To further the understanding of the Presocratics, as intelligent beings looking for a theoretical ordering of the world, has been, among the cherished aims of Gwil Owen, one to which he has repeatedly recurred and to which we owe some of his most characteristically illuminating work. To this collection in his honour I offer a set of proposals for the understanding of Heraclitus.

Epistemology: the programme

The hypothesis to be explored claims that at the heart of Heraclitus' thought there lies a remarkable and characteristic epistemology, and that it is this above all that must first be grasped if his account of the world is to be understood. It will help to begin with a statement of what would be agreed about Heraclitus' epistemology by many scholars.

I shall treat as non-controversial the position summarised in the rest of the present paragraph. Heraclitus is deeply interested in the problem of knowledge. He sharply rejects the claims to be guides to knowledge of (a) ordinary common sense; (b) popular and traditional beliefs; (c) much of traditional Greek religion; (d) the older accepted authorities, Homer and Hesiod; (e) more recent claimants of such diverse kinds as Archilochus, Xenophanes, Hecataeus and Pythagoras. Against all these, and in support of his own account of the world, Heraclitus appeals in the first place to the evidence of the senses. ‘All of which the learning is seeing and hearing, to that I give preference’ (B55).

In this chapter I suggest that some passages in the Cratylus may give us insight into certain of the difficulties in the notorious ‘dream’ passage in the Theaetetus. This is a modest point, but it has, I hope, less modest implications for our understanding of what Plato is after when he looks for a definition of knowledge. Whatever the merits of this paper, it has a certain appropriateness as my contribution to a volume honouring Gwil Owen; my introduction to working with him, when I was a graduate student at Harvard, was a class on the Theaetetus and a paper I wrote on the Cratylus. Nothing here survives, I am glad to say, of the content of my early efforts; what I do hope survive are some effects of those lessons in rigour, in patience with the text and in Platonic readiness to lay aside one's own concerns in the attempt to see Plato's as they are. This paper is offered to Gwil as a token, however inadequate, of truly Socratic teaching and friendship.

What is the dream theory doing in the Theaetetus? At 201C8, Theaetetus, having been convinced that knowledge cannot just be true belief, suggests that it may be true belief plus logos or account. This is not an unexpected suggestion to those familiar with Plato's earlier attempts to show that knowledge is the kind of improvement over true belief that results from being able to offer certain kinds of reason and grounds for what is believed.

Heraclitus and the Milesians

In recent decades there has been a tendency among scholars to question whether Heraclitus was, in the same sense as the Milesians were, a cosmologist: ‘[Heraclitus'] real subject is not the physical world but the human condition, which for the Greeks means the condition of mortality … Like [his] substitution of Fire for [Anaximenes'] Air, any changes in detail must have been designed not to improve the physical scheme in a scientific sense but to render its symbolic function more drastic.’

It would be foolish to deny that problems about mortality, fallibility and the human perspective were an important part of Heraclitus' main subject. But this is not inconsistent with his having seen himself as answerable in the first instance to the same questions as the Milesians, whatever his reservations about their would-be polymathiē:

One thing is wisdom: to understand the plan by which all things are steered through all things (B41).

Since the end of the Second World War, the academic study of Greek philosophy in Britain and North America has changed almost out of recognition. One fairly crude index of the change is the huge growth in the numbers of scholars active in the subject, the volume of their publications, and the variety of their interests. Harder to characterise is the way in which scholarly study of the ancient texts has, without losing in historical scrupulousness or historical imagination, become much more a first-order philosophical activity than it was in the first half of the century. Such changes require moving causes. This volume salutes the work of a scholar and philosopher whose influence on the development of study of Greek philosophy in the last 30 years is second to none.

There have been three major channels through which G. E. L. Owen has made his influence felt. Pride of place must go to the series of masterly essays which he has given us since the early fifties, transforming the state of the art. In Parmenides and Zeno he has shown us, in place of the dogmatic monist and the sophist of the textbooks, the inventors of philosophy as we now understand it, or more specifically of a tradition of profound and subtle metaphysical argument. He has demonstrated how Plato in the theory of Forms gave classic expression to a seductively simple picture of the relations between language and the world, but then in his later dialogues fought his way self-critically to a more penetrating understanding of their complexities.

The ancients were already well aware that the sorites is not just about heaps. It brings into question the very existence of the gods, or at least the rationality of religious belief. If modern philosophers of language (who in recent years have been much preoccupied with the sorites paradox) seldom know this, a magisterial paper by Jonathan Barnes has now made clear the range and richness of the ancient material on the subject. Some of this material looks strange by modern lights. All the more reason, therefore, why enthusiasts for the sorites should take an interest in its history.

The history begins, as everybody knows, with a memorable example fashioned by the past master of paradox, Eubulides of Miletus (4th cent. b.c.), known also for his purveying of the Liar, the Bald Man, the Nobody, and other logical delights (D.L. 11 108).

