STYLE AND INVESTIGATICI IN HERACLITUS

STYLE AND INVESTIGATICI IN HERACLITUS

1. We may be attracted by the view that the strongly marked chiastic, asyndetic and antithetical features of Heraclitus' style are in themselves attempts to repre-sent the reality of the universe in pictorial terms '. While I think this view can hardly be refuted in its entirety, I am inclined to Kahn's opinion2 that these singular and characteristic features of style result from the struggles of a humanly limited language to express a difficult theory of reality. On this view Heraclitus' apparent idiosyncrasies of style are not mimetic in intention, but semantic, sus-taining his argument with partly formalised vertebrae. Heraclitus maintains that the human condition constrains human insight: r\Qoq ydp dv0pcÓ7U£iov pèv CÒK £/£i yvcópaq, 0£iov 5è E%EI (Diels-Kranz B 78) \ Bearing in mind the reluctance of early Greek philosophers to construct technical vocabularies 4, I propose to argue that Heraclitus used various stylistic schemata as instruments of investiga-tion: submerged proto-logical formulae intended to empower naturai language towards representing a reality which effectively lay out of reach of its semantic capacities.

2. It is almost certain that Heraclitus was influenced by Iranian and Semitic ideas 5 and beyond doubt that he shows the influence of Semitic stylistic sche­mata 6. Ephesus (Lydian: ibsimis) was accessible to non-Hellenic cultural influ­ence, as the occurrence of Lydian and other non Greek words in the vocabulary of Hipponax indicates 7. There was also an Aramaic-speaking zone in its proxim-

1 C.J. Emlyn-Jones, Heraclitus and the Identity of Opposites, «Phronesis» XXI/2 (1976) 89-114 (98, 101); T.M. Robinson, Heraclitus and Plato on the Language of the Real, «Monist» LXXIV/4 (1991) 481-489 (485); P.K. Curd, Knowledge and Unity in Heraclitus, ibid. 530-549 (535).

2 C.H. Kahn, The Art and Thought of Heraclitus. An Edition of the Fragments with Translation and Commentary, Cambridge 1979, 123ff.

3 H. Diels-W. Kranz, Die Fragmente der Vorsokratiker, II, Berlin 19526 (hereafter re-ferred to as: D.-K.).

4 H. Frànkel, A Thought Pattern in Heraclitus [1936], in The Pre-Socratics. A Collection of Criticai Essays, ed. A.P.D. Mourelatos, Garden City 1974, 216.

5 M.L. West, Early Greek Philosophy and the Orient, Oxford 1971, 175ff. 6 K. Robb, Preliterate Ages and the Linguistic Art of Heraclitus, in Language and Thought

in Early Greek Philosophy, ed. K. Robb, La Salle, 111. 1983, 153-206 (178 ff.). 7 Hipponax. Testimonia et fragmenta, iterum ed. H. Degani, Stutgardiae et Lipsiae 19912

72 RANKIN

ity 8. The importance Heraclitus places upon fire as the principal cosmic constitu-ent and the perpetuai change that seems to characterise his universe might suggest a species of Irano-Hellenic eclecticism 9. Certainly he differs significantly from the attitudes of current Greek thinking. He has strong objections to blood sacri-fices and to anthropomorphic statues of the gods (D.-K. B 5). Also disedifying to him are mystery cults of various kinds, though we may note that he includes pdyoi in the list of practitioners of those that he specifically proscribes (D.-K. B 14) 10. We also observe that the word [idyoq (if it is not a later inclusion in the list of VUKXITCÓ^OI, paK^oi, À,fjvai, puaxai) " seems to have acquired the sense of 'religious imposter', rather than 'member of a sacred Median caste' as Herodotus' numerous references later indicate (Hdt. I 107, 120, 128) 12. We may suspect anachronism arising from the intrusion of some later gloss or comment, but need not embrace the suspicion. Heraclitus could have been influenced consciously or even 'osmotically' by Iranian ideas without approving every aspect of minor cults of apparently Iranian origin. We may note also his dislike of obscene phallic songs associated with the worship of Dionysus. His feeling about these is that they suggest an equation between Dionysus and Hades (D.-K. B 15). This could be an entirely Hellenic fastidiousness of the kind celebrated in Euripides' Bacchai. It is impossible to be certain that his expressed view that corpses are more worthy of rejection than dung (D.-K. B 96) is based on anything more than an empirical appreciation of their dung-like qualities. Certainly He places a low value on the taboos associated with burial in the Greek world. However, criticism of inherited customs was already part of the Ionian intellectual tradition and it goes back at least as far as Archilochus (fr. 8,96 Tard. = 5,114 W.2), whom Heraclitus himself criticises together with Homer (D.-K. B 42). His deployment of certain character-istic schemata of style, on the other hand, offers a more apparent basis for the ascription of non-Hellenic influence. These, as Robb (o.c.) argues, strongly re-semble some of the structural formulae of Semitic poetry.

