SOURCE: Edward Hussey, "Heraclitus," in The Presocratics, Duckworth, 1972, pp. 32-59.

[In the following excerpt, Hussey provides an overview of Heraclitus 's thought, particularly his concept of logos, and contends that the point of many of his paradoxical writings was to offer analogies.]

In the middle of the sixth century, the Ionian cities of the Asiatic coast had for some time been tributaries of the kingdom of Lydia. The Lydian kings were not hard masters; the last of the line, Croesus, was distinctly phil-Hellene and was sincerely admired by many Greeks. But in 546 Croesus was defeated by the invading Persians under Cyrus, and the Ionians were faced with subjection to a new and much greater power. The new Persian empire was far more efficient and on a larger scale than anything these Greeks had yet seen. The Ionian territory would form an insignificant and peripheral part of the vast areas controlled by the Great King from the region of modern Iran.

Thales, according to Herodotus (1 170), had proposed that the Ionians should form themselves into a federal state to meet the Persian danger. This novel suggestion was not adopted, and the coastal cities were reduced to submission one by one. Many Ionians emigrated; the Persians seem to have interfered in the internal politics of the cities and to have checked the spread of political equality—or it was feared that they would do so. The citizens of Phocaea, a small Ionian city whose men were famous for expert seamanship, migrated in a body and after various adventures founded the city of Elea in southern Italy. The new city was to be famous for nothing but philosophy.

Ionia remained under the Persians for seventy years. During this time, only two Ionian thinkers are known to have been active who in any way carried further the ideas of the Milesians. One was Xenophanes of Colophon, whose theological utterances were used in the last chapter. But Xenophanes, a poet and a professional reciter of poetry who wandered about the Greek world, was not an original or systematic thinker. There is nothing to suggest that he tried to improve the Milesian framework where it seemed in danger of inconsistency. His own views were in certain crucial places vague and incoherent, if we may trust Aristotle, and in one fragment (fr. 34) he takes refuge in the thought that, after all, no man can know anything for certain. Still, Xenophanes is illuminating: he casts light, as has been seen, on the Milesians, and his difficulties and deficiencies cast light on the situation as it presented itself to the other Ionian thinker of this period, Heraclitus of Ephesus.

Heraclitus seems to have been born around 540 and to have lived past the turn of the century. Nothing suggests that he ever left his native town of Ephesus. His utterances suggest a striking and original personality in a partly self-willed intellectual isolation. In later centuries a great deal of biographical fabrication was called forth by the fact that Heraclitus was revered by the Stoics and had obviously been a remarkable man. Some of the stories invented about him are still occasionally repeated as true, but there is in fact hardly any reliable information about the life of Heraclitus beyond what little can be inferred from the fragments. Of the actual words of Heraclitus there survive, cited by later writers, some eighty or ninety fragments, while there are many other passages in which his words are paraphrased or alluded to. The difficulty of interpreting his statements was already complained of by Aristotle and Theophrastus; Aristotle noted that they were frequently ambiguous, and Theophrastus remarked that Heraclitus sometimes left them incomplete, sometimes contradicted himself. Thereafter he was a byword for obscurity. So far as can be judged from our evidence, most of what he wrote consisted of isolated statements, terse and intentionally memorable prose sentences composed with great attention to rhythm and lacking for the most part any reference to a context. Lengthy explanations are avoided, and meaning is concealed in allusions, puns, portmanteau-words and ambiguities of construction. This style has often enough been compared to that of an oracle or a prophet. Yet it would be wrong to suppose that Heraclitus chooses obscurity entirely for its own sake, or for its effect on his audience. In many fragments, especially when denouncing the stupidity of ordinary people, Heraclitus speaks plainly enough; in some other cases, the fragments as cited may or must have been deprived of a context it originally had. In the only fragment in which Heraclitus offers a description of his own activity, he claims to be 'defining each thing according to its nature, and showing how it is'. The fact seems to be that Heraclitus believed his style of utterance to be uniquely suitable to his subject-matter. If his utterances are like riddling oracles to most men, this is appropriate, since he believed, as will be explained, that the truth about things was like the meaning of an oracle, or the solution of a riddle, of which most men saw only the meaningless or misleading (perhaps deliberately misleading) exterior.

Heraclitus, in fact, seems to be the first exponent of an idea which has since been endemic in philosophy: that language must be used in a highly unusual way in order to fit the nature of things, which is conceived as being radically different from anything that is generally supposed. It is no mere accident that his style of thought and of expression is often reminiscent of Wittgenstein's Tractatus Logico-Philosophicus, and the resemblance goes even deeper than has been suggested. (Some ideas about the reasons for this resemblance will be offered at the end of this chapter.)

A careful and sympathetic examination of the fragments is the necessary beginning of any attempt at an interpretation of Heraclitus. Many of them can hardly be rendered adequately in English; nevertheless, English versions have been given here of all important fragments, as an inadequate English version is better than nothing. But it must always be remembered that such versions are only second-bests, and may already, in spite of the best intentions, incorporate wrong or unwarrantable assumptions about the meaning of Heraclitus.

It is characteristic of Heraclitus that his attitude to the fundamental assumptions of the Milesians is already highly critical. Like the Milesians and Xenophanes, he believes that there is a single all-powerful deity in control of the universe; but on the questions arising from such a belief—the relation of the deity to what it controls, and the possibility of discovering the laws which it imposes—he takes up positions which appear to be subtler than those of his predecessors.

To begin with, the problem of knowledge. That men are easily deceived, that the scope of their knowledge is at best very limited, that their judgment is the creature of circumstances—these are themes of the earliest Greek literature. By contrast the Milesians must be assumed to have been highly optimistic about the possibility of true knowledge—an optimism which remained part of the Presocratic tradition. But Xenophanes intro duces a deeper pessimism:

There is no man that has seen, nor any that will ever know, the exact truth concerning the gods and all the other subjects of which I speak. Even if a man should chance to speak the most complete truth, yet he himself does not know it; all things are wrapped in appearances (fr. 34).

