ISSN 1393-614X Minerva - An Internet Journal of Philosophy Vol. 8 2004.





Chad Mohler




Kant claims that things-in-themselves produce in us sensible representations. Unfortunately, this “transcendental affection” appears to be inconsistent with Kant’s prohibition against applying the category of causality to things-in-themselves.

This paper gives an account of transcendental affection that does not require it to be seen as a type of causation. Transcendental affection, properly understood, is the logical relation of the ground of things-in-themselves to the consequent of an affected subject. This relation is what one gets when one de-schematizes causation, revealing the underlying hypothetical form of judgment.

So conceived, transcendental affection no longer poses a potentially debilitating problem for the interpreter of Kant who contends that things-in-themselves enjoy an independent objective existence. The paper, then, is a partial defense of such an interpreter against the Kantian interpreter who contends that the thing-in-itself is merely a limiting concept useful for the regulation of thought.



I.  Introduction: Two Opposed Views


There are two common interpretations of Kant’s doctrine of the thing-in-itself in his Critique of Pure Reason. One of these, which I shall call the phenomenalist reading, holds that Kant took the thing-in-itself to be not an existing object, but the mere limiting concept of an object considered in abstraction from all conditions of sensible perception. 1 The thing-in-itself, on this view, is such a limiting concept in two distinct ways. First, the thing-in-itself, being the concept of something unperceivable and hence by Kant’s lights unknowable, places constraints on what we can know. Kant makes the thing-in-itself a limit of our epistemological capacity by first restricting our knowledge to that of things considered as they appear to us and by then construing the thing-in-itself as the concept of a thing considered apart from the manner in which it so appears.


The thing-in-itself is also a limiting concept in a second sense, in that it is the concept of an ultimate logical ground for all judgments about an object of experience. The thing-in-itself is the conceptual unity under which all these judgments fall and by which they are all conditioned — that is, the thing-in-itself as conceptual ground is the condition for the assertion of the judgments. For the phenomenalist, of course, this conditioning is merely conceptual. On the phenomenalist view, we make a judgment about an empirical object as if it were grounded by an actually existent thing-in-itself, all the while bearing in mind that the thing-in-itself is in actuality only a concept, not an object.


Because all judgments about the experienced object are so conditioned by the thing-in-itself and not vice versa, the thing-in-itself is, with respect to the experienced object, unconditioned — that is, there is nothing that could be said about the object that could serve as the condition for asserting something about the corresponding conceptual thing-in-itself. The thing-in-itself, taken in this sense as the ground for all judgments about an empirical object, is the logically most fundamental concept of the object, and as such, it constitutes the most general concept that could be invoked in any line of reasoning concerning and limited to that object. It is thus the logical limit of concepts involved in any series of judgments exclusively about the object. In light of this construal of the thing-in-itself, we can understand in a phenomenalist way one of Kant’s remarks in the B Edition Preface to the Critique, where he says, “For what necessarily forces us to transcend the limits of experience and of all appearances is the unconditioned, which reason, by necessity and by right, demands in things in themselves, as required to complete the series of conditions” (Bxx). 2  The thing-in-itself is the unexperienced, unconditioned unity that we think (again, conceptually) as the ground of all judgments which we make about an empirical object and that we thus consider as completing any “series of conditions” which we might employ in syllogistic reasoning limited to that object. 3 Just as reason seeks to extend itself to the Kantian transcendental ideas of the thinking subject, the world, and God, the phenomenalist holds that reason also strives in its consideration of a single empirical object toward yet another totality, that of the thing-in-itself.


In keeping with this phenomenalist interpretation, we must be careful not to hypostatize the concept of a thing-in-itself as an actually existing object, just as Kant in the Ideal of Pure Reason showed us we must not objectify as a deistic God the unconditioned unity of the series of all objects of thought in general. The lack of existent objects corresponding to the concept of thing-in-itself is what makes this interpretation of Kant a phenomenalist one. Our inability on this view to assert unqualifiedly that things-in-themselves exist means that, at least as far as the First Critique is concerned, Kant can only be justified in according the ontological status of existence to the objects of one’s possible experience, that is, to possible appearances and mental acts. 4  Roughly put, there is nothing that we can say is lurking “behind the appearances.”


