‘“I think” is the Sole Text of Rational Psychology’: Comments on Ian Proops’s The Fiery Test of Critique

Abstract

I focus on two main points in Ian Proops’s reading of Kant’s Paralogisms of Pure Reason: the structure of the paralogisms in the A edition of the Critique of Pure Reason, and the changes in Kant’s exposition of the paralogisms from A to B. I agree with Proops that there are defects in the A exposition and that Kant attempted to correct those defects in B. But I argue that Proops fails to give its due to what remains fundamental in both editions: Kant’s criticism of the rational psychologist’s confusion between the subjective (albeit universally subjective) standpoint thinkers have on themselves just in virtue of thinking, and the objective, metaphysical standpoint on a thinking thing. In short, Proops fails to give sufficient attention to Kant’s opening statement in the Paralogisms of Pure Reason: ‘“I think” is the sole text of rational psychology’.

I first want to express my admiration for Ian Proops’s book. He describes his method as combining a ‘contextualist’ approach, placing Kant’s view in its own historical context; and a ‘reconstructive’ approach, making use of contemporary logical tools to analyse and assess Kant’s arguments. He makes superlative use of both methods.

My comments will focus on Proops’s discussion of the Paralogisms of Pure Reason. I will give a brief overview of some central agreements and disagreements and discuss in more detail two points raised by Proops: the structure of the paralogisms in A and the changes from the A to the B edition of Kant’s Paralogisms.

1. Overview

In the Paralogisms of Pure Reason, Kant refutes four arguments he attributes to rationalist metaphysicians, purporting to prove that we, as thinking beings considered merely as such, are souls: thinking substances, simple, conscious of our own identity at different times, and conscious that our existence as souls is distinct from that of our body. Those features of the soul, according to the rationalist view reconstructed by Kant, justify the claim that the soul is immaterial and immortal. Kant intends to show that those arguments, while apparently valid, turn out to be fallacies ‘due to form’: paralogisms. Kant’s official characterisation of the ‘fallacy due to form’ proper to paralogisms is that what passes for a middle term in the apparently valid syllogism does not in fact count as a middle term. Once disambiguated, it becomes clear that what passed for a middle term is not the same term in the major and the minor premise. Kant calls the fallacy a sophisma figurae dictionis (a ‘sophism of a figure of speech’). Proops shows, however, that not all the fallacies denounced by Kant in the Paralogisms obey this model. Some of them (at least one in A and one in B) correspond, rather than to the model of a sophisma figurae dictionis, to that of what Kant’s Logic Lectures call an ignoratio elenchi, in which the fallacy does not consist in an equivocation on the middle term, but rather in the fact that the proponent of the inference gives its conclusion a stronger meaning than is supported by the premises. I agree with Proops in one case (that of the first paralogism in A). I do not agree in the other (the second paralogism in B). I will discuss both cases.

At the end of the day, the important question is: does Kant retain anything at all of the rationalist view of the soul as it is supposed to follow from the inferences Kant reconstructs and shows to be ‘false due to form’? I agree again with Proops that in both editions of the Critique, once the fallacies ‘due to form’ have been revealed, the corrected conclusion is radically deflationary with respect to the rationalist claims. This is true for both the A and the B edition paralogisms even while Kant’s reconstructed argument is not the same in both editions.

Proops makes one exception to the deflationary character of the conclusion: the second paralogism in B, the paralogism of simplicity. There, Proops maintains, Kant admits the ‘nervus probandi’ of the rationalist position (‘A thought can only be the effect of the absolute unity of the thinking being’: A353, cited by Proops p.142) ¹ and thus allows a conclusion more amenable to the rationalist position than he did in A: the soul is an entity that is simple, albeit not the conclusion that it is a simple substance.

Setting aside the latter case, I agree with Proops’s assessment of the deflationary character of the conclusions Kant allows once he has disambiguated the premises. ² I also agree with him that Kant came to realise that there was a fundamental flaw, indeed several fundamental flaws in his reconstruction of the rationalist fallacy in A. But I do not read the flaws the way Proops does.

Kant, according to Proops, offers misleading explanations of his own reconstruction of the rationalist arguments. Proops’s complaint concerns two main points: Kant’s account of the general structure the paralogisms in the A edition of the Critique, and Kant’s account of the changes from A to B.