The thought that reality is uniformly concealed from men by an impenetrable veil of 'seeming' (dokos) is quite new. It is natural to connect the appearance of this thought with the ideas of Anaximenes and Xenophanes on the unity of the universe. If God is all things, then appearances are certainly deceptive; and, though observation of the kosmos may yield generalisations and speculations about God's plans, true knowledge of them could only be had by a direct contact with God's mind. There are in this fragment the roots of various philosophical doubts which have often been advanced against scientific explanations.

Heraclitus agreed, as will be seen later, with Anaximenes and Xenophanes that the universe was a unity. Consequently, he necessarily agreed with them that reality was to some extent 'hidden':

The nature of things (phusis) is in the habit of concealing itself (fr. 123);

Latent structure is master of obvious structure (fr. 54).

But on the possibility of true knowledge for men his ideas were more complex and less totally pessimistic than those of Xenophanes, and they can be recovered from the fragments which spread themselves round this topic.

First, there are the remarks which associate knowledge with God, and emphasize the gulf between God and man:

It is characteristic of God, but not of man, to have discernment (fr. 78);

A man is considered silly by God (or: a god), as a child is by a man (fr. 79);

Of all whose words I have heard, none has reached so far as to know that the wise is different in nature from everything else (fr. 108);

One thing, that alone is wise, is unwilling and is willing to be called by the name 'Zeus' (fr. 32);

One thing is wisdom: to be skilled in the plan upon which all things are controlled throughout the universe (fr. 41).

The first two of these fragments are clearly expressed; it is important that they do not exclude all possibility of human knowledge. Men may, uncharacteristically, come to understanding, just as children can be educated. The next two fragments make wisdom the exclusive property of God, who as the underlying and controlling unity in the universe can be said to be 'different from everything else'. The name 'Zeus' is appropriate because Zeus is traditionally the strongest god, and his association with the thunderbolt was important for Heraclitus. There is also a characteristically Heraclitean reference to the verb zēn (to live) in the form of the word 'Zeus' used by Heraclitus here. He wishes to suggest that wisdom is the attribute of what is the only truly living thing. Such allusions depending on similarities between words are frequent in Heraclitus, for whom they contain important truths. (So too in fr. 79, the word used for 'god' is daimon, which can also be taken to mean 'he who knows'.) The name 'Zeus' is, on the other hand, inappropriate because of its traditional associations.

It may be that the 'wisdom' of which these fragments talk is to be distinguished from mere knowledge. The adjective sophos had, traditionally, practical connotations: a 'wise' man was one who could advise, act or perform well in some respect. The wording of the last fragment quoted above is unclear, and it is possible that it means that wisdom is knowing how to control the universe, not merely knowing how it is controlled.

However that may be, it is clear from other fragments that Heraclitus thought that some measure of true knowledge was attainable by men, and that he said something of how it was to be attained:

All that can be learnt of by seeing and hearing, this I value highest (fr. 55);

Men who love wisdom must be knowers of a great many things (fr. 35);

Much learning does not teach men understanding (fr. 40: part);

Bad witnesses to men are eyes and ears, when they belong to men whose souls cannot understand their language (fr. 107).

Here again Heraclitus takes up a position which is a refinement of that of the Milesians. He reaffirms the importance of first-hand inquiry, as in other places he attacked the dependence on the authority of tradition. But the results of such inquiry are not an end in themselves: it is necessary to have understanding (noos), in order to be able to interpret the evidence of eyes and ears. The step from the obvious to the latent truth is like the translation of utterances in a language which is foreign to most men. Heraclitus offers two other similes for this step. Fr. 56 says that men, in regard to knowledge of perceptible things, 'are the victims of illusion much as Homer was', and goes on to explain this remark by telling the story of how Homer was baffled by a riddle put to him. In fr. 93 he remarks, significantly:

The prince whose the oracle at Delphi is neither tells nor conceals: he gives a sign.

To reach the truth from the appearances, it is necessary to interpret, to guess the riddle, or divine the meaning of the oracle. But though this seems to be within the capacity of men, it is something most men never do. Heraclitus is very vehement in his attacks on the foolishness of ordinary men, and of what passes for knowledge among them. They are compared to sleepers in private worlds of their own (frr. 1, 2, 73); to children who believe what their parents tell them (fr. 74, cf. fr. 79); to dogs who bark at strangers (fr. 97); to deaf men, 'as good as not there' (fr. 34); to idiots who take fright at any sensible utterance (fr. 87). Their opinions are toys with which they childishly amuse themselves (fr. 70). The points made by these different similes are clear, but sometimes the remarks are even more explicit:

Most men do not have thoughts corresponding to what they encounter; they do not know what they are taught, but imagine that they do (fr. 17);

Other men are unaware of what they are doing when awake, just as they are of what they forget about in sleep (fr. I, part);

It is mere appearances that the most reliable of them keep hold of (fr. 28, part).

Together with these fragments go those which assail particular individuals by name. Heraclitus singled out for attack all those who had a wide reputation for knowledge or wisdom among his contemporaries: Homer, Hesiod, Archilochus, Hecataeus, Xenophanes, and Pythagoras (frr. 40, 42, 57, 129). It is perhaps significant that the Milesians were, so far as is known, spared, though there is implied criticism of many of their views in many fragments, and of Anaximander in particular in at least one. These wholesale denunciations are certainly suggestive of mania, not because they are so sweeping but because so much energy has clearly gone into the making of them. It must be remembered that Heraclitus is not writing to be read only, but to be heard and to resound in the memory.

If Heraclitus is to be so insistent on the lack of understanding shown by most men, it would seem only reasonable that he should offer further instructions for penetrating to the truth. The talk of riddle-guessing suggests that some kind of revelation, beyond human control, is necessary, and this is perhaps confirmed by another cryptic fragment:

If one does not expect it, one will not find out the unexpected; it is not to be tracked down and no path leads to it (fr. 18).

