Kant on Philosophy as Conceptual Analysis

Lewin, M. 2023. Kant on Philosophy as Conceptual Analysis, Con-Textos Kantianos 18, 11-20. https://dx.doi.org/10.5209/kant.91509. [1]

Abstract: For Kant, philosophical investigations are inherently analytic. The proper method of philosophy is analysis, and the object of analysis are concepts. Hence, Kant’s short description of philosophy as “rational cognition […] from concepts” (KrV, A 837/B 865) can be substituted by “philosophy is conceptual analysis”. The article shows that Kant follows a representationalism about concepts and a combination of intensional and extensional feature semantics. Against the claim that Kant is a proponent of the concept-judgement-inversion, it is argued that concepts are being articulated in form of singular terms, propositions, and sets of propositions alike. Kant interlinks different kinds of concepts with different kinds of definitions, as the reconstruction of his theory of definitions reveals. Philosophy foremost deals with what Kant calls ‘given’ empirical and pure concepts. Such concepts are object of analysis, i.e., explication and exposition. The article ends with answers to five possible objections. Is philosophy all about analysis?

Keywords: Metaphilosophy; methodology; conceptual analysis; theory of definitions; representationalism; judgements.

1 Introduction

Some scholars purport that analytic philosophy owes much to Kant. For example, the critical turn, the analytic-synthetic-distinction, and the concept-judgment-inversion. It has been claimed that early analytic philosophers contrasted their theories with the Kantian thought, and that conflicting changes within analytic philosophy since the second half of the 20^th century resulted partially from the lack of engagement with Kant.[2] The assessment of Kant’s importance in this regard is dependent on what one understands by ‘analytic philosophy’. The term ‘analytic philosophy’ alone does not render any clear-cut insight into the methodological and doctrinal commitments of the members of the signified philosophical movement. Dummett’s suggestion of three tenets of analytic philosophy – (1) analysis of the structure of thought; (2) the study of thought is separated from the study of psychological process of thinking and (3) the only proper method is the analysis of language – is neither an empirical linguistic explication of this term nor a necessary conceptual exposition (see Dummett 1978, p. 458). ‘Analysis’ could mean ‘decomposition’ or ‘replacement’ and be carried out in form of ‘definition’, ‘explanation’, ‘exposition’, ‘explication’, and ‘demonstration’.[3] These operations may be dependent on the kind of object that a word or concept refers to. The matter of analysis could be linguistic entities, but also concepts, values, and things directly. As Williamson (2022, pp. 14-24 and Lewin and Williamson (2023)). shows, analytic philosophers sometimes drop the third and can even drop the first tenet. The observation that ‘philosophy’, including the concept of wisdom, has been understood in many incommensurable and conflicting ways, adds even more terminological confusion.

The title ‘analytic philosophy’ is so abstract and problematic in context of the rational demand of correct and precise naming in science that it could be even a good title for Kant’s philosophical programme, as I have polemically argued in a joint debate article with Williamson (see Lewin and Williamson (2023)). If one turns to Kant, one finds definite and decisive expositions of the concepts of philosophy and analysis. Moreover, one gets to know that philosophical investigations are necessarily and inherently analytic – i.e., the proper way to do philosophy is to do it analytically. The object of analysis are concepts. Hence, philosophers are conceptual analysts. This is not only explicit in the methodological section of the Critique of Pure Reason, the Doctrine of Method, but also in the corresponding and further passages in the transcripts of his logic lectures, which Kant constantly developed throughout 40 years of lectureship.[4] To defend the thesis that for Kant, conceptual analysis is the proper philosophical method, I will give a synoptic overview by explaining what he understands by concepts, how they relate to judgements and inferences, how different kinds of concepts are being defined, and what philosophers do with concepts.[5] After that, I will discuss whether analysis is the alpha and omega of Kant’s critical concept of philosophy. Namely, one could object that (1) philosophy is not only about analysis; (2) Kant also speaks of a synthetic method; (3) we do not analyse when we form synthetic judgements; (4) analysis can lead to dogmatic metaphysics, the object of Kant’s critique, and ergo cannot always be the proper method of dealing with concepts in philosophy; and (5) Kant himself does not seem to explicitly apply the method of conceptual analysis.

2 What are Concepts in Kant’s View?

A concept, as the Latin ‘conceptus’ suggests, is a cognitive cluster consisting of several interrelated cognitive units. Kant has called these units, following the mentalist tradition since Descartes, ‘representations’. Representations are the medium of all mental operations, and different mental capacities are responsible for different kinds of representations. Kant gives the following taxonomy of representations at the beginning of the Transcendental Dialectic:

The genus is representation in general (repraesentatio). Under it stands the representation with consciousness (perceptio). A perception that refers to the subject as a modification of its state is a sensation (sensatio); an objective perception is a cognition (cognitio). The latter is either an intuition or a concept (intuitus vel conceptus). The former is immediately related to the object and is singular; the latter is mediate, by means of a mark, which can be common to several things. A concept is either an empirical or a pure concept, and the pure concept, insofar as it has its origin solely in the understanding (not in a pure image of sensibility), is called notio. A concept made up of notions, which goes beyond the possibility of experience, is an idea or a concept of reason (KrV, A 320/B 376f.).

From this follows that all concepts are representations, i.e., they belong to the genus ‘representation’, but not all representations are concepts. Kant divides representations into singular and universal (see Log, AA 09: 91). While the former concern singular objects in sensation, the latter contain marks or serve as marks that can be attributed to several objects via the faculties of judgment, understanding, and reason. For example, the sensation of the colour red or the immediate sensible intuition of a flower are singular representations. But the concepts of the red colour and the flower contain marks that are mutual to many other objects of cognition, such as roses or dahlias. Concepts are therefore always universal representations, and the expression ‘universal concept’ is a mere tautology (cf. ibid.).[6]

Kant’s theory of representations may be considered in the wake of representationalism of the early modern European philosophy, but with transcendentalist’s modifications.[7] A representation is always a representation of something, or, in Kant’s words, of an ‘object in general’. This is the genus; the species are empirical objects and objects that are not given in experience – although they may be, in case of such pure concepts as ‘pure will’, ‘virtue’, ‘wisdom’, and ‘philosophy’, at least partially given in concreto, as ectypes, imperfect examples of conceptual archetypes.[8] Hence, one can have concepts, i.e., universal representations, of supersensible objects. The analysis of concepts of such objects is anytime possible and often necessary, but in some cases prone to fallacies (see the last section of this article).

