Attempting to Exit the Human Perspective: A Priori Experimentation in Kant’s Critique of Pure Reason

  • Rachel ZuckertEmail author
Open Access
Part of the Synthese Library book series (SYLI, volume 416)


I consider a problem for Kant’s transcendental idealism if one construes it as a claim that human beings know from a particular, specifically human perspective. Namely, ordinarily when we speak of someone seeing from a perspective, we understand other people to have other perspectives, and think that people can change their perspectives by moving away from them, to a different one. So one may recognize that one’s own perspective is a perspective by comparing to others, by seeing a former perspective from a new vantage point. But Kant denies such plurality and variability for the perspective he identifies; it is the human perspective as such. Thus, one may worry that Kant’s view is incoherent: Kant claims that we can know only from one perspective, yet, in order to recognize that perspective, he himself must stand “outside” of it. I consider a potential Kantian response to this charge, in the form of an interpretation of the Dialectic section of the first Critique. When one attempts to know things that lie beyond the human perspective — to exit it — one falls into contradictions and empty thinking. These failed attempts to exit the human perspective constitute its horizon, a limit recognizable without one needing truly (but impossibly) to occupy a different perspective. Such failed attempts, I argue, are some of the confirming results of the a priori experimentation Kant proposes in the Preface to the Critique: his hypothesis of transcendental idealism is shown to identify the dividing line between successful and failed, productive and contradictory attempts at human knowledge.


A priori knowledge Transcendental idealism Kant’s theoretical philosophy Kant’s critique of rationalist metaphysics Philosophical methodology Perspective 

1.1 Introduction

As thematized in this volume, philosophers often speak of perspectives, meant not (usually) in the literal senses of painting technique or of viewer’s perception of a segment of a visual field from a particular spatial position, but in the also familiar, though more metaphorical sense of having a “point of view”, thinking of things in one’s own way, framed by one’s particular modes of attentiveness, organizing principles or interests, and so forth. In his Critique of Pure Reason, Kant uses the closely related term “standpoint” to describe the human epistemic condition. Summing up his argument that space is an a priori form of intuition, a framework for human sensibility, and so formative of objects as they appear to such sensibility, he writes:

We can accordingly speak of space, extended beings, and so on only from the human standpoint [Standpunkt]. If we depart from the subjective condition under which alone we can acquire outer intuition, namely that through which we may be affected by objects, then the representation of space signifies nothing at all.1

Here Kant comes quite close to saying that human knowledge – at least knowledge of sensibly presented objects – is inevitably from, formed by, a human perspective. More generally, one could use this metaphor to gloss Kant’s central philosophical doctrine, transcendental idealism: that human “cognition reaches appearances [i.e., objects as they appear to us, in sensible experience] only, leaving the thing in itself as something actual for itself but uncognized by us” (Bxx). One might say, then, that on Kant’s view human beings can know, even with necessity, how things will be from our perspective – i.e., the spatio-temporal realm of appearances – but only within and for our own perspective.

My paper concerns a question brought out by this way of characterizing Kant’s position, specifically by a disanalogy between it and our ordinary way of thinking of perspectives. Ordinarily we take perspectives to be plural and variable: different people have different perspectives; individuals can also change their standpoints and thereby look (from “outside”) at their own previous perspective, recognizing it as such.2 But, at least with respect to the human standpoint centrally at issue in his transcendental idealism, Kant denies this.3 There are and can be no plural perspectives among human beings, nor can any individual move away from the standpoint she occupies. For the human standpoint as such is both universal, shared by all human beings, and necessary for each individual human knower.4 Thus one may ask: what does it mean to identify one’s mode of knowledge as a particular perspective if one cannot exit it, if one does not encounter others who occupy different ones? How could one recognize one’s own standpoint as such without such plurality or possibility to exit?

These questions do not, I think, result merely from pressing a metaphorical expression. Rather, they reformulate a central question concerning the Kantian philosophical enterprise, raised in various forms from early on in Kant’s reception. Hegel and other post-Kantians accuse Kant of self-contradiction for similar reasons: Kant claims that human beings can know only from one perspective, yet it would seem that in order to recognize that perspective as such, he does and must stand outside of it.

In this paper, I investigate one Kantian response to such concerns: in brief, the Dialectic section of the Critique. In the Dialectic, Kant portrays rationalist metaphysics as a failed attempt to know things beyond experience, and I shall suggest that these failed metaphysical views can be understood as attempts to exit the human perspective; one discovers, in the course of the Dialectic, that these attempts lead to cognitive failure. Thus, on Kant’s view, we do not have to inhabit another perspective or – impossibly – stand outside our own perspective to establish its limits. The limits of our perspective can be established from within, from the epistemic problems that arise when we attempt to transcend it. I begin with a slightly more extensive discussion of the problem, before turning to propose this view of the Dialectic.

1.2 The Problem, in Some More Detail

As just noted, when we speak of someone occupying a particular standpoint or having a particular perspective in everyday life, we take such perspectives to be plural and variable: we understand other people to have other perspectives, from their own, different spatial locations, and that one can change one’s perspective, moving away to another (location or, by metaphorical extension, attitude or theoretical position). One may recognize that one has a specific perspective then, either by recognizing that others see things differently or by varying one’s own position, looking back at one’s original perspective, and so seeing its location and limitation. It seems prima facie unclear, then, what a “perspective” (in this everyday sense) would mean in cases of universal, unaltered, perhaps unalterable agreement: are there “perspectives” on well-established mathematical theorems or basic facts such as ‘human beings need to eat to survive’? How would one establish that this agreement is actually a universally shared perspective? By comparison to what?

