The emergence of philosophical interest in cognition

LES RAN] Va a LÀ GN oro StuoieS IN ANC fm SCH THE EMERGENCE OF PHILOSOPHICAL INTEREST IN COGNITION J. H. LESHER 1. Introduction In the famous ‘autobiographical’ passage of Plato’s Phaedo Socrates mentions in passing some of the questions his predecessors used to ask concerning the physical substances and processes involved in thought: ‘Is it blood that we think with, or air or fire? Or is thought due to something else, namely the brain’s providing our senses of hearing, sight, and smell, which give rise to memory and opinion, and ultimately, when memory and opinion have acquired stability, to knowledge?’ (964-B). A few Presocratic fragments' and the summaries of early views of the soul given by Aristotle and Theophrastus’? confirm the accuracy of Socrates'' (or Plato's) listing. And in so far as Presocratic interests appear to have been focused on the physical elements and processes © J. H. Lesher 1994 An earlier version of this paper was read at the 1992 Ancient Philosophy Colloquium at Princeton University. I am indebted to the commentator on that occasion, Dominic Scott, for his very helpful comments. Murat Aydede, Rachel Barney, Patricia Curd, Brad Inwood, Charles Kahn, Andrè Laks, Mitsi Lee, Gary Meltzer, Joergen Meyer, Michael Rohr, and Eva Stehle have also made helpful suggestions and criticisms. ' Cf. Empedocles frr. 105, 106 (on blood and knowledge); Anaximenes fr. 2 and Diogenes of Apollonia fr. 5 (on intelligence and air); and Heraclitus fr. 118 é E A i N (‘the dry soul is a flash of light, wisest and best’). Alcmaeon is often identified as the author of the view that the brain plays a role in knowledge (cf. A 11 DK), but fr. 16 of Parmenides also links the thoughts of men with conditions in each of their ‘limbs’ or organs, and Censorinus (A 108 DK) attributes to Anaxagoras the specific thesis ‘cerebrum unde omnes sunt sensus’. The idea of a progression from sensation to memory to opinion to knowledge might also have been Alemaeon's, but if mortals do actually gain knowledge it is not clear what we are to make of the contrast drawn in his fr. 1 between divine certainty and mortal conjecture. ? Arist. Metaph. T's; De anima 1. 2, 3. 3; Theophrastus, De sens. 3, 4, 10, 25, dtaebsast

attendant to thought, the philosophical importance But in what follows of their 3 I argue that both of these views are untenable: there was a greater degree of philosophical interest in accounts has often been discounted.’ But in this century several scholars have claimed to find evicognition among the Presocratics than has often been supposed, dence in the Presocratics' teachings of an emerging ‘intellectual’ but the view that there was an identifiable process of evolution in sense of the main terms of the Greek vocabulary of knowthe meaning of the key terms of the Greek cognitive vocabuledge—yiyvwaxw, youn, yrw@ats, vóos, voéw, ofda, and ouvému. Since, lary—as a set of (Homeric) perception-orientated expressions in their view, at least some of the Presocratic thinkers had articugradually acquired a more intellectual cast—seems to me at odds lated a distinction between perceptual knowledge and a kind of with a sizeable body of evidence. In Section 2 I explain what I knowledge acquired through thought or reflection, it would be a regard as the main difficulties facing the ‘developmental’ view. In mistake to suppose that philosophical interest in the concepts of Sections 3 and 4 I offer a different line of argument in support thought, perception, and knowledge emerged only as a conseof the thesis that both Heraclitus and Parmenides countenanced a quence of the sceptical challenges raised by the sophists of midform of knowledge gained not simply (if at all) through sensefifth-century Athens.* perception, but also (or perhaps instead) through one or more processes of reflection. But, in my view, their accounts were characterized more by a continuity in terminology and paradigms ’ Cf. D. W. Hamlyn, Sensation and Perception (London, 1961), 8: ‘the interests of these [Presocratic] philosophers were purely physiological or physical in characfor knowledge than by any semantic novelties. ter. ‘They made no attempt to philosophize about the nature of the concept of sense-perception, nor in general did they raise epistemological issues.’ The classic account of the Presocratics as material psychologists is J. I. Beare’s Greek Theories 2. The developmental view of Elementary Cognition (Oxford, 1906). Beare conceded that one can find the occasional non-materialist-orientated comment in early Greek philosophy, but vowed ‘to have as little as possible to do’ with them since they were ‘really aimed at little more than the clearing up of common ideas or words. Thus Plato's We can, as a consequence of overlapping argumentation, identify Theaetetus is largely occupied with an endeavour to determine the meaning of three main theses within the view that early Greek expressions émornpy or knowledge’ (p. 6). for knowledge show a pattern of development away from an * The main proponents: Bruno Snell, Die Ausdriicke fiir den Begriff des Wissens in der vorplatonischen Philosophie (Philologische Untersuchungen, 29; original conception of knowledge as virtually identical with sense- Berlin, 1924); see also his Discovery of the Mind: The Greek Origins of European perception towards one in which knowing had little or nothing to Thought (New York and Evanston, 1960), an English trans. by T. G. Rosenmeyer do with perceiving. of the 2nd edition of Die Entdeckung des Geistes (Hamburg, 1948), with a chapter 1, Etymologically, and in the earliest period of Greek literature added on ‘Human Knowledge and Divine Knowledge’; and ‘Wie die Griechen known to us, many of the terms of the Greek cognitive vocabulary lernten, was geistige Tätigkeit ist’, in Journal of Hellenic Studies, 93 (1973), 172-84, repr. in Der Weg zum Denken und zur Wahrheit (Hypomnemata, 57; Gôttingen, 1978), 21-43 (subsequent citations refer to pages of the reprint); Kurt von Fritz, ‘Nois, voeiv, and their Derivatives in Pre-Socratic Philosophy (Excluding Anaxagoras)’, Classical Philology, 40 (104s), 2233-42. ibid. 41 (1046), tose. repro ADD HWeatrEea were linked directly with the operation of the faculties of sense, especially the sense of sight; as a consequence, during this early period there was little if any awareness of the possibility Mourelans cat) Dir Braves. 4 Collection af Onitival Bouave Garden Cuy. NY 1974), 23-88 (subsequent citations refer to pages of the reprint) and Herman Fränkel, Dichtung und Philosophie des frühen Grechentums (Munich, 1962), English trans. by M. Hadas and J. Willis, Early Greek Poetry and Philosophy [EGPP] (New York and London, 1973), and "Nenophanesstudien‘, Hermes, 60 (1925), 174-02, repr. in his Were und Formen frakeniechischen Denkens (Munich, 1960) (a portion of this study was translated into English by Cosgrove and Mourelatos and included in the latter’s The Pre-Socratics as “Nenophanes' Empiricism and his Critique of Knowledge’, pp. 118-31). Many of the views advanced in these studies were subsequently incorporated in accounts given by Guthrie, Kahn, Mourelatos, and Robinson, to mention only a few. * In *Noús, wei’ (pp. 36, 61) von Fritz indicates that he adopted the accounts of pardaro and yıyrwarw given by Snell in Die Ausdrücke. In his Discovery (p. 313 n.) and ‘Wie die Griechen lernten’ (p. 41) Snell cites with approval the account of vos and roéw given by von Fritz; in ‘Wie die Griechen lernten' (p. 22n.) Snell also cites in concurrence the results of the earlier studies by Frankel. In his ‘Xenophanesstudien' Frankel states that Snell incorporated his interpretation of Xenophanes’ fr. 34 in Die Ausdrücke (Mourelatos, The Pre-Socratics, 123) and alludes to Snell's account of oída and toropin (p. 124n.); and in EGPP Frankel cites both Snell’s and von Fritz’s studies. As we shall see, there were also some important differences, even the occasional inconsistency, among their accounts.

that knowledge might be acquired not immediately in senseperception, but with an expenditure of effort and through some process of reflection (in short, during this early period ‘knowing’ was essentially just a matter of ‘having perceived’). 2. According to von Fritz and Frankel, in the teachings of Heraclitus and Parmenides these (still perception-orientated) terms for knowledge acquired the association with reflection and logical inference that would remain characteristic of their meaning in later periods. The ‘immediate’ or ‘sudden-intuitive’ quality so characteristic of voos and voeiv in Homer, for example, was still evident in the voos and voei of the earliest philosophers—the element of deductive reasoning or inference being even less characteristic of Xenophanes’ and Heraclitus’ use of these terms than it was of Homer's.* With Xenophanes the idea emerged that vóos was an exceptional quality possessed by only a few and, in a related development, it became the business of vöos to grasp the larger law governing the cosmos.’ While this represented a new subject-matter, the function of vdos in Heraclitus was still closely related to the basic Homeric meaning (‘realizing the meaning of a situation’); and gaining insight into the cosmos was still a matter of realizing through perception. What was new in Heraclitus was the idea that through our contact with the natural world we could gain an insight of vastly larger scope and significance.* Similarly, for Parmenides voos still enjoyed an intuitive relationship with reality (although error might occur when mortals mistakenly analysed the nature of ro é6v). As a result of these developments, both an insight into the unity of the cosmos and a capacity for logical inference became characteristic of both voos and voeiv. Yet the earlier meaning of voos—as ‘that in us by which we enjoy a direct view of reality —did not completely disappear. 3. In the last of his several studies on this topic Snell maintained that a ‘higher’ or ‘intellectual’ way of knowing emerged at a certain point in ancient Greek thought as a consequence of a newly developed appreciation of the value of seeking and enquiring.” In Homer, he argued, the main verbs for knowing and * Von Fritz, 'Noús, voeiv’, 26, 33, 40-1 ” Ibid. 34-7. * Ibid. 38-40. ” In ‘Wie die Griechen lernten’ Snell states his objective as ‘zeigen, wie die nachhomerische Zeit den Gedanken entwickelt hat, daf man suchen und forschen rı=n Philosophical Interest in Cognition 5 understanding—yiyvwoxw, voéw, oída, ouvimu—had shown little or no connection with reflection, and no awareness of a need for intellectual effort. These elements appear for the first time when Archilochus urges his Ovyós to ‘know what rhythm holds man in its sway’ (fragment 67a Diehl) and Solon describes as ‘the most difficult knowledge conceiving of the unseen measure that alone holds the limits of all things’ (fragment 16 Diehl). The process of evolution comes to term in a philosophical setting when Heraclitus articulates the idea of a Adyos grasped through search and reflection, and contrasts (in his fragment 40) this insight with the knowledge of accumulated experience. Appreciation of the value of enquiry was a distinguishing feature of Ionian thought generally,'” but in Heraclitus (according to Snell, but not von Fritz) voos began to operate as an organ of thought. However (as von Fritz had also maintained), logical reasoning became one of the characteristic functions of vóos as a consequence of Parmenides' use of the techniques of logical deduction.'' The developmental view, it should be said, contained a few solid kernels of truth. The Greek knowledge-verb olda is in fact a second perfect formed from the root perceptual verb *eiöw ‘I see’ (thus ofa, translated as ‘I know’, means literally ‘I have seen’); the present of the Greek y:yvwoxw was probably derived from an older and simpler aorist form éyvw, with its root element, yvw-, in all likelihood meaning ‘note’ or ‘notice’; and the Greek ovvimu began its life meaning merely ‘coming or bringing together’ in a physical way, only later coming to designate intellectual comprehension and ‘concurrence’. It would also clearly be a mistake to read the earliest uses of voéw and voeiv as though informed by the muß, um zu höherem Wissen zu kommen’ (pp. 20-1). Cf. Discovery, 139 ff. In his IA “si earlier Die Ausdriicke (p. 28) he had explained the ‘intellectual’ or ‘philosophical’ sense of ‘know’ (‘y:yvwoxew ... in seinem prägnanten, philosophischen Sinne') as merely an extension of the act of synthetic unification that constituted simple sense-perception. This argument does not appear in his later studies, which is perhaps just as well. The essential (and therefore constant) synthetic character of the perceptual act could in itself hardly account for the onset of a new way of thinking. 1 Discovery, 139 ff. !! In Discovery, 147 ff., Snell also ascribes to Parmenides the idea that men acquire knowledge of the truth through divine revelation, and cites various passages in Parmenides’ fr. 1 to support that claim. I argue against this view in sect. 4 below.