Kant distinguishes between two kinds of concepts, empirical and pure concepts, whereby the later can be subdivided into pure concepts of the understanding and pure concepts of reason. The empirical concepts obtain their universal form via three interconnected logical operations: comparison, reflection, and abstraction (cf. Log, AA 09: 94). I compare red roses and red dahlias, I reflect on what is common to them, and abstract it from all other possible marks to get the concepts ‘red colour’ and ‘flower’. These concepts, in their turn, are parts of other concepts. I can think of many other different flowers, and I can identify something as a flower knowing what necessarily belongs to the concept of flower. Empirical concepts are therefore abstracted copies of objects in experience, which is also true regarding concepts of made-up or invented objects.[9] The situation is somewhat different when it comes to pure concepts. To the domain of pure concepts belong the categories (substance, causality), preadicabilia (force, action), notiones (the substantial, the simple), ideas (virtue, freedom, infinity of the universe, philosophy), and concepts of mathematical objects (such as a triangle). On the one hand, as Kant claims in the Logic Jaesche and as far as one can trust a not fully authorized transcript – the logical operations of comparison, reflection, and abstraction “are the essential and universal conditions for generation of every concept whatsoever” (Log, AA 09: 94). But on the other hand, Kant claims that pure concepts are not borrowed from experience, which means that they cannot be a result of abstraction from sensible objects and combination of representations.[10] This causes a serious dilemma.[11] One way out of this dilemma is the claim that they are – except of mathematical concepts – given, either by the nature of the understanding or reason.[12] The other way is to suggest other operations, like Kant did regarding the ideas of reason claiming that they are not merely reflected but inferred (cf. KrV A 310/B 366). Or like in case of mathematical concepts, suggesting that they result from an intuitive synthesis.

Another important theoretical element of the Kantian concept of concepts is the distinction between their intension and extension, which can be traced back to the Aristotelian logic, Logic of Port Royal, and e.g., Meier’s Doctrine of Reason. Kant follows a certain ‘feature semantics’, which suggests that concepts consist of multiple so-called ‘marks’. Marks are representations that serve as grounds of cognition – they identify objects of cognition and the intension and extension of concepts. For instance, we can identify something as a planet because it is in orbit around a star, has a certain amount of mass, a round shape, and is a dominant object in its orbit. All these marks belong to the intension or content of the concept of planet. The concepts of Mercury, Earth, Mars, Jupiter etc. contain these marks, ergo they belong to the extension of the concept of planet, or under this concept. As a rule, the intension and extension of a concept “stand in inverse relation to one another. The more a concept contains under itself, namely, the less it contains in itself, and conversely” (Log, AA 09: 95). The concept of celestial bodies has a greater extension than the concept of planets, as it can be applied to more objects, but it is at the same time ‘thinner’, it contains fewer concrete marks that are common to multiple objects. The interrelation of marks, the intension and extension of concepts enable coordination and subordination of concepts. This is what we do when we cognize things, and this is what we do when we judge and infer.

3 Concepts, Judgments, and Inferences

A judgement, for Kant, is “the representation of the unity of the consciousness of various representations, or the representation of their relation insofar as they constitute a concept” (Log, AA 09: 101). An inference is “the derivation of one judgment from the other” (ibid., 114) with or without an intermediate concept. Judgements and inferences are therefore cognitive operations with content and extension of concepts, i.e., with representations and marks that are contained in and under a concept. Some scholars, such as Sluga (1980) and Brandom (1994, pp. 79-82), have suggested that historical roots for the Fregean concept-judgement-inversion or the thesis of priority of propositions over concepts can be found in Kant. While Lemanski recently casted some doubts on this and offered a different historical reconstruction, I will suggest four points that follow from the theoretical reconstruction of Kant’s theory of concepts (see Lemanski 2021, pp. 116-170 and (2013)). Brandom cites a passage from the first Critique where Kant claims that judgements are the genuine operations of the understanding and “the understanding can make no other use of […] concepts than that of judging by means of them” (KrV, A 68/B 93). But this passage does not give much plausibility to Brandom’s conclusion that for Kant, “any discussion of content must start with the contents of judgments, since anything else only has content insofar as it contributes to the contents of judgments” (Brandom 1994, p. 80). Firstly, as Kant claims here, concepts have a generative priority. We cannot form any judgement without representations that form a concept (which is itself also a representation). Representations, including concepts, are the data of all cognitive operations. Secondly, there is no passage in Kant, where he would claim that concepts are denoted only by singular terms. Propositions and sets of propositions, such as definition or an inference, express one concept as a reflected set of combined representations. This follows from the mentioned definitions of judgements and inferences and corresponds to Kant’s own procedures – there is no discrepancy between saying and doing. For example, the set of propositions

a science of reason, not as to mere form but also as to matter; a science a priori of the necessary laws of thought, not in regard to particular objects, however, but to all objects in general; – hence a science of the correct use of the understanding and of reason in general, not subjectively, however, i.e., not according to empirical (psychological) principles for how the understanding does think, but objectively, i.e., according to principles a priori for how it ought to think (Log, AA 09: 16)

expresses one concept, “the concept of logic” (ibid.). Thirdly, judgments deal with the content and extension of concepts. If logic is the subject matter of philosophical investigations, then the primacy is given to the analysis of the form of judgments, types of judgments, and fallacies regardless of the conceptual content. But outside of logic, in natural sciences, epistemology, ethics etc. rational agents use judgements to articulate intensions and extensions of concepts. For example, logic determines that ‘all bodies are divisible’ is a categorical judgement, but the actual matter of judgement is the assertion that a certain mark, ‘divisibility’, necessarily belongs to a certain concept, the concept of body. In the universal judgement “all human beings are mortal” the rational agent claims that the concept of human beings falls entirely under the extension of the concept of mortality. Seen from this angle, the content of concepts has more relevancy than the form in which this content is being dealt with.

