The Paralogisms and Kant's View of the Self

. . . io dismento nostra vanitate
trattando l'ombre come cosa salda

--Dante Alighieri, La Divina Commedia, Purgatorio, Canto XXI


In his ``Paralogisms of pure reason'' (Chapter I of Book II of Division II of Part II of the Transcendental Doctrine of Elements in the massively hierarchical Critique of Pure Reason), Immanuel Kant returns to a topic discussed earlier in the Critique: the ontological status of the self as thinking being (that is, the soul). Kant disproves a number of dogmatic assertions about the nature of the self, claiming them all to be due to a confusion between representation and thing in itself. In the case of the first three, we are confusing our representation of the soul (by nature non-predicative, simple, and identical) with the soul itself. In the case of the last, we confuse appearances with the things in themselves, resulting in a false belief that external objects are of doubtful reality.

There are two versions of the paralogisms. The version contained in the A edition of the Critique is significantly longer than that in the B edition. In the B Paralogisms of Pure Reason, Kant focuses his attention on the analyticity (and thus empirical unsuitability) of the propositions of the paralogisms. The A edition, on the other hand, provide a more varied assault on the paralogisms; in particular, the discussion of the fourth paralogism is something of a precursor of the B edition's ``Refutation of Idealism''. We shall discuss further the connections and differences between the two versions of this Chapter at the appropriate locations below.

The self as subject

The first paralogism states quite simply that ``I, as thinking being (soul), am substance.'' (A348). In his criticism of the paralogism, Kant employs a technique he will go on to use in the following paralogisms, especially in the B Critique. He grants that the soul fills the logical requirements of the concept `substance': ``I cannot be used as the determination of another thing'' (A349) . However, according to Kant, this does not prove that the soul persists. For ``We must rather ground the persistence of a given object on experience if we would apply to that object the empirically usable concept of a substance'' (A349). That is to say, the pure category of substance can prove nothing of ``objective significance'' (A348) without an intuition grounding it. Since, as Kant has demonstrated in the Transcendental Deduction, there is no intuition of the transcendental self (restated at A350), such a self cannot be said to persist.

In the B paralogisms, Kant changes the definition of substance. Whereas in the A edition he allowed that the soul might be ``a substance only in the idea but not in reality'' (A351), here he maintains that the non-predication of the self ``does not signify that I as object am for myself a self-subsisting being or substance'' (B407). In the A edition Kant attacks the deduction I am substance; therefore I endure, here he attacks I am always subject and never predicate; therefore I am substance. Thus, by revising the first paralogism, Kant returns substance to its proper use: a concept applicable only to appearances, and not to transcendental objects.

The self as simple

The second paralogism draws from Kant a longer critique (at least in the A edition). It states that ``the soul, or the thinking I, is such a thing [viz. a simple]'' (A351). To Kant, this principle ``seems to withstand even the sharpest testing and the greatest scrutiny of inquiry'' (A351).

In a move paralleling the first paralogism, Kant makes the distinction between simplicity of representation and the simplicity of a thing. Thus ``I am simple signifies no more than that . . . it [viz. the representation I] is an absolute (though merely logical) unity'' (A355). To the psychological proof of the simplicity of the self, Kant responds that the I ``signifies only a Something in general (a transcendental subject), the representation of which must be of course simple . . . certainly nothing can be represented as more simple than that which is represented through the concept of a mere Something''. Here we see the transcendental self as it appears in the B deduction: it is neither appearance nor thing in itself (B158), but a mere placeholder---an idea, under Kant's strict definition of that term.