7 deductivist conception of knowledge later championed by Parmenides. Moreover, in ancient Greek, as in other languages ancient and modern, what people commonly claimed they knew was one or more features of the immediate environment disclosed to them in sense-perception, most notably through their sense of sight. As Aristotle explained (Metaph. 980°), we hold the sense of sight in high regard precisely because it ‘reveals many distinctions and phanes), that vóos was a special capacity possessed by a select few, would not, moreover, have been entirely without precedent. most enables us to know’. When Penelope asks Odysseus how he will learn if she excels all But in many other respects the use of early Greek expressions for knowledge fails to conform to the terms of the developmental vóos not the ‘mind’ all persons would have, but that ‘wisdom’ or sey, moreover, it was not only possible, but of considerable importance to Homer's story, that not every person perceived was thereby known for who he or she really was, and not every situation encountered was immediately well understood.'* The idea expressed by Heraclitus (perhaps also by Xenowomen vóov «ai emippova pre (Od. 19. 325-6), she must mean by view.” ‘high intelligence (and shrewd cunning)’ she and a few others Thesis (1) above, shared by all the proponents, almost certainly exaggerated the closeness of the tie between perceiving and knowing in the Homeric epics. While ‘Homeric knowing’ was commonly a matter of becoming aware of the features of one’s immediate environment, it is now recognized that Homer also allowed for knowledge on occasions when what was known either was not or could not be directly perceived.'* In various scenes of the Odyspossess in great measure. In both epics Homer sets apart paragons of vóos who ‘know how to voÿoa before and after'—and who can as a consequence give wise counsel—from all those who suffer from limited natural endowments or heaven-sent confusion.'* It would, moreover, be difficult to understand how Heraclitus (in fragment 40) could possibly have communicated a challenge to the conventional view of vöos unless vóos in the sense of ‘wisdom’ or ‘understanding’ already had some currency in common parlance. In short, in Homer vóos, voeiv, yıyyaorw, and €idévar often 1? I ignore in this assessment the question of the etymology of vdos. It is perhaps worth noting that the perceptually orientated etymologies which von Fritz characterized as the ‘most probable’ (from ‘sniff’ and ‘smell’) have not gained support in recent years. Among other possibilities: the proto-IE root *nes-, as seen in the Greek verb véopa ‘return’ (in accounts by D. Frame and P. Frei); the IE root *snu-, based on a parallel with the Gothic snutrs ‘wise, clever’ (Schwyzer and Frisk); and the IE root *nej- (McKenzie) based on a Sanskrit verb meaning ‘lead’ (cf. Greek jyéopa: ‘lead’ and ‘think’). In any case, etymology is not crucial; on any etymology we shall still have to determine what the term meant during the period designate a knowledge that is neither identical with nor an immediate result of sense-perception. Thesis (2) similarly overstates the extent of the perceptual orientation of these terms as they were employed by the earliest philosophers. Frankel, for example, argued that the presence of the visual term idev (‘saw’) in Xenophanes B 34 requires that we give to eióws and oíde the particularly pointed meaning of ‘a knowing rooted in vision’. But if so, then Xenophanes' thesis: when the Homeric epics were being created. 13 Edward Hussey, “The Beginnings of Epistemology: From Homer to Philolaus’, in S. Everson (ed.), Companions to Ancient Thought, i. Epistemology (Cambridge UP, 1990), 13-14, calls attention to various passages in Homer where people are said to have knowledge about events lying either in the future or in the distant past, about general truths, or other matters not open to direct observation. Cf. Il. zo. 203-4 ‘We know (iöpev) each other’s lineage and parents, for we have heard the tales told in olden days by mortal men, but not with sight of eyes have you seen my parents nor I yours (oper 8’ ob’... ides)’; Il. 13. 665 ‘He embarked knowing full well (ed eiöws) his deadly fate, for often had the old one, passages just quoted), we cannot suppose that he wished to deny knowledge to all mortals, and claiming that knowledge about this matter (the names of all the Greek captains at Troy) was available only to direct observers hardly qualifies as a blanket repudiation of all knowledge (cf. Plato, Theaet. 201 Bff.). The limited epistemological implications of the invocation have been pointed out by Hussey (loc. cit.). '* I have defended this thesis at some length in ‘Perceiving and Knowing in the Iliad and Odyssey", Phronesis, 26 (1981), 2-24. The paradigm case is, of course, good Polyidus, told it to him’; Il. 6. 447-9 ‘For of a surety I know (ed... ofa) . . . the day will come when sacred Ilios will be laid low.’ For a parallel critique of the Snell-von Fritz view of Homeric knowledge see E. Heitsch, ‘Das Wissen des Xenophanes’, Rheinisches Museum, 109 (1966), 193-235. Often cited in support of the ‘knowing’ = ‘having seen’ thesis is the famous invocation of the Muses at Il. 2. 484ff., where the goddesses are said to know all things since they are present (mäpeore), whereas mortals hear only rumour and know nothing (xAéos .. . oúde re ue). But since Homer routinely credits individuals with knowledge (as in the LOALPSE Odysseus, who remains unrecognized long after he has been perceived. I discuss the character of the knowledge which Penelope and the other members of his household gain of Odysseus, as well as the methods they employ in doing so, in sect. 4 below, Cf. Il. 1. 572; 3. 109; 18. 250; Od. 24. 452; also Hesiod, WD 793-4: ‘On the great twentieth, in full day, a wise man (‘aropa Hóra) should be born. Such a one is very sound-witted (uáda yäp re vóov merukaauévos écriv) (trans. Evelyn-White).

kai 76 pèv oùv oades oùris avip Dev oùdé res ¿oras ides audi dev re al dooa Àéyw repi mavrwv is essentially linked with 76 &öv, it does not follow that véos must 8 enjoy some intuitive insight into the true nature of reality (or that ei yap Kai rà pddiora ruxoı Teredegpévov ciróv, abrés Guws odk olde: Sdxos 8" émi maa réruxTau. .. and the certain truth no man has seen, nor will there be anyone who knows both about the gods and such as I say about all things; for even if one were to succeed in speaking of what is brought to pass, nevertheless he himself would not know. But opinion is allotted to all. TECSUTPR must be downgraded to the much less interesting claim that ‘no one has or ever will enjoy a perceptually grounded knowledge’ about such (obviously non-evident) matters as the gods and the truth of various views of how ‘all things’ are. The mere absence of perceptual knowledge, moreover, could not possibly have prepared the way for Xenophanes’ conclusion that only opinion is allotted or available to all. Thus, if fragment 34 is to make any sense as an argument, its ‘know’ cannot mean just ‘know through sense-perception’.'° Similarly, when Heraclitus maintains (fragment 56) that éénmérqvrar ... of avOpwrror mpòs Tv yvow Tay pavepaw (‘people are deceived in their knowledge of evident things’) we are to infer, according to von Fritz, that the missing knowledge is perceptual in character—‘the ‘‘hidden harmony” [is] quite directly visible 9 voéw means simply ‘know by direct acquaintance’). Alternatively, and more plausibly, these remarks can be read as asserting that whenever we speak and think, our words and thoughts must have To éév as their subject-matter (even though mortals do not realize this and believe in many incorrect ways that what-is-not is a feature of the real world), since (according to B 6. 1-2 and 8. 35- 7) ro édv is the only thing ‘available’ for thought and speech to be about. It is particularly difficult to see how vóos could be characterized as a faculty of intuitive insight into reality without making the vóos attributed to know-nothing mortals (cf. B 6. 6-9) not so much deplorable as impossible. It must, moreover, be regarded as at least peculiar that, as Thesis (2) would have it, Parmenides chose to include deductive inference—an operation he expressly contrasts (at B 7. 3-8.1) with the faculties of sense—among the EOarEU functions of a faculty he is supposed to have linked with them. If voos and voeiv had enjoyed the close tie with sense-perception which Thesis (2) asserts, we should expect Parmenides to have spoken out in some fashion against them, for example, by having the goddess urge not that we keep our vónua away from some specific way of thinking (B 7. 2), but rather that we should keep for him who is able to see it’. But this is neither an obligatory nor an especially plausible reading to give to Heraclitus’ remark; on the contrary, what people have failed to grasp about the most away from all (supposedly perception-based) véqua. Here, as elsewhere, when we read early Greek poetry and philosophical evident things could be the way or ways in which these things are non-evidently connected with one another.'” Moreover, when Parmenides asserts in different ways that voos tive vocabulary implied some degree of engagement in sense- 1e Frinkel’s ‘empiricist’ reading of the terminology of fr. 34 is discussed (and rejected) by Heitsch, ‘Das Wissen des Xenophanes’, and Jonathan Barnes, The Presocratic Philosophers (London, Henley, and Boston, 1979), i. 1 38-9. Except where noted to the contrary, I follow the texts of the Presocratics given in H. Diels and W. Kranz, Die Fragmente der Vorsokratiker, 6th edn. (Dublin and Zurich, 1966). argument armed with the assumption that the terms of the cogni- AESTRI EPDLITS !? The ‘intelligence that steers all through all’ (fr. 41), the ‘deep measure of the ‘* It appears to have been assumed by each of the proponents of the developmental view that advances in philosophical thought and common-sense opinion necessarily go hand in hand with semantic changes; that is, that the novel beliefs and ideas of a people are necessarily reflected in new meanings for the terms in their language. Witness the statement made by W. von Humbolt (quoted with approval by Snell at the outset of Die Ausdriicke): ‘Die Geisteseigentiimlichkeit und die Sprachgestaltung eines Volkes stehen in solcher Innigkeit der Verschmelzu ng ineinander daß, wenn die eine gegeben wäre, die andere müßte vollständig aus ihr abgeleitet werden können. Denn die Intellektualität und die Sprache gestatten und beförden nur einander gegenseitig zusagende Formen. Die Sprache ist gleichsam die äußerliche Erscheinung des Geistes der Völker; ihre Sprache ist ihr Geist und ihr Geist ihre Sprache; man kann sich beide nie identisch genug denken.’ soul’ (fr, 45), ‘how all things are one’ (fr. 50), ‘the back-stretched connection’ (fr. 51), and ‘the unity of day and night’ (fr. 57), etc. cannot plausibly be regarded as perceptible properties of the cosmos. Heraclitus’ central thesis about knowledge has been aptly summarized by Malcolm Schofield: ‘we need to redescribe or find the meaning in what we know already’ (‘Heraclitus’ Theory of Soul and its Antecedents’, in S. Everson (ed.), Companions to Ancient Thought, ii. Psychology {Cambridge UP, 1991), 18). perception, we encounter many anomalies.'® Wdats Von Humbolt's identification of the forms of thought with forms of speech might have enjoyed some degree of validity in the context of German idealist philosophy (where a cunning world spirit might be assumed to be at work co-ordinating all