These questions are, of course, the ones I wish to press about Kant’s claim to identify a universally shared, necessary human perspective on objects of experience, Kant’s famous philosophical “Copernican Revolution”. According to this transcendental idealist “altered method of our way of thinking”, Kant writes, he can explain the possibility of a priori knowledge: “we can cognize of things a priori only what we ourselves have put into them” (Bxviii). More precisely, we can know objects a priori insofar as they “conform to the constitution of our faculty of intuition”, and therefore “the experience in which alone they can be cognized … conforms to [certain a priori] concepts” (Bxvii). Kant elaborates this approach in the Transcendental Aesthetic and Analytic sections of the Critique: we can know a priori about objects of experience, concerning both their spatio-temporal character – their conformity to the a priori conditions of human sensibility – and their conformity to the categories, the a priori concepts of the understanding (such as unity, negation, cause and substance). Thus, Kant claims, we can know that “nature (in the empirical sense) [is] … the combination of appearances … in accordance with law … indeed in accord with its original laws, in accordance with which experience itself first becomes possible” (A216/B263). Correspondingly, Kant argues, these claims – what we can know a priori – are correct from the human standpoint alone. They concern objects as appearing to us, given the nature of human cognition, and otherwise – as he strongly puts it concerning space (at A26/B42, quoted above) – are “nothing”.

Henceforward, I will refer to this complex of doctrines as “the human perspective” on Kant’s view. Namely, (1) human beings must employ together the cognitive faculties of understanding and sensibility – must use both concepts and sensible intuitions – to know objects; (2) each of these human cognitive faculties furnishes and necessarily relies upon a priori cognitive representations (the categories and space/time, respectively); (3) objects known by such human beings – the objects that (as it were) come into view for such a perspective – are correspondingly and therefore law-governed, spatio-temporal objects of experience, or “appearances”. These three aspects of the human perspective may be glossed in turn as describing (1) the subject who knows, (3) the object known thereby, and (2) the constitutive structure of (1) that is carried over to or in some way determinative of (3). Perhaps one could therefore call (2) the structure of the perspective (i.e., it describes how the subject is oriented to objects, and correspondingly how objects will appear to or come into view for that subject). And again, the question I aim to raise concerning this complex of doctrines is: given the universal character of this “standpoint”, the necessity for every human knower that she uses these cognitive capacities to know objects, in what sense is this a perspective, and how can it be identified as one?

I note first that Kant is well aware of the philosophical-methodological utility of invoking plurality or variability of perspectives, to become aware of one’s own perspective as such. In the Paralogisms chapter of the first Critique, Kant explicitly invokes the conception of an observer at a different “standpoint” in order to make clear that the constant (potential) presence of self-consciousness in one’s experience, the persisting identity of one’s representation of one’s “I”, does not prove that one’s self is a persistent substantial entity.5 Kant here proposes an alternative perspective – or stages a “thought-experiment” of inhabiting another perspective – in order to make one aware that certain facts (the persistence of one’s “I”) are features of one’s own perspective, and do not necessarily hold of things independently of that perspective. Earlier in the Paralogisms, Kant also brings out the difficulty with this methodology if deployed to try to identify necessary, shared features of a human perspective as such: “It is obvious”, he writes, “that if one wants to represent a thinking being, one must put oneself in its place, and thus substitute one’s own subject for the object one wants to consider (which is not the case in any other species of investigation)”.6 If there are necessary features of the human mode of knowing as such, then any attempt by a human subject to inhabit a different perspective from that one, to look at it from outside, will bring those necessary features along with it. One will “substitute” oneself, place oneself, along with all the necessary features of human knowing, within that alternative perspective.

Thus, I suggest, the ordinary way of making sense of, and recognizing, a “perspective” or “standpoint” appears to be ruled out for the purported Kantian human standpoint – and Kant acknowledges as much. But he does, I think, propose (at least) two ways of identifying the human perspective in the Transcendental Aesthetic and Analytic.

First, Kant extends the notion of plurality of perspectives to refer to other, non-human beings. He suggests that the necessary, universally shared human perspective may be distinguished from, and recognized by comparison to, that of possible other finite intellects that might have other forms of intuition (other than space and time),7 or, more prominently, a divine mode of knowing: intuitive intellect or intellectual intuition.8 These distinctions seem to me pedagogically helpful (as it were) for bringing readers to understand the kind of position Kant is proposing, to see that he wishes to argue that human beings have particular modes of access to the world, or (in Henry Allison’s terms) “epistemic conditions” governing our approach to objects.9

The gesture to other potential modes of knowing makes the general shape of this proposal clear. But these comparisons do seem to me to be mere gestures. By themselves, they are not decisive for establishing that space, time, and the categories – the lawfulness of nature – are factual merely for, and derived from, the human perspective. For, in Kantian terms, it is unclear whether these purported alternative perspectives are “real possibilities”: though we can (somewhat) imagine such entities and their possible alternative modes of knowing, we do not know that there really could be such alternative modes of knowing. Or, in the everyday terms I sketched above: our inability really to conceive of such intellects, of what their modes of knowing would be like, of how things could be from another, entirely different sort of perspective, would follow again from the inescapability of the universal, human perspective. If we think in a serious way about thinking (knowing) beings, we will “substitute our own subjects” for such beings. Why, in short, should we take it that there really could be such other “perspectives”?

Of course one can always consider whether there might not be some radically other way of looking at the world, whether one’s beliefs, modes of thinking and so forth might be mistaken. So too can one wonder whether there might be radically other sorts of beings, with radically other modes of knowing. But unless something further is said, these considerations are somewhat idle – perhaps a salutary reminder to be epistemically humble, but not providing significant ground for restricting one’s claims, demoting them (as it were) to a mere perspective.

Second, Kant provides more substantive, direct arguments that space and time are and must be conceived as forms of human intuition in the Transcendental Aesthetic. Likewise (if more complicatedly), he argues in the Transcendental Deduction that objects of experience are constituted as such by the categories – that, in other words, such objects are substances with attributes, subject to causally governed changes, and so forth – on the grounds that otherwise human experience would be impossible. Thus, Kant concludes, objects of experience may legitimately be so characterized only from a human perspective, only under the restriction that we are discussing objects as presented to human knowers.

