SciELO - Scientific Electronic Library Online

vol.15 issue2Sophokles' Philoktetes and the ascent to political friendshipThe sensible and the reasonable: Plato and Rawls author indexsubject indexarticles search
Home Pagealphabetic serial listing  


On-line version ISSN 2413-3086
Print version ISSN 1561-4018

Phronimon vol.15 n.2 Pretoria  2014




Heraclitus' usage of όστις in fragments DK B 5 and B 27



Andrew J. Mason

Unaffiliated Australian scholar, residing in Japan, working and publishing mainly in the field of ancient Greek philosophy




In their basic senses the difference between όστις and ός is straightforward, corresponding to that between whichsoever/whosoever and which/who. But in usage it is often much subtler and at times negligible, both because όστις can like ός be used for a definite referent and because in certain constructions ός can like όστις denote an indefinite one. There are five certain uses of όστις in Heraclitus' surviving fragments, and it is notable that in each case translators do not render the term in its basic, indefinite sense but in a sense akin to ός. While in most cases this is clearly right, I question it with regard to B 5 and B 27, and explore what these fragments may have to say to us beyond prevailing interpretations if όστις is read as a true indefinite pronoun. In the case of B 27 this may, I argue, force us to revise our understanding quite radically. The paper also examines Heraclitus' use of other terms with a similarly indefinite reference (ός with άν, όκόσος, όκοΐος, όσος), for the purpose of establishing whether he would have used one of these in B 5 and B 27 instead of όστις if he intended indefinite objects there. I argue against this, since the connotations of these terms would be inapposite in these fragments in comparison with the root sense of όστις.

Key concepts: Heraclitus; hostis; hos; unity of opposites; expectation



In terms of its construction, the indefinite or general relative pronoun όστις simply involves the addition of the indefinite pronoun τις to the relative pronoun ός. What Liddell, Scott and Jones call its radical sense is, accordingly, anyone who, anything which, or whosoever, whichsoever (e.g. "whoever wants this cup", Iliad 23.667). As the Middle Liddell notes, here όστις is to ός as the Latin quisquis is to qui. In the second sense LSJ list, however, όστις is not so clearly distinguished from ός, since it too can refer to a "definite object", particularly in Ionic Greek (cf. Smyth 1956: §2496b). LSJ state that this occurs "properly only when a general notion is implied", and refer among other examples to Herodotus 3.120, where δι' όντινα has the sense "one through whom" rather than "the one through whom". LSJ note as an exception Thucydides 6.3, where όστις directly refers to the altar which. With reference to this passage the Middle Liddell calls the second sense of όστις "hardly different from ός". But to some degree this holds even with the other examples, for ός too can refer not to its antecedent noun itself but to the notion implied in it or the class of thing it belongs to: "a thing which" (e.g., Plato, Republic 359c), or "one (of those) who" (Euripides, Orestes 920). Moreover, ός can embrace the first sense of όστις as well, notably when used with particles or conjunctions (κε, κεν, άν; δή, δήποτε). Smyth adds that όστις rather than ός is used after a negative (1956: §2496a), and that, when the antecedent is indefinite, όστις is "preferred" when the verb is indicative, ός when it is subjunctive or optative (§2508).

While context can help us determine how these words are used, contextual clues are scant in Heraclitus both because of the way his texts are cited by ancient sources and the very way he writes. In addition, Heraclitus has a penchant for exploiting linguistic ambiguity, and I will propose in this paper that on two occasions this is true of his use of όστις. In one case, B 27, this may lead us to revise quite radically our understanding of his meaning. In the surviving fragments we find the following forms of όστις:1

1. όστις, singular masculine nominative (B 57).

2. ότεω, singular masculine dative, Epic form (B 15).

3a οϊτινες, plural masculine nominative (B 121).

3b οϊτινές (B 5).

4 άσσα, plural neuter nominative, Epic and Ionic form (B 27).

1, 2 and 3a are clear examples of sense II, where όστις is "hardly different from ός". Indeed, they are most like the example from Thucydides, where the object is quite definite. In B 57 the referent is Hesiod, "he who" (or perhaps "one who") failed to recognise the oneness of day and night. In B 15 it is Dionysus, "for whom" the participants of the Lenaean festival rant and celebrate. In B 121 it is the adult Ephesians, "they who" expelled Hermodorus from their city. Note also that in these cases Heraclitus uses the word not just with a definite referent but a personal one. It is generally assumed that sense II applies in 3b and 4 as well. But is this the case?