3. We can have no absolute certainty on such points of cultural influence, but we see in the fragments how frequent was his use of these strongly com-

(Lipsiae 1983'), XXVIIf., cf. frr. 1;2;3;7; 16; 36,6; 39,5; 41,2; 42b,lf.; 43,3; 47,lf.; 49; 72,7; 79,18; 95; 107,21; 119; 124; 148; 154; 164; 170; 173; 178; 186; Id., La lingua dei barbari nella letteratura greca arcaica, in «Actes du colloque international "Langues et Peuples" (Gressoney-Saint-Jean, Chàteau Savoia, le 8 mai 1988)», Aosta 1989, 75-82 (78ff.).

8 G. Deeters, Lydia, in RE XIII/2 (1927) 2122-2161. y A. Momigliano, Alien Wisdom: the Limits of Hellenization. Cambridge 1975, 125f. 10 H. Tarrant, Greek Thinkers and the Question of Alien Influence, in «Proceedings of the

First Australian Congress of Classical Archaeology», ed. J.-P. Descoedres, Oxford 1990, 625f. " G.E.R. Lloyd, Magic, Reason, and Experience: Studies in the Origins and Develop-

ment of Greek Science, Cambridge 1979, I3f. 12 Lloyd, o.c. 13 n. 20.

STYLE AND INVESTIGATION IN HERACLITUS 7 3

pressed and antithetical schemata, which resemble the well-known patterns in Semitic rhythmic poetry l3. Not ali of the fragments generally attributed to Heraclitus contain examples of these schemata. His use of them seems to have been delib­erate, and the present argument is that his decision to employ them was philo-sophical, and not merely a craving for exotic ornament. His statements about the nature of reality are intuitive, not demonstrable. If we understand the logos we can have some intuitive understanding of this reality. Logos includes conscious rational narrative 14; articulated rationale which empowers the former, and a 'meta-rational' 'meta-articulation' of reality (D.-K. B 1,2). Logos is both what Heraclitus says, and what he is talking about. No surviving statement of Heraclitus explicitly equates fire with logos, or with psyche, but there are reasonable grounds for supposing, if not 1/1 identity 15, at least considerable overlap in function and commonality of attributes between them 16. 'Logos' in the sense of what Heraclitus actually says to us, is the nearest to proof of his theory that we can hope to have.

4. Although there are evidences of traditional hexametric rhythm in the fragments of Heraclitus, for example: tydxiq aùxoiaiv papxupEi TcapEÓvxac, dneivai (D.-K. B 34) 17, the dominant pattern in Heraclitus' prose is chiastic and antitheti­cal, together with assonances, anaphorai, alliterations, paronomasiai, homoioteleuta, parallelisms and asyndeta. These features of his style are typically illustrated by such fragments as D.-K. B 10: rjvjvdxj/iEC, óÀ.a Kaì ox>x oÀ.a, aup^EpópEvov 5ia(()£póp£vov, ODVCXÒOV 8ia5ov 18, and D.-K. B 62: d0dvaxoi 0VT)XOÌ, 0vr|xoì d0dvaxoi. Such structures seem to anticipate some of those favoured by Gorgias. Chiasmus was a feature of other Ionie writers such as Democritus (D.-K. B 108, 177, and, indeed, of Plato who also provides numerous examples 19. It is difficult to suppose that the frequency of chiasmus in the surviving fragments of Heraclitus is due simply to its attractiveness as a mnemonic device. It is embodied in sharp paradoxes and apparent contradictions. When Heraclitus adapted the forms as distinct from the contents of Semitic Wisdom Literature to convey his theory of reality 20, he took over a style more suited to the antithetical character of his

