Navigation – Plan du site
Dire et vouloir dire dans les arts du langage anciens et tardo-antiques

What is wrong with lekta? Ancient critics of Stoic logic and language

Qu’est-ce qui ne va pas avec les lekta ?
Ada Bronowski

Résumés

On examinera la théorie stoïcienne des lekta en suivant les critiques formulées contre elle provenant de trois perspectives différentes : celle des Péripatéticiens, de Sextus Empiricus et celle formulée au sein même de l’école par Sénèque. Ces critiques se concentrent sur des questions relatives à une théorie du langage, mais une lecture minutieuse révèle que le cœur du problème réside dans un rejet profond de l’ontologie stoïcienne, constituée en partie, par les lekta. Les réactions des critiques tendent toutes à confirmer leurs propres positions, soit en rejetant la possibilité de présence réelle des lekta, soit en considérant cette présence comme négligeable. Ces confirmations de la validité de ses propres convictions se font par le biais de déformations ou de raccourcis faits à la théorie stoïcienne. En dévoilant ces mécanismes d’escamotage, c’est pourtant bien l’importance de la théorie des lekta qui est percée à jour, et les sérieux défis qu’elle pose aux théories rivales, en remettant en question le contrôle supposé de l’homme sur ses pratiques linguistiques, et en complexifiant la relation de la logique à l’éthique.

Haut de page

Texte intégral

Introduction

1A leitmotiv in the post-Aristotelian debates about the production of meaningful speech in antiquity turns on most, if not all the main schools of thought attacking the Stoics for their insertion of the items they call ‘lekta’ into the context of that debate. There are morphological echoes in the word ‘lekton’ which tie the term to its roots in the Greek verb ‘legein’, (‘to say’), and which at first blush, rightfully place lekta, whatever they may be, in the context of a discussion about language for the modern reader, as indeed for the ancient reader. In speaking of ‘lekta’, the Stoics – we, then as now, may legitimately expect – are talking about something related to language, something to do with ‘what is said’ or ‘what can be said’, which are the literal translations of the Greek term ‘lekton’, the substantivized form of the verbal adjective derived from ‘legein’. For if the Stoic lekta were not involved in the production of meaningful speech, why would the Stoics have called these things ‘lekta’?

2The inclusion of lekta in the ancient discussions on theories of language by rival schools, tends, what is more, to confirm that lekta are indeed part of a Stoic theory of language. And yet, the attacks are unanimous in this: that the Stoic lekta are additional items, which do not contribute anything to a theory of language but merely create difficulties. The general tenor of the attacks can be summed up in the damning view that the Stoics do not in fact, have a theory of language, since whatever it is that they are proposing with their introduction of lekta does not stand scrutiny. But, whilst there is a generalised targeting of the Stoic lekton, the nature of the complaints varies from one school to another.

3For one school of thought, it is redundant to say that words express lekta, because, so goes the theory, words already express thoughts; thus, either the term ‘lekton’ is just another word for a thought, or the Stoic lekta simply have no place in a theory of language. Another branch of critique insists that a sentence (not words) cannot possibly express one lekton since by the time we have uttered one word of the sentence, the next is no yet in existence, and the previous word is out of existence; thus, either there is no such thing as a single lekton which a whole sentence should somehow express, or a lekton cannot correspond to anything more than the signification of separate words. Yet another branch affirms that for lekta to be part of an analysis of the production of meaningful speech, there first needs to be a proof of the existence of lekta, but there is no such proof, since the Stoics themselves say that lekta subsist or obtain but precisely do not exist.

4These critical perspectives represent a sample mosaic of the different views found across the board of the main schools of thought in antiquity, from the Platonists, the Peripatetics, and the Epicureans. But, as we shall see, the questioning of the presence of lekta comes also from within the Stoic school. The effect of these variations is twofold: on the one hand, the different critiques bring to light the specificity of the view advocated by the school of thought the critic represents. On the other hand, the accumulation of different complaints, reflecting different concerns from one author to the next, leaves the Stoic account ever more stretched at the seams. For there is no single feature which focuses everyone’s attention. Rather, there is always something different, which is wrong for different people. That a doctrine strikes different chords of dissonance with different people is in itself, to be expected, but what appears peculiar in the case of the Stoic lekta is that, when assembled, the critiques appear in fact incompatible one with another. It is not clear that there is any agreement, between its detractors, of what a lekton is in the first place, that everyone finds so problematic. Are the Stoic lekta attacked for having a correspondence to words, or for having a correspondence to a sentence? Are the Stoics attacked for having granted existence to lekta or for having considered them to be mental items? It would seem that the lekta are to be found, for different authors, on one or the other side of these alternatives. But whichever the side, the critics always find fault with lekta. There is enough, in sum, to indicate a peculiar degree of confusion with regard to what critics generally understand of the nature of a lekton.

5However, confusion does not mean ignorance. Rather, what appears at first as confusion reflects on the one hand, the debates ignited by the introduction of lekta. For example, whether a sentence expresses something different from the addition of the singular significations of its component words, and what such a whole greater than its parts would actually come down to. Relatedly, a question which becomes urgent through the discussion of lekta is what ontological status the signification of a word has and whether that status is different from what a sentence gets said as a whole. On the other hand, the confusion arising from the variety of critiques reflects the original, multi-faceted complexity characterising the Stoic lekton. The lekton lends itself to a variety of attacks because it is, or seems to be, proteiform in its original characterisation. We shall see that this is not correct. The lekton plays roles in different accounts within Stoic doctrine: it appears in logic through its status as an incorporeal, distinct from corporeal items, and it appears in language as the incorporeal thing which gets said when we speak (to take the two most relevant roles for our present inquiry). But for the Stoics, these are not separable roles, as we shall see. At the heart of all the critiques is the disregard for the inseparability of the roles of lekta in logic, ontology and language. What makes lekta seem proteiform through the lens of the critics comes down to a rejection, or misrepresentation of the original Stoic theory. The various critics pick the Stoic theory apart, whereby each critic focuses on one aspect, disregarding the rest. Whether it is legitimate to do so, that is, to focus on one claim, by disregarding the others, is a central question, but it is, to all intents and purposes, a different question from that of assuming ancient commentators were simply ignorant of the Stoic views concerning lekta.

6Given that it is far more likely that any ancient critic is more familiar with Stoic doctrine than any modern commentator, we should be very wary of decreeing that an ancient critic is incorrect in his characterisation of the Stoics, and in particular in the role and nature of lekta imputed by a critic to the Stoics. Rather, we should proceed on the working hypothesis that everyone is correct – though some more than others. That is to say that, however contradictory or confusing, or reductive in appearance, the collection of accounts of lekta may be, they each add a small piece to the puzzle. As one goes through the different critiques, one realises that each unfolds from commitments which characterise the specific school of thought it belongs to. The explicit rejection, and often derision of the Stoic lekta frames, for each school of thought, its own perspective. Through the contrast brought out with lekta, a commentator highlights and confirms his own view or the view of his school which he is defending against the Stoics. The problem each school identifies with lekta turns out then, to be a different problem depending on who is problematising the contrast. It is, each time, a problem for the school in question concerning a specific point which is important to confirm for its own internal line of thought. The problem thus exposed is then revealing of two aspects: on the one hand, it reveals and confirms the kind of commitments a school of thought is bound to, such that it cannot admit of lekta in addition; on the other hand, it reveals that admitting lekta in addition means that one has to also make distinctive commitments which allow for and indeed justify the presence and need for these lekta. In discovering what these commitments are, we shall be able to see what is, in effect, so wrong with lekta from the point of view of the critics, and what is so right about lekta for the Stoics.

  • 1 The aim of the present inquiry is thus to uncover and recover Stoic doctrine about lekta from purel (...)

7The question of what is wrong with lekta is tightly connected with the question of what lekta are. The texts that we have at our disposal, the majority of which are critical of the presence of Stoic lekta, force us to go about an investigation into lekta and their role in language by following a path of negativity first. The aim however, is to show, by the end of this inquiry, that what first appears to be consistently negative accounts reflect, in fact, the layered complexity which is at the heart of the Stoic notion of a lekton. We shall focus here, on three main directions of critique, which shall correspond to the main parts of our inquiry: the argument from redundancy articulated by the Peripatetic critique, which shall lead us to a first reflection on some misrepresentations of the Stoic lekta, before examining secondly, the argument from indistinguishability put forward by Sextus Empiricus, and finally the argument from triviality, whose sharpest upholder is an in-house Stoic, namely Seneca. Each perspective will enable us to move closer, through concentric circles, to a positive account of the Stoic lekton1.

The Argument from Redundancy

8An emblematic line of attack is found in the opening pages of the Aristotelian 5th century commentator Ammonius in his commentary on Aristotle’s De Interpretatione (Int.). Ammonius embarks on the commentary of the first lines of the Aristotelian text (Int.1, 16a3-9), by giving a first summative account of the Peripatetic view, which is immediately put in the perspective of a frontal attack on the Stoics:

  • 2 Amm. in de Int. 17. 24-28: “πρότερον ἡμᾶς ὁ Ἀριστοτέλης διδάσκει διὰ τούτων, τίνα ἐστὶ τὰ προηγουμέ (...)

“Aristotle first teaches us, in these lines, what is principally and directly signified by them [names and verbs], namely thoughts, and that these thoughts are the intermediaries, through which things are signified. And there is nothing else to consider in addition to these, which would come between the thought and the thing, but that is precisely what the Stoics thought fit to posit and call a lekton.”2

  • 3 Ammonius is well aware of this: in the first pages of his in de Int. (5.29-7.14), he recalls how th (...)

9Ammonius appeals to the authority of Aristotle, and appears to be reporting, like a good student, the conclusions drawn from the teachings of his master. He provides us accordingly, with the schema of a linguistic triad composed of the three necessary and sufficient elements involved in the production of meaningful speech as they are put forward by Aristotle: words (here, as discussed in the previous lines by Ammonius, specifically names and verbs), thoughts and things. But Ammonius’ report of Aristotle’s De Interpretatione might appear overhasty in its schematisation. After all, Aristotle speaks neither of thoughts tout court, but only eventually of the ‘thought in the soul’ (en tē psuchē noēma, Int. 16a9-10), after having first thrown into the discussion talk of the well-known though mysterious ‘affections in the soul’ (en tē psychē pathēmata, Int. 16a3) which then become ‘affections of the soul’ (pathēmata tēs psuchēs, Int. 16a7)3; nor does Aristotle speak, in the lines Ammonius refers to, of ‘what is signified by words’, but rather, Aristotle speaks of words being ‘symbols’ (‘symbola’, Int.16a4), and then ‘signs’ (‘sēmeia’, Int. 16a6) of the affections in the soul. Crucially, Aristotle says of the latter that they are ‘likenesses’ (‘homoiōmata’, Int. 16a7) of things. In Ammonius’ reconstruction, the difficult notion of a likeness is presented in terms of a mediation: thoughts (noēmata) mediate between words and things such that words signify things through thoughts. Words, that is, directly signify thoughts, and indirectly things. Thoughts are the go-between. As a stark contrast to this view, Ammonius presents the Stoics as having inserted lekta as intermediaries between things and thoughts.

  • 4 See Simplicius, in Cat. 41.28-9, who appears to be quoting from Iamblichus, who presumably draws hi (...)
  • 5 Dexippus, in Cat. 9.24-5: “προηγουμενῶς μὲν τὰ νοήματα, κατὰ δεύτερον δὲ λόγον καὶ τὰ πράγματα σημα (...)