These three points prove that it is highly problematic to speak of a primacy of judgments over concepts in Kant. Concepts – as it is widely believed at the time of Kant – are extralinguistic entities, general mental representations that can be articulated in form of singular terms, propositions, and sets of propositions.[13] The correct use of words, precise naming and flawless construction of judgments and inferences are very important to Kant, but these are just different instruments for dealing with concepts.

4 Kant’s Theory of Definition

Against Brandom’s claim that in Kant, “any discussion of content must start with the contents of judgments” (Brandom 1994, p. 80) speaks another, fourth point: Kant’s theory of definitions and ‘given concepts’. Apart from an over six decades old article by Lewis White Beck, one can scarcely find attempts at a systematic reconstruction of this theory in Kant scholarship.[14] Important feature of Kant’s theory of definitions is the view that different kinds of concepts require different kinds of definitions. Kant subdivides concepts into given and made. Made concepts result from either pure or empirical synthesis. Concepts of mathematical objects are produced via pure intuitive synthesis. A rational agent can draw a triangle in her imagination or on the paper, combine its different parts into one figure, experiment with it, and “exhibit originally the exhaustive concept” of the triangle “within its boundaries” (KrV, A 727/B 755), i.e., give what Kant calls ‘definition’ in the narrower sense. “Definitions”, in Kant’s understanding, “have the highest degree of analytic distinctness, they are the final purpose of all our concepts” (V-Lo/Wiener, AA 24: 844). They occur only in mathematics, where the concepts are synthesized entirely a priori. It is, following Kant’s idiosyncratic view, wrong to speak of definitions in philosophy – for philosophers, a definition is an ideal that one seeks to achieve at the end of an investigation. What philosophers produce are rather ‘imperfect definitions’, which he suggests calling ‘explications’ and ‘expositions’.

Besides of concepts of mathematical objects, concepts of invented empirical objects are being made. Like the pure synthesis, a posteriori synthesis starts with parts that are being put together to one object, e.g., the chronometer or a Minotaur. An inventor or a fantasist does not define such concepts, she simply declares what the concept of her invention or fantasy contains.

The counterpart to made concepts are given concepts.[15] The most concepts that rational agents articulate belong to this category. As concerns made concepts, “the parts of the cognition precede the whole cognition”; in case of given concepts, however, “the whole precedes the parts” (V-Lo/Wiener, AA 24: 845). The mentioned concepts of flower and planet are examples of empirical given concepts. To understand their intension and extension we have to analyse them, i.e., break them down in parts and find out which marks they contain. Kant calls this procedure explication. A rational agent makes the implicit components of a concept explicit, or what she understands by a certain term, proposition, or set of propositions. In Kant’s view, such procedure can never “exhibit originally the exhaustive concept within its boundaries” (KrV, A 727/B 755) as rational agents can make new observations and distinctions and disagree on which marks necessarily belong to a concept (KrV, A 728/B 756).[16]

Similarly,

no concept given a priori can be defined, e.g., substance, cause, right, equity, etc. For I can never be certain that the distinct representation of a (still confused) given concept has been exhaustively developed unless I know that it is adequate to the object. But since the concept of the latter, as it is given, can contain many obscure representations, which we pass by in our analysis though we always use them in application, the exhaustiveness of the analysis of my concept is always doubtful, and by many appropriate examples can only be made probably but never apodictically certain (KrV, A 728f./B 756f.).

Hence, Kant suggests calling the procedure to acquire analytic distinctness in case of a priori given concepts exposition. Rational agents expound marks of a given concept, which concern an object that is either not or at least not entirely given in experience. In this context, the genuine and only source of marks is the understanding – “we cannot go out of the understanding and seek elsewhere” (V-Lo/Wiener, AA 24: 917). An example is the concept of virtue, which contains the marks: “first – lawfulness for actions, then lawfulness for right actions, then in addition freedom. {Which stands in conflict with the inclinations […]}” (V-Lo/Dohna, AA 24: 758).

Ergo, the claim that philosophers first look at the content of judgement does not correctly render the procedure that Kant describes. Philosophers start with an empirical or pure given problematic concept, which may be articulated in form of a term, a proposition, or a set of propositions, and try to elucidate it by analysis of marks, i.e., the network of representations which it consists of.

5 Analysis as the Proper Method of Dealing with Concepts in Philosophy

The distinction between made and given concepts reflects the difference between synthesis and analysis. Kant states: “I cannot explain virtue synthetically. For I am supposed to say what we all think under the concept of virtue, not what I perhaps understand under this concept in accordance with my own caprice” (V-Lo/Wiener, AA 24: 845). The synthetic method of combination of parts of a concept into a whole is suitable for mathematicians. Kant calls it “cognition from the construction of concepts” (KrV, A 837/B 865) – a mathematician can construct a concept by using the pure forms of intuition. Synthesis is also the right method for inventors – they can use sensible intuition to explicate which parts of the invented object they have combined. In both cases, either the pure or sensible intuition goes hand in hand with the generation of concepts.

This is different with discursive cognition from given empirical or pure concepts. The concept of planet contains an integrated whole of various representations, but I do not have the parts prior to the whole, neither in empirical intuition nor in an intellectual intuition (as in case of an intellectus archetypus) nor in my understanding. The faculties of the understanding and the sensibility must work hard to help a rational agent explicating all necessary marks of the concept of planet, either by enumeration of marks or by judgements and inferences. Two or more rational agents must fixate that they are speaking of the concept of one and the same object and that their analysis corresponds to the object in experience. In this way, there is no room for a personal ‘caprice’, i.e., free creation of concepts from synthesis of arbitrarily chosen different parts, there is no room for ‘concept-poetry’ in rational cognition from concepts. The analytic method is the one that brings necessity and clarity into given concepts.

