Where Do All These Ideas Come From? Kant on the Formation of Concepts Under the Guidance of Pure Reason

1.1 Given and made concepts

The idea that the entire repertoire of possible concepts can be traced back to some basic concepts runs through the entirety of modern philosophy. According to Descartes, such basic concepts denote “simple natures,” such as “cognition” (cognitio), “extension” (extensio), and “existence” (existentia).^[9] According to Locke, there are simple ideas from which more complex ideas are composed.^[10] And Leibniz sees the possibility of a complete analysis of concepts leading to simple concepts as their basic elements, which are recognized by intuition.^[11] To be sure, Kant rejects the idea of innate concepts as well as the idea that something simple could be given in intuition.^[12] However, he also follows this tradition in a certain sense when he speaks of derived or compound concepts. These concepts play a role in various contexts of his philosophy. Before I turn to an example, namely the Paralogisms chapter of the first Critique, where Kant tries to derive some rational–psychological concepts by “composition,”^[13] I would first like to briefly introduce Kant’s proposal for classifying all (theoretical) concepts according to “given” and “made” concepts, as found in his Logic. By doing so, one can also gain a complete overview of the totality of all possible concepts.

In his doctrine of concepts, presented in the Logic, Kant basically distinguishes given from made concepts with respect to the “matter” of concepts (their “object”^[14]), where the given concepts “are given either a priori or a posteriori.”^[15] Kant specifies only the given concepts in more detail there, but Reflexion 2853 further distinguishes made concepts. In this Reflexion, he distinguishes “arbitrarily” (willkürlich) and “naturally” (natürlich) made concepts, the first being “invented” (gedichtet) and the second being concepts “inferred by reason,” called “ideas.”^[16] Accordingly, all concepts can be divided into the following types: (a) given concepts are either (aa) categories (“notions”) as a priori given concepts or (ab) concepts of experience as a posteriori given concepts; (b) made concepts are either (ba) ideas as naturally made, i.e., here: “inferred by reason,” concepts, or (bb) arbitrarily made concepts.^[17] Adding Kant’s division of concepts with respect to the “origin of concepts in regard to their matter,” which is not considered in (general) logic but “in metaphysics,” the concepts of kind (aa) are called “intellectual” concepts, of kind (ab) “empirical” concepts, and of kind (bb) “arbitrary” concepts.^[18]

However, the classification of all (theoretical) concepts according to the distinction between given and made concepts is problematic, especially with regard to the classification of ideas (ba). This is because an exact identification of the ideas qua “naturally” made or “inferred” concepts with the transcendental ideas, with which the “Transcendental Dialectic” of the Critique of Pure Reason deals in an outstanding way, turns out to be implausible. It is true that right at the beginning of the first book of the “Transcendental Dialectic,” Kant calls the “concepts from pure reason” “inferred concepts.”^[19] In this respect, with reference to Reflexion 2853, they could be understood as made concepts, which, unlike the arbitrarily made concepts of type (bb), are not composed arbitrarily, but “contain something … borrowed from the nature of the understanding”^[20] and are intellectually made concepts. However, this interpretation is opposed to Kant’s statement “that transcendental ideas are just as natural to [human reason] as the categories are to the understanding.”^[21] Kant’s starting point for the derivation of ideas is the “form of the syllogisms” or the “function of reason in its inferences,”^[22] i.e., reasoning as a logical function of reason. With this, Kant proceeds analogously to his explanations on the clue of the “Discovery of all Pure Concepts of the Understanding.”^[23] There, the starting point for the formation of the categories, i.e., of the concepts given a priori (aa), is set in the logical function of understanding (judging). Both the categories as well as the transcendental ideas are thus derived by Kant in the Critique of Pure Reason from the logical functions – the “nature” of understanding or reason.