Thesis (3) more plausibly holds that for Heraclitus and Parmenides neither voéw nor yıyyworw designated a knowing gained exclusively in sense-perception. But when Snell proposes that an appreciation for a distinctly intellectual form of knowledge emerged in the archaic period when poets became aware of the difficulty of acquiring knowledge and of a need for effort and enquiry, it is difficult, I think, to feel we have the whole story. That it might be hard to know, or that one might need to undertake an enquiry in order to learn, appear to have been ideas fully as intelligible to Homer as they were to Archilochus and Solon.'? And what the Ionian historians demonstrated was not the value of reflection, but rather how wide travel and first-hand observation might enlarge one’s knowledge about the world, its people, and their customs.* What we miss in Hecataeus (to the extent we can know his views at all), as later in Herodotus, is any inkling that we might not be able to discover one or more important truths about the world by reading them directly off its face, or that we might need to engage in one or more processes of reflective thought that would take up where perception left off. But, for all the reasons just given, it is unlikely that we shall be able to explain the emergence of such an idea within the framework of the developmental view. aspects of reality), but it is in general implausible to assume that everything commonly regarded as true about x (or philosophically theorized about x) must also represent an essential element in the meaning of the term ‘x’. " Cf. Od. 13. 312-13: ‘Hard it is for a mortal to know you, goddess (dpyadéov oe Bed, yvóva: Bporw), when he meets you, no matter how clever he might be.’ At Il. 10. 211 ff. Nestor explains that his companions might learn a good deal about Trojan plans if they were to go among them; and at Od. 1. 94ff. (and often elsewhere) Athena directs Telemachus to go to Sparta and Pylos ‘in order to learn’ (mevodpevov) what he can about the return of his father. In ‘Wie die Griechen lernten’, 37, Snell considers Odysseus’ various attempts to find out what is going on around him, who these people are, and what sort of vóos they have, etc., but claims that none of that amounts to the ‘systematic investigation’ of Hecataeus. Perhaps not, but it is nevertheless ‘knowledge gained as the result of human endeavor’ (Discovery, 142). 2% Among many similar remarks: ‘Wishing to get clear knowledge of this matter (cagés ri eldéar) from a point where that was possible, I took ship to Tyre in Phoenicia where | heard there was a very holy temple of Heracles. There | saw it .. At Tyre I saw yet another temple of Heracles ... Therefore, the things I have discovered by enquiry clearly show (rà foropnyéva dy Aoi cagéws) that Heracles is an ancient god’ (Hdt. 2. 44). Philosophical Interest in Cognition J. H. Lesher 11 3. Perceiving and knowing in Heraclitus and Xenophanes Any attempt to show that a Presocratic philos opher embraced some sort of distinction between intellectual ledge TGAY will have to confront the and perceptual knowextraordinary Aristotle and Theophrastus that the Presocratics, testimony of without exception, actually equated thought (and knowle dge) with sensation: Kat of ye dpyaîo: 76 ppoveiv Kai ro alabáveoda: raùròv elval g¢aow (And the ancients say that thinking and sensing are the same thing). (De anima 427°21-2) Ta UrodapBdvew dpovnow pev rH aíodyow (These thinkers suppose thought [or knowledge] to be sensation). (Metaph. 1009? 12-13)?" Clearly, if the Presocratics failed even to disting uish thought from sensation, it would be difficult to imagin e how they Wperison BAEULSRIEBMYONDSETA could also have supposed that any instance of knowledge constituted anything more (or other) than a form of senseperception. | But Aristotle’s remarks must be read with care. On both occasions his characterization of the ancients’ view is linked with the claim that they all regarded both though t or knowledge ($pövnaıs) and sensation (aioßnoıs) as a physical alterat ion (dAAoiwais) taking place within the human body: They all look upon thinking as a bodily process that like is known as well as perceived by like. like perceiving, and hold (De anima 427°26-8) And in general it is because these thinker s suppose knowledge to be sensation, and this to be a physical alteration, that they say that what appears to our senses must be true. (Metaph. 1009%12-1 5) The view expressed in the passages quoted from Empedocles Parmenides, and Homer is that human thought is physically determined, either by ‘what is present’ to the individual (rapeor -.. watts in Empedocles, fragment 106: én’ ap dyna ... vóos from Homer, Od. 18. 136-8: and eLéorn ... &Modpoudorra apparently a loose quotation of Il. 23. 698) or else by mat ** One would normally translate dpoveiv as ‘to think’, but since Aristotle speaks here of dpdunoıs as necessarily true, it appears he has knowled ge in mind. Theo res tus never states the thesis in similarly comprehensive terms, but he does claim that Parmenides ‘says that perception and thought are the same thing’ (De sens. and suggests that Empedocles might have done the same (23).

physical ‘nature’ (uerébuy ... ro dpoveiv in Empedocles, fragment 108: and xpdow pedéwv ... vóos, pedéuv duos ... ppovéea, and ro mhéov ... vonpa in Parmenides, fragment 16). Not only do these remarks not amount to a complete identification of thought and knowledge with sensation, but such a thesis would actually be incompatible with the clear implication of Homer’s description of Hector as still having thoughts while lying insensibly on the ground. Nor would the complete identification of thought with sensation sit easily with the view expressed by Parmenides that our thoughts are a function of the physical condition or disposition of our ‘limbs’ or ‘organs’. Elsewhere (GC 325713), Aristotle confirms that he took the ancients’ ‘identification’ of sensation with thought and knowledge as much less than complete, since he there speaks of the followers of Parmenides as ‘passing over and disregarding sense-perception on the ground that one should follow reason’. In short, the only way in which Aristotle could have consistently believed that the Presocratics ‘had identified thought with sensation’ was that since they had spoken of both processes as instances of physical alteration, they must have considered thought and sensation to be essentially the same natural phenomenon—a familiar, if very misleading, Aristotelian way of speaking of two things as ‘identical’ or ‘the same’.”? Aristotle fan consistently regard thought (or knowledge) and sensation as the same’ (in so far as both are physical alterations) and as quite different from one another in other respects. o. In general, I would assume, we could grant that a distinction had been drawn between two forms of knowledge, one perceptual and the other (at least to some extent) intellectual, if either (1) in speaking about the faculties of sense-perception generally a philo- 13 that some observed phenomena would need to be classified according to larger organizing categories in order to be understood, or that an inference would need to be drawn from the observed phenomenon in order for its hidden nature or causes to be discovered). In fact, as many have thought, both (1) and (2) arguably apply to Heraclitus. In fragment 107 he characterizes the faculties of sight and hearing in general terms as ‘bad witnesses’ for those with uncomprehending souls, and repeatedly asserts that even when people have acquired perceptual experience they still have not gained knowledge (fragments 1, 17, 34, 56, 72). He also held that nature, or the nature of things, was hidden (fragments 55, 123) and that its truth could be conveyed to us only if we were able to puzzle it out (fragment 93).”? His claim in fragment ro1 that he ‘enquired of’ or ‘searched for’ himself, may also have been his way of referring to a process of reflection in which he both posed and answered the most important questions.” Through the repeated juxtaposition of opposing pairs of qualities (experienced yet inexperienced, having learnt but not knowing, having heard the word but not got the message, having been caught and yet left behind, having been overlooked and yet taken along, being present but absent, in contact yet isolated, asleep while awake, familiar yet strange, obvious but hidden, both expected and unexpected, and common although private), Heraclitus appears to have sought to provoke his audience into reflecting on their present understanding of the world, to consider the possibility that it might have a nature drastically different from the one they think it has,?* and that they might need to uncover 22 I follow Marcovich (and others) in reading fr. 93 as a statement not about sopher were to characterize them as incapable (in at least some (or not only about) Heraclitus’ oracular style, but about the nature of the world ing about a particular instance of knowledge, a philosopher were (although I would not necessarily state this in Marcovich’s terms of the ‘Logos emitting ojuara'). See M. Marcovich, Heraclitus: Greek Text with a Short Commentary (Mérida, Venezuela, 1967), 51-2. cases) of providing us with genuine knowledge, or (2) while speakto maintain that some distinct contribution by the intellect would be needed before such knowledge could be acquired (for example, holding that information acquired through sense-perception would need to be interpreted in order for it to have any meaning, or 2 Cf. Metaph. 1018°6-7 raùrà dé€yerat wai dv i ovoia pia, which is also the of as ‘one’ (1016°8-9 ra de sp can bebe spoken hi i two things i which primary sense in "ptas Aeyöpeva Ev div 7 odoía pia). ** A detailed, and I think persuasive, case for reading SL qua: on the model of ‘enquiring’ of the oracle is made by Guthrie, History of Greek Philosophy, i (Cambridge UP, 1962), 416-109. ** Similarly Charles H. Kahn, The Art and Thought of Heraclitus (Cambridge UP, 1979), 124: ‘The only hope of “getting through” to the audience is to puzzle and provoke them into reflection.’ For the implications on the objective side see A. P. D. Mourelatos, ‘Heraclitus, Parmenides, and the Naive Metaphysics of Things’, in E. N. Lee, A. P. D. Mourelatos, and R. M. Rorty (eds.), Exegesis and Argument (Phronesis, suppl. 1; Assen, 1973), 16-48; and more recently, P. Curd, ‘Knowledge and Unity in Heraclitus’, Monist, 74 (1991), 531-49.

