What Is Philosophy?

Giorgio Agamben, Lorenzo Chiesa

• Publisher: Stanford University Press
• Year: 2017
• ISBN:
  9781503604056
• Edition: 1
• Topic: Philosophy
• Subtopic: Phenomenology in Philosophy

In attempting to answer the question posed by this book's title, Giorgio Agamben does not address the idea of philosophy itself. Rather, he turns to the apparently most insignificant of its components: the phonemes, letters, syllables, and words that come together to make up the phrases and ideas of philosophical discourse. A summa, of sorts, of Agamben's thought, the book consists of five essays on five emblematic topics: the Voice, the Sayable, the Demand, the Proem, and the Muse. In keeping with the author's trademark methodology, each essay weaves together archaeological and theoretical investigations: to a patient reconstruction of how the concept of language was invented there corresponds an attempt to restore thought to its place within the voice; to an unusual interpretation of the Platonic Idea corresponds a lucid analysis of the relationship between philosophy and science, and of the crisis that both are undergoing today. In the end, there is no universal answer to what is an impossible or inexhaustible question, and philosophical writing—a problem Agamben has never ceased to grapple with—assumes the form of a prelude to a work that must remain unwritten.

On the Sayable and the Idea

1

It is not the unsayable but the sayable that constitutes the problem philosophy must at each turn confront again. The unsayable is in fact nothing else than a presupposition of language. As soon as there is language, the named thing is presupposed as the non-linguistic or the unrelated with which language has established its relation. This presupposing power is so strong that we imagine the non-linguistic as something unsayable and unrelated, which we somehow try to grasp as such, without realizing that in this way we are simply trying to grasp the shadow of language. In this sense, the unsayable is a genuinely linguistic category, which can be conceived only by a speaking being. This is why, in a letter to Martin Buber of July 1916, Walter Benjamin could speak of a “crystalline elimination of the unsayable in language”: the unsayable does not take place outside of language as something obscure that is presupposed, but, as such, it can be eliminated only in language.

I shall try to show that, on the other hand, the sayable is a non-linguistic but genuinely ontological category. The elimination of the unsayable in language coincides with the exhibition of the sayable as a philosophical task. For this reason, unlike the unsayable, the sayable can never be given before or after language: it arises together with it and, however, remains irreducible to it.

2

The Stoics designated an essential element of their doctrine of the incorporeals with the term sayable, λεκτόν, on the definition of which the historians of philosophy have not yet reached an agreement. Before starting an investigation of this concept, we should therefore first locate it in the philosophical context that pertains to it. Modern scholars tend anachronistically to convert ancient categories and classifications into modern ones and to treat sayability as a logical concept. At the same time, they know perfectly well that the division of philosophy into logic, ontology, physics, metaphysics, and so on, by the grammarians and the scholiasts of late antiquity lends itself to all sorts of equivocations and misunderstandings.

Let us consider Aristotle’s treatise on Categories, or predications (but the Greek term κατηγορίαι means in juridical language “charges, accusations”), which is traditionally classified among his logical works. However, it contains theses that undoubtedly have an ontological character. Ancient commentators therefore debated what the object (σκοπός, the purpose) of the treatise was: words (ϕωναί), things (πράγματα), or concepts (νοήματα). In the prologue to his commentary, repeating arguments by his teacher Ammonius, Philoponus writes that according to some (such as Alexander of Aphrodisias) the object of the treatise is only words, according to others (such as Eustatius), only things, and according to still others (such as Porphyry), only concepts. According to Philoponus, Iamblichus’s thesis (which he accepts with some specifications) for which the σκοπός of the treatise is the words insofar as they mean things through concepts (ϕωνῶν σημαινουσῶν πράγματα διὰ μέσον νοημάτων [Philoponus 1898, pp. 8–9]) is more correct. From here follows the impossibility of distinguishing logic from ontology, in the Categories, where Aristotle treats things and entities insofar as they are signified by language, and language insofar as it refers to things. His ontology presupposes that being is said (τὸ ὂν λέγεται . . .) and is always already in language, he stresses continually. The ambiguity between logical and ontological is so consubstantial in the treatise that, in the history of Western philosophy, categories will be presented both as kinds of predication and as kinds of being.

א. Our classification of Aristotle’s works derives from the edition Andronicus of Rhodes produced between 40 and 20 BC. We owe to him both the collection of Aristotle’s so-called logical writings in an Organon and the notorious location μετὰ τὰ φυσικὰ of the lectures and notes we today call Metaphysics. Andronicus was convinced that Aristotle was a deliberately systematic thinker and that his edition thus faithfully reflected the author’s intention, but we know that he projected onto Aristotle Hellenistic ideas that were totally alien to a classical mind. The modern editions of Aristotle, however philologically updated, unfortunately still mirror Andronicus’s erroneous conception. We thus continue to read Aristotle as if he really systematically composed a logical ὄργανον, treatises on physics, politics, and ethics, and, finally, the Metaphysics. It is possible to read Aristotle only starting from the destruction of this canonical articulation of his thought.