The second interpretation of the Kantian thing-in-itself, by contrast, reads the First Critique as asserting the existence of objects considered apart from the sensible conditions under which we perceive objects — there are actual things underlying appearances, on this account. This view, which I shall refer to as the realist interpretation, holds that such objects can be considered in one of two ways: either empirically, as given through their appearances to the sensibility of the observer; or transcendentally, in abstraction from all conditions of the sensibility. 5  In this latter sense, the objects are thought as they are “in themselves.” Though the objects can be thought in this way, we can know nothing about them as they are in themselves, for knowledge can be had only of things for which there are available intuitions, which for humans are exclusively sensible in nature. The object thought about in the above non-sensible, abstract manner is only the unknown “something=X” behind the appearances that that something generates in us. A thing-in-itself on this view, then, is just an existent object considered apart from any sensible means we have of perceiving it. 6 There is a grave threat, however, to any interpretation of Kant which incorporates the above realist construal of the thing-in-itself. The danger is that with things-in-themselves interpreted realistically, Kant’s assertion of the existence of such entities is a dogmatic claim with no justification; and if we have to give up the existence of things-in-themselves, we have to give up the existence of objects “beneath the appearances” altogether. The argument with which the phenomenalist so impugns the realist goes roughly as follows:


(1)      If a thing produces sensible representations in me, it must affect me in some causal way.

(2)      But I can only apply the category of causality to empirical objects, not to things-in-themselves.

(3)     Therefore, I cannot say that it is things-in-themselves that produce sensible representations in me.

(4)   Kant gives us no reason, other than things-in-themselves producing sensible representations in us, to believe that things-in-themselves exist.

(5)   So Kant’s view does not provide us with any legitimate reason to think that things-in-themselves exist; Kant’s view does not give us reason to believe that there are objects considered apart from the conditions of sensibility.


Given this argument, says the phenomenalist, any further insistence on the realist’s part that existence can be attributed to objects considered apart from the manner in which we perceive them is dogmatic and unjustified.


Unlike the phenomenalist, the realist is committed to the doctrine of transcendental affection: that is, to the assertion that objects considered transcendentally as things-in-themselves do, in fact, affect us in such a way as to cause us to have the sensible intuitions that we do. This is the problematic aspect of the realist’s account, for as the phenomenalist's argument points out, it requires that the realist illegitimately apply the category of causality to things-in-themselves. If, as Kant maintains, such an application cannot be made, and if transcendental affection is, in fact, a causal relation, the conclusion of the phenomenalist's argument straightforwardly follows.


That argument does not threaten the phenomenalist. For the phenomenalist who allows that we can only justifiably say that appearances and mental acts exist, there are no non-sensible objects that we can unequivocally assert as existing. This being the case, on the phenomenalist view we cannot consider the objects of empirical intuition — the only objects there are on this account — transcendentally, for there is nothing to them outside of appearance (at least, nothing that we could know). Thus, for the phenomenalist, there is no transcendental affection that we need dogmatically to posit.


Of course, rejecting transcendental affection creates for the phenomenalist problems of her own. Among other difficulties, the phenomenalist faces the challenge of giving some reasonable explanation of what Kant actually means in the numerous instances in the First Critique where he speaks of things-in-themselves affecting us. In this paper I want to focus not on how the phenomenalist might do this, but, rather, on the phenomenalist’s challenge to the realist interpreter of the First Critique. I will say how, despite the attractiveness of the phenomenalist’s argument, the realist can nonetheless evade the phenomenalist’s indictment of his transcendental affection.



II.    The Phenomenalist’s Argument in Detail


 In numerous passages in his Critical writings, Kant affirms that things-in-themselves appear to us because they affect us in some causal way. The following are examples of such references to this “transcendental affection”:


The faculty of sensible intuition is strictly only a receptivity, a capacity of being affected in a certain manner with representations... The non-sensible cause of these representations is completely unknown to us... We may, however, entitle the purely intelligible cause of appearances in general the transcendental object, in order to have something corresponding to sensibility viewed as a receptivity (A494/B523).

How things may be in themselves, apart from the representations through which they affect us, is entirely outside our sphere of knowledge (A191/B236).

In the process of warning [the sensibility] that it must not presume to claim applicability to things-in-themselves but only to appearances, [the understanding] does indeed think for itself an object in itself, but only as transcendental object, which is the cause of appearance and therefore not itself appearance, and which can be thought neither as quantity nor as reality nor as substance, etc. (because these concepts always require sensible forms in which they determine an object) (A288/B344).