10The account Ammonius gives of the basic Aristotelian picture emphasises more explicitly certain aspects which, in the original Aristotelian account, do not seem fixed in such precise terms. Ammonius emphasises the directness of reference of words to their corresponding thoughts, the “principle and direct” relation of signification, and the fluidity with which thoughts lead to things. This fluid correspondence from thought to thing is not a main point of emphasis in the Aristotelian text, but it becomes an important point of interpretation in the Peripatetic tradition. That names signify things through thoughts is a formula which is echoed throughout the corpus of Peripatetic commentaries – especially on Aristotle’s Categories for which a tradition develops of beginning the commentary by treating a preliminary set of questions concerning what the categories are: signifying words, things signified by words, or thoughts. Boethus is cited, who himself refers the view back to “the ancients”, stating that “the only things which are uttered and signified are thoughts”4. This view which thus focuses on the priority of the relation of names with thoughts, is the basis for the further view that via thoughts, as the primary relation, things are, derivatively, also signified. Thus, Dexippus, in Cat. 9.24-5, writes that: “primarily it is thoughts which are signified, and at a secondary level, also things”5. We find here almost word for word, Ammonius’ schematic presentation of the linguistic triad, with the same insistence on the “primary” relation of words to thoughts. This sweeping bird’s eye view of the question helps put into perspective Ammonius’ presentation. For Ammonius is thus at once rehearsing what is, by his time, consecrated Peripatetic lore, and at the same time defending it again, by highlighting its most important feature, which is threatened by the Stoics’ introduction of lekta.

11The fluid correspondence from thought to thing plays a key role in the contrast Ammonius builds with the Stoics, anticipating, through it, the rejection of the Stoic view. For rather than questioning the centrality of the role of an individual’s mind and thoughts in the linguistic process, Ammonius focuses on the solidity of the correspondences that govern the Aristotelian schema for it. There are no chinks, and no room to insert something additional in between words, thoughts, and things. The Stoics are presented as attempting to break apart this compact triad. The mention of the Stoics serves to reinforce the validity of the proposed Aristotelian model. And this is so, much at the expense of the Stoic theory of lekta. For Ammonius, the Stoics intercalate a superfluous intermediary into the solid Aristotelian schema. Ammonius does not, that is, treat the introduction of lekta as offering a completely different schema (which is, in effect, what the Stoics do).

  • 6 Ammonius, in de Int. 19.1-3, appeals directly to the Aristotelian text, saying that when Aristotle (...)
  • 7 Ammonius, in de Int. 20.7: “τὸ δέ σύμβολον ἤτοι σημεῖον τὸ ὅλον ἐφ' ἡμῖν ἔχει.”, ‘a symbol or a sig (...)
  • 8 Thus, Ammonius explains that Aristotle identifies assertoric sentences as the subject of ‘interpret (...)

12Ammonius’ presentation is revealing as to the attitude taken up by a tradition of rejection of the notion of a lekton: lekta do not fit in the general schema of signifying speech, established on a certain interpretation of a basic Aristotelian model. The core of that interpretative tradition, crystallised in the Boethan formula mentioned above, is that it roots language to thoughts, consolidating thereby the relation between a word and a thought. Ammonius, in the following pages of his commentary, magnifies the view by drawing first on a distinction between natural and conventional, according to which words are by convention6; it then follows that the relation of words to thoughts is analysed as “up to us”. It is up to us, that is, in our power, to establish the rules, consecrating such and such a sound as the word apt to signify such and such a thought7. That the direct relation of words to thought is thus further glossed by Ammonius in terms of its being up to us, underlines the central thesis of the Peripatetic account of language, namely that language is the crowning capacity of human beings. The capacity, that is, to set words to the thoughts which things, naturally (and thus, not under a person’s control) bring about in us. It is this control over the relation of words to thoughts, which guarantees its directness and fulfils our capacity for knowledge8. From this perspective, the redundant intercalation of lekta into the schema, is, at the very least, taken to overburden the word-to-thought correspondence, with, now, a word-to-thought-to-lekton correlation.

13We are still not clear at all about what a lekton is; more importantly, we are still not clear about whether Ammonius has a clear notion of what a lekton is. But that Ammonius seems to think the Stoics propose to intercalate the lekton between words and thoughts is a first indication of the nature of the threat the lekta represent to the Peripatetic account. Their presence jeopardises the direct relation between words and things, that is to say, given the way Ammonius understands that indissoluble relation, that the presence of lekta jeopardises the control a community of speakers is considered to have over the relation of words to their thoughts.

14The tradition Ammonius represents is committed to the central role of thought such that it is the association of thoughts which is then reflected in an association of words. The association of thoughts itself, refers to a relation between things, e.g. Socrates and illness. Socrates and illness are ‘things’, that is, they are extra-mental and extra-linguistic, but they determine our thoughts of them. This is what Ammonius describes as a relation – that between things to thoughts – which is not ‘up to us’, since we cannot control it (see fn.7). The linguistic process serves to signify these things, as they are represented by our thoughts. It does so thanks to the logical tool of predication, by which we can say something about something. Crucially, however, predication has, itself, no ontological correspondence outside of language. It is, to continue to speak like Ammonius, in our control. For it merely is the way we express the association our thoughts make with reference to an association of things (e.g. the association of a substance with a property). Language expresses thoughts. It has a direct correspondence with thoughts. Only indirectly is it associated with external things, in virtue of the relation of thoughts to things.

15By framing lekta as part of the linguistic process, Ammonius casts lekta as doubling up on the role thoughts play with respect to words. From the Peripatetic perspective, if there are lekta which are something distinct from our thoughts, then the linguistic triad explodes. For the insertion, or intrusion of lekta, threatens the transitivity from words to thoughts to things, which is the cornerstone of the Peripatetic account of language. To question the directness of that relation is to question the very solidity of the account of the production of meaningful speech, as also to question the privileged creative place of human beings in that production. But language is a characteristic of human beings, therefore, the Stoics must be wrong about adding lekta to the picture. The Stoics’ lekta must come down to nothing other than items redundant on thoughts, or redundant on words.

16This is, in fact, the most expedient means, in the Peripatetic corpus of texts, to both acknowledge and disparage the threat to the theory presented by the lekta. The lekta get assimilated to already identified items, only under another name. A series of texts thus draw out a grid of terminological comparisons, in which the Stoics appear as merely terminological innovators, and in which lekta are presented sometimes as the equivalents of words, (phonai), with a new name:

  • 9 Themistius, in An.pr. 92.1-3: “οἱ δὲ Στωϊκοὶ καινοτέραν βαδίζοντες τὰ μὲν πράγματα τυγχάνοντα ὠνόμα (...)

“The Stoics, who tend to make innovations, call things (pragmata), ‘tunchanonta’ [that which obtains], because we want to obtain things, whereas they call thoughts (noēmata), ‘ekphorika’ ‘enunciated’, because we utter out loud what we think in our minds, and they call words, ‘lekta’.”9

17Or as the equivalents of thoughts:

  • 10 Simplicius in Cat. 10.2-4: “σαφῶς εἶπεν ὁ Ἀριστοτέλης, τὰ δὲ λεγόμενα καὶ λεκτὰ τὰ νοήματά ἐστιν, ὡ (...)

“Aristotle says this very clearly, thoughts (noēmata) are the things said (ta legomena) and lekta (the things which can be said), as was also the view of the Stoics”10.

  • 11 E.g.: Cicero Fin. 3.5; 4.7; Galen, Inst. Log. 4.6, or Plutarch Comm. Not. 1068D, Plat. Quaest. 1009 (...)
  • 12 The same strategy is illustrated by Themistius’ semantic treatment of the other Stoic notions he me (...)

18Since the Aristotelians cannot admit of a third element outside of the mind, and yet distinct from the subject, they treat lekta as redundant on one or the other, oscillating between identifying lekta with words or with thoughts. The appeal to the Stoics’ taste for innovative vocabulary is a hackneyed recourse which is often made by the critics of the Stoics11. But to consider that the Stoic account must in some way be reconciled with the Aristotelian account, creates the difficulties and misunderstandings, which are betrayed by the very insistence within the tradition of the Aristotelian commentators, to box in the Stoic account. In both these two passages, what is striking is the literality with which the Peripatetic commentators approach the Stoic terms. They cite the peculiar terminology all the better to give literal explanations which belittle the (aspirations for) conceptual innovation. Thus, for the term ‘lekton’ in particular, both Themistius and Simplicius consider it purely from its morphological origins, relating to ‘saying’ (legein), rather than treat the term as designating a full-blown concept in itself, that of the thing there to be said, independently of the actual successful production of meaningful speech. This strategy of by-passing the status of concept-term which the Stoics give to the lekton, by re-interpreting the use of the word ‘lekton’ as a mere descriptive qualifier is a subtle means of critique; we shall see it at work again in Seneca, when it comes to translating the term into Latin, (infra §51-57)12.

  • 13 See Andreas Graeser (1978), “The Stoic Theory of Meaning”, in John Rist (ed.), The Stoics, Berkeley (...)

19What lies behind the Peripatetics’ appraisal of the tension between the Aristotelian and Stoic accounts are two very different ways of thinking about language and rationality. For what the Stoics propose is to re-model the relation of the mind to the world13. Thus, what is at stake is a different account of ontology. The difficulties the Peripatetics encounter in deciding whether lekta are linguistic items or mental items arise from the impossibility for the Peripatetics of accepting first and foremost into ontology, and only following from that, into their linguistic triad, additional items - additional to what, on this tradition, we have been calling ‘things’ (substances and their properties), thoughts and the language a subject controls, (in the way we have been speaking of this control as the direct relation of words to thoughts which a community of speakers regiments). The Stoic lekta prove to be incompatible with the Peripatetic model, because the lekta the Stoics introduce are something else, distinct from language and the mind. Thus, the more clearly we see where the difficulties lie in the critical discussions of lekta for the Peripatetics, the more certain we can be of the direction to take for a more faithful understanding of lekta.

Learning from the Peripatetics’ Mistakes

20If the lekta are something distinct from our thoughts, then there are different things in reality apart from substances, properties, words and thoughts, which are lekta. They are the things which we get said when we speak. But, being distinct from us, from our use of language and from our thoughts, they must be there to be said, whether we actually speak, or not. Similarly, they are also the things which we think, when we have thoughts. But since they are distinct from our thoughts, they are there to be thought, whether anyone ever thinks them or not. But what are these lekta that we can get said, that we can think of, but which are there, regardless of what we, humans, do, say, or think? It is not the words that we use which can determine the presence of a lekton, or what a lekton is, since, as the Peripatetics note, words represent specific and speaker-subjective regimentations of sounds. But if the lekta are independent of us (independent of our minds, and independent of our linguistic practices), then lekta cannot have a correspondence to words, nor a correspondence determined by the associations of thoughts that we might have. What is it that we get said when we speak?

21The lekton that gets said or that gets expressed when we speak, that is, when we utter words, cannot be determined by the uttered words, and yet it is expressed by our uttering them. The only solution must be that the lekton does not correspond to words, but to what the string of uttered words ultimately gets said. There is, in other words, a distinction between the utterance of words (which is dependent on us, on the rules of a linguistic community) and what is said through this utterance, what is communicated that is, which is distinct enough from the specific words used such as for instance, to be translatable into other languages. Thus ‘It is sunny’ and ‘Il fait beau’ may get one and the same thing said, though juxtaposing the significations of the separate words would not have led us to that conclusion. The words we utter signify certain things, but they are not the kind of things or thing which we get said by uttering words.

22This is precisely the distinction the Stoics make between to utter and to say on the basis of a distinction between the objects of saying and uttering:

  • 14 Diogenes Laertius, Lives of the Eminent Philosophers, Dorandi, T. (ed.) (2013), Cambridge [D.L.], 7 (...)

“to say is different from to utter, for it is sounds which are uttered, but things which are said, and which actually are lekta14.

  • 15 It is in the middle form that the verb is indeed commonly used in the sense of ‘to utter’: e.g. at (...)
  • 16 See John D. Denniston (1950), The Greek Particles, Oxford, OUP, on ‘δὴ’.

