Info

auth. Anne-Françoise Schmid, trans. Sylvia.

pub. Dictionnaire d’Histoire et de philosophie des sciences, Paris: PUF, 1999.

The difficulty of accounting for the concept of “concept” is that it is everywhere in the tradition ­— habitually made to ascend to the Platonic Idea ­—, but that it only takes up a sense that is functionally and systematically distinct from the Idea with Kant. The history of the concept thus marries the meanders of that of the Idea. The minimal difference that can be proposed is that mode of relation to the object of knowledge is always at play in the concept, while the Idea is this object itself, preferably known immediately. Morris Weitz’s hypothesis, in Theories of Concepts (1988), is that one can determine a theory of the concept with each philosopher. Richard L. Schwartz, Der Begriff des Begriffes in der philosophischen Lexikographie. Ein Beitrag zur Begriffsgeschichte (1983), treats the term of concept without systematically distinguishing it from the idea, as is the case in many encyclopedias.

We call “concept” a universal, mediate, and inferential mode of representation of the relation to the object of knowledge. The “concept” is specific to knowledge, philosophical or scientific. The concept is in effect distinguished from the idea in that it is presented as the result of an act of conception (from the Latin concipere, taking together) or as an objective entity being related to something other than itself. The concept is reduced neither to the psychological although it is the result of an act of thought, nor the empirical although it is not confounded with the item of though, nor the linguistic, nor its own definition although it is the correlate of meaning of some linguistic expressions, nor the ideal although it is also universal. It is a characteristic of the concept of “concept” to make distinct orders intervene and be a skeleton key that permits access to each of them. It is only with it that what is at stake is the status of the theoretical, not that of the thinking subject or the object of thought. Depending on the perspective, it can be described either as the point of condensation or accumulation of theoretical problems, or as the elementary unit of theoretical construction.

These characteristics make it such that it is not easy to describe something that is concept, distinct from the idea, the word, the thing, thought, the Idea, and which would be describable within the whole of the philosophical tradition as a specific function. The word is of a Latin origin and is missing in Greek — or is only represented by a constellation of terms susceptible to having other significations; this is moreover also the case in Latin, where we just as well find conceptus, intentio, notio. This situation draws the possible solutions of the treatment of the concept of concept: either it derives from the idea, or it possesses a history that is distinct from and complementary to it. It is this second solution that Morris Weitz bolsters in Theories of Concepts (1988), which postulates the existence of a veritable theory of the concept across the philosophical tradition, the characteristics of which would only be broken by the Wittgenstein of the Investigations and some — rare — contemporary authors. The criterion he proposes is that the “concept” is in a way “closed” by a set of necessary and sufficient conditions allowing to identify it, a criterion that even the big changes proposed by Frege in the conception of concept would not modify.

Thus, we can say that Plato had in view in the Ideas something equivalent to the concept when he made it the correlate of a definition — which is not confounded with it, particularly in the so-called Socratic dialogues; but it will for all that not be confounded with the Idea or the form which is rather the very condition of the concept: without idea, there is no possible concept, this being the “disposition” (according to Weitz) to be distinguished, to pose the question “what is… ?” But the knowledge of the idea as such is not inferential, and therefore not conceptual.

It is indubitably Aristotle who was the first to invent the concept as entity (definitional entity according to Weitz) conceived at the same time as abstracted from the empirical, a both mediate and general mode of knowledge, and a mode of classification between genre and species, as intention and extension. All these characteristics seem to go without saying together: if the concept results from what is common between several entities 〈étants〉, it is normal that it would be more general than each of them and therefore be mediate knowledge. The classification between genre and species goes together with Aristotelian syllogistics. Each of these characters will be important in what follows of the history of the concept: it will undergo modifications in terms of the manner in which we think on the one hand abstraction and its status — whence discharges the classical distinctions between “formal concept” (from the side of the Idea) and “objective concept” (from the side of the object) —, on the other hand the identification of the logical to the syllogistic and the mathematic.

