Kant’s Philosophy of Language of Philosophy: On Philosophical Terminology

1.1 The search for appropriate terms

Towards the end of 1773, in reference to what would end up becoming the Critique of Pure Reason (CPR), Kant wrote to his former student and trusted correspondent Marcus Herz the following words:

I doubt that many have tried to formulate and carry out to completion an entirely new conceptual science. You can hardly imagine how much time and effort this project requires, considering the method, the divisions, the search for exactly appropriate terms.^[1]

In the CPR, published almost eight years later, Kant wrote:

In the great wealth of our languages, the thinking mind nevertheless often finds itself at a loss for an expression that exactly suits its concept, and lacking this it is able to make itself rightly intelligible neither to others nor even to itself.^[2]

We may ask: what is it, in Kant’s view, that makes a term appropriate or suitable for a philosophical concept?^[3] The answer to this question, as Kant notes, is especially challenging for intelligibly developing new philosophical thoughts. This makes the topic potentially attractive and exciting for anyone considering an original contribution to the discipline but struggling to find a way of expressing their thoughts. From our historical vantage point, Kant himself has come to be considered one of the most innovative individual contributors to philosophy. A reading of his insights regarding the language of philosophy is warranted, if only by its historical significance.

It is commonplace to consider Kant’s use of terminology original and complex. This fact is reflected not only by the sheer number of published lexica or dictionaries of Kantian terminology,^[4] but also by a wealth of depictions of Kant’s philosophy in the secondary literature. For instance, Yovel states that “[n]ot only was Kant a philosophical revolutionary, he also had to invent the language by which to express his innovative ideas; and in both ways he had to act as pathbreaker”,^[5] and, as Caygill puts it, “[i]t is widely recognized that Kant transformed philosophical language”.^[6] I hold any reader of Kant can in principle agree with these kinds of statements.

For this reason, it might strike us as paradoxical that Kant very rarely coins the philosophical terms he uses.^[7] In Adorno’s words, who considered this fact to be an advantage, Kant’s writing “is relatively free of neologisms and arbitrarily coined terms, and even in certain areas it adheres to the transmitted philosophical terminology.”^[8] Is this intentional on Kant’s part? Does Kant have a deliberate criterion when choosing terms for philosophical concepts?

Kant’s avoidance of coining new terms is not accidental. In fact, surveying Kant’s reflections on this subject reveals a thinker who consistently expresses his preference for borrowing traditional terms, words from common language or extant words from extinct languages, instead of coining new terms. Therefore, the “search for exactly appropriate terms” from the quoted letter to Herz is to be regarded as a search within the trove of history. However, Kant’s espousal of preexisting terminology is not, as we will see, attributable to a kind of uncritically adopted linguistic conservatism.

This paper argues, rather, that Kant’s position on the language of philosophy is ultimately a consequence of two tenets of his philosophical methodology. According to these – developed in section 2 – philosophy is a discipline whose concepts cannot be securely defined nor represented concretely by its terms. More precisely, first, one cannot securely define a philosophical concept and, as it were, tag it with its corresponding term; second, neither is there anything in terms that inherently attach them to specific philosophical concepts, since the latter’s representations are abstract. Section 3 shows that, consequently, as Kant argues in the Inquiry concerning the distinctness of the principles of natural theology and morality (1764) – in German: Untersuchung über die Deutlichkeit der Grundsätze der natürlichen Theologie und der Moral (henceforth UD) –, the relationship between a term and a concept in philosophy is only held together by ‘linguistic usage’ (Redegebrauch). Taking this dependence on linguistic praxis into account, Kant’s remarks in the CPR on finding suitable terms for philosophical concepts within tradition or ordinary language are seen as consistent with and grounded on his philosophical methodology.

1.2 Preliminary considerations

One way of investigating the philosophy of language in Kantian philosophy is historical. This approach underscores, for instance, the reconstruction of Kant’s influence on Herder’s treatment of the problem of the origin of language,^[9] or a defense from Hamann’s charge that the CPR lacks an understanding of language.^[10]

Another approach is systematic. An early example is Strawson’s (1966) formulation of the principle of significance, namely, the claim “that there can be no legitimate, or even meaningful, employment of ideas or concepts which does not relate them to empirical or experiential conditions of their application.”^[11] The principle relates Kant’s theory of concepts from the CPR with a notion of what it is to be semantically meaningful, since without it “we shall not merely be saying what we do not know; we shall not really know what we are saying.”^[12] This suggests that there should be an implicit philosophy of language present in the CPR, and several studies have emerged putting forward this position.^[13]

Now, instead of determining what Kant’s philosophy implies for language, another approach emphasizes examining Kant’s explicit assertions about language and determining the systematic concerns that they answer to. Adorno’s (1973) and Simon’s (2003) works more closely resemble this strategy, and this paper follows suit. As it turns out, Kant is quite outspoken regarding language when arguing on the appropriate terminology for conducting philosophy or when contrasting the expressions of philosophical concepts with those of mathematical and other non-philosophical concepts.