13 W.G. Watson, Classical Hebrew Poetry: a Guide to its Techniques, Sheffield 1984, 202ff.

14 W.K.C. Guthrie, A History of Greek Philosophy, I, Cambridge 1962, 419-435. 15 J. Wilcox, Barbarian Psyche in Heraclitus, «Monist» LXXIV/4 ( 1991 ) 624-637 (632). 16 G.S. Kirk, Heraclitus. The Cosmic Fragments, Cambridge 1954, 37-40; U. Hòlscher,

Anfdngliches Fragen. Studien zur fruhen griechischen Philosophie, Gòttingen 1968, 131ff.; M.C. Nussbaum, VvxA *'« Heraclitus, I, «Phronesis» XVII (1972) 1-16 (3ff.).

17 F. Schleiermacher, Heraklit der Dunkle von Ephesus, Berlin 1839, 3; E. Norden, Die antike Kunstprosa, I, Stuttgart 19585, 44.

18 Robb, o.c. 192; B. Snell, Die Sprache Heraklits [1926], in Gesammelte Schriften, Gòttingen 1966, 129-151.

19 J. Denniston, Greek Prose Style, Oxford 1959, 74f. 20 Robb, o.c. 178.

7 4 RANKIN

arguments than that of the narrational hexameter. Parmenides amply illustrates the problems involved in an attempt to use this metre for philosophical argument. It may be asked why Heraclitus did not employ paroemiac lengths, rather than the fully developed hexameter, or else the generally useful elegiac couplet, or even the Xéfyq £Ìpopévr|, which he also seems to have avoided 21. Unexpectedly for a thinker who regarded reality as in ultimate terms a unity, he sometimes used vivid similes to reinforce his teaching (D.-K. B 53)22. It would appear, however, that his principal requirement was for schemata which could make his sentences into representational instruments whose structures matched those of the difficult argu­ments he wished to express. The reality with which he had to deal was a 'proc-ess' " : not only fire in various transformations, but a reality which involved a fusion of mutually tensioned and mutually identifying opposites. It is not easy to describe such a 'field' of being (if we use that term from physics), in naturai language which is morphologically modular and in its syntax analytical. Logos in the sense of a philosophical explanation of reality involves 'division': ÓKOÌCOV

èycò 5ir|Y£ijpai Kaxd fyvoiv 8iaipécov EKaaxov m i §pdi,(x)v ÒKCOC, E^EI (D.-K. B 1). Logos is forced to use the language of division (Siaipécov) in order to talk about the unity and the tensioned fusion of opposites.

5. Heraclitus needed a calculus to catch the universe on the move. A com-parable problem was later to absorb Plato's attention. He attempted to solve it by means of the eide. Heraclitus' solution was to insert into his sentences formulale blocks of words matching the shape of his chosen stylistic schemata. These sche­mata might attempt to represent reality in symbol without claiming to give a picture of it. Nature, after ali, loves to hide (D.-K. B 123). He strengthened this formulale dimension by adopting the 'aphoristic style' which also belonged to the tradition from which he took the schemata 24. This has been described by Robb (o.c. 176) as an 'orational' rather than declarative mode 2 \ Orational utterances and statements, frequent in Semitic 'Wisdom Literature', tend to be individuai and gnomic rather than narrational or periodic: each item can be considered as residing independently within its own semantic cartouche. Contradiction between any number of such items is not of primary importance 26. Heraclitus' use of this

21 W. Schmid-O. Stàhlin, Geschichte der griechischen Literatur, I/I, Munchen 1929, 745-755 (751f.).

22 W. Kranz, Gleichnis und Vergleich in der friihgriechischen Philosophie, «Hermes» LXXIII (1938) 99-122 (111).

23 K.R. Popper, Back to the Pre-Socratics, «Proceedings of the Aristotelian Society», n.s. LIX (1958-1959) 1-24 (17).

24 J. Barnes, Aphorism and Argument, in Robb, o.c. 91-109 (98). 25 M. Schofield, Heraclitus' Theory of Soul and its Antecedents, in Companion to An-

cient Thought 2 (Psychology), ed. S. Everson, Cambridge 1991, 13-24 (24). F. Waismann, Language Strato, in Logic and Language (Second Series), ed. A.G.N. 26

STYLE AND INVESTIGATION IN HERACLITUS 75

mode was far from extreme: his aphoristic sentences are by no means bereft of the particles and linking words which suggest continuity of argument27. Yet the sur-vival of so many eminently quotable utterances tells its own story of the commu-nicative strength of aphoristic style allied to schemata.

