in Heraclitus, I Author(s): Martha C. Nussbaum Reviewed work(s): Source: Phronesis, Vol. 17, No. 1 (1972), pp.

1-16 Published by: BRILL Stable URL: http://www.jstor.org/stable/4181870 . Accessed: 06/12/2011 02:59
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

BRILL is collaborating with JSTOR to digitize, preserve and extend access to Phronesis.

http://www.jstor.org

V'YXH in Heraclitus, I
MARTHA C. NUSSBAUM

In

the fragments of Heraclitus*, frequent mention is made of 4uxiC; to understandwhat Heraclitus means by +uynwould seem to be central to any attempt to discuss his ideas about human life and death. In order to understand Heraclitus' own usage, however, we must first attempt to review the meaning of the word 4u ' as traditionally used, and particularly as used in the poems of Homer, of whose ideas and influence Heracitus is harshly critical.' It will also be useful to sketch since the pre-Heracitean history of the word ?o6yoq, its history parallels, in certain important respects, that of 4uZ', and since the two notions will be seen to be vitally related in Heraclitus' 4u i fragments. The first section of this paper wil deal briefly with the history of these words, and go on to investigate the role which, for Heracitus, +Uzl plays in the living man, and the way in which this role may be seen as dependent upon Heraclitus' ideas about language. The iuyC in Homer2is that which leaves a man at death to continue existence as a shade in Hades. Without +uX', a man cannot live; it is the single factor the presence or absence of which differentiates the living man from the corpse. But it is mentioned as present only insofar as it may potentially depart, and is thus the characteristic sign of human vulnerability and mortality. It becomes plain, upon examining the passages in which it is used, that the term 4uX, alone, implies the presence of none of those faculties which we would regard as characteristic of human life. Tu ' is a necessary condition for human life, for consciousness,thought and emotion; but it is not a sufficient condition. In speaking of his faculties, the Homeric man distinguishes a number of "organs"with separate functions and locations: Ou~tu,xVp,irop, voO4,etc. He does not refer explicitly to anything wich connects ypev?, them, or in virtue of which he is a single being.3
* For modern works referred to in this article, please see Bibliography on p. 15. 1 Explicitly in DK 22 A 22, B 42, B 56; implicitly, as I hope to show, in many other fragments. 2 See the thorough discussion in Snell, Discovery, ch. I, esp. pp. 8-12. Also Snell, Gnomon 77 ff. and B6hme passim. 3 See Snell, Discovery, p. 8.

The +uXC vaguely, a principle of breath; we know this from its is, etymology, and from the fact that when it leaves a man it is breathed out, flies away, or departs through a wound. But its function in the living man remains undefined, and it is mentioned only in connection with death. When one's life's breath leaves, it is called 4uXf But when a hero wishes to say "as long as the breath of life remains in me", he
says:
eL5 6 xX'Otrp- / 'v a
sacaC. ZAv-1 (K

89-90). A hero may fight

p't

4uXyj (X 161) or risk his su

in battle (I 321) or discourse about the

irrecoverability of the ~iqu once it is lost (I 408). But he is never

aware of doing anything by means of it in life; only once is it mentioned as being present in a living man at all. And the sole point of mentioning its presence there is to declare the man mortal and vulnerable.Agenor, attempting to persuade his 9,uLOnot to fear Achilles, says (D 569):
xM yocp 0h2v Tovx eV 8i 'LOX UX
9tu~evoca. tputOG xjpw O;t

&V e' pa.' &V?5pG

xmax&, 7toL

Each mortal man has within him a single 4uv', and it is this which characterizeshim as mortal. The gods are not credited, either in Homer or in subsequent literature, with the possession of su 4
4 This seems to hold true even after the word 4uX' has acquired a wider usage or and can, on occasion, be substituted for Ou[L6q, even for v6oq. In the nonphilosophic literature of the fifth century, +uZ' is ascribed in only four cases to a god. Aristophanes, Frogs, 1468 - puataocat yxp OV7rtp i +uXA O0,et - seems for metrical and contextual reasons to be a quotation from tragedy, and the fact that it is Dionysus, a god, who quotes it may be yet a further dimension to the r' joke. In line 1472, Euripides addresses Dionysus as X [LCapc awv,&pw7rwv. The Oceanids (Aesch. Prom. 693) are characterized throughout more as females than with q6XeLvmay have as immortals, and the poet's desire to juxtapose Puxn allowed him to admit this irregularity. A more startling exception is at Pindar, Pyth. 3.41, where Apollo speaks of enduring in his iuZ. This must be understood, I think, as connected with Pindar's ideas concerning the fundamental similarity of gods and men (Nem. 6.1 ff.) and the divinity of +uX' (frs. 116, 127). The fourth exception is Euripides, fr. 431 (following the attribution by Clement, though Stobaeus attributes it to Sophocles): gpC, y&p&vgpmcov 6vouq lnipXeaL
xaxl a' Ac yUvatxac, ,&XX& &e&v&vc

n6vrovgpXeroct. rapraact, x&TrI X4~ux& For several reasons I believe the fragment to be Euripidean. Stobaeus mentions a Phaedra play of Sophocles. Could it instead be the first Phaedra of Euripides, which so shocked contemporary sensibilities? And could this not be one of its more shocking lines? In any case, though absence of context prevents sure interpretation, it seems worth noting that this humanization of the gods takes place in connection with gpw,, which, for Heraclitus too, is, next to death, the prime disturber of the 4uX' (see our Part II, later in this volume). To be vulnerable to