The dichotomy of given and made, as found in the doctrine of concepts of the Logic, thus turns out to be insufficient to address the specificity of pure concepts of understanding and reason.^[24] That is because either both kinds of concepts are given concepts, since they are “borrowed” from the “nature” of understanding or reason,^[25] or both are made concepts, since they are “reflected” or “inferred” a priori concepts. They are then understood as concepts derived from the functions of judging or reasoning, which as the logical functions alone constitute the logical “nature” of understanding or reason and which are accordingly exclusively given a priori.^[26]

Nevertheless, Kant’s classification of the transcendental ideas in Reflexion 2853 as made concepts is comprehensible. That is because the categories are “merely reflected concepts:” “they contain nothing beyond the unity of reflection on appearances.”^[27] That which they describe, the “synthesis of a manifold,”^[28] is already presupposed in every cognition of an object. And insofar as knowledge of objects proceeds in judging, the categories can be taken directly from the functions of judgment. In contrast, reason never refers directly to objects, but to the judgments of the understanding.^[29] And its logical function, reasoning, already has a less immediate character.^[30] Moreover, the derivation of the transcendental ideas is once again more complex than that of the categories. I will come back to this. At this point, it may suffice to ascribe a kind of intermediate position to the transcendental ideas in the context of the distinction between given and made concepts: On the one hand, they arise from the given “nature of reason,” and on the other hand, they still have to be “inferred” in a certain way and to that extent are “made.” They could therefore be called “quasi-given concepts.”

The arbitrarily made concepts (bb) can also be subdivided again into (bba) mathematical concepts that contain “an arbitrary synthesis which can be constructed a priori,”^[31] (bbb) predicables (Prädikabilien) as “derived concepts a priori,”^[32] and (bbc) arbitrarily composed concepts that “depend upon empirical conditions.”^[33] In these concepts, there are clearly fewer problems resulting from the underlying distinction between given and made. Thus, the determination of mathematical concepts (bba) as arbitrary (made) concepts also coincides with numerous statements by Kant, not only in the Logik, but also in the Critique of Pure Reason.^[34] The fact that the mathematical concepts are to be characterized as made concepts refers here to the possibility of their a priori construction in pure intuition.^[35] However, it is to be noted here that the forms of intuition themselves and the formal concepts of space and time are obviously given,^[36] which is why there can be no question of a completely arbitrary construction of these concepts.

The predicables (bbb), in turn, as derived concepts of the understanding, can also be counted among the intellectual concepts, since some of them are generated merely from the combination of the categories “with each other.”^[37] Moreover, there are predicables that are generated by a combination of categories “with the modis of [pure] sensibility.”^[38] Since in the case of predicables a special combination is necessary, they can well be understood as made concepts (bb). Finally, in their distinction as arbitrarily made concepts, as in the case of composed concepts (bbc), the problem arises that the elements to be combined are not (arbitrarily) constructed or “made” ones, but given ones – in the case of predicables: the categories and “modis of pure sensibility,” and in the case of composed concepts: empirical concepts (ab).

The classification of all (theoretical) concepts on the basis of the distinction between given and made from the Logik thus turns out not to be completely coherent. This distinction, however, draws attention to two points that are important with respect to a general characterization of the ideas of pure reason. First, Kant assumes that concepts can be given a priori if they arise from the logical “nature” of understanding or reason, as the example of the categories (aa) demonstrates. And second, a priori concepts can also be made concepts, as the examples of mathematical concepts (bba) and predicables (bbb) demonstrate.

1.2 Objective reality of concepts

For a coherent classification of all (theoretical) concepts, it is more promising to consider the referentiality (objective reality) of the individual types of concepts.^[39] Looking at the categories (aa) and the mathematical concepts (bba), this question can be answered in short order because both types of concepts represent special cases with respect to the question of their objective reality. The categories have a special kind of objective reality, because they “lie at the ground of all experiential cognition.”^[40] According to Kant, they are the only given concepts whose objective reality can be demonstrated a priori. In the case of mathematical concepts, the proof of objective reality is secured by the procedure of mathematical construction. It is true that they are not concepts of objects of empirical knowledge. However, they have objective reality, because their object can be given a priori in pure intuition and they “can be applied to empirical intuitions.”^[41] They are therefore those made concepts whose objective reality can be shown a priori.