Philosophy is not so much concerned with given empirical concepts as with concepts given a priori, such as causality, force, virtue, freedom, reason, and philosophy. Kant insists that no object in experience corresponds to such concepts, ergo – they cannot be arbitrarily generated out of material of empirical intuition. Three humans with the predicate ‘wise’ do not render the concept of wisdom. And five examples of existing philosophies do not give the idea of philosophy – they are mere ectypes of a pure archetype. Also, we do not possess the capacity for intellectual intuition that would give us all the parts of a concept at once. For Kant, theorists of intellectual intuitions want to free themselves from the hard discursive work with concepts, with the arduous step-by-step exposition of marks – a procedure, that would sooner or later lead to “the death of philosophy” (VT AA 08: 398). Thus, as Kant claims, “the whole of metaphysics and morals has to do with analytic definition” – “these two sciences are the true objects” (V-Lo/Wiener, AA 24: 917) of analysis, although definitions remain the ideal of cognition and one must content oneself with expositions. There is no room for arbitrary synthesis or ‘concept-poetry’ in rational cognition from pure concepts either.

6 Is Philosophy Only About Analysis?

A careful and not easily convincible reader may have the following objections: (1) philosophy is not only about analysis; (2) Kant also speaks of a synthetic method; (3) we do not analyse when we form synthetic judgements; (4) analysis can lead to dogmatic metaphysics, the object of Kant’s critique, and ergo cannot always be the proper method of dealing with concepts in philosophy; and (5) Kant himself does not seem to explicitly apply the method of conceptual analysis. The answers, however, give the defended thesis even more colour and contrast.

(1) Philosophy, in Kant’s view, is “rational cognition […] from concepts” (KrV, A 837/B 865), it deals with given concepts, and such concepts are subject of analysis. But conceptual analysis is not the only mark of the concepts of philosophy and philosophical method. In other words, conceptual analysis is the proper method of philosophizing, but it is not identical with the intension and extension of the concept of philosophy. Some other necessary marks are ends of reason, rational, science, singularity, and systematicity.[17] None of them contradicts the mark ‘conceptual analysis’ – it is part of one systematic and coherent conceptual unity.[18] For example, Kant counterposes rational cognition against historical cognition (cf. KrV, A 835-7/B 863-5). Historical cognition of philosophy, e.g., of the Kantian philosophy, is, considered objectively, rational cognition. But subjectively, it is mere historical knowledge of a given fact. The philosopher has done the work for his reader – all the marks of the concepts that interest her have been explicated or expounded. She can grasp and learn them well, which will spare her a lot of time. She does not have to perform the analysis herself. Philosophical cognition as rational cognition from concepts requires a self-wrought analysis – the givenness must be taken as a problem and not as ready fact.

(2) One of the tasks of philosophy is to order its cognitions, which it obtains via analysis of concepts, into one coherent system. And here, as Kant suggests, one can make use of a synthetic method, which he also calls progressive method (see Log, AA 09: 149; cf. Prol AA 04, 263f.). This method was used in the first Critique – elements of cognition, sensibility, understanding, and reason have been investigated and ordered step-by-step into the whole of elements of transcendental philosophy. In the Prolegomena, Kant follows the analytic or regressive method – the whole of elements of transcendental philosophy was taken as a given fact. Hence, this distinction between progressive and regressive method concerns the matter of presentation of a system of cognitions, not a concrete cognition, which, in philosophy, is a product of conceptual analysis.

(3) But do philosophers also analyse when they produce synthetic judgements? The only answer coherent with the previous reconstruction is that in case of critical philosophical cognition, synthesis can be either only an auxiliary method of analytic investigation or the object of analysis. Philosophers can analyse a judgement by suggesting and examining different synthetic predicates, as in case of hypothetic reasoning, thought experiments, or polemic (e.g., to make in controversial arguments about objects of traditional theoretical metaphysics the opponent aware that her position is as dogmatic as a counter position – cf. KrV, A 769-82/B 797-810). Such synthesis is a conscious deliberative act, but the superordinate question is analytic: Is one concept part of intension and extension of the other concept? Does all that happens, happens due to a cause? Are humans good or evil by nature?

Analytic and synthetic judgments are objects of analysis alike. The predicate ‘unmarried man’ in the analytic judgement “bachelor is an unmarried man” is the explication of the concept of bachelor. It makes, unlike the tautology “bachelor is a bachelor”, the concept of bachelor more explicit, explicates the implicit identity, and hence contributes to cognition of this concept (cf. Log, AA 09: 111). The synthetic judgement “all bachelors are virtueless” suggests that the subject ‘all bachelors’ falls entirely under the extension of the predicate ‘virtuelessness’. The philosopher can intervene and claim that the concept “all bachelors are virtueless unmarried men” is wrong by examining counterexamples in experience (“some men are very virtuous yet unmarried”) and analysing the concepts of marriage and virtue. Philosopher, as any rational agent, encounters and generates synthetic judgements, but her job is to question and analyse them.

(4) But is analysis not also the method of dogmatic metaphysicians, and ergo partially flawed? This question itself is an example of an object of philosophical analysis. Does the concept of analysis include the concept of partially flawed performance of analysis? This is obviously a synthetic and non-necessary mark. The performance of analysis is, to focus on one example, neither dependent on nor linked to a possible hypostasis or the fallacy of subreption that Kant discusses in the Paralogisms chapter of the first Critique. Every concept consists of multiple representations. Ideas of reason, as Kant states in the passage on the progression of representations, are “made up of notions” (KrV, A 320/B 376f.). The analytic exposition of these notions does not amount to the assertion of cognition of a real object. Hence, the notions ‘the simple’, ‘the substantial’ etc. add nothing to the cognition of the soul, but they can be nevertheless the result of an analysis – as possible constituents of the concept soul, without which it would be not only empirically, but also logically empty. It is, ergo, “entirely permissible to think the soul as simple” (KrV, A 771/B 799) and

one can quite well allow the proposition The soul is substance to be valid, if only one admits that this concept of ours leads no further, that it cannot teach us any of the usual conclusions of the rationalistic doctrine of the soul, such as, e.g., the everlasting duration of the soul through all alterations, even the human being’s death, thus that it signifies a substance only in the idea but not in reality (KrV, A 351).

