Akademik

Anaxagoras and the atomists
Anaxagoras and the atomists C.C.W.Taylor ANAXAGORAS In the course of the fifth century BC the political and cultural pre-eminence of Athens attracted to the city a considerable number of intellectuals of various kinds from all over the Greek world. This phenomenon, the so-called ‘Sophistic Movement’, is fully described in the next chapter; here it suffices to point out that, in addition to the discussions of moral and theological questions for which the sophists are more widely known, the activities of many of them included popularization and extension to new areas, such as the study of the origins of civilization, of the Ionian tradition of general speculative enquiry into the natural world (see Chapter 2). Anaxagoras stands out from his sophistic contemporaries as a truly original thinker, who sought not merely to transmit the Ionian tradition, but to transform it radically in a number of ways, and in so doing to enable it to meet the challenge of Eleatic logic, which had threatened the coherence of the cosmological enterprise. An Ionian from Clazomenae on the central coast of Asia Minor, Anaxagoras was a contemporary of Protagoras and Empedocles. Aristotle says (Metaphysics 984a11–12: DK 59 A 43) that he was older than the latter, and (probably) that his writings are later than those of Empedocles (the interpretation of the crucial sentence is disputed). It is reliably attested that he spent thirty years in Athens and that he was closely associated with Pericles, though there is some dispute among scholars on when the thirty years began and ended, and whether they were a single continuous period or discontinuous. Socrates in the Phaedo (97b– 98c: DK 59 A 47) describes reading Anaxagoras’ book as (probably) quite a young man, but implies that he was not personally acquainted with him; some have taken this as evidence that Anaxagoras had already left Athens for good by about the middle of the century, but the evidence is weak. It is clear that, in common with other intellectuals, his rationalistic views on matters touching on religion (in his case, his materialistic accounts of the nature of the sun and other heavenly bodies) made him unpopular in certain circles, and there is a tradition (questioned by Dover [6.6]) that he had to flee from Athens (with the assistance of Pericles) to escape prosecution. He is said to have died at the age of 72, probably in the early 420s. He appears to have written, as did Anaximander, a single comprehensive prose treatise, referred to by later writers, such as Simplicius, by the traditional title On Nature. In the Apology (26d, DK 59 A 35) Socrates states that it was on sale for a drachma, about half a day’s wage for a skilled craftsman, which indicates that it could be copied in well under a day. The surviving quotations from it (almost all preserved by Simplicius), totalling about 1,000 words, therefore probably represent quite a substantial proportion of it. In what follows I shall be concerned with two central topics of this work, the nature of the physical world and the nature and cosmic role of mind. The Physical World For all post-Parmenidean thinkers the central challenge was to show how natural objects, including the world order itself, could come to be, change and cease to be without violating the Eleatic axiom that what is not cannot be. Parmenides had argued that that axiom excluded coming to be (for what comes to be comes from what is not), change (for what changes changes into what it is not) and ceasing to be (for what has ceased to be is not). Anaxagoras’ contemporary Empedocles met this challenge by redescribing change (including coming and ceasing to be) as reorganization of the four elements, earth, air, fire and water. Those elements satisfy the Parmenidean requirement in its full rigour, since they are eternal and changeless. What we observe and call change, coming to be and destruction is in reality nothing but reorganization of these elemental components; hence neither organic substances, such as animals and plants, nor their components, bones, hair, blood, leaf tissue, etc., strictly speaking ever come into being or cease to be. Put anachronistically, coming into being reduces to elemental rearrangement, and what is reduced is thereby eliminated from a strict or scientific account of the world. Anaxagoras agreed with Empedocles that what is conventionally regarded as coming to be and destruction is in fact reorganization of basic items. He asserts this fundamental thesis in fragment 17: The Greeks are not correct in their opinions about coming to be and destruction; for nothing comes to be or is destroyed, but they are mixed together and separated out from things which are in being. And so they would be correct to call coming to be mixing together and destruction separation. The language is strikingly reminiscent of Empedocles’ fragment 9: Now when they [i.e. the elements] are mixed and come to light in a man or a wild animal or a plant or a bird, then they say it has come to be, and when they separate, then they call that dismal destruction; they do not call it as they ought, but I too assent to their usage. But there is a crucial difference, in that Anaxagoras rejected Empedocles’ core belief in the primacy of the four elements. Even if we accept (as I shall assume) that Anaxagoras’ book was written later than Empedocles’ poem on nature, it must be a matter for conjecture how far Anaxagoras arrived at his view of what was physically basic through conscious opposition to the views of Empedocles. What is, however, indisputable, is that Anaxagoras’ view of the physically basic constituted a radical departure from that of Empedocles (and a fortiori from that of his Ionian predecessors); that divergence, moreover, marked a fundamental innovation in the conception of physical reality and of the relation between reality and appearance. For Anaxagoras’ account of what is physically basic we may begin with fragments 1 and 4. Fragment 1, according to Simplicius the opening sentence of Anaxagoras’ book, describes the original state of the universe, in which everything that there is was so mixed up together that nothing was distinguishable from anything else. What these things were fragment 4 tells us; they were ‘the wet and the dry and the hot and the cold and the bright and the dark and a lot of earth in with them and an infinite number of seeds, all unlike one another’. In this list we see: first a list of the traditional opposite qualities, as in Anaximander for example; second, earth, one of Empedocles’ four elements; and third, an infinite number of seeds. ‘Seeds’ is a biological term, denoting roughly what we would call the genetic constituents of organisms; the seed of a kind of plant or animal is what develops into a new instance of that plant or animal type, and, as Vlastos [6.19] points out, the process was ordinarily conceived as one in which the seed, seen as ‘a compound of all the essential constituents of the parent body from which it comes and of the new organism into which it will grow’ (p. 464), develops by assimilating more of the same kinds of constituent supplied by the environment. That these constituents were identified by Anaxagoras with the organic stuffs, flesh, blood, fibre, etc., which compose organisms of different kinds, is suggested by fragment 10: ‘How could hair come to be from what is not hair, and flesh from what is not flesh?’ For the naked embryo to develop into the hirsute adult, the seed must have contained hair, the presumably minute quantity of which was supplemented by the amounts of hair contained in the nourishment which the growing animal assimilated. In Anaxagoras’ primeval mixture, then, we find qualities, namely the opposites, and stuffs mingled together without any categorial distinction. The stuffs include the four Empedoclean elements; earth is mentioned in fragment 4, and air and aithēr, the bright upper atmosphere, (traditionally conceived as a form of fire) in fragment 1, while the principle of fragment 10 (‘F cannot come to be from what is not F’) implies that water is a constituent in the mixture too. But the elements have no special status relative to other stuffs; earth is no more primitive than bone or flesh (contrast Empedocles frs 96 and 98). In fact the central and most novel feature of Anaxagoras’ world-picture is that it contains no elemental stuffs. Relative to substances such as trees or fish, and to their parts, such as leaves and fins, all stuffs are elemental, since substances come to be through rearrangment of stuffs. But relative to other stuffs, no stuffs are elemental, since every stuff is a component of every stuff; ‘so everything is in everything, nor is it possible for them to be apart, but everything has a share in everything’ (fr. 6), ‘in everything there is a share of everything, except mind’ (fr. 11). Some interpreters (Cornford [6.5], Vlastos [6.19]), finding a literal reading of these statements intolerably uneconomical, have urged a restricted reference for the second occurrence of ‘everything’, interpreting ‘in everything there is a share of everything’ as ‘in every substance there is a share of every opposite’. On this view the basic items of Anaxagoras’ ontology are the opposites, stuffs such as flesh and earth being ‘reduced’ to clusters of (opposite) qualities as in Berkeley and Hume. (Schofield even describes stuffs as ‘logical constructions’ of opposites ([6.17], 133).) The texts in which these statements occur contain no hint of any such programme. They give no justification for restricting the reference of ‘everything’ more narrowly than to the ‘all things’ which were together in the original mixture, which undoubtedly include the stuffs air and aithēr (fr. 1) and on the most natural reading earth and an infinite number of seeds (fr. 4). Moreover, the idea that qualities are ontologically more basic than stuffs also lacks support from the fragments. Those, to repeat, present the picture of the original state of things as a mixture of constituents of all kinds, every one of which is equally a constituent, not only of the mixture, but of every other constituent. Further, they attest that the ‘everything in everything’ principle holds in the present world order as much as it did in the original state (fr. 6). Can sense be made of these claims on the generous interpretation of ‘everything’ which is here adopted? Before proceeding to that question we should consider another restriction on the generality of ‘everything’ proposed by Cornford [6.5]. Observing correctly that the concept of seed is a biological one and that the biological processes of nutrition and development are particularly prominent in the fragments and testimonia, Cornford restricts the ‘everything in everything’ principle to organic substances, interpreting it as ‘in every organic substance there are seeds of every organic substance’. While the indefinite variety of observable biological transformations provides grounds for accepting that every organic substance can come from every other, and must therefore (by the principle that F cannot come from what is not F) be a constituent of every other, there is no ground to extend this to non-organic substances. To use his example ([2.15], 280), since we never observe acorns turn into emeralds, there is no reason why Anaxagoras should have believed that acorns contain portions of emerald. On the other hand, Aristotle reports (Physics 203a23–4, DK 59 A 45) that Anaxagoras held that every part is a mixture in the same way as the whole (i.e. the universe) because he saw that anything comes to be from anything, and Lucretius cites the coming to be of gold, earth and other non-organic stuffs along with, and explained by the same process as, organic generation (I.830–42, DK 59 A 44). The comprehensive character of traditional Ionian explanation makes it plausible that Anaxagoras should have accepted the universal thesis. Xenophanes had already noticed the transformation of animals and plants into stone by fossilization (DK 21 A 33), and the transformations of stone into earth and earth into water by erosion, of water to wine, wine to animal tissue, etc. were matters of common observation (see Simplicius’ commentary on the passage from the Physics cited above (DK 59 A 45)). It is therefore highly plausible that Anaxagoras should have held that we can have no reason to say of any two things that they cannot be transformed into one another by some chain of causation, however long. We know that the atomists, following