that nature in a quite novel way.?® And to the extent to which we can assume that his own method of instruction would exemplify attached to sense-perception and thought. Did he suppose, for example, that we ought to attempt to amass experience through 14 15 what he regarded as the most appropriate approach to learning, perception before seeking to reflect on that experience? we can conclude that he took the most important knowledge to be assign acquired not simply in perception, i.e. not simply in hearing the individual words which made up his Aöyos (or in gaining an holding that the only significant advance towards knowledg e took place when we began to reflect on the various non-perc eptible awareness of individual perceptible qualities of various kinds), aspects of nature? but rather through a process of reflection on the overall structure The consensus in recent accounts is that Heraclitus regarded the acquisition of extensive experience through sense-perception and meaning of what he said (or of what is perceived generally).?’ Thus, alongside Heraclitus’ distinctly materialistic way of speaking of wisdom and intelligence alluded to in Socrates’ remarks in a fairly minimal significance to perceptual Or did he experience, as the indispensable first stage in the learning process: we shall have to carry out extensive enquiries into nature to find out what the Phaedo?* was the view that knowledge came in two forms: it looks and sounds like: and then we must reflect on what one obvious, easy, and common, and the other subtle, difficult, means, what its hidden principles of unity might be, and rare. But it is not yet clear what relative importance Heraclitus ze Cf. Plato, Rep. 5245: ‘But if some contradiction is always seen with it, so that it no more appears to be one than its opposite, there would forthwith be need of something to judge between them, and it would compel the soul to be at a loss and enquire, by arousing thought in itself (dv év aùrw Yuyn amopeiv ral Enreiv, xwotaa év &aurn Tv Evvorav)', which may be as good an explanation of the aporetic character of Plato’s early dialogues as it is of Heraclitus’ style. The view that paradox played a key role in Heraclitus’ approach to instruction forms a main theme in M. M. Mackenzie’s ‘Heraclitus and the Art of Paradox’, OSAP 6 (1988), 1-37. 17 The extent of the innovation represented by this conception of language comprehension is explained in Martha Nussbaum's “Puy in Heraclitus (I and 11), Phronests, 17 (1972), 1-16, 153-70. 28 The main texts are fr. 118 (on the wisdom and superiority of the dry soul), and 117 (on the lack of awareness characteristic of those having a wet soul). According to Sextus (Heraclitus A 16), Heraclitus linked the achievement of a rational understanding of the cosmos with coming into contact through senseperception with the rational Adyos which rules the world. On some occasions in the history of philosophy the identification of material substances and physical it all etc.” Fragments 35 (‘lovers of wisdom must be enquirers into very many things’) and 22 (with its allusion to ‘seekers of gold’) have been read as assigning a positive value to enquiry and the experien ce gained through it; while fragment 40 (the famous put-dow n of the polymaths) has been regarded as an assertion that senseexperience cannot suffice for knowledge—genuine wisdom must be more than disconnected bits of information one picks up along * Cf. Edward Hussey, The Presocratics (New York, 1972), 37: ‘He reaffirms the importance of first-hand inquiry ... But the results of such inquiry are not an end in themselves: it is necessary to have understan ding (noos), in order to be able to interpret the evidence of the eyes and ears’ (similarly Kahn, Heraclitus, 108: ‘polymathié -.. is not enough ... the mere accumulation of informatio n will not yield understanding, unless it is accompanied by some fundamental insight’); cf. also G. S. Kirk, The Cosmic Fragments (Cambridge UP, 1954), 61; Robinson, Heraclitus: Fragments (Toronto, 1987), 182; Marcovich , Heraclitus, 28; Barnes, Presocratic Philosophers, i. 147-9; G. S. Kirk, J. E. Raven, and M. Schofield, The Presocratic Philosophers, zndedn.[KRS] (Cambridge UP, 1983), 187. The main dissenter was F. M. Cornford, who (in From Religion to Philosophy [New York, 1957] and Principium Sapientiae [Cambridge, 1952]) portrayed Heraclitus as an embodiment of ‘the mystic temperament’, successor to the shaman, and a fierce processes related to cognition has formed part of a larger effort to reduce all cognitive phenomena to physical ones. But since Heraclitus identifies various matters as involved in acquiring knowledge (e.g. paying attention, reflecting on opponent of Ionian enquiry and rationalism. But the implausibi certain matters, interpreting, etc.) that are not themselves linked with physical (1955), 65-76, repr. in D. J. Furley and R. E. Allen (eds.), Studies elements or states, a reductionist programme seems unlikely. Heraclitus does Philosophy, ı (London and New York, 1970), 42-55. Uvo Hölscher also claimed that ‘the paradox of the unity of opposites was something not to be proven [as in adopt a form of cognitive materialism (in so far as he proposes that certain material elements are causally linked with thought and knowledge), but there is no good reason to think that he sought to reduce thought and knowledge to physical states or to account for them solely in terms of material conditions. For the same reasons (as well as for others relating to the limited credibility he assigned to the Sóta account) Parmenides’ explanation of vóos in terms of the physical condition of our bodily organs (fr. 16) can hardly be regarded as his entire account of the nature of thought and knowledge. lity of Cornford’s view of Heraclitus as mystic was made clear in Vlastos’s review in Gnomon, 27 in Presocratic Parmenides], but to be grasped through intuition alone. B 18 cautions us that insight into the paradox is not the acquired result of investigation; there is no “method of inquiry”’ (‘Paradox, Simile, and Gnomic Utterance in Heraclitus’, in Mourelatos (ed.), The Pre-Socratics, 233). Guthrie appears to have sought to straddle the fence on this issue, claiming that Heraclitus saw no purpose in foropin (History of Greek Philosophy, i. 417), and that he held that ‘accurate perception of phenomena is the necessary preliminary to the discovery of the Logos’ (ibid. 429-30).

the way. While this view of Heraclitus has some initial plausibility, it is not as well founded as has been supposed, and accepting things happen’ (fragment 1) and are ‘one’ (fragment 50). We should read fragment 40, then, as the statement that in spite of all 17 their vast learning? none of these four paragons of wisdom it does bring some odd consequences tn its train. We can begin with the disparagement of roÀvuabin in fragment managed to grasp the common and unifying Aöyos at work in the cosmos. 40: hd è moAvpadin voor’ où Siddorer Hoiodov yap av édidate xal IluBayopnv, abris re Eevopaved re kat Evxaraiov. Much learning does not teach wisdom; for it would have taught Hesiod and Pythagoras, and again both Xenophanes and Hecatacus. The argument consists of a sweeping negative thesis followed by several counterfactual cases cited to prove its point. Judging from some similar ‘inferences from failure in the best case’ arguments,*' Heraclitus appears to be trying to drive a wedge between two matters normally regarded as identical or synonymous. The argument allows the first item to be realized to an optimal degree, and then claims that even in this best case it still does not constitute a bona fide case of the second. Since (so it is implied) even the greatest paragons of learning cannot be credited with ‘wisdom’, then, contrary to conventional opinion, the two states or qualities of mind must be distinguished from one another. | When in fragment 114 Heraclitus links ‘speaking with vdos’ with ‘basing themselves firmly on that which is common to all things’, and in fragment 41 identifies wisdom (or ‘the wise —7ò oop6v) with ‘knowing the intelligence that steers all things’, he thereby links the desired vóos with the grasp of the Adyos he elsewhere describes as ‘common’ and as that ‘in accordance with which all 3° I follow Marcovich in omitting the éxew in Athenaeus and Clement, which would (implausibly) turn vóos into a natural endowment ( = ‘having sense’, ‘having good sense’), arguably not teachable by anything. In this respect Heraclitean vdos differed from the éiveois which Theophrastus (De sens. 25 ff.) claims Alcmaeon distinguished from perception, since it served to separate human beings from the lower animals. Alcmaeon’s £éveois was, therefore, not ‘Heraclitean understanding in the sense of a grasp of some complex matter acquired through reflection, but It has sometimes been thought that since the -uad- element of modvpabin also appears in the ôfis dro) udßmaıs trio of fragment 55, the state of awareness disparaged here in fragment 40 consisted in the body of information gained from sense-perception.* But this is unlikely, for several reasons. While (as we shall see) a robust regard for the value of information gained through senseperception can be plausibly attributed to Xenophanes—perha ps even to Pythagoras, whom Heraclitus credits with both mrodvpabin and ioropin in fragment 129—the same cannot be said for the Hesiod who in standard poetic fashion identified the divinities as his sources (cf. Th. 22 Hoiodov kaAnv ¿Sidatar dov, and WD 1 ff.). Second, fragment 55 does not actually express a preference for what is learnt in sense-perception (i.e. ‘whatever comes from sight, hearing, learning’), but rather a preference for Sowr— "whatever (or as many) from which’ or ‘whatever (or as many) of which’ there is ‘seeing, hearing, learning’. We are not told the identity of these ‘many’, nor what they are preferred to, but one natural choice would be the many persons, things, and events there are that populate the cosmos.** Since Heraclitus elsewhere criticizes those * Rob Bolton has argued that the argument contained (or at least assumed) two additional premisses: (1) the teachings of the polymaths were thoroughly incompatible with one another, and (2) if (1), then none of them could have qualified as a paragon of vóos (‘Nature and the Human Good in Heraclitus’, in K. J. Boudouris, Jonian Philosophy (Athens, 1989], 49-57). While it is intriguing to think that Heraclitus might have developed an early form of the sceptical argument from the existence of disagreement among experts, there is a serious obstacle in the path of this reading; if their teachings had been (collectively ) riddled with contradiction, they could not even have qualified for moAupaßin, as of course they must if they are to show that it où 5:ddoxer voor. » Cf. Snell, Discovery, 144: “This delight in a wealth of experience which was rather ‘understanding’, as in the title of Locke’s Essay: ‘the faculty by which one sO prominent in archaic Greece ... met thinks and knows’. Heraclitus [fr. 40].’ Cf. also K. Pritzl, ‘On the Way to Wisdom in Heraclitus’, 2 Tronically, a form of argument first on display in Xenophanes. Cf. B 34. 374: ‘even if, in the best case (rà padAcora), one were to succeed in speaking of what is brought to pass, still one would not oneself know; but opinion is allotted to all’, its first adversary in the person of Phoenix, 39 (1985), 303-16; Kahn, Heraclitus, 106; P. K. Curd, ‘Knowledge and Unity in Heraclitus’, Monist, 74 (1991), 531-49, at 540 ff. ** Hippolytus, to whom we are indebted for fr. 55, read it as embodying a Cf. also B 2. 17 ff. in support of his claim that athletic excellence is not worth the contrast between ‘things seen’ and ‘things unseen’, and degree of honour accorded to it: ‘nor if there were someone swift of foot—which for the former over the latter (similarly Snell, Discovery, 145). is most honoured of all governed’, men’s deeds of strength—would a city be better as expressing a preference Yet a preference for the seen over the unseen sits poorly with the idea expressed in fr. 54 that the dppovin ddavis davepjs xpeírrwv, It appears that Hippolytus wished to swallow this