These lines of argument are too complex for me to discuss in any detail. I will therefore just sketch a worry – loosely inspired by (a perhaps odd combination of) Strawson and Heidegger10 – about taking them as decisively identifying a human perspective as such, i.e., as entailing that the cognitive claims established are local to, or even just descriptive (solely) of some particular sort of knower and objects as known thereby. That is, these Kantian lines of argument might be consistent with the conclusion that this sort of object – natural, sensibly intuitable, temporal objects – must be so, for any knower who would be acquainted with them.11 Therefore, we need not say that these objects only appear so, or that they are so only for a specific sort of knower. Rather, any knower able to know about such objects will need to be furnished with the appropriate cognitive capacities – here a priori forms of intuition (space and time), and a priori concepts such as substance or cause – in order to recognize those facts about those objects.12 Kant’s restriction of such claims to a human standpoint might then seem like an over-interpretation, an unjustified specification of the results he has in fact achieved (from analysis of the character of objects of experience). Certain concepts and principles may apply to a restricted range of objects, or to objects only under certain descriptions – objects of sensible nature, say – but it may not be clear exactly why one needs to add that such objects are such only from or particularly for the human perspective.13

There are many possible responses to this sort of worry. Kant himself seems inclined to (what we now call) an argument to the best explanation. That is, if one grants that human beings are furnished with a priori representations (that is, the second item on the list of three points defining the human perspective above, what I called the “structure” of the perspective) and that objects must conform to such a priori forms in order to “appear” to us – so Kant argues – it would be a miraculous coincidence that those objects also just happen to be so, independently of our cognitive demands, our modes of apprehending them, their modes of appearing to us. The more plausible, efficient explanation is that they are so in virtue of appearing to us (see Kant 1772). The recent renewed interest in metaphysical readings of transcendental idealism, particularly of Kant’s doctrines concerning sensibility, could also be seen as responses of the following sort: Kant’s arguments concerning space and time establish that all spatio-temporal objects simply cannot be things in themselves, given the (dependent, relational, or otherwise metaphysically questionable) nature of space and time. Space and time must be understood to exist only in their relations to us; knowledge claims concerning them (or objects “in” them) hold only to or from a human standpoint (see, notably, Langton 1998 and Allais 2015). Here I explore a different possible response, namely the Dialectic of the Critique of Pure Reason.

1.3 The Transcendental Dialectic: An Alternative Perspective

The Transcendental Dialectic comprises, roughly, the second half of the Critique of Pure Reason.14 Here Kant is concerned with what he takes to be the central subject matters of traditional “special” metaphysics: soul, God, and world-whole. He aims to explain how human beings come to formulate conceptions of these “transcendent” entities – i.e., things that transcend human sensible experience – and to show definitively that philosophers’ claims to know a priori about such entities are mistaken. In particular, he devotes the Paralogisms chapter (quoted above) to a priori arguments that purport to elucidate the nature of the soul, as simple, immaterial substance. Kant objects that all such arguments are either tautological – based upon and merely re-describing formal features of (the representation of) self-consciousness, but no entity that purportedly underlies it – or invalid (insofar as they do claim to establish truths about some such entity, they illicitly import contents from sensible experience). The Antinomies chapter concerns a priori responses to questions concerning the world-whole: most famously, is it thoroughly deterministic, or is there a place for freedom? But also, is there a beginning of time, an end of space? Is there a smallest, most basic component of material objects, or are they infinitely divisible? Is there a necessary being grounding all the contingent elements of the world, or is it turtles all the way down? Kant argues that there are, always and systematically, two conflicting answers to such questions, both of which are supported by a priori rational argument. In pursuing such questions, therefore, we fall into a “contradiction of reason” (A408/B435). Finally, Kant devotes the last chapter of the Dialectic, entitled “The Ideal”, to explaining the origin of the idea of God and to arguing that a priori proofs of God’s existence fail, most famously because existence is “not a real predicate” (A598/B626).

These arguments contribute, in obvious and direct ways, to Kant’s philosophical enterprise in the Critique. Kant announces that this work concerns the possibility of metaphysics (Axii, Bxv); it therefore appropriately includes consideration of rationalist special metaphysics. This discussion belongs as well to Kant’s project of “critique”, i.e., of reason’s “self-knowledge” (Axi-xii). For Kant argues that reason formulates the ideas of soul, God, and world-whole as a result of its “drive” to attain complete, satisfactory, systematic explanation, or (in his terms) to find the “unconditioned”, the ultimate totality of conditions from which all (experiential or other) truths could be derived, on which all experiential things could be grounded.15 And of course insofar as Kant shows that we cannot know about soul, God, or world-whole by pure reason, independently of experience, he supports his transcendental idealist conclusion concerning human knowledge as such, i.e., that it is restricted to objects of experience.16

But in the B Preface, Kant claims that the results of the Dialectic not only support his restriction of human knowledge to objects of experience, but also serve as a “splendid touchstone of” his characterization of the a priori knowledge that we do have (on his view) as characteristic of the human perspective – as comprising “what we ourselves have put into” objects of experience (Bxviii). For, Kant proposes, one might see the Critique as an “experiment”.17 The hypothesis tested is his own transcendental idealism, namely (again) that human a priori knowledge can be explained if we take objects of experience to “conform to the constitution of our faculty of intuition” so that “the experience in which alone they can be cognized … conforms to [certain a priori] concepts” (Bxvii). By contrast, Kant writes, the arguments treated in the Dialectic concern “objects” not considered as appearances, but “insofar as they are thought merely through reason”.18 In showing that those arguments fall into “contradiction” (or otherwise fail), Kant claims, he also confirms his hypothesis: “The experiment decides for the correctness of [the] distinction” between appearances and things in themselves, thus for his construal of objects of experience as appearances, i.e., as objects for or within the human perspective alone (Bxix).