Yet the application of this remark is uncertain. It must be left undecided how men achieve true insight, though the necessity of some revelation is plausible. The true wisdom, as has been seen, is closely associated with God, which suggests further that in advancing in wisdom a man becomes like, or part of, God. Some hints as to how this is done will be found later in the chapter.

If Heraclitus does not tell men how to achieve true wisdom, he does the next best thing. He communicates to them, in a suitably oracular style, the truths that his own insights have shown him. These truths are the content of what Heraclitus calls 'the logos' 'logos' is for him clearly a technical term in some fragments, but its meaning is not immediately obvious.

The word 'logos', in ordinary Greek of this period, has a family of meanings: 'word', 'story', 'reckoning', 'proportion' are all possible renderings in different contexts, and all are relevant to its use by Heraclitus. The three fragments where it seems to be used as a technical term are as follows:

Of this logos which is so always men prove to have no understanding both before they have heard it and immediately on hearing it; though everything comes to be according to this logos, they are like persons who have no experience of it … (fr. 1, part);

Though the logos is common (xunou), most men live as if they had a private source of understanding (fr. 2);

Having heard not me, but the logos, it is wise to concur that all is one (fr. 50).

The last of these fragments shows that the logos is independent of what Heraclitus himself may have to say. The first suggests that it is the expression of the cosmic law; and this is borne out by a further fragment, which also sheds further light on what is meant by calling the logos 'common':

Those who speak with understanding (xun noōi) must make themselves strong with what is common (xunōi) to all, as a city does with its law, and far more strongly than that. For all human laws are nourished by the one divine law … (fr. 114, part).

The analogy here intended is that the law of a city makes it strong because it stands above private interest; it is therefore 'common' to all both in the sense of 'impartial' and in the sense that all alike share its benefit. In the same way, the logos is equally true and equally accessible for all, and overrides private points of view; as we should say, it is objective, not subjec tive. But the law which the logos expresses, the 'divine law', is far more rigorous and exceptionless than any human law, and therefore to 'speak with understanding' involves being more carefully 'law-abiding' than any good citizen.

The minimal sense that can well be given to the word logos here is therefore something like: 'the true account of the law of the universe'. But if that were all, it would be difficult to explain why Heraclitus chooses to signify this by the single word logos. Given his habits, it may be suspected that there are further layers of meaning concealed in the word. A clue may be found in a development of meaning in logos around this time. In the first half of the fifth century the sense of 'reason' or 'reasoning' appears to be well established. This sense is presumably a development from the meaning 'proportion', which is already attested in Heraclitus (fr. 31). What is reasonable or unreasonable is in or out of proportion in some sense. Though there is no direct evidence, it is likely that this development was already proceeding at the end of the sixth century, and that Heraclitus is playing upon it here. If this is correct, his thought is that the logos expresses a proportion or analogy in the universe; and, therefore, that the logos is reasonable and the law it expresses, in virtue of this proportion. The reasonableness of the logos would further resolve a problem to which Heraclitus does not directly offer a solution: what public evidence of its own truth does the logos carry with it? It cannot be a merely personal revelation.

All this remains unconfirmed speculation, unless it can be shown that Heraclitus does indeed make appeals to proportions and analogies in his account of the Universe. In fact, some such appeals have already been quoted: all similes can be seen as analogies, and there is a direct statement of proportion in fr. 79 (man: god:: child: man). In the rest of this chapter, it will become evident that the idea of analogy as a guide to the truth was indeed present in the mind of Heraclitus, and that the attempt to create analogies between words and things, sentences and states of affairs, is one of the principles of his extraordinary style. Above all, it is their value in providing analogies that explains the point of a large number of fragments in which Heraclitus describes various paradoxical states of affairs drawn from everyday experience. To these fragments we now turn.

Consider the following remarks:

Sea is purest and most unclean water: for fish, drinkable and life-giving; for men, undrinkable and deadly (fr. 61);

Of the bow the name is 'life', the work is death (fr. 48); [The word bios could mean either 'life' or 'bow'];

A road is, upwards and downwards, one and the same (fr. 60);

Sickness it is that makes health pleasant and good, and so with hunger and satiety, weariness and rest (fr. 111).

These four fragments have been chosen to stand here because in them the actual words of Heraclitus are not in any doubt. They are representative of a whole group of fragments of which, as here, the point is drawn in a single way from ordinary experience. Each of the first three presents for consideration a single familiar subject, and then proceeds to show that within the unity of that subject there coexist opposites of some kind. The fourth is less typical in not giving an explicit unity, but instead demonstrating that the opposites named are dependent upon each other for their most essential qualities. Another important difference is that in the fourth fragment cited the opposites in question seem to be thought of as successive, not as coexistent.

It is necessary to remember that to look at these fragments in an abstract way is to impose upon them a pattern that may be misleading. Heraclitus had no abstract vocabulary at his command. At least one of the fragments presented, the 'road', can be read as asserting, not the coexistence of opposites in a single subject, but the identity of opposites. 'The road up' is the opposite of 'the road down': yet they are identical. Other fragments make it clear that Heraclitus did indeed wish to say that certain opposites were not merely coexistent or mutually interdependent, but identical with one another:

[Hesiod] did not know what day and night are; they are one (fr. 57);

The same thing is in us as the living and the dead, the awake and the sleeping, the young and the old; these change to become those, and those change back again to become these (fr. 88).

Here the identity of certain opposites is deduced from the fact that they change into one another, which is a strong form of mutual interdependence. If the ideas of Anaximenes and Xenophanes are recalled, it is clear what is Heraclitus' motive, at least in part, for these unifications. He is concerned as would be expected, to work out a way in which the universe can be a true unity, while leaving room for the diversity of the perceptible world. Here again he picks up the problems that we find in Xenophanes and tries to work out a better solution. Xenophanes 'gave no clear account', according to Aristotle; 'he simply surveys the whole world and says that the one is God' (Metaphysics 986^b21-5).