On the other hand, in the case of predicables (bbb), the question of objective reality cannot be decided a priori as in the case of categories and mathematical concepts. On the contrary, in this respect, they do not differ from the arbitrarily composed concepts (bbc) and the empirical concepts (ab). In all of these concepts, according to Kant, it is necessary that there be an empirical presentation of the concept. Kant makes this point particularly clear in his late Preisschrift:

The possibility of a thought or concept rests on the principle of contradiction, e.g., that of a thinking immaterial being (a spirit). The thing of which even the mere thought is impossible (i.e. the concept is self-contradictory), is itself impossible. But the thing of which the concept is possible is not on that account a possible thing. The first possibility may be called logical, the second, real possibility; the proof of the latter is the proof of the objective reality of the concept, which we are entitled to demand at any time. But it can never be furnished otherwise than by presentation of the object corresponding to the concept; for otherwise it always remains a mere thought, of which, until it is displayed in an example, it always remains uncertain whether any object corresponds to it, or whether it be empty, i.e., whether it may serve in any way for knowledge.^[42]

The objective reality of a concept can be spoken of precisely when it demonstrably refers to an object and thus claims, in the case of intuitions and concepts, to contribute to the knowledge of this object or, in the case of judgment, to afford knowledge of this object. With regard to concepts, Kant says here that the criterion for the objective reality of a concept is not the mere absence of contradiction of the concept, i.e., the mere conceivability of the object designated by it (as in the case of the concept “spirit,” Geist), but the possibility of the presentation of the object designated by it in intuition. Accordingly, a concept has objective reality if its object is not only logically possible, i.e., conceivable without contradiction, but furthermore demonstrably real, i.e., can be represented in intuition.^[43]

The formation of empirical concepts (ab) by the logical operations of understanding (“comparison,” “reflection,” and “abstraction”) guarantees eo ipso their objective reality, since in the case of empirical concepts the representations from which such concepts are “derived” are representations of a given manifold of empirical intuition.^[44] In short, empirical concepts are produced on the occasion of perceptions. An “example” of an empirical concept already exists at the time of its genesis. If, on the other hand, empirical concepts or concepts given a priori are combined to new concepts (bbc and bbb), the objective reality of these new concepts must be shown separately, if their objects are to be regarded as objects of experience.^[45]

In the context of the question of objective reality, the ideas of pure reason are an exception in the set of all possible concepts. For the transcendental ideas are “merely sophistical (dialectical) concepts” if they are understood as “concepts of certain objects.”^[46] Accordingly, the ideas of pure reason are those a priori concepts to which no objective reality can be ascribed. While categories and predicables can be represented in intuition, the ideas of pure reason are characterized by the fact that no object can correspond to them in experience, since they go “beyond the possibility of experience.”^[47] They transcend the realm of possible experience constituted by the cooperation of pure forms of intuition and categories, since they represent the “unconditioned”^[48] or the “supersensible.”^[49] Through this representation of the “unconditioned,” they produce the “transcendental illusion” of being related to special objects, which is why they are to be treated in a “dialectic” according to the system of critique.^[50]

The two distinctions mentioned – that between given and made concepts and the question of objective reality – allow for a characterization of the ideas of pure reason as follows: These ideas are concepts that have their origin in pure reason, and are therefore a priori concepts, but their objective reality can never be proven. With regard to the first distinction, an intermediate position could be identified, although it remains open whether all or only some ideas of pure reason are quasi-given concepts, i.e., whether there are not also made ideas. In what follows, I will argue that some ideas of pure reason are quasi-given concepts from which further ideas can then be derived. This two-step procedure is similar to that of deriving predicables from categories. Most importantly, this procedure allows Kant to fulfill one important claim of his “Transcendental Dialectic,” namely to explain the origin of the entire a priori vocabulary of the metaphysica specialis.

3.1 Transcendental idea(s) of rational psychology

The sketch above of the derivation of the transcendental ideas has pointed out three essential steps which we should also find at work in the case of the transcendental ideas of rational psychology. The derivation goes as follows:

1. The first step is the correlation to one of the forms of syllogism in order to state the “origin” of the concepts of rational psychology. Here, the categorical syllogism is relevant, which is characterized by the fact that the major premise is a categorical proposition, i.e., a relation of inherence is thought. As we have seen, for the derivation of concepts of reason from this relation of inherence, a “totality of conditions” must be thought as a “totality” of features of a subject.