Even in such borderline cases of transcendental ideas, the task of the critical philosopher is to find out which marks have been thought and must be thought in the given concepts of soul, world, and God. Kant does it throughout the Transcendental Dialectic and assigns all 12 categories to the logical content of the transcendental ideas.[19] There is no room for concept-poetry in this regard either, as the aim of the critical philosopher is not fabrication of new concepts via synthesis of marks with the freedom of a fictioneer: “The idea cannot be attained by composition, […] for the whole is prior to the part” (Log, AA 09: 92). Kant goes even further. Transcendental ideas are concepts given not only historically by the community of traditional metaphysicians, but by the nature of reason that is disposed to metaphysics (KrV, B 21f. and Prol, AA 04: 362-4). The historical discovery of such concepts is in truth their discovery in our own reason[20] – as there is only one reason, the historian of philosophy is an “archaeologist of reason” (cf. FM/Lose Blätter, AA 20: 341). The interplay between the natural rational givenness and analysis of what has been given is also relevant for the “indispensably necessary regulative use” (KrV, A 644/B 672) of transcendental ideas. The mark ‘simple’ belonging to the concept of soul is relevant for the regulative conception of unity of faculties and forces of the mind; the mark ‘purposiveness’ belonging to the content of intellectus archetypus (original reason as the source of the planful creation) can be helpful for teleological theories and observations; the mark ‘infinity’ for the ongoing exploration of space.

Hence, Kant’s account of conceptual analysis is in line with his project of critique: no innatism, no concept-fabrication, no arbitrary synthesis of marks – analysis of given concepts.[21]

(5) The signpost does not follow the path to which it points – does Kant follow the method of conceptual analysis, which he has identified as the general philosophical method? A somewhat misleading passage at the beginning of the Transcendental Analytic, The Analytic of Concepts of the first Critique seems to cast doubt on this:

I understand by an analytic of concepts not their analysis, or the usual procedure of philosophical investigations, that of analyzing the content of concepts that present themselves and bringing them to distinctness, but rather the much less frequently attempted analysis of the faculty of understanding itself, in order to research the possibility of a priori concepts by seeking them only in the understanding as their birthplace and analyzing its pure use in general; for this is the proper business of a transcendental philosophy; the rest is the logical treatment of concepts in philosophy in general (KrV, A 65-66/ B 90-91).

Does it mean that transcendental philosophers can examine the faculty of understanding without the mediating service of concepts? Do they inquire into the faculty of understanding or into the concept of the faculty of understanding? At this point, Kant introduces a difference between the analytic of concepts and analysis of concepts. The goal of the transcendental analytic is to analyze (Kant uses the German word zergliedern, to dissect) “the entirety of our a priori cognition into the elements of the pure cognition of the understanding” (KrV, A 64/ B 89). This is a procedural remark that provides the rationale for a particular project within the transcendental philosophy. The name of the partial project, ‘analytic of concepts’, requires a terminological clarification, which is provided in the second to last quote: ‘analytic of concepts’ designates a certain part and procedure in the transcendental philosophy, while ‘analysis of concepts’ refers to the philosophical method in general. Furthermore, the passage is implicitly about pure concepts, i.e., the twelve categories of the pure understanding, and not about concepts in general. The analytic of concepts as a part of the program of transcendental philosophy provides “the less frequently attempted” analysis of the faculty of understanding – which means, not the analysis of pure concepts per se, such as ‘causality’ or ‘substance’, as in traditional ontology, is in the focus of transcendental philosophers. The object of analysis is the pure faculty of the understanding or the entirety of the elements of the pure cognition of the understanding. This object is not given directly in a sensible or (the in Kant’s view impossible) intellectual intuition. It is given as an “idea of the whole” (ibid.), i.e., according to the taxonomy of representations, as a concept. Ideas are a priori given concepts and such concepts are object of analysis. Hence, what Kant does is entirely in line with his view of conceptual analysis as the proper philosophical method in the Doctrine of Method and the corresponding methodological passages in his logic lectures. He exposes pure elements, categories and principles, as marks of the idea of pure faculty of the understanding. It would be an absurdity to claim that Kant dissects the faculty of the understanding itself and not its concept.

7 Conclusion

As I have shown, for Kant, the medium of rational cognition are representations, which make up the structure and content of a concept. Concepts originate either from comparison of representations, reflection, and abstraction or from additional operations, such as inference or synthesis, or from the nature of the understanding and reason. Concepts are being articulated either in form of singular terms, or propositions, and sets of propositions. Rational agents deal with intension and extension of concepts by focusing on marks, i.e., representations that are contained within or under a concept and serve as grounds of cognition of its content and extension. Concepts are either made or given – in the former case parts are given prior to the whole, in the latter the whole is given prior to parts. Made concepts are objects of mathematicians and inventors. They can be synthetically produced in recourse to either pure or empirical intuition. Mathematicians define their concepts, inventors declare them. Given concepts are either empirical or pure concepts. Rational agents explicate the marks of the former using observations of given objects and expound the interrelated marks of the latter by the understanding and reason alone. As philosophy is rational cognition from concepts, and neither an invention of concepts based on arbitrarily combined intuitions nor a cognition of all parts given entirely in an intellectual intuition, the only proper method of philosophy is analysis. Philosophical cognition bears entirely on conceptual analysis. Philosophy is inherently analytic.

Relating to what has been said in the introduction, Kant’s concept of philosophical method may cause a dilemma for theorists and historians of analytic philosophy. If ‘analytic philosophy’ is “a phrase in a living language” and “the attempt to stipulate a sense for it that excludes many” (Williamson 2022, p. 23) philosophers causes nothing more than a terminological confusion, two options are left: One must either count Kant in, because he has given one of the clearest accounts of philosophy as conceptual analysis or must at least admit that the label ‘analytic philosophy’ is too vague and empty. The demand to specify, replace or abandon it is justified.