And we indeed, rightly considering objects of sense as mere appearances, confess thereby that they are based upon a thing in itself, though we know not this thing as it is in itself but only know its appearances, viz., the way in which our senses are affected by this unknown something. The understanding therefore, by assuming appearances, grants also the existence of things in themselves, and thus far we may say that the representation of such things as are the basis of appearances, consequently of mere beings of the understanding, is not only admissible but unavoidable (Prolegomena, §32).


There are a number of interesting points to observe about these passages. First, while in some of them Kant refers to transcendental objects as the causes of the appearances we have, in others he explicitly posits things-in-themselves in that affective role. It is clear, however, that in these particular passages, insofar as he is referring in all of them to some non-sensible entity which is the underlying, unknown origin of the mental representations that a subject has, he uses ‘transcendental object’ and ‘thing-in-itself’ synonymously. In the third passage, Kant even equates the two terms.


Furthermore, the passages make it clear that the things-in-themselves affecting the subject are themselves not objects of knowledge, for there corresponds to them no sensible intuition. As the Prolegomena passage suggests, the things-in-themselves doing the affecting are objects considered in abstraction from those very sensible conditions under which they appear to us. Considered in this way, we can know nothing about them because without sensible intuitions of them, we cannot determine their properties by subsuming the concept of the object under other concepts. As Kant says in the B Transcendental Deduction, “Our conclusion is therefore this: the categories, as yielding knowledge of things, have no kind of application, save only in regard to things which may be objects of possible experience” (B148). Things-in-themselves, as objects considered apart from the sensible conditions of possible experience, are not the sorts of objects to which the categories can be applied to yield any determinate knowledge.


The very non-applicability of the categories to things-in-themselves is the origin of the problems for the realist that the phenomenalist’s argument brings to the fore. As the above passages indicate, Kant appears to construe the manner in which things-in-themselves affect the subject in a causal way, which would imply that the realist is committed to affirming the truth of the first premise of the phenomenalist’s argument, that things-in-themselves’ production of sensible representations in a subject requires some sort of causal affection of the subject by those things-in-themselves. But, by the above admission that categories are not applicable to things-in-themselves, the realist must accept the second premise as well. The first two premises imply the preliminary conclusion in (3): we cannot say that things-in-themselves cause the sensible representations that we have, since such a judgment requires that the category of causality be definitely applicable to things-in-themselves.


Now the realist interpreter of Kant is in trouble, for the only reason Kant gives for allowing us to assert the existence of things-in-themselves as subject-independent objects is that, on the realist view, subject-independent objects considered as things-in-themselves are what produce sensible representations in us. 7  If the realist really is committed to the argument the phenomenalist describes, he cannot justifiably posit subject-independent objects considered as things-in-themselves as the causes of the appearances he has, so that he has no justifiable reason on the basis of Kant’s view for saying that things-in-themselves exist at all.


It appears, then, that unless the realist is willing to assert dogmatically that things-in-themselves exist, she must admit that we really have no reason for saying there are objects “beneath the appearances.” Her realism under such a concession would dissolve away into a mere phenomenalism that asserts the “empirical reality” of objects without the transcendental substrate of existent, non-sensible things-in-themselves underlying those real empirical entities. Considered apart from their relation to us, subject-independent objects on this view could not be justifiably said to be anything at all.


III.         The Realist Vindicated: Affection De-Schematized


Fortunately for the realist, she is not forced to subscribe to this phenomenalist vision of the world. Despite Kant’s characterization of affection in the above passages as a causal relation between the subject and things-in-themselves, we can construct a Kantian view of transcendental affection that is non-causal in nature and thus does not rely on a misapplication of the categories to subject-independent objects considered apart from the conditions of their sensible perception.


First, it is important to note that, Kant’s above causal descriptions of affection notwithstanding, it follows from many positions of Kant in the Critique that transcendental affection is definitely not a causal relation. We have already mentioned, of course, that Kant does not allow the categories, including that of causality, to be applied to things-in-themselves. We can better understand why we are prevented from considering things-in-themselves in any causal connection with the subject by examining a few other sections of the Critique. As Kant notes in his discussion of the Second Analogy, any causal relation between two objects requires that those objects be determined in time (A190/B234). Kant’s transcendental idealism commits him to the view that subject-independent objects do not have temporal properties apart from our perception of those objects. In particular, things-in-themselves, then, are not determined in time, so that we cannot properly postulate a causal relation involving them.