Two elements in this Homeric picture will be of importance to our discussion of Heraclitus: the tendency to mention 4uxn only in "negative contexts" (i.e. its functioning is noticed only when it goes wrong), and the absence of any sense of a central faculty connecting the disparate faculties of the living man. Snell has shown that the same is true of the Homeric notion of body: it is spoken of as a collection of parts, and there is no word to describeits unity.5 We shall now observe that the Homeric picture of language presents similar features: MOC, frequently mentioned, are a series of unorderedunits.6Aoyo4,implying order or connection, is employed, as is fuX', primarily in "negative contexts." And there is no sense of a central faculty by virtue of which one learns to use language properly.7 in Guthrie, discussing the meaning of ?o6yog Heraclitus, outlines "the ways in which the word was currently used in and around the time of Heracitus."8 Now it is true, accepting Guthrie'sown dating, that most of the meanings he lists are current within a generation or two after Heraclitus' death. Several examples he cites - those from early works of Pindar - even fall, probably, within his lifetime, although it is unlikely that he ever became acquainted with Pindar's work or with the social milieu in which he worked. What is interesting, however, is that if we examine the works of those writers known to have been read by Heracitus (Homer, Hesiod, Archilochus, Xenophanes), as well as those of other poets distinctly prior in date (Solon, Theognis, etc.), we find, instead of an impressive diversity of usage, a singular unanimity. A6yo4in early writers is not used frequently. When it is used, it always means a story, or some sort of connected account told by a specific person. And, in the vast majority of cases, this account is a falsehood, a beguiling tale, one which is intended to deceive the hearer or to make him forget something of importance." Common formulae with Xoyo4
Ipc; is almost to be vulnerable to death. Elsewhere the notion of mortality remains central to the notion of 4uyX. Cf. also Plato, Phaedrus, 246 c 7 ff. 5 Snell, Discovery, pp. 5-8. 6 Cf. Fournier, pp. 211-12: "Dans Homere gTrog etait un vase vide de pens6e: il pr6sentait les paroles prononc6es comme un fait, un objet, ou un instrument... Il Il peut etre compl6ment de 3azeXv. est utilise comme arme. . . C'est quelque chose de passif et d'inerte". 7 On the history of X6yoq,see also Hoffmann; Verdenius, pp. 81-82; Boeder, p. 85; Fournier, pp. 53 ff., 217-219. 8 Guthrie, pp. 420-424; also Kirk, HCF, p. 38. 9 Cf. Boeder, p. 20: "Jedesmal eignet diesen ),6yot etwas Beruckendes und sie gelten daher als mittel der Bezauberung und Ablenkung und Irrefuhrung. Sie

are: 4i8ee& Ca
6yOLaL (cX 56,

cljiuxouq TE xo6youq (Hes.

Op.78,

788) and oc'IpMoLaL

H. Herm. 317, Hes. Th. 890, Theognis 704; cf. also 0 393; Theognis 254, 981, 1221). Less frequently, Xoyoq designates simply a tale or story, without regard to content; a legendary nature, however, is often suggested (Hes. Op. 106, Tyrtaeus 9.1, Theognis 1055, are Xenoph. B 1, 14; B 7, 1). Ao6yoL personified in the Theogony: hateful Eris gives birth to NeLxs& re +eu8aex xr Aoyouq ApXoytcLoyx Te (229). And the only mention of an explicitly true ?6yoqin the works of the authors mentioned is in Archilochus 35, at line 12: A6yo[Lv]uv We '[60' oik}]M' 7rop[%. are reminded that the word for truth itself in Greek is a privative term, designating the absence of concealment, or "'uncoveredness".10 A6yoqby itself does not mean only falsehood, any more than +uX' means merely the shade of the dead man. But men's notions of language at this time appear to have been simple. As a man speaks of his life in terms of separate faculties, but fails to notice their unity until this unity is destroyed, so he speaks of his words as separate units, and usually refers to connected statement only when the connecting has been done improperly, so as to produce a falsehood. When the impressive effect of Odysseus' speaking is described,his words (9nza)are compared to wintry snowflakes (P 222) - apparently an early case of semantic atomism. And T 248 ff., a section of Aeneas' speech to Achilles, gives us an even clearer idea of the Homeric man's lack of awarenessboth of syntax and of the mental processesby which language is learned and understood:
atp~7r'