Parmenides, appealed to the Principle of Sufficient Reason, arguing, for example, that since there was no more reason for atoms to have one shape rather than another, and since they obviously had some shapes, therefore they must have all possible shapes (Simplicius, Physics 28.9– 10). Similarly, Anaxagoras may have reasoned that since some transformations are observed to occur, and there is no reason for one transformation to occur rather than another, all transformations must be assumed to occur. I shall take it, then (1) that Anaxagoras drew no systematic distinction between stuffs and qualities and (2) that he believed that every amount or bit of any stuff (or quality) contains quantities or bits of every other stuff (or quality). On the assumption that if a given amount of a given stuff (amount A of stuff S) contains amounts B, C, D…of stuffs X, Y, Z…then A is larger than B, and larger than C and larger than D etc.,…it immediately follows that there is no smallest quantity of any stuff. For any quantity of any stuff, however small, contains smaller quantities of every stuff, and so on ad infinitum. Anaxagoras asserted this conclusion explicitly: ‘for there is no smallest part of what is small, but always a smaller’ (fr. 3), and according to Simplicius deduced it from the premiss that everything is in everything and is separated out from everything. But if there is a portion of every stuff in every stuff, what distinguishes one stuff from another? Anaxagoras’ answer is given at the end of fragment 12; ‘Nothing is like anything else, but each single thing most clearly is and was that of which it contains most’. This dark saying is explained by Aristotle in Physics 187b1–7 (not in DK): Therefore they say that everything is mixed in everything, because they saw everything coming into being from everything. And they appeared different and were called by different names from one another on account of the quantitatively predominant component in the mixture of infinitely many components; for there is nothing which is as a whole pure white or black or sweet or flesh or bone, but the component each thing has most of, that is what the nature of the thing appears to be. That is to say, every stuff contains amounts of every stuff, but in different proportions, and in each stuff the component of which there is the largest amount gives its character and name to the whole. Thus a lump of earth contains, in addition to earth, ‘seeds’ of every other stuff and quality, but it contains more earth than any of the others (the other ‘seeds’ may be thought of as impurities in the sample of earth). So any sample of earth is not pure earth, but predominantly earth; in general, to be a sample of S is to be predominantly S. (The texts leave it indeterminate whether something predominantly S must contain more S than all other components put together, or merely more S than any other component; I shall assume that the weaker condition is sufficient.) This doctrine may seem to threaten Anaxagoras with a dilemma. Either it commits him to the existence of samples of pure S, which is inconsistent with the doctrine that everything contains a bit of everything, or it is empty. Taking the first horn, it is clear what it means to say that a sample of S is predominantly S. Analyse the original sample, by whatever physical process is available, into its components S, A, B, C… Continue the analysis until you reach pure samples of each component. Then you will discover that the amount of pure S is larger than the amount of pure A, larger than the amount of pure B, etc. But now it is false that everything contains a bit of everything; analysis will have succeeded in doing what Anaxagoras explicitly says (fr. 8) it is impossible to do, namely separate from one another the things in the cosmos ‘and chop them off with an axe’. Prima facie Anaxagoras should prefer the other horn. According to this there are no pure samples of any stuff; every sample of every stuff, however small, will contain as impurities amounts of every other stuff. But now what does it mean to say that gold contains more gold than hot, sweet, blood, vegetable fibre…? It can’t mean that it contains more pure gold than pure hot…since there are no such pure stuffs. It means that it contains more gold than hot etc., i.e. more stuff that contains more gold than hot etc., and the stuff that that stuff contains more of is the stuff that contains more gold etc., and so on for ever. That is to say, we can never give a complete specification of what it is that gold contains most of; gold just is what contains more gold than anything else, and so on for ever. In semantic terms, we have no account of what F means if all we can say is ‘“A is F” means “A is predominantly F”’. But in order to understand the name of a stuff it is not necessary that it should be possible, even in principle, to isolate pure samples of that stuff. As Kripke1 has shown, the names of stuffs are proper names whose reference is fixed by those observable properties which typically, though contingently, characterize that stuff. Thus gold is that stuff, whatever it is, which is yellow, shiny, malleable, etc. The specification of what stuff it is which has those properties is the task of the best available theory, in modern terms the theory of elements, which identifies gold as the element with atomic number 79. The only resource available to Anaxagoras to identify stuffs is via their constitution; thus gold just is that stuff which when analysed yields more samples of yellow, shiny, malleable stuff than red, warm, sticky, liquid stuff (and so on for every stuff- description). Analysis goes on for ever, in principle at least; even when the technical limit is reached of whatever process of physical separation has been employed, we know a priori that every sample of yellow, shiny, malleable stuff contains infinitely many samples of every kind of stuff, but always more of yellow, shiny, malleable stuff than of any other. An objection to the attribution of this theory to Anaxagoras is that it seems flatly to contradict Aristotle’s evidence (DK 59 A 43, 45, 46) that in Anaxagoras’ system the elements were ‘the homoeomerous things’. In Aristotelian terminology a homoeomerous substance is one whose parts are of the same nature as the whole, e.g. every part of a piece of flesh is a piece of flesh, as opposed for example to a plant, whose parts are leaves, roots etc., not plants. In general, stuffs, which we have seen to be among Anaxagoras’ basic things, are in Aristotelian terms homoeomerous. Hence Anaxagoras is committed to holding that every part of a piece of gold is a piece of gold, which contradicts the account given above, according to which a piece of gold contains, in addition to pieces of gold, portions of every other substance and quality. (This contradiction is the basis of Cornford’s interpretation of ‘everything in everything’ as ‘every opposite in every substance’.) This difficulty seems to me illusory. One possibility (adopted by McKirahan [2.7], 208, n. 38) is that in identifying Anaxagoras’ basic substances as ‘the homoeomerous things’ Aristotle means merely to identify them as stuffs, i.e. the things which in Aristotle’s theory are homoeomerous, without attributing to Anaxagoras the thesis that those stuffs are in fact homoeomerous. This may well be right. It is, however, possible that Anaxagoras may have maintained (the texts are silent) that stuffs and qualities are indeed homoeomerous, despite containing portions of every stuff and quality. He could do so consistently if by ‘homoeomerous’ he meant ‘having every part of the same kind as the whole’, and if by part he understood what is produced by division. He might then have maintained that however minutely one divided up a lump of gold, what would be produced would be fragments of gold, the other stuffs and qualities being separable, if at all, not by division, but by other processes such as smelting. That would be, in effect, to distinguish parts, separable by division, from portions, separable, if at all, otherwise than by division. (It is not necessary for this hypothesis to suppose that Anaxagoras marked that distinction by any explicit distinction of terminology.) I emphasize that this suggestion is offered merely as a possibility, and that I am not maintaining that it has positive textual support. The crucial point is that the interpretation of the ‘everything, in everything’ doctrine which I have defended above is not inconsistent with Aristotle’s statements that Anaxagoras’ basic things were homoeomerous. That doctrine is neither empty nor viciously regressive; it is an ingenious construction which allows Anaxagoras to maintain consistently two of his fundamental theses: (1) there is a portion of every stuff in every stuff, (2) each stuff is characterized by the character of its predominant portion. Its crucial flaw is its lack of explanatory force; the character of a stuff is ‘explained’ by its principal component’s having precisely that character, which is in turn ‘explained’ by its principal component’s having precisely that character, and so on ad infinitum. A central element in explanation, the simplification of a wide range of diverse phenomena via laws connecting those phenomena with a small range of basic properties, is absent. Nor is this an oversight, since the effect of the principle ‘What is F cannot come from what is not F’ is precisely to exclude the possibility that the ‘explanation’ of something’s having a property should not contain that very property in the explanans. The slogan ‘Appearances are the sight of what is non-apparent’ (fr. 21a) thus proves to state a central, and quite startling, Anaxagorean doctrine. At first sight it appears to state the empiricist axiom that theories about what is unobserved must be based on observation, and it was presumably in that sense that Democritus is said by Simplicius to have approved it. But in fact Anaxagoras’ claim is much stronger; he is asserting that the observable phenomena literally do give us sight of what is unobserved, in that the very properties which we observe characterize the world through and through. (This was presumably the point of the remark of Anaxagoras to his associates recorded by Aristotle (Metaphysics 1009b26–8, DK 59 A 28), that they would find that things are just as they supposed.) This does not contradict fragment 21, where Anaxagoras is reported by Sextus as declaring that the weakness of the senses prevents us from judging the truth, and as supporting this claim by citing the imperceptibility of the change produced by pouring a pigment drop by drop into a pigment of a different colour. Rather, the two fragments complement one another. The senses are unable to discern the infinite variety of components in any observable thing, and hence to detect in them the microscopic rearrangements whose accumulation eventually produces an observable change (fr. 21); yet the nature of those components has to be what is revealed by observation at the macroscopic level (fr. 21a). Just as there are in Anaxagoras’ theory no elements, i.e. basic stuffs, so there are no basic properties. It cannot, therefore, be the task of theory to devise an account of the world sufficient to explain the phenomena, since the phenomena must ultimately be self-explanatory. Theory has, however, the more limited task of explaining how the observed world has come to be in the state in which it is; this brings us to another central Anaxagorean concept, that of Mind. Mind In the famous passage of the Phaedo cited above, in which Socrates describes his intellectual progress, he states that he was dissatisfied by the absence of teleleogical explanation from the theories of the early philosophers. Anaxagoras promised to make good this deficiency, since he claimed that the world is organized by Mind. Socrates, assuming that this organization by a cosmic intelligence must aim at the best possible state of things, eagerly perused Anaxagoras’ book for an account of that state and how it was attained, and was all the more disappointed to discover that in his cosmology Anaxagoras made no use of teleology, remaining content, like his predecessors, with purely mechanistic explanations. The evidence of the fragments of Anaxagoras’ views on Mind is consistent with this passage. The most important piece of evidence is fragment 12, which contains a number of theses about the nature and activity of Mind, as follows: Nature Mind is (a) unlimited (b) self-directing (c) separate from everything else (d) the finest