellectual consciousness of my existence, in the representation I am, which accompanies all my judgments and actions of my understanding, then no consciousness of a relation to something outside me would necessarily belong to this." (CPR xl, n.) Human beings are not capable of intellectual intuition; rather, our intuition is sensible. As discursive cognizers we therefore depend on what is given to us in sensibility. A non-discursive cognizer like a divine mind would, as Kant speculates in the Critique of the Power of Judgment ( § § 76, 77), not need to rely on what is given from outside in order to determine its existence but rather could cognize it spontaneously and independently. For the sceptic to overcome the argument of the Refutation, she would have to claim to overcome her own cognitive finitude. But with such a claim, the burden of proof would surely lie with the sceptic. 26 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.