The Paralogisms of the Soul

Kant discusses the nature and limits of our self-knowledge most extensively in the first Critique, in a section of the Transcendental Dialectic called the “Paralogisms of Pure Reason”. Here Kant is concerned to criticize the claims of what he calls “rational psychology”, and specifically, the claim that we can have substantive metaphysical knowledge of the nature of the subject, based purely on an analysis of the concept of the thinking self, or as Kant typically puts it, the <I think>.

I think is thus the sole text of rational psychology, from which it is to develop its entire wisdom…because the least empirical predicate would corrupt the rational purity and independence of the science from all experience. (A343/B401)

There are four “Paralogisms”. Each argument is presented as a syllogism, which consists of two premises and a conclusion. According to Kant, each argument is guilty of an equivocation on a term common to the premises, such that the argument is invalid. Kant’s aim, in his discussion of each Paralogism, is to diagnose the equivocation, and explain why the rational psychologist’s argument ultimately fails. In so doing Kant provides a great deal of information about his own views concerning the mind (See Ameriks (2000) for extensive discussion). The argument of the first Paralogism concerns our knowledge of the self as substance; the second, the simplicity of the self; the third, the numerical identity of the self; the fourth, knowledge of the self versus knowledge of things in space. We’ll take these arguments in turn.

Substantiality (A348-51/B410-11)

Kant presents the rationalist’s argument in the First Paralogism as follows:

  1. What cannot be thought otherwise than as subject does not exist otherwise than as subject, and is therefore substance.
  2. Now a thinking being, considered merely as such, cannot be thought otherwise than as subject.
  3. ∴ A thinking being also exists only as such a thing, i.e., as substance.

Kant’s presentation of the argument is rather compressed. In more explicit form we can put it as follows (see Proops (2010) and Rosefeldt (2017)):

  1. All entities that cannot be thought otherwise than as subjects are entities that cannot exist otherwise than as subjects, and therefore (by definition) are substances. (All M are P)
  2. All entities that are thinking beings (considered merely as such) are entities that cannot be thought otherwise than as subjects. (All S are M)
  3. ∴ All entities that are thinking beings (considered merely as such) are entities that cannot exist otherwise than as subjects, and therefore are substances (All S are P)

The relevant equivocation concerns the term that occupies the ’M’ place in the argument – viz. “entities that cannot be thought otherwise than as subjects”. Kant specifically locates the ambiguity in the use of the term “thought” [Das Denken], which he claims is taken in the first premise to concern an object in general, and thus something that could be given in a possible intuition. In the second premise the use of “thought” is supposed to apply only to a feature of thought and, thus, not to an object of a possible intuition (B411-12).

The major premise talks about a being that can be thought of in every respect, and consequently also as it might be given in intuition. But the minor premise talks about this being only insofar as it considers itself as a subject relatively only to thinking and the unity of consciousness, but not at the same time in relation to intuition, through which it is given as an object for thinking. Hence the conclusion is inferred per sophisma figurae dictionis, and hence through a fallacy. (B411)

The representation ‘I’ has the formal feature of non-predicability—there is no possible thought that features it as a predicate. But this fact is not sufficient for showing that the referent of <I> is a subsisting and substanding being. Why not? Because <I> can never be applied to such a being on the basis of a sensible intiuition, which could only present the subject as it appears and not as it is in itself. In particular, our sensible intuition in inner sense can present nothing that persists, so it cannot present a persisting soul or absolute subject either (eg. A350).

Hence, against the rational psychologist, Kant argues that we cannot make any legitimate inference from the conditions under which the representation <I> may be thought, or employed in a judgment, to the status of the ‘I’ as a metaphysical subject of properties. Kant makes this point explicit when he says,

the first syllogism of transcendental psychology imposes on us an only allegedly new insight when it passes off the constant logical subject of thinking as the cognition of a real subject of inherence, with which we do not and cannot have the least acquaintance, because consciousness is the one single thing that makes all representations into thoughts, and in which, therefore, as in the transcendental subject, our perceptions must be encountered; and apart from this logical significance of the I, we have no acquaintance with the subject in itself that grounds this I as a substratum, just as it grounds all thoughts. (A350)

Since, Kant denies that we have any intuition, empirical or otherwise, of ourselves as subjects, we cannot, merely in reflection on the conditions of thinking of ourselves using the first-person concept, come to have any knowledge concerning what we are. No amount of introspection or reflection on the content of the first-person concept <I> will yield such knowledge.

