Ontology and Language, between Chrysippus and Agamben
by Nicoletta di Vita, trans. by Kurt Lampe
1 Philosophy and Language
Of the many connections between philosophy and language in the ancient world, one has received particularly copious attention from researchers. I am speaking of the manner in which philosophy has integrated language into its own tonalities, which today—because of a forgetfulness that is hard to admit—we have difficulty in understanding.
The phrase logon didonai (giving a logos), which can be found in many places in Plato, is probably the most complete form in which this encounter is expressed.1 It originates in the domain of music, from which all reflections on language are likewise descended. As Johannes Lohmann has demonstrated, the relations among sonorous tonalities were originally called logoi (Lohmann 1970: 8–10).2 By “giving a tonality,” these logoi create a harmonic relation between the musical form and the emotional tonality, between the sound given externally and the Stimmungen (pitches or moods). It is precisely this “acoustic-musical phenomenon,” this need for harmonization, that soon become, according to Lohmann’s hypothesis, a “model, among the Greeks, for every kind of knowledge and experience” (1970: 8).
It is widely agreed that all later Greek philosophy ended up making logon didonai its proper task (Ildefonse 1997: 38). The result was that all philosophical reflection on logos was configured as reflection on correct or orthos logos, dedicated to an orthē philosophia. When Plato, in the Cratylus and Sophist, makes words and utterances his object of study, he is researching in what form language can be correct. He is testing the medium of his own dialogues, the form assumed by philosophical exercise—in order to evaluate not the power of language but the power of philosophy, of its relation to reality. Similarly, when Chrysippus claims that only the dialektikos—in other words, loosely speaking, the logician—possesses virtue,3 and therefore we must work on language (dialegesthai) because it leads to knowledge,4 it may well be that the goals of his appeal to theoretical work on language are entirely external, strictly speaking, to language itself.
It is possible to read all of this as a decisive step toward heteronomy, as some have already done; in other words, toward the determination of language’s status on the basis of its function specifically for thinking: classical philosophy imposed its own norms on language, according to its own needs. Thus Joly has spoken of a “linguistic blockage” (1986: 105–36),5 of a shadow projected by philosophy onto language. The latter has primarily been investigated with regard to either “its instrumental role in the expression of thoughts” or “its mediating function in the representation of extralinguistic reality” (Coseriu 2003: 14), the goal being “to provide a system of expression that apprehends ‘things as they are’” (Imbert 1993: 308). In fact, it has often been proposed that the late emergence of linguistic disciplines (grammar, philology, linguistics), that is, of disciplines explicitly dedicated to the investigation of language solely for its own sake, was due to such an original “stranglehold (mainmise) on language by philosophy” (Ildefonse 1997: 15).
This situation is in some respects so obvious that it has been possible to speak of a kind of “pillaging” of language’s fundamental structures by philosophy. In his well-known article from 1958, Émile Benveniste suggested that when Aristotelian categories were held up to the light, one could see the skeleton of common Greek verbal forms of the period: Aristotle “thought he was defining the attributes of objects,” he says, but “he only posits linguistic beings: it is language that, thanks to his proper categories, permits him to recognize and specify those attributes.” He goes on, “It is what we can say that delimits and organizes what we can think. Language furnishes the fundamental configuration of properties that the mind recognizes in things” (Benveniste 1966: 70). Therefore the “stranglehold,” so to speak, would be indirect: a tacit but intense reflection on language will have been disguised as an inquiry into the categories of thought and subjugated to those categories.
1.1
At least two points might follow from the foregoing considerations. The first is that language’s implication in philosophy has not necessarily caused its own innermost traits to be neglected. To the contrary, it has on each occasion allowed us to glimpse language’s ethical, epistemological, political, or religious nature, and therefore to escape language’s isolation by apprehending it in the actual constellation of its multiple relationships. It has opened the possibility of approaching language as it participates in all things, by revealing the complexity of its nature.
As Eugenio Coseriu fully appreciated years ago, the disciplines that make language their object “begin” properly “at the point where philosophy of language stops” (2003: 13). They have stopped posing the question surrounding language in order to begin giving, each from its partial perspective, specific answers.
There is thus something more in philosophy’s remarkable affection for language. Moreover, this leads us, with our second observation, to the other way of connecting logos and philosophy.
2 The Task of Philosophy: An “Unsayable” Passion
In a statement attributed to the Stranger in Plato’s dialogue the Sophist, we find a rather eloquent expression: “If we were deprived of this [logos]—and this would be most serious—we would be deprived of philosophy” (260a). The intuitions of the Italian philosopher Giorgio Agamben and his reading of the ancient world appear to set out from just this point: not only that we understand language via a philosophical perspective, but also, in the ancient world as in today’s, that philosophy should be understood precisely in the relation it establishes with language.
In a text originally published in 1984, Agamben had posed the question about the proper task of philosophy very clearly, when he wrote: “The task of philosophical exposition is to come to the aid of speech [parola] with speech, so that, in speech, speech itself does not remain presupposed to speech, but comes to speech as speech” (2010: 18). The reference in this curious formulation is to “coming to the aid of” (boēthein) in Plato’s Phaedrus (278c6). In the Platonic passage we read that the only person who merits the title “philosophical” is one who, in their own words, can give a reasonable account of their use of language (278d4). This means, in terms Agamben describes as “modern,” giving an account “of the very fact of language,” of this presupposition that remains like an a priori of every linguistic act. But because this presupposition also and always precedes the person who speaks and the very act of speech, it is a truly “unsayable” and “inexpressible” foundation of that act (2010: 30–4). So the question is the following: How can we philosophically conceptualize and attempt to express this taking-place of language, this “originary impossibility?”
This formulation is obviously close to the Wittgenstenian maxim, according to which “was gezeigt werden kann, kann nicht gesagt werden” (“What can be shown, can not be said,” Tractatus 4.1212): it is not possible to say language itself with language. The mediation of Martin Heidegger is equally manifest: he had assigned to this taking-place, which we cannot express in or with language, Dasein’s opening to itself, “die primäre Entdeckung der Welt” (“the primary discovery of the world,” Heidegger 1962: H138). “Not how the world is,” comments Agamben, “but that the world is”—not what one says “propositionally inside of language [linguaggio], but that language is” (2010: 80).