It is this proposed, indirect function of the Dialectic – to confirm the conclusions in the Analytic – that I shall explore here.19 For one may well wonder, why and how does not being able to know about immortal souls, God, or the world-whole show that empirical-causal explanations or geometrical proofs and the like are knowledge from or for the human perspective alone? I will propose that the Dialectic – the “experimental” results Kant attains there – may be seen as an answer, of a kind, to the problem I sketched above. That is, I shall suggest that in the Dialectic, Kant portrays human knowers as, in fact, exiting the human perspective, sort of: rationalist metaphysicians attempt, but ultimately fail, to exit the human perspective.20 Thus the Dialectic portrays an alternative perspective from which one can regard one’s original perspective and recognize it as such – though it also ultimately affirms that the human perspective is the only successful one for human knowers. To be clear, I propose that Kant’s view is not that any other perspective is incoherent, ultimately inconceivable; the alternative perspective is conceivable, even “inhabitable” by the human subject (at least the philosophically minded one). But our attempts at knowledge from within it turn out to fail. I will now discuss the two sides of this proposal in turn – the way in which the portrayed metaphysicians exit the human perspective, and thereby recognize it as such, and then the qualifications on that alternative perspective (why I characterize it above as a mere “attempt”, or a “sort of” alternative).

1.4 An Alternative Perspective

As just quoted, Kant describes rationalist special metaphysics as treating objects not as appearances, but as “thought merely through reason” (Bxviii). I propose that this brief description may be understood as referring to an alternative human perspective, taken up by the rationalist metaphysician, and portrayed or even enacted in Kant’s presentation of arguments concerning soul, world or God, in the voice (as it were) of the metaphysician. In this alternative perspective (corresponding to the three points concerning “the human perspective”, above), (1) the human knower employs the cognitive faculties of understanding and reason, not sensibility; (2) therein uses the categories (a priori concepts of the understanding) to attempt to describe objects of the a priori ideas of reason, i.e., (3) objects (soul, world-whole, and God) that are non-sensible and so transcendent, definitively beyond or outside of sensible experience. In particular, because these ideas are formulated by reason, their objects are understood as totalities or as the “unconditioned”. That is, each idea of reason (of soul, world-whole, or God) identifies a purported ultimate and total ground for some aspect of experience – whether the subject or an aspect (spatio-temporal or conceptual) of known objects (see A333-4/B390-91). In so doing, reason formulates a conception of a thing that cannot be presented in experience, which never presents ultimate grounds, nor complete totalities (at least that we could know or recognize as such). I will henceforth refer to this complex as “the alternative perspective”.

This alternative perspective can, I suggest, allow the human knower (and Kant describing the same) to identify “the human perspective” as a perspective, as a specific – not unique – way to approach objects. One may recognize that using concepts to judge sensibly given objects is not the only form knowledge may take if one also can do something else, i.e., think of objects using pure reason. In particular, the rationalist metaphysician who attempts to know about the soul by pure reason takes herself to have intellectual intuition (and so not to require sensible intuition to present objects about which to think). For she takes it that her self-consciousness, her non-sensible (intellectual) representation of “I think”, presents her immediately with a particular, given entity – herself, her soul.21 In Kant’s treatment of the project of proving God’s existence, the rationalist metaphysician again attempts to perform a cognitive function Kant denies to the human perspective: to “synthesize” concepts with one another directly (and justifiably), without recourse to sensible experience as a ground for linking the two concepts. For, Kant maintains, in arguing a priori that God exists, rational theologians aim to connect two, non-identical concepts, both attained by considering what the “unconditioned condition” for contingent beings might be: the concept of necessarily existing being and the concept of the being with the “highest reality” (ens realissimum).22

In thus subtracting sensibility and laying claim instead to different cognitive abilities (to synthesize concepts and intellectual intuition), these two rationalist metaphysical projects illuminate that the human perspective is a specific perspective, deploying specific cognitive faculties/types of representations: understanding together with sensibility, concepts as applied to and synthesizing sensibly given intuitions. Since these are, moreover, cognitive attitudes that human beings may take, projects in which human knowers may engage, in which philosophers in the history of philosophy have engaged – and which Kant enacts in the Dialectic – they make good on the gesture (in the Analytic) at possible alternatives to the human perspective.

Kant’s discussion in the Antinomies chapter is more obviously concerned with the failures of the alternative perspective (to which I turn shortly). But the position taken up by the rationalist metaphysicians here also, arguably, illuminates why one might take the human perspective not just to be a specific perspective, but also a limited one, which gives a view of “mere” appearances, not self-standing, independent things. The alternative perspective portrayed in the Paralogisms and Ideal chapters may already appear superior to the human perspective, in that the cognition therein enacted appears independent of sensibility, and thus both more “native” to human thinking (more centrally part of who we are as knowers, less dependent on external information) and immune to experiential refutation. But in the Antinomies chapter, this superiority is more directly at issue. For here the rationalist metaphysician (the alternative perspective) focuses on the very objects of experience, treating them as subject to reason’s demands to specify their ultimate conditions (their spatio-temporal extent, their ultimate substantial components, their causal and existential foundations). This demand brings out the insufficiency of objects of experience by the standards of reason: none of them will ever count as such an ultimate ground, all are ineliminably contingent, dependent on something else. In order to take them to have fully grounded existence, all must either be conceived of as parts of an infinite (and so not-experienceable) whole or as furnished with a further, ultimate, non-experiential grounding, in a different sort of object (God as first cause and necessary being, or perhaps monads—see A416/B443-A418/B446). Consequently, the human perspective, in focusing exclusively upon objects of experience, is concerned exclusively with objects of lesser, dependent metaphysical status – things that by their very nature depend upon something outside that perspective for their “real”, ultimate, grounded being, that are not (in this sense) things in themselves.23

1.5 Only Sort of an Alternative Perspective

One may immediately object the following, however: can it be that on Kant’s view, human beings actually can occupy the alternative perspective? Can it be, as just suggested, that this alternative perspective is somehow superior to the human perspective, in furnishing knowledge of what things in themselves are like? Kant’s locutions occasionally suggest so. In the B Preface passage concerning the experiment, for example, he contrasts the human perspective and its limited knowledge – knowledge (only) of “what we ourselves have put into” objects – with the conception of objects by pure reason. The latter (Kant implies) does not comprise knowledge of what we have “put into” objects, rather it conceives them in themselves.24

Yet this cannot quite be right, given Kant’s emphatic denial of our knowledge of things in themselves, his restriction of human knowledge to the human perspective. And of course in the Dialectic Kant not only enacts the rationalist metaphysical project – the alternative perspective – but also, indeed dominantly, argues that it is a failure. This is the principal reason for my qualifications (above): the alternative perspective is only sort of an alternative perspective. It is not fully a perspective on objects, first, because its cognitive attempts are failures, its point of view empty, fragmentary, contradictory. And, I add now, Kant’s arguments against its attempts at knowledge articulate the ways in which it fails such that the human knower is, as it were, pointed back to the human perspective.