Thus, the history of the concept is determined both by its possibility of definition and by its relation to the empirical. This double characterization makes the status of the concept hesitate between the universal which is its comprehension — in what is the correlate of a definition —, and the general — in what is formed in an act of abstraction of the empirical given which forms its extension. This situation accounts for doctrinal tendencies of the metaphysical interpretation of the concept: either closer to the word and nominal singularity, therefore partially arbitrary with regard to individuals of the empirical, or acquiring a form of autonomy or reality bringing it closer to the idea. Theories of the concept from the Middle Ages: nominalism, conceptualism, realism are choices among these possibles.

It is probable that the traditions that perfectly distinguish the concept from the idea are rather of an Aristotelian origin; nevertheless, the constellation of terms which gravitate around “conceptus” is found to be just as well represented in the philosophies which first and foremost distinguish themselves from Aristotle. But this debate is only secondarily pertinent here: what is important is that at a certain moment, the philosopher can no longer explicitly confound the concept and the idea, at least insofar as they seek to comprehend science: Kant’s work systematically makes the distinction between concept and idea and does this with the goal of comprehending the Newtonian scientific edifice. Or rather, the idea is indeed the concept, but which is not susceptible to being limited by experience: it a has a value that is regulatory and non-constitutive of science as such. We can make the hypothesis that the concept appears as individuated and functional when it entails comprehended the sciences and the relationships of philosophy to the sciences: thus, with Aristotle, at the time when scientific domains appear in their identity, with Kant with Newton as the outcome of the Copernican revolution and the apparition of modern sciences, with the Moderns — positivism and logical positivism — with the diversification of sciences and the apparition of a logic more mathematical than Aristotle’s, and when on seeks to identify the theoretical even more than the speculative. And in all of contemporary epistemology, the notion of concept — in Hempel, Nagel, Toulmin, but also in the important French tradition Cavaillès, Lautman, Bachelard, Canguilhem, etc. — becomes central and functional, while the idea is not so much an object of epistemology but is rather close to the “view of the mind”. The concept becomes the guarantee that one is engaged in science as it is done and as it is thought in the scientific consciousness.

Kant isolates concepts by straightaway refusing to make of them a generalization of the givens of sense like the empiricists do, but by giving them an intrinsic value, which the rationalists did know how to do: he thus suppresses a classical debate and replaces the problematic between simple and composite ideas with the distinction of pure concepts and empirical concepts. This supposes the isolation of not only of the concept, but of sensibility as such, rigorously passive and indeterminate. The problem of scientific knowledge will thus firstly be, at every level, empirical and transcendental, that of the synthesis between heterogeneous elements, among which the concept is one. This new situation will profoundly modify the notion of concept: it becomes a rule of identification of the given, and, to do this, can no longer “resemble” in any way the given, to which it no longer applies directly but across a schematism. Thus vanishes the paradoxes of the application of ideas to things, as they were recognized by a certain number of empiricists, Berkeley in particular: how can one have a general idea of the triangle, if this generality — or this universality — implies in its representation that it is neither scalene, nor rectangle, nor isosceles, etc. The paradox can only be lifted by supposing that none of these determinations are given as such in intuition: each is a conceptual determination that only tells us something of an empirical intuition across a schema of the transcendental imagination, faculty of synthesis. We see there that Kant’s conception of the concept finds itself between two worlds: that of the “predicate” as in the entire tradition, supposing that the concept says something of a subject, and is thus directly predicable of phenomena. But at the same time, the concept takes all its originality insofar as it is the component of a judgement, which can be a scientific law. Kant will modify the conditions of philosophizing in particular by showing through the very construction of his philosophy that concept only has a meaning within the frame of a theory of judgement alone capable of elaborating the condition of comparability and homogeneity of faculties — philosopher will later speak of proposition rather than judgement. In a famous article, Wilfrid Sellars shows that concepts always imply laws, without which they would be void and indistinguishable.