If this problematic appears peculiarly modern, the merit of Agamben’s research consists above all in delineating its archaeology: he discovers the matrix of this demand, which belongs to the twentieth century, in what Plato had already designated as the mark, as Agamben says, “of authentic philosophical exposition,” but which only Stoicism, through its original teaching, would have brought to perfect expression.
2.1
It is in fact with a presentiment of impossibility that Agamben first turns to the Stoics. The passage concerns the Stoic theory of the passions, with an innovative take on the Heideggerian theory of Dasein. Dasein faces the opening—which is an opening to language—with Angst, anxiety. “This opening has always been traversed by a negativity and a malaise” (Agamben 2010: 82), because Dasein is never master of this opening, of this event that assails and determines it (and which “coincides with the proper place of man’s being, with his Da,” 2010: 83).
Heidegger’s thesis that this originary opening is an emotional tonality finds an analogue in the ancient world. It is worth noting that Heidegger himself indicates that Aristotle’s theory of the passions was the object of his Rhetoric and not of a “psychological” treatise (Heidegger 1962: H38). Even more interesting for us is that, as Agamben emphasizes, the key locus is ultimately to be found in Stoicism (Agamben 2010: 84).
In Chrysippus we discover a fundamental connection between language and passions, logos and pathē. Only humankind can experience passions, because passions are actually a type of logos, a krisis or “judgment.” They must be defined in relation to logos, never in opposition to it (DL 7.111). It is here that Agamben rad icalizes the Stoic theory and understands passions as an excess of logos, that is, as its “emergence.” As both origination and “passion,” language remains excessive for individuals; the Stoics give the following definition: “Passions are an excessive impulse or one that exceeds the measure of logos” (SVF III.377).6 Here Agamben obviously displaces to the background the ethical and practical horizon that usually guides analysis of the Stoic theory of the passions. Moreover, he does not directly address the complex way in which the Stoics defined passion: not only as a krisis and a logos but also as “a movement of the soul that is irrational [alogon] and contrary to the nature of the soul” (SVF 3.462)7 —something that was already deemed “contradictory” in antiquity (SVF III.456-90, LS 65J). That being said, Agamben nevertheless seems to grasp a critical moment of the declension of passions in terms of logos: “Ὁρμή [“impulse”],” he specifies, “comes from ornumi, which has the same root as Latin orior and origo, and means . . . an emergence, an origination [uno scaturire, una origine] that surpasses the measure of language” (2010: 84). And if passion is not other than logos itself, “the excessive origination must be that of language itself” (2010: 85).8 Thus he can conclude that “the theory of the passions, of Stimmungen, has always been the place where western man thinks his fundamental relation to language” (2010: 85). And this thought is always the thought of an impossibility: the excessive origination of language, which encompasses the very fact of its existence, appears inapprehensible. It escapes us because of its very excessiveness. Yet for the philosophy that has made language its fundamental theme, it never ceases to exhibit its peculiar demand, which its relation to the passions pinpoints and outlines, but never manages to satisfy.
In antiquity as today, the question can ultimately be formulated as follows: Is it possible to elaborate “a discourse which, neither being reduced to a metalinguistics nor halting in the face of the unsayable, says language and exposes its limits” (Agamben 2010: 34)—which would situate itself at the level of philosophy as “vision of language” (Agamben 2010: 35)? “An ancient tradition of thought,” Agamben long ago announced, “sets forth this possibility” (Agamben 2010: 34).
3 Factum Linguae
It is perhaps surprising that Agamben once again refers to ancient Stoicism in order to clarify what is at issue in the “very fact of language.” This is indirect at first, when he traces Augustine’s and Gaunilo’s observations about a situation where someone hears a word unknown to them, and understands it as a word, and not as a simple sound (1982: 45–7). In Augustine’s case, the example, taken from De Trinitate (X.1.2), concerns a “dead word” (vocabulum emortuum), specifically temetum (an ancient word for vinum, “wine”). Augustine contends that the desire to investigate the meaning of the word testifies to a fundamental condition: that this term is presumed to be a meaningful sign, that the hearer is conscious of being in the presence of something endowed with meaning. Similarly, in his response to Anselm, Gaunilo once again presents the case of someone who, while not knowing the meaning of the voice they hear, still recognizes it as meaningful. Gaunilo’s cogitatio secundum vocem solam signifies precisely the comprehension of voice qua indication of an event of language (Pro insipiente 0345b). But a similar observation was already implied by Stoic texts, which Agamben does not cite. According to Sextus Empiricus’s testimony (M 1.154-8), the Stoics, in order to understand what logos is, imagined a circumstance where foreigners (barbaroi), having heard the local language, grasped nothing of its meaning. Nevertheless, they will not have doubted that they were in the presence of a language event, and this is because they heard the voice. Diogenes Laertius also attributes to the Stoics a discussion about the fact that a term like blituri, although deprived of all meaning, still is easily recognizable as part of a linguistic act (DL 7.57 = LS 33A, SE M. 8.133).
But what is this consciousness of a level that, despite being linguistic, remains below the threshold of any question about language in order to recognize it as such? Which is not yet in language, but all the same grasps and expresses it? In other words, what is the place of that which, as Agamben has nicely phrased it in a recent work, “is neither only something linguistic nor simply something factual” (2015: 58)?
4 The “Sayable” alongside Thought, Language, and Things
This question receives a decisive formulation in a recently published chapter “On the Sayable and the Idea,” where Agamben discusses the Stoic sayable at length alongside other ancient and medieval theor ies (2016: 57–122). This happens through, on the one hand, the demarcation and deepening of an interpretation of the Platonic Idea, to which Agamben had already dedicated an essay in 1984, and on the other hand, through an analysis of one of the most important topics in Stoic theory, namely that of the lekton or “sayable.” The reason is simple: the unsayable, already for Walter Benjamin, is to be “eliminated in language” (Benjamin 1995: 325–6, italics my own). It cannot remain outside of language “like an obscure presupposition.” “The elimination of the unsayable in language coincides,” Agamben observes elsewhere, “with the exposition of the sayable” (2016: 59).
