The Online Library of Liberty

A project of Liberty Fund, Inc.

• Preface
• Acknowledgments
• The Perfectibility of Man
• One: Perfection and Perfectibility
• Two: From Olympus to the Form of the Good
• Three: the Godlike Man: Aristotle to Plotinus
• Four: Christianity Rejects Perfection: the Elements
• Five: Pelagius and His Critics
• Six: Perfectibility Within Christianity: the Ascetico-mystical Tradition
• Seven: Perfectibility Within Christianity: Protestant and Heretical
• Eight: Perfecting By Social Action: the Presuppositions
• Nine: Governmentalists, Anarchists and Geneticists
• Ten: the Perfecting of Man By Scientific Progress
• Eleven: Progress By Natural Development: From Joachim to Marx
• Twelve: Progress By Natural Development: From Darwin to Teilhard
• Thirteen: Perfection Renounced: the Dystopians
• Fourteen: Mysticism and Humanity
• Fifteen: the New Mysticism: Paradise Now

THE PERFECTIBILITY OF MAN

ONE

PERFECTION AND PERFECTIBILITY

When, in everyday life, men are accounted perfect, this is most commonly in relation to their performance of a task or in a role. In such contexts, “perfect” acts as a superlative. A forger can be good, very good, excellent, or perfect at imitating signatures; an accountant can be good, very good, excellent, or perfect at drawing up balance-sheets. A perfect secretary, a perfect forger, a perfect accountant, all attain to the highest possible standards in the tasks they undertake. But this by no means implies that they are perfect in the performance of other tasks, in other roles. The perfect forger does not necessarily keep perfect accounts, nor is the perfect accountant always a perfect secretary. Even more obviously, he need not be a perfect man.

Indeed, it is a serious question whether perfection in this sense—let us call it “technical perfection”—has any intelligible application to man as such. That it has, Martin Foss, for one, denies. “Society,” he writes, “simplifies and abbreviates its members to executors of their social purposes, their social professions. . . . If they are adequate to their purpose in the social scheme, they are called perfect. So we have perfect typists, perfect lawyers, perfect accountants. But are there also ‘perfect men’? I do not think so.”1* This is a pleasantly rapid way of rejecting the perfectibility of man. Being human is not a profession, nor does it by itself fulfil a social purpose. If, then, there is no other kind of perfection, or none applicable to human beings, except technical perfection, it is natural to conclude that man as such—as distinct from a secretary qua secretary or an accountant qua accountant—cannot, by the very nature of the case, attain to perfection.

This deduction admits of a number of replies. The nineteenth-century philosophical anarchist William Godwin, at least in his later writings, argues that if each man is perfectible in a task, then it follows that man as such is perfectible. “Putting idiots and extraordinary cases out of the question, every human creature,” so Godwin maintains, “is endowed with talents, which, if rightly directed, would shew him to be apt, adroit, intelligent and acute, in the walk for which his organization especially fitted him.” Each individual man, that is, has a set of talents which enable him to be trained to perform a particular task. And this makes it possible, Godwin continues, for each man to produce “something as perfect in its kind, as that which is effected under another form by the more brilliant and illustrious of his species.” Each and every man, that is, can be so trained as to be technically perfect in some particular respect. And since man as such is a mere abstraction, since mankind is made up of individual men, to demonstrate that each and every man is perfectible is to demonstrate that man is perfectible.2

At first sight, however, there are two powerful arguments against deducing the perfectibility of man from the capacity of each man to perfect himself in some particular task. A man, we pointed out, can perfect himself in one role while being imperfect in another. He may be a perfect accountant but a dreadful public-speaker. There is an enormous gap, too, between his being a perfect accountant and his being a perfect man. It is one thing to say, then, that each and every man can perfect himself at something, quite another to say that he can perfect himself in every task he is called upon, as a human being, to perform. Secondly, a man can perfect himself not only as a secretary or an accountant but as a forger, a blackmailer, a torturer, an informer. “All men are perfectible; for even the worst of men can perfect himself as a procurer.” Something seems to have gone wrong with that argument. When moralists, theologians, philosophers, dispute about whether man is “perfectible,” they take it for granted that the perfection in question does not include perfection in vice. To be perfectible in a task, it again follows, is not the same thing as being perfectible as a man.

Plato’s republic is designed to meet both these objections. Plato allots to each man one task and one task only, that task in which his talents and skills enable him to perfect himself. “Each individual should pursue that work in this city,” writes Plato of his ideal republic, “for which his nature was naturally most fitted, each one man doing one work.”3 So whereas in our society a man may be a perfect accountant but a failure as a father, in Plato’s republic such a man will not be permitted to act as a father: that will be somebody else’s task, somebody whose talents and skills enable him to perfect himself in the task of rearing children.

As for the second objection, Plato’s republic is an ideal society. An ideal State will not contain forgers, blackmailers, or procurers. So although in our everyday imperfect States there are men whose talents and skills are wholly devoted to perfecting themselves in these deplorable occupations, in an ideal State each citizen will be allocated to a morally respectable task, chosen to accord with his talents. Perhaps, then, we need do no more than slightly modify Godwin’s view: man is perfectible if and only if each man has talents and skills which would fit him for the performance of the task which would be allocated to him in an ideal society.

Obviously, however, to talk thus of an “ideal” society—a perfect society, that is—is to make use of another, and different, sense of perfection, not technical perfection. A perfect society cannot be defined as one which performs its social task perfectly: it sets social tasks, it does not have a social task. More concretely, in Plato’s republic social tasks are set by a particular class of men—the “philosopher-kings” who act as its governors. No doubt, they are expert as rulers. But if they were only expert in a technical sense—good at keeping the society in order—it would be unsafe to rely upon them not to encourage such occupations as blackmailing, informing, torturing. Technically expert rulers, indeed, commonly provide more than enough employment for such skilled professions. The expertness of Plato’s rulers is a by-product of, not the sole evidence for, their perfection. They are not perfect because they rule perfectly, they rule perfectly because they are perfect, as a consequence of their having seen “the form of the good.” So, in the end, the whole structure of Plato’s republic rests on there being a variety of perfection over and above technical perfection—a perfection which consists in, or arises out of, man’s relationship to the ideal.