and purest of all things (e) all alike, the greater and the less Activity Mind (f) takes thought for everything and has the greatest power (g) controls everything which has a soul (h) directed the entire cosmic rotation, initiating it and continuing it (i) knew all the mixtures and separations of everything (j) organized whatever was, is and will be. The first problem is, what is the reference of Anaxagoras’ term Nous? Is it mind in general, instanced in different individual minds (as in ‘the concept of mind’), or a single cosmic mind? The answer is that it is probably both. The specification of mind given by (a)–(e) seems to be an attempt to differentiate mind as a constituent of the universe from all other constituents. Mind is the finest and purest of all things, it is self-directing (as opposed to other things, which (according to (f), (g) and (j) are directed by mind), and it (alone) is separate from everything else, whereas everything else contains a portion of everything else. (Compare fr. 11, ‘In everything there is a portion of everything except mind, but there are also some things which contain mind.’) But the account of mind’s activity, most especially (h), strongly suggests the activity of a single supreme mind, which organizes the cosmos as a whole. It is clear, too, that that is how Plato represents Socrates as understanding Anaxagoras, especially Phaedo 97c: ‘Mind is what organizes and is the cause of everything…the mind which organizes everything will organize and arrange each thing as is best’. The characteristics listed in (a)–(e) are characteristics of all minds, both ‘the greater and the less’ (i.e. presumably the supreme cosmic mind and subordinate minds, including but not necessarily restricted to human minds), which are explicitly stated in (e) all to be alike. The activities listed in (f)–(j) are activities of the cosmic mind, though (g) may also perhaps refer to an individual mind directing each ensouled thing, doubtless under the overall direction of the cosmic mind. (Aristotle says (On the Soul 2–4, DK 59 A 100) that Anaxagoras sometimes identified soul with mind and attributed the latter to all animals, but appears unsure of what precisely he meant, while the pseudo-Aristotelian work On Plants reports that he regarded plants as a kind of animals and attributed consciousness and thought to them (815a15ff., DK 59 A 117).) Assuming that the fragments refer both to the cosmic mind and to individual minds, they are inexplicit as to the relation between the former and the latter. The minds of humans and of other animals are clearly subordinate to the cosmic mind, but it is unclear what the model of subordination is, i.e. whether particular human and other minds are parts of the cosmic mind, or agents operating under its direction. The only assertion which Anaxagoras supports by any argument is (c): mind cannot be a constituent of any stuff, for if it were it would (by the ‘everything in everything’ principle) be a constituent of every stuff. Why should it not be? Empedocles had maintained that ‘everything has intelligence and a share in thought’ (fr. 110); why should Anaxagoras have demurred? The reason which he gives in fragment 12 is that if mind were a constituent of anything, the other constituents would prevent it from exercising its directive function. Mind has to be external to what it controls, as the rider has to be external to the horse. It is hard to see any force in this argument. We think of organisms as self-directing, and assume that some part of the organism functions as a control mechanism. Why should the mind of a human or animal not be a built-in control mechanism for the animal, or the cosmic mind such a mechanism for the cosmos as a whole? It is problematic precisely because what is implied by the description of mind as unlimited. All stuffs exist eternally (fragment 17), and are therefore temporally boundless, and are unlimited in amount (fragments 1 and 3). Perhaps (a) is simply to be read as making the same claims for mind, but the opening of the fragment appears to contrast mind with the other things, and it is at least tempting to look for a sense of ‘unlimited’ (apeiron) in which mind alone is unlimited. Such a sense may be suggested by fragment 14, which states that mind is where all the other things are. Mind is not, as we have seen, a constituent of anything else, but it knows and controls everything, and is here said to be where everything else is. The picture seems to be of mind as everywhere, pervading everything without being part of anything. This would differentiate mind from the other stuffs, for though every stuff is contained in every stuff, there are some places where it is not, namely those places which are occupied by other stuffs. Mind, on the other hand, if this suggestion is right, is not excluded from any place by the presence of any stuff in that place. This, together with the description of mind as the finest and purest of all things, may suggest that Anaxagoras was groping towards the conception of mind as immaterial, but it would be anachronistic to suggest that that conception is clearly articulated in the fragments. The fragments provide scant information on how the cosmic mind directs and organizes the cosmos. Fragment 12, supplemented by some secondary sources (DK 59 A 12 (Plutarch), 42 (Hippolytus) and 71 (Aetius)), indicates that a cosmic rotation separated out the original undifferentiated mass by centrifugal force into the main elemental masses, and also attests that the original rotation is continuing and will continue to a greater and greater extent. But what the connection is between the rotation and other kinds of natural change, e.g. the generation and development of plants and animals, and how mind is supposed to organize the latter, remains obscure. In particular, our extant evidence, consistently with Socrates’ complaint in the Phaedo (see above) says nothing about how natural change, of whatever kind, is directed towards the best. It is none the less likely that Anaxagoras held the cosmic mind to be divine. The explicit statements to this effect in ps.-Plutarch’s Epitome and in Stobaeus (DK 59 A 48) are not confirmed by similar assertions in the fragments. However, the description of its activity in fragment 12, as ‘taking thought for everything and having the greatest power’, ‘controlling’ (kratein) everything ensouled and the whole cosmic rotation, ‘knowing everything that is mixed together and separated out’ and ‘organizing’ (diakosmein) everything is irresistibly suggestive not only of traditional divinities such as Zeus but also of the cosmic divinities of Anaxagoras’ philosophical predecessors, the divine mind of Xenophanes which ‘without labour controls all things by the thought of its mind’ (DK 21 B 25, cf. 26) and the holy mind of Empedocles ‘darting through the whole cosmos with swift thoughts’ (DK 31 B 134).2 In this respect, as in many of the details of his astronomy and cosmology, Anaxagoras preserves some of the features of earlier Ionian thought. The conventional picture of him as a child of the fifth-century enlightenment is to that extent one-sided, yet it is not altogether inaccurate. In fact the two aspects are complementary; Anaxagoras represents in a striking way the vitality of the Ionian tradition, specifically its adaptability to the rigour of Eleatic thought and to the critical spirit of the later fifth century. That feature is, if anything, even more pronounced in the thought of the atomists, especially that of Anaxagoras’ younger contemporary Democritus. THE ATOMISTS Atomism was the creation of two thinkers, Leucippus and Democritus. The former, attested by Aristotle, our primary source, as the founder of the theory, was a shadowy figure even in antiquity, being over-shadowed by his more celebrated successor Democritus to such an extent that the theory came to be generally regarded as the work of the latter, while Epicurus, who developed and popularized atomism in the third century BC, went so far as to deny that Leucippus ever existed. Nothing is known of his life. Even his birthplace was disputed, some sources associating him with one or other of the two main centres of early Greek philosophy, Miletus and Elea, others with Abdera, the birthplace of Democritus. Of his dates all that can be said is that since he was certainly older than Democritus he lived during the fifth century. No lists of his works survive, and only a single quotation is ascribed to him by a single ancient source (Stobaeus). Only a little more is known about Democritus. He came from Abdera, on the north coast of the Aegean (also the birthplace of Protagoras), and is reported as having described himself as young in the old age of Anaxagoras, i.e. probably in the 430s. He is traditionally said to have lived to a very great age (over 100 years on some accounts), and may therefore be supposed to have lived from about the middle of the fifth till well into the fourth century (though some scholars dispute the accounts of his longevity). He is quoted as saying that he visited Athens (where no one knew him), and is said to have had some slight acquaintance with Socrates. Of his works, which according to the list preserved in Diogenes Laertius’ Life were many and encyclopaedic in scope, including a complete account of the physical universe and works on subjects including astronomy, mathematics, literature, epistemology and ethics, none survive. Ancient sources preserve almost 300 purported quotations, the great majority on ethics (see below), but also including some important fragments on epistemology preserved by Sextus. Our knowledge of the metaphysical foundations and physical doctrines of atomism relies on the doxographical tradition originating from Aristotle, who discusses atomism extensively. The precise relation between Leucippus and Democritus is unclear. Aristotle and his followers treat Leucippus as the founder of the theory, but also assign its basic principles to both Leucippus and Democritus; later sources tend to treat the theory as the work of Democritus alone. While it is clear that the theory originated with Leucippus it is possible that the two collaborated to some extent, and almost certain that Democritus developed the theory into a universal system. Physical Principles According to Aristotle, the atomists, like Anaxagoras, attempted to reconcile the observable data of plurality, motion and change with the Eleatic denial of the possibility of coming to be or ceasing to be. Again like him, they postulated unchangeable primary things, and explained apparent generation and corruption by the coming together and separation of those things. But their conceptions of the primary things and processes differed radically from those of Anaxagoras. For the latter the primary things were observable stuffs and properties, and the primary processes mixing and separation of those ‘elements’. For the atomists, by contrast, the primary things were not properties and stuffs but physical individuals, and the primary processes not mixing and separation but the formation and dissolution of aggregates of those individuals. Again, the basic individuals were unobservable, in contrast with the observable stuffs of Anaxagoras; consequently their properties could not be observed, but had to be assigned to those individuals by theory. Since the theory had to account for an assumed infinity of phenomena, it assumed an infinite number of primary substances, while postulating the minimum range of explanatory properties, specifically shape, size, spatial ordering and orientation within a given ordering. All observable bodies are aggregates of basic substances, which must therefore be too small to be perceived. These corpuscles are physically indivisible (atomon, literally ‘uncuttable’), not merely in fact but in principle; Aristotle reports an (unsound) atomistic argument, which has some affinities with one of Zeno’s arguments against plurality, that if. (as for example Anaxagoras maintained) it were theoretically possible to divide a material thing ad infinitum, the division must reduce the thing to nothing. This Zenonian argument was supported by another for the same conclusion; atoms are theoretically indivisible because they contain no void. On this conception bodies split along their interstices; hence where there are no interstices, as in an atom, no splitting is possible. (The same principle accounts for the immunity of the atoms to other kinds of change, such as reshaping, compression and expansion; all require displacement