2. The second step fixes this thought of the “totality” of features of a subject. In the case of the categorical syllogism, it is not immediately obvious how in an ascending series of syllogisms an ever more general premise and finally a pure concept of reason are found in this regard.^[95] The prosyllogistic procedure here merely leads to more and more general concepts of the logical subject, none of which, however, constitutes an “unconditional subject,” i.e., a subject that could not itself function as a predicate again. Only the notion of a thing would remain, from which all properties have been abstracted (“the concept of the substantial,” Substantiale).^[96] But how should the idea of a specific “totality” be derived from this notion? Kant himself, in his derivation of the cosmological ideas, explicitly identifies this “substantial” with the “transcendental subject,”^[97] and at the beginning of the Paralogisms chapter, he refers directly to the transcendental concept of “I think.”^[98] And indeed, in a certain sense, “I think” fulfills the condition of giving a kind of “totality in the series of premises”^[99] in the case of a series of categorical prosyllogisms. For all concepts are representations of a thinking subject: the “I think” is “the vehicle of all concepts in general”^[100] and the “I as subject … cannot again be thought as the predicate of some other subject.”^[101] This representation is nothing other than the “transcendental concept of a subject that contains nothing manifold,”^[102] for all definite predicates are abstracted from, and only its pure predicatability is thought. Since in an ascending series of syllogisms to every more general premise the “I think” can be added again, the transcendental concept of a subject thus corresponds exactly to the sought concept of a most general subject. Thus, when pure reason seeks a “totality” of all possible “conditions” in categorical inferences, it finds this completeness in the relation of all concepts to “mere apperception.”^[103]

3. But the “I think” must not only designate the “transcendental unity of self-consciousness”^[104] if it is to be the foundation for rational psychology. And as emphasized above, the mere thought of a specific totality is not yet sufficient to be called a transcendental idea. Rather, in a third step, this totality must still be taken as an object of cognition. That the totality relevant here refers to the subject already became clear in the reconstructed second step. There are then still two tasks open to make the “I think” a specific object of cognition: It has to be determined by categories – like any object of cognition – and its objective reality has to be demonstrated – due to the lack of reference to intuitions. The first task is fulfilled by the “topics of the rational doctrine of the soul.”^[105] The second task would have to be fulfilled by correct inferences of pure reason^[106] – and it is Kant’s claim to have shown the “fallacy”^[107] of rational psychology meticulously in the chapter on the paralogisms.

Since this second task no longer belongs to the topic of concept formation under the guidance of pure reason, I will not discuss it further here.^[108] The “topics” (Topik), on the other hand, is important for us here, since the four “transcendental predicates”^[109] of “I think” are listed there. Since the thought of the specific totality of categorical reasoning is supposed to designate a particular object of cognition, with this thought “first a thing, I as a thinking being, is given.”^[110] Accordingly, at the top of the “topics” is the characterization of the thinking being by the category of substance, then follow the characterizations of it as “simple,” as “numerically identical,” and as existing (independently of other objects) “in space.”^[111] The exact correlation of the respective categories is not further justified by Kant in the context of the “topics.” It is obviously oriented first of all to the determinations of the “soul” by contemporaneous rational psychology.^[112] Most importantly, this correlation remains incomprehensible without Kant’s further explanations in the Paralogisms chapter. And indeed, in the second edition, Kant gives a decisive clue right at the beginning: It is the “analysis of the consciousness of myself in thinking in general” that is “falsely held to be a metaphysical determination of the object [=the soul].”^[113] This “analysis” leads by means of the “modi of self-consciousness” (as “mere functions,” not already as “categories”^[114]) to four “identical” (analytic) propositions: “in every judgment I am always the determining subject”;^[115] “the I of apperception, consequently in every thought, is a single thing [Singular]”;^[116] I am identical with myself “ in everything manifold of which I am conscious”;^[117] “I distinguish my own existence, that of a thinking being, from other things outside me.”^[118]

If these “trivialities”^[119] are used to determine by categories that which the thought of the specific totality of categorical reasoning takes as an object, then the result is the characterization of “I think” as a simple, identical substance that exists independently of other objects in space.^[120] Thus, for the production of the transcendental ideas of rational psychology, a further step beyond the above three general steps is needed: the use of the results of an analysis of self-consciousness, which, however, are not merely understood as platitudes of transcendental logic, but as characterizations of a specific object – precisely of a “thinking being.”

