SOURCE: A. A. Long, "Epicurus and Epicureanism," in Hellenistic Philosophy: Stoics, Epicureans, Sceptics, University of California Press, 1974, pp. 14-74.

[Long offers a broad view of Epicurus's thought in the excerpt below, moving from biography and history to epistemology, and culminating with his ethical teachings.]

It has often been said that Epicurus was primarily a moralist, and if by this we mean someone who strives by theory and practice to advocate a particular way of life the description is appropriate. Epicurus thought that he could trace the causes of human unhappiness to mistaken beliefs in his society, beliefs about the gods, the destiny of the soul, and the objects in life which are truly valuable. Ultimately all his teaching has the aim of discrediting such beliefs and replacing them with those which he holds to be true. By his adherents Epicurus was regarded as a 'saviour', as the bringer of 'light', words which we naturally associate with Judaism and Christianity. But Epicurus was not a preacher, even if he sometimes preaches. He wished ardently to persuade, and to convince; it would be quite wrong to try to make him into a purely academic philosopher. But he was a philosopher. Arguments and evidence are the instruments by which he hoped to persuade those who would listen, and it is with the theory rather than the practical aspects of Epicureanism that I shall be concerned here. Beginning, after some introductory remarks, with Epicurus' theory of knowledge I propose to consider the details of his system in an order which seems to be both coherent and representative of his own methodology. Ethics proper is dealt with last, for other topics have ethical implications which can be noted en passant and moral conclusions are the ultimate goal of Epicurus' philosophy.

(i) Life and works

Epicurus was born on the island of Samos in 341 B.C. (D.L. [Diogenes Laertius] x 14). His father, who held Athenian citizenship, had settled there some ten years earlier. The first philosophical influence on Epicurus may have come in Samos itself from Pamphilus, a Platonist (Cic. N.D. [De natura deorum] i 72; D.L. x 14). But Epicurus' own philosophy is strikingly at odds with Platonism, and perhaps while still an adolescent he began an association with Nausiphanes on the neighbouring island of Teos (Herculaneum papyrus 1005) which nipped in the bud any positive allegiance to Plato. Nausiphanes was a Democritean (D.L. i 15; Cic. N.D. i 73), and it is likely that Epicurus first became acquainted with the basic principles of atomism through the teaching of Nausiphanes. In later life Epicurus denounced Nausiphanes in highly vitriolic language (D.L. x 7-8). It is not clear what prompted these attacks, but they are typical of Epicurus' attested attitudes towards other philosophers.

At the age of eighteen Epicurus went to Athens to do his two years of military and civilian service alongside the comic poet Menander (Strabo xiv 638). We know little in detail of his activities during the next fifteen years. He may have taught for some time as an elementary school teacher in Colophon, a small town to the north-west of Samos on the Persian mainland, where his family had now taken up residence (D.L. x 1; 4). Later he established his own philosophical circle first in Mytilene (on Lesbos) and then in Lampsacus (D.L. x 15), a port near the site of ancient Troy, returning to Athens at the age of thirty-four in 307/6. Here he re mained for the rest of his life. The return to Athens indicates that Epicurus was now confident of attracting followers in the main centre of philosophy. Between Athens and Piraeus Epicurus bought a house the garden of which came to stand as the name of the Epicurean school.

The community which Epicurus founded differed in important respects from the Academy and Lyceum. Its modern analogue is not a college or research institution but a society of friends living according to common principles, in retreat from civic life. Friendship has particular ethical significance in Epicureanism, and the Garden provided a setting for its realization. Women and slaves were admitted, and scraps of several private letters are preserved in which Epicurus expresses deep affection for his friends and followers. It is doubtful whether the Garden during Epicurus' lifetime offered much that might be called formal training to would-be Epicureans. Those who committed themselves to Epicurus were not so much students 'reading for a course' as men and women dedicated to a certain style of life. Seneca quotes the revealing maxim: 'Act always as if Epicurus is watching' (Ep. 25, 5). The similarity to George Orwell's 'Big brother is watching you' could scarcely be more misleading. Epicurus clearly inspired the strongest regard in his associates and personified the values of his own philosophy. But if the Garden lacked the formal curriculum of the Academy we can safely assume that its members devoted much time to reading and discussing Epicurus' books; his Principal doctrines (see below) were probably learnt by heart; some members must have been engaged in the preparation and copying of works both for internal consumption and for dissemination to Epicureans outside Athens; and Epicurus' chief adherents, such as Metrodorus, will have engaged in advanced study with the master himself.¹ Book xxviii of Epicurus' On Nature refers to Metrodorus in the second person, and the fragments which survive record parts of a discussion between the two philosophers on problems of language and theory of knowledge. Epicurus kept in touch with his followers outside Athens by correspondence, and the opening of his Letter to Pythocles is worth quoting for the attitudes it reveals of Epicurus himself and one of his disciples:

Cleon brought me a letter from you in which you continue to show good-will towards me matching my own love for you. You are trying not ineffectively to memorize the arguments which are directed at a life of sublime happiness, and you ask me to send you a brief summary of the argument about astronomical phenomena so that you can easily get it by heart. For you find my other writings difficult to remember even though, as you say, you are always using them. I was delighted to receive your request and it caused me joyous expectations.²

Consistent with these principles Epicurus preferred the company of a few intimates to popular acclaim (Sen. Ep. 7, 11). He did not however withdraw completely from civic life. In a letter cited by Philodemus Epicurus says that he has participated in all the national festivals (UsfenerJ 169); his slogan 'live quietly' was not a revolutionary denunciation of contemporary society but a prescription for attaining tranquillity. Opponents of Epicureanism vilified the founder as a libertine and voluptuary, but this is inconsistent both with his teaching on pleasure, as we shall see, and with his own professed attitudes. He claimed to derive great pleasure from a subsistence diet which cheese would turn into a feast (Us. 181f.). On his death in 271 B.C., Epicurus bequeathed his house and garden to his follower, Hermarchus, for the benefit of the Epicurean community, and succeeding heads of the school probably nominated their own successor. On the twentieth of every month Epicurus' memory and that of Metrodorus were celebrated at a festival within the Garden. This and other arrangements which are recorded in Epicurus' will (D.L. x 16-21) throw an interesting light on the character of the man himself.

Epicureanism has rightly been called 'the only missionary philosophy produced by the Greeks'.³ Before he took up residence at Athens, Epicurus had established a following in Lampsachus and Mytilene, and his disciples helped to propagate the Epicurean gospel throughout the Mediterranean world. Antioch and Alexandria are two major cities in which Epicureanism established itself at an early date. Later, it spread widely into Italy and Gaul. Cicero in the middle of the first century B.C. could write, and it gave him no pleasure to do so, 'The [Roman] Epicureans by their writings have seized the whole of Italy' (Tusc. iv 6-7). This was a time when Epicureanism briefly claimed the allegiance of some prominent Romans including Calpurnius Piso and Cassius. Julius Caesar may have been sympathetic and Cicero's Atticus was an Epicurean. The fortunes of the movement fluctuated. Political opposition was not unknown, but the main antagonists were first rival philosophers, especially Stoics, and later Christianity.

In the Roman world Epicureanism seems to have been at its strongest immediately before the fall of the Republic. But it suffered no sudden decline. Seneca quotes with approval many Epicurean moral maxims; Lucian's Alexander, written in the second century A.D., gives a fascinating account of Epicurean and Christian reactions to persecution in the area south of the Black Sea. And most remarkable of all, about A.D. 200 in the interior of modern Turkey, at a place called Oenoanda in antiquity, an old man named Diogenes had erected a huge philosophical inscription carved on a great stone wall. Between 1884 and the present day many fragments of his work have been recovered, and it constitutes a summary of Epicurus' teaching which Diogenes bestowed on his countrymen and humanity at large for their happiness.⁴ Apart from adding valuable information to our knowledge of Epicureanism, Diogenes' inscription proves the vitality of Epicurus' gospel five hundred years after the foundation of the Garden.

Epicurus himself was a prolific writer. Diogenes Laertius, who records forty-one titles of Epicurus' 'best books', says that his writings ran to three hundred rolls (x 26), and that he exceeded all previous writers 'in the number of his books'. Many of these consisted of short popular tracts and letters. Epicurus' major work was the series of thirty-seven books On Nature, a treatise On the criterion or kanôn, and a collection of ethical books which included On lives; On the goal; On choice and avoidance. He also wrote polemical works Against the physicists, Against the Megarians, and Against Theophrastus. Many of the letters, as we know from our own evidence, summarized points of doctrine or discussed these in some detail. Of all this writing only a small fraction has survived. Three letters are preserved which Diogenes Laertius included in his Life of Epicurus. The longest and most important of these, To Herodotus, gives a compressed and difficult summary of the main principles of atomism. Astronomical phenomena are the subject of the Letter to Pythocles, and the third letter, To Menoeceus, presents a clear if somewhat over-simplified account of Epicurean moral theory. In addition to these letters, Diogenes gives us a collection of forty Kuriai doxai, 'Principal doctrines', and a further set of maxims (Vaticanae sententiae) survives in a Vatican manuscript. Excavation at Herculaneum during the eighteenth century brought to light many charred rolls of papyrus which originally formed the library of some wealthy Roman. He was probably an adherent of Epicureanism, since most of the papyri which have been unrolled and read are fragmentary works by Philodemus of Gadara, an Epicurean philosopher and poet contemporary with Cicero. The rolls also contain fragments of some of the books of Epicurus On Nature. These are formidably difficult to read and reconstruct, but an invaluable supplement to earlier knowledge. Much work remains to be done on them.⁵

For our information about details of Epicurus' doctrine we are heavily dependent upon secondary sources. The most important of these is the Roman poet Lucretius, who wrote more than two hundred years after Epicurus' death. It is perhaps misleading to describe Lucretius as a secondary source. His poem, De rerum natura, is a work of genius which preceded the Aeneid and challenges it as a literary masterpiece. Lucretius, whose life and character are virtually unknown to us, was a fervid proponent of Epicureanism who presents Epicurus' teaching as the only source of human salvation. But Lucretius is no mere panegyrist. His six books set out in great detail Epicurean arguments concerning the basic constituents of things, the movement of atoms, the structure of body and mind, the causes and nature of sensation and thought, the development of human culture, and natural phenomena. At the same time, there is no reason to regard Lucretius himself as an original thinker. His work amplifies and explains points that we can find in Epicurus' own writings. Even where Lucretius reports theories, for instance the swerve of atoms (ii 216-93), which cannot be checked against Epicurus' own words, he was probably drawing on original sources which we cannot recover. Epicurus' own immediate successors were not noted for any major innovations. Certain refinements were doubtless made, and Philodemus' treatise On signs (preserved partially on papyrus) incorporates logical work by Zeno of Sidon (c. 150-70 B.C.) which may well go beyond anything worked out by Epicurus himself. But for the most part Epicurus' own writings remained canonical throughout the history of the school.

After Lucretius the best secondary sources are Diogenes Laertius, Cicero, Seneca and Plutarch. Cicero and Plutarch intensely disliked Epicureanism, and their criticism is of interest for understanding the adverse reception which the school often encountered. Seneca, though officially a Stoic, concludes most of his first Moral letters with an Epicurean maxim which he recommends to his correspondent, Lucilius. Sextus Empiricus, to whom Epicureanism was the most congenial of the dogmatic schools of philosophy, provides a useful supplement to our direct knowledge of Epicurean empiricism. Finally, as I have already mentioned, we have substantial fragments from the inscription of Diogenes of Oenoanda.

(ii) The scope of Epicurus' philosophy

Epicurus' philosophy is a strange mixture of hardheaded empiricism, speculative metaphysics and rules for the attainment of a tranquil life. There are links between these aspects of his thought, some of which are clearer than others. But one thing which certainly unites them is Epicurus' concern to set the evidence of immediate sensation and feeling against the kind of logical analysis which is characteristic of Platonic and Aristotelian methodology. Epicurus rejected many of the fundamental principles in terms of which Plato and Aristotle described the world. But more important than his disagreement concerning what is to be said about the world is his dismissal of certain logical and metaphysical concepts which are basic to Plato and Aristotle. Epicurus recognized the distinction between universal and particular; but he did not regard universals as having existence in their own right, like Plato; nor apparently was he interested, as Aristotle had been, in classifying things under genera and species. He did not set up principles such as Plato's same and different, or Aristotle's substrate and form, for the analysis of objects and their properties. Philosophers who proceed in this way, he held, are merely plying with words, setting up empty assumptions and arbitrary rules. He did not deny that philosophy uses language and logic as its tools (Us. 219). But he vehemently rejected the view that linguistic analysis by itself can tell us anything about the world which is true or relevant to a happy life. The value of words is to express those concepts which are clearly derived from sensations and feelings. These latter give us our only hold on facts and the only secure foundation for language.

One might suppose from this that Epicurus would have dispensed with metaphysics altogether. In fact, his account of what exists does not stop short at the objects of which we are made aware by immediate sensations and feelings. Our senses report to us things which we call sheep, grass, cats etc., but for Epicurus all such things are compounded out of atoms and void, neither of which is something that we can sense or feel. In asserting atoms and void to be the ultimate entities which constitute the world, Epicurus is making a metaphysical statement. This is not something which he can prove or verify directly from sensations with or without the help of experiment. He has to establish it by setting up certain axioms and assuming the validity of certain methods of inference.

The first atomist explanation of things was advanced more than a century before Epicurus began his philosophical career. Epicurus clearly believed it to be a theory for which he could offer new and improved proof. But while providing an elegant and economical answer to such questions as 'What is the structure of physical objects?' or 'How are bodies able to move?', the atomist theory attracted Epicurus on other than purely theoretical grounds. If all events and all substances are ultimately explicable by reference to atoms necessarily moving in empty space, both divine causation as popularly conceived and its sophisticated equivalents—Plato's Forms and Demiurge or World-Soul, Aristotle's Prime Mover and Heavenly Intelligences—become superfluous. Epicurus held that beliefs in divine management of the cosmos and of human destiny were a major cause of human failure to live a tranquil life. On an atomist analysis of the world, supposing this to be demonstrable, consequences would follow which could not fail to affect beliefs about a man's own place in the world.

Epicurus often asserts that philosophy has no value unless it helps men to attain happiness. This applies with particular force to his moral theory, but there is no necessary connexion between atomism and hedonism. The claim that pleasure is the only thing which is good as an end is compatible with all manner of metaphysical hypotheses. Epicurus has various ways of establishing his hedonism, none of which draws direct support from atoms and void. In this he differs markedly from the Stoics whose moral theory is intrinsically related to their metaphysics. But Epicurus thought he could show the validity of hedonism by appeal to immediate experience which, less directly, he held to support atomism. If labels can be usefully applied to a philosopher, Epicurus should be called an empiricist. That at least is what he would like to be remembered as, and empiricism provides the clearest internal connexion between his different ideas.

(iii) Theory of knowledge

If you fight against all sensations, you will have nothing by reference to which you can judge even those which you say are deceptive (K.D. [Kuriai doxai/Principal doctrines] xxiii).

The foundation of Epicurus' theory of knowledge is sense-perception. He starts from the fact that all men have sensations (aisthêseis), and asserts, without proof, that these must be caused by something other than themselves (D.L. x 31). It does not of course follow from this assertion that sensations are caused by things external to the percipient, and Epicurus would acknowledge that a feeling such as hunger (a pathos in his terminology) has an internal cause. But he takes it as self-evident that sensations of colour, sound, smell etc. must be caused by actual objects which possess these properties. 'We must suppose that it is when something enters us from things which are external that we perceive … their shapes' (Ep. Hdt. [Letter to Herodotus] 49). This statement at once raises questions which the Sceptics did not hesitate to ask about mirages, hallucinations and the like. But Epicurus has an answer to put forward, as we shall see later.