23The Stoics distinguish between two very different actions, uttering and saying on the basis of the objects of these actions. It is clear from the wording, that the emphasis on the object is key. For the distinction is first set out between ‘to say’ (‘legein’) in the active form and ‘to utter’ (‘propheresthai’), in the middle form15. But when the explanation for the distinction is given, the same verbs appear in passive constructions. Reverting thus to a passive formulation enables the objects of uttering and saying to become the grammatical subjects of ‘to be uttered’ and ‘to be said’: respectively the sounds (phonai) are the subject of ‘are uttered’ and the things ‘ta pragmata’ are the subject of ‘are said’. The grammatical re-adjustments culminate with the final relative clause, in which the true nature of these ‘things’, which are the object of saying, is revealed to be lekta. The use thus of ‘things’ (pragmata) in this context, is not to be confused with the ‘things’ the Peripatetics refer to, which, whatever else may be said about them, are distinct in this, that they are indirectly related to speech, via their relation to thoughts. Here, the ‘things’ which the Stoics are zeroing in on, are more properly called ‘lekta’. Many precautions are taken to introduce lekta as the objects of saying. First, there is the use of ‘tunchanein’, which has a more emphatic force, compared with just ‘einai’, as ‘manifestly is’, compared to just ‘is’, when it comes to indicating an identification, as here between pragmata and lekta. The emphatic intention is further corroborated with the adverbial sequence “dē kai” (‘and actually’)16.

  • 17 D.L. 7.57 identifies this sub-class of signifying sounds in the line immediately preceding the quot (...)
  • 18 D.L. 7.55: ‘phonē is the air struck by an impulse’.
  • 19 D.L. 7.56; Ps-Plut. Plac. 902F-903A: sound, phonē is said to act, poiei, in that it moves from the (...)
  • 20 S.E. M. 10.218 : “τῶν δὲ ἀσωμάτων τέσσαρα εἴδη καταριθμοῦνται ὡς λεκτὸν καὶ κενὸν καὶ τόπον καὶ χρό (...)
  • 21 For the Stoics, a body is first and foremost determined by its capacity to act or be acted upon: Di (...)

24What we say, or what we get said, is a lekton. What we utter, are sounds. And sounds, of which a sub-class are ‘signifying sounds’, that is articulate words17, are a different kind of item from lekta. We know from the Stoics, that sounds are corporeal: uttering produces corporeal items, the phonai, which, following the Stoic definition, are the air affected in a certain way, having ‘been struck’18. The speaker is the cause of this striking of the air, as it is the speaker who has struck the air in this manner19. On the other hand, in saying, that is, getting a lekton said, the speaker is not the cause of lekta. Indeed, saying, on the strict sense, does not bring about or produce anything, in the sense in which uttering does. For lekta are, in every way opposed to the corporeal sounds uttered, and as such are incorporeal items. They figure as one of the four incorporeals alongside void, place and time20. As incorporeals, the first determinations of lekta are that they lack what bodies have, namely the capacity for agency, or that of being passive receptors of an action21. Thus, as what gets said when we say something, and not merely utter something, the reality of lekta complexifies our understanding of ‘to say’: saying has nothing to do with the utterance of words, and in fact, nothing to do with uttering. For what we get said is an incorporeal thing. Thus, ‘saying’ for the Stoics belongs to a different sphere from the actions, which involve bodies, and which imply the activity and passivity of bodies. To say and to utter are, crucially, distinct because what we utter is not what is said. We utter a series of sounds, but what is said is one incorporeal thing. It is not, therefore, the surface grammar of utterances, which determines what is said. Rather, it follows, it is what is said, which eventually determines the grammar. Given the lekton’s status as an incorporeal, it is there, present in reality, independently of the corporeal sentences we can utter, or not. But it also follows that, in order to get said what there is to say, there is a strict requirement for what a correct utterance is, one which satisfactorily gets a lekton said.

25The Stoics claim that when we utter corporeal sounds, we thereby, in certain cases (the successful cases in which something gets said), express an item which is not corporeal, the lekton. More precisely: words are uttered, but what gets said is not what the words are significant of (because it is not words which get said in the relevant sense of ‘to say’), but rather what, in effect, a certain arrangement of words, forming thus a complete sentence, actually gets said. There are clear difficulties with such a claim. How can saying be distinguished from uttering, if indeed, saying is not an action like uttering is? How do corporeal beings, who are the speakers, manage to express an incorporeal, without in fact having any causal interaction with it? How can separate corporeal words end up expressing, when arranged in a certain way, one incorporeal lekton? The Peripatetics give one kind of answer to these questions, by rejecting from the start the possibility of an additional item in ontology. They therefore react to the Stoic account by collapsing it onto their own. A different approach is developed by a relentlessly contrarian critic, Sextus Empiricus (S.E.), who asks these questions, though not with the aim to prove the superiority of his own rival theory, for as a practitioner of scepticism, his aim is precisely to show that there is no theory to give. In his tenacious analysis of the Stoic position, he confirms its originality and the ontological commitments it relies on.

The Argument from Indistinguishability

  • 22 S.E. M. 8.81: “τούτου δὲ τοῦ πράγματος τὰ μέρη μὴ συνυπάρχειν, ὥστε μηδ' αὐτὸ ὑπαρκτὸν εἶναι.”, ‘th (...)

26The first line of Homer’s Iliad, ‘Sing, Goddess, the wrath of Achilles son of Peleus’, on the Stoic view, gets one thing said, one lekton, though there are, in English, nine words involved, and five in Greek. Sextus Empiricus, who focuses on this example at M.8.80-1, proceeds to show that the Stoic claim cannot be sustained, for the reason that each word, which composes the sentence is uttered one at a time: in uttering one word, the next is not yet uttered, and the preceding already uttered, so that, not being able to be uttered all at once, not being able to “sunyparchein”, be present together, nothing can be said through them22.

27S.E. shows that saying, as the Stoics understand it, is incompatible with uttering, because we do not actually say anything, by uttering sounds. Therefore, S.E. concludes, there are no such things as lekta, since to be said, a lekton must be uttered. But it cannot be uttered, therefore nothing is said, or there is no such thing as a lekton. S.E. appeals to an argument from the absurd: any utterance is constituted by a string of sequential sounds which, necessarily, can only come about and disappear one after the other – such are the limitations of being a body (as sound is), it comes in and out of being. Pushed to the absurd, on this line of argument, speech could never be more than the piecemeal uttering of one single phoneme at a time. However, the refutation is not aimed at defending an absurd position, but rather at bringing out an absurdity in the Stoic position, which associates a single incorporeal entity, a lekton, which S.E. here speaks of, as in the earlier passage from Diogenes Lartius, as ‘a thing’ (pragma), with a string of corporeal words: thus, saying one thing by means of uttering many words.

  • 23 The Stoics, in the relevant contexts, use specifically the verbs ‘hyparchein’ (‘to be present’ or ‘ (...)

28S.E. however, does not do justice to the Stoic position. He treats uttering and saying as if they were identical, rejecting from the start, the Stoics’ attempt at distinguishing them, in terms of the distinction in ontological categories of their objects. By keeping to an ordinary interchangeability between uttering and saying, S.E. shows that the upshot of a distinction is absurd and paralysing as far as any production of meaningful speech is concerned. But in highlighting the absurdity of taking saying and uttering to be interchangeable, he thereby demonstrates that there is something wrong with this identity assumption – which the Stoics go to great lengths to deny. S.E. also loosely fluctuates between different levels of ontological categories. Thus, he blurs distinctions, by misusing terminology. For he speaks of uttered words, which are corporeal, as ‘being present’ or, ‘obtaining’, for the Greek huparchei, which is a term that the Stoics reserve for lekta, and incorporeals in general, to demarcate their mode of reality from that of bodies23. Lekta are present, or obtain, whilst bodies exist. S.E. then moves from claiming that parts of (corporeal) sounds do not hyparchei or sunyparchei, that is, cannot be present, or obtain together, to concluding, that the incorporeal lekton, (which S.E. designates here as ‘pragma’), and which the words are meant to express, cannot hyparchei, obtain, as a consequence of the failure to ‘obtain together’ of the signifying words. At M. 8.81, of the first line of the Iliad, S.E. says:

  • 24 S.E. M. 8.81: “ἀλλὰ τούτου γε τὴν σημαντικὴν φωνὴν ἀμήχανόν ἐστι προφέρεσθαι διὰ τὸ οὗ τὰ μέρη μὴ σ (...)

“But it is not possible to utter the sound which is significant of this, because that whose parts do not obtain together, does not itself obtain.”24

29When we utter the first half of the verse, the second is not yet in existence, but even when we utter the first word, the second word is not yet in existence. It is not, moreover, because S.E. is merely speaking loosely, or that he is not aware of the technical and precise use of ‘obtaining’, or ‘being present’, in these contexts, that he seemingly makes such oversights. For it is clear that S.E. has precisely this technical terminology in mind in relation to lekta, and that he is, in effect, mocking the Stoics’ efforts to make a distinction between the way lekta are described (in terms of obtaining) and the way the utterances, meant to express lekta, are described (as corporeal items). The use of ‘hyparchei’ is not easily put down to an indifferent, common use, which would be simply interchangeable with ‘einai’. For it is a term, which is associated in particular with lekta in a discussion of the status of lekta. S.E. is knowingly referring to the ontological status of incorporeals, characterised by ‘hyparchein’. The critique is therefore made in full awareness of the Stoic position. It is an attack on the seeming incongruity of grasping at incorporeal lekta, through corporeal means. It is impossible, S.E. is saying, that a lekton obtain if the parts of the utterance which expresses it, come in and out of existence in alternation, assuming, thus, that the parts of an utterance should mirror presumed parts of a lekton.

30But the utterance is corporeal. Different rules govern its existence from the way lekta obtain. The utterance, in proceeding from a speaker to an interlocutor, acts: ‘the voice acts (poiei)’ (Diogenes Laertius 7.56). But this is not how lekta obtain. We have already seen that they cannot themselves be involved as agents or patients of a causal process. Therefore, the question of one word being uttered at a time cannot be the appropriate angle to take in questioning the possibility of there being such things as lekta. For the very notion of a lekton breaks with the one-to-one correlation between the existence of a word and the obtaining or subsisting of a lekton. Otherwise, a lekton might just as well be another term for ‘word’, or at best consist of the semantic shadow of words. On either view, a lekton is indistinguishable from a word. Once again, the critique comes down to rejecting an ontological framework which accommodates additional items, i.e. the obtaining incorporeals. Similarly to the Peripatetic approach, S.E. ends here, with stating that there is no room in reality for these additional items.

31And yet, the Stoics hold fast. Why? Perhaps the most pernicious attack comes from within the Stoic school itself, asking precisely this question: why be committed to the distinction, at an ontological level, between single words on the one hand, and what a sentence expresses on the other? The Stoic philosopher who gives one of the more precise accounts of the Stoic analysis of meaningful speech, displaying an accurate knowledge of its principles and the place and role of the lekton, but who at the same time, is, for that very same reason, one of the more damning critics, is Seneca. As with the previous critiques, it will become apparent that the attack relies on the rejection of an in-grained commitment to a specific ontological structure. To reject the latter indicates both a much more stark departure from his school’s fundamentals than Seneca would have us believe, but also, by contradistinction, shows how systematic and consistent the original Stoic theory is.

The Argument from Triviality

  • 25 The Stoics famously defended the interdependency of the three parts of philosophy, logic, physics a (...)
  • 26 Seneca, Ep. 117. 30: “transcurramus sollertissimas nugas et ad illa nobis aliquam opem sunt latura (...)