In his solution to the Third Antinomy, Kant makes a similar point with respect to the acting subject. Considered according to its intelligible character, “which does not itself stand under any conditions of sensibility,” “this acting subject would not ... stand under any conditions of time; time is only a condition of appearances, not of things in themselves” (A540/B568). Just as things-in-themselves cannot be determined in time, the acting subject, merely intelligibly considered, is also not temporally conditioned, so that what it does cannot have “a place in the series of those empirical conditions through which the event [of its action] is rendered necessary in the world of sense” (A540/B568). The actions of the intelligible subject, in other words, cannot be a part of any series of empirical causes and effects.


Kant’s talk of intelligible agents in his solution of the Third Antinomy and the non-applicability of the category of causality to atemporal things-in-themselves allows us to see that transcendental affection cannot be a causal relation of the empirical type. But what type of relation is it, then? It is important to note, first, that Kant allows us to describe affection in an empirical manner. For instance, he describes color as “modifications of the sense of sight, which is affected in a certain manner by light” (A29). We as empirically considered objects are affected by another empirical entity, light, in such a way that we perceive certain hues. Later on, in the Third Analogy, Kant says, “Each substance... must therefore contain in itself the causality of certain determinations in the other substance, and at the same time the effects of the causality of that other” (A213/B260). Each empirical object, according to this claim, stands in mutual causal interaction with other empirical objects; and it is this “dynamical community,” Kant goes on to say, that allows our minds, empirically considered, to be affected by empirical objects in such a way that we perceive them as coexisting: “The light, which plays between our eye and the celestial bodies, produces a mediate community between us and them, and thereby shows us that they coexist” (A214/B260). As further evidence that Kant allows us to consider affection in empirical terms, Kemp Smith cites at least 26 passages in Kant’s Opus Postumum referring to empirical causes of sensations. 8


Now, in my initial description of the realist’s interpreter’s position, I distinguished two ways of viewing objects on Kant’s view. On the one hand, we can concern ourselves exclusively with the way that we experience objects as empirically appearing to us; on the other hand, we can consider objects in abstraction from the conditions of sensibility, as things-in-themselves, and inquire as to how they transcendentally appear to us. Likewise, we can consider affection in two different ways. I have already mentioned above the empirical sense in which we can consider affection, as a physical causality between empirical objects and the empirically considered self. But we can also take affection in a transcendental sense by considering the causal relation of empirical affection in abstraction from all sensible conditions.


A determination of what is left to the causation of empirical affection after we abstract from it all conditions of sensibility requires that we first briefly review a few passages from the Transcendental Analytic where Kant describes the functions of judgment and their connection to the schematized categories. For Kant, judgments basically consist of the relation of one concept to another concept, where the latter concept, in turn, is itself related to an object of appearance (cf. A69/B93: “In every judgment there is a concept which holds of many representations, and among them of a given representation that is immediately related to an object”). In this way, judgments serve as round-about ways of representing the objects of appearance — they ascribe to a concept associated with such an object a concept-predicate that presents that object as having a certain property. The indirectness of the concept-predicate’s relation to the object, through the object’s associated concept and not directly to the object itself, is what prompts Kant to call a judgment “mediate knowledge of an object” (A69/B93). Other judgments can represent the object of experience as having additional properties; and insofar as these other concepts are related to the same object-concept, the representations are unified under the concept of one object. This explains why Kant refers to judgments as “functions of unity among our representations; instead of an immediate representation, a higher representation, which comprises the immediate representation and various others, is used in knowing the object, and thereby much possible knowledge is collected into one” (A69/B94).


After describing the general character of judgments, Kant goes on to introduce four tripartite divisions presenting the various forms that a judgment can take. These “forms of judgment” are a catalog of the various logical features that a judgment in its representation-unifying function can exhibit. Of particular interest to us, under the heading of “Relation,” is the hypothetical form of judgment. Kant exemplifies this form with the judgment, ‘If there is a perfect justice, the obstinately wicked are punished’ (A74/B99). A judgment of this type, Kant notes, is composed of two propositions whose truth value is undetermined. What is important to the hypothetical type of judgment is not the truth but the relation of the two propositions. As Kant says, “It is only the logical sequence which is thought by this judgment” (A74/B99), and he earlier characterizes this logical sequence as the relation “of ground to its consequence” (A73/B98). Thus, what a hypothetical judgment affirms is that the existence of the state of affairs described in the first proposition of the judgment entails that the state of affairs described in the second proposition also obtains. The judgment, in other words, unifies the two states of affairs in the relation of ground to consequent; the state of affairs of the first proposition is the ground of the state of affairs of the second.