Suppose we accept that sensations cannot lie concerning their causes: in other words, that if I have the sensation of hearing there must be something sounding which causes my sensation. Does this support the further proposition that there is some object like a motorcar horn or a train whistle which corresponds precisely to the content of my sensation? For Epicurus the inference may or may not be warranted. That about which our sensations cannot deceive us is not a motor-car horn but a sense-impression (phantasia). What enters me from things outside is not a motor-car horn, if that is what I do genuinely hear, but a cluster of atoms (eidôla) thrown off the outer surface of such objects. Provided that these 'effluences', as we may call them, enter the sense organ without experiencing any change of structure the impression they produce on us will be an accurate image of the object.⁶ If on the other hand their structure is disrupted in transit, the effluences will cause us to sense something which corresponds not to some actual characteristic of the object itself but to their own modified structure.

Sensations therefore are necessarily good evidence only of effluences. This raises the problem of how we can distinguish between those sensations which report to us accurately about objects and those which do not. For we cannot get at objects independently of effluences. Epicurus tackles this problem in an interesting way. He distinguishes sharply between the sense-impressions itself and judgments, or the identification of sense-impressions with objects (Ep. Hdt. 50-1). Our sense-impressions are not judgments, nor are they dependent upon reason. We are not to say that this senseimpression is reliable, that one untrustworthy, for to do so presupposes an object which can test the validity of sensation, and our sole knowledge of objects is derived from sensations. Considered as an item of information about that which affects our senses every impression is of equal validity (D.L. x 31-2).

Nevertheless, sense-impressions can be distinguished from one another in terms of clarity or vividness. Sounds may be sharp or faint, visual images both clear and blurred. Epicurus was also aware of the fact that as we move away from the apparent source of many sensations our impressions change, and may decrease in clarity. Putting these facts together he concluded that sensations provide reliable evidence about objects if and only if they are characterized by clear and distinct impressions (enargeia, Ep. Hdt. 52, cf. K.D. xxiv). Other impressions 'await confirmation' by those which are clear. This conclusion could also seem to derive some support from Epicurus' explanation of the physical processes by which sensation takes place.

If we are near the ultimate source of our sensations the effluences which affect us are less likely to encounter disruption. It is only from a distance, supposedly, that the tower which is square looks round (Us. 247).

Epicurus does not specify conditions which establish the clarity of a sense-impression. He probably regarded this as something which would entail an infinite regress. He could take it as a datum of experience that we do distinguish within limits between that which is clear and that which is blurred or obscure. Clarity however is not a sufficient guarantee that we see things as they really are. Epicurus was grossly misled by 'clear views' when he argued that the sun is about the same size as it is seen to be (Ep. Pyth. [Letter to Pythocles] 91).

Close attention to clear impressions is the first stage in acquiring knowledge. But Epicurus did not regard it as sufficient by itself. However clear our sense-impressions may be they do not constitute knowledge. They do not tell us what something is. Before judgments about objects can be made, our sense-impressions must be classified, labelled and so marked off from one another. Epicurus proposed to satisfy these conditions by what he called prolêpseis, 'Preconceptions'⁷. These are general concepts or mental pictures produced by repeated sense-impressions which are both clear and similar in kind. They persist after particular sensations cease and constitute a record of our experience of the world. We acquire a concept or prolêpsis of man by repeated and remembered experience of particular men. Hence we are able to interpret new sensations by comparing them with preconceptions, and all our judgments about objects are made on this basis of recorded experiences, which we classify by using language (D.L. x 33). Epicurus agreed broadly with Aristotle who asserted that 'science comes to be when out of many ideas born of experience a general concept which is universal arises concerning things that are similar' (Met. [Metaphysics] A 981a5 ff.). For Epicurus, preconceptions are the foundations of judgments and language. "We should not have named anything unless we had previously learnt its form by a preconception' (D.L. ibid.). Language is a method of signifying those preconceptions which seem to us to fit the present object of experience. Because preconceptions themselves are supposed to possess 'clarity', they establish, in association with the appropriate new sense-impressions, what it is that we see, hear and so on.⁸ Error arises when we use words which signify a preconception that does not correspond with the phenomenon (De nat. [De natura/On Nature] xxviii fr. iv col. 3). This may happen through confusing unclear with clear impressions, and Epicurus also recognized that the ambiguity of many words can be a cause of misassociating senseimpressions and preconceptions.⁹

Epicurus probably thought that all other concepts, in cluding those which have no empirical reference, are derived from preconceptions. Preconceptions can be combined with one another, or they can be used as basis for inference (see D.L. x 32). But, with a few exceptions, preconceptions seem to be direct derivatives of sensation, and Epicurus recommended that the meaning of words should always be established by reference to 'the first mental image' (Ep. Hdt. 37). In this way he hoped to forge a firm bond between statements and immediate experience, though he gave no sufficient reasons why people's preconceptions should be regarded as similar and therefore identifiable by the same words.

So far Epicurus can claim to be a rigorous empiricist. But we must now note a number of curious exceptions to the principle that our ideas about the world are all derived ultimately from sense-impressions. Apart from those effluences which cause our sense-impressions, Epicurus also supposed that there are 'images' which somehow bypass the sense organs and penetrate directly to the mind. In nature these too are atomic clusters, but their density is much finer than the effluences which affect our senses. They are tenuia simulacra, as Lucretius called them (iv 722ff), and account both for dream-images, phantoms, visions of the dead, and for such ordinary objects of thought as lions. Such 'images' may be direct effluences from the surface of an object. But many of them are simply chance combinations of individual atoms; others may consist of real effluences which are compounded and then produce images of Centaurs and monsters (Ep. Hdt. 48; Lucret. iv 130-42). Instead of accounting for dreams and hallucinations by reference to images entirely created or brought to consciousness by some psychological faculty, Epicurus supposed that dreams and hallucinations too are explicable by the mind's contact with atoms that enter it from outside.

If we ask how dreams and visions are to be distinguished from sense impressions, the answer is not entirely clear. Lucretius looks for a distinction in terms of continuity. 'Real' sense-impressions are produced by a steady stream of effluences; but the mind can be moved by a single 'image' (iv 746), and thus presumably catch a momentary vision. Also, dream-images are said to move by a series of effluences perishing one after another; Lucretius' description makes one think of the staccato movement of early cinema. Clearly such criteria are inadequate for Epicurus' purpose. In effect he is saying that a hallucinated person really does see something which is there, but mistakenly takes it to correspond with an actual solid object (cf. Us. 253).

The gods are a further object of direct mental perception. Postponing for the present consideration of their physical structure, we may observe here that Epicurus posits a series of fine effluences from the gods which penetrate directly to the mind. The texts which describe these divine 'images' are difficult; Epicurus put forward theoretical reasons, as well as the evidence of such visions, to justify his claims about the divine nature (Cic. N.D. i 43-55). But he seems to have given no adequate arguments in favour of divine 'images'. It is no help to this queer thesis to invoke the supposed universality of human belief in gods. The real difficulty is however that of the grounds for verification. By the concept of 'clear' view Epicurus has a standard for verifying perceptual judgments which has some claims to be called objective. Only in a special philosophical sense could people's perception of dogs be said to depend on their beliefs. How we may conceive of gods, on the other hand, is something which cannot be assimilated to perception of empirical objects. Epicurus' theory of divine 'images' puts religious belief in the same category as empirical observation.

Some scholars have argued that Epicurus posited a special mental faculty, 'apprehension by intellect' (epibolê tês dianoias), which somehow guarantees the veracity both of impressions of the gods and the validity of scientific concepts. If Epicurus had held such a theory he would have been an intuitionist in much of his philosophical activity. This interpretation was defended at great length by Cyril Bailey, whose work on Epicurus and Lucretius has held authority in the English-speaking world. According to Bailey, Epicurus supposed that the 'clear' sense-impression, of which I have already spoken, is obtained by the 'attention of the senses', and correspondingly, clear visions of the gods and clear concepts concerning, for instance, atoms and void, are obtained by the 'attention' of the mind. If Bailey were merely arguing that we cannot be aware of any object or thought unless we 'attend' to something, he would be ascribing nothing remarkable to Epicurus. But Bailey meant much more than this. On his interpretation, Epicurus supposes that 'the concepts of science are built up step by step by the juxtaposition of previous concepts, each in their turn grasped as "clear" … by the immediate apprehension of the mind'.¹⁰

Epicurus' use of the expression 'apprehension by intellect' does not justify Bailey's view. Any explanation of Epicurus as an intuitionist is on quite the wrong tack. Probably what he means by 'apprehension', whether by the mind or the senses, is concentration or attention: we need to concentrate, if we are to grasp the images which can be received by the sense organs or the mind. I shall return to this subject in the discussion of Epicurean psychology.

In order to use the evidence of the senses as material for establishing true propositions about the world, Epicurus assumed the validity of certain axioms. One of these has been stated already: 'Sense-impressions which are "clear" provide accurate information about the external appearance and properties of objects.'¹¹ These sense-impressions confirm or bear witness against the truth of judgments about objects which we may make provisionally on evidence lacking the requisite clarity. But Epicurus also allowed a weaker form of confirmation, 'lack of contrary evidence'. And this we may state as a further axiom: 'Judgments about non-evident objects are true if they are consistent with clear sense-impressions.' This second axiom is of the utmost importance to Epicurus. If positive confirmation by clear impressions were the sole ground for true objective statements, Epicurus would be unable to advance beyond the description of sensible objects. As it is, he assumes that the validity of clear impressions is such that they provide indirect evidence of things which are imperceptible, or for which a clear view is unobtainable in the nature of the case. Here is an example preserved by Sextus Empiricus (cited n. 21): If void does not exist (something non-evident) then motion should not exist since ex hypothesi all things are full and dense; so that since motion does exist the apparent does not contradict the judgment about that which is non-evident. (As stated, of course, the argument is invalid since it assumes that void is a necessary condition of movement, having fullness as its contradictory.)

Epicurus associates the axiom concerning non-contradiction with a further proposition (X), which may be treated as an implication of the second axiom: If more than one explanation of non-evident phenomena is consistent with observation then all such explanations are to be treated as equally valid (Ep. Pyth. 87). Let us state this more formally. Suppose that p is an evident fact, q a problem requiring explanation which cannot be solved directly by reference to p, and s, t, u three different statements about q which are all consistent with p. Then, independently of any other criterion or axiom, it follows from proposition X that s, t, and u are all equally acceptable explanations. This argument is formally valid, and Epicurus applied it rigorously to all statements concerning astronomical phenomena. In order to reject any of the hypotheses s, t and u it would be necessary to introduce some further principle of verification over and above the second axiom. Epicurus declined to do this on the grounds that it would be a departure from that which is definitely knowable, that is, the state of objects as given by clear sense-impressions.

Epicurus applies this principle regularly in the Letter to Pythocles, as the following excerpt shows (94):

The repeated waning and waxing of the moon may come about owing to the rotation of this celestial body; equally it may be due to configurations of air; or again by reason of the interposition of other bodies; it may happen in any of the ways in which things manifest to us invite us to account for this phenomenon, provided that one does not become so attached to a single explanation that one rules out others for no good reason, failing to consider what it is possible for a man to observe and what impossible, and therefore desiring to discover the indiscoverable.

Up to a point this is admirable as a scientific principle. One thinks of the current debate concerning different explanations of the origin of the universe (big bang, steady state etc.) which has not yet been resolved by empirical data. Epicurus could argue, with considerable justification, that the astronomy of his own day claimed to know more than its source of evidence, the naked eye, justified. What he seems to have wholly failed to appreciate is the valid check on immediate observations which some astronomers were already trying to make by reference to systematic records and by mathematical calculations.

In application to celestial phenomena, Epicurus' use of the axiom of non-contradiction has the largely negative function of leaving open a plurality of possible explanations. But the Epicureans used the principle positively as grounds to support general statements arrived at by induction. Philodemus records this example: from the proposition 'Men in our experience are mortal' we infer that 'men everywhere are mortal'. The general statement is based on the empirical fact that we know no exception to it and therefore it is consistent with experience (On signs col. xvi). The Stoics objected to this kind of reasoning on the grounds that it presupposes the non-evident (unobserved men) to be similar in kind to the evident. The Epicureans replied that their inference does not make a presupposition that all men are mortal. It is the absence of any man known to be immortal which justifies the general inference about human mortality.

Epicurus assumes, as any scientist must, that there are certain uniformaties in nature which hold for what is evident and non-evident alike. Of course, it does not follow by the second axiom that a proposition which is consistent with some evident phenomenon must also be true concerning something non-evident. But science cannot operate merely by propositions which are necessarily true. It must proceed by empirical generalizations which are rejected as and when new evidence refutes previous hypotheses. The Stoics held that all inferences must be established by arguments which are deductively valid. But deductive reasoning by itself can never be sufficient to establish a scientific statement. For the premises which entail a deductive conclusion about observable data must either be empirical generalizations, or be ultimately based upon statements of this form. At some point the scientist must make an inductive inference on the basis of evidence, and for Epicurus that point is reached when observation seems to support the belief that no instances are likely to be found which will contradict a general statement.

Epicurus and Lucretius often appeal to 'analogy' or 'similarity' to support an inference from the visible to the invisible. Thus Epicurus takes the (allegedly) observed fact that no parts can be distinguished in the smallest visible magnitude to support the inference that the same is true of the smallest invisible magnitude, the minimum part of an atom (Ep. Hdt. 58f). The main subject of Philodemus' On signs is 'analogical inference', and we might suppose that this requires a further axiom for its justification. None of our secondary sources nor Epicurus himself gives any independent discussion of 'analogy', and it is not needed. The justification for inference by analogy is provided by the two axioms already discussed.¹² By the first axiom we are justified in asserting that 'clear impressions' give evidence that men are mortal. All the men of whom we have reliable experience are similar in respect of mortality. From this positive evidence Epicurus infers that men of whom we have no experience are equally similar in respect of mortality. The inference is justified by the second axiom which allows us to assert p if there is no evidence against it. Philodemus even states that it is 'inconceivable' that there should be something which possesses nothing in common with empirical evidence (On signs col. xxi 27ff). The Epicurean test of the 'conceivable' is sense-perception and the problematic mental 'images' already discussed.

Epicurus' method of indirect proof can be illustrated by copious passages from Lucretius. It is the poet's regular practice to refute a proposition by appeal to what is clear, what we actually see, and thus infer the contradictory of the rejected proposition. I give just one example. In Book i Lucretius argues towards the atomist thesis by stages. He begins by dismissing the proposition 'Something can be created from nothing'. For, if this were so, every type of thing could be produced out of anything; nothing would require a seed. But we see (videmus) that cultivated land produces better yield than uncultivated land, which proves the existence in the soil of primary bodies stirred to birth by the plough. 'If there were no primary bodies you would see (videres) each thing coming to birth much more successfully of its own accord' (159-214).

Like this Lucretian argument Epicurus' methodology seems imprecise and informal when judged by the criteria of stricter logic. I have no doubt that it is proper to describe as 'axioms' the two principles concerning confirmation and non-contradiction. But Epicurus does not call them axioms. He almost certainly knew Aristotle's Analytics, if a fragment of Philodemus has been correctly deciphered.¹³ But although Aristotle's inductive methodology may have been an influence on Epicurus, the later philosopher did not share Aristotle's interest in logic for its own sake; and he seems to have thought that any kind of demonstrative science, based upon deductive reasoning, was mere word-play. Since most of Aristotle's Analytics is concerned with the analysis of deductive argument in the form of syllogisms and with specifying the sufficient conditions of necessary truths, Epicurus cannot have liked what he read. Above all, he rejected any kind of logical inquiry which was not applied to the understanding of empirical data. He did not, apparently, see that empirical science, if it is to be well-grounded, cannot advance very far solely on the guidance of 'clear' sense-impressions.