Let us immediately make clear that Agamben’s reading of this issue in no way resembles those to which we are accustomed. The theory of the lekton is not investigated either on the basis of its inclusion among the incorporeals or in its difficult relation to the theory of propositions. It is instead viewed, one might say, from the point of view of its “demand” (esigenza), as a function of its position in the general conception of philosophy. Agamben is obviously fascinated by a certain undecidability in situating the fourth Stoic incorporeal, and the conclusion he reaches is, for that reason, particularly interesting. In his opinion, it is the ancient Stoics who, even if they did not fulfill it, at least recognized philosophy’s fundamental task. They managed to find a way of expressing the fact of language, the demand for a philosophical presupposition and thematization.
The interpretation Agamben offers is all the more interesting because it is by no means the product of second-hand analysis or a hasty glance at the sources. To the contrary, his appropriation of Stoic thinking is always conducted with extraordinary philological care.
4.1
The suspicion that the Stoics could provide an important perspective for our inquiry derives from a passage in Ammonius’s commentary on Aristotle’s On Interpretation:
Here Aristotle teaches us what is primarily and immediately signified (σημαινόμενα) by them [nouns and verbs], namely thoughts (νοήματα), and by their mediation, things (πράγματα); and he claims that it is unnecessary to conceive of any other intermediate . . . as the Stoics suppose with the name lekton. (SVF II.168 = LS 33N)
Alongside language, concepts, and things, Ammonius claims that the Stoics posited a fourth element: the lekton. In other words, something else, which cannot be included among things, among thoughts, or even, properly speaking, among words. Its entire place is indeterminate, and so we must, as Agamben himself writes, trace its “cartography” (Agamben 2016: 62).
In order to argue for what was probably a long-standing intuition, he has recourse to a passage from Sextus Empiricus’s Against the Professors (8.11-12 = LS 33B),9 which he describes as “the most complete and at the same time the most problematic” for our question (2016: 63). In this passage, which recent historiography has considered a veritable locus classicus (Frede 1994: 118), Sextus clearly distinguishes “three aspects” that, he says, “are conjoined” (τρία φάμενοι συζυγεῖν): the sēmainon, or signifying sign, that is the phōnē or vocal sound, which he stipulates is a body; the tunkhanon, that is the object, that which is at issue on each occasion, which is also a body; and finally the sēmainomenon, which is not a body, “subsists alongside our thought” (τῇ ἠμετέρᾳ παρυφισταμένου διανοίᾳ), and is not understood by “barbarians,” even when they hear the vocal sound. According to Sextus, this is properly the “signified thing” (τὸ σημαινόμενον πρᾶγμα), which is “incorporeal” (ἀσώματον), “revealed by the voice” (ὑπ’αὐτῆς ⟨τῆς φωνῆς⟩ δηλοῦμενον), true or false, and which, finally, the Stoics call lekton.
The passage is well chosen. Without needing to refer to any “fourth” thing, Sextus expresses all the complexity contained in what is neither word nor thing, but that, in Ammonius’s testimony, has to do with both.
The passage is also interesting because it offers, in a way, a concise abstract of what we might call a “classical” theory of the parts of logos.10 This theory recognizes a tripartition that, in its broad outline, can still be accepted today; more importantly, versions of it can be found in many ancient authors. But in setting this passage before us once more, Agamben takes some important steps. He translates the second τὸ of the expression αὐτὸ τὸ πρᾶγμα τὸ ὑπ’αὐτῆς δηλοῦμενον as “inasmuch as” (in quanto)—“the thing itself inasmuch as it is revealed by” the vocal sound—and emphasizes the description of the σημαινόνενον πρᾶγμα (“the signified thing”) as αὐτὸ τὸ πρᾶγμα, “the thing itself.”
4.2
It is well known that among modern commentators there is no agreed reading of either this passage or the lekton in general. In Sextus’s formulation, the lekton is entirely reduced to the sēmainomenon, the signified, which absorbs many of the features generally assigned by the sources only to the incorporeal.
Isnardi Parente has shown, with regard to the “signification” posited as incorporeal, that “it would be hard to discern what distinguishes it from a simple concept” (2005: 180). That the passage is rather obscure is confirmed by Michael Frede, who remarks, in “The Stoic Notion of a Lekton,” that “there must be something wrong with Sextus’ account” (1994: 119). The difficulty derives not only from the confusion of lekton and signified but also, as Brunschwig too elucidates (Brunschwig 1995: 8), from the fact that this is “the only text which attributes to the Stoics the view that there is also a lekton corresponding to an expression like ‘Dion’” (Frede 1994: 119). Rather than “Dion,” Sextus should have written, based on comparison with a text by Seneca (Letters 117.113 = LS 33E), something like “Dion is walking” in order to satisfy the propositional character assigned to the lekton by the Stoics.11
As for the status of the lekton, there really is no agreement. For example, some have suggested a reading we might call “Fregean,” according to which lekta would be “independent contents of thought,” that is, entities independent of both thought and language (Frege 1892: 22–50). By contrast, others have insisted on the ontological dependence of lekta (Long 1996), which themselves depend on rational impressions (logikai phantasiai)—in other words, on that which the Sto ics propose as the site and origin of all human knowledge.
An echo of this controversy reaches Agamben as well. Following Schubert’s arguments (1994: 15–16), he proceeds to ask himself to what extent the sēmainomenon should be understood as the objective content of thought, as in Frege, or as subjective content, more specifically as “the concept present in the mind (mente) of a subject (like the Aristotelian noēma, according to Ammonius)” (Agamben 2016: 63–4).
If we inflect this question a little and reorient it to our own interests, this amounts to asking whether the lekton should be situated among thoughts or among things—whether its place is a logical place or exclusively ontological.
Although Sextus has identified the lekton with the signification of the utterance, and therefore emphasizes its metaphysical status, nevertheless, as Émile Bréhier also says (whom Agamben quotes): “If the ‘signified’ is a ‘sayable,’ we by no means see that every sayable is a signified’” (Bréhier 1962: 15)—in other words, that the fact of being sayable is identical with the fact of being signified. Conflating the two is inappropriate, and the issue remains, as Agamben himself says, “aporetic.” He suggests that it remained so for Augustine, who, by distinguishing res (the bodily thing), dictio (verbum and per verbum, that is, speech in its semantic dimension), and dicibile (the sayable), called the latter “speech and not speech,” not a verbum properly speaking, but ex verbo—and did not thereby escape from this confusion (Agamben 2016: 65–6, citing De Dialectica 5.8).