The hypothetical form of judgment is particularly important for our purposes, for, as it turns out, it is the form which, in its application to the manifold of pure intuition, is the category of cause and effect (A79-80/B105-106). Properly schematized, this pure concept of the understanding can be applied to the objects of possible experience, supplying us with the concept of causality under which the empirical objects involved in empirical affection can be appropriately thought. The schematization of the pure concept produces a “temporalized” category: that is, a pure concept of the understanding translated into temporal terms that allow that category to be applied to empirical objects. For causality, the condition of this temporalization, the schema, is “the succession of the manifold, in so far as that succession is subject to a rule” (A144/B184). How does this schema relate pure concept to empirical appearance? As mentioned above, the pure concept of causality, as a form of judgment applied to the manifold of pure intuition, gives a logical relation between ground and consequent. The schema, then, associates this ground and consequent with two events “of the manifold” of empirical intuition occurring successively in time in accordance with the law of cause and effect developed in the Second Analogy. This schematic association between the rule-governed succession of empirical events and the ground-consequent relation of the pure concept is what enables us to apply the category of causality to empirical objects in general, and to the empirical objects involved in empirical affection in particular.


In my attempt at developing an account of affection transcendentally considered, I have thus far given a rudimentary view of the relationship between judgments, categories, and schematized categories in the Analytic. In the context of that explanation, I have shown that underlying the schematized/temporalized category of causality that applies to objects involved in empirical affection is the hypothetical form of judgment. Now, it is this logical relation of ground to consequent that I want to argue we can count as affection considered in the transcendental sense. Just as we considered objects transcendentally by characterizing them in abstraction from the conditions of sensibility through which they become empirically manifest to us, we can do the same with affection. As Kant says at the beginning of the Analytic, “If we abstract from all content of a judgment,” we obtain the forms of judgment (A70/B95). Thus, considered in abstraction from the spatiotemporal conditions that characterize it, a judgment involving the subsumption of objects involved in empirical affection under the category of causality becomes simply the assertion of a logical relation between ground and consequent. This is the relation of affection considered transcendentally.


Let us consider this relation and its connection to empirical affection in more detail. The objects involved in empirical affection are the empirical object that affects a subject’s sensibility and the (empirically considered) subject so affected. The empirical object and the subject stand in a causal relation to each other that can be captured in a hypothetical judgment — for instance, “If light strikes my eyes, I will see some color.” The light’s striking my eyes causes me to perceive a particular color. If we abstract this causal judgment in hypothetical form from all conditions of sensibility, we are left with the bare assertion of a logical relation of ground to consequent. In the selected example, this assertion is that some relation (we-know-not-what) between myself and the light empirically affecting me — both considered apart from the conditions under which I and it are empirically perceived — is the ground for my affected mental state, considered apart from the way that state is perceived by me through inner sense. Alternately, insofar as we are able to characterize in a physical way the object’s affecting me, we can equate the effect of that affection with a physical state of my body. In this case, then, what is grounded in the logical judgment is this physical state of my body, considered apart from the manner in which we empirically perceive it.


Affection transcendentally considered, then, is simply a logical relation of ground to consequent, where an unknown relation between an empirically affecting object and an affected subject, considered as things-in-themselves, is taken to be the ground for the existence of a certain affected mental/physical state in that subject, considered apart from the manner in which we perceive that state and that subject. The logical relation of transcendental affection entails that if the unknown relation between the empirically affecting object and affected subject obtains, the mental/physical state of the subject will obtain as well (where here, again, the object, subject, and subject’s state are taken as objects considered apart from the conditions under which we perceive them).