In the second part of B 5 Heraclitus mocks the practice of praying to statues, which he compares with someone carrying on a conversation with houses, οΰ τι γινώσκων θεούς ούδ' ήρωας οϊτινές είσι. Marcovich translates: "not recognising what gods and heroes really are" (1967: 459). Let me say first that in broad terms this, and comparable translations such as Kahn's (1979: 81) and Robinson's (1987: 13),2 gets the thrust of the whole sentence right. Pace Osborne (1997: 83-4), Heraclitus is not just mocking the practice when it is performed by those ignorant of the nature of gods and heroes. Her rendering, "as if someone, who knew nothing of what gods and heroes are like, were to converse with the houses", ignores the explanatory function of οΰ γινώσκων. To pray to statues, being comparable to trying to converse with houses, is in itself to fail to understand the nature of gods and heroes.

But how does όστις function in this statement? All the translators referred to above take it as "hardly different from ός", and there is no discussion of the point in their commentaries. While this also applies to Patrick's translation, "knowing nothing of gods or heroes, who they are", note that Patrick supplies a comma before οϊτινές (1889: 131). This seems justified, since οϊτινές is not grammatically governed by γινώσκων like θεούς ούδ' ήρωας. If so, it may free up an alternative translation of the final phrase in keeping with sense I: "whoever/whatever they are", somewhat along the lines of the tragic formula Ζεύς, όστις ποτ' έστίν, "Zeus, whoever he may be".3 We are thus left with two questions: i) does οϊτινές denote who or what? ii) does it denote a definite who/what or an indefinite whoever/whatever?

Regarding i), most scholars take B 5 as following in the footsteps of Xenophanes' critique of religious anthropomorphism,4 and it can be argued that this entails an attack on the very assumption of personal gods with whom one could converse. Thus in B 93 Heraclitus depersonalises Apollo ("the lord whose oracle is in Delphi") and at the same time emphasises that Apollo does not speak, or more precisely, "neither speaks [λέγει] nor conceals [κρύπτει], but gives a sign [σημαίνει]". To understand the sign's import one must think for oneself, not press the god or engage him in further conversation in order that he reveal his import. A quest after personal motives is precisely what is impertinent here. More tellingly, in B 32 Zeus is deemed a fitting name of Heraclitus' cosmic god, the neuter and strictly impersonal Hen or "One", only on condition of his radical depersonalisation. Since Zên, besides denoting Zeus in Ionic, also means "to live", and since life is a recurring theme in the many aliases Heraclitus gives the One (πυρ άείζωον in B 30, βίος/βιός in B 48, αίών, life-force, in B 52), the sense is that the One is willing to be called Zeus only if Zeus is thought as impersonal (and for just that reason supremely wise) life-force. Again, Heraclitus does not seem to regard Dionysus and Hades as divine personalities, for they are "the same" (B 15). The indications are that he employs the gods' names as ciphers, pointers to the One or to certain essential features of the cosmos which the One both is and rules: in the case of Apollo, the proclivity of the One, or of the real generally, to self-concealment (cf. B 123); in the case of Dionysus and Hades (B 15), the unity of life and death. Heraclitus also speaks more generally of "gods" vis-à-vis "men" (B 24, B 53, B 78, B 83), yet here the theoi are not addressed as agents but serve as standards of comparison, against which man appears paltry in just those things in which he prides himself vis-à-vis animals.

All of this favours reading οϊτινές impersonally regarding gods. The same holds with regard to heroes. Marcovich (1967: 462-3) rightly infers from B 96 ("corpses are more fit to be thrown out than dung") that in B 5 "the statues of heroes (representing their bodily features) should not be [the] object of any cult, since it is only their fiery soul which is eternal and divine". The necessary corollary here is that the more fiery the soul is, the more impersonal: that is, the less marked it is by individual characteristics and the more identifiable with, indeed the more it identifies itself with, the impersonal wisdom of the One, i.e., fire as such. This point requires clarification before we can establish the sense of οϊτινές in B 5.

There has long been controversy among scholars about whether Heraclitus identifies soul with fire or air. Without reviewing this extensively here, I endorse the view of Robinson (1987: 158-9, 187) and Betegh (2007: 21) that soul covers a "spectrum" from damp air through drier, clearer air, to fire. In Robinson's view, the first corresponds to the polloi, the middle to the best and wisest men (who become spirits after death), and the last to the gods. I agree that the drier it gets, the more soul becomes a principle not just of life but of rationality, approximating to the supreme wisdom of the One/fire.5 I would add that it also becomes more impersonal, and that anhedonia would be a primary mark of such soul, over against the damp soul associated with pleasure in B 77. Betegh's illuminating analysis can help us take this further. Betegh draws out the significance of the distinction between soul (as a cosmic mass) and individuated souls in B 36 (2007: 9) by rehabilitating a generally denigrated commentary by Sextus Empiricus, particularly Sextus' claim that for Heraclitus the soul in us is "separated from its natural unity with the surrounding [rational] medium", a portion of which "resides" in us "as a stranger" (Against the Mathematicians, 130). For Betegh (2007: 27) this can explain the "shift" from soul to individuated souls in B 36 as "a partial dying". Soul in us is not fully what it is in itself. The one point I would question is Betegh's view of the "medium" as "the greatest individuated portion of soul", i.e., the all-steering cosmic god (2007: 12). This need not follow, given the impersonality of the One over against the individuated "bits" of soul. It is, I think, to this mass of impersonal soul that the souls of departed heroes are released, in distinction from those souls which die and become water.