I want to dispute Proops’s assessment on both points. I will consider them in the next two sections, starting with the structure of the A paralogisms (section 2), and continuing with the purported differences between A and B (section 3). I will submit that Proops’s complaints stem from his failing to give its full due to Kant’s statement: ‘“I think” is the sole text of rational psychology’ (A343/B401).

2. The structure of the paralogisms in A

Kant offers the following recapitulation of the structure of all four paralogisms in A:

If one wants to give a logical title to the paralogisms in the dialectical syllogisms of the rational doctrine of the soul, insofar as they have correct premises, then it can count as a sophisma figurae dictionis, in which *the major premise makes a merely transcendental use of the category in regard to its condition, but in which the minor premise and the conclusion, in respect of the soul that is subsumed under this condition, make an empirical use of the same category.* Thus the concept of substance in the paralogism [of substantiality] is a pure intellectual concept which in the absence of conditions of sensible intuition is merely of transcendental use, i.e., of no use at all. But in the minor premise the very same concept is applied to the object of inner experience, yet without previously establishing it in concreto and grounding the condition of its application, namely persistence; and hence an empirical, although unreliable use is being made of it. (A402-3. Bold letters between asterisks flag the part of the quote Proops calls upon in his own analysis. See below)

Proops quotes part of the first sentence. He offers a schematic representation of the structure Kant describes and concludes (p. 138): ‘This form fits none of the paralogisms in either edition’. (Proops’s schema is discussed below, 2.2.)

I think the structure Kant describes does fit at least the first three paralogisms. I will focus on the first but also show how the account I propose can be extended to the second and third.

2.1 Kant’s formulation of the paralogism

That, the representation of which is the absolute subject of our judgements, and hence cannot be used as the determination of another thing, is substance.

I, as a thinking being, am the absolute subject of all my possible judgements, and this representation of myself cannot be used as the predicate of other things.

Thus I, as a thinking being (soul) am substance. (A348. Bold letters are Kant’s emphasis)

2.2 Proops’s cleaned-up formulation

Any entity whose representation cannot be used as a determination of another thing is a substance.

The thinking I is an entity whose representation cannot be used as a determination of another thing.

Therefore, the thinking I is a substance. (p. 127)

It seems clear that the syllogism is valid. Its only defect is that, in Proops’s words, ‘its conclusion is simply too weak to advance the rational psychologist’s project’ (p. 128). All it shows is that ‘the thinking I’ is what Proops calls ‘substance₀’, a far cry from the ‘substance₁’, the absolute metaphysical subject of inherence which the rationalist wants to claim the soul is proved to be. ³ But this makes the ‘fallacy due to form’ an ignoratio elenchi, not a sophisma figurae dictionis. Moreover, so understood, the first paralogism in A does not seem to fit the description given by Kant at A402-3, which Proops schematises as follows (p. 138):

Major premise: All M^T are P^T

Minor premise: The I, as a thinking being, is M^E

Conclusion: Therefore, the I, as thinking, is P^E

This indeed does not seem to fit the first paralogism in the cleaned-up formulation Proops gave of it, which I have granted. In that presentation, there was a middle term: the subject of the major premise was the predicate of the minor premise – substance_0. The logical fallacy was only that the metaphysician ‘extended’ the conclusion by giving the term ‘substance’ the meaning, not of substance₀, but of substance₁.

But I submit that a finer-grained analysis of the first paralogism does fit Kant’s description and moreover allows us to see that the first three paralogisms in A also fit that description.

2.3 A closer look at A402-3

According to A402-3, cited above, in the major premise ‘a merely transcendental use, namely no use at all’ is made of the category. The explanation that immediately followed Kant’s statement of the first paralogism helps clarify Kant’s meaning:

We have shown in the analytical part of the Transcendental Logic that pure categories (and among them also the category of substance) have in themselves no objective significance at all unless an intuition is subsumed under them, to the manifold of which they can be applied as functions of synthetic unity. Without that they are merely functions of a judgment [Funktionen eines Urteils] without content. (A348-9) ⁴

The concept of ‘That, whose representation is the absolute subject of our judgments …’ in the major premise of the first paralogism is just such a concept of substance, reduced to the concept of something whose representation can only be used as subject and not as predicate of something else – without any reference at all to an intuition subsumed under it, much less synthesised in conformity to it. It looks like the use of the category in the major premise does fit Kant’s description at A402-3.

