¹ SE M VII 236, 230 (SVF 1 58), in which the definition is explicitly attributed to Zeno; cf. also DL VII 46, 50; Cic. Luc. 18, 77; SE M VII 248, 400.

² Cic. Luc. 77 (SVF I 59).

³ Cf. Anon, . in Plat. Theat., xi 27–31Google Scholar (ed. H. Diels, W. Schubart, p. 9). I follow the text kindly given to me by F. Decleva Caizzi and D. Sedley.

⁴ SE M VII 373 (SVF II 56), ε⋯ γ⋯ρ κηρο⋯ τρ⋯πον ⋯ ψɛχ⋯ φανταστικ⋯σ π⋯σχοɛσα. Dl VII 45 (SVF II 53). τ⋯ν δ⋯ φαντασ⋯α εἷναι τ⋯πωσιν ⋯ν ψɛχῇ, το⋯ ⋯ν⋯ματος οἰκε⋯ως μετενηνεγμ⋯νον ⋯π⋯ τῶν ⋯ν τῇ κηρῇ ὑπ⋯ το⋯ δακτɛλ⋯οɛ γινομ⋯νων.

⁵ Ar. Did. ap. Euseb, . PE XV 20, 2Google Scholar (SVF I 141): Z⋯νων τ⋯ν ψɛχ⋯ν λ⋯γει αἰσθητικ⋯ ⋯ναθɛμ⋯ασιν [Thus Wellman followed by Diels and Arnim: αἴσθησιν ἢ ⋯ναθɛμ⋯ασιν codd.] 'Hρακλε⋯τος. βοɛλ⋯εnu;οσ λ⋯ρ ⋯μφαν⋯σαι ὅτι αἱ ψɛχα⋯ ⋯ναθɛιώμεναι νοερα⋯ ⋯εἰ γἱνονται, εἴκασεν αὐτ⋯σ τοῖς ποταμοῖς. The text does not allow us to be sure whether the quotation ⋯ναθɛμιώμεναι – γ⋯νονται should be attributed to Heraclitus or to Zeno. The opinions of scholars on this point differ; for possible corrections of the text, see Conche, M., Héraclite. Fragments (Paris, 1986), pp. 452–4Google Scholar, which, however, does not include the one proposed by Festa, N., I frammenti degli Stoici antichi, 11 (Bari, 1935), pp. 119–20Google Scholar. It seems to me, however, that a comparison with the Plutarch passage (see note 6) supports the attribution to Zeno. This opinion, albeit for other reasons, is shared by Barnes, J., ‘Robinson's Heraclitus’, Apeiron 13 (1988), 102–3Google Scholar. Indeed Plutarch objects that the exhalation should always be new, and, therefore, not able to retain the imprints, καιν⋯ν ⋯ε⋯ ποιεῖ τ⋯ν ⋯ναθɛμ⋯ασιν. This would confirm the reading νεαρα⋯ proposed by Meerwaldt, J. D., ‘Cleanthea’ I, Mnemosyne 4 (1951), 40–69, p. 54CrossRefGoogle Scholar, instead of νοερα⋯ in the Arius Didymus passage.

⁶ Comm. not. 1084f–1085a.

⁷ See also DL VII 157 (SVF I 135).

⁸ See SE M VII 228, 400 (SVF I 484).

⁹ Cf. Cic. Varro 40: ‘de sensibus… quos iunctos esse censuit e quadam quasi impulsione oblata extrinsecus.’

¹⁰ SE M VII 231; cf. DL VII 50.

¹¹ Cf. Cic. defato 43: ‘sic visum obiectum imprimet illud quidem et quasi signabit in animo suam speciem.’ This confirms that Chrysippus is defending Zeno's doctrine, as the argument ii defato 40 also suggests; on this subject, cf. my ‘Le cause antecedent in Cicerone De fato 40’, in Matter and Metaphysics, ed. Barnes, J., Mignucci, M. (Naples, 1989), 399–424Google Scholar.