Simplicity (A351-61/B407-8)

Kant’s discussion of the proposed metaphysical simplicity of the subject largely depends on points he made in the previous Paralogism concerning its proposed substantiality. Kant articulate the Second Paralogism as follows:

  1. The subject whose action can never be regarded as the concurrence of many acting things, is simple. (All A is B)
  2. The self is such a subject. (C is A)
  3. ∴ The self is simple. (C is B)

Here the equivocation concerns the notion of a “subject”. Kant’s point, as with the previous Paralogism, is that, from the fact that one’s first-person representation of the self is always a grammatical or logical subject, nothing follows concerning the metaphysical status of the referent of that representation.

Of perhaps greater interest in this discussion of the Paralogism of simplicity is Kant’s analysis of what he calls (in the A-edition) the “Achilles of all dialectical inferences” (A351). According to the Achilles argument, the soul or mind is known to be a simple unitary substance because only such a substance could think unitary thoughts. Call this the “unity claim” (see Brook (1997)):

if a multiplicity of representations are to form a single representation, they must be contained in the absolute unity of the thinking substance. (A352)

Against UC, Kant argues that we have no reason to think that the structure of a thought, as a complex of representations, isn’t mirrored in the complex structure of an entity which thinks the thought. UC is not analytic, which is to say that there is no contradiction entailed by its negation. UC also fails to be a synthetic a priori claim, since it follows neither from the nature of the forms of intuition, nor from the categories. Hence UC could only be shown to be true empirically, and since we do not have any empirical intuition of the self, we have no basis for thinking that UC must be true (A353).

Kant here makes a point familiar from contemporary functionalist accounts of the mind (see Meerbote (1991); Brook (1997)). Our mental functions, including the unity of conscious thought, are consistent with a variety of different possible media in which the functions are realized. Kant’s point is that there is no contradiction in thinking that a plurality of substances might succeed in generating a single unified thought. Hence we cannot know that the mind is such that it must be simple in nature.

Numerical Identity (A361-66/B408)

Kant articulates the Third Paralogism as follows:

  1. What is conscious of the numerical identity of its Self in different times, is to that extent a person. (All C is P)
  2. Now the soul is conscious of the numerical identity of its Self in different times. (S is C)
  3. ∴ The soul is a person. (S is P)

The interest taken in establishing the personality of the soul or mind by the rational psychologists, stems from the importance of proving that not only would the mind persist after the destruction of its body, but also that this mind would be the same person, and not just some sort of bare consciousness or worse (e.g. existing only as a “bare monad”).

Kant here makes two main points. First, the rational psychologist cannot infer from the sameness of the first-person representation (the “I think”), across applications of it in judgment, to any conclusion concerning the sameness of the metaphysical subject referred to by that representation. Kant is thus once again making a functionalist point that the medium in which a series of representational states inheres may change over time, and there is no contradiction in conceiving of a series of representations as being transferred from one substance to another (A363-4, note).

Second, Kant argues that we can be confident of the soul’s possession of personality in virtue of the persistence of apperception. The relevant notion of “personality” here is one concerning the contrast between a rational being and an animal. While the persistence of apperception (i.e. the persistence of the “I think” as being able to attach to all of one’s representations) does not provide an apperceiving subject with any insight into the true metaphysical nature of the mind, it does provide evidence of the soul’s possession of an understanding. Animals, by contrast, do not possess an understanding but, at best (according to Kant), only an analogue thereof. As Kant says in the Anthropology,

That man can have the I among his representations elevates him infinitely above all other living beings on earth. He is thereby a person […] that is, by rank and worth a completely distinct being from things that are the same as reason-less animals with which one can do as one pleases. (An 7:127, §1)

Hence, so long as a soul possesses the capacity for apperception, it will signal the possession of an understanding, and thus serves to distinguish the human soul from that of an animal (see Dyck (2010), 120).

Relation to Objects in Space (A366-80/B409)

Finally, the Fourth Paralogism (here I primarily discuss the A-edition version) concerns the relation between our awareness of our own minds and our awareness of other objects distinct from ourselves, and thus as located in space. Kant describes the Fourth Paralogism as follows:

  1. What can be only causally inferred is never certain. (All I is not C)
  2. The existence outer objects can only be causally inferred, not immediately perceived by us. (O is I)
  3. ∴ We can never be certain of the existence of outer objects. (O is not C)

Kant locates the damaging ambiguity in the conception of “outer” objects. This is puzzling since it doesn’t play the relevant role as middle term in the syllogism. But Kant is quite clear that this is where the ambiguity lies and distinguishes between two distinct senses of the “outer” or “external”:

Trancendentally Outer/External:
a seperate existence, in and of itself
Empirically Outer/External:
an existence in space

Kant’s point here is that all appearances in space are empirically external to the subject who perceives or thinks about them, while nevertheless being transcendentally internal, in that such spatial appearances do not have an entirely independent metaphysical nature, since their spatial features depend at least in part on our forms of intuition.