Even in the case of the ordinary citizen, indeed, technical perfection is not, for Plato, enough. There is, he no doubt says, an “image” of justice—of moral perfection—in “the principle that he whom nature intended for a shoemaker should attend to shoemaking and nothing else.” But it is only an image. True justice “does not concern a man’s management of his own external affairs, but his internal management of his soul, his truest self and his truest possessions.”4 In other words, technical perfection does not automatically carry human perfection with it; if men ought to seek technical perfection, this is only as an outward expression of their moral perfection, their willingness to submit their passions to rational control.

A not dissimilar analysis of the relation between technical and human perfection is sometimes to be met with in Christian thought, especially, although not solely, in the Reforming tradition. “No one,” Luther writes, “is without some commission and calling”—a set of tasks it is his responsibility to perform.5 In most cases, this commission and calling is made clear to a man by the station into which he is born. “This means,” Luther tells us, “that a servant, maid, son, daughter, man, woman, lord, subject, or whoever else may belong to a station ordained by God, as long as he fills his station, is as beautiful and glorious in the sight of God as a bride adorned for her marriage.”6 Men serve God best, Luther argues, by remaining in their vocation, however “mean and simple” it may be—although he does, somewhat inconsistently, allow that if a boy has special abilities it is the duty both of parents and of the State to make it possible for him to perfect those abilities by education.

At once, however, that difficulty arises which we have already met, although in a somewhat different context. Human beings may be born into stations which only a God considerably more broad-minded than the God of Christianity would find “beautiful and glorious.” Luther himself gives a list of “sinful” callings: “robbery, usury, public women, and as they are at present, the pope, cardinals, bishops, priests, monks and nuns.”7 This list is no doubt somewhat controversial; not everybody would wish to include popes along with prostitutes. But the principle is what is important. Luther has to grant that the calling into which human beings are born—as a woman may be born, let us say, a temple-prostitute—is not necessarily the one, the perfect performance of which will make them “beautiful and glorious in the sight of God.”

Furthermore, although he is so emphatic that men should work diligently at their calling—and Calvin emphasizes this even more strongly—Luther does not maintain, just as Plato did not maintain, that they will reach perfection as men in this way. Indeed, Luther vehemently rejects the view that, in their earthly life, men can achieve perfection. He draws, too, an important distinction between technical and vocational perfection. The Christian, according to Luther, must use his vocation primarily as a way of serving his fellow-men; technical expertness is desirable only as a means to that end.* So although technical perfection plays a part in Lutheran moral theology, it is only a small part; Luther would certainly not identify the perfectibility of man with his ability to perfect himself in any secular task or set of tasks.

A man’s performance in his vocation, indeed, is to Luther important only because it demonstrates his obedience to God, and to God’s plan for man—just as it is important to Plato because it demonstrates his willingness to submit himself to the rule of the philosopher-kings. This attitude is even more clearly expressed in the moral theology of Karl Barth. “Faithfulness in vocation,” he writes, “means positively that in my vocation as it is I seek, either well or badly, to do satisfactory work to the best of my ability, skill and conscience, . . . giving myself to my own particular concern, remembering always that it is no accident but part of the plan and providence of God that it is my concern, and that God summons me to do justice to it.”8 In other words, technical perfection—in so far as perfection rather than conscientious application is what is called for—matters to Barth only as a special case of obedientiary perfection, absolute obedience to the will of God. This is man’s real task, the task all men have to undertake: the practice of a calling does no more than exemplify it.

Obedientiary perfection, however, still has its problems: what guarantee is there that submission to God’s will cannot lead men into imperfection? God, to judge from the Old Testament, makes strange demands upon men, bids them to act in ways which scarcely accord with our everyday ideas about moral perfection. Can the temple-prostitute really be confident that it is not God’s will that she should continue in that station in life? But this, it might be said, is at worst the problem of determining in what God’s will consists. There can be no denying that to do what God wills, once that will has been determined, is the path to perfection. But why should this not be denied? Some theologians, like the thirteenth-century Duns Scotus, have been prepared to reply that it is self-contradictory to suppose that God would ever will men to do anything except what is good. Nothing is meant by “good,” on their view, except “action in accordance with God’s will”; the supposition cannot intelligibly be entertained, therefore, that it might be good to disobey God.* But most theologians have been unhappy with the suggestion that “to do what God wills is good” means no more than “to do what God wills is to do what God wills.” They have suggested, rather, that God is perfect, and must be obeyed just for that reason. But obviously, God’s perfection is not obedientiary perfection; God obeys nobody, not even himself. So, if this view is to be maintained, perfection has to be defined, when it applies to God, as something other than obedientiary perfection. And the perfection of men’s conduct when they obey God will then lie not in their obedience as such but in the fact that their conduct reflects the perfection of God.

Summing up, to identify the perfectibility of man with his capacity to perfect himself technically in a task is to encounter insuperable difficulties—unless perfectibility is wholly divorced from moral perfection. It is only if the task can be thought of as one set by a perfect Being or in a perfect society that there is any plausibility in supposing that a man can perfect himself as a human being merely by perfecting his ability in a task. Two consequences then follow. The first is that the perfection of such a Being or such a society does not itself consist in its perfect performance of a task. The second is that a man’s perfecting himself in the performance of a task is not in itself sufficient to ensure his perfecting himself as a man: it is important only in so far as it bears witness to his perfection in some other, more fundamental respect, in obedience, in submission, in rational control.