That is, in his critical analysis both of the proofs for the existence of God and arguments concerning the nature of the soul, Kant argues in effect that the alternative perspective fails in its attempts to know its supersensible objects precisely because it excludes sensibility – precisely in the way in which it attempts to exit the human perspective. On Kant’s analysis, the attempted proofs of God’s existence all succumb to conceptual fragmentation: for human knowers, there is no way, justifiably, to synthesize two different concepts, unless one can invoke sensible experience. But of course the alternative perspective – especially when attempting to know God as necessary being – excludes sensibility.25 Hence, Kant claims, there remains an ineliminable gulf between the concepts to be synthesized.26 As it turns out, the alternative perspective does not present a coherent view of a unified object.

The rationalist metaphysical claim to know the self by intellectual intuition is likewise false, Kant argues: insofar as there is some intuitive – immediate, existential, singular – content to the rationalists’ representation of their selves, this content comes from temporally formed sensibility, not from awareness of the “I” of thought (itself merely an empty, formal structure of self-consciousness).27 Kant thus accuses the rationalist metaphysical arguments of the fallacy of ambiguity: their purported conclusions, their purported grasp of a supersensible object, turns on unacknowledged substitution of sensible for purportedly purely intellectual conceptual content. In the terms I have been using, the alternative perspective fails to know its object (engaging rather in tautology or fallacy), and it proves to be dependent (in an unacknowledged way) on the human perspective, specifically on sensibility; it is dependent, then, precisely on that which it claims to exclude.

The failure of the alternative perspective is even more centrally at issue in the Antinomies chapter, where metaphysical reason is portrayed as enmeshed in unavoidable contradictions. Contradiction, of course, is a cognitive failure; if one is led to assert both p and not-p, something has gone wrong, by reason’s own most basic standard. A perspective plagued by contradiction is clearly a failed perspective; it cannot provide an integrated “view” of or approach to objects.

Kant provides a more complicated explanation of this failure than in the two other cases. The antinomies do not arise because of the exclusion of sensibility, for here the alternative perspective attempts to find its own proper objects, things conceived by reason, in or at the basis of sensibly presented objects of experience. But there are two ways to perform such an identification of sensible and rationally conceived objects. As in the antithesis positions – which Kant also names “dogmatic empiricism” (A470-71/B498-9) – one may insist that spatio-temporal objects of experience are fully self-standing things, as conceived by reason.28 Or, with more traditional rationalists (the thesis positions), one may claim that spatio-temporal objects must be grounded in some more basic, rationally conceived things. Hence, necessarily, two opposed positions – the antinomial conflict – as well as Kant’s resolution to it: if one distinguishes objects of experience from objects thought by reason, the contradiction disappears.29 Indeed Kant claims more strongly that one learns here to distinguish appearances from things in themselves, i.e., to accept the restriction of human knowledge to the human perspective as such (Bxx; I return briefly to this characterization of the results of the experiment in the Coda).

The experiment that Kant enacts in the Dialectic thus shows what happens when one tries to exit the human perspective: cognitive failure. We must therefore qualify – but not dismiss – my above characterization of the alternative perspective: it is not, fully or stably, an alternative perspective accessible to or inhabitable by human beings. However, I propose that it remains a “position” from which one can recognize the specific, limited character of the actual human perspective, universally shared and necessary as it may be.30 In trying and failing to know oneself through intellectual intuition, or to know God by synthesizing concepts, one recognizes that one is specifically and ineluctably dependent on sensibility, as an element of cognition distinct from thought. And in recognizing (through antinomial contradictions) that the sensible objects with which the human perspective is concerned do not sustain the demands of full rational explanation – that they are not, and cannot be derived from unconditioned, absolute grounds – one recognizes the restrictedness of the human perspective, its insufficiency to answer all questions about, to provide full grounding for, the only objects it does know. Thus, though there is no actual variability or plurality of (successful) human cognitive standpoints on Kant’s view, nonetheless human beings can attempt to exit the human perspective. And through engaging in such attempts, and recognizing their failures – through philosophical experimentation – one can recognize the specificity and the limitations of the human perspective.31

1.6 Coda

In closing, I’d like to acknowledge two limitations of the above proposals. I have suggested that Kant’s philosophical experiment can address the question how one can identify one’s own, necessary perspective, without really being able to leave it, not (incoherently or inconsistently) from outside but from within. There is another important aspect of this project that I have not discussed explicitly, however, which may seem all the same to take one outside the human perspective. That is, Kant’s explanations of the origins of the ideas of reason (soul, God, and world-whole) and his diagnoses of where the alternative perspective goes wrong in trying to know them – at all of which I have merely gestured here – are also crucial to the “experiment” I am describing. In these explanations, Kant aims to show that the alternative perspective is, as it were, constructed out of materials (such as the structure of self-consciousness or the serial relations of conditioned to condition that connect objects in experience) constitutive of the human perspective, or by purporting to subtract some such materials (i.e., primarily, sensibility). It is in this way as well not really an alternative perspective – not really other, but rather a newly (and, Kant wishes us to discover, badly) reorganized version of the human perspective.32 But there is one signal element in these explanations and in the aspirations of the alternative perspective, which is not such a material: reason itself. To identify the human perspective as such, to see its limits, dependence, and insufficiency, human beings must have a faculty that, as Kant writes, “has a natural propensity to overstep all these boundaries” (A642/B670), that demands absolute explanations, grasp of totalities and necessary grounds, that drives toward the unconditioned, and so (according to Kant) pursues illusory, deceptive self-transcendence. But is this self-transcending drive really accommodated within the human perspective?