The reader will be able to extend the criticism of Epicurus' methodology for himself. What should be added is a warning against taking at face value offhand remarks in ancient writers, which would imply that Epicurus had no interest in logic and scientific method. These views are to be found in many modern hand-books and they are incorrect. Epicurus, in order to shock, sometimes writes as if he despised all learning; but this is rhetoric, an expression of contempt for what he regarded as pedantic and positively harmful in the culture of his own day. Fortunately, sufficient of the twenty-eighth book of On Nature survives to give us a glimpse of Epicurus when he is not merely summarizing or exhorting. In this work Epicurus discussed induction, using Aristotle's technical term epagôgê, problems of meaning and ambiguity, the distinction between universal and particular, problems connected with the designation of individuals, and linguistic puzzles of the sort propounded by the Megarians. Unfortunately the text is too badly damaged to let us see how he treated all of these subjects in detail.¹⁴ But it gives us sufficient evidence to judge that parts of the following statement by Cicero are grossly misleading: 'Epicurus rejects definitions [he did not]; gives no instruction concerning division [classification into genus and species]; fails to show how an argument is to be constructed; does not point out how sophisms are to be resolved nor how ambiguities are to be distinguished' (Fin. [Definibus] i 22). It is salutary to remember that Epicurus wrote thirty-seven books On Nature, and that we can observe Cicero's prejudice by studying fragments from just one of these.

(iv) The structure of things

The nature of the universe is bodies and void (On Nature i).

Epicurus claimed to be self-taught, but atomism was more than a century old in Greece by the time he reasserted its central principles. First Leucippus and then Democritus, in the second half of the fifth century, had argued that what really exists is ultimately reducible to two and only two kinds of thing: the full (indivisible bodies) and the empty (space). How such a thesis came to be propounded is itself a fascinating story, but it belongs to the history of Presocratic philosophy. I will touch on it here only in so far as it is essential to understanding Epicurus' atomism.

Our starting-point is once again the Letter to Herodotus. There, in a few paragraphs of highly succinct argumentation, Epicurus discloses the essential features of the atomist theory (38-44). The problem which that theory purports to solve may be stated as follows: What principles derived from empirical evidence are necessary and sufficient to account for the physical world as it presents itself to our senses? The answer is highly economical: an infinite number of indivisible bodies moving in infinite empty space.

Epicurus arrives at this answer by a series of metaphysical propositions, which he then uses to support inferences about the underlying structure of the changing objects of experience. (A) Nothing can come out of nothing. (B) Nothing can be destroyed into nothing. (C) The universe never was nor will be in a condition which differs from its present one.¹⁵ The first two propositions are established indirectly by what I have called the second axiom. It would controvert experience to suppose that nothing pre-exists or survives the objects which we observe to grow and decay. Things are seen to grow out of something; they do not just emerge at random. Secondly, there is something into which things pass away. Otherwise there would be no limit to destruction and everything would have perished into non-being. The third proposition (C) is treated by Epicurus as analytic. Since the universe embraces all that there is, nothing exists outside the universe which could cause it to change. (He does not consider the possibility that any internal cause of change might bring about different conditions of the universe as a whole at different times.)

From (C) it follows that any explanation of things in general which holds now is eternally valid.

It is an evident fact that bodies exist; empty space must therefore also exist, since bodies must be in something and have something through which to move.¹⁶ Epicurus next asserts that apart from bodies and void nothing can be conceived of as an independent entity: all things must be reducible to body and mind. Bodies are of two kinds, compounds and the units out of which compounds are formed. From (B) it follows that one class of bodies, non-compounds, must be limited with respect to change and destruction. Epicurus expresses this thus: 'And these bodies [sc. non-compounds] are indivisible and changeless, if all things are not to be destroyed into non-being but are to persist secure in the dissolution of compounds; they are solid in nature and cannot be divided at any place or in any manner. Hence the first principles must be bodies which are indivisible' (Ep. Hdt. 39-41).

Question-begging though this argument is, in Epicurus' abbreviated formulation, its main points are wholly clear. We do not see atoms, but what we see, birth and death, growth and decay, is taken to require the existence of bodies which are themselves changeless and wholly impenetrable.

What else is to be said about atoms and empty space? Epicurus proceeds to argue that the universe is 'unlimited' in itself and also in the number of atoms which it contains, and in the extent of empty space. If the universe were limited it would have extremities; but there is nothing to limit the universe. And if the universe itself is unlimited its constituents must also be unlimited. For a limited number of atoms in infinite empty space would not be sufficient to hold one another together; they could not form the plurality of compounds which we experience; and an unlimited number of atoms could not be accommodated in limited space (Ep. Hdt. 41-2).

Since all the objects of experience are compounded out of atoms and void, Epicurus held that the atoms themselves must have innumerable, though not infinitely, different shapes, in order to account for the variety of things. Besides shape, all atoms are necessarily subject to continuous movement, a fact which will require further discussion shortly. They also possess weight, and of course bulk or mass. All other properties of which we have experience are accounted for by the arrangements which come into being when a plurality of atoms and void combine. Atoms as such are not hot or cold, coloured or resonant, and so forth (Ep. Hdt. 42-4; 68-9).

It is now time to consider in more detail what Epicurus meant by the 'indivisibility' of the atom. As has been shown already, the concept of an atom is arrived at by elimination of a contradictory hypothesis, that bodies are ultimately divisible into non-being. If Epicurus supposed that the infinite divisibility of a body must lead to its reduction to nothing at all he was guilty of an elementary fallacy. Infinite divisibility implies nothing about reduction to sheer non-existence. Lucretius however sets out an argument for indivisible bodies which avoids this fallacy, and which, we may presume, goes back to Epicurus himself.

The argument runs thus: body and empty space are mutually exclusive, otherwise there would not be two kinds of real things (i.e. all things could be reduced to either body or empty space). Body therefore cannot include as part of itself empty space. Anything which does include as part of itself empty space must be bounded by that which is solid—body. Created things are of this kind, that is, compounds of body and empty space. But the particular bodies which help to form such compounds must consist of that which is wholly solid and indivisible. For nothing can be divided unless it contains within itself empty space. And nothing can contain within itself empty space unless it has components which are themselves wholly indivisible (Lucret. i 503-35).

The force of this argument turns on the assumption that empty space is a necessary condition of divisibility. The earlier atomists had spoken of empty space as 'non-being' and of body as 'being': empty space is non-body. If we take Epicurus' reference to non-being as a legacy of this earlier usage his argument becomes compatible with the passage summarized from Lucretius. Just as what-is cannot be ultimately reduced to what-is-not, so body cannot be reduced to non-body, i.e. empty space.

The Epicurean atoms cannot be split into smaller bodies. They are physically indivisible. But they are not the smallest units of extension. The atom itself consists of minimal parts which are not merely physically unsplittable but indivisible in thought: nothing beyond these minima can be conceived of. Epicurus supposed that there is a finite number of such minimal parts for every atom. Atoms vary in size, and their size is determined by the number of their minimal parts. Again, atoms vary in shape and their shape is determined by the arrangement of their minimal parts.

This doctrine of minimal parts raises many difficulties and will only be discussed briefly here. Epicurus apparently regarded each atom as something composed of minimum units of magnitude which are not separable from one another, and therefore not separable from the whole atom which they compose (Ep. Hdt. 56-9).¹⁷ The notion is obscure and may be clarified by a concrete analogy. Suppose we take a cubic centimetre of solid metal and mark it off in three dimensions by millimetres. Then, the minimal parts stand to the atom as each cubic millimetre to the whole cube of metal, with this proviso: each square millimetre must be taken as the smallest unit which can be distinguished on any surface of the whole cube. The Epicurean atom is the smallest magnitude which can exist as a discrete independent body. Epicurus seems to have thought that this left something to be accounted for—namely, the atom's boundary points, or the fact, to put it another way, that the atom is a three-dimensional object and as such possesses shape. He sought an explanation by reference to its having minimal parts.

In giving the atoms minimal parts Epicurus almost certainly modified earlier atomism. The atoms of Leucippus and Democritus were physically indivisible but also, in all probability, without parts, and therefore theoretically indivisible as well. Simplicius, the Aristotelian commentator, distinguishes Epicurus from the earlier atomists thus: Epicurus, he says, appealed merely to the changelessness of his primary bodies, whereas Leucippus and Democritus also referred to their smallness and lack of parts (Us. 268).

Our knowledge of early atomism is largely derived from Aristotle. Modern research has shown that he was right to connect the fifth-century atomists with the slightly older Eleatic philosophers, Parmenides and Zeno.¹⁸ This is not the place to offer any detailed account of the Eleatics. Very briefly, Parmenides, in his poem The way of truth, had set out arguments of quite remarkable subtlety concerning what can be said about that which exists. He concluded that the following predicates are inadmissible: subject to creation and destruction, subject to divisibility, subject to change of place, and change of quality; what exists is 'whole, immobile, eternal, all together, one and continuous'. Zeno reinforced Parmenides' arguments by seeking to show that the proposition 'Things are many and subject to motion' leads to insoluble dilemmas. Zeno's puzzles turn chiefly on the notion of partition or divisibility. The interpretation of them is extremely difficult and controversial. One point only concerns us here. Zeno argued that if a unit of magnitude can be divided at all it must be infinitely divisible. This conclusion was unpalatable to the early atomists who wished to give an explanation of the world which would be compatible, as far as possible, with Parmenides' logic. Hence they adopted as their primary bodies partless, and therefore indivisible, units of magnitude, which satisfied most of the predicates deduced by Parmenides to belong to what really exists. Lacking parts, the atoms of Democritus would not even be theoretically divisible.

Epicurus modified this doctrine by ascribing parts to the atom but making these parts themselves minima, i.e. physically and theoretically indivisible. He seems to have supposed that it is a necessary condition of the atom itself, the minimum discrete body, that it possesses parts—but parts which rule out theoretical divisibility to infinity for the atom itself. Minimal parts satisfied this condition, and they also provided a means of accounting for differences in the shape, size and weight of particular atoms. Lucretius asserts that 'things which are not augmented by parts do not have the diversities of properties which creative matter must have' (i 631-3). Finally, Aristotle had pointed out a number of difficulties in the notion of a partless atom which Epicurus' new thesis may have been intended to resolve (Physics vi 231b25-232al7; 240b8-241a6). We shall come back to minimal parts in the next section.

(v) The motion of atoms and formation of compound bodies

One modification of earlier atomism by Epicurus has just been discussed. He also differed from Democritus concerning the motion of atoms. Both agreed that the atoms are always in motion, but Democritus almost certainly supposed that the course which any one atom takes relative to any other is wholly random.¹⁹ In fact, most of our evidence for the motion of Democritus' atoms clearly refers to motion derived from collisions with other atoms, and we can only speculate about what he might have said of the original motion of an atom. He probably did not attribute weight to the atom, and if so, cannot have used weight as a cause of motion. Furthermore, it is unlikely that he would have thought it proper to ascribe any direction to the movement of atoms since he said of infinite void that it has neither top nor bottom, centre or extremity (Cic. Fin. i 17). Void provided Democritus with a condition which he certainly regarded as necessary for motion, and he may have simply taken it as a fact about the atom that it does necessarily move in the void.

Against Democritus Epicurus held that weight is a necessary property of the atom. His reasons for this assertion cannot be established by any direct testimony, but they are almost certainly founded on the hypothesis that a weightless body cannot move. Aristotle devoted considerable attention to the analysis of weight as a determining factor of motion. Indeed, in the De caelo Aristotle defines heavy and light as 'the capacity for a certain natural motion' (307b32), and a little later says: 'There are certain things whose nature it is always to move away from the centre, and others always towards the centre. The first I speak of as moving upwards, the second downwards' (308al4). Now Epicurus recognized that in an infinite universe one cannot strictly speak of a centre, nor of up and down (Ep. Hdt. 60). But the thought that one could speak of up and down relative to some fixed point, and that in this relative sense the natural motion of atoms is downwards as a consequence of their weight. Any other motion than perpendicular fall requires other factors than weight alone to account for it. In all probability, Aristotle's discussion of weight as a determinant of movement influenced Epicurus' modification of Democritus.

If an atom is unimpeded by collision with other atoms its speed and direction of motion are invariant. Epicurus grasped the important fact that differences of weight make no difference to the velocity of bodies falling in a vacum: 'When they are moving through the void and encounter no resistance the speed of the atoms must be equal. Neither will heavy ones travel faster than light ones …' (Ep. Hdt. 61). At what speed then do free-falling atoms move? Epicurus sometimes expresses this in the graphic phrase 'as quick as thought'. But that does not help us very much. In fact, he seems to have supposed that time, like extension, is not infinitely divisible. Just as the atom is the minimum discrete body and consists of minimal parts, so time is divisible into 'minimum continuous periods' which themselves consist of indivisible temporal units, 'times distinguishable only in thought' (Ep. Hdt. 62). Epicurus probably supposed that the time which an atom takes to move the minimum distance, that is the minimum of extension, is the minimum temporal unit. This temporal unit, being indivisible, is not such that a movement can take place during it. 'Has moved' not 'is moving' is the relation which a moving body has to the indivisible units that constitute time and space (Us. 278). As Aristotle saw, such a theory turns movement into a series of jerks (Phys. vi 231b25—232al7). The atom has to hop, as it were, from one set of spatial units into the next. For there is no time or space in which its progression from one unit into the next can be said to occur. The minimum distance by which an atom can alter its location at any moment is the measure of any one minimal part.

Epicurus could have avoided these consequences if he had seen that the infinite divisibility of any quantum, whether of space or time, can be asserted without its entailing the consequence that these quanta are, as a matter of fact, divisible into infinite parts. He chose instead to predicate physical indivisibility of the atom and a limit to the theoretical division of its parts. Having wrongly concluded that the extremities of the atom must be accounted for by positing minimal spatial units, he accepted as a corollary the indivisibility of time and movement as well. He countered Aristotle's objection, that this makes differences of speed impossible, by arguing that apparent differences in the speed of the compound bodies which we see can be explained as a function of the collective movements at constant speed of individual atoms within each compound; the speed of a moving compound is determined by the collisions which occur between its internal atoms. The more its atoms tend to move in the same direction over a short period of time the greater the speed of the compound body. If the movement of some atoms in one direction is balanced by a movement of others in a different direction the compound body will be stationary (Ep. Hdt. 47; 62).²⁰

Let us now return to consider the falling atom in more detail. Given that this fall is at constant speed for any atom, and in the same direction, how can a world be formed which consists of atoms in conjunction, atoms which have collided and formed compound bodies? Oddly enough, no word of the answer to this problem survives by Epicurus himself. But his theory can be reconstructed from Lucretius and other later writers. Lucretius writes as follows: 'At this point I wish you to learn this too: when bodies are being carried straight down through the void by their own weight, at an undetermined time and at undetermined places they push a little from their course—only as much as you might call a change of direction. If they were not accustomed to swerve, all things would fall down through the deep void like raindrops, nor could collision come about nor would the atoms experience blows. And so nature would never have created anything' (ii. 216-24).

This swerve which atoms make at no determined time or place has always intrigued the readers of Lucretius. It apparently builds into the universe, as Epicurus conceives of this, a principle of relative indeterminacy. The movements of an atom, and therefore any consequences of its movement, are not entirely predictable. Our further information about the swerve in this context merely confirms Lucretius' words, and they must be taken at face value.²¹ It follows then that an atom, independently of any secondary motion which may result from collisions, has both a unidirectional movement and an unpredictable tendency to deviate from this.