Therefore Agamben’s suggestion, which may surprise us a little, is that the “philologically correct” analysis of the passage in Sextus is the only one to offer a genuine answer to this aporia.
4.3
We can follow Agamben’s argument by distinguishing two moments, whose fundamental starting point is Ammonius’s indication of the presence of a fourth element and whose conclusion is that the lekton does not belong to the domain of thoughts. A first observation is that Sextus explicitly distinguished the Stoics from those who located “the true and the false” (τὸ ἀληθές τε καὶ ψεῦδος) in “the movement of thought” (περὶ τῇ κινήσει τῆς διανοίας, M 8.11 = LS 33B1). Therefore the lekton, to which, according to Sextus’s statement, the true and the false belong, can in no way coincide with thought. Moreover, Sextus placed the lekton “alongside” thought—it “subsists alongside thought” (παρυφισταμένου διανοίᾳ), he says, separating it from things in the mind. The second observation concerns the unusual description of the lekton as αὐτὸ τὸ πρᾶγμα, which Sextus mentioned near the passage’s beginning, when he says, “the signified is the thing itself inasmuch as it is revealed by it [the vocal sound].” Now, Agamben deliberately translates this as “the thing itself” (la cosa stessa). But does this mean that the lekton belongs to the world of “things?”
The solution is once again to be found through philological attentiveness. By choosing the word pragma, Sextus has not used an equivalent for tunkhanon, the real event or object, that which “happens to be” (ἅ τυγχάνει ὄντα). The pragma is certainly a “thing” (cosa), but its semantic range also encompasses, Agamben asserts, “that about which one speaks,” “that which is in question,” which is “at stake (in causa),”12 as in the Latin res, which signifies “what is at stake in a discussion or a procedure” (2016: 64).13 Similarly, the verb used here (ὑφεστάναι)—which, we might add, Chrysippus employs for the past and the future in opposition to the present (SVF II.509 = LS 51B)—signals explicitly that we are not dealing with a thing in the proper sense of the term, namely as existent (endowed with huparxis).
This is why Agamben’s translation of the end of the passage is well founded: the lekton is of course a “thing,” but only inasmuch as it is revealed by the phōnē or vocal sound—and this means: in its being-revealed by the phōnē, considered solely inasmuch as, to the extent that the phōnē displays it. When Sextus stipulates that the lekton is “the thing inasmuch as it is revealed by” the vocal sound, he precisely indicates, Agamben claims, with an unexpected clarity, “what is in question in speech and in thought: the res which, through speech and thought, but without coinciding with them, is at stake between a human being and the world” (2016: 65).
The secret of the lekton is entirely hidden in such a “being in question” of language in language itself, which is of course even more mysterious. Let us therefore question once more the place of this “inasmuch as.” What does it mean that the lekton, neither thought nor thing, expresses precisely “that which is in question” in human speech?
5 The Platonic Idea as Sayability
A glimmer of illumination comes from Plato. It is not by chance that Agamben has chosen the passage in Sextus; his heavy emphasis on the words αὐτὸ τὸ πρᾶγμα, which appear in an entirely negligeable position in the passage as a whole, is not without purpose. The reason, if we look closely, is rather obvious. In his essay on Plato from 1984, Agamben had already addressed the Idea as “the thing itself” (2010: 15–20). I will not fully retrace his detailed arguments there and in his chapter from 2016; it will suffice to summarize briefly the confrontation he orchestrates between the passage from Sextus and one from Plato’s Seventh Letter (342a8-d1).
The Platonic passage is relatively well known. There Plato indicates the five elements necessary for achieving epistēmē, the science or understanding of “each being,” taking as his example the circle. Epistēmē itself, as the text explains, is the fourth element. The others are given as onoma (name), eidōlon (image), logos, and a curious “that through which each being is knowable (gnōston) and truly is.”
We should note that once again Agamben’s translation deliberately departs from what is commonly accepted. Plato’s text is usually printed and translated as follows: πέμπτον δ’αὐτὸ τιθέναι δεῖ ὃ δὴ γνωστόν τε καὶ ἀληθῶς ἐστιν ὄν, “as fifth, we must posit that very thing which is to be known and is truly being” (342a8-b1, italics my own). But in his 1984 essay, Agamben had discussed the need to revise the editions of Burnet and Souilhé in order to restore the manuscript reading (which was the version used by Marsilio Ficino as well). If we retain the passage in its current form, the result will be that the fifth is simply the thing that is the object of knowledge (Agamben 2010: 13). But this would create a pointless circularity by putting in the fifth place that which was presupposed by the entire discourse (in other words, “each being”). It would also confirm Aristotle’s interpretation, since he “sees in the Idea a kind of pointless duplication of the thing” (Agamben 2010: 13). By contrast, the manuscripts, according to Agamben (who relies on the work of other scholars),14 actually read δι’ὁ rather than δεῖ ὃ; in other words, “through which” or “by which” (2010: 14). Therefore Agamben claims that the fifth element is the Idea, whose “technical denomination” in the passage is αὐτὸς ὁ κύκλος (“the circle itself, this circle here”)—a formulation very close to Sextus’s expression for the lekton, αὐτὸ τὸ πρᾶγμα, and, one might say, a rather special way of being a “thing” (Agamben 2016: 68). The interpretive context is soon clarified: in order to analyze the formula chosen by Plato for the Idea, Agamben postulates that autos, though it functions adjectivally here, still retains the anaphoric force of its pronominal usage.15 In “αὐτὸς ὁ κύκλος,” he says, the αὐτὸς only designates something, as with all anaphoric pronouns, “inasmuch as this thing has already been signified by means of another term endowed with sense.” The anaphoric pronoun, lacking any virtual reference, acquires one through its relation to a term that precedes it.16 Unlike the case where autos appears between the article and the noun (ὁ αὐτὸς κύκλος, “the same circle”) and expresses identity (Latin idem), in the singular Platonic formulation αὐτὸς ὁ κύκλος, it expresses ipseity (Latin ipse). As Agamben puts it, it does not refer to kuklos alone (ὁ αὐτὸς κύκλος), but to the syntagm ho kuklos (αὐτὸς ὁ κύκλος) (Agamben 2016: 79–82). Moreover, Agamben observes that the article ho is also by origin an anaphoric pronoun. This is to be connected, via a meticulous analysis of the passage, with the expression κύκλος ἐστι τι λεγόμενον, “circle is something said” (Agamben 2016: 79–82).