3

Similar considerations apply to the Stoics’ notion of the sayable. In modern studies, it is taken for granted that the λεκτόν belongs to the sphere of logic, but this makes assumptions (such as the identity between σημαινόμενον and λεκτόν, meaning and sayable) that are far from certain. Let us consider Ammonius’s remarks, who critically defines the λεκτόν from an Aristotelian standpoint: “Aristotle teaches what the things primarily and immediately signified (σημαινόμενα, that is, by names and verbs) and the concepts (νοήματα) are, and, through them, the things (πράγματα), and affirms that we should not think another intermediary in addition to them (that is, the νόημα and the πρᾶγμα), such as that which the Stoics suppose by the name of sayable (λεκτόν)” (Ammonius 1897, p. 5). That is, Ammonius informs us that the Stoics inserted—uselessly, in his opinion—a third element between the concept and the thing, which they called sayable.

The passage in question comes from Ammonius’s commentary on Περὶ ἑρμηνείας. Here Aristotle defined the process of “interpretation” by means of three elements: words (τὰ ἐν τῇ ϕωνῇ), concepts (or, more precisely, the affections of the soul, τὰ παθήματα ἐν τῇ ψυχῇ)—of which words are the signs—and things (τὰ πράγματα)—of which concepts are the resemblances. Ammonius suggests that the sayable of the Stoics is not only not linguistic, but neither a concept nor a thing. It does not take place in the mind or simply in reality; it does not belong to either logic or physics, but somehow lies between them. We should map out this specific location between the mind and things as it may properly be the space of being and the sayable may coincide with the ontological.

4

The richest and, at the same time, most problematic source with which every interpretation of the doctrine of the sayable should begin is a passage from Sextus Empiricus’s Against the Logicians:

Some placed the true and false in the signified thing (περὶ τῷ σημαινομένῷ), others in the word (περὶ τῇ ϕωνῇ), and still others in the motion of thought (περὶ τῇ κινήσει τῆς διανοίας). And the Stoics stood for the first opinion, saying that three [things] were connected with one another, the signified (σημαινόμενον), the signifier (σημαῖνον), and the object (τυγχάνον, “what happens to be,” the existing thing that is at each turn in question). The signifier is the word (ϕωνή)—for example, “Dion”; the signified is the thing itself insofar as it is manifested by it (αὐτὸ τὸ πρᾶγμα τὸ ὑπ’ αὐτῆς δηλούμενον), which we apprehend as what subsists beside (παρυϕισταμένου) our thought, and which foreigners do not understand even when they hear the word; the object is the externally existing substance (τὸ ἐκτὸς ὑποκείμενον) (e.g., Dion himself). And of these, two are bodies, namely, the word and the object, while one is incorporeal, namely, the signified and sayable thing (τὸ σημαινόμενον πρᾶγμα καὶ λεκτόν), which becomes true or false. (8.2 ff.; 1842, p. 291)

The signifier (the signifying word) and the object (the thing that corresponds to it in reality; the referent in modern terms) are evident. What is more problematic is the status of the incorporeal σημαινόμενον, which modern scholars have identified with the concept present in the mind of a subject (like the Aristotelian νόημα, according to Ammonius) or with the objective content of a thought, which exists independently of the mental activity of a subject (like Frege’s “thought”—Gedanke) (Schubert 1994, pp. 15–16).

Both interpretations project onto Stoicism the modern theory of signification and, in this way, omit to tackle a philologically correct reading of the text. The fact that foreigners do not understand the σημαινόμενον when they hear the word could lead us to assimilate it to sense or a mental image (in Frege’s sense); but, opposing the Stoics to those who place the true and the false “in the motion of thought,” Sextus implicitly rules out that the σημαινόμενον could be identified with the thought of a subject. After all, the text clearly says that the σημαινόμενον is not identical with thought, but “subsists beside” it. Even the following passage, which seems to evoke something similar to what moderns call meaning (at least in the sense of Bedeutung or denotation), requires a more careful interpretation. The σημαινόμενον is here defined as the “thing itself” (αὐτὸ τὸ πρᾶγμα) insofar as it is manifested by the word (τὸ ὑπ αὐτῆς δελούμενον—we should notice the repetition of the article τὸ, which I have rendered as “insofar as”).

Like the Latin res, πρᾶγμα means first and foremost “what is in question; what is at stake in a trial or in a discussion” (from here follows its Italian translation into cosa, which derives from the Latin causa), and only subsequently also “thing” or “state of affairs”; but the fact that this passage is not about a thing in this second sense is clear because of its difference from the τυγχάνον, what at each turn happens to be (ἃ τυγχάνει ὄντα), the event or the real object. However, this does not mean that the “thing itself” is simply the meaning, or the signified, in the modern sense, that is, the conceptual content or the intentional object indicated by the word. The thing itself, αὐτὸ τὸ πρᾶγμα, indicates what is in question in the word and in thought; the res that, through thought and the word—but without coinciding with them—is at stake [è in causa]¹ between humans and the world.