The fact that Kant has made good progress here on the way to explain the origin of the entire a priori vocabulary of rational psychology can be shown quite well. For at least the characterization of the thinking being as a simple substance that exists independently of other objects in space is familiar in German rational psychology of the eighteenth century.^[121] Accordingly, the psychologia rationalis in Baumgarten’s Metaphysica begins with a statement of the “nature of the human soul” (natura animae humanae), where substantiality, simplicity, and its existence in space are emphasized at the very beginning.^[122] Crusius’ quite differently conceived “Pneumatology” starts from the beginning with “spirits” (Geister) that are eo ipso substances distinct from the body.^[123] That they nevertheless exist in space and that they are simple is essential for Crusius, and he devotes extensive discussions to both features.^[124] And Feder, who is comparatively reserved about metaphysical speculations, also presupposes the substantiality of the soul in his exposition of the “Pneumatology” and considers its “simplicity” to be the most important feature.^[125] Only the characterization as an “identical” being does not come up explicitly in these texts. The reason for this is that Kant distinguishes “identity” and “personality” (Persönlichkeit) here in the rational–psychological context. Since Kant presents personality as a derived concept of reason,^[126] I will return to it in Section 3.2.

3.2 Derived ideas of rational psychology

With the “topics of the rational doctrine of the soul,” besides the concept of a subject that does not contain anything manifold (= soul), four further concepts of reason are obtained: the concept of the substantiality of the soul, the concept of the simplicity of the soul, the concept of the identity of the soul, and the concept of the existence of the soul that is independent of other spatial objects. Directly following the “topics,” Kant explains that further concepts of reason can be derived in turn from these concepts:

From these elements, at least through composition, spring all the concepts of the pure doctrine of the soul, without any other principle being cognized in the least. This substance, merely as an object of inner sense, gives us the concept of immateriality; as simple substance, it gives us that of incorruptibility; its identity, as an intellectual substance, gives us personality; all these points together give us spirituality; the relation to objects in space gives us the interaction with bodies; thus it represents the thinking substance as the principle of life in matter, i.e., as a soul (anima) and as the ground of animality, and this – limited by spirituality – is immortality.^[127]

According to this claim, further concepts of rational psychology can be produced by mere “composition.” In order to understand this claim better and to make it somewhat plausible, the six “compositions” can be reconstructed in a first step – in the most possible closeness to the text and paying attention only to the “elements” and their “composition” – as follows (where: 1 = “soul as substance”; 2 = “soul as simple”; 3 = “soul as identical”; 4 = “soul as existing independently”; + = “composed with”):

1. “[t]his substance, merely as an object of inner sense, gives us the concept of immateriality”: 1 + inner sense = immateriality;^[128]

2. “[t]his substance … as simple substance … gives us [the concept] of incorruptibility”: 1 + 2 = incorruptibility;

3. “[the] identity [of this substance], as an intellectual substance, gives us [the concept of] personality”: 1 + 3 = personality;

4. “the relation [of this substance] to objects in space gives us [the concept of] the interaction with bodies; thus it [=pure doctrine of the soul] represents the thinking substance as the principle of life in matter, i.e., as a soul (anima) and as the ground of animality”: 1 + 4 = interaction (Kommerzium – as principle of life);

5. “all these points [=immateriality, incorruptibility and personality] together give us [the concept of] spirituality”: (a) + (b) + (c) = spirituality;^[129]

6. “this [=as the ground of animality] – limited by spirituality – [gives us the concept of] immortality”: (d) + (e) = immortality.

Taking this reconstruction as a starting point, a general observation can be pointed out in a second step. Of the six derivations, (b), (c), and (d) are said to be combinations of 1 with the other elements, i.e., 2, 3, and 4. The last two derivations (e) and (f), on the other hand, are combinations of already derived concepts, whereas (f) even contains a concept (spirituality) derived from derived concepts. All these five derivations therefore consist of more or less simple compositions of the original elements – “without any other principle being cognized in the least.” The first derivation (a), however, seems to be an exception, insofar as here the first element (soul as substance) is thought “as an object of inner sense.” So, it is not composed with one of the other elements, but with inner sense. With this, however, it is only emphasized that the substance thought here is a thinking being and as such could also only be accessible to inner sense.^[130] But with that it also cannot be anything material, perceptible through the outer senses.^[131] It is therefore to be called non-corporeal (“immaterial”).