who hold bards and poets in high esteem (fragments 104, 57, 56, and 28a—perhaps also a reference to Hesiod), it would be natural to regard fragment 55 as a statement to the effect that he for one (éyw) regards the many persons, objects, and events he can perceive and learn about as more estimable authorities than singers and poets (cf. the similar contrast of person and authoritative reality contained in his fragment 50: ‘listening not to me but to the Adyos’). Neither udßnoıs nor the cognate verb pavBavw, moreover, is restricted in meaning to what one learns through sense-perception, but both range across what a person might learn from perception, or by practice, or through instruction from another.” Indeed, to be compatible with the variety of approaches represented by the four paragons, udßnoıs would need to be ‘learning’ in a fairly broad sense. There is, then, no good reason to interpret the failure of all the polymaths to gain wisdom as in itself an indictment of a perception-based approach to knowledge. Furthermore, if we assume that Heraclitus was favourably Philosophical Interest in Cognition 19 disposed towards ioropin, we shall have to infer that when in fragment 129 he describes Pythagoras as having practised ioropin ‘most of all men’, he is actually commending Pythagoras for this positive achievement, even though (1) Heraclitus goes on to say that the wisdom Pythagoras constructed for himself out of these writings was a ‘much learning’ that was also ‘bad art’, and (2) he elsewhere calls Pythagoras ‘prince of swindlers’ (fragment 81a). And while a positive reading may be theoretically possible, that is not the most natural sense to assign to these two references to seekers and travellers: Those seeking (of 5:{jpevor) for gold dig up much (moAAy») earth and find little. (fr. 22)** One could never discover the limits of soul, though one were to traverse every road—so deep a measure does it possess. (fr. 45) Two other fragments speak of 'enquiry' in more ambiguous terms: Men who are lovers of wisdom must be enquirers into [or perhaps knowers of*”] very many things (ed uaAa moAAwv toropas). (fr. 35) inconsistency (between fr. 55: ‘honouring the visible more than the invisible’ and fr. 54: ‘he praises and extols the unknown part of it and the unseen part of its power more highly than the known part’). Catherine Osborne explains that in Hippolytus’ view Heraclitus’ teaching was that there was no significance in the opposites of conventional morality—thus prefiguring the Noetian view that god possesses all attributes (Rethinking Early Greek Philosophy: Hippolytus of Rome and the Presocratics [Ithaca, NY, 1987], 133-82). Osborne’s thesis may provide us with a coherent explanation of what led Hippolytus to select the particular fragments he did, but if we are to accept that Hippolytus understood Heraclitus correctly, we shall have to conclude that Heraclitus did not really mean all those nasty things he said about the Ephesians, Pythagoras, Hesiod, and Homer (since he meant to abolish all contrasts between negative and positive), and we shall have to attribute to Heraclitus an interest in precisely the issue that divides Hippolytus from Noetus some seven centuries later, 35 Cf, Xenophanes B 10: ‘since from the beginning all have learnt (nepaßnxaoı) according to Homer’, It is difficult to see how pavOdvw can be a matter of ‘learning by personal experience’ in Herodotus when the Cymeans learn of what the Mytilenaeans are about to do with the person shipped off to them for safekeeping (1. 160 ds ¿uadov raúra mpyocóueva), or in Parmenides when the goddess promises the youth that he will ‘learn (u00%0ea,) the opinions of mortals in which there is no true trust. Snell argued (correctly, it seems to me) that pavdávo originally involved accustoming oneself to behave in a certain manner (cf. Il. 6. 444: ‘learning always to be valiant’). A _pd070 that takes place through trial and practice is also evident from Alcman (fr. 109 Diehl) and Pindar (Isth. 1. 56) through to Aristotle (cf, NE 1103°33 raüra moroûvres pav?dvopev), but learning by doing is itself not synonymous with learning through sense-perception. I enquired of myself (&dılnoaunv ¿uewuróv). (fr. 101) But in each there is enough of a twist to suggest that neither is enquiry of the familiar Ionian sort.** It must also be counted of 1% Marcovich proposes to read this comment in a positive vein: ‘little but precious and worth searching for, because it is gold’ (Heraclitus, 38), but the correspondence of ‘those who dig up many things’ with those who have ‘learnt many things’ makes so positive a gloss difficult to accept. * It is not even certain that forwp here must mean ‘one who practices ioropin’ (cf. the use of iorwp in the passage quoted from Hesiod’s Works and Days in n. 15 above). The term was used generally to refer to wise men, judges, or observers. ** Cornford took fr. 35 as ‘an ironical sneer’ at Pythagoras and the other ‘much learners’ (From Religion to Philosophy, 186n.), and cited the characterization of Hesiod in fr. 57 as mAeiora elöeva as equally contemptuous. A similar reading (of fr. 35 as ‘heavily ironical’) was adopted by Guthrie (History of Greek Philosophy, i. 417). Most commentators have taken it as a positive characterization—do faropin, by all means, but then go on to do something more. So far as I can see there was neither sneer nor irony in characterizing Hesiod as the one person of whom most ‘feel sure’ (r{oravrau) that he knows the most things. But if it is hard to be certain about when Heraclitus’ comments are laden with irony, we can be sure that he speaks of those who think they are merely in one state (presence, experiencing, knowing, encountering, etc.) when in reality they are still in the opposite condition. In this sense he could have spoken of those who are currently learning about nature as needing to learn far more than they have to date.

some relevance that no fragment or testimonium depicts Heraclitus as actually engaged in the ¿oropín he is supposed to have valued as the essential initial stage in the acquisition of knowledge. He does of course allude to the orderly arrangement of the cosmos (and the measures, limits, and exchanges among various natural phenomena), but this is not in itself support for foropin as a form of research, i.e. for the gathering of factual information about the natural world through wide travel and direct observation. The inclusion of Xenophanes in the list of fragment 40 may provide some leverage in the question of Heraclitus’ attitude quential, but, fragment 40 notwithstanding, were kindred Jonian spirits in many ways.*! 21 the two thinkers But if so, then what are we to make of the unfavourable mention of Xenophanes in fragment 40? There is one (and so far as I can see, only one) plausible candidate for an offending view: Xenophanes’ theory, as reported by Hippolytus, of the wide-scale destruction of human civilization in alternating eras of drought and flood: j Xenophanes thinks that a mixture of the land with the sea comes about, but that in time [the land] becomes freed from the moisture ... and he says that all men will perish when the land sinks into the sea and towards ‘oropin. Partly as a consequence of two loosely phrased comments by Plato (Soph. 242 C—D) and Aristotle (Metaph. A 5), Xenophanes became known to posterity as the founder of Eleatic philosophy. But many of the surviving fragments of his poetry, as well as a great many of the testimonia, link him with the debates of Ionian philosophy.’ Both the merits and the seriousness of purpose of Xenophanes’ most novel view, his ‘unified cloud theory’, have been questioned,*” but many of his teachings display what might fairly be regarded as the defining characteristic of unfolds in time (xpdvw). Heraclitus’ own view that the cosmos becomes mud, at which time generation begins again, and this change comes about in all worlds [or world arrangements]. (Xenophanes A 33 (5-6) ) Xenophanes’ account draws on several ideas expressed first by Anaximander: an alternating cycle of opposites (here of drought and flood, or the wet and the dry), the notion of a plurality of ‘world arrangements’ (mda. rois kócuors), and of a process that Ionian scientific thought—seeking to account for the marvels of goes through a cycle of change as opposites pass over into one nature by reference to an entirely natural set of substances and another stands in the same tradition (although he takes pains to forces. The testimony that Heraclitus was actually a pupil of Xenophanes (Heraclitus A 1 (5); 1a) is late and probably inconse- 80 that strife zs justice). distinguish it from Anaximander’s view by holding in fragment But absent from Xenophanes’ theory (as stated in A 33) are two elements Heraclitus insisted upon as vital to a correct understand# In fr, 28 Xenophanes joins in the Ionian debate about the earth's shape and extent, asserting that while we see its upper limit here at our feet, its lower portion stretches down ‘without limit’ (és ämeıpov—as I read it, ‘indefinitely’ not ‘infinitely’ far). In contrast with earlier claims that the earth floated on water or was surrounded by air, Xenophanes' thesis at least had the virtue of being universally confirmed by experience—what would have been observed at every location would only have been earth resting on more earth below. In frr. 29 and 33 he identifies earth and water as, in effect, Milesian apxai, since all things were said to come from them. In what is more consistently read as a piece of geophysics, fr. 27 identifies (the) earth as that from which all things come and that in which they all end up, which fits nicely with the view expressed in fr. 30, that the sea is the source of all winds, clouds, and all other waters (which arise from and descend to return to it through the compression and vaporization described in A 46, from Aetius). Various testimonia credit him with accounts of the (cloud-based) nature of the sun, moon, eclipses, lightning, meteors, comets, and St Elmo's fire (known popularly as the Dioscuri). * Cf. e.g. the comments made by Guthrie (History of Greek Philosophy, i. 383): ‘He was not only, or even primarily, a natural philosopher ... he was a poet who took the didactic function of poetry seriously’; see also ibid. 390, 393, 395. ing of the cosmos: (1) Xenophanes apparently never realized that * In his frr. 11 and 12 Xenophanes described Homer and Hesiod as authorities not to be trusted (cf. Heraclitus B 57, 104, 106); he presented account of a supreme being that in various respects both did (in his frr. 23-6) an and did not resemble Zeus (cf. Heraclitus B 32); he noted (in fr. 34) the limited range of human knowledge, and may have connected this view with a restriction of genuine knowledge to the divine (cf. Heraclitus B 78, 83); he criticized his fellow citizens for their pursuit of wealth and an extravagant lifestyle (fr. 3; cf. Heraclitus B 125a); he viewed Pythagoras (in his fr. 7) with much the same contempt that we hear from Heraclitus (frr. 81, 129); he ridiculed the superstitious beliefs in conventional religious ritual (fr. 17; cf. Heraclitus B 5, 14, 15); mankind for its generally deficient understanding of the gods embedded and he criticized (frr. 1, 14-16, 30-2; cf. Heraclitus B 5). Xenophanes was also credited with a number of scientific ideas which have parallels in Heraclitus—e.g. of a sun created anew every day (A 33, 41; cf. Heraclitus B 6) and of an eternal and imperishable cosmos Heraclitus B 30).