8& yeaa'

?OL

7, i7OXe POTrV3p

8 9VL [0d&OL

?7t?()V & 7UOkU' CV'rQoOL), VO[L6q9VIXOX 6,50C to6Ovx'e7r0xouaLacL. 07r7r0Z6Vx'e'UMa&N97rOq,

The tongue, herding the single words here and there like sheep, is the only acknowledgedlinguistic organ, both here and elsewhere.And the general effect of Aeneas' succeeding arguments is to demonstrate the heroes' contempt for speeches."l Speaking requires little skill; one
verschleiern, was im Blick bleiben sollte, schieben anderes vor und stellen es in ein giinstiges Licht." 10 Krischer, pp. 163-4, and Boeder, who reminds us that the false x6yot of the Theogony are sisters of Lethe. 11 Benardete (p. 2) has observed that this contempt for speeches is evident in (T the hero's description of speeches as the work of &v&pwtrot 204, etc.) and of deeds as the work of &v&pe5(I 189, etc.). "The hero's contempt for speeches is but part of his contempt for &v,p7o L."

simply says the sort of words one hears, and anyone, even a woman, can do this (T 252). Fighting, not speaking, is a genuine test of skill, and proves the hero's worth.'2 There is no recognition here that the learning of language involves anything more than the ear and the tongue. And the Homeric man has no term at all for any faculty of connected reasoning; even voos, generally considered a "rational" faculty, has been shown by von Fritz to stand, in Homer, for a nondiscursive, a-logical faculty of insight, more nearly similar to the sense of sight than to what we think of as "reason".13 In general, then, Homeric man fails to recognize explicitly that in virtue of which he is a single individual. His use of the first person shows that he is conscious of the self, and that he is somehow aware that his limbs and faculties form a unity. But he cannot explain what connects his separate faculties; and though he implicitly acknowledges the centrality of +u ' as a necessary condition for consciousness, he has not yet acquired a notion of its activities and its role. His understanding of language reveals similar limitations: he is aware of words rather than of syntax, of the ear and the tongue rather than of the discursive reasoning and connection-making necessary for the proper learning of language. Heracitus sees deficiencies in this view, and attempts to formulate a more complex picture of human life and language, conceiving the role of 4uXn as that of a central faculty connecting all the others, and ascribing to it the power of connected reasoning and language-learning for which his predecessors have no explanation.
12 This passage, and others like it, seem to indicate that Kirk (HCF, p. 32) insists too emphatically upon the parallelism of word and action in early Greek thought when he denies the existence of a "sharp distinction between the two, at any rate until the development by the sophists and rhetoricians of the contrast . . ." His analysis assimilates Heraclitus to his predecessors, ),6yoq-9pyov and ignores certain definite distinctions between the Homeric and the Heraclitean views of language and its relationship with action. Cf. also Verdenius' remarks (p. 97) on early examples of the X6yo4-9pyov contrast. 13 Von Fritz, "NOOE in Homer," pp. 79-93. Also von Fritz, "NOON in PreSocratics," 225ff., where he says: "There is absolutely no passage in Homer in which this process of reasoning is so much as hinted at, when the terms V6osand vo?tv are used. On the contrary, the realization of the truth comes always as a sudden intuition: the truth is suddenly 'seen."' D. G. Frame's thesis, The Origins of GreekNOOX, contains an excellent study of the etymology and semantics of v6oqwhich, while supporting von Fritz's general conclusions, provides a much more likely etymology and a far more precise and thorough study of contexts and associations in the early epic tradition.

The prominencewhich Heraclitus gives to 4uXqamong the faculties of the living man can clearly be seen in those fragments in which iuXy takes the place of fire, the central element of the cosmic Xoyoq, the in of 6Oyo4 human existence (36, 77), and in many other fragments as well. Fragment 67a is not indispensable for an understanding of Heraclitus' +uX' doctrine; and if one denies its authenticity, one must still, I think, admit that it is in no way inconsistent with what we learn from other fragments. But I believe, with Pohlenz, Diels, Kranz, that it is genuine; and since Kirk, and others, and against Marcovich,14 it presents with greater clarity than any other single fragment the central role of tuxy in the living man, it seems appropriate to discuss it briefly before moving on to a consideration of specific aspects of this central role as they can be seen in other fragments. Fragment 67a is cited by an early twelfth-century scholiast, Hisdosus, in his commentary on Chalcidius'translation of the Timaeus (Codex Parisinus Latinus 8624), with regard to 34b, the discussion of the world-soul's position in the center of the cosmos. The scholiast remarksthat there are others who believe the sun to be the center; just as the anima, with its seat in the heart, diffuses energy throughout the limbs, so the heat proceeding from the sun gives life to all living things:
cui sententiae Heraclitus adquiescens optimam similitudinem dat de aranea ad animam, de tela araneae ad corpus. sic(ut) aranea, ait, stans in medio telae sentit, quam cito musca aliquem filum suum corrumpit itaque illuc celeriter currit quasi de fili perfectione15 dolens, sic hominis anima aliqua parte corporis laesa illuc festine meat quasi impatiens laesionis corporis, cui firme et proportionaliter iuncta est.