In a third step, the other derivations can also be considered one by one. However, they are in need of explanation. The most unproblematic derivation is certainly the second one (b): Every simple substance is by definition imperishable (also: incorruptible, indestructible). It does not matter whether it is conceived as material or immaterial – even an absolutely simple body would be incorruptible.^[132] The concept of corruption (corruptibilitas) used here is that of a “division” (Zerteilung, divisio physica) or “separation” (Zertrennung) of the parts,^[133] which obviously cannot be applied to an absolutely simple thing. The conceptual composition here is thus merely a connection of the concept of substance with the concept of the simple.

The third derivation (c), on the other hand, is more difficult, at least in a terminological respect. Apart from the fact that “personality” is presented here as a derived concept, but at a later point is apparently used synonymously with “identity” (i.e., element 3),^[134] the interjection “as an intellectual substance” is remarkable. For obviously a composition of “substance” (element 1) and “identity” (element 3) would not yet lead to “personality” in the sense of an identity extending over time, of which this substance is also conscious.^[135] An identical substance could just as well be a material substance, and then the feature “personality” would not be attributed to it. Therefore, the concept of substance used here must explicitly be that of an intellectual substance, that is, a thinking substance that possesses a “memoria intellectualis.”^[136] If a thinking substance is aware not only of its particular states, but also of its identity in changing states, then it possesses a corresponding faculty, which is called “personality” in early modern philosophy.^[137] The composition here is thus the connection of the concept of (thinking) substance with the concept of identity, so that this substance is conscious of its states and its identity.

The fourth derivation (d) is problematic, since on the one hand, the concept of an “interaction with bodies” (Kommerzium mit Körpern), and on the other hand, the concept of a “principle of life” results from it. As far as the concept of interaction with bodies is concerned, it seems to be only about the composition of the concept of substance (element 1) with the property of something which exists in space but independent of spatial things (element 4). This composition results in a thing which is not spatial itself, but is in connection with spatial things (bodies). What is in question here is the exact nature of the connection in which the thinking substance, distinct from spatial things, stands with bodies. Kant does not problematize this question – a classical controversial issue in the metaphysics of philosophy in early modern times – further here, but in using the term “commercium” merely gives the indication that the thinking substance stands in a “real community”^[138] with bodies. Moreover, he specifies the thinking substance directly here as “principle of life” or as “ground of animality.” Thus one would have to speak of a living thing if a thinking substance exists and can interact with material things. This corresponds at least to the views of traditional German metaphysics.^[139] With regard to Kant, this is on the one hand plausible, since it corresponds to his relevant precritical considerations.^[140] On the other hand, it becomes particularly clear here that Kant is merely trying to reconstruct given terms from mainstream philosophy. For he himself defines the concept of life in his critical philosophy by using the concept of the faculty of desire (Begehrungsvermögen) and the concept of law (Gesetz) and precisely no longer in the terms of the metaphysics of substances.^[141]

The fifth derivation (e) also seems to be rather unproblematic. Because in it the concepts of immaterial substance, incorruptibility, and personality are connected (“composed”) with each other. The concept of the spiritual is consequently defined using the predicates “immateriality,” “incorruptibility,” and “personality.” But as clear as this definition may look, it seems to be arbitrary. Baumgarten, for example, defines “spirit” simply as intellectual substance (substantia intellectualis),^[142] which is already given by the derivation (c). If one adds the common premise that matter cannot think,^[143] then the other two predicates “immateriality” and “incorruptibility” could also be directly inferred. However, the concept of spirit (Geist) is defined, for instance, by Crusius in such a way that it refers to a substance that can also think and will independently of its connection to the body.^[144] And indeed, even in the late Preisschrift, Kant himself defines “spirit” as a soul which can continue to live (think and will) even without connection to the body.^[145] Thus, the concepts of immateriality and incorruptibility are added to the concept of a person, making this derivation comprehensible as well.

Finally, the sixth derivation (f) can be reconstructed insofar as it starts from the concept of “interaction with bodies” (result of (d)), but this is “limited.” According to the traditional view, only those living beings are immortal whose “ground of animality” is a spiritual substance. To this belongs the quality of personality, which is denied to animals.^[146] It is true that animals also have life, since their body is connected with an immaterial and incorruptible soul. However, it is only the feature of personality, consequently spirituality, which distinguishes a soul as “immortal,” such that the thinking substance can also exercise its abilities (to think and to will) independently of its connection with the body. Thus the composition here is that between the concept of a thinking substance connected with bodies and the complex concept of a spiritual substance according which the second is not only understood as a type of the first (like “human” and “animal”), but also a separation of the thinking substance from the body is considered. And only in the case that the thinking substance is a spiritual substance, can it continue to exercise its abilities and in this respect be called immortal.