Perhaps an answer to this question could be found in Kant’s account in the Appendix to the Transcendental Dialectic of reason as providing regulative principles for empirical investigation, and indeed of the ideas of reason as opening up a “focus imaginarius” – aspirational end points or goals that orient or unify the human perspective in a way similar to the focal points of literal, visual perspectives (on this account, see Massimi 2017, 2018; Zuckert 2017). In other words, it appears that for Kant, if appropriately directed or understood, reason and its self-transcendence are constitutive of the limited, human perspective as such. Perhaps so – but then again one may wonder what it can mean to have a standpoint, a perspective, defined by an orientation to something beyond itself.

Second, I hope to have suggested not just how Kant can explain that one is aware of the human perspective (despite its universality and necessity), but also the meaningfulness of taking the human perspective as a perspective: it is a specific form of cognition, of a specific kind of object, that may be taken as, according to the standards of reason, insufficient, and limited. None of this amounts, however, to showing that the objects of such cognition are metaphysically or otherwise dependent on the human perspective, nor that claims about them are true only for, localized to, that perspective. As noted above, however, Kant claims that his experiment does show this:

on the assumption that our cognition from experience conforms to the objects as things in themselves [i.e., as carried out by the alternative perspective in the Antinomies], the unconditioned cannot be thought at all without contradiction, but … if we assume that our representation of things as they are given to us does not conform to these things as they are in themselves but rather that these objects as appearances conform to our way of representing, then the contradiction disappears. (Bxx)

On Kant’s description, the contradictions in the antinomies do not show (or do not merely show) that an alternative perspective, that of rationalist special metaphysics, fails. Rather, Kant claims that the conflicts arise if one denies the distinction between appearances and things in themselves, that is, denies that there is a human perspective at all. Thus, he famously claims, the antinomies are an “indirect proof” of transcendental idealism: the contradiction disappears if one recognizes that appearances are appearances, i.e., only so for the human perspective. This claim raises questions for my description of the alternative perspective as a perspective. For example, does this re-description (as I hope) free Kant from his residual-rationalist, apparently dogmatic claim that reason (what I call the alternative perspective) would provide human beings with knowledge of things in themselves? Or is the alternative perspective insufficiently robust to delimit the human perspective as such, unless it is taken so to transcend perspectival knowledge altogether?


  1. 1.

    Kant, Critique of Pure Reason, A26/B42, my emphasis. Citations to the Critique will be, as customary, to the A/B page numbers of the first and second editions. Translations are from Kant (1998).

  2. 2.

    On perspective as a metaphor used in philosophy, I have profited from Conant (2005), as well as Moore (1997), though the latter is a deeper treatment than I can properly engage with here.

  3. 3.

    Kant does not deny, of course, that individual human beings could have different empirical standpoints (either literal or metaphorical).

  4. 4.

    This disanalogy is heightened by the fact that Kant uses “standpoint” to describe space, so that the very idea of moving to a different spatial position, or of different people occupying different spatial positions, is ruled out. (Here Kant follows Leibniz, who takes space to be merely phenomenal, yet describes the truly-existing monads as having their own distinctive “perspectives” on the world-whole.) In line with Strawson’s objections to Kant (1966, e.g., p. 41), and with Conant (2005), one might accuse Kant therefore of illicitly (perhaps even self-contradictorily) extending an intramural experiential concept (of occupying a spatial position) to extra-experiential use. I will not be able to take up this specific version of the problem here.

  5. 5.

    A 362-3. Kant uses “Gesichtspunkt” here, and then glosses the same observer position as “Standpunkt” at A364.

  6. 6.

    A 353-4. “In its place” translates “an seine Stelle”. These passages may seem to conflict with one another, since in the first, Kant suggests that one can regard a thinker from outside, precisely not “substituting one’s own subject” for that thinker’s self-consciousness. But we may note that in the thought experiment the outside observer is furnished with the same necessary features of human knowledge (specifically, time as a form of sensibility but also, presumably, its own self-consciousness). Thus, the observer does not in that respect occupy an “external” (i.e., non-human) perspective. The staged outside observer also does not aim to judge “my self” as a knower, but in “outer intuition”, that is, as a substantial thing persisting in time, which supports or underlies that thinking or (in short) as object.

  7. 7.

    See A27/B43, following the characterization of space as belonging to the human standpoint.

  8. 8.

    There has been discussion about whether intuitive intellect and intellectual intuition are the same – e.g., Förster (2009) – but I think the present line of argument is independent of a decision on this issue.

  9. 9.

    I think it is not coincidental that Allison’s interpretation, which emphasizes the epistemological (not metaphysical) dimensions of transcendental idealism – and so is close to thinking of Kant as identifying a human “perspective” – also emphasizes the Kantian contrast between human and (purported) divine cognition (see Allison 1983, chapter 2, especially 19–24).

  10. 10.

    Heidegger (1990) and Strawson (1966). Obviously, I am presenting a very broad-brush picture here.

  11. 11.

    Unsurprisingly, Heidegger and Strawson get to this conclusion differently, Strawson proposing a Kantian “metaphysics of experience” (analysis of what objects must be like, if there is to be experience), while Heidegger takes Kant to engage in a metaphysical (“ontological”) investigation of human subjectivity, which also reveals the ontology of this sort of object. Heidegger does emphasize the finitude of the subject, a Kantian theme connected to the topic of this paper, but finitude need not be understood in terms of the perspectival limitation of claims, with which I am concerned. For a recent realist interpretation of Heidegger and his reception of Kant (somewhat along these lines), see Kinkaid (2018).

  12. 12.