¹² Cic. Varro 41 (SVF 1 60). Cf. SE M VII 232–5, in which the Stoics clearly distinguish between an affection such as an itching hand, or a scratch in the hand, and a presentation.

¹³ DL VII 46 (SVF II 53); cf. Cic. Luc. 77 (SVF I 59): ‘ex eo quod esset sicut esset impressum et signatum et effictum.’

¹⁴ Cic. Varro 41: ‘id autem visum cum ipsum per se cerneretur.’ The meaning of Zeno's statement is made clear by SE M VII 257: “This presentation being self-evident and striking (⋯ναργ⋯σ οὖσα κα⋯ πληκτικ⋯), all but seizes us by the hair, they say, and pulls us to assent, needing nothing else to achieve this effect or to establish its difference from other presentations.’ The passage refers to the younger Stoics, but since the definition of the cognitive presentation given at 247 is that of Zeno and at 253 it is said that older Stoics (⋯ρχαι⋯τεροι) invoked it as criterion of truth, we could reasonably assume that all the Stoics shared, on this subject, Zeno's view.

¹⁵ Bonhoeffer, A., Epictet und die Stoa (Stuttgart, 1890), p. 163Google Scholar, believes that if καταληπτ⋯ν has to be given a passive sense ‘unter visum aber nicht sowohl die Vorstellung, als vielmehr das Vorgestellte Object selbst zu verstehen.’

¹⁶ SE M V II 409 (SVF II 85).

17 DL VII 51: ἔτι α⋯ μ⋯ν εἰσι λογικα⋯, αἱ δ⋯ ἄλογοι. λογικα⋯ μ⋯ν αἱ τῶν λογικ⋯ν ζᾠων.…αἱ μ⋯ν οὖν λογικα⋯ νο⋯σεις εἰσ⋯ν.

¹⁸ Cf. SE M viii 397 (SVF II 91); see also Arcesilaus in Plut, . Col. 1122b–dGoogle Scholar, which refers to Zeno's position.

¹⁹ Cf. Cic. Luc. 145; Numen, ap. Euseb, . PE xiv 6, 13Google Scholar. Moreover the ambiguity of Zeno's conception of presentation is to be found even in the adjective καταληπτικ⋯ς, which in Greek can be both active and passive in sense; on this subject, much discussed by critics, see Bréhier, E., Chrysippe et l'ancien Stoïcisme ² (Paris, 1951), pp. 88ffGoogle Scholar; Sandbach, F. H., ‘Phantasia Kataleptike’, in Problems in Stoicism, ed. Long, A. A. (London, 1971), 9–21, p. 10Google Scholar.

²⁰ Cf. Frede, M., ‘Stoics and Skeptics on Clear and Distinct Impressions’, in The Skeptical Tradition, ed. Burnyeat, M. (Berkeley, Los Angeles, London, 1983), 65–93, p. 68Google Scholar.

²¹ Cf. Cic.Luc. 145 (SVF I 66).

²² Cic. Varro 40 (SVF I 60): ‘sed ad haec quae visa sunt et quasi accepta sensibus, adsensionem adiungit animorum, quam esse volt in nobis positam et voluntanam.

²³ Cf. SE M VII 237. The Stoics' opponents use τ⋯πωσις and ἔτεοἰωσις indifferently to indicate the presentation, showing that there was no difference for them between Zeno's position and Chrysippus' interpretation (see 232). In fact, that this was in part legitimate, is confirmed by the definition of the cognitive presentation given by Zeno and never abandoned by the later Stoics, in which the characteristics of the presentation as an imprint are emphasized.

²⁴ Cf. SE M VII 239.