Kant then uses this distinction not only to argue against the assumption of the rational psychologist that the mind is better known than any object in space (this a claim famously argued by Descartes), but also against those forms of external world skepticism championed by Descartes and Berkeley. Kant identifies Berkeley with what he calls “dogmatic idealism” and Descartes with what he calls “problematic idealism” (A377).

Problematic Idealism:
we cannot be certain of the existence of any material body
Dogmatic Idealism:
we can be certain that no material body exists – the notion of a body is self-contradictory

Kant brings two arguments to bear against the rational psychologist’s assumption about the immediacy of our self-knowledge, as well as these two forms of skepticism, with mixed results. The two arguments are (what I am calling) the arguments from “immediacy” and “imagination”. We’ll take these in turn.

i. The Immediacy Argument

In an extended passage in the Fourth Paralogism (A370-1) Kant makes the following argument:

external objects (bodies) are merely appearances, hence also nothing other than a species of my representations, whose objects are something only through these representations, but are nothing separated from them. Thus external things exist as well as my self, and indeed both exist on the immediate testimony of my self-consciousness, only with this difference: the representation of my Self, as the thinking subject, is related merely to inner sense, but the representations that designate extended beings are also related to outer sense. I am no more necessitated to draw inferences in respect of the reality of external objects than I am in regard to the reality of the objects of my inner sense (my thoughts), for in both cases they are nothing but representations, the immediate perception (consciousness) of which is at the same time a sufficient proof of their reality. (A370-1)

I take the argument here to be as follows:

  1. Rational Psychology (RP) privileges awareness of the subject and its states over awareness of non-subjective states.
  2. But transcendental idealism entails that we are aware of both subjective and objective states, as they appear, in the same way – viz. via a form of intuition.
  3. So either both kinds of awareness are immediate or they are both mediate.
  4. Since awareness of subjective states is obviously immediate then awareness of objective states must also be immediate.
  5. ∴ We are immediately aware of the states or properties of physical objects.

Here Kant displays what he takes to be an advantage of his Transcendental Idealism. Since both inner and outer sense depend on intuition, there is nothing special about inner intuition that privileges it over outer intuition. Both are, as intuitions, immediate presentations of objects (at least as they appear). Unfortunately, Kant never makes clear what he means by the term “immediate” [unmittelbar]. This issue is much contested (see Smit (2000)). At the very least, he means to signal that our awareness in intuition is not mediated by any explicit or conscious inference, as when he says that the transcendental idealist “grants to matter, as appearance, a reality which need not be inferred, but is immediately perceived” (A371).

It is not obvious that an external world skeptic would find this argument convincing, since part of the grip of such skepticism on us relies on the (at least initially) convincing point that things could seem to one just as they currently are, even if there really is no external world causing one’s experiences. This may just beg the question against Kant (particularly premise (2) of the above argument). And certainly Kant seems to think that his arguments for the existence of the pure intuitions of space and time in the Transcendental Aesthetic lend some weight to his position here. Thus Kant is not so much arguing for Transcendental Idealism here as explaining some of the further benefits that come when the position is adopted. He does, however, present at least one further argument against the skeptical objection articulated above – viz. the argument from imagination, to which we’ll now turn.

ii. The Argument from Imagination

Kant’s attempt to respond to the skeptical worry that things might appear to be outside us while not actually existing outside us appeals to the role that imagination would have to play to make such a possibility plausible (A373-4; cf. Anthropology, 7:167-8).

This material or real entity, however, this Something that is to be intuited in space, necessarily presupposes perception, and it cannot be invented by any power of imagination or produced independently of perception, which indicates the reality of something in space. Thus sensation is that which designates a reality in space and time, according to whether it is related to the one or the other mode of sensible intuition.

What follows is my reconstruction of this argument.

  1. If problematic idealism is correct then it is possible for one to have never perceived any spatial object but only to have imagined doing so.
  2. But imagination cannot fabricate – it can only re-fabricate.
  3. So, if one has sensory experience of outer spatial objects, then one must have had at least one successful perception of an external spatial object.
  4. ∴ It is certain that an extended spatial world exists.