the processes of cosmic change must have a periodic character; i.e. that the successive eras of drought and flood occur within fixed limits or measures manifesting overall equality and (hence) justice:* and (2) neither here in A 33 nor anywhere else did Xenophanes speak of earth and water, and the wet and the dry, as interdependent (or, as Kahn puts it, co-variant) elements within a larger continuum (cf. Heraclitus B 36, 76a, 76c, 126). Since failing to see how the various opposites in nature are essentially ‘one’ was in Heraclitus’ view the cardinal philosophical sin (cf. contrasts the conventional, highly mythic view of the rainbow with the natural cloud phenomenon which is actually ‘there to be 22 fragments 8, 10, 50, 51), Xenophanes’ failure to grasp the larger significance (as Heraclitus would have understood it) of the cycle of destruction and regeneration would warrant his presence in the list in fragment 40 just as fully as Hesiod’s failure to grasp the reciprocal relationship between day and night testified to his ignorance about nature (fragment 106). Hippolytus also reports that Xenophanes arrived at his theory of human destruction in alternating ages of drought and flood through discoveries of the fossilized remains of sea creatures at various inland locations throughout the Mediterranean region: And he asserts that there are proofs for these ideas: that shells are found inland and in mountains; and he says that in Syracuse imprints of fish and seals were found, and in Paros the imprint of coral in the deep of the marble, and on Malta slabs of rock containing all sorts of sea creatures. (Xenophanes A 33 (5)) Other fragments and testimonia display Xenophanes’ curiosity about phenomena in distant locations (e.g. the presence of water in underground caverns, month-long ‘eclipses’ (at northern latitudes?), mountains and volcanic eruptions in Sicily, St Elmo’s fire on ships at sea, etc.), as well as divergent conceptions of the gods from Thrace to Ethiopia, and differing customs from Lydia to Egypt. In fragment 32 (on Iris, the rainbow) he explicitly * The view that (at least temporal) periodicity loomed large in Xenophanes' science and theology was presented by H. A. T. Reiche in ‘Empirical Aspects of Xenophanes' Theology’, in J. P. Anton and G. L. Kustas (eds.), Essays in Ancient Greek Philosophy, i (Albany, NY, 1971), 88-110. Xenophanes mentions a number of astronomical bodies (the sun, moon, and stars) whose movements are clearly periodic, but what he discusses is why they appear and disappear (and in one case why the sun only appears to move in a circle). There is no mention (and no implication either) of a natural principle of justice or equal measures of natural change in any Xenophanes fragment or testimonium. 23 looked at’. In his most extended comment on knowledge (fragment 34) Xenophanes denies that anyone has achieved, or will do so, 70 aades ‘about the gods and such things as he says about all things’; to speak of knowledge in terms of 76 oades is implicitly to testify to the importance of observation for knowledge.‘ Although Xenophanes did not in any surviving fragment use the word ioropin, his commitment to it is evident. (He does at least say in fragment 18 that ‘by searching mortals find out better — Inroüvres édevpiorovow duewvov.**) Thus, if Heraclitus had been favourably disposed towards ioropin, he could only have applauded the line of research that led Xenophanes to his theory of worldwide destruction—the one theory, of all of his teachings we know anything about, that can possibly account for his presence in the list of benighted polymaths in fragment 40. On balance, then, we make better sense of Heraclitus’ teachings if we take his view to be that we ‘learn vdos’ by reflecting on the cryptic and often highly misleading picture of the cosmos conveyed to us through sense-perception and ordinary speech. What we can thereby learn is how the many apparently separate, even fiercely opposed, elements in nature in reality enjoy a deep symbiotic unity. It is at least a fair guess that he acquired his appreciation for the value of the reflective path not—as the view supposed—when he climbed aboard the toropin bandwagon, but rather when the failure of the Ionian developmental scientists to uncover the hidden unities in nature convinced him that travel and observation would never lead to the discovery of + important truths ‘even if one were to traverse every road’, M There are three reasons to think this: (1) In Homer, knowing in a way that is odga often involves direct exposure or experience (e.g. Il. 7. 226). (2) A contrast between direct and indirect approaches to knowledge is implied by Alcmaeon’s contrast of divine cagrveta with mortal rexuaípeadas in his fr. 1. (3) Herodotus also expresses the idea that knowing in a way that is aagés involves first-hand observation (see the passage quoted in n. 20 above), + As I read it, fr. 18 was not the earliest ‘hymn to progress’ but rather an assertion that enquiry is a ‘better way of discovering’—i.e. better than the approach rejected in B 18. 1 (gods intimating all things to mortals). For a defence see ‘Xenophanes on Inquiry and Discovery: An Alternative to the “Hymn to Progress" Reading of Fr. 18", Ancient Philosophy, 11 (1991), 229-48,

4. Parmenides’ way of knowing divisions, comes into being and is destroyed, moves about, and In Parmenides B 7. 3-6 an unnamed goddess commands a youth to take what sounds very much like a reflective path to deciding (by means of or on the basis of her account) the 24 25 readings—they erroneously suggest that what there is admits of knowledge:** develops over time. The alternative she proposes is judging or multiple-event* testing process she has presented. It is clear from B 8. 15 that each of the decisions facing the youth represents the same binary ‘is or is not’ choice stated at the unôé a’ ¿os moAumeıpov 680v kara THVdE Bido0w, vwpuéy áoxorov Supa Kai Nxrjeooav dxoury outset: either ‘it is and is not available for not being’ or ‘it is not and is properly involved in not being’. The goddess proclaims kal yAacar, xpiva: Se Aóyw moAvSypw ¿hey xov both early (B 8. 1-2) and late (B 8. 50-1) that there is only one ef ¿uédev pnôévra. true account or way of thinking: that ‘it is’. During the course of Nor let custom born of much experience force you to ply along her critical examination, each of the various respects in which one this road might imagine that what is might not be (a time—either past or An aimless eye and echoing ear And tongue, but judge by discourse** a much-contested testing*” future—in which what is either was not or will not be, a place Spoken by me. where what is is not, or an instance of what is moving to where it As I understand the meaning of this difficult passage, the goddess here calls on the youth to resist the pull of custom or “force of habit’ to think about ‘what is’ or ‘the existent’ (ro édv) in the terms made available to him (as to all mortals)** through language and perceptual experience. Our sense of sight is ‘aimless’ (lit. ‘untargeted’) and our sense of hearing and faculty of speech (the ‘tongue’ of mortals) are ‘echoing’ in so far as they give false # After quoting these lines Sextus comments: “The thinker . . . proclaimed the cognitive reason to be the standard of truth in things that are, and gave up paying attention to the senses’ (M. 7. 114, trans. Gallop). 4 Following the translation given by A. H. Coxon, The Fragments of Parmentdes (Assen and Maastricht, 1986), 58. The non-empirical character of this judgement has led many to assign to Aóyos the meaning of ‘reason’ (i.e. ‘judge by means of your faculty of rational thought’), but this would move forward the earliest use of Aóyos in this sense by about a century. See the summary in Guthrie, History of Greek Philosophy, i. 419-24. . + Cf. Pindar, Nem. 8. 20 veapà $’ E£eupövra Ödper Baodvw és ¿deyxov, amas xiv8vvos; for a parallel ‘testing’ use of the related verb ¿Xéyxw cf. Bacchylides fr. 25 Edmonds. For additional details see ‘Parmenides’ Critique of Thinking’, OSAP 2 (1984), 1-30; and Coxon, Fragments of Parmenides, 192. “ Coxon identifies the way rejected here with the one repudiated at B 6. 7, and is not, or in some respect becoming what it is not) is considered. But in each case, in order to think that what is is involved in not being, in some respect, we must also think that in that same manner what is not is—which violates the principle stated at the outset of fragment 7 (following Gallop): ‘never shall this prevail: that things that are not are’.*° “ So I (now) understand the meaning of roAvönpıs, in the light of the parallel between its moAv- element and the moda ofuara at B 8. 2-3: the testing process comprises five issues (whether rò &öv can admit of generation, destruction, division, movement, and temporal development), and each of these is decided in the same way, in favour of ‘it is’. Parmenides organizes, as it were, a five-event elenchus, a veritable philosophical pentathlon (while önpıs is more commonly a matter of strife or conflict in battle, at Nem. 11. 26 Pindar describes Aristagoras as a better wrestler ‘than his competitors’—snptwvrwr). 5° ] omit in this reconstruction of Parmenides’ argument two premisses frequently attributed to him: (1) that in each of these incorrect forms of mortal opinion one is thinking ‘it is not’, and (2) it is impossible ‘to think, speak of, or come to know of something that does not exist’. (1) is certainly correct, and it is true that Parmenides also says things along the lines of (2) (B 2. 6-7, 3. 1, 6. 1-2, 8. 8-9). But it is not completely clear that these comments were ever intended as premisses rather than conclusions. As they are usually construed, they would: (1) render incorrect mortal opinion not so much deplorable as impossible, and (2) regards the ‘know-nothing mortals’ mentioned at B 6. 4 as those (Ionian) philosophers who attempted to acquire knowledge based on perceptual experience. I do not see, however, that the mere reference to a ‘way’ (of enquiry) unambiguously seeks a highly persuasive truth, and has an impressive argument against the ‘is least equally misleading to the non-philosopher. It is, then, possible that the actual argument, and to think of the various announcements about what can be then, and in some respects but not in others. by the end of the argument it will be shown that one cannot think of, speak of, or know ‘what is not’, since by then it will have become clear that what is there to be targets philosophical enquiry, and in any case speech and sensory experience are at opposing view here is that held by anyone who operates according to the common-sense belief that what there is exists here but not there, now but not require that Parmenides’ entire argument rest on an extremely controversial, indeed extremely counter-intuitive, ‘law of thought’. Since Parmenides manifestly not’ way of thinking along the lines I have suggested, I prefer to regard it as his thought and known as proleptic announcements of what the goddess will prove—i.e.