The atomic swerve is also important to Epicurus' theory of human action. But that will require discussion later. The effect of the swerve with which we are concerned here is the collision between two or more atoms which it may bring about. Since every atom is solid through and through, the effect of a collision between atoms is a momentary check on atomic movement followed by a rebound (at the same speed), and hence a further change of direction. But it may sometimes happen that colliding atoms in spite of their tendency to rebound become intertwined and form a temporary and apparently stable compound. The compound so formed is in fact a dynamic entity, a collection of atoms moving both in their normal downward manner and from the effect of blows or swerves. But it will often present the appearance of something stable. Lucretius reports that the different densities of objects are determined by the relation between the atoms and void which they contain. Iron consists of close-packed atoms which are unable to move and rebound any great distance. Air, on the other hand, is composed of atoms which are interspersed with large areas of void (ii 100-8).

We have seen that for Epicurus all properties of things beyond size, shape, weight and movement are secondary. That is to say, they are properties which cannot be predicated of atoms but only of the compound bodies which atoms may form. This does not mean that colour, sound etc. are merely human ways of ordering and interpreting sense-impressions. Democritus had argued thus, but Epicurus did not agree (Ep. Hdt. 68-71). His discussion of secondary qualities is condensed and obscure, but seems basically to amount to this: colour, sound etc. cannot exist independently of bodies, nor are they 'parts' out of which compound bodies arise. Rather, any secondary property which is a permanent attribute of some object (compound body) is a constituent of the object in the sense that the object would not be what it is without this attribute. We might illustrate this point by saying that a man does not arise out of a combination of hands, legs, colour and so forth. He arises from a combination of atoms and void. But the effect of this combination is the production of hands, legs, pink or black colour etc., and these, or some of them, are necessary attributes of any man.

Although Epicurus' basic distinction between bodies is the simple disjunction, atom or compound, he may have supposed that certain compounds function as molecules or basic complexes which serve as 'seeds' for the production of more complex things.²² Lucretius writes of'seeds of water' or 'seeds of fire', and though he may mean simply 'that out of which water or fire is composed' the specific property of water or fire is something which only arises through the combination of atoms. It is probable, though not certain, that Epicurus would have regarded a pool of water as a compound of smaller compounds—molecules of water. If the molecule were broken down we should of course be left only with atoms and void. Epicurus firmly rejected the four-element theory which persists through Greek philosophy in various forms from Empedocles to Neoplatonism.

I have called the compound body a dynamic entity. This description applies not merely to its internal atoms but also to those atoms which constitute its surface. An object which persists over any length of time does not retain the same atoms throughout that time. Epicurus supposed that atoms are constantly leaving the surface of objects and having their place taken by further atoms which bombard the object and may then get caught up on its structure.²³ The notion that atoms are constantly leaving the surface of objects is fundamental to the explanation of sensation. As observed already, we sense something when effluences enter our sense organs. If these are 'real' images they are simply the outer surface or skin of objects which is 'sloughed off, to use a Lucretian metaphor, in a continuous stream of 'films'.

Epicurus took over his fundamental principles of atoms and void from earlier atomists but we have seen that he was no slavish imitator. The atomist system seemed to him to provide an explanation of the structure of things which was both compatible with empirical data and psychologically comforting, in that it did away with the need for divine causation and any form of teleology. Whether or not men find it more or less comforting to suppose that the world is determined by a supernatural being or beings, seems to be very much a matter of personal temperament. Lucretius praises Epicurus for delivering mankind from 'the weight of religion' (i 62ff.). He means popular religion, superstitious beliefs in the gods as direct arbiters of human destiny and fears of divine anger as expressed in thunder and lightning. But Epicurus cannot have taken popular superstitions as his only target. He was also concerned to reject the sophisticated theology of Plato and probably Aristotle too.

How he did so and what he put in its place will be the subject of the next few pages. Before discussing this, however, we should notice how Epicurus rejects out of hand the method of teleological explanation which bulks so large in Plato and Aristotle. Plato makes Socrates complain that Presocratic thinkers fail to show why it is best for things to be as they are (Phaedo 99a-d).²⁴ What Socrates is alleged to have found a defect—the concentration on mechanical explanations—Epicurus regarded as a positive merit. Things are not 'good for anything', he argued; this is merely a piece of learned superstition. There is no purpose which the world as a whole or things in particular are designed to fulfil. For design is not a feature of the world; it is manifestly imperfect.²⁵ Given the fact that the number of atoms is infinite and that their shapes are immensely various, it is not remarkable that similar combinations of things arise. Indeed, Epicurus held that the number of worlds is infinite, some of which are like our own while others vary (Ep. Hdt. 45). All of them however are ultimately explicable by reference to the purposeless combination and separation of discrete and inanimate physical entities moving in empty space.

Epicurus' cosmology denies the foundation of the Platonic and Aristotelian world-picture. The fossilized Aristotelianism of the schoolmen came under attack in the Renaissance, and Epicurean atomism was given a new relevance by the French mathematician and opponent of Descartes, Pierre Gassendi. The history of later science has amply vindicated Epicurus' rejection of final causes. But it is arguable that Epicurus' renunciation of teleology, in its historical context, went too far. His principle of explanation in terms of accidental arrangement of atoms will hardly serve to account adequately for such phenomena as biological reproduction. Why, to give the question an Aristotelian tone, does man produce man? Lucretius, it is true, offers an answer to this question: the characteristics of a species are transmitted to the offspring through its own seed (iii 74Iff.), and he repeatedly emphasizes that each thing has its own fixed place; that there are natural laws determining biological and other events (i 75-7; ii 700ff; iii 615ff. etc.). But the basis of these laws does not seem to rest firmly on anything implied by Epicurus' atomist principles. His physical theory has to explain too much by too little. These complaints are legitimate if we judge Epicurus by Aristotelian standards. But in making them it is necessary to remember that Epicurus did not set out to be a purely disinterested investigator of things. According to his own words, 'the purpose of studying nature is to gain a sharp understanding of the cause of those things which are most important' (Ep. Hdt. 78). By 'most important' he means fundamental to human well-being.

The two subjects on which he thought it most important to have correct beliefs were theology and psychology. Having now discussed Epicurus' basic physical principles I turn to consider how he applied these to his treatment of the gods and the human mind.

(vi) The gods of Epicurus

That which is sublimely happy and immortal experiences no trouble itself nor does it inflict trouble on anything else, so that it is not affected by passion or partiality. Such things are found only in what is weak (K.D. i).

Nothing disquieted Epicurus more profoundly than the notion that supernatural beings control phenomena or that they can affect human affairs. That there are gods he did not deny. But he repeatedly and vociferously rejected the belief that gods are responsible for any natural events. His rejection of this belief is expressed most pointedly in contexts concerning astronomical phenomena.

Moreover we must not suppose that the movement and turning of the heavenly bodies, their eclipses and risings and settings and similar movements are caused by some being which takes charge of them and which controls or will continue to control them, while simultaneously enjoying complete bliss and immortality. For occupation and supervision, anger and favour, are not consistent with sublime happiness…. Nor again must we suppose that those things which are merely an aggregate of fire possess sublime happiness and direct these (celestial) movements deliberately and voluntarily (Ed. Hdt. 76-7).

The object of Epicurus' polemic in these lines is any theology which ascribes divine control to the heavenly bodies. By denying the survival of the personality in any form after death Epicurus thought he could remove the source of one basic human anxiety—fear of divine judgment and eternal punishment. And I will discuss this feature of his philosophy in the next section. But he held that it was equally false and disturbing to credit gods with any influence over human affairs here and now. Thereby he denied the foundations of popular Greek religion. The notion that human wellbeing and adversity are dispensed by the gods was fundamental to popular Greek religion. Language reflected the belief: a happy man was eudaimôn, 'one who has a favourable deity'; kakodaimôn, 'unhappy', literally means 'having a harmful deity'. For the majority perhaps, belief in the gods of mythology had generally been a matter of civic or private ritual rather than any inner experience. But many Greeks, educated and uneducated alike, subscribed to mystery cults which promised salvation to the initiated, and fears of pollution and divine intervention remained strong. Theophrastus' portrait of the Superstitious Man, overdrawn though it is, would lose all point if it had no basis in everyday experience.²⁶

But Epicurus' attack on divine management of the cosmos was more specifically directed, we may suppose, against the cosmology of Plato and Aristotle. Plato, in his later works, constantly refers to the regularity of celestial movements as evidence of intelligent direction by divine beings. In the Timaeus we are told that the purpose of sight for men is to observe the 'revolutions of intelligence in the heavens, so that we may use their regular motions to guide the troubled movements of our own thinking'; we are to 'imitate the invariant movements of God' (47b-c.) This concept of divine ordination of the heavens was developed by Plato with much more detail in the Laws, his last major work.²⁷ There he defends the thesis that the heavenly bodies have as their cause virtuous souls or gods (x 899b). Their virtue is proved by the regularity of the movements which they cause. Disorderly movements, whether in the heavens or on earth, must be accounted for by a soul which is bad (897d).

Not only, according to Plato, are the stars directed by gods. Mankind and the universe as a whole are the 'possessions' of the gods (902b-c). In the Laws Plato emphasizes that the control which gods exercise over men is providential. But he states equally strongly that it is absolute. Plato thought he could legislate for a reformed religion by banishing the discreditable gods of tradition and replacing them with new gods whose excellence was manifested in the mathematical perfection of celestial physics. But to Epicurus Plato's astral gods were quite as repugnant as the traditional Olympian pantheon. His refusal to regard the movements of the stars as a consequence of divine intentions is an explicit rejection of Plato's own language. In the Epinomis, the title of which means 'after the laws', astronomy is made the key to this new theology.²⁸ The heavenly bodies are there said to be the source of our knowledge of number, which itself is the foundation of intelligence and morality. This queer assertion is put forward in all seriousness, and it leads to a reiteration of the divinity of the stars. These beings possess wondrous powers of mind; they know our own wills; they welcome the good and loathe the bad (984d-985b).

Epicurus regarded such beliefs as a prime source of human anxiety. To him they seemed to compound old superstitions and to make the heavenly bodies, with their watching brief over human affairs, an object of utter terror. He can have been little less disquieted by Aristotle's theology. For Aristotle too regarded the heavenly bodies as intelligent, divine beings whose movements are voluntary. We know that Aristotle expressed such views in an early work designed for popular consumption.²⁹ It is also true that Aristotle's views about celestial movements developed during his lifetime and that he does not in any extant treatise make the heavenly bodies personal arbiters of human destiny. But it is a basic doctrine of his Metaphysics that all movement and life are ultimately dependent upon the Unmoved Mover, pure Mind or God, whose activity of eternal self-contemplation promotes desire and motion in the heavenly bodies, each governed by its own intelligence. Aristotle approved of popular beliefs in the divinity of the heavens, even though he denied their traditional mythological trappings (Met. 1074a38ff.). The Unmoved Mover, like Epicurus' own gods, is not personally concerned with the universe. But unlike Epicurus' gods the Unmoved Mover is the prime cause of all things. He does not determine human affairs by his own fiat. They are none the less dependent upon events, such as the sun's diurnal rotation and seasonal change, of which he is the ultimate cause.³⁰

Epicurus rejects divine management of the world by an argument based upon the meaning and implications of the words 'sublimely happy' and 'immortal'. He accepts that these predicates, traditionally ascribed to the gods, express real attributes of divine beings (Ep. Men. 123). But he denies that sublime happiness and immortality are compatible with any involvement in 'our affairs' (Cic. N.D. i 51-2). In his view happiness, whether human or divine, requires for its full realization a life of uninterrupted tranquillity or freedom from pain. For the moment we must accept this concept of happiness as a dogmatic assertion, for its full analysis belongs to the field of ethics. Epicurus' argument concerning the gods' indifference to the world picks up this concept of happiness, applies it to the gods and thereby removes from them any actions or feelings which take account of human affairs. Seneca summarizes this position: 'Hence god dispenses no benefits; he is impregnable, heedless of us; indifferent to the world … untouched by benefits and wrongs' (Us. 364).

The argument makes three assumptions. First, that there are gods. Secondly, that the gods are sublimely happy and immortal. Thirdly, that their happiness consists in uninterrupted tranquillity. We must now consider how Epicurus justified these assumptions.

Since he was disposed to combat all beliefs about divine control of the world it may seem surprising that Epicurus accepted even the existence of supernatural beings. For, if the gods could be shown to be fictions, all activities associated with them would necessarily also be ruled out. Epicurus argued however that the universal beliefs of mankind establish the fact that gods exist. 'What people or race is there which lacks an untaught conception of gods?' (Cic. N.D. i 43). The argument claims that a belief in gods exists independently of institutionalized religion or custom. It is therefore something natural. Of course the belief might be natural and false. But Epicurus uses a basically Aristotelian premise (E.N. Nicomachean ethics 1172b35) to reject this objection: 'That about which all agree must be true.'

The same principle, consensus omnium, is used to establish the properties of the gods. All men are also said to have a natural belief that the gods are immortal, sublimely happy and of human shape.³¹ Epicurus held that these common beliefs are 'preconceptions' derived from experience—visions which people have when awake and all the more when asleep. He argued that these visions must, like all sensations, be caused by something real, that is to say atomic configurations or images (effluences) which come from the gods themselves and which penetrate to our minds. The theory is naïve and fails to take account of other factors which might have led men to believe in gods. If as a matter of fact human beliefs about gods were as consistent with each other and as similar as Epicurus claims, the hypothesis of mental perception of divine images would offer a reason for the consensus omnium: we are all acted upon by the same kind of external images. But Epicurus assumes the consensus omnium, and then seeks to explain it in psychophysical terms.

According to Lucretius the properties of sublime happiness and immortality were inferred by men from their mental images of the gods:

They endowed the gods with eternal life, because images of the gods were constantly supplied with unchanging form, and also because they believed that beings possessed of such strength could not be vanquished by any chance force. And they supposed that the gods enjoyed supreme happiness because no fear of death troubled any of them, and because in dreams the gods were seen to do many remarkable things without any expenditure of effort (v 1175-82).

Lucretius proceeds to argue that early men, having acquired a belief in gods, supposed them to be the agents of astronomical and meteorological phenomena through ignorance of the real cause of these things. Elsewhere he argues at length that the gods cannot have had any desire or ability to create the universe, the imperfections of which are clear evidence that it is not under divine direction (v 156-94). The gods, like true Epicureans, dwell in sedes quietae ('tranquil resting-places') enjoying a life free of all trouble (iii 18-24).

Besides the natural conceptions of mankind Epicurus himself gave other reasons for justifying his account of the gods' nature. He defended their anthropomorphic appearance by the argument that this is the most beautiful of all shapes and therefore the shape which belongs to beings whose nature is best (Cic. N.D. i 46-9). Not all of his statements about the gods can be reduced to the supposed evidence of natural conceptions.³² But while recognizing that the images which men receive from the gods are 'insubstantial' and difficult to perceive, he clearly hoped to show that primary knowledge of the gods is something similar in kind to the direct acquaintance with physical objects which we obtain through our sense organs.

How is the physical structure of the gods to be conceived of and what manner of life do they lead? The evidence which bears on these questions is difficult and must be dealt with summarily here. The basic problem is the everlasting existence of the gods. Atoms and void are imperishable, for atoms possess a solidity which is impervious to all 'blows' and void is unassailable by 'blows' (Lucret. iii 806-13). But the ordinary objects of experience, because they are compounded out of atoms and void, are not of a kind to resist destruction indefinitely. So Lucretius writes:

By (natural) law they perish, when all things have been made weak by outflow (of atoms) and give way to the blows which come from outside (ii 1139-40).