This is a fundamental difference: that the predicate be in the subject (“praedicatum inest subiecto” — according to Leibniz’s major principle) or in the judgement makes the difference between a “phenomenology” where the predicates tell us something of the subjects and a more “inferential” conception of the concept which is no longer necessarily linked to the concept defined as genre and specific difference. Hegel’s philosophy of sciences — which, we know, gives such an importance to the concept — can be understood as a way of overcoming this contradiction of the phenomenological and the inferential in the movement of the concept. For a “genealogical” conception of the concept (to take up one Darwin’s terms which he applies to classifications), a larger conception of the concept will progressively be substituted linked not only to classical predicates with one place, but also with two, three places, etc., i.e., also to binary, ternary, etc. relationships but also to possible worlds understood as exemplifying the concepts inasmuch as they are what, in the law, can be exemplified. This has supposed an entire formal treatment of the concept as function (Frege), as well as a theory of quantification (with Frege principally, but also with Peirce’s lesser known “logic of relatives”). In the expression F(x), “F” represents the essentially unsaturated concept and “x” the argument or occurrence. This is the origin of propositional functions: the concept appears here as both functional and unsaturated. Whence, the problem will be to understand the relation of the concept to the empirical across the proposition. This is what analytic philosophy will attempt, in its will to take note of the scientific invention of the new logic. The problem of the concept will thus take two directions: either the philosophers attempt a theory of the concept taking pains to understand how it is related to the given, this is Russell’s case — particularly with his theory of descriptions (1905) — or else, critical of an idea of the given, they will attempt to understand the systematic relationship between concepts, in the famous problem of coherence, and will tend towards a nominalist interpretation of concepts.

In contemporary epistemology, the question of concepts is often treated in terms of a theory of definition. Such an approach to the concept allows determining the relation of the concept to the empirical, since a definition shows how to eliminate a term. One of the questions which has seemed fundamental in the modern sciences, and which puts in relation the concepts of concept and law is that certain concepts seem to only be able to be formed through dispositional definitions which have the form of a conditional, noted particularly by Hempel in his article “Fundamentals of Concept Formation in Empirical Science” (1952). Such is the case of concepts of magnetism or temperature which makes impossible their immediate relating to givens of observation: the status of the concept thus seems specific and to be in a relationship with problems of deductibility rather than observability. Logical empiricism has thus sought modalities of reduction of concepts by an enlargement of definitions (Hempel, Nagel). The problem of the relation of the concept to the given or the criterion of coherence is thus no longer posed as a simple alternative. Russell will himself abandon the notion of “sense-data” when he will adopt the doctrine of “neutral monism”, the notion of the “given” can appear like a myth (Wilfrid Sellars, 1963). The idea that intuitions and concepts form representations of a different type will end up being critiqued in the analytical tradition itself (see Rorty, 1979): from the 1950s, the works of Quine, Austin, and especially Wittgenstein’s Philosophical Investigations will reformulate otherwise the question of the concept by putting in relation with language rather than the given. This is the “linguistic turn”. Theories of knowledge in the strict sense are no longer the only site where theories of concepts develop.

From this moment onwards, what had been highlighted concerning concepts had been, more than their “closure” in terms of a definition, their network character: “puzzles” in Wittgenstein, “chicanes” in Derrida, or “points of accumulation” in Deleuze. The importance of the “proposition” or the “statement” for evaluating the concept fades away, to allowing the appearance other notions, like that of language games, difference, or the plane of immanence, or indeed “conceptual character”, such as Socrates or Zarathustra. In these tendencies, the reflection on philosophy, its gestures, its decisions has taken up such an importance that the concept more than the idea is the support of what is specifically theoretical and technical in philosophy.

Currently, the term of concept is used very fluently to designate the product of a design 〈conception〉 in the sense that engineering gives to it, as the articulation of models different natures and types in view of a finalized, integrated, and concretized production: thus appears a new vision of the genealogical aspect of the concept, as belonging not so much to the “natural system” as it does to that of the artificial. The concept is thus what the engineer, much more than the scientist, will have conceived. In this sense, the concept is what opens philosophy not only to the domain of science but also to that of technique and technology.