What about the minor premise? There, Kant says, an ‘empirical use’ is made of ‘the same category’ or of ‘the very same concept’, that is, presumably, the concept of ‘that, whose representation is the absolute subject of our judgments and cannot be the predicate of something else’, which is definitional of ‘substance’. What can it mean to say that ‘an empirical use is made’ of it? We might think it means that it is used in application to an empirically given object and thus, under the condition of the object’s being experienced as permanent. ⁵ But that cannot be what the application to ‘the thinking I’ is. ‘The thinking I’ is what is represented by the first-person pronoun and concept in the sentence and proposition ‘I think’.

At the beginning of the chapter on the Paralogisms of Pure Reason, Kant wrote:

“I think” is the sole text of rational psychology, from which it is to develop its entire wisdom. (A343/B401)

From this we may conclude that what Kant calls ‘the thinking I’ is what is represented by ‘I’ in the proposition ‘I think’, and nothing else. This is what makes rational psychology rational as opposed to empirical psychology. Nevertheless, we do have, Kant also says:

…an inner experience of this proposition, … without any particular distinction or empirical determination being given in it. (A400-1/B342-3)

Proops offers the following comment:

Kant’s idea would seem to be that the ‘I think’ furnishes us with nothing beyond some highly indeterminate experience, namely: the awareness of inner perception or inner experience as such. (p. 83)

I agree. And I submit that, in the minor premise, it is in relation to this ‘highly indeterminate experience’ that the use of the category is described as empirical. But it is all too easy to mistake this minimal empirical support of the proposition ‘I think’ for the presentation in intuition of a determinate empirical object. Then one slides to the conclusion: ‘therefore, the thinking I is substance’, where one has included in the notion of substance the condition for the application of the concept of substance which was missing in both the major and the minor premise: the experience of permanence. This is explained in the second part of A402-3, cited at the beginning of the present section.

To recapitulate: in the major premise, a ‘merely transcendental use, namely no use at all’ is made of the concept of substance, reduced to the concept <something whose representation cannot be the predicate of something else (in judgement)>. In the minor premise, an empirical use is made of the very same intellectual concept, namely, a use in application to something that has a minimal empirical basis, but a use that is unzulässig: unreliable or perhaps better, improper, because what is empirical in this case (the mere perception of thinking expressed by ‘I think’) provides no objective determination of what is thereby presented as existing, because it offers no determinate temporal intuition, much less an experience of permanence. The latter is illegitimately assumed if one interprets the conclusion: ‘I, as a thinking being (soul) am substance’ as meaning that I ‘endure’, and ‘neither arise nor perish’ (A349).

This still does not make of the paralogism a sophisma figurae dictionis. It makes it, at most, an ignoratio elenchi: the conclusion, if it includes in the concept of ‘substance’ the condition of permanence and thus the imperishability of the soul, goes further than what is allowed by the premises. Here, I agree with Proops. The point seems to be confirmed by Kant’s explanation of the rationalist’s mistake:

In all our thinking the I is the subject in which thoughts inhere only as determinations, and this I cannot be used as the determination of another thing. Thus, everyone must necessarily regard himself as a substance [my emphasis] but regard his thinking only as accidents of his existence and determinations of his state.

But now what sort of use am I to make of this concept of substance? That I, as a thinking being, endure for myself, that I naturally neither arise nor perish – this I can by no means infer, and yet it is for that alone that the substantiality of my subject can be useful to me; without that I could very well dispense with it altogether. (A349)

However, to preserve the appearance of validity of the syllogism while also giving its due to Kant’s explanation at A402-3, I suggest the following friendly amendment to Proops’s schema (p. 138, cited above under 2.2):

All M are P _{(transcendental use)}

S is M _{(empirical use, without sensible condition: unzulässig)}

S is P _{(empirical use, with sensible condition: extended)}

Here again the syllogism can be read as an ignoratio elenchi, extending the ‘improper empirical use’ in the minor premise into a full-blown empirical use in the conclusion. But I submit that we can also see it as a sophisma figurae dictionis if we see the equivocation as concerning the use of the middle term. The same concept has a different use. And the conclusion builds on this equivocation in the use of the very same concept. It is, admittedly, a very unusual kind of sophisma figurae dictionis, in which the equivocation is in the use of rather than the middle term itself.