²⁵ Ibid. 259. Both Cic. Varro 42, and Aristo's polemic against the Academics in D L VII 163 (SVFi 346) might be adduced to confirm that this was Zeno's position. Moreover the Epicureans also accuse the Academics of ablepsia; on this subject, see my Opinione e scienza. II dibattito tra Sloici e Accademici nel terzo e nel secondo secolo a. C. (Naples, 1986), pp. 187–9Google Scholar.

²⁶ Cf. SE M VII 152.

²⁷ Different interpretations and different translations have been given of the expression ⋯μετ⋯7pi;τωτος ὑ7gamma;οɛ in the definition of knowledge, depending whether ‘reason’, ‘argument’, or ‘reasoning’ is used to represent λ⋯7gamma;ος. I prefer ‘argument’‘, because it allows us to distinguish the κατ⋯ληφ7iots;ς from the ⋯πιτ⋯μη For while reason represents the qualifying element in both (even in cognition the assent must be a strong assent if it is to be criterion of truth), unchangeability by arguments is peculiar to that knowledge which only the wise man possesses; see my review of Long, A. A., Sedley, D. N., The Hellenistic Philosophers, Elenchos 10 (1989), 231–46, pp. 239–41Google Scholar.

²⁸ Cf. DL VII 53 (SVF II 87); Sen. ep. 120,4, 5, 8; SE M IX 393–5; xi 250–1; VIII 58–60; III 40–2.

²⁹ Aet. plac. IV 11(SVF II 83).

³⁰ DL VII 51 (SPT II 83).

³¹ ibid. 52 (SVF II 84).

³² Cf. SE M VIII 56–8.

³³ ibid. XI 250.

³⁴ Cic. Varro 42: ‘quia nihil quod cadere in earn posset relinqueret quodque natura quasi nonnam scientiae et principium sui dedisset.’

₃₅ DL Vn 51: λογικα⋯ μ⋯ν αἱ τ⋯7nu; λογικ⋯ν ζῴω7nu;, ἂλογοι δ⋯ τ⋯ν ⋯λ⋯γν

³⁶ The metaphor is present in Cic. Luc. 38, where Zeno is not explicitly quoted, but Aicesilaus' polemical argument in Plut. Col. 1122c, τ⋯ δ ⋯7rho;μητικ⋯ν - γε⋯σεων γινο7mu;⋯νην attests the Zenonian origin of the comparison. In fact Zeno frequently made use of the scales and weights comparison, cf. Epict. i 17, 10 (SVFi 48); Stob. II 22, 12 (SVF I 49).

³⁷ Cic. Varro 41.

³⁸ Cf. Cic. Varro 40 and de fato 39–43, where Chrysippus answers an objection put forward by opponents against the Stoic claim, probably Zeno's, that assent is in man's power.

³⁹ On this subject, cf. my Opinione e scienza (n. 25 above), pp. 34ff.

⁴⁰ Cf. Anon, , in Theaet. II 52–III 15Google Scholar, which treats the first part of the Theaetetus as if it were a refutation of Stoic epistemology.

⁴¹ At Theaet. 209c6–7 Socrates expresses the view that whoever makes a judgement about Theaetetus has to possess an imprinted memory trace of Theaetetus' distinguishing marks. This could be taken as Plato's upholding of the wax tablet model of knowledge. But, firstly, Plato has already rejected that model of knowledge, because it cannot account for mathematical errors; and, secondly, at 208b12–e6 Plato is no t suggesting that a true judgement concerning Theaetetus amounts to a knowledge, because, as McDowell, J., Plato, Theaetetus (Oxford, 1973), p. 256Google Scholar, points out, ‘any judgement concerning a particular thing requires that its maker have a thought, an d hence the ability to produce a form of words, which distinguish that thing from everything else’.

⁴² Chrysippus will not maintain the definition of presentation as an imprint, because it does not allow either memory or the τ⋯χναι to take place; cf. SE M VII 373, and Mansfeld, J., ‘Intuitionism and Formalism: Zeno's Definition of Geometry in a Fragment of L. Calvenus Taurus’, Phronesis 28 (1983), 59–74, pp. 67ff.CrossRefGoogle Scholar

⁴³ 192b6–7 (trans. McDowell).