The topic of this paper is therefore accordingly restricted to Kant’s philosophy of language of philosophy. Despite its terminological focus, this paper is not theoretically nor methodologically philological, neither does it provide an etymological, conceptual or a statistical analysis of Kant’s own use of terms. My concern is to examine and evaluate Kant’s arguments about what makes terms appropriate for referring to philosophical concepts – I do not ask whether he himself applies them consistently.

Simon (2003) examines Kant’s conception of the language of philosophy, and argues that it is a result of Kant’s critical insight.^[14] I accept his characterization, but do not believe his explanation can be entirely correct. As we will see, Kant’s open-ended conception of language predates his critical turn, and thus must be comprehended on the basis of philosophical insights that both predated and survived his radical reform of philosophy.

There are two types of passages in Kant’s published writings that are relevant. First, there are Kant’s remarks on the problematic nature of definition and representation of philosophical concepts. Next, we have arguments about his choice of terms. My approach is to use the former passages to elucidate the latter. It is important to briefly spell out some particularities of each.

When it comes to the question of definition and representation of philosophical concepts, we benefit from the fact that Kant’s reflections belong to a main argument on philosophical methodology in the UD and in the CPR,^[15] as well as in the Log.^[16] Kant’s placement of the topic in his published works can be called systematic. In fact, Kant’s argument from the UD is reworked into the Discipline of Pure Reason chapter of the CPR. Thus, for section 2, I have adopted the methodological assumption that stitching both the UD and the CPR together provides a more complete picture of Kant’s stance on the subject, despite each work being chronologically on opposite sides of Kant’s critical turn. This move is not unfounded – as Caimi has shown, a comparison between the 1764 and the 1781 (and 1787) texts reveals that “[t]he steps we have found out in the Untersuchung über die Deutlichkeit are to be also met in the Critique”.^[17] Kant’s arguing points and their order remain for the most part unaltered. The account from the CPR is less detailed, but up-to-date with his critical terminology.^[18] These facts are taken to legitimize – and even require – resorting to said pre-critical passages for clarifying parts of Kant’s critical philosophy, contrary to Simon’s (2003) approach.

Let us now consider Kant’s reflections on choosing philosophical terms. Although a recurring topic, their occurrences are scattered and relative to specific terms; nothing like a general defense of a criterion for choosing terminology occupies a systematic place anywhere in the Kantian corpus. As such, the depiction of Kant’s stance in section 3 is necessarily a reconstruction from the fragments. However, at the time it was published, the CPR presented Kant’s largest and most complex terminology to date, and it thereafter arguably served as the lexical blueprint for the rest of his writings. For no other work did he amass so much effort in seeking adequate words to express his thoughts, and thus we are to pay special attention to the deliberations on terminology in the CPR. In section 3, I examine Kant’s most thorough discussion justifying the choice of a philosophical term – ‘idea’ – from the CPR.

2.1 Definition

What is involved in defining a concept? In the CPR, Kant says that “[a]s the expression itself reveals, to define properly means just to exhibit originally the exhaustive concept of a thing within its boundaries.”^[19] It is this notion of definition that in the CPR Kant argues cannot be secured for philosophical concepts, so it is worth taking a closer look at what he means by the three key terms: (1) ‘exhaustive’, (2) ‘original’, and (3) ‘boundary’.^[20]

Kant clarifies these as follows. “Exhaustiveness”, he writes, “signifies the clarity and sufficiency of marks”,^[21] meaning that the features that are used to define something are sufficient and clear.^[22] Exhaustiveness on its own is not enough to secure a definition. Why? Suppose one intended to define ‘vertebrate’, but instead provided a definition of ‘animal’. The definition offered could be clear and sufficient to cover all vertebrates (since they are animals), but too general and thus would include non-vertebrate animals. Thus, Kant says, the sought-for definition requires “boundaries”. In his words, these signify “the precision, that is, that there are no more of these than are required for the exhaustive concept”.^[23]