Moreover, Sextus indicates that we are most separated from the medium when we are asleep. When awake, by way of perception we can regain contact with it and therewith rationality, and to some extent win the dry or even fiery form of soul. What would distinguish us then from other human souls is not simply a different set of personal traits. Fiery soul is distinguished by its transcendence of all the limiting quirks and failings, desires and passions, hopes and fears, and so on, that hold soul back from attaining its highest form, and by its resolve to view things, and indeed to comport itself to life, sub specie aeternitatis, with no investment other than in what is and what must be.

Returning to B 5, the point would now seem to be that, whomever we pray to or worship, whichever god or hero it be, we should recognise that we are relating ourselves to an impersonal "one and the same" which we may discern in any of them, but which we should not confuse with any one of them. This seems to open up a new perspective on question ii). On a first hearing of the reading "without any recognition of gods and heroes, whoever/whatever they are", the final words may seem (as with the examples from tragedy in note 3) to express an uncertainty which is at odds with the strident nature not only of B 5 but of Heraclitus' proclamations generally. Yet they need not have that sense. Rather, in a way that exploits the ambiguity of whoever/whatever, Heraclitus would be saying that, no matter whom we pray to, we are bound to misrecognise it if we overinvest the "who" and seek a personal relationship with it, rather than seeking to instantiate ourselves the impersonal wisdom we should properly recognise there. Whether it be a god or hero, and whichever one it be in any given case, it is always at bottom this, and we have in ourselves the same "stuff" that can attain this form if we let it.

While reading όστις in its radical sense gives an additional nuance to B 5, it does not radically change its meaning. We can still agree with Kahn (1979: 267) that Heraclitus' purpose here is to shock his fellow Greeks into the realisation that their religion normatises actions which would otherwise "count as stupidity or insanity". But B 27 is another matter.

B 27 reads: άνθρώπους μένει άποθανόντας άσσα ούκ έλπονται ούδέ δοκέουσιν.6 There seems to be universal consensus that here too όστις (άσσα) has sense II. Translations most commonly take the form: "There await men when they die things which they neither expect nor even imagine."7 Substantive variations pertain mostly to the rendering of έλπονται and δοκέουσιν.8 I have no objection to "expect" and "imagine". Whatever the awaiting things are, not only do men not think they will come to pass, but it never enters their heads that they might. "Neither expect nor even imagine" conveys this more than adequately. The issue for me lies elsewhere. On the usual reading the saying concerns men's misguided expectations about the afterlife,9 and it is a matter of specific things to come which, it seems, Heraclitus at least knows all about. But let us listen to the saying again with άσσα as an indefinite pronoun, and with the phrase introduced by it, which is the grammatical subject, placed at the beginning for emphasis: "Whichever things men do not expect (for themselves)10 or even imagine await them when they die." The very different sense the saying now yields is not lost if we put the sentence in existential form: "There await men when they die whichever things they do not expect for themselves or even imagine."

On this construal B 27 does not assert, while remaining silent about them, that there are certain things that await all of us in common when we die. Nor does it restrict the object of men's expectations to the afterlife. What it is really talking about is the living: not just what they expect but the way they expect and go about their lives in expectation. Life, it seems, is in an essential sense about learning how to expect properly, and what we do not learn in life we will have to come to terms with after death. Heraclitus even seems to give a clue to what this involves: not expecting for oneself, from a narrowly personal and self-interested standpoint. Again the theme of wisdom in impersonality seems to come to the fore. Assuming that for Heraclitus what afterlife we may have is impersonal, determined in its rank11 by the kind of person we have been but itself divested of personal traits, his message in B 27 would be that we do not have to wait for death to win this impersonality. We can gain the requisite distance from our egoic investments in life and let this guide the manner in which we comport ourselves to the future.