Aristotle opens Physics I by stating that an inquiry into nature (peri phuseōs), like other inquiries, should begin with an account of the relevant principles (archai). He does not tell us what he means by ‘nature’ – for that we have to wait until book II – and he does not tell us what he means by a ‘principle’ in this context, but as we read on we may come to think this omission unimportant. For straightway at the beginning of chapter 2 he appears to place himself in the tradition of a series of writers on nature (peri phuseōs) whose views on the ‘principles’ (archai) were perfectly well known. Thus Thales held that there was one ‘principle’, namely water, while Anaximenes selected air and Heraclitus fire; Empedocles again held that there were four principles (earth, water, air, fire), Anaxagoras that there were infinitely many, Leucippus and Democritus that there were just atoms and void, and so on. So Aristotle, it would seem, is preparing to offer us his answer to the question to which these answers had already been propounded by his predecessors: he is preparing to list the ultimate ingredients of the world, and to given an account of how the world is made up from those ingredients. Perhaps this characterization of what the older physicists were up to is rather oversimplified, but I think it is not worth elaborating their problem now.

In his published writings Gwil Owen has from time to time given us some characteristically stimulating comments on the Cratylus. Although he has not devoted an entire essay to the dialogue, it has often been among his favourite topics of discussion. I recall in particular a seminar which he conducted on it soon after his arrival in Cambridge in 1973. I hope he may enjoy this further contribution to the conversation.

The progress of the main argument of the Cratylus is clear enough, although not always well understood. Socrates in conversation with Hermogenes propounds a theory according to which a name (i.e. a noun, a verb, or an adjective – but not a word of primarily syntactic function) counts as a genuine or correct name if and only if its nominatum is a real thing and its constitution is naturally suited to its nominatum. The principal test of a name's suitability is its capacity to teach or disclose to us the nature or essence of its nominatum: simply by considering the constitution of a name we should be able to tell what it is the name of (386D–391A).

Socrates then attempts to explain just how a name could be so constituted as to disclose its nominatum. He does so by means of the hypothesis that his own language, Greek, consists of genuine names in the sense required by the theory.

The account he offers of the way in which Greek names can be construed as genuine names is developed in two stages.

In chapter 11 of his de Interpretatione Aristotle tries to explain why certain features of a thing go together to make up a unity, whereas others do not. ‘For example,’ he says, ‘a man is perhaps an animal and two-footed and tame, and from these there does come to be some one thing’ (2ob16–8). But ‘from white and man and walking,’ he adds, ‘there is not one thing’ (2ob18–19).

The reasoning in this chapter is intricate; but without tracing out the intricacies we can perhaps say that the unity Aristotle is seeking here is the oneness of an individual substance. In this passage he is unwilling to count as a unity the parasitic oneness that is enjoyed by features only accidentally compresent in a single substance. Though it may be true to say, he remarks later on in the chapter, ‘The white is musical’ (that is, presumably, ‘The white person, is musical’), still, he warns, musical white and white musical (contrast: two-footed animal) are not one thing (21a7–14).

Aristotle later softens his position. He allows, in Metaphysics v 6, for example, that there is such a thing as an accidental unity (hen kata sumbebēkos). The musical and the just, he says, make up an accidental unity because musicality and justice are accidents of one substance (1015b21–2).

These studies are offered to Gwil Owen on the occasion of his 60th birthday with respect, gratitude and affection. Their writers are all either pupils of his or younger scholars who, while not formally his pupils, would wish to acknowledge the stimulus of his talk and thought at a formative stage in their own philosophical histories. The volume contains fifteen chapters, all concerned in one way or another with aspects of the role played by reflection upon language in the thought of Plato, Aristotle, and other ancient Greek philosophers. It would have been as easy and as appropriate to persuade a quite different team of authors to write essays in Gwil Owen's honour on some quite different subject – say, Greek science and philosophy of science – no less close to his heart. So this book is dedicated to him with the good wishes, expressed by many to the editors, of a much greater number of pupils and others, on both sides of the Atlantic, than are assembled between these covers. Our thanks go to all who have helped us by their co-operation or advice, particularly our publisher Jeremy Mynott, who has made the project possible.

The second half of the Parmenides has been a source of puzzlement to generations of scholars, inspiring a wide variety of interpretations. Thanks to the groundbreaking work of Ryle and Owen, one can see this material today as offering serious reflections of conceptual and metaphysical nature. In this chapter I wish to locate the key conceptual problems that Plato addresses in this passage, and argue that Plato is here also defending and revising his theory of Forms.

General remarks

All attempts at interpretation have to come to grips with the strange structure of the passage. On the surface at least, it seems that the material is arranged into eight arguments; the arguments taken pair-wise contradict each other. This organisation is, however, not very tight. Thus it suggests that it might serve for Plato more as a frame of exposition rather than as the logical back-bone of content. For one thing, after the second argument we find a longer passage on time which begins by stating that we are starting for ‘the third time’ (155E4); commentators, taking the surface structure too seriously, have been treating this as an ‘appendix’ to the second argument. Further evidence for the hypothesis that the over-all structure is not meant very deeply is provided by the fact that the arguments are not of equal length. The second one is the longest, and the last four take up much less space than the first four.