I believe that the general content of this simile can be attributed to Heraclitus, although I would not agree with Kranz that every particular is likely to be genuine. Kranz' argumentsfor verbatim acceptance are largely based on the dubious analogy with Tertullian, de anima, 14,
14 Pohlenz, p. 972; DK I ad. Ioc.; Kranz, pp. 111-113; Kirk, Archiv, p. 76; Marcovich, pp. 576-9. 15 I have retained the de fili perfectione of the manuscript rather than accepting Diels' persectione. Persectio seems to occur nowhere else; and the usage de aliquo dolere with the meaning "to mourn for" or "to grieve over something which is no longer" is perfectly good Latin (e.g. Cic. Att. 6, 6, 2; Hor. Ep. 1, 14, 7), and is an easy and natural extension of the ordinary meaning, "to grieve over." Thus I would translate, "grieving over the wholeness of its thread (which is no longer, whole, being broken)."

an analogy which Marcovich uses to prove just the opposite point, and one which, as I will show, is confusing rather than enlightening. In addition to the arguments cited by Pohlenz in favor of the ascription of this simile to Heraclitus (the absence of such a reference in Chalcidius, the scholiast's general awareness of ancient literature), I would point also to the consistency of the fragment's general import with that of other Heraclitean fragments, and to the interesting suggestion Hisdosus gives in cui sententiae . .. adquiescens that his Heraclitean source contains some sort of a further analogy between solar energy That Heraclitus draws an analogy between cosmic fire and and fu the Puxn' the "fiery" element in man, is certainly an accurate observation, and is one of the only points about Heraclitus' theory of 4uyI on which there is general agreement. Such an observation would not have been possible, I think, for a man who knew Heracitus' work only through the pneumatic theory of the Sceptics. And yet this, Marcovich alleges, is the case with Hisdosus, although he gives us no concrete reason for rejecting Pohienz' assessment of the scholiast's knowledge of the ancients. Indeed, the arguments used by Marcovich to deny the fragment's authenticity are based on highly dubious analogies with later doctrine. The fragments of Straton of Lampsacus (110-111), which he considers an indispensable basis for the spider analogy, give us an extremely passive notion of 4u ' as the seat of perceptual 7rah', drawn to that If Hisdosus were familiar with these fragments, and ap' o6 7t6tovl. were constructing his "Heracitean" simile on the basis of this familiarity, it is difficult to see how it could have occurred to him to mention the simile in a passage of commentary relating views concerning the energizing and active properties of Qu Moreover, Straton gives no account of Puxas microcosm, which Hisdosus' source apparently does. Nor does Marcovich argue convincingly when he declares that Hisdosus became familiar with Heraclitus through the Sceptics and their pneumatic-diffusion theory of vu As an example of this theory, he cites Tertullian's de anima 14,5, which Diels and Kranz cite in support of the fragment's authenticity. This passage, which attributes to "Straton, Aenesidemus, and Heracitus" an analogy between 4uX and Archimedes' water-organ, obscures the issue here more than it clarifies it. Tertullian is rarely reliable when he paraphrases; and his ascnption of doctrines to a group of people of widely divergent dates and views is a confusing and inaccurate polemical expedient. This pneumatic doctrine bears even less resemblance to fr. 67a than do the

fragments of Straton. Hisdosus does not describe the Heraclitean +u as generally diffused, a breath rushing through the body; the implication of the whole passage is, rather, that 4ux' is analogous to the sun, a central, definitely locatable, possibly fiery, source of energy. If Hisdosus used the sources Marcovich describes, it is amazing that he should have produced a picture of suX' which is so much closer in its general import to other known fragments of Heraclitus that it is to the alleged sources themselves. It is also worth remembering that Hisdosus claims to quote directly from Heraclitus, and seems to have no polemical point in view, as does Tertullian, which might cause him willfully to present a false picture. Even Tertullian, however, quotes correctly in the single reference to Heraclitus where he actually claims to quote (de an 2: fr. 45). We can, I think, conclude that Hisdosus' source was one which did not assimilate Heraclitus to the Sceptics, and which was generally more faithful to his words than those "sources" upon which Marcovich claims the fragment is based. We may accept as Heraclitean the comparison between spider and +uX', though not necessarily in all its detail, and the suggestion of a further analogy between 4uX' and cosmic fire. What, then, can we learn from the spider-simile itself about Heraclitus' notions of +uXy 's role in life? First, that it is the central lifefaculty, upon which the others depend, and through which they operate. A man is not, as in Homer, a loosely-joined collection of limbs; he does not react to one sort of stimulus in his 4upok, and to another in his 9pe'v. All stimuli are referred to 4uy', which holds the body and its faculties together. It is an active faculty, and not merely the seat of the -ta't. It responds to stimuli, but these stimuli remain external to it. It is locatable, and not generally diffused, but its location is not fixed, as it is in the very similar analogy of Chrysippus (II 879 A). Unlike Tertullian's flatus in calamo, it is apparently self-moving, and capable of directing its movement. And there is the further implication that, since it is the essential animating force in the body, nothing done merely to the body will be sufficient to produce death. If the simile is accurate, we will expect to see death explained as something which happens specifically to 4uXn, and neither merely to the body, nor to the creature as a whole. And, indeed, this prediction is borne out by the evidence of fragments 36 and 77, as we shall see. It would not be wise to push the details of the analogy further. The interesting suggestion, embodied in firme et proportionaliter, about the