Closely related to technical perfection, however, is a more philosophically complex concept of perfection—teleological perfection—which has often been invoked as the test of human perfectibility. This is the perfection which consists in a thing’s reaching its “natural end.” Its great exponent is Aristotle. Every form of activity, Aristotle argues, is directed towards an end. The art of sculpting, for example, has as its end the depiction of the human figure, the art of medicine has as its end health. It is impossible to suppose, Aristotle goes on to contend, that although sculptors and doctors each seek the end natural to the forms of activity they undertake, man as such has no end which he naturally seeks. The only question, for Aristotle, is in what that natural end consists. He identifies it with eudaimonia—“happiness” or “well-being.” Man is perfectible, then, if and only if he is capable of attaining to well-being.9

The soundness of Aristotle’s argument is more than questionable. Tasks are not set by Nature but by men, within particular forms of social organization: it would not be in the least surprising if man as such, unlike the sculptor or the carpenter, had no task to perform, no end to pursue. This, indeed, is precisely what we should expect. Tasks, forms of activity, differentiate men, diversify their responsibilities; there are sculptors, and sculptors who produce figures in marble rather than abstract arrangements of junk-metal as their “natural end,” only in societies which call upon men to undertake, and encourage them to perfect themselves, in such activities.* No doubt there are natural differences between human beings; men cannot give birth to babies nor can women fertilize a womb. But even the “functions” of men and women, beyond this elementary point, are fixed by men, not by Nature. It is not Nature who decides whether women shall work in industry, in agriculture, or in the home, or what they shall take as their objective in those pursuits. The so-called “natural ends,” indeed, should more accurately be described as “conventional” ends.

The distinction between technical and teleological perfection might at first sight appear to be a distinction without a difference. Technical perfection, it can plausibly be argued, is simply perfection in the methods necessary and sufficient to produce a particular end. A man is technically perfect in a task if and only if he is capable of attaining the objective to which that task is conventionally directed, i.e. its teleological perfection. But this is not quite the situation.

In some instances, no doubt, the two modes of perfection are but opposite sides of the same penny. We should not reckon a forger “technically perfect” unless he can actually produce a perfect imitation of a signature, the “conventional end” of forgery. But we might describe a sculptor as “technically perfect”—capable, that is, of perfectly performing such tasks as the carving of marble in the likeness of a human figure—and yet deny the perfection of the works he produces. He lacks, we say, “inspiration” or “genius.” And on the other side, an artist might produce the works of art towards which his activity is directed without having perfected himself technically: technically rough, they yet delight the eye. A doctor might succeed in keeping his patients healthy only because they have an excellent constitution and are fortunate in the kind of work they undertake. Health can be a “natural gift,” as distinct from the product of the doctor’s skill, even when that skill is directed towards maintaining health. To sum up, whereas technical perfection entirely depends, as Kant argued, on talent and skill,10 men can achieve teleological perfection—the end towards which their efforts are directed—as a gift, or by luck, rather than as a result of effort.

This is a fact of great consequence for Christian theories of perfection. Aquinas took over the Aristotelian analysis of perfection and developed it systematically. Everything whatever, he says, moves by its own nature towards a particular condition, the condition in which it can rest. That condition is the thing’s “perfection.” The perfection of man as such, according to Aquinas, lies in the vision of God—or, more precisely, of the Divine Essence.11 But at the same time Aquinas argues that men can achieve that end only by the grace of God, not by the mere exercise of talents and skills. The vision of God is at once man’s natural end and a supernatural gift, just as it is the sculptor’s natural end to produce great works of art, and yet to do so requires a kind of inspiration which is something over and above the talents and skills he directs towards that end.* So technical perfection is not the same thing as teleological perfection, however closely the two are sometimes related to one another. A man can be technically perfect without attaining to teleological perfection; he can perform perfectly his religious and moral duties, so far as that involves the skilled use of his abilities, he can make himself expert in ritual and in Christian knowledge, without being vouchsafed the vision of God; and he can attain to that perfection without being technically perfect.

The presumption, as we saw, which lies beneath the concept of teleological perfection is that every kind of thing, including man, has a natural end in which alone it can find perfect satisfaction. There is another way of putting this metaphysical assumption. Everything, it may be suggested, contains unrealized potentialities; “becoming perfect” consists in actualizing the potentialities. A thing’s “natural end” is to actualize, to realize, its potentialities. So in his Critique of Judgement Kant suggests that what he had previously called “practical perfection”—task-perfection—is more accurately described as “utility.” To perfect oneself technically is to make oneself useful but not necessarily to perfect oneself. A thing perfects itself, he now says, only when it attains an end inherent in the thing itself, what it has it in itself to be, not merely an end which someone has chosen to set up as its objective.12

This explains some at first sight very odd philosophical remarks about perfection, as when Aquinas writes that “everything is perfect so far as it is actual” or Spinoza that “by perfection . . . I shall understand . . . reality.”13 The language, and the metaphysical presumptions, are Aristotelian. For Aristotle, what actually is must be better than what merely can be. To cite his examples, it is better actually to see than merely to possess the power of seeing, a building is better than the mere capacity to build.14 In each case the merely potential, according to Aristotle, is incomplete, formless, imperfect. The actual is “perfect,” then, in so far as it is the realization of, or the giving form to, a potentiality.