⁴⁴ Cf. Fine, G., ‘Plato on Perception: A Reply to Professor Turnbull’, Oxford Studies in Ancient Philosophy, Suppl. Vol. 1988, pp. 15–28Google Scholar.

⁴⁵ Burnyeat, M., ‘Plato on the Grammar of Perceiving’, CQ 26 (1976), 29–51, p. 36CrossRefGoogle Scholar, observes: ‘take away judgement and all that is left to perception is to be an unarticulated encounter with sensible things.’

⁴⁶ Cf. G. Fine, art. cit. (n. 44).

⁴⁷ That the Greeks found it difficult to distinguish these two senses of the verb ’to be’ has been pointed out by many critics, incuding Kahn, C., ‘Why Existence does not Emerge as a Distinct Concept in Greek Philosophy’, AGP 58 (1976), 323–34Google Scholar; Cooper, J., ‘Plato on Sense Perception and Knowledge (Theaet. 184–6)’, Phronesis 15 (1970), 123–46, pp. 123–4CrossRefGoogle Scholar, Burnyeat, , art. cit. (n.45), p. 45Google Scholar, excludes that οὐσια could ever have the meaning of existence in this section, because ‘no defence is offered of the idea that knowledge, let alone truth, presupposes a grasp of being in the specific sense of existence, reality or essence.’ Nevertheless I believe that Plato ascribes to ‘being’ the existential meaning as well, when he prevents sense perception from formulating any type of judgement and from expressing itself in an articulate content.

⁴⁸ Indeed neither Arnim, nor Hülser include SE M VII 154 among their Stoic testimonia, because they consider the position expressed there, as reflecting Arcesilaus' criticism not Zeno's position.

⁴⁹ Cic.Varro 42.

⁵⁰ Luc. 77–8; cf. also de nat. deor. I 70 (SVF l 63).

⁵¹ Cf. SE M VIII 9; DL x 32; contra Long and Sedley, , op. cit. (n. 27), i. 85Google Scholar; Striker, G., ’Epicurus on the Truth of Sense Impressions’, AGP 59 (1977), 125–42Google Scholar.

⁵² Cf. Frede, M., Die stoische Logik (Göttingen, 1974), p. 41Google Scholar. Moreover this use shows the permanence of a linguistic practice which has its origin in the identity of language, thought and reality, held by earlier thinkers, as Calogero, G., Storia delta logic a antica, I (Bari, 1967)Google Scholar, has well pointed out, and which can still be found in Platonic expressions such as ⋯ντως ⋯ντα and ⋯ληθ⋯ς ⋯ντα.

⁵³ In my view it is not only Epicurus who used the attributes ‘true’ and ‘false’ in this sense, as Long and Sedley, , op. cit. (n. 27), i. 85Google Scholar believe, but Zeno as well.

⁵⁴ That Plato uses αἴσθησις where the Stoics use øαντασ⋯α is not a significant difference. Besides, the Stoics ascribe to αἴσθησις several meanings, cf. DL VII 52 (SVF II 71).

⁵⁵ ‘Observations on Perception in Plato's Later Dialogues’, in Essays in Ancient Philosophy (Minneapolis, 1987), 3–8Google Scholar; cf. also Fine, art. cit. (n. 44), Burnyeat, art. cit. (n. 45), Cooper, art. cit. (n. 47), who agree in emphasising the different conception of sense perception expressed by Plato in the Republic.

⁵⁶ Cf. Aristocles ap. Euseb. PE (fr. 53 Caizzi).

⁵⁷ 189e–190a; cf. also Soph. 236eff.; Phlb. 38c.