As Émile Bréhier observed, the specification “the signified and sayable thing” does not imply that σημαινόμενον and λεκτόν are the same thing, and that the fact of being sayable is the same as the fact of being signified. In his edition of the fragment, Armin inserted a comma between τὸ σημαινόμενον πρᾶγμα and καὶ λεκτόν, which enables us to affirm both the identity and the difference of the two terms. Bréhier in fact concludes that “in general, if the signified is something expressible (this is how he translates λεκτόν), it does not follow in any way that everything expressible is also a signified” (Bréhier 1997, p. 15). Here the interpretation of the syntagm “the thing itself” (αὐτὸ τὸ πρᾶγμα) becomes all the more decisive: what is in question is the thing itself in its being manifested and sayable; but how should we understand and where should we locate such a “thing itself”?

א. Augustine’s Dialectic bequeathed to us an analysis of linguistic signification in which the influence of Varro and Stoicism is evident. Augustine (Dialectic, 5) distinguishes in the word (verbum)—which “in spite of being a sign, does not stop being a thing”—four possible elements. The first is given when the word is uttered with reference to itself, as in a grammatical discourse (in this case verbum and res coincide); in the second—which Augustine calls dictio—the word is not uttered to signify itself, but something else (non propter se, sed propter aliquid significandum); the third is the res, that is, the external object, “which is not the word or the concept of the word in the mind [verbi in mente conceptio]”; the fourth, which translating literally the Stoic term Augustine calls dicibile—“sayable”—is “whatever is perceived from the word by the mind, not the ears [quicquid autem ex verbo non auris, sed animo sentit et ipso animo continetur inclusum].”

Augustine must have found it hard to distinguish between the dictio (the word in its semantic aspect) and the dicibile, since he soon after tries to clarify the difference without really succeeding: “What I have called dicibile is a word, and yet it is not a word, but what is understood in the word and is contained in the mind [verbum est nec tamen verbum, sed quod in verbo intelligitur et animo continetur]. What I have called dictio is a word, which, however, signifies at the same time two [things], that is, both the word itself and what is produced in the mind through the word [verbum est, sed quod iam illa duo simul, id est et ipsum verbum et quod fit in animo per verbum significat]” (ibid.).

We should not lose sight of the nuances through which Augustine tries to define this difference—for instance, resorting to different prepositions. What is question in dictio is something (the signified, or meaning) that remains inextricably linked to the signifying word (it is a word—verbum est—and, at the same time, what is produced in the mind—in animo—through the word—per verbum); on the other hand, the sayable is not properly a word (verbum est nec tamen verbum), but rather what is perceived from the word (ex verbo) by the mind. The aporetic location of the sayable between the signified and the thing is here evident.

5

The phrase “the thing itself” appears in a decisive passage of Plato’s Seventh Letter, a text whose influence on the history of philosophy we are still far from appreciating. A comparison of the Stoic source quoted by Sextus with the philosophical digression of the Letter shows surprising affinities. For convenience, let us here refer to the text of the digression:

For every entity there are three [things] through which science is necessarily generated; fourth is science itself, and as fifth we must posit that same thing through which (each entity) is knowable (γνωστόν) and truly is. The first is the name, the second the defining discourse (λόγος), the third the image (εἴδωλον), and the fourth science. If you wish to understand what I have just said, consider an example, and thereby think about all things. There is something called a circle (κυκλός ἐστί τι λεγόμενον), whose name is the same we have just uttered; second is its λόγος, made of names and verbs: “that which at all points has the same distance from the extremes to the center”: here is the λόγος of what is named “round,” “circumference,” or “circle.” Third comes that which is drawn and rubbed out, or turned on a lathe and broken up—none of which things can befall the circle itself (αὐτὸς ὁ κύκλος), around which the other things mentioned have reference, for it is something of a different order from them. Fourth comes science, the intellect, and true opinion about these things; and all this should be thought of as a single thing, which does not dwell in words (ἐν φωναῖς), or in bodily shapes, but in souls (ἐν ψυχαῖς), from which it is clear that it is something different from the nature of the circle itself and from the three [things] mentioned above. (342a8–d1)

Not only do the words that open the digression—“for every entity there are three [things] through which science is necessarily generated”—duly correspond to the “three [things that] were connected with one another” with which Sextus’s Stoic quotation begins, but the “three” here mentioned (the σημαῖνον or the signifying word—e.g., “Dion”—the real object, τυγχάνον, and the σημαινόμενον) correspond to just as many elements present in Plato’s list. The first, the signifying word (φωνή), corresponds exactly to what Plato calls the “name” (ὄνομα; e.g., “circle,” which he in fact locates ἐν φωναῖς); the second, the τυγχάνον, corresponds to the circle that “is drawn and rubbed out, or turned on a lathe and broken up,” that is, to what at each turn presents itself and happens.

The identification of what, in Plato’s list, corresponds to the σημαινόμενον and the sayable is more problematic. If we identify it with the fourth element, which “does not dwell in words, or in bodily shapes, but in souls,” this would be consonant with the incorporeal status of the “signified thing,” but would entail that it should be identified with thought or the mind of a subject—whereas the Stoic source rules out any coincidence with a “motion of thought.” We are left with the fifth element—the idea—whose technical denomination (the circle itself, αὐτὸς ὁ κύκλος) the...