But there is something more than a little strange in thus identifying perfection with the realization of potentialities. Suppose a man is potentially a liar. When he actualizes that potentiality, has he thereby perfected himself? At this point, it is important to recall that the general concept of perfection does not have written into it any suggestion of moral excellence. A man can be a perfect scoundrel or a perfect idiot just as he can be a perfect saint; he can commit a perfect crime, be a perfect forger, or have a “perfectly rotten time of it.” But, as we have already pointed out, when we speak of “perfectibility,” as distinct from perfection simpliciter, the situation is different; to assert that man is perfectible is to assert that he can become, in some sense taken to be absolute, a better person. To the extent to which an analysis of perfection is directed towards helping us to answer the question whether human beings are, or are not, perfectible, it must not allow the response: “they are perfectible all right: there are plenty of men who are potential villains and who actualize that potentiality perfectly.”

If perfection is to imply becoming better, and yet is still to be defined in terms of the actualization of potentialities, it must be supported by a very special theory of evil. Of any actual person—let us say, a repulsive bigot—it has to be said that “in so far as he is actual he is perfect.” But what of his bigotry? This, it is then necessary to argue, is not actual. As Augustine puts the point: “Evil has no positive nature; what we call evil is merely the lack of something that is good.”15 Similarly, Descartes takes it to be self-evident that blindness and error are not “real”; they are, he says, simply the lack or absence of a power which we possess by nature.16 And Leibniz, so Bertrand Russell has suggested, makes his moral philosophy plausible only by moving backwards and forwards between a moral definition of perfection, for which evil is positive, and a teleological definition for which it is negative.17

On this view, then, the bigot does not actualize a human potentiality, he does not “realize his nature,” by his bigotry. Rather, he fails to realize his nature, since he is deprived of some good which is potential in it. All potentialities, then, are for good. It is more than a little surprising how often perfectibilists have taken this for granted; the “release of potentialities” is calmly identified with the release of potentialities for good. In everyday life, of course, it is not in the least degree paradoxical to say of some particularly nasty child: “he is a potential criminal” or, even more specifically, a “potential murderer.” Nor would it seem absurd, though the point might be disputed, to suggest that all men are potentially criminals. But if Augustine and Descartes and Leibniz are correct, all such judgements are mistaken: criminality is not a potentiality, capable of being actualized, but only a defect, the imperfect actualization of a potentiality. There are, indeed, a great many episodes in the history of perfectibilism which can only be understood by remembering that evil is assumed not to be “actual,” and potentialities all to be for good.

So far, we have distinguished three different modes of human perfection: technical perfection, which consists in performing, with the maximum efficiency, a specialized task; obedientiary perfection, which consists in obeying the commands of a superior authority, God or a member of the élite; teleological perfection, which consists in attaining to that end in which it is one’s nature to find final satisfaction. Abstractly separable, they may, in the writings of a particular perfectibilist—or anti-perfectibilist—be variously conjoined and variously disjoined. All three rely, in some degree, on the concept of a function, an allotted task, an end to be pursued, whether set by other men, by society, by God, or by Nature.

Indeed, the Greek word teleios, commonly translated as “perfect,” is etymologically related to telos (end)—the relationship between perfection and the achievement of an end is, as it were, written into it. The English word “perfect,” however, ultimately derives, by way of Middle English, from the Latin word perficere, the roots of which, in turn, are facere, “to make,” and a prefix per suggesting “thoroughly.” The perfect, that is, is etymologically definable as the “thoroughly made,” the “completed.” Between the definition of perfection in terms of ends and the definition of it in terms of the “thoroughly made” or the “complete” there are, of course, close links. If a thing is badly made or incomplete, it may, in consequence, be unable to fulfil its function. A pair of secateurs may be useless because a screw has been put in crookedly or a spring is missing; a doctor may be unable to cure his patients because his sight is poor.

It is not necessarily the case, all the same, either that what fulfils its function must be well-made and complete or that what is well-made and complete must fulfil its function. We can say of a house: “The workmanship is poor, and we have never bothered to have the terrace completed, but it suits our purposes admirably.” Or alternatively: “The workmanship is excellent, and the house is now complete, but as a place to live in, it is dreadful.” In such judgements, we make use of criteria of good workmanship and of completeness which are independent of our criteria for suitability to an end. We can judge an Etruscan artifact, a bronze statue, to be well-made, complete, and so far perfect, without believing that it ever fulfilled its intended function, whether as a fertility god or as a funereal consolation to the dead.

The perfect, it might therefore be suggested, is best defined in Newman’s manner as “that which has no flaw in it, that which is complete, that which is consistent, that which is sound.”18 Perfection is thus cut loose from any connexion with an end; one can simply look at a person and describe him as perfect by seeing him to be free of flaws. “Perfect,” in this sense, is an adjective applying to objects, to persons, to States, not (necessarily) to the performance of tasks or to the ends towards which those performances are directed. By what criterion, however, are we to determine whether a characteristic counts as a flaw? The idea of a flaw, as it is normally employed, is a relative one, not absolute. So the English, but not the Greeks, consider it a flaw in food that it is lukewarm; the English, but not the Japanese, consider it a flaw in a garden that it lacks flowers; the English, but not all African tribes, consider it a flaw in a woman’s beauty that she has pendulous ears. Not everyone would admit, even, that it is necessarily a flaw in a thing that it is unfinished, incomplete. “In everything, no matter what it may be,” writes Yoshida Kenkō in his thirteenth-century Essays in Idleness, “uniformity is undesirable. Leaving something incomplete makes it interesting, and gives one the feeling that there is room for growth. Some one once told me, ‘Even when building the Imperial Palace, they always leave one place unfinished.’”19

In Christian theology, however, it is ordinarily supposed that there are absolute flaws—moral and metaphysical. An important form of absolute moral flaw is sin. About the nature of sin there are no doubt disputes, disputes about whether, for example, the mere experiencing of an illicit sexual impulse constitutes a sin, or only the “entertaining” of the impulse. When one Christian moral theologian asserts, and another denies, that men can ever, in this life, be free of sin, they may be disagreeing about the nature of sin rather than about man’s moral capacities. But they would agree, at least, on one point: unless man can be sinless he cannot be perfect. So, in practice, disputes within Christianity about the perfectibility of man very often turn out to be disputes about the possibility of sinlessness. Let us call that sort of perfection which consists in freedom from moral flaw—including sin—immaculate perfection.