I have mentioned a bit about what transcendental affection is; now I will describe what it is not. To begin with, affection considered in abstraction from the conditions of sensibility is not a relation of determinate objects to determinate objects. The subject, object, and subject’s state involved in the logical relation of affection are, like all objects considered as things-in-themselves, merely indeterminate “somethings=x,y,z” about which we can know nothing at all. We cannot even know that the “x”, “y”, and “z” somethings are distinct. There may be no one-to-one mapping at all between the empirical objects involved in an affection relation and those objects considered as things-in-themselves. All the realist can say about objects considered transcendentally, and still remain in agreement with the limits Kant establishes concerning the unknowability of things-in-themselves, is that there is some transcendental substrate underlying the appearances of empirical objects to us.


In characterizing the transcendental object, Kant makes exactly this point. He asks, “What, then, is to be understood when we speak of an object corresponding to, and consequently also distinct from, our knowledge?” He responds, “It is easily seen that this object must be thought only as something in general=x, since outside our knowledge we have nothing which we could set over against this knowledge as corresponding to it” (A104). The passage from which this quote comes is the A edition Transcendental Deduction, in which Kant is probably more interested in using the term ‘transcendental object’ to signify the conceptual unity that keeps our judgments “from being haphazard and arbitrary” (A104) than he is in using that term to refer to things-in-themselves. Nevertheless, in light of the fact that things-in-themselves are, like the transcendental objects described here, objects as they are considered “distinct from our knowledge,” this passage lends weight to the claim that Kant took things-in-themselves to be merely indeterminate somethings. This claim is further reinforced in a similar passage in the “Phenomena and Noumena” section of the Critique, where Kant equates things-in-themselves with noumena and refers explicitly to their indeterminate character. He says there that the understanding “limits sensibility by applying the term noumena to things in themselves (things not regarded as appearances). But in so doing it at the same time sets limits to itself, recognising that it cannot know these noumena through any of the categories, and that it must therefore think them only under the title of an unknown something” (A256/B312; underlining added).


The prohibition against determinate objects in the logical relation of transcendental affection keeps our characterization of affection in line with what Kant claims about the limited use the pure concepts of the understanding do have apart from their schematized application to sensible intuitions. For instance, at the end of the section on schematism, Kant claims,


Now there certainly does remain in the pure concepts of understanding, even after elimination of every sensible condition, a meaning; but it is purely logical, signifying only the bare unity of representations. The pure concepts can find no object, and so can acquire no meaning which might yield a concept of some object. (A147/B186)


The concepts, divorced from sensible intuitions, do have a function, but it is only a logical one that yields no knowledge of objects. That is exactly the functional role that the concept of cause, in its bare hypothetical judgment relation of ground to consequent, can be thought to play in transcendental affection. Kant makes similar remarks about the limited role of the pure concepts of the understanding at another point where he talks specifically about what the category of cause would amount to apart from its schematization:


If I omit from the concept of cause the time in which something follows upon something else in conformity with a rule, I should find in the pure category nothing further than that there is something from which we can conclude to the existence of something else... the concept would yield no indication how it applies to any object.” (A244/B302)


Again, this grounding of the existence of something in the existence of something else, without specifying those somethings as objects and without even specifying that the former “something” is distinct from the latter “something else”, is exactly what the logical relation of transcendental affection does.


Admittedly, this relation does not tell us much about anything at all, and we might feel disappointed that we could not within the limits of Kant’s Critical philosophy come up with anything more substantial to posit as the relation of transcendental affection than this. Still, keeping in mind that transcendental affection and empirical affection are the same relation considered in two different ways, we have succeeded in characterizing transcendental affection in a way that allows the realist to say that things-in-themselves affect the subject: a thing-in-itself, in some unknown relation to the subject-in-itself, is the ground for the subject's affected state, insofar as the subject and that state are considered apart from the conditions through which they are known. It is the dual aspect description of affection that allows us to characterize affection in one sense (the empirical sense) as a causal relation without at the same time making an illegitimate application of the categories to things-in-themselves when we characterize that relation transcendentally, as a bare logical, non-causal relation. In this way, we can restore to a respectable place in Kant’s view the transcendental substrate “behind the appearances,” the ground that momentarily threatened to give way beneath the realist’s feet.


Construing transcendental affection as the logical relation of the ground of things-in-themselves to the consequent of an affected subject certainly seems to be consistent with the numerous passages in the Dialectic in which Kant specifically characterizes things-in-themselves as the ground of appearances. For instance, in the section on the Paralogisms, Kant says,


... all that can be done is to indicate [how in a thinking subject outer intuition is possible] through the ascription of outer appearances to that transcendental object which is the cause of this species of representations, but of which we can have no knowledge whatsoever and of which we shall never acquire any concept. In all problems which may arise in the field of experience we treat these appearances as objects in themselves, without troubling ourselves about the primary ground of their possibility (as appearances). But to advance beyond these limits the concept of a transcendental object would be indispensably required (A393; emphasis added).