32In his Letter to Lucilius 117, Seneca presents the Stoic analysis of language as mapping onto fundamental divisions in ontology. He concludes his presentation with an impassioned complaint against the Greek Stoics – or what we tend to call, in modern scholarship, Stoic orthodoxy in comparison with, for instance, Seneca’s heterodoxy. He judges very harshly these distinctions which he argues throughout his account are imposed by logic but are but “trifling matters” (“nugaeEp. 117.30) compared to questions of life and death, which Seneca steers his reader to consider as not only, in fact distinct from questions of logic and ontology (a move which, in itself, is a fundamental breach with his Stoic forefathers25), but also as the only real questions of philosophy, those which pertain to ethics: what to do if you lose all your possessions, if your children, your country, are in danger, and so on26.

“Let us pass over all these overly clever trivialities and hurry on to things which will bring us some help.”

33However, though this is ultimately the view Seneca formulates, his own loyalty to his school compels him to go through the original Stoic commitments.

34The Stoics consider that there is a difference in ontological categories between what is designated by a noun like ‘wisdom’ and what is designated by a verbal phrase like ‘being wise’. Despite the semantic affinities between wisdom and being wise, these are two distinct things, with distinct ontological status:

  • 27 Seneca, Ep. 117.12: “Sic aliud est sapientia, aliud sapere.”

“Wisdom is one thing, to be wise is another.”27

35Wisdom is a corporeal thing. For the Stoics it is the corporeal soul in a certain state.

  • 28 Seneca, Ep. 117.2: “quod bonum est corpus esse, quia quod bonum est facit, quidquid facit corpus es (...)

“a good is a body, since a good acts, and whatever acts is a body. […] Wisdom is a good; it follows necessarily that we also say that wisdom is a body”28

  • 29 D.L., 7.58: ‘a noun (prosēgoria) signifies a common quality, like a human being or a horse’, a qual (...)
  • 30 See above n. 19.
  • 31 This much can be agreed upon also from distinct and parallel testimonies, for example an often cite (...)

36Corporeal things of this kind, i.e. items qualified in some way (here the soul, in the state of wisdom), are designated by nouns29. These are the signifying words (phonē sēmantikē) which, as suggested above, are uttered by speakers. But [being wise], which is what is expressed by the expression ‘being wise’ (the thing-lekton is distinguished here from the expression through the different punctuation marks), is, to follow Seneca’s wording, an “attribute” of wisdom, (“accidens”, Ep. 117.3). As such, it is not a thing which acts in any way, in contrast to wisdom. Hence unlike wisdom, [being wise] fails the test for corporeality. It is, by contrast to wisdom, an incorporeal. An incorporeal is a different kind of thing altogether in Stoic logic and ontology: it has its own distinct ontological status. It is a lekton, the incorporeal content of meaningful speech, which the Stoics distinguish from speech itself, precisely in that speech is corporeal30, but what it gets said, namely a lekton, is not. The lekton is something incorporeal which is one thing that can be expressed in different languages, and can even not be expressed at all, but is there, ready to be expressed31.

37Seneca goes into the details of this distinction in his Letter 117, challenging the Greek Stoics for betraying their commitment to linguistic naturalism, by which they purportedly distinguish themselves from the Peripatetic line. Seneca transforms the logical distinction into a form of imposed convention, in contrast to what is natural, (which would amount to not making the distinction). He renews, through this twist, with the topical adage of nature versus convention. The move is surreptitious: it is not a recasting of the question of the origin and function of language – in the way that the decidedly conventionalist view is taken up by the Peripatetics. For the Peripatetics, as we saw above, the direct relation of words to thoughts, puts all the onus of stipulating language usage to a convention regimented by different linguistic communities. Seneca revisits the debate through an internal critique of Stoicism for having forfeited its naturalist commitments. For it is against nature, i.e. it goes against the way people talk, says Seneca, to make a distinction between wisdom and being wise. For people talk about having wisdom or being wise interchangeably, and do not thereby mean something different. Seneca thus suggests a dismissal of the special status of lekta; they are not distinct ontological items. In so doing, he rejects two foundational Stoic claims: firstly, the reality of a distinction between a body, like wisdom, and a state-of-affairs, (which characterises a body but is not identical with that body), like [being wise]; secondly, the distinction, in terms of ontological categories, between what we say and how we say it, namely through the utterance of precise sequences of words.

  • 32 See in particular, S.E., P.H. 3.14, and Clement, Strom. 8.9.26.4.

38Seneca thus contributes in complex ways to the long tradition of interpreting and criticising the Stoic theory of lekta. Firstly, as we see, he does so by casting doubts on its validity. Secondly, as we shall see in the following, he also amplifies the critique by reporting the criticism brought forward against the theory by rival schools, in particular by the Peripatetics. His testimony to this debate shows on the one hand, its importance within the broader picture of the discussions in antiquity about logic and language. In the Aristotelian commentaries, there is an intentional marginalisation of the Stoic position by different means: a parenthetic mention, or a re-assimilation of Stoic views to Peripatetic schemata as we saw above. From Seneca, as also from other more neutral sources32, we are able to appraise to its more far-reaching dimensions, the purport of the debate between the Stoics and the Peripatetics as consolidating the contrasting positions from both sides. The debate is framed, not so much by the pointed question about the production of meaningful speech, but rather in more comprehensive terms, by the radical differences between the ontological structures each school relies on. Seneca’s testimony also serves to put his own criticism in perspective, in that his defence of the Stoics against the Peripatetics exposes a certain frustration Seneca has with the role of logic in Stoic doctrine, but ultimately also shows the importance of that role as guaranteeing the solidity and consistency of Stoic doctrine. His critique thus tapers down into soft-sabotage, for instance through a reluctance to give a consecrated translation in Latin of the term ‘lekton’, or ultimately dismissing the whole discussion as a detail in logic with no deep incidence on ethics.

  • 33 See Aristotle, Cat.4, 2a34-b6, and on this: Wolfgang R. Mann (2000), The Discovery of Things, Princ (...)

39The complexity of Seneca’s position lies in his loyalties. He demarcates himself from the official critics from rival or generally antagonistic schools of thought, who, within the strict context of logic, deny the reality of lekta. This is at the base, the discussion between the Stoics and the Peripatetics: for the latter, saying something of something, e.g., of Socrates, that he is wise, is nothing other than articulating the relation between the property wisdom and the substance Socrates. The predicative formulation ‘is wise’ is a tool for expressing that relation, which can be interchangeably expressed as wisdom’s belonging to Socrates33. The discussion thus, is grounded in views on logic and ontology: it turns on whether language is revelatory of ontological distinctions (for the Stoics), or a tool at the service of expressing the subordinated relations between substances and properties via associations of thoughts (for the Peripatetics). Seneca is in a different position: he is a Stoic (as he reminds us by including himself in the first-person pronoun ‘we’, the Stoics at Ep. 117.2, and 117.6 and repeatedly throughout the letter). Thus, in the face of rival schools, he defends the tenets of his school. He defends, that is, the coherence and consistency of the Stoic position against the Peripatetics. As far as logic is concerned, the Stoic theory is vindicated by Seneca, who gives us one of the more complete accounts of lekta left to us, at Ep. 117-12-3.

  • 34 Zeno of Citium, the founder of the Stoic school, was a student of Diodorus Cronus, who is associate (...)

40He premisses the defence with a mini-historiography of the view, which he refers back to ‘the old Dialecticians’ (Ep. 117.12). This reference bears a subtle negative insinuation, which echoes Seneca’s first comment on the Stoic distinction between wisdom and being wise. Thus, a few lines earlier, at Ep. 117.6, Seneca explains that the Stoics are “weighed down by an original chain (primo vinculo)” which forces them to abide by the distinction. Seneca suggests in this way, that the distinction is neither useful nor consonant with other commitments of the school, namely respect of ordinary language, or the imperative to bring solutions to how to live. But the Stoics are forced, according to Seneca, to integrate the distinction into their doctrine, because of this original chain, which they cannot rid themselves of. When Seneca, a few paragraphs down, points to the legacy of the “Old Dialecticians”, as the origin for the Stoic distinction, it seems as though he is indicating these Dialecticians, as constituting the primum vinculum, the original shackle or chain, which is holding down the Stoics34.

  • 35 It is, for instance, this broad affiliation which underlies Cicero’s mention of the “dialectici”, i (...)

41The subtlety of Seneca’s position is twofold: at a first level, he understands the Stoic distinction in logic. Thus, when it comes to school loyalties, Seneca defends Stoic logic against the Peripatetics. At a second level, he traces the Stoic view back to ancestors of the Stoics, who, legitimately, can be said to have some influence on Greek Stoicism. We need not assume that Seneca is actually referring to the direct relations of the Dialectical school with the founder of the Stoic school, Zeno, (who had been, amongst others, also a student of Diodorus Cronus, see n. 34). Seneca might be more generally correlating the Stoics with the Dialectical school(s), in view of the importance of logic, i.e. dialectic, in Stoicism35. What Seneca is intimating however, is that, having inherited the distinction, there is no actual doctrinal need to commit to it, since other foundational doctrines of stoicism clash with it – amongst them, is the commitment to a linguistic naturalism, which Seneca seems to take to be something akin to common parlance. Seneca comes out of these considerations as more Stoic than the Stoics, explaining and defending the logic, whilst at the same time, motioning towards a need to purify the doctrine from external elements which weigh down on it, disburdening Stoic doctrine from an inheritance, which is alien to its purpose.

42Seneca’s critique thus consists in urging the Stoics to submit their analysis of language to their ethical tenets: what to do in order to be wise, not to fear death, and be good. Part of the answer is to forfeit logical distinctions, and embrace ordinary, natural language, as the means to collapse the ideology into the goal.

43The strategy employed is to stake out the immediate application of the distinction. Seneca claims that, even in acknowledging the logical and ontological difference between wisdom and being wise, this does not lead to any real, useful distinction. It is universally acknowledged that wisdom is a good, indeed it might even be the good, but it then follows, according to the Senecan presentation of the Stoic account that:

  • 36 Seneca, Ep.117. 1: “quod Stoicis placet, sapientiam bonum esse, sapere bonum non esse.”

“Wisdom is a good but to be wise is not”36.

44This is an absurd outcome, for Seneca. For when we say that wisdom is a good or that being wise is a good, we are saying one and the same thing. The “extremely clever” (“sollertissima”) distinction the Greek Stoics put so much analytic effort into, Seneca shows to be but a “quaestiuncula”: an insignificant piece of tedious pedantry. For it turns out that the distinction is only a surface distinction, because what counts is the way we speak:

  • 37 Seneca, Ep.117.3: “Quid ergo? […] non dicimus: bonum est sapere?”

“Do we not say that it is good to be wise?”37

45There are two levels of attack here. Seneca is on the surface of things referring to the run-of-the-mill way in which people speak. The appeal to common parlance focalised in the use of “dicimus yields an almost naïve form of criticism, as if to say: ‘but surely this is how we speak: ask anyone, we all would just say it is good to be wise, or wisdom is good, indifferently’. We obviously mean the same thing.

46The original Stoic thought behind the distinction is however, very far from the question of common parlance, as Seneca himself suggests. For Seneca re-enacts an imaginary dialogue with the Greek Stoics, who appear to retort to the common parlance view channelled by Seneca, that surface expression covers complex ontological relations of dependencies:

  • 38 Seneca, Ep.117.2: “Dicimus referentes ad id ex quo pendet, id est ad ipsam sapientiam”

“Yes, we say it referring to that on which [being wise] depends, namely on wisdom itself.”38

47The Stoic analysis of language is shown to refer to the fundamental constituents of ontology, thus loading the notion of ‘saying’ with far greater purport than a mere matter of expression (as my italics in the quotation bring out). To say is to reveal the inner structure of reality. Seneca gives a deliberately heavy-handed articulation to this view. He thereby ridicules it. Instead of getting close to nature, the Greek Stoics block the path by positing so many hurdles, which separate what we say from the real things we mean to say. These hurdles are represented by the accumulated series of prepositions and pronouns: ad quod (to that), id ex quo (on which), id est (that which), ad ipsam (to which). This much is clear from Seneca’s presentation of the view: for him, linguistic naturalism is on the side of common parlance. He shows the Stoics to be taking grammar to reveal the structure of reality, but this, for him, only leads to a further estrangement of language from nature.