I think this reading of B 27 is at least plausible, and it is surprising that the possibility of όστις having its root sense here is never touched on. It is true that the ancient authors who preserve the saying understand it as a pronouncement on the afterlife. Yet rather than taking this as evidence in favour of the usual reading, I think we are entitled to have qualms here. As Marcovich says (1967: 401), Plutarch seems to give B 27 the very unlikely sense that after death "good things beyond any expectation and imagination" await men. But Clement's view that Heraclitus "manifestly agrees" with the view articulated by Socrates in the Phaedo (that after death good and bad men will receive appropriate fates) also seems dubious as a reading of this saying. Marcovich (1967: 4012) spells out the conditions on which Clement's view, which he broadly endorses, may hold: by "men" Heraclitus cannot mean all men but the bad majority; the things they deludedly expect after death must be good things, and the things that actually await them must be the bad things they deserve. Yet is it so clear that άνθρωποι has a "negative implication" in B 27? Marcovich cites B 1 in support, without mentioning that it speaks later of τους άλλους άνθρώπους, the rest of mankind, who are set off against those men who act and speak wisely (including Heraclitus). This distinction is still relevant in B 27, but the point is that B 27 reads more like a general maxim about mankind than an attack on the polloi.12 If Plutarch and Clement's readings can be questioned in these ways, I suggest they can be questioned in others, down to the basic assumption that Heraclitus is speaking fundamentally about the afterlife here.

I will say more later about the substance of my alternative construal of B 27, but at this point it is worth bringing in the issue of Heraclitus' highly unique style. Robinson (1987: 94) takes B 27 to make the "straightforward" point that "life after death is different from what people expect or imagine", yet from a stylistic point of view the very straightforwardness of this point seems to me to count against it. Hölscher (1993: 238) admirably describes Heraclitus' sayings as "discoveries: insights that dawn upon the thoughtful soul like the solution of a riddle" in a manner akin to a Goethian appergu.13 Heraclitus seeks to instruct his listeners in the matters of philosophy in much the way that Grace Kelly sought to instruct her leading man in matters of love in one of her films: "The answer would be worth much more to you if you found it for yourself." It is not his way to hand things to us on a platter, nor to content himself with truisms. Heraclitus proclaims himself the bearer of a truth never before attained (B 108), and with this emerges a no less sui generis apophthegmatic way of averring it in keeping with its own predilection for self-obfuscation. At the same time, and indeed for just this reason, he has the option of seeming to advert to the truisms of folk wisdom while quietly introducing a twist which changes everything, if only it is caught sight of. Let us suppose that he carefully constructed this particular saying so that such a discovery might leap out at us from under the straightforward sense which we only too readily expect and imagine the saying to have. Where else would this discovery lurk but in άσσα, which as it happens forms the exact centre of the saying in terms of syllable count (excluding length)? Moreover, in a Heraclitean saying economy is everything; it is very much a matter of saying as much as possible in as few words as possible, with each word given as much semantic weight as it can bear. Can it be ruled out that άσσα is used in order to say something more and different than Heraclitus could say using ά, the neuter plural of ός?14

In summary, then, either Heraclitus never uses όστις in its basic sense, or we may have to radically revise our understanding of one of his fragments, and make some adjustments to our understanding of another. If we follow the commentators and assume the former, then we must explore how, if not through όστις, Heraclitus expresses "whosoever" or "whichsoever" if he needs to.

Let us start with ός. This is used as the relative pronoun "who" in B 56 (referring to Homer) and B 39 (to Bias). In B 102 ά, the neuter plural accusative, is used twice to distinguish "some things" which men deem unjust from "others" they deem just. LSJ cite this as an instance of ός as a demonstrative pronoun in oppositions. In two other fragments ός is used in conjunction with άν, with a sense that is more or less identical to the radical sense of όστις. B 85: "It is difficult [for soul] to fight passion, for whatever [ό άν, singular] it wants it buys with soul." B 97: "Dogs bark at whomever [ών άν, plural] they don't recognise." We might conclude that Heraclitus would similarly use ός + άν in B 5 and B 27 if he needed an indefinite pronoun there. This may be somewhat hasty, since we only find this construction twice. However, there are several other terms Heraclitus employs to convey a similar sense. The most frequent is the adjective όκόσος, Ionic for όπόσος, which in general differs in denoting a quantitative indefiniteness:

B1: "all other men are oblivious to whatever [όκόσα, plural: all the things, however many or however great or small] they do while awake, just as they forget whatever they do while asleep."

B21: "death is whatever [όκόσα] we see while awake, as sleep is whatever we see while asleep."

B108: "Of the however many [όκόσων: genitive plural] accounts I have listened to, none gets to the point of recognising that which is wise, distinguished from all things."

B110: "It is not better for men to get whatever [όκόσα] they desire."

B114: The divine law (or perhaps simply the divine) "rules to whatever extent [τοσούτον όκόσον] it wishes".

Another is the qualitative counterpart όκοΐος (όποιος):

B1: "words and deeds such as [όκοίων, genitive plural: of whatever kinds] I describe".

B5: όκοΐον with εϊ: "as if one were to carry on a conversation with houses".

B17: "Most people do not understand such things [τοιαυτα] in the ways [όκοίοις: dative plural] that they actually encounter them."15

B31b: the sea that pours out from earth "is measured in the same amount as [όκοΐος] existed before it became earth".