of matter within an atom, which is impossible without any gaps to receive the displaced matter.) It is tempting to connect the assumption that bodies split only along their interstices with the Principle of Sufficient Reason, which the atomists appealed to as a fundamental principle of explanation (arguing for example that the number of atomic shapes must be infinite, because there is no more reason for an atom to have one shape than another (Simplicius, Physics 28.9–10, DK 67 A 8)). Given the total uniformity of an atom, they may have thought, there could be no reason why it should split at any point, or in any direction, rather than any other. Hence by the Principle of Sufficient Reason, it could not split at all. Atoms are in a state of eternal motion in empty space; the motion is not the product of design, but is determined by an infinite series of prior atomic interactions (whence two of Aristotle’s principle criticisms of Democritus, that he eliminated final causation and made all atomic motion ‘unnatural’). Empty space was postulated as required for motion, but was characterized as ‘what is not’, thus violating the Eleatic principle that what is not cannot be. We have no evidence of how the atomists met the accusation of outright self-contradiction. As well as explaining the possibility of motion, the void was postulated to account for the observed plurality of things, since the atomists followed Parmenides (fragment 8, 22–5) in maintaining that there could not be many things if there were no void to separate them. The theoretical role of the void in accounting for the separation of atoms from one another has an interesting implication, recorded by Philoponus (Physics 494.19–25 (not in DK), On Generation and Corruption 158.26–159.7, DK 67 A 7). Since atoms are separated from one another by the void, they can never strictly speaking come into contact with one another. For if they did, even momentarily, there would be nothing separating them from one another. But then they would be as inseparable from one another as the inseparable parts of a single atom, whose indivisibility is attributed to the lack of void in it (see above); indeed, the two former atoms would now be parts of a single larger atom. But, the atomists held, it is impossible that two things should become one. Holding atomic fusion to be theoretically impossible, and taking it that any case of contact between atoms would be a case of fusion (since only the intervening void prevents fusion), they perhaps drew the conclusion that contact itself is theoretically impossible. Hence what appears to be impact is in fact action at an extremely short distance; rather than actually banging into one another, atoms have to be conceived as repelling one another by some sort of force transmitted through the void. Again, though no source directly attests this, the interlocking of atoms which is the fundamental principle of the formation of aggregates is not strictly speaking interlocking, since the principle of no contact between atoms forbids interlocking as much as impact. Just as impact has to be reconstrued as something like magnetic repulsion, so interlocking has to be reconstrued as quasi-magnetic attraction. If this suggestion is correct (and it is fair to point out that no ancient source other than Philoponus supports it) it is a striking fact that, whereas the post- Renaissance corpuscular philosophy which developed from Greek atomism tended to take the impossibility of action at a distance as an axiom, the original form of the theory contained the a priori thesis that all action is action at a distance; consequently that impact, so far from giving us our most fundamental conception of physical interaction, is itself a mere appearance which disappears from the world when the description of reality is pursued with full rigour. Chance and Necessity While the broad outlines of the views of the atomists on these topics can be fairly readily reconstructed, there is much obscurity about the details. The atomists’ universe is purposeless, mechanistic and deterministic; every event has a cause, and causes necessitate their effects. Broadly speaking the process is mechanical; ultimately, everything in the world happens as a result of atomic interaction. The process of atomic interaction has neither beginning nor end, and any particular stage of that process is causally necessitated by a preceding stage. But exactly how the atomists saw the process as operating is obscure. This obscurity is largely attributable to the fragmentary nature of the evidence which we possess, but it may be that the statement of the theory itself was not altogether free from obscurity. The fundamental text is the single fragment of Leucippus (DK 67 B 1): ‘Nothing happens at random, but everything from reason and by necessity.’ The denial that anything happens ‘at random’ (matēn) might well be taken in isolation to amount to an assertion that all natural events are purposive, since the adverb and its cognates frequently have the sense ‘in vain’ (i.e. not in accordance with one’s purpose) or ‘pointlessly’. If that were the sense of ‘not matēn’ then ‘from reason’ (ek logou) would most naturally be understood as ‘for a purpose’. These renderings are, however, very unlikely. The majority of the sources follow Aristotle (On the Generation of Animals 789b 2–3, DK 68 A 66) in asserting that Democritus denied purposiveness in the natural world, explaining everything by mechanistic ‘necessity’.3 A reading of Leucippus which has him assert, not merely (contra Democritus) that some, but that all natural events are purposive, posits a dislocation between the fundamental world-views of the two of such magnitude that we should expect it to have left some trace in the tradition. Moreover, the attribution of all events to necessity, a central feature of the mechanistic Democritean world-view, is itself attested in the fragment of Leucippus. We ought, then, to look for an interpretation of the fragment which allows it to be consistent with Democritus’ denial of final causation. Such an interpretation is available without forcing the texts. Sometimes (e.g. Herodotus VII.103.2, Plato Theaetetus 189d) matēn is to be rendered not ‘without purpose’ but ‘without reason’ (‘in vain’ and ‘empty’ have similar ranges of application). Given that construal of matēn, ‘from reason’ is to be construed as ‘for a reason’, where the conception of reason is linked to that of rational explanation. The first part of the fragment (‘Nothing happens at random, but everything from reason’) thus asserts, not universal purposiveness in nature, but a principle which we have already seen to be pervasive in atomism, the Principle of Sufficient Reason. Instead of a radical discontinuity between Leucippus and Democritus, the fragment, thus construed, attests commitment to a principle basic to atomism. The second half (‘and by necessity’) makes a stronger claim, which links the notion of rational explanation to the notions of necessity and of cause. The stronger claim is that whatever happens has to be happen, cannot but happen. This amounts to a specification of the reason whose existence is asserted in the first half of the sentence; nothing happens without a reason, and, in the case of everything which happens, the reason for which it happened was that it had to happen. But the claim that whatever happens happens ‘by necessity’ is not just the claim that whatever happens has to happen, though the former implies the latter. For the concept of necessity is not a purely modal concept requiring elucidation via its connection with other such concepts, such as possibility and impossibility. Rather, necessity is conceived as an irresistible force bringing it about that things have to happen. This is indicated both by the causal force of the preposition hypo (rendered ‘by’ in the expression ‘by necessity’), and also by the fact that Democritus is reported as identifying necessity with impact and motion ((Aetius I.26.2, DK 68 A 66) on the interpretation of this see below). Impact and motion, then, take over the determining role traditionally assigned to Necessity, when the latter is conceived (as in Parmenides and Empedocles) as an ineluctable, divine cosmic force (cf. Plato, Protagoras 345d5 ‘Against necessity not even the gods fight’). Nothing, then, just happens; every event occurs because it had to occur, i.e. because it was made to occur by prior impact (namely, of atoms on one another) and prior motion (namely, of atoms). So there can be no chance events, i.e. no events which simply happen. On the other hand, we have evidence that the atomists assigned some role to chance in the causation of events, though precisely what role is not easy to determine. Aristotle (Physics 196a24–8, DK 68 A 69), Simplicius (Physics 327.24–6, DK 68 A 67; 330.14–20, DK 68 A 68) and Themistius (Physics 49.13–16 (not in DK)) all say that Democritus attributed the formation of every primal cosmic swirl to chance (indeed Aristotle finds a special absurdity in the theory that while events in a cosmos occur in regular causal sequences, the cosmos itself comes into being purely by chance). Cicero (On the Nature of the Gods I.24.66, DK 67 A 11) says that heaven and earth come into existence ‘without any compulsion of nature, but by their [i.e. the atoms’] chance concurrence’, while Lactantius (Divine Institutions I.2.1–2, DK 68 A 70) baldly attributes to Democritus and Epicurus the view that ‘everything happens or comes about fortuitously’. Aetius I.29.7: ‘Democritus and the Stoics say that it [i.e. chance] is a cause which is unclear to human reason’ may be read either as asserting or as denying that Democritus believed that there are genuinely chance events. Read in the latter way it attributes to Democritus the view that we explain an event as due to chance when its real cause is unknown; on the former reading the view attributed to Democritus is that chance is itself a real cause of events, but an unfathomable one (the position mentioned by Aristotle without attribution at Physics 19605–7). A passage from Epicurus’ On Nature (fr. 34.30 in Arrighetti), which one might hope to be our most authoritative source, is similarly ambiguous. There Epicurus describes the atomists as ‘making necessity and chance responsible for everything’, a formulation which is ambiguous between two positions; (1) ‘necessity’ and ‘chance’ are two names for a single universal cause, (2) necessity and chance are distinct but jointly exhaustive causes of everything.4 The passage of Lactantius is of little weight; he states that the fundamental question is whether the world is governed by providence or whether everything happens by chance, and says that Epicurus and Democritus held the latter view. It is plausible that he took their denial of providence to commit them to that view, since he himself took those alternatives to be exhaustive. This passage, then, gives no independent ground for the attribution to either philosopher of the thesis that literally everything happens by chance. We are still, however, left with those passages attesting Democritus’ belief that every cosmic swirl, and therefore every cosmos, come into being by chance. That might be thought to be confirmed by the statement in Diogenes Laertius’ summary of Democritus’ cosmology that he identified the cosmic swirl itself with necessity (IX.45, DK 68 A 1). On this interpretation the statement that everything happens by necessity is confined to events within a cosmos, and states that all such events are determined by the atomic motions constituting the swirl. The swirl itself, however, is not determined by itself, nor by anything; it just happens. Eusebius (Praeparatio Evangelica XIV.23.2, DK 68 A 43) also reports Democritus as ascribing the formation of worlds to chance, and goes further by reporting him as holding that the pre-cosmic motion of the atoms was also random (‘these atoms travel in the void hōs etuchen (literally “as it chanced”). On this view necessity governs, but is local to, a world order, which itself arises by chance from a pre-cosmic state where there is no necessity. The recognition of pure chance is, however, inconsistent with the Principle of Sufficient Reason, which we know the atomists accepted. It therefore seems preferable to look for some interpretation of the evidence which is consistent with that principle. That interpretation is provided by the first reading of the Aetius passage cited above, namely that the ascription of events to chance is a confession of ignorance of their causes, not a denial that they have causes. Some features of the evidence support this suggestion. Diogenes’ summary of the cosmology of Leucippus (IX.30–3, DK 67 A 1) concludes with the sentence, ‘Just like the coming into being of worlds, so do their growth, decay, and destruction occur according to a certain necessity, the nature of which he does not explain.’ In line with his famous dictum, then, Leucippus held that all events including the formation of worlds happen according to necessity, but was unable to say what it is that necessitates cosmic events. It is then plausible that either he himself or Democritus said that such events may be said to occur by chance, in the sense that we are (whether merely in fact or in principle is indeterminate) ignorant of their causes. Simplicius’ evidence suggests just that; in Physics 327. 24–6 his attribution to Democritus of the view that the cosmic swirl arises by chance is avowedly his own inference from the fact that Democritus did not say how or why that occurs. In Physics 330.14–20 he says that although Democritus appeared (edokei) to have made use of chance in his account of the formation of worlds, in his more detailed discussions (en tois merikōterois) he says that chance is not the cause of anything. That suggests that he merely seemed to ascribe cosmogony to chance (perhaps by speaking of it as a chance occurrence in the sense of an occurrence whose cause is unknown). Explanations of specific kinds of events and of particular events were governed by the principle that there are no chance events, but no attempt was made to offer explanations of the fundamental cosmic processes themselves. That need not imply that they are literally uncaused, but that they might as well be treated as such, since their actual causes are of a degree of complexity outstripping the powers of the human mind to discover. For the atomists, then, everything happens of necessity; the identification of necessity with the mechanical forces of impact and motion may have been due to Democritus. But what exactly was his view on this? Aetius (I.26.2, DK 68 A 66) reports him as identifying necessity with ‘impact and motion and a blow of matter’. Are impact and motion given equal status in this identification, or is it taken for granted that motion is always caused by prior impact? On the former construal some motion may be either uncaused, or attributable to a cause other than impact. In favour of the first alternative is Aristotle’s evidence (Physics 252a32–b2, DK 68 A 65) that Democritus held that one should not ask for a cause of what is always the case. He might then have said that the atoms are simply always in motion. But while that principle allows him to exclude the question, ‘What causes the atoms to be in motion?’, the Principle of Sufficient Reason requires that the question, ‘Why is any particular atom moving with any particular motion?’ should have an answer, and it might appear inevitable that that answer should refer to a prior atomic collision. We have, however, to recall the evidence from Philoponus that atoms never actually collide or come into contact, with its implication that the basic physical forces are attraction and repulsion. Attraction, as we saw, explains, not atomic motion, but the immobility of atoms relative to one another, since the relative stability of atoms in an aggregate has to be explained, not by their literal interlocking, but by their being held together as if interlocked by an attractive force operating over the tiny gaps between the atoms in the aggregate. In addition, some form of attraction may also have explained some atomic motions; Sextus cites Democritus (Adversus Mathematicos VII.116–8, DK 68 B 164) as holding that things of the same kind tend to congregate together, and as illustrating that phenomenon by examples of the behaviour of animate (birds flocking together) and inanimate things (grains of different sorts being separated out by the action of a sieve, pebbles of different shapes being sorted together by the action of waves on a beach). That this principle was applied to the atoms appears from Diogenes’ account of the cosmogony of Leucippus, where atoms of all shapes form a swirling mass from which they are then separated out ‘like to like’. The separation out of atoms of different sizes could adequately be accounted for by the stronger centripetal tendency of the larger, itself a function of their greater mass. But the context in Diogenes, where the atoms have just been described as of all shapes, with no mention so far of size, suggests that ‘like to like’ is here to be understood as ‘like to like in shape’. Aetius’ report of Democritus’ account of sound (IV.19.3, DK 68 A 128) asserts that atoms of like shape congregate together, and contains the same illustrative examples as the Sextus passage; it is plausible, though not explicitly asserted, that this same principle accounts for the formation of aggregates of spherical atoms, for example flames. We have, then, evidence that Democritus’ dynamics postulated three fundamental forces: a repulsive force which plays the role of impact in a conventional corpuscular theory, and two kinds of attractive force, one of which draws together atoms of the same shape and another which holds together atoms of different shapes in an atomic aggregate. It is plausible that he applied the term ‘necessity’ to all three, regarding them alike as irresistible. It must, however, be acknowledged first that the evidence for this theory is fragmentary and also that even if it is accepted we have no idea whether or how Democritus attempted to unify these forces into a unified theory. Stated thus baldly, the theory has obvious difficulties; for example, if two atoms of the same shape collide, do they rebound or stick together? If all atoms have both attractive and repulsive force there must be some yet more basic principles determining what force or combination of forces determines their motion. Our sources give no hint of whether Democritus had so much as considered such questions. Epistemology and Psychology While we have no evidence to suggest that Leucippus was concerned with epistemological questions, there is abundant evidence of their importance for Democritus. It is quite likely that the latter’s epistemological interests were stimulated at least in part by his fellow citizen and elder contemporary Protagoras (see below). Our evidence is highly problematic, in that it provides support for the attribution to Democritus of two diametrically opposed positions on the reliability of the senses. On the one hand, we have a number of passages, including some direct quotations, in which he is seen as rejecting the senses as totally unreliable; on the other, a number of passages ascribe to him the doctrine that all appearances are true, which aligns him with Protagorean subjectivism, a position which he is, however, reported as having explicitly rejected. The former interpretation is supported mainly by evidence from Sextus, and the latter mainly by evidence from Aristotle and his commentators, but we cannot resolve the question by simply setting aside one body of evidence in favour of the other. The reasons are that: (1) in the course of a few lines (Metaphysics 1009b7–17, DK 68 A 112) Aristotle says both that Democritus says that either nothing is true, or it is unclear to us, and that he asserts that what appears in perception is necessarily true; (2) in Adversus Mathematicos (VII.136, DK 68 B 6) Sextus ascribes some of Democritus’ condemnation of the senses to a work in which ‘he had undertaken to give the senses control over belief. Prima facie, then, the evidence suggests that both interpretations reflect aspects of Democritus’ thought. Was that thought, then, totally inconsistent? Or can the appearance of systematic contradiction be eliminated or at least mitigated? The former interpretation is based on the atomists’ account of the secondary qualities, whose observer-dependence Democritus seems to have been the first philosopher to recognize. Our senses present the world to us as consisting of things characterized by colour, sound, taste, smell, etc., but in reality the world consists of atoms moving in the void, and neither atoms nor the void are characterized by any secondary quality. We thus have a dichotomy between how things seem to us and how they are in reality, expressed in the celebrated slogan (fr. 9) ‘By convention sweet and by convention bitter, by convention hot, by convention cold, by convention colour, but in reality atoms and the void.’ Further, the distinction between the reality of things and the appearances which that reality presents has to be supplemented by an account of the causal processes via which we receive those appearances. Atomic aggregates affect us by emitting from their surfaces continuous streams of films of atoms which impinge on our sense organs, and the resulting perceptual states are a function of the interaction between those films and the atomic structure of the organs. For example, for an object to be red is for it constantly to emit films of atoms of such a nature that, when those films collide with an appropriately situated perceiver, the object will look red to that perceiver. Hence we are doubly distanced from reality; not only phenomenologically, in that things appear differently from how they are, but also causally, in that we perceive atomic aggregates via the physical intervention of other aggregates (namely the atomic films) and the action of those latter on our sense organs. A number of fragments stress the cognitive gulf which separates us from reality: (fr. 6) ‘By this principle man must know that he is removed from reality’; (fr. 8) ‘Yet it will be clear that to know how each thing is in reality is impossible’; (fr. 10) ‘That in reality we do not how each thing is or is not has been shown many times’ and (fr. 117) ‘In reality we know nothing, for truth is in the depths.’ This evidence immediately presents a major problem of interpretation. On the one hand fragment 9 and associated reports stress the gulf between appearance and reality, claiming that the senses are unreliable in that they misrepresent reality. That dogmatic claim presupposes that we have some form of access to reality, which enables us to find the sensory picture unfaithful to how things are in fact. On the other hand, fragments 6, 8, 10 and 117 make the much more radical claim that reality is totally inaccessible, thereby undercutting the thesis that there is a gulf between appearance and reality. Fragment 7, ‘This argument too shows that in reality we know nothing about anything, but each person’s opinion is something which flows in’ and the second half of fragment 9, ‘In fact we know nothing firm, but what changes according to the condition of our body and of the things that enter it and come up against it’ attempt uneasily to straddle the two positions, since they draw the radically sceptical conclusion from a premiss about the mechanism of perception which presupposes access to the truth about that mechanism. We might conclude that Democritus simply failed to distinguish the dogmatic claim that the senses misrepresent reality from the sceptical claim that we can know nothing whatever about reality. An alternative strategy is to look for a way of interpreting the evidence which will tend to bring the two claims nearer to consonance with one another. We can bring the two claims closer to one another if the ‘sceptical’ fragments are interpreted as referring, not to cognitive states generally, but specifically to states of sensory cognition. These fragments will then simply reiterate the thesis that we know nothing about the nature of reality through the senses, a thesis which is consistent with the slogan stated in the first half of fragment 9 and which dissolves the apparent tension internal to fragment 7 and the second half of fragment 9. Support for that suggestion comes from consideration of the context in which Sextus quotes fragments 6–10, namely that of Democritus’ critique of the senses; of this Sextus observes, ‘In these passages he more or less abolishes every kind of apprehension, even if the senses are the only ones which he attacks specifically.’ It thus appears that Sextus understands Democritus as referring in these fragments to the senses only, though in his (i.e. Sextus’) view the critique there directed against the senses in fact applies to all forms of apprehension. This is confirmed by the distinction which Sextus immediately (Adversus Mathematicos VII.135–9) attributes to Democritus between the ‘bastard’ knowledge provided by the senses and the ‘genuine’ knowledge provided by the intellect (fr. 11). The latter is specifically said to be concerned with things which fall below the limits of sensory discrimination, and we must therefore suppose that the atomic theory itself is to be ascribed to this form of knowledge. This is supported by those passages (ibid. VIII.6–7, 56) in which Sextus associates the position of Democritus with that of Plato, in that both reject the senses as sources of knowledge and maintain that only intelligible things are real; for Plato, of course, the intelligible things are the Forms, whereas for Democritus they are the atoms, which are inaccessible to perception and, consequently, such that their properties are determinable only by theory. Thus far the prospects for a unified interpretation of Democritus’ epistemology look promising. The position expressed in the fragments cited by Sextus is not general scepticism, but what we might term theoretical realism. The character of the physical world is neither revealed by perception nor inaccessible to us; it is revealed by a theory which, starting from perceptual data, explains those data as appearances generated by the interaction between a world of imperceptible physical atoms and sensory mechanisms also composed of atoms. But now, as Sextus points out (ibid. VIII.56 (not in DK)) and Democritus himself recognized (in the famous ‘complaint of the senses’ (fr. 125)) scepticism threatens once again; for the theory has to take perceptual data as its startingpoint, so if the senses are altogether unreliable, there are no reliable data on which to base the theory. So, as the senses say to the mind in fragment 125, ‘Our overthrow is a fall for you.’ Commentators who (like Barnes [2.8]) read fragment 125 as expressing commitment to scepticism (despairing or exultant, according to taste) on the part of Democritus, naturally reject the unitary interpretation proffered above. On this view fragments 117 and 6–10 are not restricted to sensory cognition, but express a full-blooded rejection of any form of knowledge, which must be seen as superseding the distinction between appearance and reality of fragments 9 (first part) and 11 and the claim to ‘genuine knowledge’ in the latter. Yet Sextus presents 6–11 in a single context (Adversus Mathematicos VII.135–40) without any suggestion of a conflict within the collection. Moreover, in Outlines of Pyrrhonism I.213–4 (not in DK) he points out that, though the Sceptics resemble Democritus in appealing to phenomena of conflicting appearances, such as the honey which tastes sweet to the healthy and bitter to the sick, in fact Democritus uses those phenomena to support, not the sceptical position that it is impossible to tell how the honey is in fact, but the dogmatic position that the honey is itself neither sweet nor bitter. (I interpret the latter as the assertion that sweetness and bitterness are not intrinsic attributes of the structure of atoms which is the honey (see above).) Sextus, in short, sees Democritus not as a sceptic, but as a dogmatist. Indeed, Sextus does not cite fragment 125, and it is possible that he did not know the text from which it comes; VIII.56 shows that he was aware of the problem which is dramatized in the fragment, but he clearly saw it as a difficulty for Democritus, rather than as signalling Democritus’ rejection of the basis of his own theory. At this point we should consider in what sense the theory of atomism takes the data of the senses as its starting-point, and whether that role is in fact threatened by the appearance—reality gap insisted on in fragment 9. According to Aristotle (On Generation and Corruption 315b6–15, DK 67 A 9; 325b24–6, DK 67 A 7) the theory started from sensory data in the sense that its role was to save the appearances, i.e. to explain all sensory data as appearances of an objective world. Both Aristotle (On Generation and Corruption) and Philoponus (his commentary, 23.1–16 (not in DK)) mention conflicting appearances as among the data to be saved; the theory has to explain both the honey’s tasting sweet to the healthy and its tasting bitter to the sick, and neither appearance has any pretensions to represent more faithfully than the other how things are in reality. All appearances make an equal contribution to the theory. That is a position which atomism shares with Protagoras, but the latter assures the equal status of appearances by abandoning objectivity; in the Protagorean world there is nothing more to reality than the totality of equipollent appearances. For Democritus, by contrast, the reconciliation of the equipollence of appearances with the objectivity of the physical world requires the gap between appearance and reality. Without the gap a world of equipollent appearances is inconsistent, and hence not objective. But there is no ground for denying equipollence; qua appearance, every appearance is as good as every other. Hence the task of theory is to arrive at the best description of an objective world which will satisfy the requirement of showing how all the conflicting appearances come about. So far from threatening the foundations of the theory, then, the appearancereality gap is essential to the theory. But in that case what is the point of the complaint of the senses in fragment 125? Surely that text provides conclusive evidence that Democritus believed that the gap threatened the theory, and hence (assuming that he understood his own theory) conclusive evidence against the interpretation which I am advancing. I do not think that the text does provide such evidence, for the simple reason that we lack the context from which the quotation comes. The point of the complaint need not (and given the nature of Democritus’ theory certainly should not) be the admission that the theory is selfrefuting. It is at least as likely to be a warning against misunderstanding the account of the appearance-reality gap as requiring the abandonment of sensory evidence. We may imagine an anti-empiricist opponent (Plato, say) appealing to the gap to support the claim that the senses are altogether unreliable, and should therefore be abandoned (as is perhaps indicated by Phaedo 65–6). In reply Democritus points out that the attack on the senses itself relies on sensory evidence. Sextus does indeed align Democritus with Plato in this regard (Adversus Mathematicos VIII.56). It is my contention, however, that when we put the Aristotelian evidence of the atomists’ acceptance of the appearances as the starting-point of their theory together with all the other evidence, including the fragments, we have to conclude that the picture of Democritus as a failed Platonist is a misunderstanding. The atomists’ distinction between appearance and reality does not involve ‘doing away with sensible things’; on the contrary, appearances are fundamental to the theory, first as providing the data which the theory has to explain and second as providing the primary application for the observationally-based terminology which is used to describe the nature and behaviour of the entities posited by the theory (cf. [6.46]). A final objection, however, comes from Aristotle himself, who describes Democritus as concluding from conflicting appearances ‘that either nothing is true, or it is unclear to us’ (Metaphysics 1009b11–12). This is a very puzzling passage, for a number of reasons. Aristotle is explaining why some people go along with Protagoras in believing that whatever seems to be the case is so, and in the immediate context (1009a38ff.) cites the phenomena of conflicting appearances and the lack of a decisive criterion for choosing between them as conducing to that belief. But at 1009b9 he shifts from the thought that conflicting appearances lead to the view that all appearances are true to the sceptical account of those phenomena, namely that it is unclear which of the appearances is true or false, ‘for this is no more true than that, but they are alike’. This, he says (i.e. the belief that none of the appearances is truer than any other) is why Democritus said that either nothing is true, or it is unclear to us. So Democritus is represented as posing a choice of adopting either the dogmatic stance that none of the appearances is true, or the sceptical stance that it is unclear (which is true). Yet in the next sentence Aristotle says that because he and others assimilate thought to perception they hold that what appears in perception is necessarily true, the position which we have already seen him attribute to Democritus in a number of places. So unless Aristotle is radically confused, the disjunction ‘either none of the appearances is true, or it is unclear to us’ must be consistent with the thesis that all perceptions are true. If ‘it is unclear to us’ is read as ‘it is unclear to us which is true’, then the claims are inconsistent. I suggest, however, that what Democritus said was to the effect that ‘either nothing is true, or it (i.e. the truth) is unclear’. The first alternative he plainly rejected, so he maintained the second. And that is precisely what he maintains in fragment 117: the truth (about the atoms and the void) is in the depths, i.e. it is not apparent in perception, i.e. it is unclear (adēlon) in the sense that it is not plain to see. That he used the term adēlon to apply to atoms and the void is attested by Sextus (Adversus Mathematicos VII.140, DK 68 A 111), who cites Diotimus as evidence for Democritus’ holding that the appearances are the criterion for the things that are unclear and approving Anaxagoras’ slogan ‘the appearances are the sight of the things that are unclear’ (opsis tōn adēlōn ta phainomena). The truth, then, i.e. the real nature of things, is unclear (i.e. non-evident), but all perceptions are true in that all are equipollent and indispensable to theory. If that is what Democritus held, then it may reasonably be said that ‘true’ is the wrong word to characterize the role of appearances in his theory. ‘All appearances are equipollent’ is equally compatible with ‘All appearances are false’, and in view of his insistence on the non-evident character of the truth it would surely have been less misleading for him to say the latter. Though there are some difficult issues here, I shall not argue the point, since I am not concerned to defend Democritus’ thesis that all appearances are true. I do, however, accept that he actually maintained that thesis and have sought to explain why he did and how he held it together with (1) his rejection of Protagorean subjectivism and (2) the views expressed in the fragments cited by Sextus.5 In conclusion, it should be observed that the persuasiveness or otherwise of the atomists’ account of the secondary qualities cannot be separated from that of the whole theory of perception of which it is pan, and that in turn from the theory of human nature, and ultimately of the natural world as a whole. As presented by the atomists, the theory is entirely speculative, since it posits as explanatory entities microscopic structures of whose existence and nature there could be no experimental confirmation. Modern developments in sciences such as neurophysiology have revised our conceptions of the structures underlying perceptual phenomena to such an extent that modern accounts would have been unrecognizable to Leucippus or Democritus; but the basic intuitions of ancient atomism, that appearances are to be explained at the level of the internal structure of the perceiver and of the perceived object, and that the ideal of science is to incorporate the description of those structures within the scope of a unified and quantitatively precise theory of the nature of matter in general, have stood the test of time. Democritus’ uncompromising materialism extended to his psychology. Though there is some conflict in the sources, the best evidence is that he drew no distinction between the rational soul or mind and the non-rational soul or life principle, giving a single account of both as a physical structure of spherical atoms permeating the entire body. This theory of the identity of soul and mind extended beyond identity of physical structure to identity of function, in that Democritus explained thought, the activity of the rational soul, by the same process as that by which he explained perception, one of the activities of the sensitive or non-rational soul. Both are produced by the impact on the soul of extremely fine, fast-moving films of atoms (eidōla) constantly