48The Greek Stoics do indeed take grammar to be revelatory of real distinctions. The contrast therefore, which Seneca is making by insisting on the naturalism of common parlance, is a contrast about how to understand the relation of language to reality. For on the one hand, linguistic naturalism is conceived as both borne out of, and at the same time, providing access to, ontological complexity, and on the other hand, the Senecan view is that linguistic complexity, (involving synonymy, paronymy, syntactical norms etc.), is merely linguistic variation, and does not itself imply further distinctions in logic and ontology. Seneca silences the former view on linguistic naturalism, by making the Greek Stoics appear to have themselves attempted to rely on common parlance to give an account of linguistic naturalism, just as Seneca does, with the result that the Greek Stoics end up giving an overly complicated account of it. Seneca thus posits an alternative Stoic view without seemingly breaking the confines of Stoic doctrine. The question then is who is a better Stoic, asking which account of linguistic naturalism, of Greek Stoicism or Senecan Stoicism, is the one best capable of articulating life, and language, according to nature.

49For Seneca, the convoluted interpretation of a simple sentence such as ‘it is good to be wise’ is the ultimate proof that the Greek Stoics twist language to give unnatural explanations of what is, in effect, perfectly clear and natural. He thus motivates a deep-seated criticism common in the tradition of anti-Stoic thinkers: that the Stoics distort language needlessly, “verba torquere” as Seneca writes himself (at Ep. 117.5). It is a critique with a double-edge. For, at base, the Stoics are accused of making pettifogging distinctions, which have no relevance to the real problems of life and philosophy. But the critique very quickly extends to accusing the Stoics of making distinctions, on the basis of language, which have no correspondence in reality. Seneca himself goes to that extent: for he both acknowledges the distinction, explains its origin, all the better to subsequently reject it bluntly.

50For Seneca defends the Stoic distinction against the Peripatetics. He does so by having recourse to an analogy, modelling the distinction between wisdom and being wise on the difference between a field and having a field:

  • 39 Seneca, Ep. 117.12: “aliud es ager, aliud agrum habere, quindi?”

Is it not the case that a field is one thing, and having a field is something else?”39

51There are two ways of reading Seneca’s analogy, and both play a role in his defence-but-not-quite of the Stoic position. A weaker reading towards which all of Seneca’s comments point to, consists in putting the difference down to a difference in degrees of independence of status: a field is the field it is regardless of anything else in addition, whereas having a field depends on there being a field in the first place, and on there being someone to possess it. In the same way, wisdom is what it is in and for itself, whereas being wise depends on wisdom being there in the first place. The stronger reading of the analogy takes in the consequences of this difference in independence. The field is a corporeal entity, [having a field] is incorporeal. Incorporeality corresponds to an utterly different status in ontology. It is in virtue of that status that [having a field] is the sort of thing which can stand in a certain relation both to the field and to the possessor.

  • 40 The grounds for this backpedalling are far from clear, and therefore all the more fragile for his c (...)
  • 41 With n. 11 above, see also Galen, Μeth. Μed. X, 155,7-8; Simplicius, In Cat.. 310.14-312.37, and mo (...)

52When it comes to wisdom once more, Seneca abandons the analogy, on the grounds that the field is external to its possessor whilst, in the case of wisdom, wisdom is internal to its possessor. Therefore, though there is a distinction between a field and having a field, there is ultimately no such similar distinction between wisdom and being wise. Seneca thus backpedals on his initial acknowledgment of the distinction40. For Seneca, ultimately, the Greek Stoics got carried away by language, to make distinctions where there is nothing to distinguish. Seneca thus contributes to the reputation, which the Stoics gain for paying too much attention to words rather than to things. He bolsters thus the view that the Stoics are terminological innovators, rather than philosophical or conceptual innovators, as one of the Peripatetic critiques puts it as well, according to which the Stoics are talking about the same notions as the Aristotelians, only that the Stoics call them by different names41.

53The Greek Stoics emerge from this first Senecan critique shrouded with a little ridicule: common parlance shows their distinction to be overruled as non-existent – this is the deflationist interpretation of the “quaestiuncula”. At the same time their own technical explanation of ordinary language is shown to be pedantic and alienating. Through the repeated “dicimus”, Seneca stages them as uselessly overloading the elements of speech: according to the Stoics we do not really say what we say. When we say that to be wise is good, what we actually say is something quite different, though nobody other than Stoic logicians realises as much.

54The critique is, at a third level, more pernicious still, as Seneca seems to show that the Greek Stoics are incapable of explaining ordinary language. Ordinary language is imprecise with respect to reality and therefore needs a technical unnatural language to express it. Seneca thus presents the Stoics as twistedly claiming that ‘saying X, actually means to refer to P which depends on Y’. But the Stoics are par excellence the philosophers who claim not only to have a complete system grounded on the principles of nature, but also that they are exclusively concerned with the understanding of nature – and their analysis of language is no exception to this fundamental commitment. Seneca is therefore questioning here a fundamental Stoic claim: the consistency of the naturalistic stake the Stoics originally make, as a basis for the relation between ontology and language.

55For the original Stoic theory, the relation is one of revelation: language reveals reality. The distinction in form, between the noun ‘wisdom’ and the verbal phrase ‘being wise’ reveals a real distinction, in nature, i.e. in the structure of reality, which has consequences with regard to moral choices: e.g., ‘What do I want? Wisdom or [being wise]? A body, or an incorporeal lekton becoming true of me? Seneca’s rephrasing of the Stoic position, namely his collapsing of the ontological theory (the distinction between a body, wisdom, and its attribute, [being wise]) into ethical doctrine concerning the nature of the good makes the Greek Stoics seem to be merely cavilling.

  • 42 Seneca, Ep. 117.1: “quod Stoicis placet”.
  • 43 Half of Seneca’s Letter 117 is a long and bombastic rant about the superfluous contribution of logi (...)

56But in Seneca’s great haste to reach the “important questions about life and death”, he takes a few shortcuts. For though Seneca presents the claim that wisdom is good but [to be wise] is not, as a Stoic tenet42, this is not an original Stoic formulation, but a simplification. The question is whether this simplification is legitimate. It is based on Seneca’s dissatisfaction with the Stoic commitment to the interrelatedness of the parts of philosophy (cf. n. 25). He identifies inconsistencies when the logic is applied to the ethics and the ethics to the logic. And this is so, because language, which is the intermediary between logic and ethics, is bound to the logical structures, when it should be, in fact, bound to ethics43. Language should be disengaged from its role as relay of the fundamental structures of reality, and be deflated into a surface grammar, which reflects external phenomena – and is therefore natural – rather than internal hidden structures. Only then will language be properly natural.

  • 44 Seneca does so for notions, and corresponding concept-terms, which are important to him and which h (...)

57These hidden structures are the lekta: the logical keystones of the Stoic explanation of language. As an ultimate move in the reshuffling of the linguistic natural order, Seneca downgrades the place and role of lekta by not considering the very word ‘lekton’ fit for a unique translation. The term itself of ‘lekton’ might best be left transliterated, as we have done so far in these pages (and as Cicero does with the Greek term ‘katēgorēma’, a kind of lekton, in the passage quoted in fn. 35). This is so, precisely because already in the Greek sources, it is a consecrated, concept-term. It is thus only an appropriately consecrated concept-term, which would be a suitable translation; that is to say, a term, which is symbolic and connoted, rather than descriptive44. But Seneca does neither the former nor the latter.

58Seneca comes to discuss lekta properly speaking, two pages into Letter 117, when he comes to defend the Stoic position against the Peripatetic lekta-deniers. But, in fact, the whole letter is about the notion of what gets said when we speak, e.g. [being wise] in uttering ‘is wise’, (a katēgorēma, which is one kind of lekton), which corresponds to what gets said about something, as opposed to when we utter the word ‘wisdom’ by which we name the corporeal entity, wisdom. Precisely because Seneca means to dismiss the importance of lekta, and the validity of the distinction between being wise and wisdom, he does not even give the term ‘lekton’ a consecrated translation in Latin. He merely gives an emphatically tentative string of approximate translations.

  • 45 Seneca, Ep. 117.13: “Dico deinde: Cato ambulat. Non corpus inquit est quod nunc loquor, sed enuntia (...)

“Then I say: ‘Cato walks.’ He says: ‘What I am now saying is not a body but something which can be enunciated about the body; some people call this an effatum, others call it an enuntiatum, still others call it a dictum45.

  • 46 Cicero’s ‘enuntiatum’ becomes canonical, presumably alongside a suggestion from Varro (‘proloquium(...)
  • 47 Seneca thus echoes one etymological explanation for the Greek term: kategorema: “κατὰ τινος ἀγορευό (...)

59It is not so much the problem of translation, as the approach to translation, which is of interest. By not following up on the need for a consecrated term in Latin, Seneca at once diminishes the importance of the notion of a lekton, and subtly attacks the legitimacy of the term itself, and what it designates. Suffice it to contrast this with Cicero’s clear-cut translation of ‘lekton’ as ‘enuntiatum’ (De Fato, 19) raised to the status of concept-term as it is with the Greek Stoics46. Seneca adopts a descriptive mode of transmission of doctrine: what a thing is like, rather than what it is. The first suggestion of ‘enuntiativum’ would seem to be his own, as it does not appear in other Latin authors, but the immediate listing of other renderings contributes to blur the contours of the specific notion in question. Indeed, given these multiple choice translations, Seneca himself ends up reverting back to a periphrasis, describing the distinction between wisdom and being wise as that between a thing, and “that which is said about a thing”: “de illo loquor47.

60The use, or rather excuse of translation, is a means to revisit the original Greek concepts, and propose a novel, revisited and simplified version in Latin. The simplification is justified as a form of naturalisation of the original doctrines. These are seemingly overburdened by commitments taken in the process of elaboration of the theories. But once the theories are properly set out, these commitments appear to be superfluous. In this way, Seneca can present his interpretation of Stoicism as faithful to the original naturalist intent of the Greeks, leaving out elements of doctrine, not because he does not understand them or is not aware of them (he is clearly extremely well informed), but rather because, having properly grasped the purport of the Stoic injunction to live according to nature, he is in the best position to show how the doctrine is true to its goal, despite appearances. Descriptions rather than conceptualisations, literality rather than metaphorical or technical language. In the name of the theoretical naturalism which grounds Stoic philosophy, and which guarantees the interrelation of logic, physics and ethics, Seneca pushes towards cancelling out the link between logic and language. That is to say, that he ends up reversing the natural order which the Greek Stoics discovered in the dependency of language on logic and ontology. For the Greeks, linguistic naturalism was in effect, discovered with the discovery that surface grammar is moulded onto deep logical distinctions. With Seneca, linguistic naturalism is now accounted for by not delving any deeper than surface grammar, which, though it might be derivative on ontological concerns, is enough to go by, in order to realise the moral injunction of living, in all practical dimensions, including speech, according to nature.

Conclusion

  • 48 E.g., Gabriel Nuchelmans (1973), Theories of the Proposition. Ancient and Medieval conceptions of t (...)
  • 49 E.g., Michele Alessandrelli (2013), Il Problema del Lekton nello Stoicismo Antico: origine et statu (...)
  • 50 E.g., John Rist (2006), “Zeno and the Origins of Stoic logic”, 1978¹, in Jacques Brunschwig (éd.) L (...)