Agamben can therefore conclude:
The αὐτὸς, since it refers to a term already anaphorized by the article, resumes the circle in and by its being-said, in and by its being in language . . . . The circle itself—the Idea—is neither one of the four elements, nor yet simply other than those four. It is that which is in question each time in each of them, and at the same time, it remains irreducible to them: it is that through which the circle is sayable and knowable. (Agamben 2016: 82)
5.1
Let us step out of this philological labyrinth for a moment and ask: What does the passage of the Seventh Letter, interpreted in this way, tell us about the place of the lekton? According to Agamben, there are intriguing affinities between this passage and that of Sextus. The initial listing of elements is similar, and the use of the name “Dion” in Sextus’s example recalls the Dion of the Seventh Letter—whereas, Agamben notes, Aristotle had substituted the names Choriscus or Callias in his arguments. Moreover, there are also correspondences between the individual elements: to Sextus’s phōnē/sēmainon corresponds Plato’s onoma, which indeed he describes as being en phōnais; to Sextus’s tunkhanon corresponds Plato’s eidōlon, since both are particular occurrences (Agamben 2016: 67–8).
It remains to discover the place of the sēmainomenon/lekton. We have already denied that it is thought, so it cannot respond to epistēmē in the Seventh Letter. The lekton can only correspond to the fifth element in the Platonic passage—to the Idea.
As is often the case in Agamben’s texts, the argument comes full circle with impeccable coherence. Did not Sextus describe the lekton with the expression αὐτὸ τὸ πρᾶγμα? If we accept Agamben’s contention that Plato’s Idea is expressed in the anaphoric movement of αὐτὸς ὁ κύκλος, the operation that results is extraordinary: for him, the Stoics situated the sayable in the same place as the Idea. This does not mean that they equated the lekton with the Platonic Idea, but that they conceptualized the sayable, this strange fourth element, in the same place between language and reality that Plato had attributed to the Idea—an uncomfortable place, difficult to express, yet fundamental for Platonism just as for Stoicism.
“The history of post-Platonic philosophy,” Agamben says, “can be read, starting with Aristotle, as the history of various attempts to eliminate the Idea or think it differently” (Agamben 2016: 68). In Aristotle’s On Interpretation, for example, this occurs through an examination of language, whose approach is similar to those we have just analyzed, but adds as its fourth element gramma (writing or letter), and deliberately omits the Idea (Arist. Interpr. 16a3-7).
Modern commentators—Agamben still refers explicitly to Schubert (1994: 15)—have remarked upon this proximity of the Idea and the sayable, albeit in the negative.17 This nevertheless corroborates Agamben’s contention: the Stoics borrowed from Plato the Ideas’ mode of existence, modeling the lekton on it. Moreover, just as the lekton, like every incorporeal, has its “being” as paruphistasthai, “subsisting alongside” thought, similarly the Idea is before all else a paradigm—para-deigma, what shows itself “alongside (para) things.”
“The thing itself,” Agamben had said in discussing the Platonic Idea, “is not a thing: it is sayability itself, the opening that is in question in language, which is language, and which, in language, we constantly presuppose and forget” (2010: 18). Thus the merit of the Stoics is to have given this fourth element, which they rehabilitated, its proper denomination: the “sayable.”
It is easy to see here how not only the Stoic theory of the lekton but also the Platonic theory of the Idea receives a new interpretive possibility. Was it not said, in this same passage of the Seventh Letter, that the Idea is not “sayable” (rhēton) like other objects of learning (341c5-6)?
But Agamben invites us to re-read the passage in the Parmenides where Plato, speaking about the Ideas, says that they are “the things you can most grasp with logos” (ἐκεῖνα ἃ μάλιστά τις ἂν λόγῳ λάβοι, 135e3). The Idea, in its proximity to the lekton, is what is most (malista) sayable, the place where sayable and unsayable meet—and the Stoics deserve credit for illuminating, with their terminology, the paradoxical place of the fact of speaking, of this inasmuch as, this unsayable sayability.
Before all else, language is a fact, a fact that escapes us, and yet demands expression. That the lekton is a fact, the expression of an event—this is well known; Agamben specifies which fact we are dealing with, namely the fact of speaking.18
6 Discussion
At this point we can articulate some critical reflections. Almost all of them derive from the fact that the original question does not really appear to have been resolved by this comparison of lekton and Idea. The place that has been identified for the lekton remains, despite everything, a problematic position, which clearly exhibits what we understood at the beginning of our discussion as the “excess” of logos, the impossibility of apprehending what always precedes us.
6.1 The Logical Value of the Lekton
One might ask—and this is not only the first but perhaps also the most important question—whether the lekton should not in every case be understood against the linguistic background of discussions of logic, in which it actually appears, and from which it takes its very name.19 It is well known that Chrysippus generally spoke of the lekton as an axiōma (“proposition”), even where it is an incomplete axiōma,20 and thus always highlighted its logical value. Moreover, the most canonical ancient definition, which one might call foundational for our interpretation of the Stoic lekton, runs as follows:
The lekton is what subsists in accordance with a logical impression (κατὰ λογικὴν φαντασίαν ὑφιστάμενον), and a logical impression is one in accordance with which the object of the impression (τὸ φαντασθὲν) can be represented discursively (λόγῳ). (LS 33C = SE M. 8.70)
This means that the lekton, which some modern readers understand and translate as “what is said” (Long 1971: 77), should be understood as something like “the sense of the statement, which the statement exhibits discursively.” Some scholars even go so far as to claim that it only exists “in the moment of its utterance” (Ildefonse 1997: 138). But neither the sense of kata (“in accordance with”) nor the place of to phantasthen (“the object of the impression”) and its relation to the lekton have truly been made clear (Alessandrelli 2013: 65–75). Most importantly, what does it mean to assign the lekton to “the domain of logic?”