    I take it that this could describe a contemporary naturalist approach to a priori knowledge: human beings justifiably deploy a priori concepts (innate cognitive modes of interpreting experience such as object-permanence) because this is accurate to the world – and this “fit” between innate capacities and world is in turn to be explained in evolutionary terms (that human beings have evolved cognitively to fit their environment). Of course such a line of argument is subject to Nietzsche’s objection that falsities might be evolutionarily just as useful to the human animal as accurate conceptualization.

  13. 13.

    The “neglected alternative” objection (prosecuted by Adolf Trendelenburg in several works in the 1860s), namely that Kant does not rule out the possibility that space and time are both forms of our intuition and characterize things in themselves, occupies similar territory. I do not think the approach I explore here can address this way of formulating the problem.

  14. 14.

    According to Kant’s own divisions of the work, this is not accurate: organizationally (if not in terms of page numbers), the Transcendental Dialectic is something like an eighth of the work. For Kant divides the book into two (unequal) halves: the Method, a section at the end of the work concerning what philosophy is to do after Kantian critique, and the Elements, the much longer first half of the work, which includes the Transcendental Aesthetic, containing Kant’s arguments that space and time are a priori forms of intuition, and the Transcendental Logic. The Logic is in turn divided into two subsections, the Transcendental Analytic, concerning the a priori concepts of the understanding (categories) appropriately and necessarily employed with respect to objects of experience, and the Transcendental Dialectic, concerning the ideas of reason that purport to present objects beyond experience.

  15. 15.

    Bxx; “drive” translates “treibt”. See also A327/B383-4 and A332/B389.

  16. 16.

    One might wonder whether the field of potential non-experiential knowledge is larger than rationalist specialist metaphysics: in showing that we cannot know these things (soul, world-whole, God) a priori, has Kant thereby shown that we cannot know anything about anything non-experiential? Here Kant’s transcendental-psychological etiology of the ideas of reason proves crucial: it aims to show that these are the only rational (non-practical) ideas human beings are able to form of transcendent things. Investigation of this line of thought lies beyond the scope of this paper, however.

  17. 17.

    Bxvi. One suggestive element of this description of Kant’s project is that it proposes a new view of Kant’s philosophical methodology: as experimental, trying out hypotheses, seeing which best fits the philosophical “facts”, rather than as the “apodictic” demonstrative argument to which Kant lays claim (but which have not been found so decisive by most of his readers).

  18. 18.

    Bxviii. Kant’s contrast in this immediate textual location is to the knowledge we have of objects in accord with “what we ourselves have put into them”, whereas I use a more specific contrast (to “intuition” and “experience”) from earlier in the passage. I return to Kant’s broader contrast below.

  19. 19.

    I should note that the subject matters of the Dialectic also identify the “meaning” of Kant’s assertion in the Analytic that human cognition is only from the human perspective, in another sense: it matters that objects of experience are such only for the human perspective, so as, famously, to “make room for faith” (Bxxx). If objects of sensible experience are (mere) appearances, then one may believe – perhaps has to believe – that sensible nature is not all there is, that empirical scientific description does not exhaust truths about things. In particular, Kant claims, we are permitted to believe that our agency may be free and that God exists – as we are required to do for moral reasons. Hence the desirability of the limitation connoted by identifying a perspective, even one shared by all human beings.

    Given that morality requires this different view of oneself, one might think that it provides an alternative perspective from which we can recognize the limits of the human (cognitive) perspective (I owe this suggestion to Alix Cohen). This, however, seems to me to mistake the structure of Kant’s philosophical position: as clearly stated in the “make room for faith” passage, theoretical (cognitive) limits must first be established, on theoretical grounds, in order to permit moral self-reconception as an alternative perspective.

  20. 20.

    Here I qualify the claims of the preceding section that insofar as human beings share the same standpoint, we are incapable of exiting it or occupying an alternative one.

  21. 21.

    This is how I read Kant’s references to intuition in the Paralogisms, e.g., at B411-12, though this is a more controversial interpretive and reconstructive claim than I can defend here. I note also that my characterization of this representation as a purported intellectual intuition – an immediate presentation of a particular (object), given by intellect, not sensibility – does not include an element often associated (by Kant and by scholars) with intellectual intuition, namely that such an intellect would produce its objects. I think these two aspects of purported intellectual intuition can in fact be separated: both characterize God’s purported intellect, on Kant’s sketchy characterization, but it is not clear to me why non-divine intellects could not have a different form of intuitive intellection. So Förster (2009) also concludes, on different grounds.

  22. 22.

    Insofar as Kant’s portrayed rationalist metaphysician seems to perform (somewhat) different cognitive actions in the three chapters of the Dialectic, one might think that there are three (or more) alternative perspectives portrayed in the Dialectic. I do not have a worked-out account of how all three are deeply connected, but I think my general point is unchanged either way – some “experimental” alternative fails.

  23. 23.

    To be clear, the metaphysical status – of dependency or contingency – at which I gesture here is not dependence on human knowers (i.e., it is not idealism). Rather, these objects are dependent (on other things, of some other, less flimsy kind). Insofar as these are the only objects known by the human perspective, then, that perspective does not provide a view of metaphysically independent things. As I discuss in the Coda, it would require further argument to move from this dependence to idealism (dependence of appearances on human knowers/perspective).

  24. 24.

    Bxviii. One might call this the “residual rationalism” in Kant’s thought: if we were to know things in themselves, we would know them by reason alone. Perhaps he thinks that insofar as we have a positive conception of things as separate from experience (not just the “negative” noumenon of B307, i.e., an abstraction from objects of experience), this conception is gained from reason (in the form of the ideas of reason). But I think he also believes that reason (not sensibility) articulates the independent-ness of things in themselves. Langton (1998) is still the most worked-out account (to my knowledge) of Kantian idealism as a form of such residual rationalism.

  25. 25.

    I gloss over considerable detail here, and proceed blithely on the assumption that Kant’s criticisms of rationalist metaphysics are successful.

  26. 26.

    So I construe Kant’s arguments (e.g., A611/B639-A613/B640) that we cannot prove that the necessarily existing being is the ens realissimum or vice versa.