relation of +u ' to body cannot be accurately assessed in translation; one supposes that proportionaliter translates some phrase involving but it is hard to know whether the translator has grasped the XO6yoq, in appropriate sense of the word 6Oyoq its context; without the original, one cannot judge. We see, then, in general, that +uX( is the essential life-faculty in the living being (in men specifically, if we accept hominis anima as genuine). Two gaps in the Homeric picture of man are filled simultaneously, as Heraclitus accounts for the life-activity of that which was previously noticed only in terms of death, and also accounts for the unity of the individual, who had been described as a collection of limbs and faculties. Now we will see that his 4utxdoctrine answers yet a further need: the need to account for language, and the ability to use it, in some more satisfactory way than had the epic poets and their followers. Fr. 107: xocxo' p'upA V9p@tOlaLV 4pOL XMX &rA papous The syntax of the participle CxOvtwv has been adequately explained as similar to that of certain Homeric examples discussed by Classen,'8 and Marcovich, Kirk, and Guthrie17 are probably right in understanding it as conditional, not causal. The central problem in this fragment lies in the interpretation of the word Pappapouq. It has become usual for commentators, following Diels, to adopt a metaphorical interpretation: the "barbarian" PUxocE are incapable of understanding the "language of the senses".18 Now it is clear that, to a certain extent, this is right; it is not, for Heraclitus, the senses which are deceptive, but the +uyact which err in interpretation. But an understanding of 5apf3pou4 as referring to a "language of the senses" is, first of all, more metaphorical than necessary, and, secondly, philosophically uninteresting, since it tells us nothing about why or how the 4uycd make their mistake. A more literal interpretation will both be truer to contemporary usage of the word, and will also give far more insight into Heraclitus' theory. The word Ocppapos the early fifth century means only one thing: in not speaking (or understanding) the Greek language. Thucydides (I. 3)
Classen, pp. 174-5. Marcovich, p. 47; Kirk and Raven, number 201, p. 189; Guthrie, p. 415. 18 DK ad. loc.; Kirk and Raven, p. 189; Marcovich, p. 47. Compare Holscher's translation (p. 138): "wenn sie verworrene Seele haben." Verdenius' translation, "wenn man die Sprache der W6rter nicht versteht" (p. 98), is much nearer the mark, though perhaps unnecessarily tricky.
16
17

observes that the word was not used until a sense of Greek unity had arisen, with reference to which fO'poCXpoL, could be defined as a too, single group. Liddell and Scott remark that it is not until after the Persian war that the term even began to be extended to mean generally rude or uncomprehending.In the period which concerns us it refers only to linguistic understanding, and has no implications of general inability to comprehend. Aeschylus in the Persians, has the Persian characters of the drama refer to themselves as r3&prapoL 337, 391, (255, etc.), a fact which proves that even shortly after Heraclitus' time the term still retained a very narrow sense, and was not pejorative in connotation. Thus it seems that Heracitus is not making the rather bland and vague statement that sense-data are deceptive if you have a crude and uncomprehending4uxB, but the far more interesting statement that your senses will deceive you if you do not have an accurate understanding of your own language. Addressing his remarks in the Greek language to a Greek-speakingaudience, he seems to say, "You divide mankind into two groups, Greeks and fpo3xppQL, describing yourselves as those who understand the Greek language, and all other men as those who do not. But I intend to show you that, in fact, you have no better understandingof your language than those whom since you call r3&pcxpoc, linguistic understandingis dependent not only upon birth in a certain locality, but also upon the proper training of the 4ux ." We noted that the Homeric man seemed unaware of the processes by which he learnedand understoodlanguage. He spoke what he heard; nothing that happened between the ear and the tongue was of essential importance. Heraclitus insists that sense data must be referred to +UXI,which is capable of understandinglanguage; only then will the process be complete. The meaning of this emphasis on language in connection with perception is not immediately clear. And yet it is evident that for Heraclitus, who so frequently contrasts men and animals (frs. 4, 9, 29, etc.), who places great emphasis on learning how to understand the connections among things, for whom wisdom consists in speaking the truth as well as in action (fr. 112), the way men understand their language is of central importance; and their errors in understanding the nature of the world can be understood, from a slightly different viewpoint, as failures to understand the structure of their language. It is, I think, in failures to grasp language that Heraclitus tends to the essential find the source of the more general error, and Moyoq,
10