The idea of a “metaphysical” flaw is even more complex. Consider, for example, the following passage from Descartes, in which he sets out to show that a corporeal body cannot be perfect:

Descartes here takes it for granted that what in an immediately preceding passage he calls “simplicity and unity” are perfections, and that divisibility is an imperfection. Similarly, metaphysicians often presume that it is an imperfection in a thing for it to be in any way dependent for its existence on anything else, to be complex, to be finite, to pass away. The possession of such characteristics is taken to demonstrate that the thing is not really complete. When it has a cause, or when it has distinguishable parts one of which might be taken away from it, something essential for its existence lies outside it. An article on perfection in the Dictionnaire des Sciences Philosophiques21 will further illustrate the point:

The ideal of perfection, thus understood—metaphysical perfection or, as Kant calls it, “theoretic” perfection—removes it, we should at first be inclined to conclude, far beyond human reach. For how can a man set out to become less finite than he is, or more of a first cause, or less of a temporal being? What hope has he of becoming “a being to whom nothing is lacking”? But human ambition is boundless. And so, under the influence of this metaphysical ideal of perfection, men have set out to become more like a first cause, in the sense of not allowing themselves to be affected or influenced by anything which happens to them; they have sought to become less finite and less temporal by freeing themselves from all concern with the changing and by uniting themselves with God; they have tried to persuade themselves that they “lack nothing” by rejecting as worthless all that this world contains. The achievement of perfection has thus been identified with the development of a capacity for standing aside from life, rising above it to union with a Being, or a Universe, supposed to be infinite and eternal. For only such a life, it has been supposed, can be metaphysically flawless, metaphysically immaculate.

The difficulty that to define perfection in such exalted terms removes it far beyond human aspirations has, however, often been recognized by less mystically-minded, more practical, moralists. When absolute perfection is defined in a manner which makes it no longer meaningful to suppose that any finite being could be absolutely perfect, it is commonly supplemented by a doctrine of “relative perfection,” perfection relative to a humanly-attainable moral ideal, human as distinct from metaphysical perfection. The problem, then, is to determine what constitutes this ideal of perfect humanity.

Sometimes the ideal is made concrete as a person. Human perfection is taken to consist in imitating the example set by that person—exemplary perfection. For the Stoics, Socrates served as such an ideal; Christians have naturally turned to Jesus, considered as a human figure. The great problem, it has sometimes been supposed, is to choose between these two exemplary ideals: Joseph Priestley, for example, wrote a book called Socrates and Jesus Compared.22 Kant, however, objects to any definition of perfection by reference to examples. “Nor could anything be more fatal to morality,” he writes, “than that we should wish to derive it from examples. . . . Even the Holy One of the Gospels must first be compared with our ideal of moral perfection before we can recognize Him as such.”23 Why should this be so? Why should we not say, simply, that to be perfect is to be like such and such an exemplar, Jesus or Socrates?

The point Kant is making—it derives in the long run from Plato’s Euthyphro—is that perfection cannot mean “being like such-and-such a person.” For when we say: “Socrates is a perfect man” or “Jesus is perfect,” this is not the empty tautology, “Socrates is Socrates-like,” “Jesus is Jesus-like.” We adjudge Jesus and Socrates perfect by comparing them with our ideal of moral perfection; in calling them perfect we mean that they perfectly exemplify that ideal. But if we are in possession of such an ideal, then it is only sensible to judge our own conduct, too, by direct reference to it. Contemplating the life of a particular person, no doubt, may help us to conceive our moral ideal more concretely, more vividly. In the end, however, it is by reference to the ideal that we must determine our own, or anybody else’s, degree of perfection. “Examples,” so Kant elsewhere argues, “serve for our encouragement and emulation. They should not be used as patterns.”24 To regard them as an inspiration, only, will serve to prevent us from thinking that we are obliged to imitate some personal peculiarity of the persons we take to be ideal—Jesus’ habit of speaking in parables, for example, or Socrates’ habit of falling into trances.

But the problem Kant has here raised breaks out again at the level of ideals. We have still to decide between conflicting ideals of moral perfection, between, for example, the Buddhist ideal of disengagement and the humanistic ideal of involvement. How are we then to proceed? There is no higher ideal of moral perfection by reference to which men can decide whether their ideals of moral perfection are in fact perfect. Spinoza has argued, indeed, that the appeal to ideals is always arbitrary: we arbitrarily set up ideals—the perfect house, the perfect man—and then speak of things as perfect or imperfect in relation to these arbitrary notions of ours.25 Approached in this way, the question whether man is perfectible has, as it stands, no answer. “Perfectible,” we must ask, “in relation to whose ideal of perfection?”