61What is wrong with lekta? The critiques we have examined here point, from different perspectives, to one fundamental aspect of the Stoic theory: that the Stoics do not accommodate or extrapolate lekta to supplement a question about a philosophy of language, but rather start off with a view of reality, which is constituted, entre autres, by lekta. The subsequent role of lekta in the production of meaningful speech is, to all intents and purposes, an implication of the establishment of an ontological structure founded on the presence of lekta. Language comes after ontology. In isolating a Stoic account of language from its foundations in logic and ontology, the theory breaks down, since there is no incorporeal thing there to identify as the lekton, which a sentence gets said. The influence of this critical approach to lekta on the history and development of logic has been extremely powerful since the Stoic theory of lekta was never revendicated as such. But the debates it generated in antiquity are also crucial turning-points, since the discussion of the rejection of lekta opens breaches which have only ever grown larger, concerning the status of propositions, and the reality of states-of-affairs, or facts. The ancient critics have also been very influential in shaping the reception in modern scholarship of the Stoic lekta. Their authority has led many modern interpreters in their footsteps, who have consistently interpreted lekta as either mental items48, or purely linguistic items and then, as the correlates of words49, in the wake of the Peripatetic treatment of lekta. Others have belaboured the oddity of lekta from within the Stoic system itself, to end up claiming that lekta are brought in to solve topical problems but do not belong to, nor contribute to the original comprehensive account of the world50. The latter direction of interpretation follows in the tracks of Seneca’s questioning whether every claim in Stoicism has an equal weight of relevance and importance for every other question.

62The moderns have of course, the additional impediment of having almost only the critiques to go by. The testimony of the critics however, cannot be taken at face value, nor can it be taken as utterly unreliable. Somewhere in between, we have begun to see that we can steer through the motivations and hidden agendas of the critics. We are thus able to recognise the more or less concerted misrepresentations, and the biases that underscore the tradition of critique of lekta. From it, we can salvage, as from a negative to a positive, some fundamental traits of the Stoic account. If, as with the Peripatetics, the question revolves on whether words express thoughts or something distinct from us, (which would be the lekta), then focusing on the Stoic proposal of lekta, whilst denying or invalidating the presence of lekta as additional ontological items, leads to merging the Stoic account into the Peripatetic account – and thus confirming the latter’s superiority. If, as with Sextus Empiricus, the question revolves around the possibility of distinguishing what gets said from what we utter, then, by denying the difference in ontological categories between corporeal sounds, and an incorporeal unit of meaning (which would be the lekton), the distinction falls apart and what we say is what we utter, syllable by syllable. If, lastly, with Seneca, the question revolves around whether acknowledging the presence of lekta in ontology, makes a difference in ethical behaviour, then, by dismissing the structural inter-dependence of logic and ethics, the distinction between a body like wisdom and an incorporeal lekton like [being wise] collapses.

  • 51 Stob. Ecl. 2.7.6f.
  • 52 S.E. M. 7. 38-40.
  • 53 On lekta as the objects of impulse and desire, see Stob. Ecl. 2.7.9a; on lekta as what is true or f (...)

63However, Seneca is wrong. The Stoics develop the relevance of the distinction between a body and a lekton precisely in terms of ethical goals. Thus, for instance, the Stoics distinguish between ‘what one can choose’ (haireton), which is a body like wisdom, and ‘what one should choose’ (haireteon) which is a lekton51. What we should choose, that is to say, what we should go after are lekta like [to be wise], because what we want is for that lekton to become true of us. We do not go after the body wisdom, because in effect, it is not the sort of thing we can access. Thus, the Stoics say that we can do things which are wise, say things which are true, without wisdom or truth being in our soul52. More generally, it is the incorporeal status of a lekton which enables us to enter into a number of relations to it: of saying it, thinking it, wanting it, acting in the way most propitious for it to become true or false of us53. These further applications and possibilities go beyond the scope of this article. But they follow from what, ultimately, is confirmed and reconfirmed throughout this inquiry through negativity, recovering the Stoic lekton from under a mass of critical clouds, namely that lekta are distinct from speech and from thought, and therefore, have a presence in ontology.

64What our inquiry shows is that the critics attack the Stoics for considering lekta to play a linguistic role, when in fact, lekta are not linguistic items. They are ontological items. But this is precisely what the critics cannot accept, that reality be constituted of items, which have linguistic articulations. For what this would mean – and indeed, what this comes down to for the Stoics – is that language merely reflects the structural components of reality. It is not, therefore, a human creation. Rather, human beings, endowed with reason, cannot but express that structure through the use of language, which they cannot but help acquire, though the modes of expression of a language will differ from one linguistic community to another. Thus, though lekta are legitimately discussed within the debates about language, it is because they are the kind of items which play a role in language, though their status and nature is not determined by any successful production of meaningful language. Since the Stoics grant human beings a certain degree of rationality, this involves humans having access to language, and therefore access to lekta. This invites the misleading apprehension of lekta as belonging first and foremost to a theory of language, whereas it is language which is derivative on the rational structure of the world, held up by the lekta.

Haut de page

Bibliographie

Ancient Texts

ps-Alexandre I (=Michael of Ephesus), In Sophistici Elenchi, Wallies, M. (ed.) (1898), CAG ii.3. (in S.E.)

Ammonius, In Aristotelis de Interpretatione commentarium, Busse, A. (ed.) (1897), CAG, iv.5, Berlin. (in de Int.)

ps-Ammonius, In Aristotelis Analyticorum priorum librum I commentarium, Wallies, M. (ed.) (1899), CAG iv.6, Berlin (in An.pr)

Aristotle, Categoriae et Liber de Interpretation, Minio-Paluello, L. (ed.) (1949) Oxford. (Cat.; Int.)

Apuleius, De Interpretatione, in David Londey & Carmen Johanson (1987), The Logic of Apuleius, including a complete Latin text and English translation of the Peri Hermeneias of Apuleius of Madaura, Philosophia Antiqua, Leiden, Brill.

Boethius, In librum Aristotelis de Interpretatione- secunda editio, Meiser, C. (ed.) (1880), Leipzig. (in de Int¹. and in de Int².)

Cicero, De Fato, in Traité du Destin, Yon, A. (éd.) (1933) Paris, Les Belles Lettres.

De finibus bonorum et malorum, Reynolds, L.D.,(ed.) (1998), Oxford. (De fin.)

Academica Posteriora – Academica Priora, (Academicorum reliquae cum Lucullo), Plasberg, O. (ed.), Leipzig:Teubner 1922, repr. Stuttgart 1980. (Acad. Post. – Acad. Pr.)

Tusculanae Disputationibus, Pohlenz, M. (ed.) (1903), Leipzig, repr. Stuttgart 1967.

Clement of Alexandria, Stromata I-VI, (ed.), O.Stählin, L.Früchtel & U.Treu, Berlin 1985, CGS 52² and Stromata VII-VIII, 1970, GCS 17². (Strom.)

Dexippus, In Aristotelis categorias commentarium in Busse, A. (ed.) (1888), Dexippi in Aristotelis categorias commentarium (CAG 4.2.) Berlin, Reimer. (in Cat.)

Diogenes Laertius, Lives of the Eminent Philosophers, Dorandi, T. (ed.) (2013), Cambridge. (D.L.)

Galen, De methodo medendi, in Kühn, C.G. (1821-33), Clauddi Galeni Opera Omnia, vol. 10 (1825). (Meth. Med.)

Institutio logica, Kalbfleisch, K. (ed.) (1896), Leipzig. (Inst.Log.)

ps-Galen, Historia Philosophiae, in Diels, H. (1879). (Hist. Phil.)

Gellius, Aulus, Noctes Atticae, Marshall, P.K. (ed.) (1968), 2 vols., Oxford. (Gell.)

ps-Alexandre I (=Michael of Ephesus), In Sophistici Elenchi, Wallies, M. (ed.) (1898), CAG ii.3. (in S.E.)

Plutarch, Platonicae quaestiones, in Hubert, C. (ed.) (1954), Plutarchi moralia, vol. vi.1. Leipzig. (Plat.Quaest.)

De Communibus Notitiis, De Stoicorum Repugnantis, in Cherniss, H. (ed.) and transl. (1976), Plutarch’s Moralia, vol. xiii.2, Cambridge, Mass. & London. (De Comm.Not., Stoic. Rep.)

De Virtute Morali, in Pohlenz, M. (ed.) (1929), Plutarchi Moralia, vol iii, Leipzig. (Virt. Mor.)

Seneca, Epistulae Morales, Reynolds, L.D. (ed.) (1965), 2 vols., Oxford. (Ep.)

Sextus Empiricus, Pyrrhonei Hypotyposes, in Mutschmann, H & J. Mau (eds.) (1954-61), Opera, vol. 1, Leipzig. (S.E. PH)

Adversus Mathematicos, in Mutschmann, H & J. Mau (eds.) (1955-61), Opera, vols. 2&3, Leipzig. (S.E. M.)

Simplicius, In Aristotelis Categorias Commentarium, Kalbfleisch, K. (ed.) (1907), CAG viii, Berlin. (in Cat.)

In Aristotelis Physica Commentaria, Diels, H., (ed.) (1882-95), CAG ix & x, Berlin. (in Phys.)

Stobaeus, Eclogae, in Wachsmuth, C. (ed.) (1884), Ioannis Stobaei Anthologii libri duo priores qui inscribi solent Eclogae physicae et ethicae, 2 vols., Berlin, repr. (1974) Zurich. (Stob. Ecl.)

Suda, Adler, A. (ed.) (1928-38), 5 vols, Leipzig.

Themistius, in Wallies, M. (ed.) (1884), Themistii quae fertur in Aristotelis analyticorum librum (i) paraphrasis, CAG, xxiii.3, Berlin. (in An.pr)

Secondary Literature

Alessandrelli, Michele (2013), Il Problema del Lekton nello Stoicismo Antico: origine et statuto di una nozione controversa, Florence, Olschki.

Atherton, Catherine (1993), The Stoics on Ambiguity, Cambridge, CUP.

Bronowski, Ada (2016), “Rational Empiricism: the Stoics on Reason, Experience and Katalepsis”, in Logical Analysis and History of Philosophy 19, p. 167-187.

— (2019), The Stoics on Lekta: All there is to say, Oxford, OUP.

Brunschwig, Jacques (1988), « La théorie stoïcienne du genre suprême et l’ontologie platonicienne », in Jonathan Barnes & Mario Mignucci (eds.), Matter and Metaphysics, Naples, Bibliopolis, p. 19-127.

— (1995), « La théorie stoïcienne du nom propre », 1984¹, in Jacques Brunschwig, Études sur les philosophies hellénistiques, Paris, PUF, p. 115-139.

Cooper, John M. (2004), Knowledge, Nature and the Good: Essays in Ancient Philosophy, Princeton-Oxford, OUP.

Denniston, John D. (1950), The Greek Particles, 1934¹, Oxford, OUP.

Detel, Wolfgang, R. Hülsen, G. Krüger & W. Lorenz (1980), “λεκτὰ ἐλλιπῆ in der stoischen Sprachphilosophie”, in AGP 62, p. 276-288.

Dillon, J.-M. (2014) trans. and notes, Dexius: On Aristotle’s Categories, London, Bloomsbury.

Duhot, Jean-Joël (1989), La conception stoïcienne de la causalité, Paris, Vrin.

Frede, Michael (1994), “Stoic notion of a lekton”, in Stephen Everson (ed.), Language, p. 109-128.

Gaskin, Richard (1997), “The Stoics on Cases, Predicates and the Unity of the Proposition' in Aristotle and After”, Bulletin of the Institute of Classical Studies, Summer 68, p. 91-108.