These fragments, then, are Heraclitus' analogies, drawn from experience, and aimed at demonstrating the unity, in general, of pairs of opposites. To this end, he collects examples which show opposites inextricably bound up together with each other in various ways. These examples, it is clear, took up a considerable part of his book, and have always attracted, as they were meant to, a great deal of interest. Historically, they are interesting as the first deliberately sought philosophical 'examples', and as the forerunners of the collections of puzzles which were made in the late fifth century for 'sophistic' purposes, and which served as raw material on which, in the fourth century, the creators of logic tested and refined the instruments they had devised. Some of them continue to be relevant to philosophical discussion, and they have stimulated, and continue to stimulate, many imaginations not concerned with philosophy or with cosmology.

From these illustrations of unity in opposites, the path back to the logos that tells us that 'all things are one' runs through two remarkable fragments in which Heraclitus is visibly attempting a generalisation from all the particular situations. Of men, in general, he complains:

They do not understand how what is at variance is in agreement with itself: a back-turning structure (palintropos harmoniē) like that of the bow and of the lyre (fr. 51).

The first part of this statement seems clear. Men fail to understand the general truth illustrated by the 'paradoxes', that what is at variance (from itself) is in agreement with itself. The words translated by 'is at variance' and 'is in agreement' have the primary meaning of 'is drawn apart' and 'is drawn together', and therefore suggest some kind of alternating movement.

In the second part, Heraclitus goes on to give a further description of the general situation he has in mind, and the interpretation of these words has been much debated. The word harmoniē, which here and in fr. 54 has been translated by 'structure', is a noun derived from the verb harmozein, 'to fit together'. The noun appears in Homer with reference to concrete 'fittings together', such as those of carpentry or masonry, but also in another sense, that of 'treaty' or 'covenant'; the verb shows a similar range of meaning in early Greek. The basic notion seems to be that of the mutual adjustment of two or more different components to form a structure which is more than the mere sum of its parts. This basic meaning was still alive in the late sixth century, for it is about this time that it gives off a subsidiary special meaning, that of a 'mode' in music. The thought behind this derivation seems to be that, in tuning his lyre in a certain way, the musician is mutually adjusting the strings so that the notes playable form a particular system of pitch-relationships.

Heraclitus, it is reasonable to assume, uses harmoniē in its widest sense. The translation 'structure' tries to render this breadth of meaning, but fails to capture completely the notion of 'mutual adjustment'. The ways in which the word harmoniē might naturally be applied to a bow and to a lyre are fairly clear. Applied to a bow, it might refer simply to the structure of the arms of the bow, if that were complex, as it was in the composite bow; or to the structure of the strung bow which the arms and the string were mutually adjusted. The second way of taking it is clearly preferable since it applies to all kinds of bow and since the string and the arms pull different ways when the bow is at rest and move different ways when it is drawn, so that they are an apt picture of a marriage of opposites. Applied to a lyre harmoniē might refer to the structure of the unstrung lyre, or to that of the strung lyre whether tuned or not, or to that of the lyre tuned in a particular mode. Here it is not immediately clear what the opposites are which are unified.

To choose between these possibilities, and to grasp the full point of the double simile, it is necessary to understand the adjective which Heraclitus chooses to qualify harmoniē. The witnesses to this fragment are divided between palintropos and palintonos. In either of these the prefix palin- must be taken to mean 'back', 'in a contrary direction'; but -tropos represents the notion of turning, and so of a movement which alters, while -tonos represents the notion of stretching or tension and seems to refer to a static situation. Palintropos is the better attested here directly and indirectly, and Parmenides uses the word palintropos with emphasis in a passage (DK 28 B 6) which may well allude to Heraclitus. But these external considerations are not quite strong enough to be decisive by themselves. Considerations of sense come into play, and these make the choice between the two adjectives part of a wider debate about the form of Heraclitus' thought.

If palintonos is correct, then the bow and lyre are thought of as not functioning, but at rest and in a state of tension, as indeed they both are when strung. If this is so, then the unity of opposites expresses itself most typically in a static state, an equilibrium in which the opposed forces balance each other.

If palintropos is correct, then the bow and the lyre are thought of as in use. Their proper functioning implies the movement in opposite or alternate directions of their complicated structure. This is easy to explain of the bow. To apply it to the lyre, it seems best to take into account the special musical meaning of harmoniē, and to take the movements in opposite directions as the alterations in pitch as the melody ascends and descends. If this is right, the unity of opposites expresses itself most typically in alternating movements in opposite directions.

These two opposed interpretations may be labelled for convenience the 'tension' and the 'oscillation' interpretations of Heraclitus. They both seem to go back to the fourth century at least. This is not the place to decide finally between them. (It may be added that the musical interpretation of palintropos harmoniē, as applied to the lyre, is supported by some further evidence (DK 22 A 22).)

In whichever way it is to be understood, it is clear at least that fr. 51 lies at the core of Heraclitus' thought. A generalised truth about the unity of opposites is expressed by taking the word harmoniē, capable of a wide range of meaning, in its greatest generality and propping it up, as it were, by the similes of the bow and the lyre. In another fragment there is an equally remarkable attempt to generalise along the same lines:

'Conjunctions: wholes and not wholes, the converging the diverging, the consonant the dissonant, from all things one, and from one all things (fr. 10).

It is doubtful what exactly the word translated here by 'conjunctions' was; there is a choice between two readings, but in either case the significance of the word is close to that of harmoniē. The 'conjunctions' are the particular instances of the general situation described in fr. 51. The point of fr. 10 seems to be that in making a general statement about all these instances, it itself furnishes a further instance. The pattern of fr. 10 is that of the typical 'paradox' fragment, exemplified in frs. 60 and 61. As we might put it, the concept of a 'conjunction' is itself a 'conjunction'.