Neither Plato nor Kant, however, would admit that ideals are arbitrary. For Plato, at least as he is commonly interpreted, ideals have an independent reality. Indeed, only the ideal is fully real. The triangle you or I might draw on a piece of paper is, in virtue of its imperfection, not a fully real triangle; the real triangle is the ideal triangle, the form of triangularity. Kant is not prepared, as he himself puts it, “to soar so high.” Ideals, he grants, do not have objective reality. But they nevertheless have, he says, “practical power.” They provide us with what he calls an “archetype,” they “form the basis of the possible perfection of certain actions.” “Although we cannot concede to these ideals objective reality (existence),” he writes, “they are not therefore to be regarded as figments of the brain; they supply reason with a standard which is indispensable to it, providing it . . . with a concept of that which is entirely complete in its kind, and thereby enabling it to estimate and to measure the degree and the defects of the incomplete.”26

But Spinoza’s objection is not easily set aside; the content of such standards has still to be decided. To take an example from physical culture advertisements: must the “ideal specimen of a man” have bulging muscles? If perfection is defined teleologically, the mode of answering that question is clear. The perfect specimen must have bulging muscles if, and only if, bulging muscles are essential to the body’s performance of its “true functions” or the pursuit of its “natural end.” No doubt, this answer only postpones the day of reckoning: for it has still to be determined in what the natural end of the human body consists. But at least it pretends to provide us with an objective test by which the content of perfection can be determined. If, in contrast, the “ideal” is divorced from the concept of a function—as it is, for example, in the persistent Greek presumption that to have “an ideal shape” a thing must be spherical—then we seem to be left with no objective method by which to determine what can properly be incorporated in the ideal, whether, for example, the spherical is in fact the ideal form or, as some Japanese aestheticians argue, a form to be avoided.

Often enough, as we shall see, moralists change their minds about the perfectibility of man only because the ideal changes. Much will depend, for example, on whether the ideal man is defined as one who loves nothing but God for its own sake or as one who cares, above all, for the greatest happiness of the greatest number. It is not surprising that the ideal of perfection is so often made to rest on metaphysical rather than on merely moral grounds; to identify perfection with moral perfection leaves wide open the question how it is to be determined in what moral perfection consists. There is some solace—if perhaps, in the end, little intellectual satisfaction—in the view that Nature or God has settled that question for us, once and for all.

In the light of his general theory we should certainly have expected Plato, above all men, to argue that there is an ideal of humanity which men should try to copy in their pursuit of perfection, and that they are perfectible in so far as they can succeed in this task. But a momentous passage in Plato’s Theaetetus—a passage which was to be much quoted in the centuries which followed—sets up God, not an ideal humanity, as the pattern on which man must model himself. “In the divine,” Plato there writes, “there is no shadow of unrighteousness.”27 A man perfects himself morally—which, to Plato, is identical with perfecting himself as a man—by imitating the divine righteousness. To ask whether man is perfectible, then, is to ask how far man can be “like God.” Aquinas carried this doctrine further by arguing that the perfection of all things consists in their being “like God.”28 Less metaphysically minded Christian philosophers have been content to define human perfection, and human perfection alone, in terms of likeness to God—deiform perfection. God, it is supposed, is at once a person and the ideal of moral perfection: a self-authenticating ideal in virtue of his supremacy. So there is no longer any problem in determining what man, or what ideal of man, is to be taken as the ideal. Man’s ideal form is God. Exactly how God can be imitated, however, is another question and one which, to put it mildly, theologians have not found it easy to resolve.

If, in the Theaetetus, Plato identifies moral perfection with being “like God,” a rather different ideal of perfection is suggested in his other dialogues, which identify it with harmony and order. This links it on the one side with immaculate perfection—freedom from flaw is defined as freedom from disorder—and on the other side with teleological perfection. Consider a “perfect clock.” Then clearly this must keep time perfectly, thus fulfilling its function. The teleological perfection of a clock is perfect timekeeping. To keep time perfectly its parts must be technically perfect. It must lack no parts and no part must be flawed, i.e. it must be immaculately perfect. But we could also define its perfection in still another way—by saying that the clock is a harmonious orderly arrangement of parts. Similarly, if we think of the soul as having parts—as Plato does in the Republic—then the perfection of the soul can be taken to consist in each of these parts harmoniously contributing to the perfection of the soul as a whole, playing its particular role in an ordered system. The nineteenth-century metaphysician Sir William Hamilton therefore defined perfection as “the full and harmonious development of all our faculties, corporeal and mental, intellectual and moral” and this came to be, for a time, the standard dictionary definition.29

It is in these terms that Plato defines the perfection of that ideal State, by reference to whose perfection the perfection of its individual citizens is to be determined. The ideal State is harmonious, orderly, stable, unified; the ideal citizen, by performing the tasks allotted to him, contributes to the total social harmony. Let us call this kind of perfection—the perfection of a system—aesthetic perfection. For it is often employed in the criticism of works of art, at its worst by critics like the German art historian Winckelmann for whom “perfection . . . was stately and harmonious form, almost anonymous in its regularity, unmarred by individual traits, frigid, devoid of emotions and showing no explicitly sexual characteristics.”30 I have already drawn attention to the links between aesthetic perfection, technical perfection, teleological perfection and immaculate perfection—aesthetic perfection involves the perfect performance of tasks in a flawless whole. It is no less closely related to metaphysical perfection: aesthetically perfect societies and aesthetically perfect works of art are often described in language which reminds us of the metaphysically perfect—as unified, immutable, self-sufficient.

Once more, however, these various ideals of perfection, closely associated though they may be in many perfectibilist theories, are separable—not only theoretically, but in practice. That is why it has been necessary to distinguish them and give them different names. No doubt, as Hume points out, we particularly admire an object which is at the same time elegant and capable of performing its function efficiently. But, as he also goes on to suggest, some degree of aesthetic imperfection may be necessary if a thing is adequately to perform the task for which it was designed.31 It was a leading tenet of the “functional” school of architects and designers that elegance and suitability for function must inevitably go together. But a chair which can be viewed with admiration as a harmonious shape is not necessarily the most comfortable of seats. More relevantly to our purposes, a society which values, above all else, unity, harmony, stability, may preserve an outward appearance of order only at the cost of suppressing human freedom and creative experiment. It may be by no means flawless, if we regard deception as a flaw; by no means perform its function, if we believe it to be part of the function of a society to foster enterprise. Plato’s own republic is, indeed, a case in point.