meaning of which up until his time has been connected statement, is that with which he thinks one must deal first, if one is to understand in ?O\yoq its wider, cosmic sense. Menare misled by their atomistic conception of language. Fragment 1 shows us that men's ears are bad witnesses so long as they remain on the level of 9nea and fail to graspthe ?o6yo4;they are OCiDvvrot,not making connections.The a'UvvoLtry out e`nea and ?pyo; their words are detached is units. To grasp single 6`aoc without grasping Xoyoq futile; their ears pass on Heraclitus' words, but they understand no better than before
they heard. The
&'VUZvOr,as

we see from fr. 107, are going wrong be-

cause their iuvxL do not understand Greek. Mere hearing does them little good. The sorts of error men can make about things through an inadequate understanding of language are frequently discussed in the fragments.'9Two examples must suffice here. In fr. 23, he says: "They would not know the name of Dike if such things did not exist." Such things, presumably, are unjust things.20 The word Dike, which men consider as the name of an independent goddess, could not have meaning except with referenceto its opposite. If injustice did not exist, men would have no reason to speak of justice. And yet men, seeing words as 97re, miss this essential connection, this interdependence of words. Fr. 57 gives us an even clearerpicture of popular error: Heraclitus declares that men think Hesiod knew more than anyone else; but he did not even understand day and night; for they are one.21 Now if we look in the Theogony (123-124), we find that Hesiod makes Night the child of Chaos and Erebos, and Aither and Day, in turn, the children of Erebos and Night. Hesiod did not see the illogicality of making night older than day; his genealogy cannot explain how men use the words "night" and "day"; for in ordinary usage they are interdependent and are defined with reference to each other. Thus, Hesiod's misunderstanding of words leads him to make a mistake in
19 Cf. Snell, Hermes.
20

Clement (Str. IV, 10) cites this statement of Heraclitus together with a remark of Socrates that law does not exist on account of the good. The end of the preceding chapter has discussed the relativity of opposites, and how good can, in some circumstances, come from something we generally consider bad. It seems safe to assume that Clement understood 3roaxo refer - if not to unjust to things in general - at least to crimes of some sort. Even if the opposition is not between justice and the unjust, it is probably between judgment and crimes, and a similar point can be inferred. 21 Cf. Ramnoux, p. 1 ff.
11

cosmogony, and he gives a false analysis of the nature of the visible world. There is another fragment which I believe has been misunderstood by some commentators, and can be understood more precisely in

connection with fr. 107. This is fragment 34:

X&WpOtaLV rOLXxavL Cp&TL4 CXTOTLVV

&(vvToL

&xoOixav-t0e

aTCp&here has usually been translated as "proverb."22 But the proverb in question is found by Marcovich nowhere earlier than Aristophanes' Knights 1119, where it is not clearly proverbial, and may equally well be a clever paradox set forth by the comic poet. The use of the simple verbal noun to mean "proverb" is nowhere else attested. This fragment is the only evidence given by Liddell and Scott for the existence of such a meaning. Usually, cpacx-, means simnply "what is said": report, rumor account, speech, and even language, as in Aes. Ag. 1254:
rCOxpeOvVTCX MiLZvCaL. optpTVpEL

'EX-nv'

7TLC[OGCaL YpXTLV.

In the light of the connection made by fragment 107 between successful hearing and linguistic understanding, I think we can understand cpnrt here in its more basic sense. Thus: "People who fail to make connectionls, when they hear, seem like deaf people. What they say (or speech, in their case), bears witness that although they are present, they are would be simply an ethical dative, and the paradoxical absent." ocuaroLv atreLvat a Heraclitean juxtaposition of opposites. (It did not =peovraOq seem likely that Heraclitus would have cited a proverb in support of an attack on the 7no?XA.The style seems peculiarly his own.) With this interpretation, the fragment reinforces what we know from frs. 1 and 107 about the interrelationships of hearing, understanding, and speaking. (Fr. 19 suggests a similar observation: &xoiaoL oux e'MGOawVOL oWr'?s=xZv.) Men do not learn language, as many think, simply by hear-

ing and speaking. If they remain ajV'rToL

if they have iuyjxL that

are incapable of making linguistic connections - they are no better than deaf people, since what they say shows that they have not understood what they have heard any better than if they had not been there to hear it at all. To tell whether someone understands the words he hears, we have to see whether he knows how to use them himself in connected statement. A failure to do so is evidence of a
PapDapoq +?XnT-

So far, our elucidation of fr. 107 has been confined to the issue of misunderstanding of language, and the examples of bad witness have
22

Marcovich, pp. 12-13; DK ad. loc.; Holscher, p. 139; Ramnoux, p. 38. But Kirk (Archiv, p. 75) understands the paradox as Heraclitus' own.