8 Literature

Beaney, M. (2007). “The Analytic Turn in Early Twentieth-Century Philosophy”. In: The Analytic Turn. Analysis in Early Analytic Philosophy and Phenomenology, Ed. M. Beane. New York: Routledge, 1–30.

Beck, L. W. (1956). “Kant’s Theory of Definitions.” The Philosophical Review 65, 2, 179–91.

Brandom, R. B. (1994). Making It Explicit. Reasoning, Representing, and Discursive Commitment. Cambridge/London: Harvard University Press.

Bunte, M. (2022). “Determining and Grounding: The Twofold Function of the Transcendental Dialectic”. Open Philosophy 5, 1, 396–402.

Bunte, M. (2016). Erkenntnis und Funktion. Zur Vollständigkeit der Urteilstafel

und Einheit des kantischen Systems. Berlin/Boston: Walter de Gruyter.

Caimi, M. (2012): “Application of the Doctrine of Method in the Critical Examination of Reason”. Studia Kantiana 13, 5-17.

Cicatello, A. (2018). “Der seltsame Fall der ‘Gegebenen Begriffe’ bei Kant.” In: Natur und Freiheit: Akten des XII. Internationalen Kant-Kongresses. Ed. V. L. Waibel, M. Ruffing and D. Wagner. Berlin/Boston: De Gruyter, 1183–90.

Conrad, E. (1994). Kants Logikvorlesungen als Neuer Schlüssel zur Architektonik der Kritik der reinen Vernunft. Die Ausarbeitung der Gliederungsentwürfe in den Logikvorlesungen als Auseinandersetzung mit der Tradition. Stuttgart/Bad Cannstatt: Frommann-Holzboog.

Dummett, M. (1978). Truth and Other Enigmas. London: Duckworth.

Forster, M. (2012). “Kant’s Philosophy of Language?” Tijdschrift voor Filosophie 74, 3, 485–511.

Friedman, M. (1999). Reconsidering Logical Positivism. Cambridge: Cambridge University Press.

Friedman, M. (1994). “Kant and the Twentieth Century”. In: Kant and Contemporary Epistemology. Ed. P. Parrini. Dordrecht: Springer, 27–46.

Ginsborg, H. (2006). “Thinking the Particular as Contained under the Universal”. In: Aesthetics

and Cognition in Kant’s Critical Philosophy. Ed. R. Kukla. Cambridge.

Glock, H.-J., 2008. What is Analytic Philosophy? Cambridge: Cambridge University Press.

Haag, J. and Perler, D. Eds. (2010). Ideen. Repräsentationalismus in der Frühen Neuzeit. Berlin/Boston: De Gruyter.

Hanna, R. (2001). Kant and the Foundations of Analytic Philosophy. Oxford: Oxford University Press.

Hinske, N. (1999). Tra Illuminismo e Critica Della Ragione. Studi sul Corpus Logico Kantiano. Pisa: Scuola Normale di Pisa.

Kant, I. (1992 ff.) Cambridge Edition of the Works of Immanuel Kant. Ed. P. Guyer, A. Wood.

Cambridge.

Klimmek, N. F. (2005). Kants System der transzendentalen Ideen. Berlin/New York: Walter de Gruyter.

Laiho, H. A. (2019). “Types of Representational Content in Kant”. Kantian Journal 38, 1, 31–54.

Lemanski, J. (2021). Welt und Logik. London: College Publications.

Lemanski, J. (2013). “Die neuaristotelischen Ursprünge des Kontextprinzips und die Fortführung in der fregeschen Begriffsschrift“. Zeitschrift für philosophische Forschung 67, 4, 566–586.

Lewin, M. (2023). “Introduction to ‘Kant and Metaphilosophy’”. Kantian Journal 42, 3. http://dx.doi:10.5922/0207-6918-2023-3-1.

Lewin, M. (2022). “The Faculty of Ideas. Kant’s Concept of Reason in the Narrower Sense”. Open Philosophy 5, 1, 340-359. https://doi.org/10.1515/opphil-2022-0203.

Lewin, M. (2021a). Das System der Ideen. Zur perspektivistisch-metaphilosophischen Begründung der Vernunft im Anschluss an Kant und Fichte. Freiburg/München: Alber. https://doi.org/10.5771/9783495825242.

Lewin, M. (2021b). „Kant’s Metaphilosophy”, Open Philosophy 4, 1, 292-310. https://doi.org/10.1515/opphil-2020-0190.

Lewin, M. (2020a). “Reason, Ideas, and Their Functions in Classical German Philosophy”. Vestnik of Saint Petersburg University. Philosophy and Conflict Studies 36, 1, 4-23. https://doi.org/10.21638/spbu17.2020.101.

Lewin, M. (2020b). “The Universe of Science. The Architectonic Ideas of Science, Sciences and Their Parts in Kant”. Kantian Journal 39, 2, 26-45. https://doi.org/10.5922/0207-6918-2020-2-2.

Lewin, M. and Williamson, T. (2023). “Kant and Analysis”. Kantian Journal 42, 3. http://dx.doi:10.5922/0207-6918-2023-3-3.

Lu-Adler, H. (2018). Kant and the Science of Logic: A Historical and Philosophical Reconstruction. Oxford: Oxford University Press,

Martinez, L. (2019). “La Doctrina Kantiana de la Definición en las Lecciones de Lógica”. Anales Del Seminario de Historia de la Filosofía 36, 3, 683–704.

McAndrew, M. (2022). “Kant’s Theory of Concept Formation and his Theory of Definitions”. Archiv für Geschichte der Philosophie. https://doi.org/10.1515/agph-2020-9002

Pippin, R. B. (1982). Kant’s Theory of Form: An Essay on the Critique of Pure Reason. New Haven.

Quine, W. V. O. (1951). “Two Dogmas of Empiricism”. In: Philosophical Review 60, 1, 20-43.

Schlösser, U. (2013). “Concept Formation, Synthesis and Judgment: Kant’s Theory of the Logical and Cognitive Activities of the Mind”. Self, World, and Art: Metaphysical Topics in Kant and Hegel. Ed. D. Emundts. Berlin/Boston: De Gruyter, 177–206.