27 told that we can neither ‘come to know nor tell of what is not’; (2) in fragment 4 we are told to ‘look at things that while far off are The goddess draws her conclusions together at B 8. 26-34: Necessity keeps what is within strong bonds—changeless, motionless, uncreated and indestructible, and fully developed in every securely present to [or perhaps by] véos’, perhaps a reference to way. In these arguments the goddess both invites and empowers the youth, as well as those in Parmenides’ audience, to reflect at knowledge gained through reflection; (3) in fragment 6 she alludes to ‘know-nothing mortals ... whose vóos is guided by helplessan unprecedented level of generality on the question of what ness’; (4) towards the end of fragment 8 she announces that we there is. Both the subject-matter and the logical format of subse- ‘shall learn from this point on about the beliefs of mortals, listening to the deceptive ordering of my words’ and that we are quent philosophical debate reveal that her invitation did not go being told this so that ‘no mortal insight or judgement (yvoun) will overtake and pass him by’, and (5) in fragment 10 (which could belong to the proem rather than to the Sdéa section of the unheeded.*' But if we have reason to think that Parmenides advocated a reflective path to knowledge of what there is, we do not yet have a poem)*’ the youth is told that he ‘will know the nature of the aether and all the signs in it... and learn the wandering works of satisfactory understanding of what this knowledge would have been like. The goddess does promise in the proem that the youth will learn everything (both an ‘unshaking heart of well-persuasive the round-eyed moon . . . and know the surrounding sky’. But these remarks seem to tell us more about what the youth will learn truth as well as beliefs of mortals in which there is no true trust’), than about why any of this would constitute knowledge. During the course of the goddess's presentation we hear a good deal about ‘thinking’ (voéw), opinion ($ééa), trust (miorıs), persuasion but she never says explicitly at any subsequent point that he has actually come to know these things, nor would it be easy to see why this would be the case even if she had said it. We can only one way of thinking along which enquiry can profitably take (ITabw), signs (ojpara), truth ("AAnbetn), saying (Aéyw), judging (xpivw), and the édeyxos, but none of these individually is either place, the one that holds that what is exists fully in every possible synonymous with or implicative of knowledge.’* It is, in short, reasonably assume that the youth comes to know there is one and way, but in virtue of what specific conditions or hallmarks of knowledge would it be true to say that the youth knew such things? Excluding for the moment the predictions made by the goddess in the proem, here is what we have by way of Parmenidean emeqnuieri comments about knowing (or learning): (1) in fragment 2 the ‘is thought of, spoken of, and known is only ‘what is’. In short, the real work of refuting the ‘is not’ way of thinking may have been carried out simply through the multiple proofs of fr. 8. This way of understanding the logic of the argument is meaning of a situation’): “The proper translation of the verb in Parmenides is a term like “cognition” or “knowledge”: it is paraphrased by yvúwac, “to recognize, not, however, a consequence of the elenctic or peirastic character of Parmenidean # For similar elenctic arguments in Philolaus and Gorgias see ‘Parmenides’ Critique of Thinking’, 16n.; for the elenchus procedure in Socrates, Plato, and Fortunately, many of these same characters played significant * See P. J. Bicknell, ‘Parmenides’ Fr. 10', Hermes, 96 (1968), 629-31. ** Von Fritz’s studies of vóos and voéw have often been cited in defence of the thesis that in this period voéw just meant ‘know’ (in so far as it meant ‘realize the not’ way of enquiry is said to be ‘beyond learning’*? and we are knowledge. Fr. 8 will surely be regarded as part of Parmenides’ argument even if one does not consider it, as I do, to be rhe heart of it. hard to tell the players in Parmenidean knowledge without a programme. I veanst be acquainted with”” (Charles Kahn, “The Thesis of Parmenides’, Review of Metaphysics, 22 [1969-70], 700-24 at 703n., with similar views expressed by Guthrie, Robinson, Mourelatos, Coxon (Fragments of Parmenides, 174), and many others). But von Fritz characterized the ‘realize...’ use as only one of several senses (among the others: ‘intend’, ‘plan’, imagine’), albeit the basic one. voéw, I believe, can on occasion (typically in its aorist forms) be thought to imply knowledge in so far as it amounts to ‘noting’, ‘noticing’, or ‘realizing’ a particular feature in one's environment (e.g. Il. 3. 21: ‘But when Menelaus dear to Ares Aristotle, ibid. 3on. 5 Rather than Coxon's ‘about this ‘‘track’’ she has nothing to impart’ (Fragments of Parmenides, 177). Coxon compares B 1. 28 and the drevdéa at Od. 3. 88, but his is much too weak a version of the goddess's programme (cf. B 6. 1-2, 7. 1- spotted him (èv670e), his heart was shattered’). But unless one can find reason to read the forms of voéw in Parmenides’ poem in these special ‘note’ or ‘notice’ ways 2, 8. 8-11, where she imparts a good deal about the incorrect way). In Od. 3. 7989, moreover, Telemachus says that while we can learn (mevdóueda) where the (which in the light of the high level of abstraction seems extremely unlikely), they must be understood not as knowing, but rather as thinking (or conceiving, under- Achaeans died while fighting at Troy, the site of Odysseus’ death is unlearnable (ärevbéa) since no one can say for sure where he died. standing, etc.). What I take to be the correct view is defended by Barnes, Presocratic Philosophers, i. 158-9. SASU7LETaAHRI

roles in the famous story of a journey and discovery on which Parmenides modelled portions of his account, Homer’s Odyssey.** We can get some sense of the character of Parmenidean knowledge by reminding ourselves of some of the best-known incidents in Argus recognizes his master the instant he sees him (17. 301 ws evonoev). Similarly, Eurycleia recognizes Odysseus the moment Homer’s tale of Odysseus’ return from Troy. ter of the vóos of the two swineherds (21. 195-205), Odysseus promises he will tell them the truth (212 dAndeínv xatadééw) as 28 she touches his scar (19. 468 yvò p’ émpaccapévyn). But having resorted to his usual probing examination to ascertain the charac- We first encounter our earlier hero not as a youth but well into well as show them a clear sign (217 oñua dapippadés .. . Seiw) so middle age, and not on his way to the house of a goddess but sprawled half-dead, washed up on Scheria, the far-off island home of the Phaeacians (Od. 6. 1). Sound asleep and buried under a pile of leaves, Odysseus awakes to the shouts of Nausicaa and her handmaidens. An initial moment of alarm gives way to a completely characteristic response: GAA’ ay’ éyaw aúros reipioopo: Ne wpa. (6. 126) But, come, I myself will make trial and see. From the moment of this first encounter with Nausicaa through to the end of his journey Odysseus will repeatedly ‘go to see for himself”, ‘to make trial’, or ‘check out’ what manner of folk there might be—and many are those whose vóos he will come to know that they ‘may well know him and be persuaded in their heart’ (218 éd yvwrov morwbrróv 7’ Evi Bvu®). When they see the identifying ofua (his scar from a youthful injury), they fling their arms around him and weep openly (222-5). e—iem ** The many points of correspondence are set out in E. Havelock, ‘Parmenides and Odysseus', Harvard Studies in Classical Philology, 63 (1958), 133-43; A. P. D. Mourelatos, The Route of Parmenides (New Haven and London, 1970), and Coxon, Fragments of Parmenides, * After passing over the symbolic outer limit of the threshold, Odysseus answers Arete's ‘Who art thou amoung men, and from what family?’ question by disclosing the truth (7. 297 dAnOetqy xarélefa) and gaining Alcinous’ confidence, commitment (312 ff.), and (subsequently, at 545) affection. In order to know the Phaeacian men he ‘makes trial of them face to face’ (213). Eventually (9. 19) he discloses his name, lineage, usual description, and home town, Odysseus’ gradual entry and acceptance by those in Alcinous’ household foreshadow in many ways his reunion with the members of his own family. See further T. Van Nortwick, ‘Penelope and Nausicaa’, Transactions of the American Philological Association, 109 (1979), 269-76; and Marylin A. Katz, Penelope’s Renown (Princeton UP, 1991), 136 ff. For the symbolic importance of the threshold see C. Segal, “Transition and Ritual in Odysseus’ Return’, La parola del passato, 22 (1967), 321-42. really is Odysseus (23. 70 ff.). But rather than take this vital matter simply on their say-so, she insists she will be able to identify her husband through some special ojuara: yvwoöued’ GAAjAwY Kai Adiov- fari yap quiv ajpad’, à by Kai voi kekpuupéva iSpev dn’ GAAwv. (23. 107-10) ... but if in truth It really is Odysseus and he has come home, I assure you heart, and vóos to vóos.** become more complicated. Initially, he just tells Telemachus straight out who he is (16. 188). A little later, the old hound The conversation between Penelope and Odysseus displays the same features. Eurycleia and Telemachus are dismayed when they are unable to convince Penelope that the beggar in the hall ei à’ Ereöv 89 ¿o7' 'Odvoeds xai olxov ixdverat % pda vat (cf. Od. 1. 3). As early and superficial forms of contact give way to increased intimacy, shared experiences, and probing enquiries, Odysseus comes to know the true intentions and identities of those he encounters— initially face to face, but in time heart to When Odysseus sets foot on Ithaca his interpersonal discoveries 29 Tune We will know each other fully, for we have Signs we know about, hidden from others. Odysseus invites a probing examination (meıpdleıw ¿uédev) as a way of identifying him more clearly (114 ¢pdcera: xaì áperov), and she responds (181 weıpwuevn) by asking Eurycleia to place the bed outside her bedchamber and to spread it with coverings. When Odysseus reacts angrily to the suggestion that this unique bed could be so easily moved about, Penelope gets the oÿua she had been waiting for: ‘Rs daro, ris $ ayroú Auro yovvara Kai gidov Frop, ajar’ ävayvoior rd ol éureda mebpad’ ‘OSvacets. (23. 205-6) So he spoke and her knees and own heart were loosened, As she recognized the sure tokens Odysseus had told. Having seen, tested, and come to know véos the wide world over. Odysseus returns to the woman who proves, by successfully