emitted in continuous streams by the surfaces of everything around us. This theory combines a causal account of both perception and thought with a crude pictorial view of thought. The paradigm case of perception is vision; seeing something and thinking of something alike consist in picturing the thing seen or thought of, and picturing consists in having a series of actual physical pictures of the thing impinge on one’s soul. While this assimilation of thought to experience has some affinities with classical empiricism, it differs in this crucial respect, that whereas the basic doctrine of empiricism is that thought derives from experience, for Democritus thought is a form of experience, or, more precisely, the categories of thought and experience are insufficiently differentiated to allow one to be characterized as more fundamental than the other. Among other difficulties, this theory faces the problem of accounting for the distinction, central to Democritus’ epistemology, between perception of the observable properties of atomic aggregates and thought of the unobservable structure of those aggregates. We have no knowledge of how, if at all, Democritus attempted to deal with this problem. Theology Another disputed question is whether Democritus’ materialistic account of the universe left any room for the divine. According to most of the ancient sources, he believed that there are gods, which are living, intelligent, material beings (of a peculiar sort), playing a significant role in human affairs. They are atomic compounds, and like all such compounds they come to be and perish. They did not create the physical world (of which they are pan), nor, though they are intelligent, do they organize or control it. They are as firmly part of the natural order as any other living beings. Specifically, Democritus believed the gods to be living eidōla, probably of gigantic size, possessing intelligence, moral character and interest in human affairs. While some sources suggest that these eidōla emanate from actual divine beings, the majority of sources agree that they are themselves the only divine beings which Democritus recognized. Some modern scholars (e.g. Barnes [2.8], ch. 21 (c)) interpret this as amounting to atheism, taking Democritus to have held that the gods are nothing more than the contents of human fantasy. But for Democritus eidōla are not intrinsically psychological; they are not contents of subjective states, but part of the objective world, causing psychological states through their impact on physical minds. In that case the theory must explain their source and their properties, notably their being alive. Since they are of human form, it is plausible to suggest that their source is actual humans, possibly giants living in the remote past. They are themselves alive in that, flowing from beings permeated with soul-atoms, they contain soul-atoms themselves. Consistently with this naturalistic theology Democritus gave a naturalistic account of the origin of religion, identifying two types of phenomena as having given rise to religious belief, first the occurrence of eidōla themselves, presumably in dreams and ecstatic states, and second celestial phenomena such as thunder, lightning and eclipses. Democritus’ theology thus contrives to incorporate some of the most characteristic features of the gods of traditional belief, notably their anthropomorphism, power, longevity (though not, crucially, immortality) personal interaction with humans and interest (for good or ill) in human affairs, within the framework of a naturalistic and materialistic theory. It is thus, despite the bold originality of its account of the divine nature, notably more conservative than some of its predecessors (especially the non-anthropomorphic theology of Xenophanes) and than its Epicurean successor, whose main concern is to exclude the gods from all concern with human affairs. Ethics and Politics The evidence for Democritus’ ethical views differs radically from that for the areas discussed above, since while the ethical doxography is meagre, our sources preserve a large body of purported quotations on ethical topics, the great majority from two collections, that of Stobaeus (fifth century AD) and a collection entitled The sayings of Democrates’. While the bulk of this material is probably Democritean in origin, the existing quotations represent a long process of excerpting and paraphrase, making it difficult to determine how close any particular saying is to Democritus’ own words. Various features of style and content suggest that Stobaeus’ collection of maxims contains a greater proportion of authentically Democritean material than does the collection which passes under the name of ‘Democrates’. Subject to the limitations imposed by the nature of this material, we can draw some tentative conclusions about Democritus’ ethical views. He was engaged with the wide-ranging contemporary debates on individual and social ethics of which we have evidence from Plato and other sources. On what Socrates presents as the fundamental question in ethics, ‘How should one live?’ (Plato, Gorgias 500c, Republic 352d), Democritus is the earliest thinker reported as having explicitly posited a supreme good or goal, which he called ‘cheerfulness’ (euthumia) or ‘well-being’ (euestō), and which he appears to have identified with the untroubled enjoyment of life. It is reasonable to suppose that he shared the presumption of the primacy of self-interest which is common both to the Platonic Socrates and to his immoralist opponents, Callicles and Thrasymachus. Having identified the ultimate human interest with ‘cheerfulness’, the evidence of the testimonia and the fragments is that he thought that it was to be achieved by moderation, including moderation in the pursuit of pleasures, by discrimination of useful from harmful pleasures and by conformity to conventional morality. The upshot is a recommendation to a life of moderate, enlightened hedonism, which has some affinities with the life recommended by Socrates (whether in his own person or as representing ordinary enlightened views is disputed) in Plato’s Protagoras, and, more obviously, with the Epicurean ideal of which it was the forerunner. An interesting feature of the fragments is the frequent stress on individual conscience. Some fragments stress the pleasures of a good conscience and the torments of a bad one (frs 174, 215) while others recommend that one should be motivated by one’s internal sense of shame rather than by concern for the opinion of others (frs 244, 264, Democrates 84). This theme may well reflect the interest, discernible in contemporary debates, in what later came to be known as the question of the sanctions of morality. A recurrent theme in criticisms of conventional morality was that, since the enforcement of morality rests on conventions, someone who can escape conventional sanctions, e.g. by doing wrong in secret, has no reason to comply with moral demands (see Antiphon fr. 44 DK, Critias fr. 25 DK and Glaucon’s tale of Gyges’ ring in Plato’s Republic, 359b–360d). A defender of conventional morality who, like Democritus and Plato, accepts the primacy of self-interest therefore faces the challenge of showing, in one way or another, that self-interest is best promoted by the observance of conventional moral precepts. The appeal to divine sanctions, cynically described in Critias fragment 25, represents one way of doing this, and there are some traces of the same response in Democritus. While his theory of the atomic, and hence mortal, nature of the soul admits no possibility of postmortem rewards and punishments, the theory allows for divine rewards and punishments in this life. Fragment 175 suggests a complication: the gods bestow benefits on humans, but humans bring harm on themselves through their own folly. Is the thought that the gods do not inflict punishment arbitrarily, but that humans bring it on themselves? Or is it rather that the form which divine punishments take is that of natural calamities, which humans fail to avoid through their own folly? The latter alternative would make the pangs of conscience one of the forms of divine punishment, while the former would see it as a further sanction. Either way (and the question is surely unanswerable) we have some evidence that Democritus was the earliest thinker to make the appeal to ‘internal sanctions’ central to his attempt to derive morality from self-interest, thus opening up a path followed by others including Butler and J.S.Mill. The attempt, however pursued, to ground morality in self-interest involves the rejection of the antithesis between law or convention (nomos) and nature (phusis) which underlies much criticism of morality in the fifth and fourth centuries. For Antiphon, Callicles, Thrasymachus and Glaucon, nature prompts one to seek one’s own interest while law and convention seek, more or less successfully, to inhibit one from doing so. But if one’s long-term interest is the attainment of a pleasant life, and if the natural consequences of wrongdoing, including ill-health, insecurity and the pangs of conscience, give one an unpleasant life, while the natural consequences of right-doing give one a contrastingly pleasant life, then nature and convention point in the same direction, not in opposite directions as the critics of morality had alleged. (We have no evidence whether Democritus had considered the objections that conscience is a product of convention, and that exhorting people to develop their conscience assumes that it must be.) Though the texts contain no express mention of the nomos-phusis contrast itself, several of them refer to law in such a way as to suggest rejection of the antithesis. Fragment 248 asserts that the aim of law is to benefit people, thus contradicting Glaucon’s claim (Republic 359c) that law constrains people contrary to their natural bent. Fragment 248 is supplemented and explained by fragment 245; laws interfere with people’s living as they please only to stop them from harming one another, to which they are prompted by envy. So law frees people from the aggression of others, thus benefiting them by giving them the opportunity to follow the promptings of nature towards their own advantage. The strongest expression of the integration of nomos and phusis is found in fragment 252: the city’s being well run is the greatest good, and if it is preserved everything is preserved, while if it is destroyed everything is destroyed. A stable community, that is to say, is necessary for the attainment of that well-being which is nature’s goal for us. This quotation encapsulates the central point in the defence of nomos (emphasized in Protagoras’ myth (Plato, Protagoras 322a–323a)) that law and civilization are not contrary to nature, but required for human nature to flourish, a point also central to the Epicurean account of the development of civilization (see especially Lucretius V). I conclude with a brief discussion of the vexed question of the connections (or lack of them) between Democritus’ ethics and his physical theory. In an earlier discussion ([6.46]), I argued against Vlastos’s claim [6.47] to find significant connections between the content of the two areas of Democritus’ thought. Vlastos’s position has found some recent defenders (and my views some critics), including Sassi [6.43]; these discussions seem to me to call for some reexamination of the question. It is, I take it, common ground that in composing his ethical writings Democritus had not abandoned his physical theory, and therefore that, at the very least, he would have sought to include nothing in the former which was inconsistent with the latter. I shall make the stronger assumption that he took for granted in the ethical writings the atomistic view of the soul as a physical substance pervading the body. I remain, however, unconvinced of any closer connection between physics and ethics. In particular, I see no indication that any ethical conclusions (e.g. that the good is ‘cheerfulness’) were supposed to be derived from the physical theory, or that the physical theory provided any characterizations of the nature of any ethically significant psychological state. Put in modern terms, I see no evidence that Democritus believed in type—type identities between ethical states such as cheerfulness and physical states such as having one’s soul-atoms in ‘dynamic equilibrium’ (Vlastos, in [4.64], 334). My earlier criticisms of this kind of view seem to me to stand. There is, however, one particular point on which