⁵⁸ Zeno names κατ⋯ληΨισ as criterion of truth in SE M VII 152–3, and not the cognitive presentation, but we can assume that he considered the latter the foundation of the criterion because of its absolute truthfulness and reliability, as in fact SE M VII 253 reports for the older Stoics in general.

⁵⁹ Cic. Varro 41. Scholars have discussed this passage a great deal, without coming to a satisfactory explanation. My point of view is closest to that of Frede, M., art. cit. (n. 20), p. 76Google Scholar, but I disagree with him about the propositional aspect of the presentation; see also Long, and Sedley, , op. cit. (n. 27), ii. 255Google Scholar.

⁶⁰ Cic. Varro 42. If the cognitive presentation is the foundation of cognition, it has to be αἰαθητικ⋯ (cf. Zeno, ibid., ‘comprehensio facta sensibus’). Besides, the distinction between αἰαθητι⋯ and οὐκ αἰαθητια⋯ presentations in DL VII 52 presupposes that the latter have their origin in sensory presentations as well. In fact only the presentations of the incorporeals and the other things acquired by reason are examples of οὐκ αἰαθητια⋯ presentations: the presentations of the incorporeals presuppose a mimetic activity of the subject in representing them, which cannot do without previous sensory presentations, as is suggested by the example of the trainer standing at a distance and moving to a certain drill, to provide a model for the boy (see SE M vm 409 = SVF II 85). Therefore they are conceived by a process of abstraction from bodily entities; cf. Long, , Sedley, , op. cit. (n. 27), i. 163Google Scholar, who refer to DL vn 53. Moreover the definition of the cognitive presentation emphasizes the fact that its origin lies in an existing object, ⋯π⋯ το⋯ ὑπ⋯ρϰ鎿ντοσ, but we cannot say of the incorporeals that they exist. As far as concerns ‘the other things acquired by reason’, we learn from DL VII 52 that they are ‘the conclusions reached through demonstration’. Therefore they represent a point of arrival following a more elaborate mental process, which starts from sensory presentations as well. Moreover, if the δι⋯νοια could do without αἴσθησις, the Stoics would not be able to state that all our knowledge comes from sense perception (SE M VIII 56), or that the criterion of truth is the øαντασ⋯α, which decides ⋯ ⋯λ⋯θεια τ⋯ν πραγμ⋯των (DL vu 49; cf. also 46).

⁶¹ Cf. Aetius IV, 12, 1–5 (SVF II 54).

⁶² As Long, and Sedley, , op. cit. (n. 27), ii. 240Google Scholar, point out, ‘since the Stoics regarded all perception as taking place in the ⋯γεμ¿νικ⋯ν, the distinction between ‘sensibus’ and ‘animo’ is potentially misleading.’

⁶³ ibid. i.240.

⁶⁴ Cf. Aetius rv 23, 1 (SVF II 854).

⁶⁵ Cic. defato 43 (SVF II 974); SE M VII 160.

⁶⁶ The only passage that could suggest a relationship between assent and proposition is SE M VIII 70. But see below, p. 446 with note 73.

⁶⁷ This definition is repeated twice in SE M VII 153—4.

⁶⁸ Stob. II 88, 1 (SVF II 171); Wachsmuth's correction, accepted by Long and Sedley, does not seem necessary to me, since συγκατ⋯θεσις is constructed both with the genitive and with the dative. This wavering cannot be attributed to the fact that it depends on different contexts, because we find it within the same context, as Sextus M VII 154–7 shows; cf. the remarks by Inwood, B., Ethics and Human Action in Early Stoicism (Oxford, 1985), pp. 101, 287 n. 271Google Scholar.

⁶⁹ Cf. Giusta, M., I dossografi di etica (Torino, 1967), ii. 226Google Scholar, who believes that the lacuna is broader than Wachsmuth supposed.

⁷⁰ Stob. II 86, 17(SVF III 169).

⁷¹ Plut. de stoic, rep. 1037f (SV III 175).