Aristotle believed that many (not, of course, all) natural events and facts need to be explained by reference to natural goals. He understands by a goal (hou heneka) whether natural or not, something good (from some point of view) that something else causes or makes possible, where this other thing exists or happens (at least in part) because of that good. So in holding that some natural events and facts have to be explained by reference to natural goals, he is holding that some things exist or happen in the course of nature because of some good that they do or make possible. Thus he holds that living things have many of the organic and other parts that they have because of the good it does them, so that these parts exist, and are formed, for the sake of the animal or plant itself whose good they subserve. To explain why they have them, and why as they are being formed they come to have them, one must refer to the whole animal or plant who needs them as the goal for which they exist. Aristotle gives or suggests, at one place or another, several arguments in favour of this thesis. Some of these press the analogy between artistic activity, which is admittedly goal-directed, and natural processes, thus extending explanation by appeal to goals from human action to non-human, even non-animal nature.

G. E. L. Owen's papers tend not only to transform their professed subject matter, but also to spread illumination into quite distant regions. The present distant region is no exception, as the footnotes will reveal.

Plotinus distinguished at least two kinds of thinking. There is dianoia, which is often called discursive thinking, and which is the activity of the soul (psuchē). And then there is the different activity of the intellect (nous), which is often called non-discursive thinking. It is commonly held that non-discursive thinking does not involve entertaining propositions. That is, it does not involve thinking that something is the case. Instead, one contemplates concepts in isolation from each other, and does not string them together in the way they are strung together in ‘that’-clauses. It is further supposed that Plato and Aristotle anticipated Plotinus in postulating this non-propositional thinking.

I have three aims in this chapter, The main one is to deny that non-propositional thinking is to be found in any of these three thinkers at the points where it has most commonly been detected. In order to show this for the case of Plotinus, I shall have to explain some of Aristotle's ideas about thinking and how Plotinus transformed them. He certainly did believe that there is a mystical state in which we have contact with something much simpler than any proposition. But I shall maintain that he regards this mystical experience as above the level of thinking, while thinking in its highest form he treats as propositional.

For me, G. E. L. Owen's ‘Plato on Not-Being’ radically improved the prospects for a confident overall view of its topic. Hitherto, passage after passage had generated reasonable disagreement over Plato's intentions, and the disputes were not subject to control by a satisfying picture of his large-scale strategy; so that the general impression, as one read the Sophist, was one of diffuseness and unclarity of purpose. By focusing discussion on the distinction between otherness and contrariety (257B1–C4), Owen showed how, at a stroke, a mass of confusing exegetical alternatives could be swept away, and the dialogue's treatment of not-being revealed as a sustained and tightly organised assault on a single error. In what follows, I take Owen's focusing of the issue for granted, and I accept many of his detailed conclusions. Where I diverge from Owen – in particular over the nature of the difficulty about falsehood that Plato tackles in the Sophist (§§5 and 6 below) –it is mainly to press further in the direction he indicated, in the interest of a conviction that the focus can and should be made even sharper.

2. By 256E5–6 the Eleatic Stranger (ES) can say ‘In the case of each of the forms, then, what is is multiple and what is not is indefinite in number.’ Yet it is only at 258B6–7 that Theaetetus is allowed to announce the availability, at last, of the application for ‘what is not’ that was needed in order to flush the sophist from his refuge.

Meaning and signification

Professor Owen has taught us to attend to Aristotle's, no less than to Plato's, views on the relation between language and reality. Some of his work is the distinguished result of his attention to these views. In ‘Logic and Metaphysics’ and ‘Aristotle on the Snares of Ontology’ he has argued that Aristotle's views on the proper way to inquire into Being are influenced by his views on the irreducibly different senses of ‘being’. In ‘Tithenai ta Phainomena’ he has shown how Aristotle's questions ‘What is time?’, ‘What is place?’, ‘Does anyone ever act incontinently?’ and so on are approached by methods and arguments different from those of empirical science: ‘By such arguments the Physics ranks itself not with physics in our sense of the word, but with philosophy. Its data are for the most part the materials not of natural history but of dialectic, and its problems are accordingly not questions of empirical fact but conceptual puzzles.’

This account of Aristotle's results reflects a clear and influential picture of his questions and his aims. Aristotle asks the Socratic ‘What is it?’ question to find out what a word means, to give an analysis of the concept associated with the word; he wants to set out the meaning that competent speakers implicitly grasp but cannot always state in clear, paradox-free terms. Sometimes he denies that there is just one answer to the Socratic question, because the word is homonymous; and then he means that it has more than one sense.