The two fragments that have just been considered lie at the core of Heraclitus' thought. His doctrines on the nature of God and of the observable world, and their interrelations, can be seen as applications of fr. 51. It is this fact that gives unity to his teachings and substance to his claim (implicit in fr. 1) to be following the logos in everything he says.

It was suggested in the last chapter how a new theology initiated a new kind of cosmology, and how fundamental difficulties must quickly have shown themselves. In general, it might be asked of a Milesian thinker whether God was or was not everything. If he is, then there is the problem of explaining the diversity, transience and apparent imperfection and self-contradiction of a great deal of the observable world. But if not, then there is the problem of the relations between God and the rest of the universe, and the nature and mechanism of the control exercised by the one over the other. Anaximander can hardly have provided any precise thought on these questions; Anaximenes and Xenophanes both seem to have attempted a radically unifying solution without, perhaps, becoming aware of the full depth of the problem.

Heraclitus has already been seen to be a 'unifier', and to have rested his claims upon the application of the logos. The most explicit application of the doctrine of the harmoniē to the universe as a whole is the following:

God is day night, winter summer, war peace, surfeit famine; but he is modified (alloioutai), just as fire, when incense is added to it, takes its name from the particular scent of each different spice (fr. 67).

The important pairs of contraries 'day-night' and 'winter-summer' are unified by being different aspects of God, so that God is, at least, that which provides the unity through time of certain large-scale cosmic alternations. There is some reason to think that 'war-peace' and 'surfeit-famine' are Heraclitean names for other large-scale cosmic oscillations. So the unity of the universe will consist, essentially, in the fact that all the large-scale cosmic processes are oscillations in the state of God. Just so a fire on an altar may persist; yet according as different incense is thrown on to it, it gives off a different scent, by which, at each different time, it can be identified. Because the incense is something so inessential to the fire, this simile suggests, far more than that of the bow or the lyre, the existence of a gap between the superficial appearances and the hidden unity.

Fr. 67 shows that for Heraclitus theology and physics are one. But God is not simply, as this fragment might suggest, a lawlike self-regulating mechanism. It was not natural to Greek thinkers of any period to suppose that what was self-moving and lawlike in behaviour was dead or mindless. As for the Milesians and Xenophanes, so for Heraclitus too God is of necessity something preeminently living and intelligent. Xenophanes as usual is instructive: he announces of his God that

without effort, he agitates all things by the thought of his mind (fr. 25)

and that he is

not like mortal men in his bodily form, nor in his thought (fr. 23).

The evidence that Heraclitus' God has a mind has already been presented. It consists of frr. 32, 41 and 108, which attribute wisdom to God, as the only being that knows the plan upon which all things are governed. So the mind of God thinks the plan upon which God acts, and as with Xenophanes there is presumably no hiatus between thought and action. It is natural to identify the plan, the gnomē of fr. 41, with the 'divine law' of fr. 114, which has also been quoted already.

If God has a mind, clearly further problems arise. It is natural to ask where this mind is situated, and what marks it off from the rest of the contents of the Universe. 'Hidden' as it is, it cannot transcend space and time. Anaximander and Anaximenes had done their best in this direction, by placing 'the Divine' beyond human reach in space and depriving it of ordinary physical properties. But for Heraclitus God is everywhere, in the continual changes of the world-order. In what form? Heraclitus seems to have answered: as fire.

That God is a fiery mind is suggested indirectly by the connection of human souls, in their wisest state, with fire (the evidence will be given later), and directly by the evidence which assigns to fire a central and controlling place in the universe:

Thunderbolt steers all things (fr. 64);

This world-order (kosmos) was made neither by god nor by man, but it was always and is and shall be; fire ever-living, being kindled by measures and being quenched by measures (fr. 30).

To which two fragments may be appended the Aristotelian interpretation of Heraclitus, in which fire is the material cause of everything.

It is still not clear in exactly what way the equations: 'kosmos = fire' and 'fire = mind of God' are to be interpreted. Before these questions are further treated, it will be convenient to discuss the 'external' aspect of God, in other words the observable world-order. To this fr. 30 makes a natural transition.

The doctrine of the palintropos harmoniē had interesting repercussions on the general scheme of cosmology. The controlling divinity being no longer spatially separate from the world-order, there was no longer any argument for the plurality of the kosmoi, or for the bounding of the one kosmos in space or in time. Fr. 30 sums up some of this; for the fact that the kosmos was not limited spatially there is only an argument from silence.

The kosmos was still the scene of a constant struggle between opposed forces, as in Anaximander. But the struggle was not, as in Anaximander, something beyond the plan of God, and which the divine justice had to step in and regulate. Heraclitus says, with an emphasis clearly directed against Anaximander:

But one must know that war is universal (xunon), and that justice is strife, and that all things happen according to strife and necessity (fr. 80).

In other words, the perpetual struggle of opposites and the justice that balances them are indistinguishable and both equally present in every event, which occurs as a necessary part of the divine plan. Every event, then, can be analysed into encroachments (in a sense yet to be explicated) of one opposite on another—acts of 'war' or 'strife'. Hence:

War is father of all, and king of all … (fr. 53, part).

But every such encroachment is also 'justice' because it is laid down by the divine law as part of a regular plan. This does not necessarily mean that the scores between the pairs of opposites are always evened up instantaneously, though it is clear that the plan involves the preservation, in general, of certain 'measures', those mentioned in fr. 30. In one fragment an explicit assurance is given that the divine justice will preserve the cosmic order:

Sun will not overstep his measures; for the Erinyes, the assistants of Justice, will find him out (fr. 94).

where the Erinyes are strife (eris) personified. And the inextricable combination of 'strife' between two opposites and 'justice' as a set of rules is illuminated in one further striking image:

Time (aiōi) is a child at play, playing draughts; a child's is the kingdom (fr. 52).