Now, Kant also requires that the boundaries of the definition and the boundaries of what is being defined are not accidentally coextensive. For example, it may well be true that all vertebrates have kidneys (exhaustiveness), and that all living beings with kidneys are vertebrates (boundaries). But even if the boundaries are coextensive – i. e., even if all kidney-bearers are vertebrates and vice-versa – this does not entail that the two concepts have interchangeable definitions. It is vertebrae-having, not kidney-having, that originally exhibits a characteristic mark of being a vertebrate. Thus, “original”, Kant specifies, “[signifies] that this boundary-determination is not derived from anywhere else and thus in need of a proof, which would make the supposed definition incapable of standing at the head of all judgments about an object.”^[24]

2.2 Nominal and real definition

Having clarified what Kant means by definition, we may now turn to the traditional distinction between nominal definition and real definition,^[25] which Kant adopts in the UD and in the CPR. In Kant’s understanding, nominal definition merely supplies “other and more intelligible words for the name of a thing”,^[26] whereas real definition,^[27] identified as definition proper (outlined in the previous subsection),^[28] tells us what a concept consists in. That is, a real definition,

contains in itself a clear mark by means of which the object (definitum) can always be securely cognized, and that makes the concept that is to be explained usable in application. A real definition would therefore be that which does not merely make distinct a concept but at the same time its objective reality [Realität].^[29]

Unlike his Wolffian predecessors, for Kant, definition sets the methodologies of philosophy and mathematics apart, despite both sciences employing the same faculty (pure reason) and may even cognize the same objects.^[30] In the UD, Kant holds that a philosophical investigation should not attempt to follow the mathematical method: mathematics must begin by defining its concepts; philosophy may only do so when an investigation has been conducted. In philosophy, before an investigation into the nature of a concept is conducted, the only kind of definitions that can be secured are nominal. That is, other, perhaps more intelligible words can be supplied for the name of the concept which one wants to define, but not the content of the concept. In a characterization which may strike us as reminiscent of Plato’s dialogues,^[31] Kant suggests that if real definitions in philosophy are attainable at all, they can only be determined at the end point of the investigation. By the time the CPR was published in 1781, Kant had decided that proper, real definitions for philosophical concepts, strictly speaking, cannot be secured. Instead, one can only provide nominal definitions and what Kant calls exposition, which fails to be exhaustive.^[32]

In the following two subsections, I will respectively examine two conditions which I call the ‘condition of a priori arbitrary synthesis’ and the ‘condition of in concreto representation’. A concept must satisfy the condition of arbitrary synthesis for it to be definable with a real definition,^[33] and it must satisfy the condition of in concreto representation for it to be representable essentially by signs. While both conditions are present in both the CPR and the UD, the latter condition I primarily read from the UD.^[34] This is due to the fact that, in Caimi’s words, this formulation “in the Critique of Pure Reason is not as detailed and manifest as in the Untersuchung über die Deutlichkeit”.^[35] I take this pre-critical elucidation of a tenet from Kant’s critical philosophy to be valid for the reasons outlined in subsection 1.2. Now, because for Kant mathematical concepts satisfy both these conditions, one can properly provide real definitions for its concepts and can represent its concepts essentially via its signs (or combinations of signs).^[36] Yet, philosophical concepts fail to satisfy the conditions. Ultimately, since there is nothing about the content of a philosophical concept that makes it fit for any specific term, and reciprocally, since there is nothing in a term that makes it fit for any specific philosophical concept, Kant searches for a criterion of suitability in the linguistic usage of words.

2.3 The condition of a priori arbitrary synthesis

The condition of a priori arbitrary synthesis states that the only definable concepts are those which are a priori and are not given, but rather arbitrarily (willkürlich) and synthetically produced.^[37] At this point, I beg patience from the reader: it is easier to comprehend the meaning of this condition by following Kant’s argument for it, which works by showing that any alternatives for providing real definitions fail.

To understand said alternatives, let us observe the following Kantian oppositions that govern his argument in this passage: a concept may be empirical or a priori;^[38] it may be given or arbitrary;^[39] and may be a result of analysis or of synthesis (i. e., construction in pure intuition).^[40] Bearing these oppositions in mind, the outline of Kant’s argument is as follows. (1) Kant argues that a definition cannot be provided for an empirical concept. Only a priori concepts are left. (2) A priori given concepts cannot be defined, not even by analysis. Only a priori arbitrary concepts are left. (3) Kant argues that a priori arbitrary concepts might not have a corresponding object in reality. (4) Kant concludes that only a priori arbitrary concepts produced by synthesis are left, and shows they are definable. Let us see this in more detail.