It is clear that the “primary ground of the possibility” of the appearances here is taken to be the transcendental object, as a thing-in-itself, that is responsible for those representations in the subject. Here are two more passages in which Kant characterizes things-in-themselves as the ground of appearances in a subject:


Neither the transcendental object which underlies outer appearances nor that which underlies inner intuition, is in itself either matter or a thinking being, but a ground (to us unknown) of the appearances... (A380).

If, on the other hand, appearances are not taken for more than they actually are; if they are viewed not as things in themselves, but merely as representations..., they must themselves have grounds which are not appearances. While the effects are to be found in the series of empirical conditions, the intelligible cause, together with its causality, is outside the series (A537/B565).


IV.     The Parallel Between Transcendental Affection and Free Action


This last passage, found in Kant’s discussion of the Third Antinomy, serves as an important link between the Critique’s doctrine of transcendental affection and Kant’s compatibilist stance toward determinism and free will. Considering the passage in conjunction with our account of transcendental affection suggests an original way in which we can coherently view the actions of an agent as both transcendentally free and at the same time empirically determined. This can be done because the effects of an agent’s actions can be compared favorably with the affection of subjects by objects.


Just as empirical objects can affect the senses of a subject in such a way that she ends up having particular mental representations of those objects, so an agent’s actions can be considered empirically in terms of the causal relationship between the agent and certain effects of her action. Considered in this way, the agent in acting as she does is situated somewhere in the deterministic “empirical series of conditions” mentioned in the above passage. Likewise, just as the empirically affecting objects, considered as things-in-themselves in a certain unknown relation to a subject-in-itself, are the logical ground for the affected mental state of the subject, considered apart from the conditions under which it is perceived, we can analogously consider an agent’s actions transcendentally. The agent, considered as a thing-in-itself and in some unknown relation to other things-in-themselves, is the ground for a certain state of objects that results from that person’s actions, where here that state is considered apart from the conditions under which it is perceived. As we can see, actions transcendentally considered yield us as little substantial information about the actions themselves as affection transcendentally considered yields us about the affection itself. Like transcendental affection, there may not be any one-to-one mapping from the empirically considered actors and acted-upon subjects to those various objects considered as things-in-themselves. Nevertheless, actions viewed in this way allow us to consider them as actions that are freely performed. An action considered apart from empirical conditions does not presuppose “something upon which it follows according to a rule,” as the Second Analogy dictates for empirical objects (A190), so we can think of an action in the transcendental sense as at least not empirically determined.


There is a certain parallel in the Critique, then, between actions and affection. Each can be characterized in terms of a causal relation between empirical objects/agents, but each can also be construed transcendentally as a logical relation of ground to consequent between things-in-themselves, about which we only determinately know that they are objects of experience considered apart from the conditions under which we sensibly perceive them. These objects so conceived are mere, indeterminate, possibly indistinct somethings underlying the appearances. We can say with Kant in the above quote from the Third Antinomy discussion that with respect to both actions of an agent and affection of a subject, “the intelligible cause, together with its causality,” considered transcendentally as the logical relation of ground to consequent, “is outside the series” of empirical relations, even though “the effects,” considered empirically, “are to be found in the series of empirical conditions” which includes the agent or affecting body empirically considered.


Of course, to be sure, Kant would not want to say that anything other than God or a rational human agent could act freely in the transcendental sense, contrary to what our account appears to allow with respect to non-human empirically affecting objects. Considered transcendentally, their affection of the subject would seem to be an act freely performed by those objects. However, we might be able to avoid ascribing freedom to non-human, empirically affecting objects as they are transcendentally considered if we take the transcendental ground for all non-human cases of empirical affection (that is, all cases of empirical affection where the affecting empirical object is not a rational human agent) to be God, who certainly can act freely. 9 Non-human empirical objects and their causal relationships on this account would interestingly enough be, as that transcendental ground empirically considered, various empirical manifestations of God. While a full development of this view, requiring an examination of Kant’s positive conception of God, would take us significantly beyond the First Critique and, as such, shall not be pursued in this paper, exploring such an approach is one promising and potentially fruitful way we might attempt to maintain and elaborate further the analogy between actions and affection.