Here, as probably in Anaximander, 'Time' is a name for God, with an etymological suggestion of his eternity. The infinitely old divinity is a child playing a board game as he moves the cosmic pieces in combat according to rule.

On the analogy of the bow and the lyre, it is no surprise that the order and unity of the kosmos should depend upon the existence of opposed forces. But, as was shown in discussing fr. 51, it is possible to be uncertain whether Heraclitus was there thinking of a tension between opposites, or an oscillation between them, or both. It was argued in favour of oscillation that the palintropos harmoniē is best interpreted as something permanent which is expressed in the way bow and lyre function by alternations in time—the arms of the bow are drawn back and together, but then spring forwards and apart, and the high and low notes of the lyre alternate when it is played. The same alternatives appear in the interpretation of Heraclitus' physical system. With a 'tension' interpretation, the struggle between the opposites will always be evenly balanced, gains in one region by one force being always simultaneously offset by equal gains elsewhere by the opposed force. With an 'oscillation' interpretation, the struggle may go everywhere in favour of either opposite, but alternately, the alternation being subject to a law determining the periods during which each prevails.

It should be clear that the similes of bow and lyre already favour the 'oscillation' interpretation for the cosmology, since when bow and lyre function there is oscillation in time. Nor does it count against this that Plato in a well-known passage (Sophist 242D-243A) contrasts Heraclitean tension with Empedoclean oscillation. Plato's concern here is ontology, not cosmology; he is concerned with ultimate reality, not with appearances. Most of the other evidence points in the same direction. Thus, fr. 67, already cited, mentions God as the unity behind at least two pairs of opposites which are expressed by oscillations, namely, day-night and winter-summer, and (if the report of Theophrastus is reliable) war-peace and famine-surfeit were also names for longer periods of cosmic oscillation (DK 22 A 1, B 65). And the best of our later witnesses, Aristotle (esp. de Caelo 279^b 14-17 and 280^a 11-19) and Theophrastus, agree in ascribing to Heraclitus an 'oscillation' theory of cosmic processes. Against all this, there is no single piece of clearly good evidence for a 'tension' cosmology. Two fragments about paradoxical states of affairs—the 'river' (fr. 12) and the 'barley-drink' (fr. 125) would qualify, if there was anything to show that they were meant as similes with direct reference to cosmology; these will be treated separately later.

In fr. 30 the kosmos is described as 'fire ever-living, being kindled by measures and being quenched by measures'. The relation between fire and other constituents of the kosmos is further treated:

For fire all things are exchanged, and fire for all things, as for gold goods and for goods gold (fr. 90);

The turnings (tropai) of fire: first sea, and of sea half earth, half prēstēr … (the earth?) is dissolved to become sea, and is measured in the same proportion as was before (fr. 31).

All three fragments have this in common, that they stress the lawlikeness of the cosmic processes. When fire is 'kindled' or 'quenched', which is when it is 'received' or 'given' in exchange for other things, measures are observed which are like the standard rates of exchange relating gold to goods of all kinds. Gold, being universally acceptable, provides a medium of exchange and a measure of value: so too fire in the kosmos. The details of fr. 31 are obscure, but here too a 'proportion' is preserved; presumably, as much sea is derived back from the earth as was absorbed in creating earth previously. It is, then, at least clear that in Heraclitean cosmology the components turn into one another according to certain rules which keep certain quantities constant.

The mechanism of change is slightly illuminated by evidence from Aristotle:

Some say that everything is in a state of becoming and flux, and that nothing has any firm existence, with the sole exception of one persisting thing beneath the changes, from which, by rearrangement, everything naturally comes to be; this seems to be the meaning of Heraclitus of Ephesus among many others (de Caelo 298^b 29-33);

Some [Heraclitus must be meant here] say nothing about the shape [of fire] but simply make it the thing of which the parts are finest, and then they say that other things are produced by the putting together of fire, as when gold dust is solidified in the furnace (de Caelo 304^a 18-21).

The simile of the furnace is a reminder that metallurgy and cooking were the main sources of empirical data for any Greek interested in the transformations of material things. In both, fire is the main agent of change, and the proportions of ingredients are important for the nature of the final product.

Any further reconstruction of the system of physical changes in Heraclitus is necessarily conjecture. A natural starting-point is the report by Theophrastus, preserved by Diogenes Laertius (IX 8-11, in DK 22 A 1) according to which the four main constituents, fire, air, water, earth, were related by a system of changes as follows:

Fire ⇌ Air ⇌ Water ⇌ Earth.

Yet even in Theophrastus' report there are signs that this cannot have been the whole story. The reconstruction that will now be offered is based only upon plausibilities, but it seems to fit well with most of the evidence.

Two important pairs of opposites are singled out in:

Cold things grow warm, warm grows cold, wet grows dry, and parched grows moist (fr. 126).

From this it is reasonable to infer that the interchanges cold-hot, hot-cold, wet-dry, dry-wet, were of particular importance for Heraclitus. A coherent scheme can be constructed on the assumption that all change is to be accounted for in terms of these four changes. In order to construct this scheme, it is necessary to identify earth, water, air, and some fourth thing with the combinations cold + dry, cold + wet, hot + wet, and hot + dry respectively. This identification is part of the Aristotelian theory of the physical world, but there is no reason why it should not have been invented previously; indeed, it would suggest itself immediately the two pairs of opposites had been taken to be important. For Aristotle, the combination hot + dry is identified with fire; but for Heraclitus it must not be assumed that 'fire' in this sense is equivalent to the divine fire. There will then be eight possible changes between the four components, instead of six as in Theophrastus' report….