We should acknowledge that Agamben’s reading, by firmly resisting the modern inclination to distinguish what concerns language from what belongs to another area, is exceptionally faithful to the unitary structure of Stoic philosophy. No exteriority is possibly in the Stoic system. The great value of Agamben’s analysis, without any doubt, is that he not only recognizes this principle but also brings it to completion. And if this does not yet guarantee that the lekton has the ontological value he proposes, at least it clearly makes that plausible.
It is interesting, in any event, that he is entirely aware of this line of questioning. He argues that, although the context in which the sayable appears is almost everywhere linguistic, it is necessary to attend to the traces of its non-linguistic status preserved by our sources.
He refers to two passages in this context. The first, once again from Sextus, maintains that the Stoics say the lekton λέγεσθαι δεῖ, “must be said” (M. 8.80 = SVF II.167). Is it possible to “say” something that is linguistic? A passage from Diogenes Laertius suggests an answer, and Agamben reports it. In the seventh book of his Lives, Laertius specifies that προφέρονται μὲν γὰρ αἱ φωναί, λέγεται δὲ τὰ πράγματα, ἃ δὴ καὶ λεκτὰ τυγχάνει: “words are uttered, but things are said, and those things,” he continues, “are lekta.”21 Agamben’s conclusion follows directly from this: “the act of speech,” he says, “and what is in question in it are different” (2016: 69).
6.2 The Metaphysical Origin of the Lekton
At this point I would like to concisely call to mind some interpretations of the lekton that, at least prima facie, appear similar to those we have just outlined.
That the lekton did not originally belong to linguistic theorizing has long been acknowledged by some commentators. Michael Frede, for example, has shown how the term’s predominantly linguistic usage was only imposed by Chrysippus, who oriented it toward both the proposition (axiōma) and the predicate (katēgorēma). Frede speaks of a “metaphysical notion, the notion of a fact” (1994: 113), at least when it comes to the lekton’s first occurrences.
Cleanthes, the first Stoic to have used the term (according to Clement of Alexandria, Strom. 7.9), used it “metaphysically”: he viewed it above all as the incorporeal effect of a (corporeal) cause,22 and did not yet view such an effect as an “incomplete (ellipes) lekton” in the linguistic sense—though the presence of a logical term in the context of physics is, to say the least, “remarkable” (Ildefonse 1997: 189). Note also that Brunschwig views the lekton as “defined in terms of physics, and not of logic, dialectic, or semantics, at least in the beginning” (1995: 91).
But in truth, we can see that even this is different from Agamben’s recuperative reading. The Italian philosopher approaches the lekton in relation to linguistic questions, although it is also something more. His attempt to uncover its non-linguistic and “objective” nature (in the sense we have seen) cannot be confounded with what scholars have intended by speaking of the “metaphysics” of the term lekton. For Agamben, the lekton continues to be the sayable, what can be or must be said, “the thing,” certainly, but in its sayability and inasmuch as it is grasped in the act of being said. Its value is like that of an announcement. Far from being something properly existing or having the function of signifying, it is a sign—the sign that it is possible to speak, that the thing, in entering into language, leaves and expresses the trace of its own sayability. On Agamben’s reading, the lekton is below the threshold of any action in language, a simple announcement of this possibility itself, an index of the fact of speech, factum linguae.
Despite grasping and articulating the common ground of language and ways of being, which is one of the facets of Stoicism that most interests contemporary readers (the verbal nature of the event), Agamben appears to have neglected the full scope of the lekton: in other words, from the concept of a predicate to Chrysippus’s concept of a “complete sayable.” This is because that scope is extraneous to his inquiry into the status of language—as indeed was the case for his commentary on the theory of the passions, whose practical importance he neglected, but whose philosophical force he elucidated in a novel manner. In his work we see for the first time the lekton’s role as, before all else, the sign of a demand, which only subsequently branches into its actual, well-known, possibilities, all of them internal to language. Perhaps this reading is better placed to provide an answer to our initial question about the relationship between philosophy and language, below the threshold of any linguistic discipline: it chooses to be a preliminary and almost foundational glance at the lekton, the “sayable”—mark and sign of sayability.
7 Conclusion: Philology and Critique
Does it matter for contemporary thought whether or not Chrysippus acted consciously in expressing the demand of the unsayable? What Agamben wants to show is clear: the Stoics were able to apprehend the “fact of language” before diving into its analysis and comprehension. But that they associated the “announcement of a demand” with the lekton, this fourth element that Aristotle would not understand, cannot perhaps be effectively demonstrated except through the arguments we have reconstructed here.
It is obvious that Agamben has identified a hazy area in philosophical thinking and tried to understand it from the outside, so to speak. His approach does not seem to aim at understanding the lekton, but rather philosophy through the lekton. Perhaps this finds some justification in a claim made by Hülser and many others, namely that the sources for the Stoic lekton only furnish us with a very simplified version of the theory. They are “very elementary and not very satisfying: their exposition is essentially incomplete” (Hülser 1987: 832). This inspires us to go beyond the letter of the texts, to grasp the theory’s role in the broader preoccupations of ancient philosophy, and, with it, of philosophy as a whole, on the basis of a contemporary urgency.
And yet, the most striking feature of Agamben’s proposal is its obvious emphasis on philological rigor. All his answers are grounded in the text of his sources, and every connection that might appear illegitimate on the historical plane (perhaps including that between the sayable and the Idea) is, so to speak, justified on the “literal” plane. Every external perspective is complemented and constrained by the most tenacious philological care, which remains entirely delimited and defined by the evidence.
In the short text entitled “Project for a Review” in Infancy and History, which was supposed to introduce a program suspended long ago, Agamben locates in “philology” a way of reappropriating the past. “Philology,” we read, “awakens the myth from its archetypal rigidity and its isolation, and returns it to history” (2001: 151). It is what recuperates “in a critical manner” “the fracture in speech,” which coincides with the division between “content” and its “historical transmission” (2001: 147).
What Agamben adds to the understanding of ancient thought should perhaps be grasped from this angle: a critical philology that aims at the arkhē operating in history—not in the sense of a chronological origin, as in Walter Benjamin’s famous formulation, but as a presence that still and always proposes itself and makes itself visible in the historical process, though without ever becoming identical with that process.