I now think that I took scepticism too far. This was in my rejection of Vlastos’s interpretation of fragment 33, that teaching creates a new nature by altering the configuration of the soul-atoms. My reason was that ruthmos was an atomistic technical term for the shape of an individual atom, not for the configuration of an atomic aggregate, for which their term was diathigē. Hence metaruthmizei (or metarusmoi) in the fragment could not mean ‘reshape’ in the sense ‘produce a new configuration’. But, as Vlastos had already pointed out, the catalogue of Democritean titles includes Peri ameipsirusmiōn ‘On changes of shape’ (Diogenes Laertius IX.47), which cannot refer to changes in the shapes of individual atoms (since they are unchangeable in respect of shape), and must therefore refer to changes in the shape of atomic aggregates. Further, Hesychius glosses ameipsirusmein as ‘change the constitution (sungkrisin) or be transformed’, and though he does not attribute the word to any author it is at least likely to have been used in that sense by Democritus, since neither the verb nor its cognates are attested to anyone else. It therefore now seems to me that Vlastos’s reading of the fragment is probably right. Teaching, like thought and perception, is for Democritus a physical process involving the impact of eidōla on the soul, with consequent rearrangement of the soul-aggregate. (Cf. fr. 197, ‘The unwise are shaped (rusmountai) by the gifts of fortune…’) Acceptance of that causal picture does not, of course, commit one to endorsing type-type psychophysical identities. Psycho-physical identity having been set aside, some looser connections between Democritus’ ethics and other areas of his thought may perhaps be discerned. I argued [6.46] for a structural parallel between ethics and epistemology, a suggestion which still seems to me plausible. Another vague connection is with cosmology. It is not unreasonable to suppose that Democritus saw at least an analogy between the formation of worlds (kosmoi) from the primitive atomic chaos by the aggregation of atoms under the force of necessity and the formation of communities (also termed kosmoi, frs 258–9) by individuals driven by necessity to combine in order to survive, and it may be that the aggregation of like individuals to like, which is attested as operating in the formation of worlds (DK 67 A 1 (31)), had some counterpart in the social sphere. Conclusion Atomism can thus be seen as a multi-faceted phenomenon, linked in a variety of ways to various doctrines, both preceding, contemporary and subsequent. Atomistic physics is one of a number of attempts to accommodate the Ionian tradition of comprehensive natural philosophy to the demands of Eleatic logic. Atomistic epistemology takes up the challenge of Protagorean subjectivism, breaks new ground in its treatment of the relation of appearance to reality and constitutes a pioneering attempt to grapple with the challenge of scepticism. Atomistic ethics moves us into the world of the sophists and of early Plato in its treatment of the themes of the goal of life, and of the relations between selfinterest and morality and between nomos and physis. Chapters in subsequent volumes attest the enduring influence of the atomism of Leucippus and Democritus throughout the centuries, whether as a challenge to be faced, most notably by Aristotle, or as a forerunner to Epicureanism in all its aspects, and thereby to the revival of atomistic physics in the corpuscular philosophy of the sixteenth and seventeenth centuries.6 NOTES 1 S.Kripke, Naming and Necessity, 2nd edn., Oxford, Blackwell, 1980. 2 For fuller discussion see Lesher [6.14]. 3 An apparent exception is Aetius I.25.3 (DK 28 A 32, from ps.—Plutarch and Stobaeus). After ascribing to Democritus (and Parmenides) the doctrine that everything is according to necessity, the citation continues ‘and the same is fate and justice and providence and the creator’. The reference of ‘the same’ (tēn autēn) is presumably the feminine noun anangkē; Democritus is therefore said to have identified necessity with fate, justice, providence and the creator. Apart from the authority of this testimony, its meaning is problematic. It might be taken (in opposition to all the other evidence), as ascribing purpose and moral content to necessity, but could as well be taken as explaining justice and providence away as nothing more than necessity, i.e. as saying ‘necessity is what (socalled) fate, justice and cosmic providence really are’. Since in the next section ps.—Plutarch cites Democritus’ mechanistic account of necessity as impact (I.26.2, DK 68 A 66) consistency is better preserved by the latter reading. 4 In Epicurus’ own theory, chance and necessity are distinct causes (Letter to Menseceus 133, Diogenes Laertius Lives X, sections 122–35), so if he is assuming that the atomists share his view, the position he ascribes to them is (2). But that assumption is not required by the text, which leaves open the possibility that the view ascribed is (1). 5 Richard McKim argues [6.39] that Democritus held all appearances to be true in a robuster sense of ‘true’ than that for which I argue here, namely that ‘they are all true in the sense that they are true to the eidōla or atomic films which cause them by streaming off the surfaces of sensible objects and striking our sense organs’ (p. 286). Though McKim does not discusss what it is for appearances to be true to the eidōla, I take it that he is attributing to Democritus the account of the truth of appearances which Epicurus is held by some writers to have maintained, namely that sense impressions faithfully register the physical characteristics of the eidōla which impinge on the sense organs. (See G.Striker ‘Epicurus on the truth of senseimpressions’, Archiv für Geschichte der Philosophie 59 (1977): 125–42 and C.C.W.Taylor ‘All impressions are true’, in M. Schofield, M.Burnyeat and J.Barnes (eds) Doubt and Dogmatism, Oxford, Clarendon Press, 1980:105–24.) While I am in total sympathy with McKim’s account of Democritus’ overall epistemological strategy, I am unwilling to follow him in attribution of the Epicurean theory to Democritus, since none of our evidence gives any support to the suggestion that Democritus gave that or any particular account of the truth of appearances. I agree that he probably held that, for the reason dramatized in the complaint of the senses, all appearances had to be in some sense or other true if there was to be any knowledge at all. But against McKim I hold that we have insufficient evidence to attribute to Democritus any account of the sense in which appearances are true, beyond the implicit claim that all appearances are equipollent. It is plausible to suppose that Epicurus’ account was devised in attempt to make good that deficiency. See also Furley [6.33]. 6 I am grateful to Gail Fine, David Furley and Robin Osborne for their comments on earlier drafts. BIBLIOGRAPHY Anaxagoras Texts 6.1 DK [2.2] II, sect. 59. 6.2 Mansfield [2.4] 6.3 Sider, D. The Fragments of Anaxagoras, Beiträge zur klassischen Philologie 118, Meisenheim am Glan, Anton Hain, 1981. Greek text of fragments with English translation and notes. Studies 6.4 Barnes [2.8], chs 16, 19(c). 6.5 Cornford, F.M. ‘Anaxagoras’ theory of matter’, Classical Quarterly 24 (1930): 14–30 and 83–95, repr. in Allen and Furley [2.15] II: 275–322 (page refs to [2.15]). 6.6 Dover, K.J. ‘The freedom of the intellectual in Greek society’, Talanta 7 (1976): 24–54, repr. in Dover, The Greeks and Their Legacy, (Collected Papers, vol. II), Oxford, Blackwell, 1988:135–58. 6.7 Furley, D.J. ‘Anaxagoras in response to Parmenides’, in Anton and Preus [2.17]: 70–92. 6.8 ——[2.31], ch. 6 6.9 Furth, M. ‘A “philosophical hero”?: Anaxagoras and the Eleatics’, Oxford Studies in Ancient Philosophy 9 (1991): 95–129. 6.10 Graham, D.W. ‘The postulates of Anaxagoras’, Apeiron 27 (1994): 77–121. 6.11 Guthrie [2.13] II, ch. 4. 6.12 Inwood, B. ‘Anaxagoras and infinite divisibility’, Illinois Classical Studies 11 (1986): 17–33. 6.13 Kerferd, G.B. ‘Anaxagoras and the concept of matter before Aristotle’, Bulletin of the John Rylands Library 52 (1969): 129–43, repr. in Mourelatos [2.19]: 489–503. 6.14 Lesher, J.H. ‘Mind’s knowledge and powers of control in Anaxagoras DK B12’, Phronesis 40 (1995): 125–42. 6.15 McKirahan [2.7], ch. 13. 6.16 Mann, W.E. ‘Anaxagoras and the homoiomerē’, Phronesis 25 (1980): 228–49. 6.17 Schofield, M. An Essay on Anaxagoras, Cambridge, Cambridge University Press, 1980. 6.18 Strang, C. ‘The physical theory of Anaxagoras’, Archiv für Geschichte der Philosophie 45 (1963): 101–18, repr. in Allen and Furley [2.15] II: 361–80. 6.19 Vlastos, G. ‘The physical theory of Anaxagoras’, Philosophical Review 59 (1950): 31–57, repr. in Allen and Furley [2.15] II: 323–53, in Mourelatos [2.19]: 459–88 and in Vlastos, ed. Graham (see [4.64]) I: 303–27 (page refs to [2.19]). The Atomists Texts 6.20 DK [2.2] II, sect. 67 (Leucippus), 68 (Democritus). 6.21 Luria, S. Democritea, Leningrad, Soviet Academy of Sciences, 1970. Original texts of fragments and testimonia with Russian translation and commentary. Collections of articles 6.22 Benakis, L. (ed.) Proceedings of the First International Conference on Democritus, Xanthi, International Democritean Federation, 1984. 6.23 Romano, F. (ed.) Democrito e I’atomismo antico. Atti del convegno intemazionale, (Siculorum Gymnasium 33.1), Catania, University of Catania, 1980. Studies 6.24 Barnes [2.8], ch. 17, 19(b), 20, 21(c), 23(d), 24(e). 6.25 ——‘Reason and necessity in Leucippus’, in Benakis [6.22] I: 141–58. 6.26 Bicknell, P. ‘The seat of the mind in Democritus’, Eranos 66 (1968): 10–23. 6.27 ——‘Democritus on precognition’, Revue des Études Grecques 82 (1969): 318–26. 6.28 Burkert, W. ‘Air-imprints or eidōla: Democritus’ aetiology of vision’, Illinois Classical Studies 2 (1977): 97–109. 6.29 Furley [4.63] Study 1 ‘Indivisible magnitudes’; ch. 6 ‘The atomists’ reply to the Eleatics’, repr. with emendations in Mourelatos [2.19]: 504–26. 6.30 —— ‘Aristotle and the atomists on infinity’, in I. Düring (ed.) Naturphilosophie bei Aristoteles und Theophrast, Proceedings of the Fourth Symposium Aristotelicum, Heidelberg, Lothar Stiehm Verlag, 1969:85–96, repr. in Furley [2.29]: 103–14. 6.31 ——‘Aristotle and the atomists on motion in a void’, in P.K.Machamer and J.Turnbull (eds) Motion and Time, Space and Matter, Columbus, Ohio, Ohio State University Press, 1976; 83–100, repr. in Furley [2.29]: 77–90. 6.32 ——[2.31], ch. 9–11. 6.33 ——‘Democritus and Epicurus on sensible qualities’, in J.Brunschwig and M.C.Nussbaum (eds) Passions and Perceptions, Proceedings of the Fifth Symposium Hellenisticum, Cambridge, Cambridge University Press, 1993: 72–94. 6.34 Guthrie [2.13] II, ch. 8. 6.35 Hussey, E. ‘Thucididean history and Democritean theory’, in P.Cartledge and F.Harvey (eds) Crux, Essays in Greek History presented to G.E.M. de Ste. Croix, London, Duckworth, 1985:118–38. 6.36 Kahn, C.H. ‘Democritus and the origins of moral psychology’, American Journal of Philology 106 (1985); 1–31. 6.37 Kline, A.D. and Matheson, C.A. ‘The logical impossibility of collision’, Philosophy 62 (1987): 509–15. Discussion by R.Godfrey ‘Democritus and the impossibility of collision’, Philosophy 65 (1990): 212–17. 6.38 Luria, S. ‘Die Infinitesimallehre der antiken Atomisten’, Quellen und Studien zur Geschichte der Mathematik B 2, 1933; 106–85. 6.39 McKim, R. ‘Democritus against scepticism: All sense-impressions are true’, in Benakis [6.22] I: 281–90. 6.40 Makin [2.52]. 6.41 O’Brien, D. Theories of Weight in the Ancient World, vol. I, Democritus: Weight and Size, Paris, Les Belles Lettres, and Leiden, E.J.Brill, 1981. Reviewed by D.J.Furley ‘Weight and motion in Democritus’ theory’, Oxford Studies in Ancient Philosophy 1 (1983): 193–209, repr. in Furley [2.32]; 91–102. 6.42 Procopé, J.F. ‘Democritus on politics and the care of the soul’, Classical Quarterly NS 39 (1989): 307–31; 40 (1990); 21–45. 6.43 Sassi, M.M. Le teorie della percezione in Democrito, Florence, La Nuova Italia, 1978. 6.44 Sedley, D. ‘Two conceptions of vacuum’, Phronesis 27 (1982): 175–93. 6.45 ——‘Sextus Empiricus and the atomist criteria of truth’, Elenchos 13 (1992): 19–56. 6.46 Taylor, C.C.W. ‘Pleasure, knowledge and sensation in Democritus’, Phronesis 12 (1967): 6–27. 6.47 Vlastos, G. ‘Ethics and physics in Democritus’, Philosophical Review 54 (1945): 578–92; 55 (1946): 53–64, repr. in Allen and Furley [2.15] II; 381–408 and in Vlastos, ed. Graham (see [4.64]), I: 328–50.

Routledge History of Philosophy. . 2005.