Epicurus sought to avoid this difficulty by introducing a mode of being which is not a compound body in any ordinary sense. Our main evidence is a difficult text of Cicero. After observing that the gods are perceived by the mind and not the senses, Cicero writes that they do not possess 'solidity' nor 'numerical identity', like the ordinary objects of perception, but 'an unlimited form of very similar images arises out of the innumerable atoms and flows towards the gods' (N.D. i 49, cf. 105). Elsewhere the gods are perhaps spoken of as 'likenesses' and some gods (or all of them from one point of view) are 'things which exist by similarity of form through the continuous onflow of similar images which are brought to fulfilment at the same place' (Us. 355).³³ Scholars have questioned the accuracy of these statements on the grounds that they make the atoms which flow towards the gods 'images'. But these doubts may be misplaced. 'Image' (or eidôlon in Epicurus' terminology) is distinguished from 'solid body'. There can be 'images' in this sense which are merely patterns of fine atoms lacking the density to constitute a solid body. We may perceive Centaurs in virtue of such 'images', but they are not images contrasted with a real Centaur; for there is no solid body corresponding to a Centaur. Similarly, there is no solid body which emits images of the gods. For the gods do not possess a solid body. They are called 'likenesses', in all probability, because their nature is continuously reconstituted by a moving stream of 'images', discrete arrangements of fine atoms which possess similar form.

The gods then have no numerical identity. If their substance were that of an ordinary compound body they would be subject to an irreplaceable loss of atoms and hence destructible by external 'blows'. Their identity is 'formal', a consequence of the constant arrival and departure of similar forms at the points in space occupied by the gods. It has sometimes been aptly compared with the nature of a waterfall the shape of which is determined by continuous flow. Epicurus does not explain why there should be a continuous supply of atoms patterned in the right way to constitute the form of the gods. If pressed he would probably have argued that in a universe which contains an infinite number of atoms this is not impossible in principle and that our experience of the gods—the fact that we can have more than momentary visions of them—proves the continuity of their form, and therefore the supply of appropriate atoms. Moreover we think of them as immortal.

Does this strange notion of the gods' physical structure imply that they themselves make no contribution to their own unceasing existence? The evidence discussed so far could imply that the gods are simply happy beneficiaries of a constant supply of atoms which replace those they have lost. Some scholars have adopted such a view. But Philodemus, who wrote a work On the gods (as did Epicurus himself), parts of which are preserved on a Herculaneum papyrus, seems to have argued that the gods' own excellence and powers of reason secure them from destructive forces in the environment.³⁴ It should not be supposed that this is a laborious activity for the gods. On the contrary, all our sources stress the fact that the gods enjoy an existence completely free of toil. They are able, in virtue of their nature, to appropriate the atoms which preserve their existence and to ward off atoms of the wrong kind. This seems to have been Philodemus' view, and we may reasonably credit it to Epicurus himself.

The gods' tenuity of structure is explained by the 'images' which form their bodies, a theory ridiculed by ancient critics (N.D. i 105). We may now see why the gods are not said to have body but quasi body, not blood but quasi blood (Cic. N.D. i 49). They are seen above all in dreams and are themselves dreamlike in substance, insubstantial like the 'images' which make up their being. It may seem strange that Epicurus should have credited such creatures with perfect happiness, but this becomes less odd when we study what their happiness consists in. It is something negative rather than positive. The gods have no occupations, they can be affected by no pain, they are liable to no change. They dwell in no world but in the spaces which separate one world from another (intermundia, Cic. N.D. i 18). And since it is Epicurus' claim that absence of pain is the highest pleasure, and pleasure is the essence of happiness, the gods are perfectly happy. In the later Epicureanism of Philodemus more positive things are said about the gods, including the fact that they speak Greek! But Epicurus, so far as we know, did not go in for such crude anthropomorphism.

… The Gods, who haunt
The lucid interspace of world and world,
Where never creeps a cloud, or moves a wind,
Where never falls the least white star of snow,
Nor ever lowest roll of thunder moans,
Nor sound of human sorrow mounts to mar
Their sacred everlasting calm!³⁵

Human affairs are no concern of the gods. But are the gods, or should they be, any concern of men? Epicurus seems to have held that certain forms of ritual and private devotion are appropriate because, although the gods cannot be touched by prayers or sacrifices, they provide men with a model of beatitude. Whether or not a man is benefited by the gods depends upon his state of mind when he apprehends divine 'images'. If he himself is tranquil he will attain to the right view of the gods (Lucret. vi 71-8), and a passage in the Letter to Menoeceus (124) is perhaps to be interpreted along the same lines: those whose disposition is already akin to that of the gods can appropriate and gain benefit from divine 'images'. It is in the same spirit that Philodemus writes: 'He [sc. Epicurus] appeals to the completely happy in order to strengthen his own happiness' (Diels, Sitz. Berl. 1916, p. 895).

(vii) The soul and mental processes

Death is nothing to us; for that which has been dissolved lacks sensation; and that which lacks sensation is no concern to us (K.D. ii).

The first thing which Epicurus strove to establish in his psychological theory was the complete and permanent loss of consciousness at death. All his philosophy has the ultimate aim of removing human anxiety, and the therapeutic aspect of his psychology is a most conspicuous example of this. By denying any kind of survival to the personality after death, Epicurus hoped to show that beliefs in a system of rewards and punishments as recompense for life on earth were mere mythology. It is difficult to assess precisely the strength and prevalence of such beliefs in Epicurus' own time. But apart from his vehement desire to undermine them, which must be evidence that they were not uncommonly held, we find independent confirmation in literature and the popularity of 'mystery cults'.

Traces of such beliefs are already to be found in fifth-century writers. It is not likely that they declined during the fourth century. Plato, in the Republic, writes scathingly of those who provide books of Musaeus and Orpheus, and try to persuade 'whole cities as well as individuals' to absolve themselves from their crimes by performing certain rituals. Such men hoodwink people into believing that they will reap benefits in this world and the next, while those who fail to observe the rituals will be confronted with 'terrible things' (Rep. ii 364e). Cephalus, the old man who figures at the beginning of the Republic, is portrayed as someone who has been tormented by fears of having to expiate his offences in Hades. Plato himself condemns the quacks who try to exploit such fears by offering absolution at a fee. But he ends the Republic with a myth which has judgment of the dead and rewards and punishments for earthly existence as its central feature. Similar myths are used by him in the Phaedo and the Gorgias. Both the pre-existence and the survival of the soul, after death, are central Platonic doctrines which he combines with the theory of metempsychosis.

The fear of death is common to all peoples at all times. Epicurus may have exaggerated the psychological disturbance he attributed to explicit beliefs in a destiny for the soul after death. But he did not confine his diagnosis of the fear of death to eschatological dogmas. As he himself writes (Ep. Hdt. 81), and as Lucretius writes at greater length, men also fear death who believe that it is the end of all sensation. Lucretius uses an argument to remove such fear which simply asserts that this is a true belief and therefore no grounds for anxiety: events which took place before we were born did not disturb us, for we did not feel them. By parity of reasoning, nothing can disturb us when we cease to be conscious of anything (iii 830-51). A little later he goes on:

If there is to be any trouble and pain for a man he too must exist himself at that time in order that ill may affect him. Since death removes this and prevents the existence of him to whom a mass of misfortunes might accrue, we may be assured that there is nothing to be feared in death and that he who no longer exists cannot be troubled (iii 861-8).

For Epicurus birth and death are limits which contain the existence of a person. I have not existed in another body prior to this life, nor am I liable to experience a further incarnation following this life. There is such a thing as psychê, soul, the presence of which in a body causes that body to possess life. Here Epicurus agreed with philosophical and popular conceptions. But he also insisted against Plato and other dualists that the soul cannot exist independently of the body, and that a living being must be a union of the body and the soul. Disrupt this union and life ceases. His view bears comparison with that of Aristotle who defined soul as 'the first actuality of an organic natural body' (De an. [De anima/On the soul] ii 412b5). For Aristotle too, most functions of the soul are necessarily related to the body, though Aristotle made an exception of intellect. We should not however press this comparison. Aristotle's treatment of soul proceeds by very different steps from Epicurus'. It calls for distinctions between form and matter and between potentiality and actuality which are quite foreign to Epicurus' way of thinking; in Aristotle soul is not a kind of physical substance, as it is for Epicurus. That on which they broadly agree is the mutual dependence of body and soul.

What then did Epicurus say about the soul? Or, to put the question another way, how did he explain life? The first point is that life must be accounted for by reference to something corporeal. For that which is not body is void, and void cannot do anything nor be affected by anything (Ep. Hdt. 67). The grounds for denying that soul is incorporeal show that Epicurus regards the capacity to act and to be acted upon as a necessary condition of that which animates a living being. More specifically, soul consists of atoms which act upon and are affected by the atoms constituting the body itself.

'Soul is a body, the parts of which are fine, distributed throughout the whole aggregate. It resembles most closely breath mixed with heat' (Ep. Hdt. 63). Lucretius enables us to amplify this description. The atoms which form the soul, he tells us, are very small and they are also round. Roundness is inferred from the speed of thought: the soul atoms can be stirred by the slightest impulse (iii 176ff). Furthermore, breath and heat (Lucretius adds air as well) are not sufficient to account for soul. That soul is warm and airy is clear from the fact that warmth and breath are absent from a corpse. But breath, heat and air cannot create the movements which bring sensation (iii 238-40). Something else is needed—a 'fourth nature', consisting of atoms which are smaller and more mobile than anything else which exists. They have no name.

How are we to conceive of the relation between the different kinds of atoms which constitute soul? As I observed earlier when discussing compound bodies, fire, breath and so forth are things which can only arise when certain kinds of atoms are combined. We are probably therefore to suppose that the atoms which can, in appropriate combinations, create the substances specified by Lucretius are in the soul combined in such a way that they form a body which is analysable as a mixture of fire, breath, air and the unnamed element. But the soul will not be divisible into these things. Lucretius says explicitly that 'no single element [of the soul] can be separated, nor can their capacities be divided spatially; they are like the multiple powers of a single body' (iii 262-6).³⁶

This body, in virtue of the unnamed element, can produce the movements necessary to sensation. It is the unnamed element which gives the soul its specific character. Life and vital functions in general are thus explained by reference to something which cannot be analysed fully into any known substances. Epicurus wants to avoid the objection that life has been simply reduced to an appropriate mixture of familiar substances. Life requires these and something else as well. Given the primitive notions which the Greeks had of chemistry, Epicurus was wise to refrain from attempting to explain life purely in terms of the traditional four elements.

The soul then, so constituted, is the 'primary cause of sensation'. But soul by itself cannot have or cause life. It must be contained within a body. That is to say, a living being can be constituted neither out of soul alone nor out of body alone. Placed within a body of the right kind, the soul's vital capacities can be realized. From the soul the body acquires a derivative share in sensation; there is physical contact, naturally, between the body and the soul, and the movements of atoms within the body affect and are affected by those of the soul. Epicurus illustrated the relation of body and soul by considering the case of amputation (Ep. Hdt. 65). Loss of a limb does not remove the power of sensation; but loss of the soul removes all vitality from the body, even if the body itself remains intact. By insisting that soul must be contained in the body, Epicurus denied any prospect of sensation and consciousness surviving death.

In Lucretius (iii 136ff.) and some other sources a spatial distinction is drawn between the animus (rational part) and the anima (irrational part). The animus is located in the chest; the rest of the soul, though united with the animus, is distributed throughout the other regions of the body. These two parts of the soul do not undermine its unity of substance; they are introduced to explain different functions. That in virtue of which we think and experience emotion is the animus, the mind. This governs the rest of the soul. Epicurus had no knowledge of the nervous system, and we may most easily think of the anima as fulfilling the function of nerves—reporting feelings and sensations to the animus and transmitting movement to the limbs. If Epicurus had known of the nerves and their connexion with the brain he would probably have been fully prepared to accept the notion that the brain is equivalent to the animus.³⁷

Lucretius' account of the soul inspired some of his finest poetry. It also shows much acute observation of behaviour. Before considering further aspects of Epicurean psychology, we may pause over one or two passages in which Lucretius calls attention to the relations between body and soul:

If the vibrant shock of a weapon, forced within and opening to view bones and sinews, falls short of destroying life itself, yet faintness follows and a gentle falling to the ground, and on the ground there ensues a storm of the mind, and moment by moment an unsteady desire to stand up. Therefore it must be the case that the mind is corporeal in substance, since it suffers under the blow of bodily weapons (iii 170-6).

Furthermore we perceive that the mind comes into being along with the body, develops with the body and grows old with it. Just as young children totter whose body is frail and soft, so too their powers of judgment are slight. Then, when maturity has developed bringing hardy strength, their judgment too is greater and their strength of mind increased. Later, after their body has been assailed by the tough force of age and their limbs have failed with their strength blunted, the intellect grows lame, the tongue raves, the mind stumbles, all things fail and decline at the same time. And so it is appropriate that all the substance of vitality should be dissolved like smoke into the lofty breezes of the air (iii 445-56).

Again, if the soul is immortal and can feel when separated from our body we must, I believe, cause it to be equipped with the five senses. There is no other way in which we can imagine to ourselves souls wandering below in the realm of Acheron. And so painters and former generations of writers have presented souls endowed with the senses. But neither eyes nor nostrils nor hand nor tongue nor ears can exist for the soul apart from the body. Therefore souls on their own cannot feel, nor even exist (iii 624-33).

The psychological function on which we have most copious information is sense-perception. This is due no doubt to its importance in Epicureanism as a whole. In discussing the theory of knowledge I have already described Epicurus' concept of effluences or images from external objects which enter the sense organs, or which penetrate directly to the mind, and thereby cause our awareness of something. Perception is thus ultimately reducible to a form of touch, physical contact between the atoms of the percipient and atoms which have proceeded from objects in the external world (cf. Lucret. ii 434f.). It will not be possible in this context to discuss the treatment of specific problems of optics and other matters which Lucretius deals with at great length in Book iv. But some more general aspects of the theory need consideration.³⁸ In particular, it would be desirable to know how we are to conceive of the 'sense-bearing movements' which the soul bestows on the body. Lucretius states that it is not the mind which sees through the eyes, but the eyes themselves (iii 359-69), and the same holds good for the other sense organs. The eye is an organ of the body. When it is struck by a stream of effluences which have the appropriate size, we are probably to suppose that this sets up a movement of the adjoining soul atoms, which then cause sensation in the eye itself. The same internal processes will account for feelings—burns, itches and so forth. Lucretius traces a series of movements beginning with 'the stirring of the unnamed element' which passes via the other elements in the soul to the blood, the internal organs and finally the bones and marrow (iii 245-9). These stages represent an increasing disturbance of 'everything' and life cannot be maintained if the movements or sensations penetrate beyond a certain point. But 'generally the movements come to an end on the surface of the body' (252-7).

All of this still leaves the notion of sensation or consciousness very obscure. Is it simply a kind of movement, or is it rather something which supervenes as an epiphenomenon upon a kind of movement? Epicurus and Lucretius leave us in the dark here, and no wonder! For no one has yet succeeded in giving a purely mechanistic explanation of consciousness.

The treatment of thought raises further problems. In Epicurus' Letter to Herodotus (49-50) and in Lucretius, thought is assimilated to sense-perception so far as its objects and causes are concerned:

Come now, learn what things stir the mind, and hear in a few words the source of those things which enter the understanding. First of all, I declare that many likenesses of things wander in many ways in all regions in all directions; they are fine in texture and easily become united with one another in the air when they meet…. They are much finer than those things which seize hold of the eyes and rouse vision, since they pass right through the vacant spaces of the body and stir the fine nature of the mind within and rouse its awareness (Lucret. iv 722-31).