We also find the temporal counterpart όκόταν in B 117: "whenever a man is drunk...". Finally, όσος appears in two fragments. B 55 esteems "whatever things [or as many things as, όσων] are of sight, hearing and learning". And in B 56 the boys who befuddled Homer say, not of the fish they were trying to catch but of the lice they caught while fishing: "As much as [όσα] we saw and caught we leave behind, while as much as we did not see or catch we carry with us."

So what are the implications of this survey? Does it compel us to go along with the standard translations of B 5 and B 27, and deny that Heraclitus uses όστις there in its radical sense? I do not think so. None of these other terms seems quite apposite to the specific way that indefinite objects would be in play in either fragment. In the case of B 5, the quantitative indefiniteness denoted by both όκόσα and όσα is clearly impertinent; it is not a question of however many gods and heroes are. Nor would there be a reason to use the qualitative όκοίοι (of whatever kinds they are). But what about οϊ άν? It is notable that in both uses of ός + άν the following verb is in the subjunctive,16 giving a strong sense of circumstantial dependence: whatever passion happens to want (B 85); dogs bark at anyone whom they happen not to recognise (B 97). In B 5 such dependence is moot, not only in view of the indicative verb είσι but in terms of the thought expressed. There may be room for it regarding which god or hero a person happens to pray to, but Heraclitus' point, I have argued, is that what is important is not the who but the what, the impersonal wisdom they all come down to, and there is nothing circumstantially dependent about that. I conclude that in B 5 οϊτινές is the word that best fits the sense of "whoever/whichever they are", and it seems to me that there is a good case for the word having this sense here.

In the case of B 27 these arguments seem prima facie less applicable. The things men "do not expect for themselves or imagine" might have the quantitative indeterminacy denoted by όκόσα or όσα ("any number of things which"), the qualitative indeterminacy of όκοΐος ("all manner of things which"), or the dependence on circumstances denoted by ός + άν + subjunctive. Thus it seems that Heraclitus might well have used one of these forms in B 27 if he intended an indefinite plural, and that άσσα need not be indefinite. Yet here too there are counterarguments. What men do expect may depend on the circumstances of their lives, but this does not hold in the same way for what they do not expect, if this means what they should expect. Whatever our concrete situation, we should live in view of the unity of opposites and expect whichever side of an oppositional pair is relevantly absent to have its season, and whichever is present not to last forever, rather than expecting one-sidedly, e.g., health over illness, satiety over hunger, rest over distress (B 111). Presumably for Heraclitus expecting for oneself entails this one-sidedness and blinds one to the unity of opposites, the central insight he wishes to convey.17 Similarly, quantitative indeterminacy need not be the point here. Heraclitus could be speaking of "whichever things men do not expect" on a case by case basis. It is notable that άσσα is used with the singular verb μένει. Although this is common Greek practice with a neutral plural subject, Heraclitus may be exploiting this quirk of his language with a specific purpose in mind. Viewed collectively it may be a question of things men do not expect in all their quantitative and qualitative variety, but in each case there may be one thing above all that awaits any given man as that which he did not expect for himself or imagine in his lifetime.

As a first approach to this, let us suppose that there is one core tension that most determines any particular person's life, a specific polarity of contraries whose unity he should come to understand and whose reversibility he should learn to expect. For one the core issue of his life may be why friendship leads again and again to betrayal. For another it may be why he finds himself hurting the one he loves time and again. One may seek company to escape loneliness, but find himself more lonely in that company than he ever was by himself. Another may understand very well what Kafka meant by "human intercourse tempts one to introspection", only to find the solitude he flees to haunted by the images and voices of those he left behind. In these cases whatever we do not expect assails us within life; it does not "await" us when we die. What does, then? If we can experience such reversals without yet attaining insight into the unity of opposites which makes them possible, perhaps it is just this that awaits us, this oppositional unity which we do not expect for ourselves, as if we were somehow exempt from it (cf. B 2). In that case, the specific "core tension" of any individual's life will simply be his route to the general insight into the unity of opposites, resolute inherence in which would lead one to take on the impersonal character of "that which is wise", the One. But we may remain so immured in the form of expecting-for-self that it is only in death, when we are divested of our personal selves and our personal investments in one side of a polarity, that this insight can come.