(1) In Kant’s account of empirical concepts, a definition is either attempted with or without resorting to experiment. For instance, prior to experiment, anything one can provide relative to a definition of ‘water’ is only nominal i. e., pertaining to how the word is intended.^[41] If one resorts to experiment, some characteristic marks of water can be discovered, but further observations can always take away or add some characteristic marks to the definition “and therefore the concept never remains within secure boundaries.”^[42] Thus, empirical concepts cannot be properly defined,^[43] since where the boundaries reside can never be conclusively settled. (2) “Second,” Kant writes, “strictly speaking no concept given a priori can be defined”.^[44] Philosophical concepts are given a priori concepts: e. g., substance, cause, right, equity, etc. Kant holds that there is no level of analysis to indicate that the definition exhaustively covers the content of the given concept. As he puts it, “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 […] the exhaustiveness of the analysis of my concept is always doubtful”.^[45] The reason for this, as Kant puts it in the UD, is that it is easy “for the characteristic mark of an abstracted concept to escape our attention without our noticing, for there is nothing sensible which can reveal to us the fact that the characteristic mark has been overlooked.”^[46] (3) For Kant, consequently, only a priori arbitrary concepts may be definable, but these still may or may not correspond to real objects. When such concepts do not correspond to any real objects, their definition is not a real definition, for, as previously quoted, “[a] real definition would therefore be that which does not merely make distinct a concept but at the same time its objective reality”.^[47] It is merely a “happy coincidence” if such concepts do correspond to objective reality.^[48] (4) The only way of assuring that a priori arbitrary concepts correspond to real objects – or what is the same, have objective reality – is when their definitions are synthetic constructions of their objects. In Kant’s view, “only mathematics has definitions”,^[49] and as he puts it in the UD, only in mathematics “[t]he concept which I am defining is not given prior to the definition itself; on the contrary, it only comes into existence as a result of that definition.”^[50] For instance, take the concept of a cone. Kant writes: “in mathematics the concept is the product of the arbitrary representation of a right-angled triangle which is rotated on one of its sides. In this and in all other cases the definition obviously comes into being [i. e. obtains objective reality] as a result of synthesis.”^[51] Kant concludes his argument as follows:

Since therefore neither empirical concepts nor concepts given a priori can be defined, there remain none but arbitrarily (willkürlich) thought ones for which one can attempt this trick. In such a case I can always define my concept: for I must know what I wanted to think, since I deliberately (willkürlichen) made it up, and it was not given to me […] but I cannot say that I have thereby defined a true object. For […] from the concept I do not even know whether it has an object […] Thus there remain no other concepts that are fit for being defined than those containing an arbitrary synthesis which can be constructed a priori.^[52]

Step (2) of this argument leaves philosophical concepts behind – overall, they do not satisfy the condition of a priori arbitrary synthesis. This explains why “philosophical definitions come about only as expositions of given concepts”.^[53] According to Kant, settling for an exposition instead of a definition is something that “the critic can accept as valid to a certain degree while yet retaining reservations about its exhaustiveness.”^[54] Notwithstanding, exposition of concepts is not a superficial, disposable philosophical exercise: in fact, it is employed in some crucial passages of the CPR.^[55] But in being non-exhaustive, it is not helpful in the search for exactly appropriate terms, i. e., those whose suitability make philosophical concepts precise and intelligible.

2.4 The condition of in concreto representation

The condition of in concreto representation refers to whether a particular – like an object given in intuition – can essentially model or represent something universal – like a concept. In this subsection I clarify this condition and show why for Kant it applies to mathematical concepts, but not to philosophical ones.