Goldschmidt, Victor (1972), “ὑπάρχειν et ὑφεστάναι dans la philosophie stoïcienne”, Revue des Études Grecques 85, p. 331-344.

Gourinat, Jean-Baptiste (2009), “The Stoics on Matter and Prime Matter: corporealism and the Imprint of Plato’s Timaeus”, in Roccardo Salles (ed.), God and Cosmos in Stoicism, Oxford, OUP, p. 46-69.

Graeser, Andreas (1977), “Language, Thought and Reality in Ancient Greek Philosophy”, Dialectica 31, p. 369-388.

— (1978), “The Stoic Theory of Meaning”, in John Rist (ed.), The Stoics, Berkeley, University of California Press, p. 77-100.

Inwood, Brad (2007), Seneca - Selected Philosophical Letters: Translated with Introduction and Commentary, Oxford, OUP.

Isnardi-Parente, Margherita (2005), « La notion d’incorporel chez les Stoiciens », in Gilbert Romeyer-Dherbey (dir.), Jean-Baptiste Gourinat (éd.), Les Stoiciens, Paris, Vrin, p. 175-185.

Kupreeva, Ina (2003), “Qualities and Bodies: Alexander against the Stoics”, OSAP 25, p. 297-344.

Londey, David & Carmen Johanson (1987), The Logic of Apuleius, including a complete Latin text and English translation of the Peri Hermeneias of Apuleius of Madaura, Philosophia Antiqua, Leiden, Brill.

Long, Anthony A. (1971), “Language and Thought in Stoicism”, in Anthony A. Long (ed.), Problems in Stoicism, London, Athlone Press, p. 75-113.

Mann, Wolfgang-Rainer (2000), The Discovery of Things, Princeton, Princeton University Press.

Nuchelmans, Gabriel (1973), Theories of the Proposition. Ancient and Medieval conceptions of the bearers of truth and falsity, Amsterdam, North-Holland Publishing.

Rist, John (2006), “Zeno and the Origins of Stoic logic”, 1978¹, in Jacques Brunschwig (éd.) Les Stoiciens et leur logique, 1978¹, revis. edit., Paris, Vrin, p. 13-28.

Sedley, David N. (1977), “Diodorus Cronus and Hellenistic Philosophy”, Proceedings of the Cambridge Philological Society, 203 NS 23, p. 74-120.

Haut de page

Notes

1 The aim of the present inquiry is thus to uncover and recover Stoic doctrine about lekta from purely negative discussions which, as we shall discover, serve an agenda tied, not to elucidating Stoic theory, but to promoting the view of the critic, at the expense of the Stoics. Other discussions in antiquity by rival schools of Stoic doctrine, which are more neutral do not therefore belong to this present inquiry; the debates with the Epicureans and the Platonists are cases in point, whose modalities and dynamics are different enough to demand full discussions elsewhere, see Ada Bronowski (2019), The Stoics on Lekta: All there is to say, Oxford, OUP.

2 Amm. in de Int. 17. 24-28: “πρότερον ἡμᾶς ὁ Ἀριστοτέλης διδάσκει διὰ τούτων, τίνα ἐστὶ τὰ προηγουμένως καὶ προσεχῶς ὑπ’ αὐτῶν σημαινόμενα, καὶ ὅτι τὰ νοήματα, διὰ δὲ τούτων μέσων τὰ πράγματα, καὶ οὐδὲν ἕτερον δεῖ παρὰ ταῦτα ἐπινοεῖν μέσον τοῦ τε νοήματος καὶ τοῦ πράγματος, ὅπερ οἱ ἀπὸ τῆς Στοᾶς ὑποτιθέμενοι λεκτὸν ἠξίουν ὀνομάζειν.”

3 Ammonius is well aware of this: in the first pages of his in de Int. (5.29-7.14), he recalls how the authenticity of Aristotle’s De Int. was questioned by Andronicus of Rhodes precisely because it seemed incorrect to the latter for noēmata to be interchangeable with the pathēmata tēs psuchēs; Ammonius proceeds to demonstrate the grounds, in the Aristotelian text, for the validity of the interchangeability, and the authenticity of the text. The substitution for one or the other is belaboured in the Peripatetic tradition: see Boethius, in de Int. II.11.13ff on Alexander’s response against Andronicus, and Boethius, in de Int. II.12.18ff, on Porphyry’s response.

4 See Simplicius, in Cat. 41.28-9, who appears to be quoting from Iamblichus, who presumably draws his sources from Porphyry, with a parallel in Dexippus in Cat. 9.22-3; see on this tradition, John Dillon (2014) trans. and notes, Dexius:On Aristotle’s Categories, London, Bloomsbury, p. 28 fn.16.

5 Dexippus, in Cat. 9.24-5: “προηγουμενῶς μὲν τὰ νοήματα, κατὰ δεύτερον δὲ λόγον καὶ τὰ πράγματα σημαίνεται.”; see also Simplicus in Cat. 42.2-7.

6 Ammonius, in de Int. 19.1-3, appeals directly to the Aristotelian text, saying that when Aristotle says that letters and words are not the same for everyone (Arist. Int. 16a5-6), what he thereby means are that whilst ‘things and thoughts are by nature (phusei), words and letters are by convention (thesei)’.

7 Ammonius, in de Int. 20.7: “τὸ δέ σύμβολον ἤτοι σημεῖον τὸ ὅλον ἐφ' ἡμῖν ἔχει.”, ‘a symbol or a sign is entirely up to us’. This relation is contrasted (in de Int. 20.1-2) to the relation between thought and thing, which is not up to us insofar as there is nothing that we can do about what the thought is like, since it is dependent on the thing it is the thought of for its being, and hence is a ‘likeness’ of that thought.

8 Thus, Ammonius explains that Aristotle identifies assertoric sentences as the subject of ‘interpretation’ (hermeneia), justifying thereby the title of the work, because these sentences ‘can interpret the knowledge of the soul’ (Amm. In de Int. 5.17-19), where ‘interpret’ is further glossed as expounding or announcing (for ‘exangeltikon’) the ‘knowledge which arises within us’ (in de Int. 5.15).

9 Themistius, in An.pr. 92.1-3: “οἱ δὲ Στωϊκοὶ καινοτέραν βαδίζοντες τὰ μὲν πράγματα τυγχάνοντα ὠνόμασαν, διότι τῶν πραγμάτων τυχεῖν βουλόμεθα, τὰ δὲ νοήματα ἐκφορικά, διότι, ἅπερ ἐν αὑτοῖς νοοῦμεν, ταῦτα εἰς τὸ ἔξω προφέρομεν, τὰς δὲ φωνὰς λεκτά.” The same comparative grid is referred to subsequently in Philoponus, in An. Pr. 243.4, in Ps-Ammonius, in An.Pr. 68.6, in Ps-Alexander I, (who is presumably Michael of Epehesus), in S.E., 20. 28-9.

10 Simplicius in Cat. 10.2-4: “σαφῶς εἶπεν ὁ Ἀριστοτέλης, τὰ δὲ λεγόμενα καὶ λεκτὰ τὰ νοήματά ἐστιν, ὡς καὶ τοῖς Στωικοῖς ἐδόκει.”.

11 E.g.: Cicero Fin. 3.5; 4.7; Galen, Inst. Log. 4.6, or Plutarch Comm. Not. 1068D, Plat. Quaest. 1009C.

12 The same strategy is illustrated by Themistius’ semantic treatment of the other Stoic notions he mentions, the ‘tunchanon’ and the ‘ekphorikon’, which he attempts to explain by an appeal to their morphological connotations; other attempts of a similar nature are put forward in the other passages mentioned, which give the same comparative grid.

13 See Andreas Graeser (1978), “The Stoic Theory of Meaning”, in John Rist (ed.), The Stoics, Berkeley, University of California Press, p. 88f on this passage from Ammonius, as setting out the elements of differentiation of the Aristotelian and Stoic models.

14 Diogenes Laertius, Lives of the Eminent Philosophers, Dorandi, T. (ed.) (2013), Cambridge [D.L.], 7.57: “προφέρονται μὲν γὰρ αἱ φωναί, λέγεται δὲ τὰ πράγματα, ἃ δὴ καὶ λεκτὰ τυγχάνει.”

15 It is in the middle form that the verb is indeed commonly used in the sense of ‘to utter’: e.g. at Sextus Empiricus, Adversus Mathematicos in Mutschmann, H & J. Mau (eds.) (1955-61) Opera, vols. 2&3, Leipzig. [S.E. M.], M. 7. 44: “γραμματικὸς σολοικισμοῦ τιθεὶς ὑπόδειγμα σολοικισμὸν <μὲν> προφέρεται, οὐ σολοικίζει”: ‘a grammarian who is giving an example of a solecism, utters a solecism, but he is not thereby making a solecism’.

16 See John D. Denniston (1950), The Greek Particles, Oxford, OUP, on ‘δὴ’.

17 D.L. 7.57 identifies this sub-class of signifying sounds in the line immediately preceding the quoted line above, distinguishing between non-signifying utterances, like ‘blituri’, and signifying sounds.

18 D.L. 7.55: ‘phonē is the air struck by an impulse’.

19 D.L. 7.56; Ps-Plut. Plac. 902F-903A: sound, phonē is said to act, poiei, in that it moves from the speaker to the hearer.

20 S.E. M. 10.218 : “τῶν δὲ ἀσωμάτων τέσσαρα εἴδη καταριθμοῦνται ὡς λεκτὸν καὶ κενὸν καὶ τόπον καὶ χρόνον.”, ‘they enumerate four kinds of incorporeals, namely lekta and void and place and time’.

21 For the Stoics, a body is first and foremost determined by its capacity to act or be acted upon: Diogenes Laertius, Lives, 7. 56: “πᾶν γὰρ τὸ ποιοῦν σῶμά ἐστι.”, ‘because everything which acts is a body.’; Referring to the Stoics: Sextus Empiricus, Pyrrhonei Hypotyposes in Mutschmann, H & Mau, J., eds. (1954-61) Opera, vol. 1, Leipzig. [S.E., PH], 3.38: “σῶμα τοίνυν λέγουσιν εἶναί τινες ὃ οἷόν τε ποιεῖν ἢ πάσχειν.”, ‘Some [the Stoics] say that body is that which is capable of acting or being acted upon.’. Similarly at S.E., M. 9.366; see Cicero, Acad., 1. 39; Seneca, Ep. 117.2; Aristocles, apud Eusebius, PE, 15.14.1.

22 S.E. M. 8.81: “τούτου δὲ τοῦ πράγματος τὰ μέρη μὴ συνυπάρχειν, ὥστε μηδ' αὐτὸ ὑπαρκτὸν εἶναι.”, ‘the parts of this thing cannot be present together, such that neither is the thing itself able to be present’. See a parallel passage in S.E. PH 2.109.

23 The Stoics, in the relevant contexts, use specifically the verbs ‘hyparchein’ (‘to be present’ or ‘to obtain’), or ‘hyphestanai’ (‘to subsist’) to indicate a particular mode of reality, which is different from the mode of being of bodies. It is a use which constitutes an important doctrinal element regarding the status of incorporeals in ontology. See S.E. M. 1. 157 ; 8.100; 8. 262; 8.406-410 and many others for this use of ‘hyparchein’, and with reference specifically to lekta. See on the specialised use of these verbs: Victor Goldschmidt (1972), « ὑπάρχειν et ὑφεστάναι dans la philosophie stoïcienne », Revue des Études Grecques 85; Jacques Brunschwig (1988), « La théorie stoïcienne du genre suprême et l’ontologie platonicienne », in Jonathan Barnes & Mario Mignucci (eds.), Matter and Metaphysics, Naples, Bibliopolis, p.23; Michael Frede (1994), “Stoic notion of a lekton”, in Stephen Everson (ed.), Language, p. 116-117.