But there may be an additional dimension to this, in light of B 18: "If one does not expect the unexpected one will not discover it, for it is not to be discovered and intractable" (starting from ordinary, i.e., one-sided, modes of expectation).18 By the unexpected I take Heraclitus to mean the unity of opposites, which most men, as he constantly complains, fail to recognise. Expecting the unexpected is here deemed to be transitive. It brings about something which could not otherwise occur. What? Heraclitus says only that it "dis-covers" (έξευρήσει) what is not to be discovered otherwise, brings it out of hiding. Ordinarily the unity of opposites hides in that we fail to see the "one and the same" that alternatingly adopts opposite faces, be it the cosmic god which alternates between day and night, winter and summer, war and peace, satiety and hunger (B 67), or that one and the same which alternates between living and dead no less than between wakefulness and sleep (B 88). The point is not just that we err in seeing two separate things taking turns in such cases. If we do not see the "one and the same" in both, then neither can we see that this simultaneously shows itself and conceals itself either way. Having given itself as summer, it does not then say "look, I am winter too", but simply gives itself as winter to remain concealed as their inner unity. If the alternation of opposites is thus bound up, not only with their unity, but with the essentially recondite nature of that unity, a possible inference is that the discovery of this unity arrests the alternation.

Heraclitus nowhere avows that expectation in its more ordinary forms is transitive, but let us suppose that he thought it. In the examples of "core tension" put forward above, it might be said not only that we expect one thing (e.g., friendship, and the trust it implies) but find the contrary (betrayal), but that by expecting the one we "kindle" the other, in something like the way that, as Heraclitus says (B 26), one kindles (άπτεται) the dead in one's sleep, or kindles what sleeps while one is awake - unable, perhaps, to "let sleeping dogs lie". Thus the difference between ordinary expectation (of A rather than B) and expecting the unexpected (the unity of A and B) seems to be that the former remains blindly subject to the alternation of opposites, whereas the latter masters and discharges it, puts it out of play. If this holds, the point of B 27 on the alternative reading becomes more visible. Assuming that "whatever things men do not expect for themselves" = each one's "unexpected", that specific core tension or oppositional unity which is as it were the hidden meaning of his life, it follows that if we are able to recognise and discharge this now, by expecting the unexpected, then, whatever it be in our case, it does not await us after death.

I would like finally to place B 27 in relation to B 62, as the latter may shed further light on this point. In B 62 the first two clauses assert outright that immortals are or become mortals and vice versa; the last two clauses, ζώντες τόν έκείνων θάνατον, τόν δέ έκείνων βίον τεθνεώτες, literally "living the death of those, having died the life of those", seek to clarify this claim. How they do so is the question.19 By using έκείνων twice, rather than setting off "of those" against τώνδε or τούτων, "of these", Heraclitus clearly intends the same object in each clause. To translate "of these" and "of those" is to undermine his point by insisting on a fixed antithesis between mortals and immortals. Furthermore, in juxtaposing the perfect participle τεθνεώτες with the present participle ζώντες, Heraclitus clearly means that the same subject undergoes at different times the experiences referred to in clauses three and four. Thus, instead of regarding immortals and mortals inversely as the subject of one clause and the object of the other, we must see one and the same immortal-mortal something as somehow both the subject and the object of each.

This entails a double reading of the two clauses, and the only commentator, in my view, who has not only grasped this but provided a compelling interpretation on its basis is Leon Ruiz. For him (2007: 70) Heraclitus means both "Mortals live the death of immortals. Immortals are dead in the mortal life", and "Immortals live the death of mortals. Mortals are dead in the immortal life." My only criticism is that Leon Ruiz uncritically adopts Kahn's rendering of τεθνεώτες as "dead (in)"; a verbal rather than adjectival sense is needed with the direct object τόν βίον. Since I am persuaded by Leon Ruiz's view (2007: 63) that we ourselves are meant in each case, as "fallen" immortals with the possibility of returning to immortality,20 I do not wish to deny the idea that the immortals we were are dead in the mortals we are (and vice versa). There is some independent evidence that Heraclitus thought this, like the Orphics and Plato. According to Sextus (Outines of Pyrrhonism, 3.230): "Heraclitus says that both life and death are in both our living and our dying; for when we live our souls are dead and buried in us, but when we die our souls revive and live." But it is not the direct message of the statement.

Whatever misgivings we may have about assimilating Heraclitus to the Orphic-Pythagorean tradition,21 or to later thinkers influenced by it like Empedocles and Plato, the shoe seems very much to fit in this case. Above all, this reading makes sense of the text, and explains as no one else has why Heraclitus composed it in the form he did, killing two birds with one stone. Our mortal lives take off where our immortal one ends, and in their development deepen the loss, so that we become increasingly dead to intimations of the Wordsworthean kind, but in a previous round our mortal lives will have made way for the immortal one in the same fashion. Likewise, to gain immortality is to cease to be the mortal selves we were, having likewise made way for them on some previous round. All this is said by "living their death, having died their life". In both cases there is a sharp discontinuity. Each is not only radically different than the other but dead to it. Yet the disconnect is not absolute; there is a "one and the same" enduring throughout these "turnabouts" (cf. B 88) between the mortal-personal and the immortal-impersonal. This means that we can, within mortal life, wake up to the immortal dormant within us, and Leon Ruiz provides a strong case that such "work of awakening" is the fundamental character of Heraclitus' teaching and his own experience.