Firstly, Kant holds that there are two ways of representing something universal: in concreto (concretely) and in abstracto (abstractly). When something particular is capable of representing or standing in for something universal, Kant calls the mode of representation in concreto. When a universal cannot be cognized by means of a particular, then its representation is arrived at in abstracto.^[56]

The condition of in concreto representation requires referring concepts to objects that can be exhibited in intuition. Kant draws a difference between two ways in which an object can be given in intuition: in sensible intuition, which refers to empirical content given in intuition through the senses, and in a priori or pure intuition, which refers to the form of intuition, that is, the forms of space (external and simultaneous relations) and of time (internal and successive relations.) For example, water can be given in intuition through the senses, making it the type of thing that can be pointed to, whereas a geometrical object, such as a cone, is exhibited in intuition a priori by its definition, which refers to how it is to be built in space (as seen in Kant’s cone example from the previous subsection.) Although Kant does not argue so explicitly, presumably an object that is empirically given in sensible intuition can only obscurely represent in concreto its concept: a certain degree of uncertainty in the representation is unavoidable. A representation, in Kant’s words, “by many appropriate examples can only be made probably but never apodictically certain.”^[57]

In mathematics, the condition of in concreto representation is perfectly satisfied. This is because, as seen in the previous subsection, a mathematical concept is arbitrarily (or deliberately) synthetically defined as nothing but the construction via the positing of signs of the mathematical object.^[58] This entails that the universal concept and the particular object are in complete accordance. Additionally, since in mathematical definitions “there are posited first of all not things themselves but their signs”,^[59] there is the further implication that “the significance of the signs employed is certain, for it is not difficult to know what the significance was which one wished to attribute to those signs.”^[60] In short, because the signs are deliberately used for constructing mathematical concepts, the signs are suitable for representing them in concreto.

In other words, mathematical signs can stand in for their concept without loss of generality. For example, a geometrical figure such as a circle is a sign that exhaustively represents in concreto the geometrical concept of circle. As Kant puts it, “in order […] to discover the properties of all circles, one circle is drawn; and in this one circle, instead of drawing all the possible lines which could intersect each other within it, two lines only are drawn.”^[61] In arithmetic and algebra, on the other hand, it is not the signs themselves – e. g., characters for numbers –, but the way in which the signs combine in concreto that represents the concepts. Thus, arithmetical conclusions may be arrived at without resorting to anything outside of the signs themselves.^[62]

Let us now contrast this to the case in philosophy. Kant writes:

The signs employed in philosophical reflection are never anything other than words. And words can neither show in their composition the constituent concepts of which the whole idea, indicated by the word, consists; nor are they capable of indicating in their combinations the relations of the philosophical thoughts to each other.^[63]

In other words, the mode in which philosophical concepts are represented is not in concreto, but in abstracto, severing any inherent connection between the signs and the philosophical concepts. Thus, in the case of philosophical concepts, “one is constrained to represent the universal in abstracto without being able to avail oneself of […] handling individual signs”.^[64]

Let us note an important conclusion from the failure of philosophical concepts to satisfy either the condition of a priori arbitrary synthesis or the condition of in concreto representation. First, by failing to meet the condition of a priori arbitrary synthesis, we’ve discovered why Kant considers the task of securing a real definition of a philosophical concept problematic. If the meaning of a philosophical concept could be secured via a real definition, then in principle a term could be stipulated to conventionally refer to that definition. But philosophical concepts are not fit for real definitions, so one cannot, as it were, tag a word onto a definition of a philosophical concept.^[65] We are forced to conclude that when one tries to stipulate philosophical terminology by definitions, these definitions are at best nominal – pertaining to the terms – and say nothing about the philosophical concepts they are meant to refer to: “the word”, says Kant “with the few marks that are attached to it, is to constitute only a designation and not a concept of the thing; thus the putative definition is nothing other than the determination of the word.”^[66]

Second, we’ve seen that for Kant philosophical concepts fail to satisfy the condition of in concreto representation. Instead, words represent philosophical concepts in abstracto. This means that there is nothing intrinsic to words or combinations of words to pin down which philosophical concepts they signify. They are, in the end, completely unlike a drawn square, which can essentially represent the mathematical concept of a square – but not the concept of a circle, triangle, etc. Philosophical terms are never more than mere conventional designations.

As a conclusion, when considering naming a given philosophical concept, apparently no term is more appropriate than any other term. This means that, when introducing terms in a philosophical investigation, their suitability for expressing philosophical concepts – which is what Kant says in his letter to Herz cost him unimaginable time and effort – can be sought for neither in how the terms nor the concepts are constituted. Only a third possibility remains.

3.1 Sanctioned usage

In the previous section of the paper, we have seen two inherent difficulties in trying to establish suitable terms for philosophical concepts. Kant appears to be saying that a referential relationship between a word and a philosophical concept is not ultimately held together by semantics nor by syntax. How do these issues shed light on Kant’s terminology?