Lucretius proceeds to illustrate these statements by reference to the perception of monsters and the dead. From this we might suppose that he is describing only certain kinds of mental perception. But he goes on to argue that the thought of a lion is also produced, like the sight of a lion, by simulacra leonum ('images of lions'). He also asks the question, 'How is that we are able to think of things at will?' The answer which he suggests is curious:

In a single period of time which we perceive, that is, the time it takes to utter a single word, many times escape notice which reason discovers; and so it happens that in any time all the images are available, ready in every place. So great is their power of movement, so great the supply of them…. Because they are fine the mind can only distinguish sharply those on which it concentrates. Therefore all except those for which it has prepared itself pass away. The mind prepares itself and expects that it will see what follows on each thing; therefore this comes about (iv 794-806).

We should expect thought to be explained by reference to data, images, or what not which are somehow already present in or created by the mind. But that is not what Lucretius says. He clearly implies that the 'supply' of images is external, and that the mind apprehends just those images to which its attention is directed. Thinking, on this interpretation, is analogous to noticing something which falls within the scope of vision.

Many scholars have been reluctant to take this passage at its face value. They have presumed that Epicurus must have envisaged an internal store of images in terms of which some thought at least and memory are to be explained. It has been suggested that the effluences which are received by the sense organs or the mind cause a change in the movements of the soul atoms, and this new pattern of movement persists as a memory or thought-image. But if Epicurus held such a view no evidence about it has survived, and it is not presupposed in the theory recorded by Lucretius. According to this theory, thought is a kind of internal film-show in which the mind controls the images which it permits to enter the body. Not only thought but volition also requires the consciousness of appropriate images. Lucretius observes that we walk when 'images of walking fall upon the mind' (iv 881). Then the will rouses itself, and passes on movement to the rest of the soul so that finally the limbs are activated. In dreaming, too, passages are open in the mind through which images of things can enter (iv 976-7).

We must suppose that this is Epicurus' own theory, and it is quite consistent with the strange idea that images of the gods possess objective status. But if memory is not explicable by a storehouse of images how is it to be accounted for? The few texts which bear upon this question suggest that memory is a disposition, produced by repeated apprehension of images of a certain kind, to attend to such images as continue to exist after the previously experienced object which produced them may have perished or changed in other ways. Hence we can remember the dead.

Although Lucretius' account of thought should be given full weight as orthodox Epicurean doctrine, certain forms of thinking must have been explained in other ways. There are no images of atoms and void; there is no image of the principles of confirmation and non-contradiction. Yet these are things which cannot be grasped except by thought. Epicurus himself distinguished between what he called 'the theoretical part', and 'apprehension' whether by the senses or the mind.³⁹ I have already discussed 'apprehension', and rejected Bailey's claim that it guarantees the clarity of an image. What it does involve, I suggest, is the direct apprehension of some image. In other words, 'apprehension' covers awareness of all data, whether of the senses or the mind, which possess objective existence because they are in origin images which enter us from outside. 'The theoretical part' refers to thinking which may be presumed to function purely by internal processes. That is to say, it involves inference about things like atoms and void which are unable to be apprehended directly. How such thinking takes place in detail, and whether or not it involves images, are questions which we cannot answer categorically. It almost certainly must make use of 'preconceptions', general concepts arrived at by repeated observation of particular objects. Precon ceptions however cannot be reduced to external images for there are no 'generic' images existing objectively. These must be constructions of the mind, which can be utilized in the formation of new non-empirical concepts. Epicurus may have explained them as patterns of movement in the soul, but his words on this subject, if he described it in detail, have not survived.

(viii) Freedom of action

Our next subject is one of the most interesting and controversial problems in Epicureanism. According to Lucretius and other later writers, the 'swerve' of atoms has a part to play in the explanation of 'free will' (libera voluntas). But what part? The answer must be sought, if anywhere, from a difficult passage of Lucretius. It is the second argument which he uses to prove that atoms sometimes deviate from the linear direction of their downward movement through the void:

If every movement is always linked and the new movement arises from the old one in a fixed sequence, and if the primary bodies do not by swerving create a certain beginning of movement which can break the bonds of fate and prevent cause from following cause from infinity, how comes it about that living things all over the earth possess this free will, this will, I say, severed from fate, whereby we advance where pleasure leads each man and swerve in our movements at no fixed time and at no fixed place, but when and where our mind has borne us? For undoubtedly each man's will gives the beginning to these movements, and it is from the will that movements are spread through the limbs.

You see, do you not, that when the barriers are opened at an instant of time, the horses for all their strength and eagerness cannot burst forward as promptly as their mind desires. This is because the whole stock of matter throughout their whole body has to be set in motion, so that having been roused through all the frame it may make an effort and follow the desire of the mind. So you can see that a beginning of movement is engendered by the heart, and it comes forth first from the mind's volition and then is dispatched throughout the whole body and limbs.

It is not the same as when we move forward under the pressure of a blow from the mighty strength and strong constraint of another man. For then it is clear that the whole matter of the body in its entirety moves and is seized against our will, until the will restrains it throughout the limbs. So now you surely see that although an external force pushes many men and often compels them to go forward against their will and be driven headlong, yet there is something in our breast which can fight back and resist. At its direction too the stock of matter is at times compelled to change direction through all the limbs, and although pushed forward is checked and comes to rest again.

Therefore you must admit the same thing in the atoms too; that another cause of motion exists besides blows and weights which is the source of this power innate in us, since we see that nothing can arise out of nothing. For weight prevents all things happening by blows, by external force as it were. But a tiny swerve of the atoms at no fixed time or place brings it about that the mind itself has no internal necessity in doing all things and is not forced like a captive to accept and be acted upon (ii 251-93).

This is our only detailed evidence for the relation between the swerve and 'free' will in Epicurean literature.⁴⁰ No explicit word about the swerve from Epicurus himself has so far been discovered. We know however that he attacked 'the destiny of the natural philosophers' for its 'merciless necessity' (Ep. Men. 134). And he discussed the causes of human action in a book which is partially preserved on papyrus from Herculaneum.⁴¹ The text contains so many gaps and defective lines that it is difficult to grasp a clear train of thought; more work may yield positive advances in our understanding of it. But Epicurus certainly distinguished sharply in this book between 'the cause in us' and two further factors, 'the initial constitution' and 'the automatic necessity of the environment and that which enters' (i.e. external 'images'). These distinctions should be borne in mind when one approaches Lucretius' text.

The first thing to notice is his context. Lucretius' main subject at this stage of his work is not psychology but the movement of atoms. He offers no formal argument to defend the 'freedom' of the will. Rather, he assumes it, exemplifies it by examples, and uses the assumption and examples to prove that atoms sometimes swerve.

The logical structure of the first paragraph might be expressed like this: (A) If all movements are so causally related to each other that no new movements are created by a swerve of atoms, then there could be no such thing as 'free will'. (B) For 'free will' entails the creation of a new movement at no fixed time and at no fixed place. (C) But there is such a thing as 'free will'. (D) Therefore the atoms sometimes create new beginnings by swerving.

In the next two paragraphs (as I have set out the text) Lucretius gives his examples to show that the will can create new beginnings of movement. First, he considers the case of the race-horses. When the barriers are raised they are free of any external constraint to move forward; and they do move forward (fulfil their desire to move) as soon as their will has had time to activate the limbs. This example is used to show that there is some faculty within the horses which enables them to initiate movement freely.

In his second example Lucretius considers a different case. Unlike the horses, which require a brief interval of time for their will's action to have an external effect, men may be pushed forward immediately by outside pressures; and such involuntary movements require no internal movements in the men before they occur. But men have something within them which can resist external pressures. This is a power to cause atoms within the body to change their enforced direction of movement. In this case, the will initiates movement when the body is already undergoing compulsory movement. But the power which is exercised is the same as that faculty for initiating movement exemplified by the horses.

It seems clear to me that both examples are intended to illustrate 'free will', from different starting-points. And what are we to make of the 'blows' and 'weights' mentioned in the last paragraph? Lucretius does not exclude these as necessary conditions of a 'free' action. He denies only that they are sufficient to bring it about. We have already seen that the 'will' to walk requires 'images', that is, 'blows from outside', and nothing weightless could act or be acted upon in Epicurus' system. The weight of the mind's atoms affects its reaction to external blows. But an action caused solely by blows and weights could not be 'free'. The horses' movement and the men's resistance were 'free' in virtue of an additional third factor which Lucretius calls voluntas, 'will'. The will, it should be noted, is not treated as equivalent to desire. Desire is prompted by the awareness of some pleasurable object. Lucretius, I think, regards the will as that in virtue of which we seek to fulfil our desires (cf. ii 258, 265).

If we ask where the swerve features in all this, the answer seems to be that the swerve is a physical event which presents itself to consciousness as a 'free' will to initiate a new movement. Consciousness during waking states is normally continuous. But our external bodily movements are not wholly continuous. Nor are our intentions. Lucretius, I suggest, treats some animal actions as if they were relatively discontinuous events initiated by the 'will'. But they are not wholly discontinuous. Memory, reflection, habit, these and other dispositions are not ruled out as causal factors by anything which Lucretius says. A swerve among the soul atoms need not be supposed to disrupt all or even any character traits. The swerve is not treated in a context which enables us to place its precise function in the whole history of living things. But what does seem clear is its rôle as an initiator of new actions.

Once it is recognized that other causes besides the swerve are necessary to the performance of a voluntary act, certain difficulties observed by one recent writer become less acute. If the swerve, which by definition is something random and unpredictable, were sufficient by itself to explain voluntary actions, then the bonds of fate could seem to have been broken at the expense of making actions purposeless and wholly indeterminate. Sensing this difficulty, David Furley has suggested that the swerve need not be supposed to feature in the explanation of every voluntary action.⁴² Its function, he argues, is rather to free the disposition from being wholly determined by heredity and environment. For this purpose he suggests that a single swerve of a single atom in an individual's psyche would be sufficient.

Furley's interesting arguments cannot be surveyed in detail here. But it is my own opinion that Lucretius' text is easier to interpret on the assumption that a swerve is at work in the freedom of particular actions. The theoretical single swerve which Furley postulates can hardly suffice to explain the 'beginnings of motion' which characterize each act of 'free' will. But Furley is right to object to interpretations which treat the swerve by itself as a sufficient condition. We know from Diogenes of Oenoanda that the swerve was held to be necessary by Epicureans if moral advice is to be effective (fr. 32 col. iii). But moral advice cannot be effective if it is to depend entirely on the possible occurrence of a swerve in the soul of the man being advised. I think we are to suppose that the swerve of a single atom is a relatively frequent event. It may occur when one is asleep or when one is awake, without having any observable or conscious effect. If however a man's natural disposition to seek pleasure or to avoid pain is roused by external or internal causes, and if at such times he is in a physical condition which makes it possible for him to act then, depending upon the kind of man he is, any swerve(s) among the atoms of his mind constitutes a free decision to act. If he is untrained in Epicurean philosophy he may decide to pursue objects which in the event cause more pain than pleasure. The true Epicurean's atoms may also swerve at a time when he walks down a Soho street, but having learnt that freedom from pain is more pleasurable than momentary sensations of pleasure he does not follow his companions into the night-club. Swerves help him to initiate new actions in the pursuit of tranquillity.

Before concluding this subject two general observations may be made. First, it may be asked whether the use of the swerve which I attribute to Epicurus is historically plausible. Furley believes not, but he omits one point which seems to me important. The so-called problem of 'free will' arises primarily out of two conceptions—beliefs in God's omniscience and predetermination, and beliefs in the absolute continuity of physical causation. There is every reason to attribute the second of these to Zeno, the founder of Stoicism, and probably the first as well. Zeno and Epicurus were active in Athens together for thirty years, and I find it unlikely that Epicurus developed his opposition to determinism quite independently of Stoic theories. Those theories provide conditions which favour the emergence of a concept of volition which is not completely dependent upon the state of things at the preceding instant. I conjecture that Epicurus used the swerve to defend such a concept.

Secondly, the random nature of the swerve is a difficulty for Epicurus whatever its rôle in human psychology. But it seems to me logically possible to suppose that the swerve of a soul atom is not random so far as the consciousness of the soul's owner is concerned. And Epicurus attempted to solve that problem, so I think, by making the swerve a constituent of" the 'will'. If it is not part of the will or cognitive faculty, but a random event which disrupts the soul's patterns of atoms from time to time, are not the consequences for mortality which trouble Furley still more serious? He wants the swerve to free inherited movements of atoms and make character adaptable. But this raises new problems. A man of good Epicurean character will live in fear of an unpredictable event which may change him into a Stoic or something worse. I find it easier to posit some discontinuity between the antecedent conditions of an action and the decision to do it than discontinuity between movements on which character depends.

To conclude, Epicurus used the swerve of atoms in the soul to explain situations where men are conscious of doing what they want to do. Obscurities persist, and we cannot rule them out in order to make the theory more palatable or convincing. I pass now to a less controversial subject. Epicurus' theory of pleasure has already been referred to, and we must now consider its full ethical significance.

(ix) Pleasure and happiness

Epicurus was not the first Greek philosopher whose ethics can be called hedonist. In [an earlier chapter] I referred briefly to the earlier hedonism of Aristippus whose conception of the pleasant life differs sharply from Epicurus'. Unlike Aristippus, who regarded absence of pain as an intermediate condition, Epicurus claimed that the removal of all pain defines the magnitude of pleasure (K.D. iii), and his interest in specifying the conditions which establish a life free of trouble may well have been roused by Democritus. The earlier atomist probably had no systematic ethical theory, but he is credited with a conception of happiness which consists above all in peace of mind (D.L. ix 45). It was Epicurus' primary concern to show how this state can be attained.

Pleasure was also a topic which received considerable attention from Plato, Aristotle and the Academy in general. It is highly probable that Epicurus was familiar with ideas which Plato discusses in the Philebus, and he may also have been influenced by some Aristotelian notions, most notably the distinction between pleasure 'in movement' and pleasure 'in rest' (E.N. 1154b28). The great difference between Plato and Aristotle on the one hand and Epicurus on the other turns on the relation they posit between happiness and pleasure. All three philosophers are concerned in their ethics with specifying the necessary conditions of happiness, but only Epicurus identifies happiness with a life full of pleasure. Some pleasures for Plato and Aristotle are good, and make a contribution to happiness; others are bad. For Epicurus no pleasure in itself can be anything but good since the good means that which is or causes pleasure. The fundamental constituent of happiness for Plato and Aristotle is virtue, excellence of 'soul', which manifests itself in the exercise of those activities appropriate to each faculty of the personality and in moral action. (The differences between Plato and Aristotle are less important for my present purpose than the similarities.) But for Epicurus virtue is necessary to happiness not as an essential ingredient but as a means to its attainment. This is the most significant difference between Epicurus and his major predecessors:

We say that pleasure is the starting-point and the end of living blissfully. For we recognize pleasure as a good which is primary and innate. We begin every act of choice and avoidance from pleasure, and it is to pleasure that we return using our experience of pleasure as the criterion of every good thing (Ep. Men. 128-9).

Subjective experience is a 'test of reality' for Epicurus, and it is on this evidence that he based his doctrine of pleasure.

All living creatures from the moment of birth take delight in pleasure and resist pain from natural causes independent of reason (D.L. x 137).