6. Let us consider more closely a developed specimen of Heraclitus' use of chiasmus in D.-K. B 36: (v|/\)xfiaiv [A] 0dvaxocj [B] vÒwp [C] Y£véo0ai), (USGCTI

[C] 5E edvaxoq [B] yfiv [D] y£véo0ai), (ÈK yf\q |D] \)5cop [C] yivExai), (èc, \)8axo<; [C] 5È yv%r\ [A]). The cycle of change is in four phases. The pattern of contrasted words runs: ABCCBDDCCA in a complicated chiasmus. If we write it in the following way:

ABC CBD DC CA,

we can see more clearly the role played by C in the process of change. If we leave out B as a piece of stylistic emphasis and imagine some less arresting expression of dissolution, the effect is clearer stili. Water is the class to which ali the other items belong as it belongs to them. The pattern is probably as dose to an ostensive 'proof of the proposition as Heraclitus is equipped to come. Frànkel (o.c. 217-220) saw that the logicai mechanism involved was that of the doublé proportion or geometrie mean. Chiasmus also occurs in the following fragments: D.-K. B 1, 10, 11, 18, 20, 25, 30, 31, 34, 36, 50, 53, 67, 111, 126.

7. Heraclitus also uses the schema etymologicum in D.-K. B 18 to support a philosophical position which is similarly undemonstrable in a logicai sense but has some basis in general experience: èdv pf| £À,7crìxai [A], dvÉATUoxov [Al OVJK

ècJEuprjoEi [B], àvECjEpELivrixov [C] èòv Kai a7iopov [C]. The apodosis gains credibility by the almost Aeschylean aggregation of negatives, both particle and compound. In B 62: (dGdvaxoi [Al Oviixoi [B]), (Gviixoi [B] àedvaxoi [A]), { COVXEC; [A] xòv EKEÌVCOV edvaxov [B]}, {xòv 5è EKEÌVCOV [B] piov XEOVECOXEC;

[Al}. Since A=B, we can express AB as A—.A (or perhaps VA3x [Ax=-iAx]). It is not the purpose of this discussion to argue whether Heraclitus was more inter-ested in classes of things rather than propositions about things, but AeA would be a nuli class, which of course is stili arguably 'something'. Heraclitus is talking

Flew, Oxford 1953, 11-31 (22); J.M. Moravcsik, Appearance and Reality in Heraclitus, «Monist» LXXIV (1991) 547-567 (558).