Let us pause now to draw together the threads. When we describe a man as “technically perfect,” we said, this is in relation to his efficiency in the performance of a task. But he cannot be technically perfect as a man; technical perfection applies only to a man’s performance in a specialised role. This is so, at least, unless we suppose either that men have a task set for them by a supreme legislator, that of obeying his commands (obedientiary perfection), or alternatively that there is a task inherent in their very nature, their perfection consisting in the achievement of an “end” which is “natural” to them (teleological perfection). By somewhat devious metaphysical routes, the theory of “natural ends” leads to the identification of the perfect with the actual; for a man to be perfect is for him to realize what he has it in him to become. This carries with it the conclusion that a good deal of what we ordinarily count as actual is not “really” actual: sin and evil are defined as negation or privation, as a “lack” rather than as an accomplishment. By means of such arguments, teleological perfection is identified with the actualization of potentialities.

A somewhat different approach to perfection defines it as the “complete” or “well made.” Perfection is then negatively defined (immaculate perfection) as the absence of flaw or imperfection—an imperfection sometimes identified with sin, sometimes with such metaphysical properties as complexity or self-sufficiency. Since in what is “complete” or “well made,” or “immaculate,” the parts fit together harmoniously, like the dove-tailed joints of a good piece of joinery, perfection is also defined as the harmonious working together of component parts (aesthetic perfection). Harmony and order are identified with perfection, conflict and disorder with imperfection. Finally, perfection may be defined in terms of conformity to a model, whether a model person (exemplary perfection), an ideal of moral perfection, or God (deiform perfection). God, in such instances, is usually taken to be both a person and an ideal—at once Plato’s “God” and Plato’s “form of the good.”

It will by now be clear that the question “Are men perfectible?” does not admit of any easy straightforward answer. The reply, often merely obstructive, is for once justified: “It all depends on what you mean . . .” To assert that man is perfectible may mean either:

• (1) there is some task in which each and every man can perfect himself technically;
• (2) he is capable of wholly subordinating himself to God’s will;
• (3) he can attain to his natural end;
• (4) he can be entirely free of any moral defect;
• (5) he can make of himself a being who is metaphysically perfect;
• (6) he can make of himself a being who is harmonious and orderly;
• (7) he can live in the manner of an ideally perfect human being;
• (8) he can become godlike.

Distinctions ought not to be made without necessity: the justification, if there is one, for distinguishing one from another these different ways of understanding perfectibility must be found in the story which follows. Two other possible sources of ambiguity should also be resolved. Christianity has sought to persuade men that they have two lives to live, one on this earth, the other in some infinitely more delightful, or inconceivably more horrendous, extra-terrestrial abode. When Christian theologians have denied, as they ordinarily have denied, that man is perfectible, what they have rejected as impossible is terrestrial, not celestial, perfectibility. The “perfectibility of man,” in fact, normally means his perfectibility on earth. Heavenly perfection enters our story only in so far as it has been invoked as an ideal standard, in relation to which every human achievement must be adjudged imperfect.

A second ambiguity. The question: “Is man perfectible?” can be interpreted either as asking whether any man is perfectible or as asking whether all men are perfectible. Greek perfectibilists, and such Christians as have been perfectibilists, ordinarily ascribe perfectibility only to a very few men, endowed with exceptional talents or granted an extraordinary degree of divine grace. Not a few moralists, however, have been dissatisfied with such relatively modest aspirations. Each and every human being—assuming only that he is normally constituted—is capable, so Godwin argued, of being perfected. The doctrine that man is perfectible has, in consequence, two forms. Particularist perfectibilism ascribes terrestrial perfectibility only to an élite; universalist perfectibilism ascribes it to all men. The context will usually make it clear which variety of perfectibilism is in dispute.

TWO

FROM OLYMPUS TO THE FORM OF THE GOOD

For the Homeric Greeks, the situation was clear. Not even the gods were perfect, if perfection entails freedom either from moral or from metaphysical defect. As the poet-sage Xenophanes complained, and Plato after him: “Homer and Hesiod have ascribed to the gods everything that is a shame and a reproach amongst men, stealing and committing adultery and deceiving each other.”1 Had Zeus laid down the commandment: “Be ye therefore perfect, even as your father in heaven is perfect” this would have been the sheerest hypocrisy, at least if the perfection in question is moral perfection.

In respect to metaphysical defects—or what Descartes, for example, took to be such—the Olympian gods were scarcely better off. They were, it is true, immortal and powerful, Zeus immensely so, but they were not self-created. They were born, even if sometimes by rather unorthodox mechanisms; they could suffer injury; there is not the slightest suggestion that they were indivisible or that they had an infinity of infinite attributes. Not even Zeus could stand alone in the Universe, a world complete in himself.

This is an important fact. There is a tendency nowadays to suppose it to be an a priori truth that what is divine must be perfect, both morally and metaphysically: that this is “part of what is meant” by being divine, or “part of the definition” of God. Sometimes this attitude is carried to the point of denying that the Olympian religion is “a religion at all.” That Homer should write of the gods as he does, in a tone which is unmistakably ironic, is particularly liable to shock and disturb the modern reader, persuading him that the Homeric gods were a mere literary convention who could not have been taken seriously, whether as objects of worship or as powers to be placated.

Such a view is clearly mistaken. It may be permissible to laugh at the gods; in a somewhat different sense, however, they have to be taken very seriously indeed. Some of the gods, it would seem, were more sensitive than others, more demanding, more insistent on their rights. But, in general terms, not to sacrifice to the gods, to neglect their rites, to profane their shrines, or to encroach in any way upon their privileges was to invite disastrous consequences, as Odysseus’ men learnt to their cost. There were deeds the gods would not brook: sacrilege, a breach of the laws of hospitality, the killing of a blood-relation. In fifth-century Athens, certainly, to deny their divinity was a crime; Anaxagoras was charged with impiety for declaring that the sun was not a god, but a stone. The fact remains that it was by no means impious, let alone a contradiction in terms, to deny that the gods were perfect.