In a 1790 letter, Abraham Gotthelf Kästner encouraged Kant to clearly define the terms used in the CPR,^[67] which we’ve seen his philosophical methodology opposes. In his answer from May 1793, Kant confessed to his lingering dissatisfaction with the terminology of the CPR, which he felt obstructed intelligibility. Yet, instead of following Kästner’s suggestion, Kant wrote: “I shall take the next available opportunity where a dry discourse is called for to undo this mischief and try to combine that scholastic terminology with ordinary language.”^[68] In this section we will see why Kant thinks that instead of stipulating, establishing, or defining terms, one ought to search for them. Their suitability comes from the fact that they are already in use and is sanctioned by how one continues to use them.

Kant’s first formulation of this position dates back to the 1760s.^[69] In the UD, terms are endowed with their meaning through their employment, that is, what Walford and Meerbote have translated into English as ‘linguistic usage’ (Redegebrauch).^[70] The usage of a term may be correct with respect to a philosophical concept, even without having previously arrived at a real or even nominal definition for the concept. In this vein, Kant writes:

Augustine said: ‘I know perfectly well what time is, but if someone asks me what it is I do not know.’ […] no real definition has ever been given of time. For, as far as the nominal definition is concerned, it is of little or no use to us, for even without the nominal definition the word is understood well enough not to be misused.^[71]

The position that definitions of philosophical concepts cannot (and need not) be secured also commits Kant to allowing that nothing can safeguard philosophical terms once and for all from their susceptibility to unsteady, imprecise usage. According to Caygill in the introduction to his Kantian dictionary, Kant “took a problematic and a relational view of philosophizing and philosophical language […] philosophical concepts were acroamatic and not axiomatic, by which he meant that they were the discursive outcome of an open-ended process of reflection upon philosophical problems.”^[72] It is only accumulated consistent and precise usage, according to Kant, that sanctions the terminology employed.

In this context, Kant does mention in the UD that ‘logical limitation’ could possibly prove a more precise way of determining meanings of words than usage. He does not explain what logical limitation consists of, but from its context and similar expressions in other works,^[73] a reasonable guess is that he is referring to making distinctions between otherwise similar concepts (drawing boundaries) and between different senses in which a single word can be used. Nevertheless, whether terminology has become sanctioned by its usage or has been made precise by logical limitation, Kant is clear that both can be undermined when the further use of a term is not precise and consistent. Usage is, in the end, the final authority:

In philosophy generally […] words acquire their meaning as a result of linguistic usage, unless, that is, the meaning has been more precisely determined by means of logical limitation. But it frequently happens that the same words are employed for concepts which, while very similar, nonetheless conceal within themselves considerable differences. For this reason, whenever such a concept is applied, even though one’s terminology may seem to be fully sanctioned by linguistic usage, one must still pay careful attention to whether it is really the same concept which is connected here with the same sign.^[74]

In 1781, the way in which Kant argues about choosing philosophical terminology had become crucially shaped by his view on sanctioning terms by accumulated consistent linguistic usage. In his consolidated position, when choosing terms for philosophical concepts, one must keep an eye on the words used in tradition and common use.

3.2 Terminology in the CPR

In the preliminary considerations of this paper (subsection 1.2), I’ve drawn attention to the fact that Kant’s arguments regarding choice of terms in philosophy come as preludes or justifications to his own philosophical terminology. I will now turn to Kant’s most detailed and explicit defense of the use of the term ‘idea’ (derived from Platonic terminology) which spans pages A312 – 320/B 368 – 377 of the CPR. What makes this passage significant, however, is not only that Kant lays bare his criteria for the appropriateness of this term, but that he holds that this is the criteria one ought to follow when deciding which term is suitable for given a philosophical concept. This part of the paper is meant to examine what Kant’s reasons amount to.