2.4 The same pattern applies to the second and third paralogisms

Interestingly, Proops acknowledges that the same pattern applies in the case of the third paralogism, since he says that the latter does have the form of a sophisma figurae dictionis, but one in which the equivocation consists in a change of standpoint from the major premise to the minor premise: from an objective standpoint on an entity in the major premise to the mere expression of self-consciousness itself in the proposition ‘I think’, where no object is presented, in the minor premise (p. 161). I submit this is what Proops should have acknowledged in the case of the first and second as well.

As I suggested earlier, the disagreement between Proops’s reading and mine concerns the following question: how essential is the fact that, in Kant’s own terms, ‘“I think” is the sole text of rational psychology’? In my view, Kant’s entire argument against the rational psychologist is built on that claim. The thinking being, considered merely as such, is represented to itself in the proposition ‘I think’ by a concept, ‘I’, which in that proposition can occupy only the position of subject and not that of predicate. The rational psychologist transforms that (universally) subjective representation into the representation of an object. If that is right, it is essential to Kant’s argument that ‘the sole text of rational psychology is “I think”’, considered in just that form. My sense is that Proops does not think the proposition considered in just that form is essential to Kant’s argument. I submit this is the root of his criticism of what he takes to be Kant’s inaccuracy in characterising the structure of his own argument.

3. The changes from A to B

According to Proops, the changes between A and B are ‘significant and substantive’.

Nowhere is it more obvious than in the first Paralogism, where Kant’s critical analysis changes dramatically between the two editions; nowhere is it more important than in the second Paralogism, where Kant comes to regard as flatly false his A-edition claim that one of the argument’s crucial premises – the so-called nervus probandi – lacks ground. (p. 91)

Let me briefly consider each point.

3.1 The ‘obvious’ change in the first paralogism

Kant’s formulation of the paralogism in B:

What cannot be thought otherwise than as a subject does not exist otherwise than as a subject, and is, therefore [insofern], substance.

Now a thinking being, considered merely as such, cannot be thought otherwise than as subject.

So, it also exists only as such a thing, i.e., as substance. (B410)

Proops offers the following regimented formulation:

All entities that cannot be thought otherwise than as subjects are entities that cannot exist otherwise than as subjects and are therefore (by definition) substances.

Entities that are thinking beings (considered merely as such) are entities that cannot be thought otherwise than as subjects.

So:

Entities that are thinking beings (considered merely as such) are entities that cannot exist otherwise than as subject, and are therefore (by definition) substance. (p. 105)

Proops notes that the B formulation (B410-1) is superior to the A. It is clearly a sophisma figurae dictionis, which thus fits Kant’s official characterisation of the paralogisms. The equivocation on the middle term consists, according to Proops, in the fact that in the major premise, ‘thought’ (as subject) means ‘conceived’ (as subject). In the minor premise, ‘thought’ (as subject) means ‘deployed’ (in the syntactical position of subject).

I agree with Proops about the greater clarity of the equivocation on the middle term, which makes the inference a clear case of sophisma figurae dictionis. But I wonder about Proops’s explanation of the equivocation. I agree that it concerns the term and concept ‘thought’. But Kant’s explanation of the equivocation is as follows:

The major premise talks about a being that can be thought of in every respect, and consequently also [auch] as it might be given in intuition. But the minor premise only talks about this being insofar as it is considered as subject relative to thinking and the unity of consciousness, but not at the same time in relation to the intuition through which it is given as an object of thinking. Thus the conclusion is drawn per sophisma figurae dictionis, hence by means of a deceptive inference. (B411)

Kant is quite explicit that the equivocation on the term and concept ‘thought’ consists in the fact that in the major premise, thought is considered in relation to intuition and the entity under consideration is an object of thinking but also an object that may be given in intuition. In contrast, in the minor premise, what is thought of is thought of independently of intuition and indeed is not presented as an object at all. But in Proops’s explanation, the relation to intuition disappears altogether: the distinction becomes one between, in the major premise, thinking as conceiving (subsuming an entity under a concept), and in the minor premise, thinking as deploying a representation in a particular position in a judgement.