27 H. Gomperz, Uberdie ursprùngliche Reihenfolge einiger Bruchstucke Heraklits, «Hermes» LVIII (1923) 20-56 (25f.).

76 RANKIN

about identity rather than nullity. His glosses in the second part of the fragment, between the brackets { } { }, declare his premises, viz. the uncontrovertible difference between mortai and immortai kinds: dGdvaxoi Ovrixoi, 0vr|xoì àGdvaxoi, which nevertheless are somehow identical. Grammatically and logically these pairs of words would later be classified as predicates. Here they are as sparely stated as possible, typical 'Nominalfassungen der Pràdikat' without expressed verb 28. They intend to express the ontological fusion of opposites, to express in our language and using our rules of thinking, the manner in which logos, the rational basis of such language and rules, attempts to represent reality. Merely to predicate contradictory 'A' of contradictory 'A' would be trivial29, and this is the substance of Aristotles' objection to Heraclitus' apparent position (e.g. Met. lOlOa, 10; 1012a, 24, 34; 1062a, 32; 1063b, 24). But Heraclitus' task is to adumbrate fused unities which in terms of logos have the compositional character of unified opposites. The problem with language (which is logos as an arrangement of words and also the theoretical rationale which the words describe) is that it is linear and time-bound and involves the use of a succession of linguistic items to explain reality which need not be either linear or time-bound. Heraclitus deliberately uses temporal words like d£Ì and dEÌ^coov (D.-K. B 30) to warn us that reality is not to be seen in temporal terms. In D.-K. B 67 ó Qeóq is said to be a sum of opposites: ripépri Etxjipóvri, x£tpcòv 0épo<;, 7tóÀ.£po<; £Ìprjvr|, KÓpoc, Àipóc;. This asyndetic row of juxtaposed opposites attempts a formula for the oppositional nature of reality, but it has to operate within the limits of successive, sectionalised, analyti-cal language. The stylistic schema that Heraclitus uses in this fragment is an attempt to reach beyond this limitation by a stylistic procedure which achieves quasi formalisation. Again, in Fragment B 12 (cf. B 49a, B 91): Heraclitus ex-plains the gaseous, volatile, intelligent (vospd) nature of the psyche by the fa-mous 'river' analogy: 7coxapoiai xoìaiv aPxoìoiv èp(3aivo\)Oiv EXEpa Kai EXEpa \)5axa ETuppEÌ. The concept of 'same' presents a difficulty here. Its irrelevance to reality is indicated by B 49a: 7toxapoìc; xoic; aPxol'c; èppaivopév XE Kaì OÙK ÈpPaivopEv, plus what could be an intrusive gloss: EÌpév XE Kaì OÙK £Ìp£v. Mind is continuous like the phenomenon of an actual river; but our language is not a continuum, it is linear and successive, therefore involving the use of relations such as 'same'. So we can only talk about the mind (and the river) as a succession of sections. Heraclitus' use of the word EXEpa to describe the oncoming waters of the river is to be noted: if he had wished simply to describe a continuum, he could have used dÀ.À.a as the attribute of ì)8axa, instead of Exspa, which means 'another set of two', rather than 'other'. As it is he conveys not simply the idea of 'orderly progression' 30, but a succession of beings understood as arranged in a binary

28 Schmid-Stàhlin, o.c. 75 n. 4; G. Vlastos, On Heraclitus, «AJPh» LXXVI (1955) 337-368 (353).

29 J. Barnes, The Presocratic Philosophers, London 1982, 70. 30 G.S. Kirk, Naturai Change in Heraclitus, «Mind» LXV (1951) 35-42 (36).

STYLE AND INVESTIGATION IN HERACLITUS 7 7

groupings which constitute a fused being. Logos describes this fusion in a way which enables us to apprehend it, a succession of 'sections' or 'portions' of water which yet constitute a moving stream. Fragment D.-K. B 30 applies to fire a comparable orderly progression of péxpa 31. The ouvdyiEc; (D.-K. B 10), which {are} òÀ.a Kaì ox>x òXa, rjup^EpópEvov 5ia(|)£pópEvov, a-uvaSov 8ia5ov, also (D.-K. B 8) xò dvxiCjOVjv ovjp^épov Kaì ÈK xcov 8ia<|)£póvxcov KaXXiaxr)v àppoviav, and (D.-K. B 51) orj cyuviàoiv OKCOC, 8ia(J)£póp£vov ècoDxeo ópoA.oyé£r Tca^ivxpo-Koq àppovìr) ÒKCooTCEp xó£,o\) Kaì A/upr|<;. I\)vd\|/i£c;, dppovìri, are attempts on the part of language to express the way in which beings are fused into an other level of being which nevertheless is infused with tension between the constituent beings. In terms of logos that is accessible to us, such words as ovva\\tiq connote primarily the combination of separate items in an ordered arrangement, but in terms of the logos which is the constitutive rationale of the cosmos, the implied analyticity of these words and language in general need not be considered rel-evant. The logos of the psyche, for instance, is so ldeep' that we cannot hope to reach the psyche"s boundaries (rcEÌpaxa D.-K. B 45). Our human notion of logos, however, is restricted in semantic scope. Heraclitus grapples with the problem that our language (and apprehension of logos) is anthropic, just as our perceptive orientation is peculiar to our species (D.-K. B 7): ei Ttdvxa xà òvxa Kanvòq yévoixo ptvECj àv SiayvoiEv. I suppose that he would have some sympathy with our contemporary and only partially felicitous attempts to describe cosmic singularities, 'Black Holes', in naturai language.