Thus, archaeological work explores the places where thought—this unique thought that connects and establishes “western history”—finally makes itself intelligible, freed from what makes it inaccessible, ossifies it into fixed categories, and locks away all its hidden potential. This is not simply a matter of dialogue—between us and the ancients, between the contemporary demand of the unsayable and the Stoic expression of the sayable. Stoicism comes closer to us in a concrete sense, alive with the questions it poses and the answers it gives. By suspending thought’s allegiance to chronological sequencing, we recognize Stoicism as an interlocutor like others, and better than some others. It becomes not only lively but also influential—“foundational” even, as Pohlenz said, “for western philosophy of language” (1939: 151)—but through a perspective that is original and profuse. The Stoic “revolution in logic,” whose motivations lie, as Émile Bréhier writes, “in physics” (1962: 13), becomes with Agamben an extraordinary emergence, the very first, the most foundational, for “ontological” reasons.
Notes
1 Plato Phaedrus 95d7, Cratylus 426a2, Theaetetus 175c8, Republic 533c2.
2 For a discussion of the relationship between language and music based on ratios and proportions, see also Koller (1955) and Georgiades (1958).
3 διαλεκτικὸν μόνον εἶναι τὸν σοφόν, DL 7.83.
4 To τὸ ὁμολογουμένως τῇ φύσει ζῇν (DL 7. 87 = SVF I Zeno 179 = LS 63C), “ethical life in agreement with nature.”
5 Grammar is a textbook case. Ildefonse has demonstrated that the Stoics could only “refuse, avert an autonomy” of grammar or other linguistic disciplines, precisely “for philosophical reasons” connected to their project of a “systematic philosophy” (1997: 139).
6 πάθος δὲ πλεονάζουσα ὁρμὴ ἢ ὑπερτείνουσα τὰ κατὰ τὸν λόγον μέτρα.
7 ἄλογόν τε καὶ παρὰ φύσιν κίνησιν ψυχῆς.
8 Agamben clarifies that “in the fragments of the Stoics . . . we nowhere find such an explicit assertion; however, this is the only one that does not contradict the premises of their theory of the passions” (2010: 85).
9 ἦν δὲ καὶ ἄλλη τις παρὰ τούτοις διάστασις, καθ' ἣν οἱ μὲν περὶ τῷ σημαινομένῳ τὸ ἀληθές τε καὶ ψεῦδος ὑπεστήσαντο, οἱ δὲ περὶ τῇ φωνῇ, οἱ δὲ περὶ τῇ κινήσει τῆς διανοίας. καὶ δὴ τῆς μὲν πρώτης δόξης προεστήκασιν οἱ ἀπὸ τῆς Στοὰς, τρία φάμενοι συζυγεῖν ἀλλήλοις, τό τε σημαινόμενον καὶ τὸ σημαῖνον καὶ τὸ τυγχάνον, ὧν σημαῖνον μὲν εἶναι τὴν φωνήν, οἷον τὴν 'Δίων', σημαινόμενον δὲ αὐτὸ τὸ πρᾶγμα τὸ ὑπ' αὐτῆς δηλούμενον καὶ οὗ ἡμεῖς μὲν ἀντιλαμβανόμεθα τῇ ἡμετέρᾳ παρυφισταμένου διανοίᾳ, οἱ δὲ βάρβαροι οὐκ ἐπαῖουσι καίπερ τῆς φωνῆς ἀκούοντες, τυγχάνον δὲ τὸ ἐκτὸς ὑποκείμενον, ὥσπερ αὐτὸς ὁ Δίων. τούτων δὲ δύο μὲν εἶναι σώματα, καθάπερ τὴν φωνὴν καὶ τὸ τυγχάνον, ἓv δὲ ἀσώματον, ὥσπερ τὸ σημαινόμενον πρᾶγμα, καὶ λεκτόν, ὅπερ ἀληθές τε γίνεται ἢ ψεῦδος.
10 For instance, Michel Foucault attributes to the Stoics the very invention of this triadic system of signs (1966: 57).
11 Gourinat objects, citing Priscian, that “in antiquity an isolated proper noun, employed as an answer, designates a complete proposition” (2000: 112).
12 [My translation of this phrase follows that of Chiesa in Agamben 2018. (KL).]
13 Note here that, in order to assert the ontological nature of the lekton, Agamben does not need the Heideggerian thesis, which he knows very well, about the ontological rather than logical value of the Greek verb legein. Cf. Hadot (1980).
14 The principal manuscripts for Burnet and Souilhé are Parisinus graecus 1807 and Vaticanus graecus 1. The reading δι’ὁ was only restored in 1923 by Andreae (Agamben 2010: 23).
15 On the anaphoric use of αὐτὸς, see Chantraine (1968: 143), Ildefonse (2014: 20).
16 On the status of anaphoric pronouns, see Benveniste (1966: 251–7), Jakobson (1971: 130–47).
17 Long notes that a reading of Plato was, for Zeno at least, obligatory; but Zeno’s attitude, probably following Antisthenes, was one of suspicion, especially with regard to the separate nature of Ideas and the immateriality of the soul (1996: 18–19).
18 Daniel Heller-Roazen, in the “Editor’s Introduction” to Agamben 1999, has already proposed a reading of the Stoic lekton in this direction. It is the lekton, he writes, which expresses in antiquity “the fact that something appears in language and that language itself, in this appearance, takes place.” He also effectively demonstrates the enduring nature of this element in the history of philosophy, notably in the anonymous thirteenth-century Ars Burana, where the enuntiabile is said to be extrapredicamentale, in other words beyond Aristotle’s categories and categoriality itself (predicamentum enuntiabile); then in the works of Peter Abelard and Gregory of Rimini (fourteenth century), and through the theorization of “objective” contents of thought by Alexius Meinong, which subsist (bestehen) without however existing (existieren) (Heller-Roazen 1999: 10–12).
19 Hülser says, “von Hause aus ist die Lekton-Theorie kein Thema der Physik, sondern eins der Dialektik” (1987: 833).
20 Think, for example, of the text where Sextus says that what is true (to alēthes) is a proposition (axiōma), and the proposition is immediately defined as a lekton (LS 33P; cf. SE M. 7.38 = SVF II.132, Isnardi Parente 2005: 183).