Schumann, G. (2019). „Begriffsanalyse“. In: Handbuch Erkenntnistheorie. M. Grajner und G. Melchior Hg. Stuttgart: Metzler, 383–390.

Sgarbi, M. (2016). Kant and Aristotle. New York: SUNY.

Sluga, H. D. (1980). Gottlob Frege. The Arguments of the Philosophers. London/New York: Routledge.

Strauß, G. (1996). „Wort – Bedeutung – Begriff: Relationen und ihre Geschichte“. In: Bedeutung – Konzepte – Bedeutungskonzepte. Theorie und Anwendung in Linguistik und Psychologie. J. Grabowski, G. Harras und T. Herrmann Hg. Opladen: Westdeutscher Verlag, 22–46.

Stuhlmann-Laeisz, R. (1976). Kants Logik: Eine Interpretation auf der Grundlage von Vorlesungen, veröffentlichten Werken und Nachlaß, Berlin/Boston: De Gruyter.

Wang, W. (2021). “Kant on the Formation of Empirical Concepts”. Kant-Studien 112, 2, 195–216.

Williamson, T. (2022). The Philosophy of Philosophy. Second Edition. Hoboken and Chichester: Wiley-Blackwell.

[1] Michael Lewin is Senior Researcher at the Academia Kantiana, Immanuel Kant Baltic Federal University. He can be reached at mlewin@kantiana.ru. The article is based on a keynote lecture at the conference “Existence, Cognition, Action: Kant’s Legacy for the 21st Century” at the University of Belgrade, March 1-3, 2023. I want to thank the organizer Andrija Šoć for the kind invitation and participating colleagues for their feedback and questions, some of which I have used for the section 6 of this article. I also want to thank the two anonymous reviewers for their helpful remarks. This research was supported by the Russian Federal Academic Leadership Program Priority 2030 at the Immanuel Kant Baltic Federal University.

[2] I will address the thesis of concept-judgement-inversion in Kant that has been proposed, among others, by Sluga (1980) and Brandom 1994, pp. 79-82, and recently contested by Lemanski 2021, pp. 116-170 and (2013) in the third section. Quine’s (1951) “Two Dogmas of Empiricism” had a lasting and partially transformative effect on analytic philosophy. Some even speak of a “crisis that arises from the very unsettling fact that many and perhaps even most analytic philosophers now question the defensibility and ultimate intelligibility of the very idea of analysis. But how can there be analytic philosophy without a cogent and coherent conception of philosophical analysis?” (Hanna 2001, p. 11). As Hanna points out, Friedman 1994, pp. 44f., and 1999, pp. 1-14, traced the source of this crisis back to analytic philosophy’s rejection of Kant. Within the scope of current metaphilosophical investigations, a synoptic account of Kant’s method of conceptual analysis serves both a better understanding of Kant’s concept of philosophy and a self-reflection of analytic philosophy.

[3] For the distinction between analysis as decomposition and replacement see Beaney (2007). For a recent topical introduction into conceptual analysis see Schumann (2019). For the problem of defining ‘analytic philosophy’ see Lewin and Williamson (2023), Williamson (2022), and Glock (2008).

[4] Kant’s concept of logic is rooted in the Aristotelean tradition and the historical context of Port-Royal Logic, Leibniz, Wolff, Baumgarten, and Meier (whose “Excerpt from the Doctrine of Reason” he has used for his lectures). For Kant’s concept of logic and the historical background see Lu-Adler (2018); Sgarbi (2016); Hinske (1999); Conrad (1994), and Stuhlmann-Laeisz (1976), For my theoretical reconstruction of elements of Kant’s theory of conceptual analysis, I use the Wiener Logic (1780ff., based on lectures at the time when the first Critique was prepared for publication), the Dohna-Wundlacken Logic (a transcript of lectures held in summer semester 1792); and the Jaesche Logic (1800, the only partially authorized manuscript). All cited passages have been carefully chosen and compared with other corresponding passages in the transcripts as well as in the first Critique, Prolegomena, and other published texts. In cases of doubt preference is given to published works and systematic theoretical coherence. All citations of Kant’s works in this article are from the Cambridge Edition of the Works of Immanuel Kant (Kant 1992ff.).

[5] I proceed in a different way than Caimi (2012), who has used several elements of Kant’s doctrine of method to reconstruct Kant’s critical examination of reason. Firstly, I focus on Kant’s method of conceptual analysis. As the short formula “rational cognition […] from concepts” (KrV, A 837/B 865) and its explanation suggest, analysis of concepts is what philosophers, in Kant’s understanding, do. Secondly, I keep the topic on a more general level in the sense that I do not attempt at reconstructing, e.g., Kant’s first Critique for exegetical purposes. However, as I show throughout the article and especially in part 6, Kant’s specific description of the method of conceptual analysis is in line with what he is doing. For the significance of the heuristic distinction between Kant’s conception of philosophy in general and his conception of transcendental philosophy, see Lewin (2021b). For application of Kant’s method to his conceptions of philosophy and reason see ibid. and Lewin (2022).

[6] I agree with Laiho 2019, p. 33, that from this exposition immediately follows that intuition (just as sensation) is a non-conceptual representation. I will not further engage in the ongoing debate on conceptualism vs. non-conceptualism as it leads little value to the intended synoptic account of analysis of given concepts.

[7] For a historical study on representationalism (that soon will be extended, in an editing project of other authors, to Kantian and post-Kantian philosophy) see Haag and Perler (2010).

[8] I add this specification, because Kant claims that practical ideas (e.g., wisdom and virtue) can be partially given or as they are practical conditions for acting, put into effect in experience (cf. KrV, A 569/B 597 and A 328/B 385). Moreover, in several passages Kant claims that freedom (in the guise of pure will) is the only transcendental idea, whose objective reality can be proven (cf. Log, AA 09: 93; KpV, AA 05: 4; and KU, AA 05: 468). For the different types of ideas in Kant and their functions see especially Lewin 2021a, pp. 41-113, but also Lewin 2020a, 2020b, and Lewin 2022.