8. Finally I should like to consider paronomasia (word-play), a schema which Heraclitus uses to powerful effect in his efforts to increase the capacity of lan­guage to represent reality. Its use compels naturai language to express more than one level of intention as economically as possible. Here again we have a stylistic form which is more frequently used in Semitic literature than in Greek ?2. In the surviving fragments of Heraclitus word-play is relatively frequent: Fgs. B 1, 20, 25, 28, 46, 54, 56. He uses it for his own special purpose - not for ornament or in aid of memory, but to represent by means of the polysemic character of some parts of naturai language the nature of a reality which we can think about and describe as multivalent but which need not itself be so. The play between (3ìoq ('life') and the death-giving fìlóq ('bow') (D.-K. B 48) presents almost a visual diagram of the opposed duality. The distortion of accent would have been obvious to anybody who heard the sentiment recited orally. Bvóq ('bow') would need a regressive Lesbic accent to correspond accentually with pìoq (i ife ') . The sym-

31 G.S. Kirk, Heraclitus. The Cosmic Fragments cit. 366; West, o.c. 138. 32 Robb, o.c. 167; Watson, o.c. 245ff. 33 C D . Buck, Introduction to the Study of the Greek Dialects, Boston-New York-Chi­

cago-London 1910' (19282), 79.

78 RANKIN

bolic identity would have been much more apparent to a visual reader of the sentence in text. This may lend some support to the view that Heraclitus deliber-ately departed from the orai tradition in favour of the stability of script. The acknowledged fact that very little private reading in antiquity was entirely silent34

does not detract from the likelihood of an intended visual effect. The parallel: t)vv vóto: ch'uveo (D.-K. B 114) also involves a harsh accentuai effect. Fragment D.-K. B 25: pópoi ydp péc ovEC, péc^ovac; poìpaq Xayxdvovoi, not only employs word-play but also alliteration and an ABBA chiasmus. Fragment D.-K. B 46 not only has word-play between oìr|CJi<; (as a 'mental disease': Upd vóooc,) and òpaoic; (with spiritus lenis in the originai Ionie), but also a play upon oìiicic; and oioc; (= not t)vvóq). Paronomastic word-plays in Heraclitus can be compared to test-drills penetrating through stratified subtexts in language. They reveal rather than express certain characteristics of reality. We recali that Heraclitus (D.-K. B 193) tells us that the god whose oracle is at Delphi: ovxe ÀiyEi ovxe Kpvnxei dXXd rjrjpaìvEi. The use of or|paiv£i is significant in that, unlike A.éy£i which asserts, and Kpvnxei which challenges our interpretative powers to a semantic test to destruction, it indicates factual possibilities rather than possible faets. In a Wittgenstinian 35 sense semainei pictures the possibility of what it seems to indi­cate and is 'beyond representation' 6 when we take that word to mean words and thoughts matching reality.

9. Heraclitus' greatest problem was the communication in naturai language of his intuitive views about the nature of a reality which he intuited as effectively indescribable in naturai language. It is impossible to assert effectively in naturai language the indiscernibility of identicals which are by definition mutually op-posed and which are processes as much as they are phenomena 37. Heraclitus invoked the aid of poetic imagery and allusion to extricate himself from a «maze whose centre is the discovery that there is no centre» 38. Fire, which is conceptu-ally the most fluid of ali elements, is allocated a philosophical centrality in his theory, but we cannot be certain to what extent he is speaking of an Ionian physi-cal element or using it as a 'concept' of change (dvxapoipri D.-K. B 90). His use of stylistic schemata, some of which were unfamiliar in the Greek poetic tradition, and some of which had very ancient roots 39 constituted an ingenious advance

34 W.V. Harris, Ancient Literacy, Cambridge, Mass. 1989, 36 n. 37; 84 n. 90. 5 L. Wittgenstein, Tractatus Logico-Philosophicus, London 1922, II 203; D.G. Stern,

Heraclitus ' and Wittgenstein 's River Images: "Stepping Twice into the Same River", «Monist» LXXIV/4 (1991) 578-604 (597).

36 A. Palmer, Beyond Representation, London 1992, 152-163 (157). 37 Popper, o.c. 38 J.O. Wisdom, Esotericism, «Philosophy» XXXIV (1959) 338-354 (338f.). ' E. Norden, Logos und Rhythmus [1928], in Kleine Schriften zum Klassischen Altertum,

Berlin 1966, 533-551.

STYLE AND INVESTIGATION IN HERACLITUS 7 9

towards a quasi-symbolisation of his arguments and also a considerable extension of a limited legacy of heuristic equipment4 0 .

Southampton D A V I D R A N K I N

40 J.M. Moravcsik, Heraclitean Concepts and Explanation, in Robb, o.c. 134-152 (146).