21 Agamben could perhaps have added, in addition to these passages, what Plutarch says. In speaking of the γένος τῶν λεκτῶν, he links it to what is οὐσίαν τῷ λόγῳ παρέχον (Plut. Adv. col. 1119f).
22 Chrysippus had clarified (contrary to Zeno, who left it undetermined) that the effect—that “of which the cause is a cause”—is not a body (SVF I.89). He also secured some reality for the predicate, even though it is not an attribute of a body (συμβεβηκός), by speaking of entities that “subsist” (SVF II.509; see Sedley 1999: 398).
Bibliography
Agamben, G. (1982), Il linguaggio e la morte. Un seminario sul luogo della negatività, Torino: Einaudi.
Agamben, G. (2001), Infanzia e storia. Distruzione dell’esperienza e origine della storia, Nuova edizione accresciuta, Torino: Einaudi.
Agamben, G. (2010), La potenza del pensiero. Saggi e conferenze, Vicenza: Neri Pozza.
Agamben, G. (2015), L’avventura, Roma: Nottetempo.
Agamben, G. (2016), Che cos’è la filosofia?, Macerata: Quodlibet.
Agamben, G. (2018) What Is Philosophy? trans. Lorenzo Chiesa, Stanford: Palo Alto.
Alessandrelli, M. (2013), Il problema del lekton nello Stoicismo antico. Origine e statuto di una nozione controversa, Firenze: Leo S. Olschki Editore.
Andreae, W. (1923), “Die philosophischen Probleme in den platonischen Briefen,” Philologus 78: 34–87.
Benjamin, W. (1995), Brief an Martin Buber vom 17. Juli 1916, in Briefe, 6 Bd., Frankfurt am Main: Suhrkamp.
Benveniste, E. (1966), Problèmes de linguistique générale, vol. I, Paris: Gallimard.
Bréhier, E. (1962), La théorie des incorporels dans l’ancien stoïcisme, 3rd ed., Paris: Picard.
Brunschwig, J. (1995), “Remarques sur la théorie stoïcienne du nom propre,” Histoire Epistémologie Langage 6 (1): 3–19.
Chantraine, P. (1968), Dictionnaire étymologique de la langue grecque. Histoire des mots I, Paris: Klincksieck.
Coseriu, E. (2003), Geschichte der Sprachphilosophie von der Antike bis zum Gegenwart 1: Von den Anfängen bis Rousseau, Stuttgart: A. Franke.
Foucault, M. (1966), Les mots et les choses. Une archéologie des sciences humaines, Paris: Gallimard.
Frede, M. (1994), “The Stoic Notion of a lekton,” in S. Everson (ed.), Companions to Ancient Thought, 3: Philosophy of Language, 109–128, Cambridge: Cambridge University Press.
Frege, G. (1892), “Über Sinn und Bedeutung,” Zeitschrift für Philosophie und philosophische Kritik 100: 22–50.
Georgiades, T. (1958), Musik und Rhythmus bei den Griechen, Hamburg: Rowohlt.
Gourinat, J.-B. (2000), La dialectique des stoïciens, Paris: Vrin.
Hadot, P. (1980), “Sur divers sens du mot ‘pragma’ dans la tradition philosophique grecque,” in P. Aubenque (ed.), Concepts et catégories dans la pensée antique, 309–319, Paris: Vrin.
Heidegger, M. (1962), Being and Time, trans. John Macquarrie and Edward Robinson , Harper and Row: New York.
Heller-Roazen, D. (1999), “Editors’ Introduction,” in G. Agamben , Potentialities, 1–23, Stanford: Stanford University Press.
Hülser, K. (1987), Die Fragmente Zur Dialektik der Stoiker. Neue Sammlung der Texte Mit Deutscher Übersetzung Und Kommentaren von Karlheinz Hülser, 4 vols., Stuttgart: Frommann-Holzboog.
Ildefonse, F. (1997), La naissance de la grammaire dans l’Antiquité grecque, Paris: Vrin.
Ildefonse, F. (2014), “Que nous apprend le αὐτός ilidiaque?,” in D. Doucet and I. Koch (eds.), Autos, Idipsum. Aspects de l’identité d’Homère à Augustin, 19–37, Aix-en-Provence: Presses universitaires de Provence.
Imbert, C. (1993), Phénoménologie et langues formulaires, Paris: PUF.
Isnardi Parente, M. (2005), “La notion d’incorporel chez les stoïciens,” in G. R. Dherbey and J.-B. Gourinat (eds.), Les stoïciens, 175–185, Paris: Vrin.
Jakobson, R. (1971), Selected Writings II, The Hague: Mouton.
Joly, E. (1986), Philosophie du langage et grammaire dans l’Antiquité, Bruxelles-Grenoble: Ousia-Université de Sciences Sociales de Grenoble.
Koller, H. (1955), “Stoicheion,” Glotta 34: 161–174.
Lohmann, J. (1970), Musiké und Logos. Aufsätze zur griechischen Philosophie und Musiktheorie, hg. v. A. Giannaras, Stuttgart: Musikwissenschaftl. Verlags-Gesellschaft.
Long, A. A. (1971), “Language and thought in Stoicism,” in A. A. Long (ed.), Problems in Stoicism, 75–113, London: The Athlone Press.
Long, A. A. (1996), “Stoic Psychology and the Elucidation of Language,” in G. Manetti (ed.), Knowledge through Signs: Ancient Semiotic Theories and Practices, 109–131, Brussels: Brepols.
Pohlenz, M. (1939), “Die Begründung der abendländischen Sprachlehre durch die Stoa,” Nachrichten der Göttinger Gesellschaft 6: 151–198.
Schubert, A. (1994), Untersuchungen zur stoischen Bedeutungslehre, Göttingen: Vandenhoeck & Ruprecht.
Sedley, D. (1999), “Hellenistic Physics and Metaphysics,” in K. Algra , J. Barnes , J. Mansfeld , and M. Schofield (eds.), The Cambridge History of Hellenistic Philosophy, 355–411, Cambridge: Cambridge University Press.