31 inno. rai pe hépouoiv, daov 7’ Ei Bupòs ikdvor, putting the master tester to the test, that she does indeed excel all , > # ,, o x CI A 14 meunov, ered 1’ és dddv Bioav moAubnuov dyovaat others vdov «al érippova pyri. Three characteristics of the knowledge achieved in these famous scenes” are worth our notice. (1) For some, the recognition of Saipoves, 7) kara mavr' arm! deper ciddra dara: Odysseus has come directly in the perception of his general The horses that bear me as far as mind"? might reach, physical appearance or a distinguishing (visible or palpable) mark; but for others it has come through verbal assurances; and on occasion, when these were either unavailable or insufficient to produce conviction, it has come from tell-tale indicators acquired through a trial or testing process.”” (2) In each case the discovery of the truth has been tied, on the objective side, to the disclosure of a completely persuasive indicator of his identity,*? and (3) on the subjective side, to a profound ‘change of heart’, i.e. becoming convinced that the stranger at the door is in reality the lord Were conveying me when they set me on a richly songed TH depounv: ti) ydp pe modigpaato: dépov inmor... road*? Of a goddess, who bears the knowing man among all cities. (‘Studies in Parmenides’, Wege und Formen frühgriechischen Denkens (Beck, 1955], translated and reprinted in mind or spirit which has been transported by them, and to a special ‘pathway of song’ along which he has travelled, all portend that profound knowledge is about to be imparted:* * The same features will mark Odysseus’ final recognition by his father: a promise to tell the whole story (24. 324); then a request by Laertes for a clear oía in order to be sure (329 ddpa reroi0w); then the disclosure of the scar (331) and other tell-tale remarks about the garden Odysseus had known as a child (336); and finally, Laertes’ loosened knees and heart ‘as he recognized the sure tokens’ (3445). Cf. also 19. 250 omuar’ dvayvodan. , , . E. Allen and D. J. Furley (eds.), Studies in on any affinities with a Homeric forerunner; indeed, at one point he claims that Parmenides converted ríoris to a novel, epistemic use. ® I adopt the familiar but still problematic reading xaré dvr’ dorn. Coxon (following Heyne) proposed d(v)rn(v) for what looks like ärn in Sextus, translating ‘through every stage straight on’ (Fragments of Parmenides, 158). Coxon cites Il. 8, In the proem of fragment 1 Parmenides relates how a youth (presumably himself) was at some point in the past transported perhaps even supernatural, location. References to the powers of inspired thought which have brought him to this discovery, to the R. Presocratic Philosophy, ii (London, 1975], 5). Cf. KRS 244. Fränkel does not mention the peirastic character of Parmenidean knowledge, nor does he comment Odysseus come home. by a team of maiden-guided horses and taken to a remote, (B 1. 1-4) 399-400, but as R. Renehan has recently argued in his review of Coxon (Ancient Philosophy, 12 [1992], 395-409), neither there nor elsewhere does dvrn» appear to ES mean anything other than ‘face to face’, ‘directly in front of’ (which makes no sense in the context of B 1. 3). My preference for dorn, however, is based not on the references in Homer (cited by Renehan and others) to an Odysseus who sees the cities of many men (since Parmenides would not have wanted to endorse the idea of a knowledge acquired through wide experience of the world), but rather on the appositeness of the phrase as a reference to the poet who (typically) travels from city to city (cf. Xenophanes B 45, and for xard as ‘down among’ cf. Il, 1. 229, 2. 47, and commonly elsewhere). See further J. H. Lesher, “The Significance of ara mdvr’ ¿(o)ry in Parmenides Fr. 1. 3', Ancient Philosophy (forthcoming). * DK ‘Lust’, and commonly ‘as far as my heart or desire might reach’, but cf, Od. 21. 218, quoted above; also Jl, 10. 491 rà dpovewv xará Oupdy (‘thinking these things in his Guuds); Il 2, 409 fee ydp ward Bvpòv (‘he knew in his Buyds’); compare Archilochus' injunction to his Byués to know the unseen rhythm of things (fr. 67a Diehl). fuuds can also signify ‘conscious thought’ or ‘consciousness’ generally, as at /l. 13. 671-2: ‘his @uuds left him and hateful darkness seized him’, or Od. 24. 349, when his recognition of Odysseus causes Laertes to lose Buuds (for Suds as ‘mind, soul, seat of thought’ see LSJ s.v. II. 6). The ixdva at B 1. 25 ** Other cases of ‘peirastic knowledge’ occur at Il. 1. 302 el 8 dye pv weipyoat, iva yrdwa «al oide (‘but come, make trial, so that these will also know’; cf. 8. 18); Od. 8. 213 dM' éBédw Bpev ai repnóñueva: ävryv (‘but I want to know and make trial of them face to face’); 13. 335-6 cot 5’ où mw didov ¿ori Sarjueva: ovde nußeoda: | house of the goddess. mpiv y” Erı ons didyov reiproea: (‘it is not for you to acquire knowledge or learn until you have put your wife to trial’), Cf. also Theognis 125-6 ov yap dv eidelns .. . mpiv meıpndeins. a # Since ofua is cognate with the Indic dhyéma ‘thought’, it is probably no accident that the ofua provides the means by which things can be identified and known; see further G, Nagy, ‘Séma and Nóésis: Some Illustrations’, Arethusa, 16 % Lit. ‘richly endowed with song or voice’, Cf. Taran, Parmenides (Princeton, NJ, 1965), 10; Coxon, Fragments of Parmenides, 44-5. The use of the adjective moAv- to describe the pathway of song and the ‘much-indicating’ horses is echoed in the subsequent references to the ‘multiple contests’ during the elenchus (moAvÖnpıv &eyxor) and in the ‘very many signs’ (moAAd udAa) lying along the correct way of thought. These subsequent ‘decisions’ and ‘indicators’ are, then, the informative (1983), 35-55. © A view defended by Frankel: ‘What is depicted by the image of the ascent revelations alluded to at the outset of the proem. For roAugpaoro: as not so much ‘wise’ or ‘well-directed’ (horses) but as ‘rich in telling’ or ‘highly informative’ amounts to little more than this: wise, sunlike powers (the steeds and the maidens) carry and lead on a resounding path the man who knows and proclaims; the ones, cf. $paornp ‘teller, expounder’, gpaorixds ‘indicative, expressive’, and the common meaning of the root verb ¢pd{a—' point out, indicate, show, communicate, journey is dizzily swift and violent; and it is a great moment when the mighty gates of comprehension of the world fly open before the man selected by Fate scenes given above). o . suggests that his 9vuds has reached as far as it needed to when it has arrived at the etc.’—as well as its frequent use in connection with ojuara (as in the recognition

33 found reliable.” “The couplet thus both echoes the ‘peirastic’ On that way I was conveyed; for in that direction the much-telling approach to knowledge characteristic of Odysseus’ earlier encounhorses bore me... ters and also presages the éAeyyos or critical testing process And at B 1. 28-9 she assures the youth he will learn of through which the ‘it is’ way of thinking will be checked out and quèv ‘AAnbeins eúreibéos** àrpeuës ijrop dè Bporwv Sótas, rais où Evi miarıs dAnOns. found (completely) true. By the end of the proem each of the Both a steadfast** heart of well-persuasive truth As well as mortal opinions, in which there is no true trust. a trial or testing process which has to be endured, a very persua- Odyssey’s hallmarks of knowledge will have made an appearance: sive truth, and an unwavering or steadfast heart which has taken it in. The phrase drpepes rop is often assigned an ontological rather than an epistemological significance: ‘the inner being’ or ‘the heart of reality’. But that reading is constrained, unnecessary, and unwarranted. As is clear from the third feature of Homeric knowledge just mentioned, it would have been entirely natural for the goddess to characterize the knowledge the youth will achieve in terms of his heart’s unwavering acceptance of a fully complete. She refers to the account she has given as trustworthy persuasive truth. (mıoröv Aoyov), and alludes to the strength of trust (riorios ¿axús) Finally, in the proem's difficult concluding couplet, she explains: GA Eumns xai raòra pabycea, ws ra doxovvra xpñv Soxipws elvas 31d mavrös wavra mepwvra. (B 1. 31-2) But nevertheless you shall even learn these: how the things thought of Would have had to certifiably be, all things pervading through all. As Owen explained, Soxiuws elvar is ‘to exist in a way that is assured or reliable’, i.e. a way that has been put to the test and * As in Sextus, rather than the eëxuxAéos of Simplicius. The reading was first defended at some length by Mourelatos, The Pre-Socratics, 154 ff., and has been accepted by Coxon, Heitsch, and many other recent editors. In fragments 2-8 the goddess proceeds to tell the story (épéw, uúbos, Aoyos) of the way of thinking that is tied to truth (AAnGein) and persuasion (Zed). During the course of her critical examination** (as well as in her account of the cosmos in fragment 10) ‘very many signs’ (ojuara) appear, all indicating that what exists exists fully—uncreated and indestructible, whole, unmoving, and eETente and true trust (miorıs aAnôms). In short, as the éleyyos discloses onuara, the miorıs she had promised is achieved. Put them all together and they spell (Homeric and Parmenidean) knowledge: a thoroughly tested, reliably true, hence fully persuasive way of thinking about what there is. To the list of Parmenides’ contributions to Greek philosophy we should, therefore, add what might best be described as an adaptation of a familiar ‘peirastic’ paradigm of knowledge for use * G. E. L. Owen, ‘Eleatic Questions’, Classical Quarterly, Ns 10 (1960), 84102, repr. in Allen and Furley, Studies in Presocratic Philosophy, ii. 48-81 at 51. But I (now) think that Sca ravros rávra mepwvra refers to the way in which ‘all things pervade through all’; that is, to the extent that all things (named by mortals) could legitimately be said to be, they would all have to be 70 &v— which would entail a far more positive assessment of mortal 662 than Owen would have allowed. © drpepés—lit. ‘unmoving’, ‘not shaking’, but in a psychological context ‘unwavering’, ‘steadfast’, etc.; cf. 11. 13. 279 ff.: ‘for the colour of the coward is always changing to another hue, nor does he keep steady the spirit in his breast (ovdé of drpenas Haba ¿pyróer' dv peat Bupds)’; cf. Il. 16. 163 Oupds | orideow arpopòs Earı. " As Coxon concedes (Fragments of Parmenides, 168), #rop never refers to anything other than a faculty of thought or emotion, and no article appears in Parmenides' text. Plutarch (Parmenides A 34) appears to have understood that Parmenides was contrasting two states of awareness, one in which we grasp unchangeable intelligible truths, and one in which the mind ‘consorts with all ** Coxon, Fragments of Parmenides, 176, holds that the testing consists in asking the question whether the subject ‘is or is not’. But the ‘is or is not’ disjunction is not so much the test as it is a statement of the only possible outcomes for every test. The ‘critical moments’ or decision-points in the ¿Aeyxos manner of changes, accidents, and irregularities’. respect. come at B 8. 6-7 and 8, 19 when the goddess asks what birth one might seek for ro edv, or in what way and from what ro édv could grow, and at 8. 21-5, 8. 29-32, and 8. 33-5 when divisibility, movement, and temporal development are considered only to be repudiated. Since at each juncture rò eöv checks out positively (i.e. with ‘it is’ as still true), we acquire many omuara that rò édv exists in every possible

in the context of philosophical enquiry and reflection. But, having recognized this, we might also want to view Socrates’ denial of any involvement with Presocratic ideas about knowledge with some scepticism. At least when the Socrates of Plato’s early dialogues sets out to discover the nature of the virtues by putting a series of rival definitions to the test—hoping to find a Adyos that will remain steadfast throughout the entire process of examination— his approach represents not a repudiation of earlier views of knowledge,‘ but rather a continuation and extension of them.” University of Maryland * The claim that Socrates repudiated the ‘fantastically strong standards’ of Parmenidean knowledge was made by Gregory Vlastos in his ‘Socrates’ Disavowal of Knowledge’, Philosophical Quarterly, 35 (1985), 1-31 (see esp. pp. 17 ff.). 19 Cf. the conclusion of the Gorgias: ‘when all the other proposals were refuted (EAeyxouevwv), this one alone remains steadfast (uóvos odros mpeuet): that we should be more on our guard against doing rather than suffering wrong, and that before all things a man should study not to seem but to be good . . . ler us then follow the guidance of the argument (rw Aóyw) now made manifest, which indicates to us (omuaiva) that this is the best way to live’ (526Dff.). The contrapositive of ‘knowledge entails stable possession of a belief’ is the thesis that we cannot be said to have knowledge so long as our beliefs are unstable or moving about (cf. Plato, Euthph. 11 B-E; Meno 85 c-D, 97 D-98 8). The view that knowledge requires achieving an unshakeable hold on truth and reality appears at Crat. 437A; Sym. 211A; Rep. 534 B-D, 503A. Aristotle also describes the process of acquiring knowledge in terms of a stabilizing or settling down of opinion (cf. Post. An. 72°25 ff., 89*510, 100°1-13).