In Kant’s account, first one will come up with a philosophical concept, and second, one will seek a suitable term in order to write or speak about such a concept. Next, Kant advises to survey “the great wealth of our languages” in search of a suitable, preexisting word for that concept.^[75] In fact, he even writes that

it is advisable to look around in a dead and learned language to see if an expression occurs in it that is suitable to this concept; and even if the ancient use of this expression has become somewhat unsteady owing to the inattentiveness of its authors, it is better to fix on the meaning that is proper to it (even if it is doubtful whether it always had exactly this sense)^[76]

But why is the fact that terms be preexisting so important? Kant’s response depends on one of his most nuanced claims, according to which the language used to express philosophical thoughts is decisive for making them intelligible, or in other words, poor language is detrimental to the intelligibility of philosophy. To repeat a quote from subsection 1.1, “the thinking mind nevertheless often finds itself at a loss for an expression that exactly suits its concept, and lacking this it is able to make itself rightly intelligible neither to others nor even to itself.”^[77] Thus, Kant’s justification for seeking terms within existing or historical languages is based on a reductio of the alternative, that is, coining new terms (even for new philosophical concepts.) Since Kant thinks that coining terms “is a presumption to legislate in language that rarely succeeds”,^[78] it follows that this will in turn “ruin our enterprise [of philosophy] by making ourselves unintelligible.”^[79]

But even when one does without coining terms, poor, imprecise philosophical language may cause both the language and the philosophical thoughts themselves to oppose the intention of the philosopher. In his words, a philosopher “may not have determined his concept sufficiently and hence sometimes spoke, or even thought, contrary to his own intention.”^[80]

This brings us to other requirements that the usage of philosophical language must follow, namely, avoiding (1) using terms with multiple intended meanings, (2) vague open-ended meanings, and (3) the use of synonyms in philosophy. This is supposed to prevent unsteady or imprecise usage of a term and ensure a one-to-one relation between terms and philosophical concepts. As Kant puts it:

[I]f there perhaps occurs only one single word for a certain concept that, in one meaning already introduced, exactly suits this concept, and if it is of great importance to distinguish it from other related concepts, then it is advisable not to be prodigal with that word or use it merely as a synonym or an alternative in place of other words, but rather to preserve it carefully in its proper meaning; for it may otherwise easily happen that when the expression does not particularly occupy our attention but is lost in a heap of others having very divergent meaning, the thought which it alone can preserve may get lost as well.^[81]

However, what happens when traditional or contemporary philosophers have misused a term? How does Kant treat preexisting terms for philosophical concepts with a track-record of imprecise usage? In such cases, Kant says, we are stuck with “an expression with which we cannot dispense and at the same time cannot safely use because of an ambiguity it has acquired through long misuse.”^[82] Considering this misuse an impairment to science, Kant feels the need to set the record straight, and furthermore urges philosophers to concern themselves with “carefully preserving the determination and the expression on which the concept depends”.^[83]

In particular, in this passage of the CPR, Kant names certain terms that have fallen into improper use. These include the term ‘absolute’, which for Kant “is one of the few words that in its original meaning was suited to one concept that by and large no other word in the same language precisely suits, and so its loss, or what is the same thing, its vacillating use, must carry with it the loss of the concept itself”.^[84] Furthermore, he turns his attention to terms for species of representation, which include the term ‘idea’ – Kant’s main focus of this passage. He entreats “those who take philosophy to heart”, to

take care to preserve the expression idea in its original meaning, so that it will not henceforth fall among the other expressions by which all sorts of representations are denoted in careless disorder, to the detriment of science. We are not so lacking in terms properly suited to each species of representation that we have need for one to encroach on the property of another.^[85]

Kant follows this passage with a corrective differentiation between the terms for the various species of representation: representation (repraesentatio), perception (perceptio), sensation (sensatio), cognition (cognitio), intuition (intuitus), concept (conceptus), notion (notio) and idea. The claim that concludes this passage is indicative of Kant’s vexation about imprecise or non-traditional usage of philosophical terminology: “Anyone who has become accustomed to this distinction must find it unbearable to hear a representation of the color red called an idea. It is not even to be called a notion (a concept of the understanding).”^[86]

Finally, there is another situation which deserves some attention, namely, when accumulated misuse has brought a word too far from its original significance and has successfully established a new meaning for it – when tradition overrules itself by compounded misuse. Under these circumstances, in my reading, Kant recognizes that adopting traditional terms is not preferable, and instead endorses employing a different term for the thought, so as to dispel its confused meaning.^[87] This is the only situation in which Kant argues that one is warranted to seek different terms for old concepts which may already have their own traditional terms. For instance, such is the case with the term ‘ontology’. Due to its misuse of “[presuming] to offer synthetic a priori cognitions of things in general in a systematic doctrine” – a presumption whose philosophical legitimacy has not yet been established – Kant claims, “the proud name of an ontology […] must give way to the modest one of a mere analytic of the pure understanding”.^[88]