What is unclear is why we fall to begin with. As Leon Ruiz (2007: 82) observes, we do not find anything in Heraclitus corresponding to Empedocles' claim that this is due to sins committed in our divine life. He takes the more cautious view that "for Heraclitus, human existence is the troubled dream of an immortal who has fallen asleep" (2007: 83). There is another possibility. If wisdom means recognising the underlying unity of opposites, this unity may be so close to us as immortals that the opposites get seen through too quickly. It may be that we snuff ourselves as immortals to rejuvenate the insight through a head-on encounter with opposition in its manifold forms. This may explain the transitivity of dying the life of immortals. It may also account for the cyclicality of the process. Presumably one lifetime would rarely be enough to master the unity of opposites in all its forms. In mastering one we may be broadsided by another, which we may then have to work through in a future round of mortal life, via a stopover in which our mortal selves are latent.

To close, I suggest that the fuller contextual picture offered by this approach to B 62 may give extra semantic weight to the theme of "awaiting" in B 27. What awaits us after death is not purely and simply a product of our actions and omissions in life. Leaving aside the differences, just as the Buddhist thought of karma is not simply a matter of how our present acts determine our future state in this life or the next, but how our past acts have already determined our present one, so, with Heraclitus, it may be that it is how we deal with the things that already await us when we come to life that determines what awaits us when we die. In other words, it may be that the "core tension" of our mortal (personal) lives, whatever it be in each case, is not simply of this life but something we have been despatched into this life in order to come to terms with, and something that, if we fail to do so, will still await us after death, so that we are despatched into mortal life again and again until we master and discharge it.



1 I leave out of account B 41 because the text is corrupt at this point; some see here an otherwise unattested feminine form, ότέη or ότέη.

2 Of these Robinson's "without any recognition of who gods and heroes (really) are" is technically most correct, in reflecting the different relations of τι and οϊτινές. Compare Patrick (1889: 113) and Osborne (1997: 83).

3 Aeschylus, Agamemnon 161; cf. Euripides, Women of Troy 885-7, where όστις could signify not "whoever" but "whatever": "be thou necessity of nature or man's intelligence."

4 See, e.g., Wheelwright (1959: 72), Marcovich (1967: 462); Kahn (1979: 266); Kirk, Raven and Schofield (1983: 212); Robinson (1987: 78).

5 This is as far as I can go with Robinson here. I see no real evidence of an aether doctrine in Heraclitus, nor of star-gods inhabiting aether and exemplifying soul as pure fire as distinct from aether

6 As cited by Clement, Stromateis 4.144.3. The partial citations of Theodoretus and Plutarch have, among other variations, όσα instead of άσσα.

7 Cf. Diels (1903: 70), Freeman (1948: 26), Marcovich (1967: 401), Kahn (1979: 67), Robinson (1987: 23), and Hussey (1991: 524) who adopts the passive. Marcovich and Kahn translate άσσα with the numerically ambiguous "what". Kahn also departs from the usual practice (given the deferred subject) of rendering B 27 as an existential sentence.

8 "Neither hope nor believe" (Zeller 1856: 483; cf. Van Ackeren 2006: 35; Barnes 1987: 64); "neither hope nor think" (Patrick 1889: 112; cf. Osborne 1997: 95 who adopts the past tense, and renders μένει in its intransitive sense "remains"); "look not for nor dream of (Burnet 1920: 105); "neither expect nor have any conception of" (Wheelwright 1959: 68), "do not expect or think" (Guthrie 1962: 477). Wheelwright renders άσσα "such things".

9 Nussbaum too accepts this much. On her reading (1972, Part II: 158), B 27 is a tricky way of saying that nothing awaits men after their death. While I share Nussbaum's view that men do not live on after death as the individual personalities they were in life, this reading of B 27 is impossible to reconcile with the plural άσσα.

10 έλπονται is middle-voiced.

11 Cf. B 25: "Greater deaths win greater fates."

12 A better comparison would be with B 119. Whether or not one accepts the standard translation "man's character is his fate", B 119 need not entail negative implications for man as such.

13 Or the sudden, intuitive resolution of a Zên koan, although koans generally do not contain in their linguistic composition essential clues to the insight that is meant to come through pondering them, and Zên takes a far more sceptical, if not outright negative, view of the capacity of language to disclose essential (and for it essentially ineffable) truths than Heraclitus does. On the disclosive importance of language for Heraclitus, see Nussbaum (1972, Part I: 8-15).

14 It is not simply the negatives it is used with that explains the use of άσσα, since, according to Smyth (1956: §2496a), this applies specifically "after a negative", in constructions like "no one who" or "nothing which". It is in cases like these that όστις is used instead of ός "because of its general meaning", but otherwise with the same sense as ός.