12

been concerned mainly with the mis-hearing of words. We have seen that it is perfectly correct, according to Heraclitus, to say of most Greeks that they are really 3cippapoL. And we have seen that a misunderstanding of words can lead, as in the case of Hesiod, to the formation of a false cosmogony, and thus that false hearing can result in false interpretation of the seen. But I believe that the critique implied in fr. 107 does not stop at correcting a false picture of language; I believe it also asserts the centrality of language in all human learning, and the importance of speech for the proper interpretation of any sense data. We need not confine our interpretation to the mis-hearing of words. In a very fundamental sense nothing seen or heard can be fully understood without language, since it is in learning and communicating through language that a human being learns to relate and judge that which he perceives. In fragment 3, we have a classic example of false witness of the eyes: the sun appears to be a foot wide. Now, borrowing Aristotle's analysis of this same phenomenon (deAn 428 b 2), we might say that this immediate and false picture is an "appearing", common to all animals. But the learning of the true account would be peculiar to man, and involves connected reasoning and language. Thus, in any case of perception, to move from appearance to a true account we need language. We cannot conceive of connected reasoning existing without it. This short account cannot do full justice to Heraclitus' ideas about language; it has elucidated that question insofar as it bears on the nature of the Heracitean 4uX. To summarize, then, our conclusions about fr. 107: It is with 4uXi, the central and connecting life-faculty, that man may potentially understand XGyoq, connected discourse. or Because of the central importance of language in understanding, the central life-faculty in man is, first and foremost, a faculty of language.23 Sense data are referred to +u and are interpreted according to the fUn 's degree of linguistic competence. All mortal living creatures, one

would suppose, have JuZj; only in human beings can that Pux' grasp
The term X6yo4is not restricted to the meaning "discourse" or ?o6yoq. is "statement", just as PuxZ not only a faculty of language. Both are used in a variety of contexts. And yet the fundamental importance of language in Herachtus' analysis of cognition cannot be denied. In his use of the term vwoos, insight - which has in itself, no linguistic or overtones - he gives it such overtones by introducing the verb Xeyc:
23

Cf. Hoffmann, p. 7.

13

in fr. 114 we may say that tuv vo6)Xyovto means much the same as using a Pux4 which speaks the language. v6o4,or insight, becomes a quality of the well-functioning 4u gained as a consequence of linX, guistic understanding; and v6o4is probably emphasized in 114 chiefly for the sake of the word-play with iuv6v.The fact that the two terms v6oqand +u are never juxtaposed led von Fritz24to ignore the role of svxmaltogether in his analysis of the role of v6o4.It seems clear, however, that for Heracitus su ' is far more central, and v6o4becomes subordinate to it. We see, then, that Heraclitus, taking two terms whose meanings before his time had most often been negatively defined in explicit use, has given them central positions in his philosophy: suZj as the connecting and knowing faculty, Xoyo4 the primary object of knowledge. The as creative nature of his achievement can most clearly be seen against the backgroundof non-philosophicthought, the scope of which Heraclitus seems deliberately to revise and to enlarge. Insufficient contemporary evidence makes it difficult to determine how much of his reassessment of ;k6yoq +uyJ3 in fact, his own. If he is really the first to have and is, accorded these terms such central importance, we may then be forced to ask ourselves whether he predicted the fifth century or created it; for no two words take on greater importance in the thought of succeeding generations (a fact which often has led, I believe, to a failure to appreciate the originality of Heraclitus' theories). I think we cannot assume that the novelty of Heraclitus' usage was as great as the available evidence might lead us to suppose; certainly the sensesof iuX' and ?o6yog which he developswerein some way inherent in the senses of the words as earlier used, and it is the contexts in which it occurredto earlier men to speak about 4ux( and ?O6yog, rather than what men meant by the words themselves, that were formerly restricted. Then, too, it is clear that other subsequent uses of the words in "positive contexts" cannot all be dependent upon an understanding of Heraclitus; the evolution in their usage must have been under way
24 Von Fritz, "NOOE in Pre-Socratics," pp. 230-236. Frame's study shows that the role of iuXp as life-principle cannot be ignored in any profitable study of v6oq, since its earliest traceable Indo-European meaning is that of a faculty by means of which one returns (the root being the same as that of vio,Lot, v6a-roq, etc.) from death to life and light. One wins, or keeps, one's +uX' by exercising is originally v6o; (cf. x 5: &pv.LCevoq~v -r 4uXyv xxl v6a-rovkX(pcav). v6oq, like u connected with the threat of death; and in making suxi explicitly central, Heraclitus clarifies what it is that a man's v6oo is to accomplish.