The second part of Kant's refutation of this paralogism asks: what if the self is simple? For ``the simple nature of the soul is of unique value only insofar as through it I distinguish the subject from all matter, and consequently except it from the perishability to which matter is always subjected'' (A356). As in the first paralogism, where Kant shows that the self is a substance only in an objectively useless sort of way, here he shows that the ``simplicity'' of the self is just as useless. Corporeal matter as we know it must be regarded as appearance and not as thing-in-itself; as the Transcendental Aesthetic showed, the concepts of extension and impenetrability, which define corporeality, are applicable only to appearances. However, the transcendental self, as demonstrated in the Deduction, is not an appearance; we must thus compare it to the ``Something that grounds outer appearances and affects our sense so that it receives the representations of shape, matter, space, etc.'' (A358). However, we can know nothing about this ``transcendental object'': since it is not an object for us, we cannot know that it is not in fact ``the subject of thoughts'' (A358). Thus, the simplicity of the self could not even prove that soul is inherently different from matter, and thus could not prove that it is not corruptible.

In the B version of Kant's criticism of the second paralogism, things are much simpler: ten pages have been reduced to just half a page. In part this is because much of the content has been moved to the (also thoroughly revised) Transcendental Deduction. However, part of this change can also be attributed to the distinction in the B paralogisms between subject and substance. Given this change, Kant no longer views the second paralogism as ``the Achilles of all the dialectical inferences of the pure doctrine of the soul'' (A351). Instead, his argument is a quite simple one: while the I is a simple subject (which claim ``lies already in the concept of thinking, and is consequently an analytic proposition'' (B408), it is not thereby a simple substance. Indeed, the claim that the soul is simple substance would be a synthetic a priori proposition---and, since it does not pertain to possible experience, untrue (or perhaps completely nonsensical). Thus we cannot show the soul to be ``except[ed] from the perishability to which matter is always subjected'' (A356).

The self as numerically identical

If the self cannot be shown to be simple, can it at least be shown to be numerically identical? In his criticism of the third paralogism, Kant argues that it cannot. Of course, I find myself to be numerically identical: the proposition that I am numerically identical ``really says no more than that in the whole time in which I am conscious of myself, I am conscious of this time as belonging to the unity of my Self'' (A362). Because the self is (according to the Deduction) the pure apperception uniting all my intuitions, it is to me by definition a single, numerically identical, thing. However, another observer ``will still not infer the objective persistence of my Self'' (A363), because my time---the time which is the form of my intuition, and through which I am identical---is not the same as ``the time in which the observer posits me'' (A363).

Kant provides, in the footnote on pp. A363--364, an interesting analogy. Imagine the collision of two elastic balls, one moving and one at rest. Upon the collision, the moving ball imparts to the other ``its entire motion, hence its whole state'' (A363). Likewise is it possible that, if indeed there are substances that think, the self (the I) moves from one substance to another. Since this would involve the transfer of all the I's representations---including that of its Self---the I would be completely unaware of this flux. Kant poses a similar argument on A364 proper: ``we cannot judge even from our own consciousness whether as soul we are persisting or not, because we ascribe to our identical self only that of which we are conscious''. Something (such as the nature of our supposed substance) of which we are not conscious is not proved identical by the identity of our consciousness.

As with the first two paralogisms, Kant's B Critique views this principle as resulting from a confusion between analytic and synthetic propositions. Likewise, he makes a similar distinction between representation and (metaphysical) substance: I am identical as concerns ``the subject, of which I can become conscious in every representation'' (B408), but not the person, ``the consciousness of the identity of its own substance as a thinking being in all changes of state'' (B409). The identity of the first comes from the very definitions of `subject' and `consciousness', and is ``hence an analytic proposition''. The identity of the second, the person, would be a synthetic proposition, and is therefore (since it does not concern an object of experience) outside the scope of reason.

The ideality of outer relation, crypto-Berkeleianism, and a resolution

The fourth paralogism is perhaps the trickiest; Kant certainly stumbles here. The paralogism is that ``objects of outer sense'' are of ``doubtful existence'' (A366--367). The existence of objects of outer experience is not perceived, since perception ``is properly only a determination of apperception''; thus it is only inferred from changes in my inner state (which I presumably can perceive). Thus does Kant describe idealism.