The goodness of pleasure needs no demonstration. Epicurus takes it as an obvious fact that men, like all living things, pursue pleasure and avoid pain. The attractiveness of pleasure is treated as an immediate datum of experience comparable to the feeling that fire is hot (Cic. Fin. i 30). We learn from Aristotle that his contemporary Eudoxus also inferred that pleasure is 'the good' from the allegedly empirical fact that all creatures pursue it (E.N. x 1172b9). Now it does not follow from the fact, if it is a fact, that men pursue pleasure that pleasure is what they ought to pursue. As G. E. Moore argued at great length in Principia Ethica, that which is desired is not equivalent to that which is desirable. Some scholars have indeed claimed that Epicurus is not concerned with what 'ought' to be or what is 'fitting', but only with what is.⁴³ But this claim is at best a half-truth. It would be a correct description of his view to say that we are genetically programmed to seek what will cause us pleasure and to avoid what will cause pain. And he probably held that no living creature whose natural constitution is unimpaired can have any other goals. But there is a place for 'ought' in his system because the sources of pleasure and pleasure itself are not uniform. What we ought to do is to pursue that which will cause us the greatest pleasure. Ought' here of course does not signify what we are obliged to do by any purely moral law. It signifies that which needs to be done if we are successfully to attain our goal, happiness or the greatest pleasure. It applies to means and not to ends:

Since pleasure is the good which is primary and innate we do not choose every pleasure, but there are times when we pass over many pleasures if greater pain is their consequence for us. And we regard many pains as superior to pleasures when a greater pleasure arises for us after we have put up with pains over a long time. Therefore although every pleasure on account of its natural affinity to us is good, not every pleasure is to be chosen; similarly, though every pain is bad, not every pain is naturally always to be avoided. It is proper to evaluate these things by a calculation and consideration of advantages and disadvantages. For sometimes we treat the good as bad and conversely the bad as good (Ep. Men. 129-30).

In order to grasp the implications of this important passage, we need to consider in more detail what Epicurus meant by pleasure and how he proposed to use pleasure as a guide to action. The most striking feature of his hedonism is the denial of any state or feeling intermediate between pleasure and pain. Pleasure and pain are related to one another not as contraries but as contradictories.⁴⁴ The absence of the one entails the presence of the other. If all pleasure is regarded as a sensation of some kind, this relationship between pleasure and pain makes no sense. For clearly most of us pass a large part of our waking lives with out having either painful or pleasing sensations. But the periods of our waking life in which we could describe ourselves as neither happy nor unhappy, or neither enjoying nor not enjoying something, are much smaller. Epicurus' view of the relationship between pleasure and pain should be interpreted in this light. He has, as we shall see, a way of distinguishing the pleasures of bodily sensations and feelings of elation from pleasures which cannot be so described; absence of pain is not, he thinks, an adequate description of the former. Epicurus' mistake is a failure to see that indifference characterizes certain of our moods and attitudes towards things.

His analysis of pleasure rests on the assumption that the natural or normal condition of living things is one of bodily and mental well-being and that this condition is ipso facto gratifying. This is the meaning of the statement quoted above: 'Pleasure is the good which is primary and innate.' It is possible in English to speak of 'enjoying' good health, and we may also call this something gratifying, or something a man rejoices in. Epicurus' use of the word pleasure to describe the condition of those who enjoy good physical and mental health is not therefore purely arbitrary. In physical terms this pleasure is a concomitant of the appropriate movement and location of atoms within the body. If these are disturbed pain follows. In other words, pain is a disruption of the natural constitution. Pleasure is experienced when the atoms are restored to their appropriate position in the body (Lucret. ii 963-8).

The idea that pain is a disturbance of the natural state was not invented by Epicurus. We find it in Plato's Philebus along with the notion that pleasure is experienced when the natural state is 'replenished' (31e-32b). Plato however argued that pleasure is only experienced during the process of restoring the natural condition. According to his theory, pleasure and pain are movements or processes. There is also, however, a 'third' life in which any bodily processes produce no con sciousness of pleasure or pain. This 'intermediate' condition cannot be regarded as pleasurable or painful (42c-44a).

It is interesting to find Plato attacking the theory that absence of pain can be identified with pleasure. Epicurus of course would not accept Plato's specification of an intermediate life. Like Plato, however, Epicurus holds that the process of removing pain results in pleasurable sensations. He calls this pleasure 'kinetic'. Suppose that a man is hungry: he desires to eat and the act of satisfying this desire produces 'kinetic' pleasure. If he succeeds in fully satisfying the desire for food he must have wholly allayed the pangs of hunger. From this complete satisfaction of desire, Epicurus argues, a second kind of pleasure arises. This is not an experience which accompanies a process, but a 'static' pleasure. It is characterized by complete absence of pain and enjoyment of this condition. Torquatus, the Epicurean spokesman in Cicero's De finibus, expresses the distinction between pleasures thus:

The pleasure which we pursue is not merely that which excites our nature by some gratification and which is felt with delight by the senses. We regard that as the greatest pleasure which is felt when all pain has been removed (Fin. i 37).

'Static' pleasure follows the complete satisfaction of desire. Desire arises from a sense of need, the pain of lacking something. In order to remove this pain, desire must be satisfied, and the satisfaction of desire is pleasurable. 'Kinetic' pleasure is thus (or so I think) a necessary condition of at least some 'static' pleasure, but it is not regarded by Epicurus as equivalent in value to 'static' pleasure.⁴⁵ For if freedom from pain is the greatest pleasure, we should satisfy our desires not for the sake of the pleasurable sensations which accompany eating, drinking and so on, but for the sake of the state of well-being which results when all the pain due to want has been removed:

When we say that pleasure is the goal we do not mean the pleasures of the dissipated and those which consist in the process of enjoyment… but freedom from pain in the body and from disturbance in the mind. For it is not drinking and continuous parties nor sexual pleasures nor the enjoyment offish and other delicacies of a wealthy table which produce the pleasant life, but sober reasoning which searches out the causes of every act of choice and refusal and which banishes the opinions that give rise to the greatest mental confusion (Ep. Men. 131-2).

Epicurus, in this passage, is not denying that drink, eating good food, sex, and so on are sources of pleasure. He is asserting that the pleasures which such activities produce are to be rejected as goals because they do not constitute a calm and stable disposition of body and mind. It is freedom from pain which measures the relative merits of different activities. This is the basis of Epicurus' hedonist calculus. His criticism of luxury and sexual indulgence is not grounded in any puritanical disapproval:

If the things which produce the pleasures of the dissipated released the fears of the mind concerning astronomical phenomena and death and pains … we should never have any cause to blame them (K.D. x).

He holds that the greatest pain is mental disturbance produced by false beliefs about the nature of things, about the gods, about the soul's destiny. Any pleasure therefore which fails to remove the greatest pain is ruled out as an ultimate object of choice by application of the rule—absence of pain establishes the magnitude of pleasure. Furthermore, the pleasure which arises from gratifying the senses may have a greater pain as its consequence. A man may enjoy an evening's drinking or the thrill of betting, but the pleasure which he derives from satisfying his desires for drink and gambling must be set against the feeling of the morning after and the anxiety of losing money.

Epicurus' concept of pleasure is closely related to an analysis of desire:

We must infer that some desires are natural, and others pointless; of natural desires some are necessary, others merely natural. Necessary desires include some which are necessary for happiness, others for the equilibrium of the body and others for life itself. The correct understanding of these things consists in knowing how to refer all choice and refusal to the health of the body and freedom from mental disturbance since this is the goal of living blissfully. For all our actions are aimed at avoiding pain and fear. Once we have acquired this, all the mind's turmoil is removed since a creature has no need to wander as if in search of something it lacks, nor to look for some other thing by means of which it can replenish the good of the mind and the body. For it is when we suffer pain from the absence of pleasure that we have need of pleasure (Ep. Men. 127-8).

Epicurus' analysis of desires is consistent with the principle that freedom from pain is the greatest pleasure. The desire for food and clothing is natural and necessary. Failure to satisfy this desire is a source of pain. But, Epicurus argues, it is neither necessary nor natural to desire this food or clothing rather than that, if the latter is sufficient to remove the pain felt by absence of food or clothing (Us. 456). Hence Epicurus becomes an advocate of the simple life, on the grounds that we cause ourselves unnecessary pain if we seek to satisfy desires by luxurious means. Necessary desires, he holds, can be satisfied simply, and the pleasure which we thus experience is no less in quantity even if it differs in kind. Moreover, those who seek pleasure in luxuries are likely to suffer pain unnecessarily either as a direct consequence of luxurious living or through an inability to satisfy a desire:

We regard self-sufficiency as a great blessing, not that we may always enjoy only a few things but that if we do not have many things we can enjoy the few, in the conviction that they derive the greatest pleasure from luxury who need it least, and that everything natural is easy to obtain but that which is pointless is difficult. Simple tastes give us pleasure equal to a rich man's diet when all the pain of want has been removed; bread and water produce the highest pleasure when someone who needs them serves them to himself. And so familiarity with simple and not luxurious diet gives us perfect health and makes a man confident in his approach to the necessary business of living; it makes us better disposed to encounter luxuries at intervals and prepares us to face change without fear (Ep. Men. 130-1).

Time and again in his Principal doctrines Epicurus asserts that pleasure cannot be increased beyond a certain limit.⁴⁶ So far as sensual gratification is concerned, this limit is reached when the pain which prompted desire has ceased; thereafter pleasure can be 'varied'—by 'kinetic' pleasure—but not augmented. This is something which needs to be grasped by the intellect, since the flesh itself recognizes no limits to pleasure (K.D. xviii; xx). The mind has its own pleasures the 'limit' of which is reached with the ability to calculate correctly the pleasures of sensual gratification and to assess the feelings which cause mental disturbance. Ancient critics familiar with Plato's dis tinction between the soul and the body criticized Epicurus for failing to draw a sharp distinction between bodily pleasures and the 'good' of the soul (Us. 430, 431). But when Epicurus distinguishes body and soul in statements about pleasure he has nothing like Platonic dualism in mind. Body and mind are in physical contact with one another; pleasurable sensations are 'bodily' events but they also give rise to pleasure or joy in the mind (Us. 433, 439). Unlike the body however the mind is not confined for its objects of pleasure to the experience of the moment:

The body rejoices just so long as it feels a pleasure which is present. The mind perceives both the present pleasure along with the body and it foresees pleasure to come; and it does not allow past pleasure to flow away. Hence in the wise man there will always be present a constant supply of associated pleasures, since the anticipation of pleasures hoped for is united with the recollection of those already experienced (Us. 439).

The memory of past pleasures can 'mitigate' present sufferings (Us. 437), and the same holds for the anticipation of future pleasures. Unlike the Cyrenaics who regarded bodily pleasures of the moment as the greatest, Epicurus is reported by Diogenes Laertius to have argued that the mind's capacity to look forward and back entails that both its pleasures and its pains are greater than those of the body (x 137).

The distinction between 'kinetic' and 'static' pleasures applies to both body and mind (D.L. χ 136). Corresponding to the 'kinetic' pleasure of satisfying a desire for food, drink and the like, the mind can experience 'joy' when, say, meeting a friend or solving a problem in philosophy. This pleasure, consisting in motion, is to be distinguished from the 'static' pleasure of 'mental repose' which corresponds to the body's pleasure in freedom from pain.

Since pleasure is the only thing which is good in itself, prudence, justice, moderation and courage, the traditional four 'moral virtues' of Greek philosophy, can have value only if they are constituents of or means to pleasure. Epicurus settled for the second alternative. Torquatus puts his position succinctly, opposing the Epicureans to the Stoics:

As for those splendid and beautiful virtues of yours, who would regard them as praiseworthy or desirable unless they produced pleasures? Just as we approve medical science not for the sake of the art itself but for the sake of good health … so prudence, which must be regarded as the 'art of living', would not be sought after if it achieved nothing. In fact is is sought after because it is the expert, so to speak, at discovering and securing pleasure…. Human life is harassed above all by an ignorance of good and bad things, and the same defect often causes us to be deprived of the greatest pleasures and to be tormented by the harshest mental anguish. Prudence must be applied to act as our most reliable guide to pleasure, by removing fears and desires and snatching away the vanity of all false opinions (Cic. Fin. i 42-3).

Moderation, on the same principle, is desirable because and only because 'it brings us peace of mind'. It is a means to attaining the greatest pleasure since it enables us to pass over those pleasures which involve greater pain. Similarly, Torquatus finds the value of courage in the fact that it enables us to live free of anxiety and to rid ourselves, as far as possible, of physical pain. Justice and social relationships are analysed in the same way, but I will say a little more about Epicurus' treatment of these at the conclusion of this chapter.

Although Epicurus regarded the virtues as means and not as ends, he held that they are necessary to happiness and inseparably bound up with the hedonist life:

Of sources of pleasure the starting-point and the greatest good is prudence. Therefore prudence is something even more valuable than philosophy. From prudence the other virtues arise, and prudence teaches that it is not possible to live pleasurably without living prudently, nobly and justly, nor to live prudently, nobly and justly, without living pleasurably. For the virtues are naturally linked with living pleasurably, and living pleasurably is inseparable from them (Ep. Men. 132).

This association between virtue and pleasure is strik ing, but it should not be interpreted as giving an independent value to prudence and the other virtues. The necessary connexion between pleasure and the virtues is due to the notion that pleasure requires for its attainment a reasoned assessment of the relative advantages and disadvantages of a particular act or state of affairs, a capacity to control desires the satisfaction of which will involve pain for the agent, freedom from fear of punishment and the like. The pleasure which men should seek is not Bentham's 'greatest happiness of the greatest number'. Epicurus never suggests that the interests of others should be preferred to or evaluated independently of the interests of the agent. The orientation of his hedonism is wholly self-regarding.

(x) Justice and friendship

Natural justice is a pledge of expediency with a view to men not harming one another and not being harmed by one another (K.D. xxxi).

Of all the things which wisdom secures for the attainment of happiness throughout the whole of life, by far the greatest is the possession of friendship (K.D. xxvii).

Aristotle asserted that 'man is naturally a political animal', and Plato held that the true good of the individual is also the good of the community to which he belongs. Epicurus took a very different view, which, though less forcefully than the Cynics, challenged fundamental values of Greek society. In his opinion human beings have no 'natural' leanings towards community life (Us. 523). Civilization has developed by an evolutionary process of which the determinants have been external circumstances, the desire to secure pleasure and to avoid pain, and the human capacity to reason and to plan. Learning by trial and error under the pressure of events, men have developed skills and formed social organizations which were found to be mutually advantageous. The only details of this process which survive in Epicurus' own words concern the origin and development of language (Ep. Hdt. 75-6). But Lucretius treats the whole subject at some length in Book v. Following the invention of housing and clothing, the discovery of how to make fire, and the introduction of family life, he writes, 'neighbours began to form friendships desiring neither to do nor to suffer harm' (1019-20). From this supposed historical stage in human culture Epicurus traces the origins of justice.

He describes justice as 'a kind of compact not to harm or be harmed' (K.D. xxxiii), prefacing this statement with the words: 'It is not anything in itself, an implicit attack on Plato's theory of the autonomous existence of moral values. Several longer statements on justice are preserved in the Principal doctrines, and a selection of these will serve to illustrate Epicurus' position:

Injustice is a bad thing not in itself, but in respect of the fear and suspicion of not escaping the notice of those set in authority concerning such things (K.D. xxxiv).

It is not possible for one who secretly acts against the terms of the compact 'not to harm or be harmed', to be confident that he will not be apprehended…. (K.D. xxxv).

Evidence that something considered to be just is a source of advantage to men, in their necessary dealings with one another, is a guarantee of its justice, whether or not it is the same for all. But if a man makes a law which does not prove to be a source of advantage in human relationships, this no longer is really just…. (K.D. xxxvii).

The just man is most free from trouble, but the unjust man abounds in trouble (K.D. xvii).