From fr. 31 it appears that prēstēr must be a name for the hot and wet component named 'air' in the diagram. This agrees well with the possible etymologies of the word, which would connect it for Heraclitus with two verbs meaning 'burn' and 'blow'; its normal usage in Greek is to denote a waterspout or hurricane attended by lightning. All this makes it an excellent word to use for a hot atmospheric component of the kosmos.

If this is right, fr. 31 mentions the changes water-earth, water-air, and earth-water. Further, Theophrastus reports that Heraclitus spoke of two 'exhalations', one, which was dark and moist, from the sea, and one, bright and dry, from the earth. Again, the doctrine of two exhalations appears in Aristotle's own physical system, but again there is no reason why it should not have been inspired by Heraclitus, and Heraclitus' strange theory of the sun, also reported by Theophrastus, depends essentially on the two exhalations. These will then correspond to the changes water-air and earth-fire. All the other changes required by this reconstruc tion are attested by Theophrastus except for fire-earth. Many of them, it is clear, can be connected with facts of common observation.

Whether or not this reconstruction of the system of physical changes is correct, it remains to ask what patterns these changes make in time, locally and in the kosmos as a whole. It has already been argued that the kosmos as a whole oscillated between two extremes, and the evidence for this view has been briefly described. From some rather late authorities (in DK 22 A 13) it seems likely that the oscillation had a period of 10,800 or 12,000 years. At one extreme, the whole kosmos will be 'fire', and at the other it will be water. It appears that Heraclitus used the terms 'war' and 'famine' for the oscillation from 'fire' to water and 'peace' and 'satiety' for that from water to 'fire'.

Our sources of information make no distinction between the hot, dry component that has been labelled 'fire', and the cosmic fire which is associated with God, underlies all changes, and is more like a process or agent of change than a component. In the absence of explicit statements by Heraclitus, it must be assumed that the relation between these two was left unclear, unless indeed the very existence of a hot, dry component is denied. This difficulty is not an isolated one, but goes to the heart of Heraclitus' system and there links up with difficulties derived from other directions. It is worth noticing, in any case, the conceptual ambiguity of 'fire', which in Greek as in English, may easily be thought of both as a component of the world and as an agent or process of change.

Water, too, is closely associated with change in the famous words so often attributed to Heraclitus: 'all is in flux' (panta rhei). The origin of this attribution is an undoubtedly genuine statement by Heraclitus about the paradoxical properties of rivers:

Upon those who step into the same rivers, there flow different waters in different cases (fr. 12).

The wording leaves it open whether the 'different cases' are produced by stepping in at different times or at different places or both. The remark is subtler when it refers to different times, for then it points to the fact not merely that a river contains many different pieces of water, but that the very being of a river depends upon its changing its waters constantly through time. If it did not flow, it would be no river. In other words, it has to change constantly in order to stay the same river. The same idea is conveyed in another image:

The barley-drink (kukeōn) comes apart if not stirred (fr. 125)

for the barley-drink is made by stirring up wine, honey and barley-grains together, and these components will settle into separate layers unless the drink is kept in motion. The kukeōn has to move in order to stay as it is.

Both the river and the barley-drink are subtle examples of the palintropos harmoniē. But there is nothing in them to suggest that either of them had any direct application to the system of physical changes. Yet Plato and Aristotle believed that this application was made by Heraclitus, and their testimony (Plato, Cratylus 402A; Aristotle Metaphysics 1010^a 10-15) cannot simply be dismissed. It can, however, be shown to be less convincing by the observation that both Plato and Aristotle know the river fragment in a version which can hardly be Heraclitean, and which is closely associated with the late fifth-century self-styled 'follower' of Heraclitus, the philosopher Cratylus. The suspect version states: 'You could not step twice into the same river.' To deny in this way the existence of a unity persisting through change is not Heraclitean, but it is very characteristic of Cratylus. Aristotle relates of Cratylus the anecdote that he 'rebuked Heraclitus for saying that you could not step twice into the same river; he (Cratylus) thought you could not even do so once.' Whether or not this anecdote is true, it is known that Cratylus was concerned to deny that any kind of persisting object was to be found in the perceptible world, and that he called himself a 'Heraclitean'. It seems therefore almost certain that the second version of the 'river', and the 'flux doctrine', are 'interpretations' of Heraclitus due to Cratylus. Heraclitus no doubt held that change was incessantly occurring, even in the most stable objects, as indeed fr. 126 suggests; but this is not the same as the 'flux doctrine' according to which the components of every object change completely from moment to moment.

There is little information about any further details in the cosmology of Heraclitus. In particular, nothing is known of any explanation of the movements of the heavenly bodies, except for those of the sun. Of the sun the report of Theophrastus tells that it originates, 'new every day', as a bowl-shaped object in which some hot exhalation from the earth is trapped. As a result the bowl and its contents begin to ascend, and the heat from the bowl acts upon the earth and draws up to itself further fiery exhalation, so that the sun rises steadily and becomes progressively hotter. By degrees the effect upon the sea begins to outweigh that upon the earth, so that the exhalations rising are more moist than dry, with the result that the fire receives less and less nourishment and from noon declines to its final extinction. This process repeats itself every day, a different bowl being needed for each day. Eclipses, and the seasonal variations of the sun's movement, were explained using this model. Of the origin and location of the bowls nothing is reported. The stars, too, are bowls full of ignited substances.

This strange theory allows the inference that all fire had a tendency to move upwards, or at least outwards from the centre of the kosmos, and hence that earth had a tendency to move inwards. This further suggests that there is a region beyond that of the stars where fire collects if not hindered. But all further treatment of the general arrangement of the kosmos is hindered by an almost complete lack of evidence, and it is probable that Heraclitus was not at all explicit.