V.    Conclusion


Admittedly, my account of transcendental affection brings us no closer to offering a conclusive justification of the existence of things-in-themselves than we were before we started. Transcendental affection’s logical relation of ground to consequent does not entail the existence of the ground itself any more than the claim that “If x exists, then y exists” implies the existence of x. Nevertheless, it is reasonable to expect that the phenomenalist will have equal difficulty defending his claim that things-in-themselves definitively do not exist; and insofar as I have given an account of transcendental affection that does not require an illegitimate application of the categories to things-in-themselves, I have removed one of the substantial obstacles in the way of the realist’s offering a coherent and epistemically respectable story about the existence of things-in-themselves. The realist can now more easily argue that a view of things-in-themselves and empirical objects as two aspects of subject-independent objects appearing to us offers us the best explanation of our sensory appearances. That dual aspect view would be considerably less attractive if we were forced to jettison altogether the notion of transcendental affection for being inconsistent with the rest of the Critical philosophy. The pitfall of the problematic character of transcendental affection thus overcome, the way is clear for the realist to offer a justification for believing things-in-themselves to exist, and thus to indemnify herself against the phenomenalist’s charge of dogmatism concerning that belief.


The phenomenalist, on the other hand, while perhaps holding a coherent ontology of his own, is at a substantial disadvantage when it comes to explaining Kant’s continual reference to things-in-themselves and the affection through which they are related to the subject and her appearances. Thus, as far as interpretations of Kant are concerned, it appears that the realist offers the more comprehensive of the two construals of the Critique of Pure Reason; and given that, as I have shown, the doctrine of transcendental affection, an important part of the realist view, can be accommodated in Kant’s Critical philosophy without contradiction, the realist interpretation of the First Critique comes off as the more promising of the two interpretations. 10




Allison, H. E. 1983. Kant’s Transcendental Idealism. New Haven: Yale University Press.

Kant, I. 1977. Prolegomena to Any Future Metaphysics. Trans. Paul Carus, rev. James W. Ellington. Indianapolis: Hackett.

Kant, I. 1929. Critique of Pure Reason. Trans. Norman Kemp Smith. New York: St. Martin’s Press.

Shaper, E. 1966. “The Kantian Thing-in-Itself as a Philosophical Fiction,” Philosophical Quarterly 16: 233-43.

Smith, N. K. 1992. A Commentary to Kant’s “Critique of Pure Reason. 2nd ed. Atlantic Highlands, NJ: Humanities Press International.   







1.  This view is advocated by Shaper (1966, 235-241).

2.  All references to the Critique of Pure Reason shall be to the Norman Kemp Smith 1929 translation.

3.  An example of such reasoning is the following two-syllogism concluding segment in a syllogistic series whose completion is thought of as being ultimately given by some tree T considered conceptually as an unconditioned thing-in-itself:                       


<earlier parts of the syllogistic series>


4.  Here, a “possible” appearance or mental act is either an actual appearance or mental act or an appearance or mental act satisfying the condition that if a person were in a particular place at a particular time, she could perceive that appearance or perform that act.

5.  This interpretation is defended by H. E. Allison in his 1983. The interpretation’s view of things-in-themselves is considered in most detail in Chap. 11 of that book (esp. pp. 240-241).

6.  This construal of what constitutes a thing-in-itself can be found, for instance, at A251-52, B306, and B308 in the First Critique.

7.  Kant does indeed at least give this reason for asserting the existence of things-in-themselves: cf. the passage quoted at the beginning of this section from the Prolegomena, and also B72: “Our mode of intuition is dependent upon the existence of the object, and is therefore possible only if the subject’s faculty of representation is affected by that object.”

8.  N. K. Smith, 1992 , p. 275, n. 1.

9.  This idea was suggested to me by Béatrice Longuenesse.

10.  Special thanks to Béatrice Longuenesse for helpful comments on earlier versions of this paper.



Copyright © 2004 Minerva

All rights are reserved, but fair and good faith use with full attribution may be made of this work for educational or scholarly purposes.

Chad Mohler is an Associate Professor of Philosophy and Religion at Truman State University, Kirksville, Missouri, U.S.A.


Mail to: Chad Mohler

Return to Minerva (Volume 8) Main Page                          Go to Top of This Page