This concept of justice, which recalls Glaucon's analysis in Plato (Rep. ii), is not the 'social contract' of Rousseau. Epicurus is not saying that people have an obligation to act justly because of an agreement entered into by their remote or mythical ancestors. The 'contractual' element in his concept of justice is not advanced as a basis of moral or social obligation. Epicurus' justice requires us to respect the 'rights' of others if and only if this is advantageous to all parties concerned. Justice, as he conceives of it, does imply recognition of the interests of others besides oneself. But the basis of this recognition is self-interest. The 'compact' of which he speaks has self-protection as its basis. It is an agreement to refrain from injuring others if they will refrain from injuring oneself.

Epicurus' comments on the fears of apprehension which beset the unjust man show clearly that justice is desirable for the freedom which it brings from mental distress as well as physical retaliation. This is wholly consistent with his calculus of pleasures and pains. Injustice is bad not in itself but because of the painful consequences which it involves for the unjust man. In a book of Problems Epicurus raised the question: 'Will the wise man do anything forbidden by the laws, if he knows that he will escape notice?'; and he replied, 'The simple answer is not easy to find' (Us. 18). Epicurus' comments in K.D. xxxv imply that the problem is a purely academic one. No one in practice can be confident that his injustice will be unnoticed; as Lucretius puts it, 'fear of punishment for crimes during life' is the real hell, and not the Acheron of myth (iii 1013-23). Realizing this the wise man acts justly in order to secure tranquility of mind.

In Epicurus' eyes political life is a rat-race or 'prison' from which the wise man will keep well clear (Sent. Vat. lviii).⁴⁷ He diagnoses political ambition as a 'desire for protection from men', and argues that this in fact can only be secured by a quiet life in retirement from public affairs (K.D. vii, xiv). But Epicurus' rejection of political life as a context for the attainment of happiness was not based upon misanthropy. On the contrary, he held that friendship is 'an immortal good' (Sent. Vat. lxxviii). So today, 'opting out' and communal life are practised by many who find society at large 'alienated'. Torquatus, in Cicero's De finibus, asserts that 'Epicurus says that of all the things which prudence has provided for living happily, none is greater or more productive or more delightful than friendship' (i 65). Once again we notice that the value of something other than pleasure or happiness is referred to this end. But Epicurus, when writing of friendship, uses almost lyrical language at times, as when he says: 'Friendship dances round the world, announcing to us all that we should bestir ourselves for the enjoyment of happiness' (Sent. Vat. lii).

There is no doubt that Epicurus practised what he preached. The Garden was a community of friends, and Epicurus clearly derived intense happiness from friendship. We are told that he was famous for his 'philanthropy' to all (D.L. χ 10), and on the day of his death, when racked by pain, he wrote to Idomeneus that he was happy, with the joyous memories of their conversations (Us. 138). Some of Epicurus' remarks about friendship might imply that it is compatible with altruism and self-sacrifice. But the basis of friendship, like justice, is self-interest, though Epicurus in the same breath says that it is 'desirable for its own sake' (Sent. Vat. xxiii). There is probably no inconsistency here. 'He is not a friend who is always seeking help, nor he who never associates friendship with assistance' (Sent. Vat. xxxix). One can enjoy or derive pleasure from helping a friend independently of any tangible benefit which this brings. When Epicurus writes of the 'benefits' of friendship he does not mean the pleasure which may come from helping others but the actual practical help which friends provide for each other. Apart from this, however, friendship is desirable because 'it is more pleasant to confer a benefit than to receive it' (Us. 544). One again we are brought back to pleasure as the sole criterion of value.

There is an elegant simplicity to Epicurus' ethics, a refreshing absence of cant, and also much humanity. He was born into a society which, like most societies, rated wealth, status, physical attributes and political power among the greatest human goods. It was also a slave-based society which reckoned men as superior to women and Greeks as superior to all other peoples. The good for man which Epicurus prescribes ignores or rejects these values and distinctions. Freedom from pain and tranquillity of mind are things which any sane man values and Epicurus dedicated his life to showing that they are in our power and how we may attain them. His ethics is undeniably centred upon the interests of the individual, and some have, with justification, praised the nobility of Epicurus more highly than his moral code. Yet we must see it in its social and historical context. No Greek thinker was more sensitive to the anxieties bred by folly, superstition, prejudice and specious idealism. At a time of political instability and private disillusionment Epicurus saw that people like atoms are individuals and many of them wander in the void. He thought he could offer them directions signposted by evidence and reason to a way of being, a way of relating to others, other individuals. Negative, self-centred, unstimulating we may regard it; we cannot say priggish or self-indulgent, and in antiquity many found liberation and enlightenment in Epicureanism. For a modern reader too there is much philosophical interest in the consistency with which Epicurus applies his basic principles. What he has to say about family life and sexual love is entirely based on the proposition that the greatest good is freedom from pain in body and mind. But consistency can be purchased too dearly; a few criticisms of Epicurus' hedonism will show this.

First, it may be objected that Epicurus misapplies the factual observations with which he starts. Unless pleasure is used analytically to mean merely that which is desired, it is difficult to agree that pleasure is the object of every desire. But such a usage of pleasure tells us nothing about what is desired or desirable, and Epicurus does not use pleasure in this empty way. He is claiming that the desire to attain a state of consciousness which we find gratifying is sufficient to explain all human action. And this seems to be patently false. Secondly, Epicurus can fairly be charged with failure to grasp the complexity of the concept of pleasure. As we say, one man's meat is another man's poison, but Epicurus seems to think that he can classify by reference to absence of pain the magnitude of any man's pleasure. This may often be good advice, but many will argue from their own experience that Epicurus' claims have no basis in fact. They will also reject his assertion that sharp pains are short and long pains mild.

Ancient critics complained that Epicurus has united under the term pleasure two quite different desiderata, positive enjoyment and the absence of pain (Cic. Fin. ii 20), and there is some grounds for the complaint. Such remarks as 'The beginning and root of all good is the pleasure of the stomach' (Us. 409) are much more naturally interpreted in the former sense, even if Epicurus did not intend this. In particular, it is difficult to make sense of the notion that absence of pain entails that pleasure can only be varied and not increased. If I derive pleasure from smelling a rose at a time when I am suffering no pain, it seems perverse to say that my pleasure is merely varied. For what I experience is something sui generis, a new sense of gratification which is more than a variation of my previous state of consciousness. Even if it is reasonable to call tranquillity of mind a kind of pleasure it is straining language and common sense to call it the greatest pleasure. That procedure leaves us no way to take account of experiences which cannot be assimilated to tranquillity and which do cause us intense gratification without pain.

Thirdly, under 'static' pleasure Epicurus seems to have classified two quite different things. The pleasure which follows from the satisfaction of desire is normally related closely in time to the desire. When I leave hospital, restored to health, it makes sense to say 'You must be pleased to be better'. But it makes much less sense to say this a long time later. The mental equilibrium enjoyed by an Epicurean of ten years' standing seems to be something quite different.

These and other observations could be prolonged. I have confined my comments to Epicurus' own theories, and it is hardly necessary to dwell on some of the obvious objections to egoistic hedonism as a 'moral' theory. But in this book Epicurus should have the last word, and it is eloquently expressed by his great disciple and admirer, Lucretius:

When the winds are troubling the waters on a mighty sea it is sweet to view from the land the great struggles of another man; not because it is pleasant or delightful that anyone should be distressed, but because it is sweet to see the misfortunes from which you are yourself free. It is sweet too to watch great battles which cover the plains if you yourself have no share in the danger. But nothing is more pleasing than to be master of those tranquil places which have been strongly fortified aloft by the teaching of wise men. From there you can look down upon other men and see them wandering purposelessly and straying as they search for a way of life—competing with their abilities, trying to outdo one another in social status, striving night and day with the utmost effort to rise to the heights of wealth and become masters of everything. Unhappy minds of men, blind hearts! How great the darkness, and how great the dangers in which this little life is spent. Do you not see that nature shouts out for nothing but the removal of pain from the body and the enjoyment in mind of the sense of joy when anxiety and fear have been taken away. Therefore we see that for the body few things only are needed, which are sufficient to remove pain and can also provide many delights. Nor does our nature itself at different times seek for anything more pleasing, if there are no golden status of youths in the entrance halls holding in their right hands fiery torches so that evening banquets may be provided with light, or if the house does not gleam with silver and shine with gold and a carved and gilded ceiling does not resound to the lute, when, in spite of this, men lie on the soft grass together near a stream of water beneath the branches of a lofty tree refreshing their bodies with joy and at no great cost, particularly when the weather smiles and the time of the year spreads flowers all over the green grass (ii 1-35).

Notes

¹ Epicurus probably first encountered Metrodorus, his junior by about ten years, at Lampsachus, the latter's native town.

² The authenticity of this letter has been questioned, but there is no reason to doubt its reliability as a statement of Epicurus' attitudes and doctrine.

³ N. W. De witt, Epicurus and his Philosophy (Minneapolis 1954) p. 329. The last chapter of this book should be consulted for a survey of the later fortunes of Epicureanism.

⁴ For the evidence see Bibliography.

⁵ Nearly all the Herculaneum papyri belong to the Biblioteca Nazionale of Naples; but the British Museum has substantial fragments of Epicurus On Nature Book ii.

⁶ Epicurus did not invent the 'effluence' theory of senseperception. It goes back to Democritus and still earlier, in a different form, to Empedocles.

⁷ Cicero (ND. i 44) says that Epicurus was the first to use the word prolêpsis in this sense.

⁸ Clement of Alexandria (Us. 255) reports Epicurus as saying that 'it is impossible for anyone to investigate…or to form a judgment…independently of preconception'.

⁹ For further evidence and discussion see my article in Bulletin of the Institute of Classical Studies 18 (1971) 114-33.

¹⁰The Greek Atomists and Epicurus (Oxford 1928) p. 570.

¹¹ Formal statements of these 'axioms' are to be found in Sextus Empiricus, Adv. math, vii 212-13 (Us. 247). Epicurus writes of them more informally in Ep. Hdt. 51 and K.D. xxiv.

¹² See Philodemus, On signs col. xvi ed. Ph. and E. A. De Lacey (Pennsylvania 1941).

¹³Adversus sophistas frs. 1, 3 ed. Sbordone (Naples 1947).

¹⁴ For more details see the article cited in n. 2.

¹⁵ Lucretius develops (A) and (B) at length, i 159-264; he deals with (C) at ii 294-307.

¹⁶ For Lucretius' arguments concerning void, see i 329-97. Aristotle denied the necessity of void to explain motion, as did the Stoics.

¹⁷ See also Lucretius i 599-634.

¹⁸ The most penetrating study of Democritean and Epicurean atomism is by D. J. Furley, Two Studies in the Greek Atomists (Princeton 1967). His first study deals with the notion of minimal parts.

¹⁹ This is supported by Aristotle's words, quoted by Simplicius (DK 68 A 37); cf. W. K. C. Guthrie, History of Greek Philosophy vol. ii (Cambridge 1965) pp. 400-2.

²⁰ For a detailed treatment, to which I am much indebted, see Furley, Two Studies in the Greek Atomists, pp. 111-30.

²¹ The gist of what Lucretius says is repeated by Cicero, Fin. i 18-20.

²² Good arguments in favour of this have been advanced by G. B. Kerferd, Phronesis 16 (1971) 88-9.

²³ See in particular Ep. Hdt. 48. The regular replenishment of lost' atoms explains why objects are not (normally) seen to diminish in bulk.

²⁴ Plato himself attacks purely mechanistic theories of causation, probably with Democritus in mind, in the tenth book of his Laws 889b.

²⁵ Lucretius attacks 'final' causes at length, ν 195-234.

²⁶ The reference is especially relevant because it is a contemporary one. Theophrastus' Characters are a set of short vignettes of character types: complaisance, boorishness, miserliness etc.

²⁷ The development of Plato's theology is a complex subject; for a well-balanced account see F. Solmsen, Plato's Theology (Ithaca, N.Y. 1942).

²⁸ Doubts about the Platonic authorship of this work have been expressed from antiquity onwards. If not by Plato himself, which is likely enough, it should be ascribed to a younger contemporary Academic, perhaps Philip of Opus (D.L. iii 37).

²⁹ Cic. N.D. ii 42-4. Comparison with Book i 33 of Cicero's treatise suggests the reference may be to Aristotle's third book 'On philosophy'. On Aristotle's theology see W. K. C. Guthrie, Classical Quarterly 27 (1933) 162-71; 28 (1934) 90-8.

³⁰ It was clearly Epicurean practice to attack all theological views which differed from their own; cf. Cic. N.D. i passim. Epicurus must have found much to object to in the writings of Plato's successor, Xenocrates, whose theology contained 'daimons' as well as celestial divinities and who foreshadowed the Stoics in referring the names of certain gods to natural substances.

³¹ Cic. N.D. i 45-6; see also Ep. Men. 123-4 which refers to the false conceptions of the 'many' about gods.

³² See further, K. Kleve, Symbolae Osloenses, suppl. xix. Epicurus (Cic. N.D. i 50) also inferred that there must be divine beings equal in number to mortals by the principle of 'equal balance' or 'reciprocal distribution' (isonomia). 'From this,' we are told, 'it follows that if the number of mortals is so many there exists no less a number of immortals, and if the causes of destruction are uncountable the causes of conservation must also be infinite.'

³³ The word 'brought to fulfilment' (apotetelesmenôn) should not be emended with Kühn and Usener to apotetelesmenous: it is the 'images' not the gods which are brought to fulfilment; cf. Ep. Pyth. 115, 'by the meeting of atoms productive of fire'.

³⁴De dis iii fr. 32a p. 52 Diels. For further discussion of Philodemus' evidence and theories based upon it, cf. W. Schmid, Rheinisches Museum 94 (1951) 97-156, K. Kleve, Symbolae Osloenses 35 (1959), who attributes the gods' self-preservation to the exercise of free will.

³⁵ Tennyson, Lucretius, based on Lucret. iii 18-23, itself a translation of Homer.

³⁶ See further Kerferd, Phronesis 16 (1971) 89ff.

³⁷ The nervous system was discovered by the medical scientists, Herophilus and Erasistratus, during the first half of the third century B.C.

³⁸ Epicurus himself writes about seeing, hearing and smell, Ep. Hdt. 49-53. Lucretius also discusses taste, iv 615-72.

³⁹De nat. xxviii fr. 5 col. vi (sup.), col. vii (inf.), ed. A. Vogliano, Epicuri et Epicureorum scripta in Herculanensibus papyris servata (Berlin 1928).

⁴⁰ Cic. De fato 22 and N.D. i 69 also show that the swerve was supposed to save 'free will'. See also Diogenes of Oenoanda fr. 32 col. iii Chilton.

⁴¹ This is found in Arrighetti's edition, Epicuro Opere 31 [27] 3-9. It is just possible that the swerve is implied at 31 [22] 7-16.

⁴² 'Aristotle and Epicurus on Voluntary Action' in his Two Studies in the Greek Atomists. My brief remarks here cannot do justice to the importance of Furley's wideranging treatment. Much of his argument turns on similarities he has detected between Aristotle and Epicurus.

⁴³ So Bailey, Greek Atomists, p. 483, followed by Panichas, Epicurus (N.Y. 1967) p. 100.

⁴⁴ cf. Cic. Fin. ii 17, 'I assert that all those who are without pain are in a state of pleasure.'

⁴⁵ This interpretation of the relation between 'kinetic' and 'static' pleasure seems to me to suit the evidence best and to make the best sense. It has also been argued that 'kinetic' pleasure serves only to 'vary' a previous 'static' pleasure; see most recently J. M. Rist, Epicurus (Cambridge 1972) ch. 6 and pp. 170ff.

⁴⁶ e.g. K.D. iii, ix, xviii, xix, xx.

⁴⁷ See also D.L. χ 119 and Lucretius' brilliant denunciation of ambitio, iii 59-77, related to the fear of death.