If the cosmology of Heraclitus was impressionistic, his doctrines concerning sleep and waking, life and death, and the nature and fate of the souls of men seem to be deliberately mysterious. It has been shown that Anaximenes probably, and possibly the other Milesians, made an analogy between the role of God in the world and that of the soul in the body, with the suggestion that individual souls are detached pieces of the divine stuff, and may perhaps rejoin it after death. The same analogy seems to be implicit in Heraclitus. At least it is clear that the soul is responsible for moving and controlling the body and for intelligence, and that it is best when dry:

When a man is drunk, he is led by a young child and stumbles as he goes, unaware of where he is going; for his soul is moist (fr. 117);

Dry soul is wisest and best (fr. 118).

Other pieces of evidence confirm that the soul in its best state is fiery. Aristotle reports (de Anima 405^a 25-6) that Heraclitus says that the soul is 'the exhalation from which he constructs everything else', a mysterious phrase but one which is consistent with the soul being a kind of fire. Another obscurely worded fragment, fr. 26, seems to suggest, like some early Hindu texts, that sleep is a retreat for the soul's fire into an inner citadel, the outer 'fires' accounting for sense-perception being quenched. Similarly, fragments (frr. 24, 25, 63) apparently relating to the fate of the soul after death can most easily be accounted for by taking that fate to have been decided by the condition in which the soul was at death. A good and wise soul, being fiery, would at departure ascend to the upper regions, whether it was absorbed into the divine fire or, as some evidence suggests, became a star (DK 22 A 15). A less good soul would find difficulty in ascending, because an admixture of cold and moisture would drag it down; the strange myth in Plutarch's dialogue 'On the Face which is Seen in the Orb of the Moon', according to which most souls undergo a kind of purgatory in the dark region between the earth and the moon, may possibly preserve some of Heraclitus' ideas.

Two further fragments relating to souls look important but are difficult to interpret:

The limits of soul you would not find out by going about, though you travelled every road; so deep a logos does it have (fr. 45); To the soul belongs a logos that increases itself (fr. 115).

'The logos of soul' may be expected to be the true account of its nature. A 'deep logos' is one that is profound and subtle—the metaphor is common in early Greek. That the soul has a deep logos will then explain why its limits (peirata) are not to be found, which is to say that it cannot be characterised or defined. No investigation will produce a complete account of soul, which therefore must be of unlimited complexity. The same conclusion is suggested by fr. 115; if the logos of soul 'increases itself it presumably grows greater without limit. These fragments seem then to have a mutual coherence, even if the self-increasing complexity of soul is not explained.

A further explanation may be offered which would take these fragments to reflect Heraclitus' awareness of a central difficulty in his system of ideas. First, it will be seen that the doctrine about souls offers some justification for the hypothesis suggested earlier that men can make progress in wisdom just in so far as they make themselves divine; they can do this because their souls are potentially divine and can realise their potentiality. This completes a cycle of associations, wisdom—God—fire—soul—wisdom, which must have been intended by Heraclitus. Now in connection with each member of the cycle an analogous problem presents itself; each has been noticed separately in the course of this chapter. Each member of the cycle is distinct from, and transcends in some way, the ordinary world in which it nevertheless seems to be entangled. 'Wisdom', or 'the wise', is 'distinct from everything'; yet some men are wise, and the wise God is everything else too. God is the kosmos, but is also beyond or beneath the kosmos. Fire is one of the physical components of the kosmos, but is also the agent or process or mind that controls the changes of such components. Soul is subject to physical transformation, yet is also the controller and agent of such transformation. In view of the cycle of associations, Heraclitus must have recognised all these problems as one problem, if he recognised them at all.

The evidence that Heraclitus did recognise this problem, apart from general considerations about his intelligence and intellectual honesty, consists only of fr. 115 and fr. 10. Both of these suggest, without being forced unduly, that Heraclitus was aware of an infinite regress made necessary by his thought. Fr. 10 shows that the most general description of the 'unity-in-opposites' is at the same time an example of that same kind of thing; so that a logos containing fr. 10 will be referring to itself among other things. In the same way, a logos in the mind of God or man which gives an account of that mind will have to refer to itself. Once this is admitted, we have the familiar 'map' paradox: a map of an area which includes the map itself will have, if true in all details, to contain an infinite procession of maps: a map of itself, a map of the map of itself, and so ad infinitum.

Now, the only plausible way to detach God from the kosmos, for Heraclitus, is to think of God's mind as a perfect map of the kosmos, which is itself in the kosmos but not identical with it. In this way the central problem of Heraclitus' thought can be seen to issue in the 'self-mapping' paradox suggested by frr. 10 and 115. It may still, reasonably, be objected that this is a construction without sufficient basis in the texts. But it can hardly be denied that Heraclitus takes his God to have in his mind something like a 'plan' for the running of the universe, which is perfectly carried out; which is the equivalent of a perfect map of the Universe. This way of seeing the central difficulties of Heraclitus can also help to explain the similarity in tone and approach between Heraclitus and the early Wittgenstein. Alphabetic writing, which pictured speech by arranging characters in an order, and the development at the same time of explicitly formulated general laws, must have helped the growth of the idea that all reality could be represented in language, and conversely that all language not representing reality was false or nonsensical. As the early Wittgenstein, inspired by the new 'language' of formal logic, tried to mark out the limits of significant language-use as that which depicts the world, and thereby to exhibit some truths about the structure of reality reflected in the true structure of language, and to demolish as meaningless all metaphysics, so Heraclitus seems to be using his new consciousness of sentences as formulae for exhibiting reality, suggested by his use of the term logos, to exhibit the structure of things in appropriately constructed language. In part of fr. 1 he describes himself as 'delimiting each thing according to its nature and declaring (phrazōn) how it is'. The devices he uses are not those of formal logic, but etymologies, puns, antitheses and portmanteau-words. He attacks the foolishness of ordinary men, as Wittgenstein attacks metaphysics. Both philosophers, being dominated by an ideal of how language should be used, are systematic in thought but discrete in expression. Both try to expel mystery and end with a central difficulty that leaves things more mysterious than before.