Kant maintains himself to be, of course, a transcendental idealist. That is, he holds that appearances ``are all together to be regarded as representations and not as things in themselves'' (A369). Such a system, however, lends itself easily to what he terms empirical realism. Kant can claim that matter truly, and undoubtedly, exists; this is not because he claims to have transcendent knowledge of matter, but rather because matter is a product of the self-consciousness. Thus the existence of objects may ``be proved in the same way as the existence of myself as a thinking being'' (A370)---that is to say, they can be shown to exist because they are necessary representations.

Kant sees his beliefs here as being substantially different from those of traditional idealists. He claims that ``all those psychologists who cling to empirical idealism are transcendental realists''---and here he falls into error. Kant was so concerned with the ``Copernican revolution'' of his transcendental idealism that he failed to see that its foundations had been gestating for some time. In particular, Bishop Berkeley, one of the dogmatic idealists that Kant so maligns, was something of a transcendental idealist. From A Treatise Concerning the Principles of Human Knowledge:

35. I do not argue against the existence of any one thing that we can apprehend either by sense or reflection. That the things I see with my eyes and touch with my hands do exist, really exist, I make not the least question. The only thing whose existence we deny is that which Philosophers call Matter or corporeal substance. And in doing of this there is no damage done to the rest of mankind, who, I dare say, will never miss it. The Atheist indeed will want the colour of an empty name to support his impiety; and the Philosophers may possibly find they have lost a great handle for trifling and disputation.
The primary difference between Berkeley and Kant seems to be the definitions of `substance' and of `existence' (as the latter pertains to matter). Kant holds that matter exists only if speak of it as appearance---that is, as representation. Berkeley holds that matter does not exist, because it is supposedly an unthinking substance, which for him cannot exist separate from mind. The gap in meaning between the two, whatever the words, is less than Kant would like to admit. No doubt, if Berkeley were given Kant's definitions of `matter' and `substance', he would agree with Kant's claims regarding their transcendental ideality. And, as the above-quoted passage indicates, he would probably accord with Kant in their empirical reality as appearance.

Kant no doubt realised sometime between 1780 and 1786 that the fourth paralogism read as a crypto-Berkeleian account of metaphysics. Thus Kant introduced in the B edition of the Critique a new ``Refutation of Idealism'' (B274--279). There, Kant claims that the representation of the self would not even be possible without external objects. Likewise, the criticism of the fourth paralogism states that the distinction between my self and outer objects does not even prove ``whether this consciousness of myself would even be possible without things outside me through which representations are given to me'' (B409). In fact, in the Refutation, he even claims that perception is possible ``only through a thing outside me and not through the mere representation of a thing outside me'' (B275).

How does this last statement reconcile with Kant's transcendental idealism? Transcendental idealism is, according to Kant, the doctrine that ``all objects of an experience possible for us are nothing but appearances, i.e, mere representations'' (B518). Yet in the Refutation of Idealism he argues for the necessity of ``a thing outside me'', rather than ``the mere representation of a thing outside me''. Thus pure reason has proven the existence of things that are not representation. Is this really all that bad? Pure reason, after all, proved the existence of the transcendental self (in the Transcendental Deduction), without ascribing it any properties. Likewise, the Refutation proves what seemed to underlie as an assumption the A Critique: that there is something metaphysically outside ourselves which grounds our representation. We cannot know anything about this something---merely that it exists. It is thus to Kant an important idea which refutes Berkeleian dogmatic idealism and provides the means for refuting Cartesian sceptical idealism.


The paralogisms, or at least the first three, provide a mostly consistent view of the self---one which expands on that developed in the Transcendental Deduction. Kant views the self (the transcendental self, that is) as Something which exists, but to which no properties can be ascribed. It is neither appearance nor thing in itself, but merely a placeholder for that which underlies our perception. In the B edition, the fourth paralogism refers to the Refutation of Idealism, which proves the existence of a similar Something behind our representations of objects. Unlike Locke's ``something I know not what'', these Somethings are completely unknown to us: they do not exist before intuition, and can thus not provide us with cognition. Thus Kant maintains substance as appearance, while making two important traditional substances (the mind-independent object and the self) into unknowable but nonetheless existent things.

Log in or registerto write something here or to contact authors.