Kant’s idea here is that the imagination is too limited to generate the various qualities that we experience as instantiated in external physical objects. Hence, it would not be possible to simply imagine an external physical world without having been originally exposed to the qualities instantiated in the physical world, ergo the physical world must exist. Even Descartes seems to agree with this, noting in Meditation I that “[certain simple kinds of qualities] are as it were the real colours from which we form all the images of things, whether true or false, that occur in our thought” (Descartes (1984), 13-14). Though Descartes goes on to doubt our capacity to know even such basic qualities given the possible existence of an evil deceiver, it is notable that the deceiver must be something other than ourselves, in order to account for all the richness and variety of what we experience (however, see Meditation VI (Descartes (1984), 54), where Descartes wonders whether there could be some hidden faculty in ourselves producing all of our ideas).

Unfortunately, it isn’t clear that the argument from imagination gets Kant a conclusion of the desired strength, for all that it shows (if it shows anything) is that there was at one time a physical world, which affected one’s senses and provided the material for one’s sense experiences. This might be enough to show that one has not always been radically deceived, but it is not enough to show that one is not currently being radically deceived. Even worse, it isn’t even clear that we need a physical world to generate the requisite material for the imagination. Perhaps all that is needed is something distinct from the subject, which is capable of generating in it the requisite sensory experiences, whether or not they are veridical. This conclusion is thus compatible with that “something” being Descartes’s evil demon, or in contemporary epistemology, with the subject’s being a brain in a vat. Hence, it is not obvious that Kant’s argument succeeds in refuting the skeptic, or to the extent that it does, that it shows that we know there is a physical world, as opposed merely to the existence of something distinct from the subject.

Lessons of the Paralogisms

Beyond the specific arguments of the Paralogisms and their conclusions, they present us with two central tenets of Kant’s conception of the mind. First, that we cannot move from claims concerning the character or role of the first-person representation <I> to claims concerning the nature of the referent of that representation. This is a key part of his criticism of rational psychology. Second, that we do not have privileged access to our self as compared with things outside us. Both the self (or its states) and external objects are on par with respect to intuition. This also means that we only have access to ourselves as we appear, and not as we fundamentally, metaphysically, are (cf. B157). Hence, according to Kant, our self-awareness, just as much as our awareness of anything distinct from ourselves, is conditioned by our sensibility. Our intellectual access to ourselves in apperception, Kant argues, does not reveal anything about our metaphysical nature, in the sense of the kind of thing that must exist to realize the various cognitive powers that Kant describes as characteristic of a being capable of apperception (e.g. a spontaneous understanding or intellect).


Allison, Henry E. 2004. Kant’s Transcendental Idealism: Revised and Enlarged. New Haven: Yale University Press.

Ameriks, Karl. 2000. Kant’s Theory of Mind: An Analysis of the Paralogisms of Pure Reason. New York: Oxford University Press.

Brook, Andrew. 1997. Kant and the Mind. Cambridge: Cambridge University Press.

Descartes, René. 1984. The Philosophical Writings of Descartes: The Meditations with Objections and Replies. Edited by John Cottingham, Robert Stoothoff, and Dugald Murdoch. Vol. 2. Cambridge: Cambridge University Press.

Dyck, Corey W. 2010. “The Aeneas Argument: Personality and Immortality in Kant’s Third Paralogism.” In, edited by Dietmar Heidemann, 95–122.

———. 2014. Kant and Rational Psychology. Oxford: Oxford University Press.

Horstmann, Rolf-Peter. 1993. “Kants Paralogismen.” Kant-Studien 84 (4): 408–25.

Kitcher, Patricia. 1982. “Kant on Self-Identity.” The Philosophical Review 91 (1): 41–72.

———. 2011. “‘I-Think’ as the Destroyer of Rational Psychology.” In, 180–200. New York: Oxford University Press.

Longuenesse, Beatrice. 2017. I, Me, Mine: Back to Kant, and Back Again. New York, NY: Oxford University Press.

Meerbote, Ralf. 1991. “Kant’s Functionalism.” In, edited by J-C Smith, 161–87. Dordrecht: Kluwer Academic Publishers.

Proops, Ian. 2010. “Kant’s First Paralogism.” The Philosophical Review 119 (4): 449.

Rosefeldt, Tobias. 2017. “Subjects of Kant’s First Paralogism.” In Kant and the Mind: Perception, Reason, and the Self, edited by Anil Gomes and Andrew Stephenson, 221–44. Oxford: Oxford University Press.

Smit, Houston. 2000. “Kant on Marks and the Immediacy of Intuition.” The Philosophical Review 109 (2): 235–66.

Van Cleve, James. 1999. Problems from Kant. Oxford: Oxford University Press.

Wuerth, Julian. 2010. “The Paralogisms of Pure Reason.” In The Cambridge Companion to Kant’s Critique of Pure Reason, edited by Paul Guyer, 210–44. Cambridge: Cambridge University Press.