[9] For more about the Kantian concept of empirical concept formation see McAndrew (2022), Wang (2021), Schlösser (2013), Ginsborg (2006), and Pippin (1982). For an extensive historical exegesis see McAndrew (2021).

[10] This is clear from the theoretical standpoint alone (abstraction would make pure concepts dependent on sensibility; combination would either render made concepts, which contradicts the theory of givenness of pure non-mathematical concepts or lead to concepts that are not pure enough (two examples of different wise humans cannot render the idea of wisdom)). For the corresponding passages cf., inter alia, V-Lo/Blomberg, AA 21: 253; Kant to Reinhold, 19.05.1789 – Br, AA 11: 40; and Log, AA 09: 92.

[11] McAndrew’s 2022, pp. 4-9 and pp. 11-16, reconstruction of Kant’s theory of concept formation is insofar incomplete, as the author does not consider and give much weight to the difference between empirical and pure (given) concepts, which is needed to recognize this dilemma.

[12] See section 4.

[13] Strauß 1996, pp. 23-25, associates Kant’s views with the so-called ‘classical dyadic’ linguistic paradigm, which has been prevalent in the 18th and 19th centuries (proponents are, e.g., the lexicographers and semasiologists Adelung, Reisig, Haase, and Heerdegen). This paradigm prioritizes the (extralinguistic) concept over the word (significant). For an introductory article on Kant’s philosophy of language, see Forster (2012).

[14] See Beck (1956). Beck tried to classify definition, exposition, explication, and declaration as real definitions. For Kant, in his logic lectures, real definitions aim at necessary marks, the essence of things, whereas nominal definitions concern the use of words. There is not enough textual evidence to support Beck’s classification attempt and Kant did not seem to give the real-nominal-distinction (that is rooted in the traditional logic) much weight. Moreover, in the first Critique, Kant has interlinked explication with nominal explanation (see KrV, A 728/B 756 and Log, AA 09: 143), and pointed at the problematic status of real definitions (see Log, AA 09: 143). I drop the real-nominal-distinction in this article, as it would add more confusion than benefits. For a reconstruction of Kant’s theory of definitions in an earlier phase (1770-1782), see Martinez (2019). For a very recent reconstruction attempt, see McAndrew 2022, pp. 11-16.

[15] I agree with Cicatello (2018) that the Kantian account of givenness has nothing to do with innatism. It seems rather to presuppose an unconscious synthesis. I think that Kant’s doctrine of the nature of reason (see, e.g., KrV A VII, A 298/B 354, A 339/B 397, A 669/B 697 and Prol, 04: 328, 331, and 353) could shed light on this problem. If the synthesis is performed naturally, by the spontaneity of the understanding and reason, the rational agent sees her being confronted with ready-made concepts. They appear as given because she is conscious of the whole prior to the understanding and reconstruction of the parts. A concept can be given by the own or alien reason before it becomes subject of analysis. In case of empirical concepts, the concept can appear as given because it reflects a complex sensible intuition, which is given as the whole prior to analysis of the parts (e.g., the whole of the planet in the sensible intuition before the consciousness of singular representations of which it consists). Kant seems to believe that rational agents err less if they perform analysis of pure and empirical given concepts – analysis leads to necessary truths, while synthesis of the same concepts would add the dimension of arbitrariness and “caprice” (V-Lo/Wiener, AA 24: 845) of different rational agents. This is an answer that satisfies me, but I will not further discuss it here.

[16] McAndrew 2022, pp. 13-15, points at a possible exegetical problem. Kant seems, at some passages, to consider a synthesis in relation to empirical concepts. McAndrew cites a passage from the Vienna Logic that states that to find out which marks the concept of metal contains, one turns to experience, finds, and adds certain features to the concept (see V-Lo/Wiener, AA 24: 913). In such cases, as McAndrew suggests in correspondence with this passage, given empirical concepts can be “classified as made once one has expanded or revised their content” (ibid., 13). I do not think that it directly follows from this passage. Kant states that it is possible to make the concept a posteriori, but the core of the concept is still given a posteriori. Hence, Kant is not concerned with a re-classification of the status of a concept, but with the right procedure to deal with marks of a given empirical concept. In the first Critique and the later logic lectures Kant does not use the terminus ‘making’ in relation to (non-invented) empirical concepts, although he describes the procedure of seeking marks in experience and through experiments. He seems to realize that this addition confuses and contradicts his theory of definitions, which unambiguously states that given concepts are subject of analysis, while made concepts are subject of synthesis. Kant seems to revise the whole theory of given and made concepts as he trespasses the line from the pre-critical to critical phase. For example, in the Blomberg Logic, he claims that ideas of reason are result of fabrication (see V-Lo/Blomberg, AA 24: 254 and cf. Lo/Wiener, AA 24: 914f.), which contradicts the theory of givenness of pure concepts and the corresponding passages in the first Critique and later logic lectures. In the section 6, I suggest a different interpretation of possible synthetic additions to given concepts, which is coherent with Kant’s theory of givenness – empirical exploration of marks and testing are to be regarded as subordinated methods of genuinely analytic inquiry.

[17] For the collection of marks and an attempted ‘imperfect definition’ of philosophy in Kant see Lewin (2021b) and the topical issue on Kant’s metaphilosophy Lewin (2023).

[18] The manifold of marks constitutes one logical sphere of a concept, if the different manifold agrees (Log, AA 09: 146).

[19] For the discussion of this point see Klimmek 2005, p. 69; Bunte 2016, pp. 154 ff.; Lewin 2021a, pp. 106-111, and the response in Bunte 2022, p. 399, fn. 14.

[20] By introducing the theory of ‘given concepts’ Kant is able to avoid the rationalists’ innatist conception of concepts – see the footnote 14 of this article.

[21] For examples of critical applications of the Kantian conceptual analysis to his ideas of philosophy and reason see Lewin (2021b) and (2022). For Kant’s metaphilosophy see also the topical issue Lewin (2023).