  27. 27.

    See A350-1: because the representation of the “I” is merely a form of thought, not an intuition, it does not have content such as “everlasting duration” (which would be content of sensible, i.e., temporal intuition and which is introduced in thinking of the self as enduring substance).

  28. 28.

    This position may not seem to conceive of objects by pure reason, but insofar as it takes objects of experience to be infinite (or parts of an infinite series), it re-describes them in terms that are themselves neither drawn from, nor presentable within sensible experience. For such reasons, Boehm (2014) argues that Kant could be targeting Spinoza in the antithesis positions.

  29. 29.

    Kant proffers two different sorts of resolutions (to the first and second, and to the third and fourth antinomies, respectively): either one stops demanding that appearances conform to reason’s standards of what independent things would be (because they are merely appearances) or one sees that one might be able to assert both p and not-p, but this would no longer be a contradiction, because the two claims concern different (aspects of) things.

  30. 30.

    Here one might differentiate my proposal from the following stronger, narrower construal of the experiment and its results: the experiment concerns the antinomies alone, particularly the first antinomy, and it establishes that the alternative perspective therein taken up is contradictory, completely incoherent. This interpretation could be supported textually, and might be appealing in presenting Kant as arguing more directly for his view that space and time are merely parts of the human perspective (for it would take the experiment to show that thinking otherwise is flatly incoherent). But it also, I think problematically, isolates the antinomies from the project of the Dialectic (and Appendix to the Dialectic) as a whole. The experiment would also not be able to perform the (on my view methodologically important) role of comparison case: a completely incoherent position cannot be inhabited, so could not be one from which another perspective could be recognized as such.

  31. 31.

    As Mark Alznauer noted (in personal communication), one may then ask: is this recognition true from or for the human perspective (alone)?

  32. 32.

    Though I do not have space to elaborate, I think Kant does and must aim to show this: he must explain how rationalist metaphysics is even possible as a project, given his view of the actual character and limits of human cognition. I am inclined, moreover, to think that this part of Kant’s project in the Dialectic could explain the sometimes problematic nature of Kant’s arguments. That is, Kant occasionally seems not just to use his own terminology, but to invoke some of his own substantive commitments (as established in the Analytic) in his treatment of the rationalist metaphysical project. Such use of his own doctrines seems to beg the question with respect to what one might call his primary aim in the Dialectic: to argue that (contra the rationalists) human cognition is limited in the Kantian way, Kant cannot use as premises those very claims concerning human cognitive limitations. (This problem seems to me to plague the interpretation in Grier 2001 as well, though this fine book has been crucial for my understanding of the Dialectic.) But for this secondary aim – explaining how limited human knowers could attempt to transcend their own, necessary perspective – Kant’s own understanding of the human perspective must be central. Kant seems to me not always clearly to disambiguate these two argumentative tasks; many of his criticisms of rationalist metaphysics might consequently need to be reformulated so as not to beg the question.



For comments on previous drafts of the paper, I am grateful to Mark Alznauer, Karl Ameriks, Morganna Lambeth, Katalin Makkai and participants in the “Knowledge from a Human Point of View” conference at the University of Edinburgh, especially Lorenzo Spagnesi and Michela Massimi.


  1. Allais, L. (2015). Manifest reality: Kant’s idealism and his realism. Oxford: Oxford University Press.CrossRefGoogle Scholar
  2. Allison, H. (1983). Kant’s transcendental idealism: An interpretation and defense. New Haven: Yale University Press.Google Scholar
  3. Boehm, O. (2014). Kant’s critique of Spinoza. Oxford: Oxford University Press.CrossRefGoogle Scholar
  4. Conant, J. (2005). The dialectic of perspectivism, I. Sats: Nordic Journal of Philosophy, 6(2), 5–50.Google Scholar
  5. Förster, E. (2009). The significance of §76 and 77 of the Critique of Judgment for post-Kantian philosophy, part I. Graduate Faculty Philosophy Journal, 30(2), 1–21.CrossRefGoogle Scholar
  6. Grier, M. (2001). Kant’s doctrine of transcendental illusion. Cambridge: Cambridge University Press.CrossRefGoogle Scholar
  7. Heidegger, M. (1990). Kant and the problem of metaphysics (Richard Taft, Trans.). Bloomington: Indiana University Press.Google Scholar
  8. Kant, I. (1772). Letter to Markus Herz of February 12. Reproduced in Gesammelte Schriften 20, 129–35. Berlin: De Gruyter.Google Scholar
  9. Kant, I. (1998). Critique of pure reason (Paul Guyer and Allen Wood, Trans.). Cambridge: Cambridge University Press.Google Scholar
  10. Kinkaid, J. (2018). Phenomenology, idealism, and the legacy of Kant. British Journal of the History of Philosophy, 13(2), 1–22.Google Scholar
  11. Langton, R. (1998). Kantian humility and things in themselves. Oxford: Oxford University Press.Google Scholar
  12. Massimi, M. (2017). What is this thing called ‘scientific knowledge’? Kant on imaginary standpoints and the regulative role of reason. Kant Yearbook, 9, 63–83.CrossRefGoogle Scholar
  13. Massimi, M. (2018). Points of View: Kant on Perspectival Knowledge. Synthese, special issue on the current relevance of Kant’s philosophy. Online first.Google Scholar
  14. Moore, A. W. (1997). Points of view. Oxford: Clarendon Press.Google Scholar
  15. Strawson, P. F. (1966). The bounds of sense: An essay on Kant’s Critique of Pure Reason. New York and London: Routledge.Google Scholar
  16. Zuckert, R. (2017). Kantian ideas of reason and empirical scientific investigation. In M. Massimi & A. Breitenbach (Eds.), Kant and the laws of nature (pp. 89–107). Cambridge: Cambridge University Press.Google Scholar

Copyright information

© The Author(s) 2020

Open Access This chapter is licensed under the terms of the 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 license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license 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.

Authors and Affiliations

  1. 1.Northwestern UniversityEvanstonUSA

Personalised recommendations