15 There is some ambiguity here, for τοιαύτα has an adverbial sense, "in such ways", and όκοίοις could imply the things of whatever kind that people encounter, but either way the statement's meaning is basically the same. See Kahn (1979: 102-3).

16 Contrary to LSJ's claim that this only occurs in Epic.

17 Recognition of the unity of health and illness enables understanding that "illness makes health sweet and good" (B 111). The health that is cleaved to one-sidedly and kept apart from illness does not have the same quality as the health restored after a bout of flu.

18 Marcovich (1967: 40) stresses that άνεξερεύνητον and άπορον entail only difficulty of discovery, not impossibility. I accept this; the extrapolation I add goes to the conditions under which discovery of the unexpected remains impossible.

19 Here I can offer only a fairly brief discussion of B 62.

20 Cf. Guthrie (1962: 478) on Heraclitus' embrace of the "Orphic doctrine of man" in B 62. Hussey's objections (1991: 520) to interpreting B 62 to say that "all souls alternate between states of life and death" do not perturb me. In particular, his worry that souls are immortal only in a "contrived" or "half-time" sense seems misplaced, as this is the very point of "immortals mortals, mortals immortals". Souls are confined to this half-time immortality unless the soul learns to discharge the alternation of opposites and frees itself from the round of its own alternation between "life" and "death".

21 Heraclitus arguably expresses more contempt for Pythagoras than for any other predecessor. However, Hussey rightly comments that his dismissal of Pythagoras as a fraud and plagiaristic polymath does not exclude "close agreement on some points" (1991: 529 note 32), which may well include metempsychosis in some form.



Barnes, J. 1987. Early Greek philosophy. Harmondsworth: Penguin.         [ Links ]

Betegh, G. 2007. On the physical aspect of Heraclitus' psychology. Phronesis, 52: 3-32.         [ Links ]

Burnet, J. 1920. Early Greek philosophy. London: A & C Black.         [ Links ]

Diels, H. 1903. Die Fragmente der Vorsokratiker. Berlin: Weidmannsche Buchhandlung.         [ Links ]

Freeman, K. 1948, Ancilla to the pre-Socratic philosophers. Cambridge, MA: Harvard University Press.         [ Links ]

Guthrie, WKC. 1962. A history of Greek philosophy, Vol 1. Cambridge University Press.         [ Links ]

Hölscher, U. 1993. Paradox, simile and gnomic utterance in Heraclitus. In The pre-Socratics: a collection of critical essays, Editor A. Mourelatos, 219-38. Princeton University Press.         [ Links ]

Hussey, E. 1991. Heraclitus on living and dying. The Monist, 74: 517-530.         [ Links ]

Kahn, CH. 1979. The art and thought of Heraclitus. Cambridge University Press.         [ Links ]

Kirk, GS., Raven, JE. & Schofield, M. 1983. The presocratic philosophers. Second edition. Cambridge University Press.         [ Links ]

Leon Ruiz, NE. 2007. Heraclitus and the Work of Awakening. PhD thesis, State University of New York.         [ Links ]

LSJ: Liddell, HG. & Scott, R. revised by HS. Jones. A Greek-English Lexicon. 9th edition. Clarendon Press. 1940.         [ Links ]

Marcovich, M. 1967. Heraclitus. The Greek text with a short commentary. Merida: Los Andes University Press.         [ Links ]

Middle Liddell: Liddell, HG. & Scott, R. An Intermediate Greek-English Lexicon. Clarendon Press. 1889. (Founded on the seventh edition of LSJ.         [ Links ])

Nussbaum, M. 1972. "ψυχή in Heraclitus", parts I and II. Phronesis, 17: 1-16, 153-70.         [ Links ]

Osborne, C. 1997. Heraclitus. In Routledge history of philosophy, Vol 1, Editor CCW. Taylor, 88-127. Routledge: London.         [ Links ]

Patrick, GTW. 1889. The fragments of the work of Heraclitus of Ephesus on nature. Baltimore: N. Murray.         [ Links ]

Robinson, TM. 1987. Heraclitus: fragments. Toronto Buffalo, London: Toronto University Press.         [ Links ]

Smyth, HW. 1956. Greek grammar. Cambridge, MA: Harvard University Press.         [ Links ]

Van Ackeren, M. 2006. Heraklit: Vielfalt und Einheit seiner Philosophie. Bern: Peter Lang.         [ Links ]

Wheelwright, P. 1959. Heraclitus. Princeton University Press.         [ Links ]

Zeller, E. 1856. Die Philosophie der Griechen in ihrer Geschichtlichen Entwicklung, Part 1. Second edition. Tübingen: Fues.         [ Links ]