If, unlike Proops, we do put the relation of ‘thinking’ to ‘intuiting’ at the core of the equivocation on the middle term (the relation is present in the major premise and absent from the minor premise), we have a clearer view of the similarities and differences between A and B, and also perhaps a clearer view of Kant’s complaint against the argument he attributes to the rational psychologist. In B just as in A, the crux of the argument is the minor premise. In B just as in A, the rationalist’s illusion is to have found, in the way ‘the thinking I’ or the thinking being, considered merely as such, is represented in the proposition ‘I think’, a direct path to an ‘unconditioned subject of inherence’: a substance. The difference between A and B is that in A, Kant starts, in the major premise, from the merely conceptual characterisation of a substance (‘that, whose representation cannot be predicate in a judgment’) and argues that applying this characterisation to ‘the thinking I’ as empirically given (in virtue of the minimal empirical support of the proposition ‘I think’) is making an empirical but improper use of it because no condition of intuition is present for its application. In B, in contrast, Kant starts (in the major premise) from the characterisation of a substance that includes the condition of intuition and argues that the latter condition is absent from the way ‘the thinking I’ is thought, because the condition of intuition is absent from the representation of the thinking I: the concept ‘I’ in ‘I think’. The doctrine is the same, but the way it is argued for is clearer. I submit that by downplaying the emphasis on intuition in the B major premise, Proops misses the nature of the equivocation denounced by Kant as well as the root of the rationalist illusion as Kant reconstructs it.

We find the very same structure in the second and third paralogisms as explained in the very quick exposition Kant offers at B407-8. What is under consideration in the minor premise has not changed from A to B. The major premise is missing in the abridged B exposition. But it is implicit, and the conclusion that Kant rejects is the one the rationalist metaphysician would claim to follow under an equivocation similar to the one explicitly denounced in the First Paralogism. So, for all the paralogisms, the only thing that has changed from A to B is that in B, the major premise (explicit for Paralogism 1, implicit for 2 and 3) considers the thinking of an existing object, which means thinking under the condition of the presentation of the object in experience. Kant’s argument against the rational psychologist is that this is not what is under consideration in the minor premise, which does not present the thinking of an object but the thinking being in the only way it is presented ‘merely as such’: in the expression of apperception, ‘I think’.

Is this change from A to B ‘substantive’? It is strategically important. But it is not a change in doctrine. Here again my disagreement with Proops rests on the different emphasis we respectively put on the contrast between the objective standpoint of the major premise and the (universally) subjective standpoint of the minor premise, which in my view is at the core of Kant’s denunciation of the rationalist illusion.

The same difference in emphasis is at the root of my disagreement with Proops concerning the supposed change in Kant’s view of the rationalist argument in the second paralogism.

3.2 Kant on the nervus probandi of the second paralogism

Due to limitation of space, here I must be far too quick. According to Proops, between A and B Kant’s attitude to the minor premise of the second paralogism changed. According to him, Kant accepts, in B, the minor premise he rejected in A. Contrary to Proops, I maintain that Kant accepts in B the minor premise in just the same sense as that in which he accepted it in A: as an expression of the indivisibility, for the thinker, of the act of thinking expressed in the proposition ‘I think’. The rationalist illusion consists in transforming the (universally) subjective indivisibility of the act of thinking into an objective indivisibility of the metaphysical subject of thinking: the soul, as a thinking substance. Let me add, however, that Proops’s analysis of Kant’s refutation of the ‘Achilles’ argument in A (the refutation of the rationalist’s objective argument for the indivisibility of the soul) is brilliant. My disagreement about the role of the minor premise notwithstanding, I have learnt an enormous amount from Proops’s analysis of that paralogism, as of all the others.

I cannot end these comments without emphasising again how much I admire Proops’s book. It is an incredibly rich book which will shape discussions of Kant’s Transcendental Dialectic for years to come.