⁷² Stob. II 86, 17, τα⋯της δ' ⋯ν εἴδει θεωρεῖσθαι τ⋯ς λογικοῖς γιγνομ⋯νην ⋯ρμ⋯ν κα⋯ τ⋯ν ⋯ν τοῖς λογικοῖς γιγνομ⋯νην ⋯ρμ⋯ν κα⋯ τ⋯ν ⋯ν τοῖς ⋯λ⋯γοις. οὐ κατωνομασμ⋯ναι δ' εἰσ⋯ν.

⁷³ SE M VIII 70 (SVF II 187Google Scholar). This passage has been interpreted in different ways, according to whether the word λ⋯γος is translated ‘reason’ or ‘language’. In my opinion the translation ‘reason’ is problematic. For if it were reason which makes a presentation rational, it would not be said that ‘a rational presentation is one in which it is possible to establish by reason the presented object’ (see Bury's translation). This means that reason may or may not establish the content of presentation. But in the latter case it would be difficult to call a presentation rational. However the translation ‘reason’ has been widely upheld by scholars since Zeller. It supports the thesis that a λεκτ⋯ν is the content of a thought independent of language. For a conciliation of this view with the one according to which λεκτ⋯ are the meanings of words, see Mignucci, M., π significato delta logica stoica (Bologna, 1967), pp. 88–96Google Scholar. It could be objected against the interpretation which maintains that a λεκτ⋯ν is that which is said, that the effects of causes are also λεκτ⋯. But the Stoics conceived of the effects as predicates which are true of the body affected. Hence what a cause produces is not something existent, but rather a semantic item signifying a process undergone by the body. Thus the Stoic theory of cause and effect not only does not help to clarify the position of λεκτ⋯, but complicates the problem considerably.

⁷⁴ Cf. Striker, G., κριτ⋯ριον τ⋯ς ⋯ληθε⋯ας (Nachrichten der Akademie der Wissenschaften in Gottingen, Phil. hist. Kl., 1974, pp. 47–110)Google Scholar, p. 94 n. 1.

⁷⁵ Mates, Contra B., Stoic Logic (Berkeley, Los Angeles, 1953), p. 22Google Scholar, according to whom the λεκτ⋯ν ‘is what might be called “the objective content”, τ⋯ φαντασθ⋯ν, of the presentation.’

⁷⁶ M VIII 11–12(SVF II 166).

⁷⁷ For the translation of τɛγχ⋯νον by ‘the name-bearer’, cf. Long, Sedley, , op. cit. (n. 27), ii. 197, i.201Google Scholar.

⁷⁸ SE M VIII 70 (SVF II 187); DL VII 63.

⁷⁹ III 7, 313, p. 220, 42, 17ff De Lacy.

⁸⁰ Cf. Stob. II 64, 18.

⁸¹ Cf. the examples of Admetus and Menelaus in SE M VII 254–5.

⁸² Cf. SVF II 429, 494.

⁸³ Cic, . Varro 41Google Scholar. Self-evidence provokes in the soul the same effect as the one of weights placed in the balance; cf. Plut, . col. 1122c.Google Scholar

⁸⁴ Die Fragments zur Dialektik der Stoiker (Stuttgart, 1987), Band 2. p. 832Google Scholar.

⁸⁵ Cf. Clem. Alex, (SVF II 488)Google Scholar, and Frede, M., op. cit. (n. 52), p. 15 n. 15Google Scholar.

⁸⁶ SE M VIII 85–6: φασ⋯ γ⋯ρ ⋯ληθ⋯ς ⋯ξ⋯ωμα δ ὑπ⋯ρχει τε κα⋯ τινι… ⋯ρωτώμενοι δ⋯, τ⋯ ⋯στι τ⋯ ὑπ⋯ρχον, λ⋯γοɛσι τ⋯ καταληπτικ⋯ν κινο⋯ν φαντασ⋯αν.