Having discussed the different derivations, in a fourth step we can ask about the special status of the concepts produced here. Since they are produced by certain compositions of those concepts which are listed in the “topics,” they must be clearly regarded as made concepts. And since the elements mentioned in the “topics” are a priori concepts, these derived concepts of reason are made a priori concepts. In this respect they are like predicables, which are produced by compositions from the categories. Unlike the predicables, however, the elements from which the derived concepts of reason are composed are not “ancestral concepts” (Stammbegriffe),^[147] but predicates of the supposed object of an inferred concept. This concept is indeed produced on the basis of the “nature” of reason. But the assumption that the object of this concept actually exists is based on a false inference.^[148] In the case of predicables, on the other hand, there is at least the possibility of granting them objective reality by examples in experience.

Nevertheless, the comparison with predicables is helpful in at least two respects:

First, both the case of the predicables and the case of the derived concepts of pure reason show that Kant continues the traditional view of the formation of concepts from a few basic concepts. From given a priori concepts, further a priori concepts are to be produced by compositions. While in the case of predicables these derived concepts are made directly from a combination of the categories; however, the elements for the production of the derived concepts of reason are concepts already “inferred” and produced by categorial determination. As in the case of the predicables, however, Kant gives only vague hints as to how exactly the respective production (“composition”) works, so the specific derivations remain unclear.^[149] Already in the case of the predicables, the derivations seem to be complex and do not follow a single principle. This also applies as we have seen to the production of the derived concepts of reason, since the respective elements to be composed are not always used in the same sense. Moreover, Kant’s explanation of composition as “synthesis of a manifold of what does not necessarily belong to each other”^[150] is so general that it does not help much. As in the case of predicables, therefore, there is nothing left to do but to consider the particular derived concepts and Kant’s references to their elements carefully and to ask how the composition functions in the respective case more precisely.^[151]

Second, the special systematic function of the predicables can be found in a certain sense in the derived concepts of reason as well. This is because the predicables are first relevant for a “complete system of transcendental philosophy,”^[152] that is, a reconstruction of the vocabulary of ontology of traditional metaphysics.^[153] Moreover, the predicables play an important role for particular sciences, providing, for instance, the basic concepts for the metaphysics of nature or the metaphysics of law.^[154] Thus, on the one hand, predicables have an integrating function, as Kant wants to use them to show that he can derive all common ontological concepts from his table of categories. On the other hand, some predicables also have a constructive function, insofar as their a priori character serves Kant to work out “metaphysical foundations” for certain domains of objects. This second function cannot, of course, belong to the derived concepts of reason in the same sense, since their elements are already mere ideas, i.e., problematic concepts. And Kant wants to show in the “Transcendental Dialectic” precisely that the traditional metaphysica specialis is an impossible project.

In contrast, the derived concepts of reason certainly fulfill the first function – at least with respect to the example of rational psychology. For just as the traditional ontological vocabulary is not exhausted in Kant’s twelve pure concepts of understanding, traditional rational psychology is not concerned only with the determination of “I think” as a simple, identical substance independent of other objects. Rather, it is precisely the spirituality of the substance of the soul that is thematized in rational (as opposed to empirical) psychology.^[155] The soul as spirit is essentially characterized in rational psychology by the predicates “immateriality,”^[156] “incorruptibility,”^[157] and “personality.”^[158] The connection of the soul with the body and its function as an animating principle are central themes of the philosophical accounts of traditional metaphysics.^[159] And the reflections on the immortality of the soul are the climax or crowning conclusion of rational psychology.^[160] If one adds Kant’s understanding of rational psychology, which follows Baumgarten and according to which rational psychology deals with those concepts that “are derived from the concept of the soul by a longer series of inferences of reason” (ex notione animae longiori ratiociniorum serie deducit),^[161] then one can rightly claim that he can explain the origin of the entire a priori vocabulary specific to rational psychology. This explanation leads from one of the logical functions of reason, through the production of the concept of an absolute subject, to concepts denoting the features of such a subject. Most of these concepts are produced by a combinatorial procedure that combines categorical elements – similar to the case of predicables.