24 S.E. M. 8.81: “ἀλλὰ τούτου γε τὴν σημαντικὴν φωνὴν ἀμήχανόν ἐστι προφέρεσθαι διὰ τὸ οὗ τὰ μέρη μὴ συνυπάρχει, μηδὲ αὐτὸ ἐκεῖνο ὑπάρχειν”.

25 The Stoics famously defended the interdependency of the three parts of philosophy, logic, physics and ethics, with analogies meant to bring out their mutual dependencies: philosophy is thus compared to the parts of an egg whose shell stands for logic, the white for ethics, and the yolk for physics; to a living being whose bones and sinews correspond to logic, the flesh to ethics and the soul to physics, or to a garden whose encircling fence is for logic, the earth and plants for physics and the fruit for ethics, see D. L. 7.40; S.E., M. 7.17-9.

26 Seneca, Ep. 117. 30: “transcurramus sollertissimas nugas et ad illa nobis aliquam opem sunt latura properemus.”

27 Seneca, Ep. 117.12: “Sic aliud est sapientia, aliud sapere.”

28 Seneca, Ep. 117.2: “quod bonum est corpus esse, quia quod bonum est facit, quidquid facit corpus est. […] Sapientiam bonum esse dicunt; sequitur ut necesse sit illam corporalem quoque dicere.”. See also S.E. P.H. II. 80 and M. 7. 39 on wisdom’s activity (it benefits) as Seneca also explains; on this criterion for corporeality, see also above n.22.

29 D.L., 7.58: ‘a noun (prosēgoria) signifies a common quality, like a human being or a horse’, a quality which is a state of a body, see Plut. Virt.Mor. 441C; Simplicius, in Phys. 530. 10-4, and on this question see Ina Kupreeva (2003), “Qualities and Bodies: Alexander against the Stoics”, OSAP 25, p. 297-344.

30 See above n. 19.

31 This much can be agreed upon also from distinct and parallel testimonies, for example an often cited passage from S.E. M. 8.12 who distinguishes the lekton from the expressions used to signify it. Though the passage is a great source of controversy, everyone agrees at least on this distinction, by which different language speakers might fail to understand expressions, without their failure altering in any way the presence of the lekton: thus see Anthony A. Long (1971), “Language and Thought in Stoicism”, in A. A. Long (ed.), Problems in Stoicism, London, Athlone Press, p. 77; Andreas Graeser (1977), “Language, Thought and Reality in Ancient Greek Philosophy”, Dialectica 31, p. 379; Wolfgang Detel, et al. (1980) “λεκτὰ ἐλλιπῆ in der stoischen Sprachphilosophie”, in AGP 62, p. 284; Michael Frede (1994), “Stoic notion of a lekton”, in Stephen Everson (ed.), Language, p. 118; Richard Gaskin (1997) “The Stoics on Cases, Predicates and the Unity of the Proposition”, in Aristotle and After, Bulletin of the Institute of Classical Studies, Summer 68, p. 94.

32 See in particular, S.E., P.H. 3.14, and Clement, Strom. 8.9.26.4.

33 See Aristotle, Cat.4, 2a34-b6, and on this: Wolfgang R. Mann (2000), The Discovery of Things, Princeton, Princeton University Press, p. 191ff.

34 Zeno of Citium, the founder of the Stoic school, was a student of Diodorus Cronus, who is associated with what became known as the Dialectic school (D.L. 2.106-112), one of the founders of which is Clinomachus of Thurii, who is remembered as having worked “on propositions (axiōmata) and predicates (katēgorēmata)” (D.L. 2.112), which are two kinds of lekta for the Stoics (D.L. 7.63). On this work and the links with the Stoics, see Jacques Brunschwig (1995), « La théorie stoïcienne du nom propre », in Jacques Brunschwig, Études sur les philosophies hellénistiques, Paris, PUF, p. 254f, and more generally on the dialectical school and the Stoics: David N. Sedley (1977), “Diodorus Cronus and Hellenistic Philosophy”, Proceedings of the Cambridge Philological Society 203 NS 23, p. 74-120. On Zeno’s affiliation: D.L. 7.16 and 25; Suda, Adler, A. (ed.) (1928-38), 5 vols, Leipzig, under the entry for ‘Σωκράτης’, 829. 27-33 (the Dialectical school, along with the Megaric and Eristic are considered generally as Socratic schools); Ps-Galen, Hist. Phil. 7.17-8. It is, moreover, far from clear that Seneca in fact, means to highlight such a teacher-student affiliation, but rather bring out the influence of the work of older logicians in Greek Stoicism.

35 It is, for instance, this broad affiliation which underlies Cicero’s mention of the “dialectici”, i.e. broadly speaking the logicians, when he speaks of predicates, katēgorēmata, as distinct from things themselves: in Tusc. IV. 9.21: “[eae res] quae dicuntur de quodam aut quibusdam, quae κατηγορήματα, dialectici appellant.”, ‘those things, which we say about something or things, which the logicians call ‘κατηγορήματα’. It is precisely these things, of which Seneca in our Letter 117.13 will give various translations, which are in fact, kinds of lekta.

36 Seneca, Ep.117. 1: “quod Stoicis placet, sapientiam bonum esse, sapere bonum non esse.”

37 Seneca, Ep.117.3: “Quid ergo? […] non dicimus: bonum est sapere?”

38 Seneca, Ep.117.2: “Dicimus referentes ad id ex quo pendet, id est ad ipsam sapientiam”

39 Seneca, Ep. 117.12: “aliud es ager, aliud agrum habere, quindi?”

40 The grounds for this backpedalling are far from clear, and therefore all the more fragile for his critical stance: see John M. Cooper (2004), Knowledge, Nature and the Good: Essays in Ancient Philosophy, Princeton-Oxford, Princeton University Press, p. 329, and Brad Inwood (2007), Seneca - Selected Philosophical Letters, Oxford, OUP, p. 298.

41 With n. 11 above, see also Galen, Μeth. Μed. X, 155,7-8; Simplicius, In Cat.. 310.14-312.37, and more generally on this critique, Catherine Atherton (1993), The Stoics on Ambiguity, Cambridge, CUP, p. 117-119.

42 Seneca, Ep. 117.1: “quod Stoicis placet”.

43 Half of Seneca’s Letter 117 is a long and bombastic rant about the superfluous contribution of logic to real life questions: those concerned with how to live well and overcome the real difficulties we are faced with. The rhetoric and mock anger unleashed in the second half, reappraises the same two angles of attack, alternating the criticism from an emphasis on the irrelevance and emptiness of the problem, with words like ‘disputatiuncula’ and ‘inanis’, to locutions such as ‘subtilitas vanissima’, fruitless subtleties, or subtleties articulated in vain.

44 Seneca does so for notions, and corresponding concept-terms, which are important to him and which he develops, especially in the name of their expression of naturalism: for instance his elaboration of the term conciliatio, and cognate forms, in Ep.121.14ff, for the Greek Stoic notion of oikeiōsis, that of familiarisation and appropriation of the world around us which happens by nature: all rhetorical means are enlisted, from repetitions to amplification, to emphatic paraphrasis and alliterations, with the phrase ‘constitutioni conciliatur’ appearing on every line almost, all the better to institute conciliatio as a key concept-term, symbolic of a major element of doctrine. Interestingly the word for ‘reason’ (“logos”) in this context, to which oikeiōsis, in the Greek corpus, is directed (“so that what is according to nature becomes what is according to reason”, D.L. 7.86), has disappeared in Seneca’s treatment: it is the same agenda at work as in the present discussion of Stoic logic, that of mono-maniacally converging all points of doctrine onto an expression of what is natural in ethics, see on Seneca’s reshaping of the Greek Stoics’ oikeiōsis, Ada Bronowski (2016) “Rational Empiricism: the Stoics on Reason, Experience and Katalepsis”, in Logical Analysis and History of Philosophy 19, p. 176-178.

45 Seneca, Ep. 117.13: “Dico deinde: Cato ambulat. Non corpus inquit est quod nunc loquor, sed enuntiativum quiddam de corpore quod alii effatum vocant, alii enuntiatum, alii dictum”

46 Cicero’s ‘enuntiatum’ becomes canonical, presumably alongside a suggestion from Varro (‘proloquium) which we know of only from a report in Aulus Gellius (XVI. 8. 4-8=Varro Fr. 28), as both of these terms are listed in ps-Apuleius’ commentary on De interpretatione. Cicero might also, however, be the source for Seneca’s “alii effatum”, since there is one occurrence in which Cicero uses ‘effatum’ at Acad. ii.95, though with some caution, as he says what the Greeks call an axioma ‘is more or less an effatum’(“quod est quasi effatum”), which hints at Cicero’s not being quite satisfied with this translation, see on this David Londey & Carmen Johanson (1987), The Logic of Apuleius, including a complete Latin text and English translation of the Peri Hermeneias of Apuleius of Madaura, Philosophia Antiqua, Leiden, Brill.

47 Seneca thus echoes one etymological explanation for the Greek term: kategorema: “κατὰ τινος ἀγορευόμενον” recorded in D.L. 7. 64, “what is said about something”, which is what Cicero reports as well in Tusc. iv.9.28 (cf. fn.35).

48 E.g., Gabriel Nuchelmans (1973), Theories of the Proposition. Ancient and Medieval conceptions of the bearers of truth and falsity, Amsterdam, North-Holland Publishing, p. 52-72; Andreas Graeser (1978),‘The Stoic Theory of Meaning’, in John Rist (ed.), The Stoics, Berkeley, University of California Press, p. 90f; Margherita Isnardi-Parente (2005), « La notion d’incorporel chez les Stoiciens », in Gilbert Romeyer-Dherbey (dir.), Jean-Baptiste Gourinat (éd.), Les Stoiciens, Paris, Vrin, p. 180.

49 E.g., Michele Alessandrelli (2013), Il Problema del Lekton nello Stoicismo Antico: origine et statuto di una nozione controversa, Florence, Olschiki.p. 146ff.

50 E.g., John Rist (2006), “Zeno and the Origins of Stoic logic”, 1978¹, in Jacques Brunschwig (éd.) Les Stoiciens et leur logique, 1978¹, revised edition, Paris: Vrin, p.25-6 ; Jean-Joël Duhot (1989), La conception stoïcienne de la causalité, Paris, Vrin, p. 48-49 ; Jean-Baptiste Gourinat (2009), “The Stoics on Matter and Prime Matter: corporealism and the Imprint of Plato’s Timaeus”, in Roccardo Salles (ed.), God and Cosmos in Stoicism, Oxford, OUP, p. 56.

51 Stob. Ecl. 2.7.6f.

52 S.E. M. 7. 38-40.

53 On lekta as the objects of impulse and desire, see Stob. Ecl. 2.7.9a; on lekta as what is true or false: S.E. M. 8.11-3; as the content of thought: D.L. 7. 63.

Haut de page

Pour citer cet article

Référence électronique

Ada Bronowski, « What is wrong with lekta? Ancient critics of Stoic logic and language », Methodos [En ligne], 19 | 2019, mis en ligne le 07 février 2019, consulté le 21 mars 2019. URL : http://journals.openedition.org/methodos/5276 ; DOI : 10.4000/methodos.5276

Haut de page

Auteur

Ada Bronowski

New College of Humanities, London

Haut de page

Droits d’auteur

Licence Creative Commons
Les contenus de la revue Methodos sont mis à disposition selon les termes de la Licence Creative Commons Attribution - Pas d'Utilisation Commerciale - Pas de Modification 4.0 International.

Haut de page
  • Logo DOAJ - Directory of Open Access Journals
  • Logo UMR Savoirs, Textes, Langage
  • Logo CNRS - INSHS
  • Logo Université de Lille
  • OpenEdition Journals