14

by the time he wrote.25But it does seem likely that he is the first Greek thinker to have elaborated a philosophicaldoctrine of +Uxn,and to have connected with the life-operations of +u7' the notion of linguistic understanding. In concluding this section, we ought to deal briefly with the two fragments which speak of a X6yoq In YuyxG fr. 45, Heraclitus declares, tv 6 OV. +Ux~qMrEPpa-a ou'xav ?'eupOLO, 7&aCv tL7top?UOSU6VOq OUTGr Px1hmv
XOyov gZeL. And, in fr. 115: iu(q i-t
Xo?yoq Cut6v

cu`@ov. These frag-

ments can both be interpreted in a variety of ways; I would like, however, to see them as a further reference to +ij 's capacity for learning, and to the central importance of language in all such learning.26 Fragments 101 and 116 emphasize the possibility and the importance of self-development for all men. Fragment 45 would, then reinforce the declaration of 116 that development is possible by saying that there are no limits to man's power to develop his understanding, of such basic importance is the innate linguistic capacity with which his +uynis equipped.27Because his fuxZhas the power of ?oyo4,there are no limits to its development; this capacity increases itself, for the more one understandslanguage, the better are one's tools for increasing that understanding. [To be concluded] Harvard University
BIBLIOGRAPHY: PART I

Benardete, Seth. "Achilles and the Iliad." Hermes 91 (1963). Pp. 1-16. Boeder, Heribert. "Der friuhgriechische Wortgebrauch von Logos und A letheia." Archiv fur Begriffsgeschichte 4 (1959). Pp. 82-112. Bohme, Joachim. Die Seele und das Ich im homerischen Epos. Gottingen, 1929. Classen, Joh. Beobachtungen iuber den homerischen Sprachgebrauch. Frankfurt, 1879. Pp. 174-5. Diels, Hermann and Kranz,Walther. Fragmenteder Vorsokratiker.Dublin-ZZurich, 14th edition, 1969. Vol. I, pp. 139-190.
26 Mr. David Furley has suggested to me that the evidence of Xenophanes fr. 7 makes it at least possible that Pythagoras recognized the connections between Quxnand speeches: the puppy's voice betrays the +uX' of a man. 26 Compare Holscher's notion of an inner space, to be discovered as the world (p. 143). 27 Cf. Snell, Discovery, p. 19. His analysis of the sense of a@i)q seems to be substantially correct. On fr. 115, cf. Verdenius, p. 87.

15

Fournier, H. Les Verbes "Dire" en Grec Ancien. Paris, 1946. Frame, Douglas Gordon. The Origins of Greek NOOX. Dissertation, Harvard, 1971. Fritz, Kurt von. "NOO, and NOEIN in the Homeric Poems." Classical Philology 38 (1943). Pp. 79-93. (Cited as von Fritz, "NOON in Homer"). "NOOZ, NOEIN, and their derivatives in Pre-Socratic Philosophy, I." Classical Philology 40 (1945). Pp. 225ff. Cited as von Fritz, "NOOE in Pre-Socratics." Guthrie, W. K. C. A History of GreekPhilosophy. Cambridge, 1962. Vol. I, pp. 403-492. Holscher, Uvo. Antdngliches Fragen: Studien zur friihen griechischen Philossophie. Pp. 130-172. Hoffmann, Ernst. Die Sprache und die archaische Logik. Tubingen, 1925. Kirk, G. S. "Heraclitus' Contribution to the development of a language for Philosophy." Archiv fur Begriffsgeschichte 9 (1964). Pp. 73-77. Heraclitus: the Cosmic Fragments. Cambridge, 1954. (Cited as Kirk, HCF). Kirk, G. S. and Raven, J. E. The Pre-Socratic Philosophers. Cambridge, 1957. Pp. 182-215. Kranz, Walther. "Gleichnis und Vergleich in der fruhgriechischen Philosophie." Hermes 73 (1938). Pp. 99-122. Krischer, T. "ETYMOZ und AAHOHE." Philologus 109 (1965). Pp. 161-174. Marcovich, M. Heraclitus. Merida, Venezuela, 1967. Pohlenz, M. Review in Philologische Wochenschrift 23 (1903). P. 972. Ramnoux, Clemence. H6raclite, ou l'hommeentre les Choses et les Mots. Paris, 1959. Snell, Bruno. The Discovery of the Mind. (trans. T. G. Rosenmeyer). New York, 1960. Review of B6hme, Gnomon 7 (1931). Pp. 77 ff. "Die Sprache von Heraklit." Hermes 61 (1926). Pp. 353 ff. Verdenius, W. J. "Der Logosbegriff bei Heraklit und Parmenides." Phronesis 11 (1966), Pp. 81-98.

16