Nor, equally, was it at all obvious to the Homeric Greeks that men ought to imitate the gods, even in so far as the gods did approach perfection. Quite the contrary. To set out to imitate the gods was to exhibit hubris, to be arrogant, to get above oneself. “Do not try to become Zeus,” the poet Pindar exhorted his readers. For, he continues, “mortal things suit mortals best.”2 In so far as men should guide their actions by models, it is the heroes, not the gods, on whom they should model themselves.3 If the heroes are perfect, however, it is only as exemplars of how to confront a particular situation. When, in the Odyssey, Athena exhorts Telemachus to take Orestes as his model, she does not suggest that Orestes was a perfect man; he is, rather, a perfect example of how a man ought to act in a particular situation, a perfect revenge-taker. His perfection is a technical perfection, not an immaculate perfection.

The gods, indeed, will not tolerate men thinking of themselves as perfect in any but this limited sense. That this is so may not be apparent in the Iliad, where both Greeks and Trojans are conspicuously vainglorious. By the fifth century, however, the doctrine that the gods are, in the words of Herodotus, “envious and interfering” was established as traditional orthodoxy. The gods, Herodotus depicts Solon as saying, are “envious of human prosperity.” Solon warns Croesus that “no man is ever self-sufficient; there is sure to be something missing.” And Herodotus goes on to suggest that Croesus suffered his dreadful fate “presumably because God was angry with him for supposing himself to be the happiest of men.”4 Perfect happiness is for the gods alone; to set oneself up as being godlike, whether in respect to happiness or any other respect, was, to this way of thinking, the surest way to ruin. As the chorus in Aeschylus’ Agamemnon warns us:

• In fame unmeasured, praise too high,
• Lies danger: God’s sharp lightnings fly
• To stagger mountains.5

There was, however, no metaphysical problem, as distinct from a lack of common prudence, in a man’s setting out to imitate the gods, in the sense in which there certainly is a metaphysical problem in understanding how any finite being could imitate the metaphysically perfect God of Christian theology—eternal, infinite, unchanging and indivisible. A passage from one of Pindar’s Odes is in this respect instructive. Men and gods, he says, both derive from the same mother, the earth-goddess Gaia. In intelligence and even in strength men therefore resemble the gods. If in comparison with the gods man is “as nothing,” this is only because he lacks their power, and the security which derives from their immortality.6

Since they are of the same race, should some mortal attract the roving fancy of a god or goddess, they can interbreed—although it is rash, as Ixion discovered to his cost, for mortals, even royal mortals, to take the initiative in such matters. The same passions inspire both men and gods; Eros can shoot his darts at either. In their everyday behaviour, then, the gods can be imitated. To do so, however, was certainly not the path to moral perfection. Paradoxically, it is only after philosophers and theologians had attempted to set up a God who would be more worthy of imitation than the Olympian pantheon that the question arose how it is possible for men to imitate God.

Could man find perfection, if not by imitating the Olympian gods, then at least by obedience to them, by subservience to their will? That, too, was a difficult attitude for the Olympian religion to adopt, seeing that the will of the gods was seldom unanimous. In Plato’s Euthyphro, Socrates draws attention to this problem. Euthyphro has defined piety as “doing that which is pleasing to the gods”; at the same time, he claims that he has acted piously in prosecuting his father for being responsible for the death of a slave. As Socrates points out, Euthyphro’s action might well be “welcome to Zeus but hateful to Cronos or Uranus, and pleasing to Hephaestus but hateful to Hera.”7 Euthyphro’s own action, that is, demonstrates how impossible it is to act in such a way as to please all the gods. To take another example, although in Aeschylus’ trilogy, the Oresteia, Orestes obeys the express command of Apollo in killing his mother, that is no excuse in the eyes of “the servants of the old gods”—the Eumenides.

No doubt, the will of Zeus is supreme,* and if men knew what Zeus had decided, they would know what they had to do. But there is not the slightest indication that the will of Zeus seeks to perfect those it works upon. The events recounted in the Iliad, Homer begins by remarking, express the will of Zeus, but they do nothing to perfect the nature of Homer’s human—or indeed, his divine—characters. If Helen shows signs of remorse, there is no suggestion that this is Zeus’ doing.

Man, in relation to the gods, is often enough represented, indeed, as a mere victim: “Zeus controls the fulfilment of all that is,” wrote the poet Semonides of Amorgos in the seventh century bc, “and disposes as he will. But insight does not belong to men: we live like beasts, always at the mercy of what the day may bring, knowing nothing of the outcome that God will impose upon our acts.”8 In Sophocles’ Women of Trachis the only explanation which is offered of Hercules’ suffering is that Zeus chose to impose it upon him. “Mark the malevolence of the unforgiving gods.” Thus it is that the play concludes. Perfection, so much is clear, was not to be sought in obedience to the will of such gods as these.

It is not surprising, then, that one can discern in sixth-century Greece a growing dissatisfaction with the Olympian religion—related perhaps to a fundamental change in the nature of Greek society, the decline of the heroic age. Men were no longer content with the typically military view that it is men’s task to do or die, to endure or fight, not to question or understand. Their dissatisfaction might, in principle, have taken either of two forms: a moralizing of the traditional religion, which, while preserving the humanity of the gods, would more strongly emphasize that they, and especially Zeus, are on the side of goodness or, alternatively, a rejection of the Olympian pantheon in favour of some quite different conception of the divine.

It did, in fact, take both forms. In the plays of Aeschylus, for example, there is an obvious “moralizing” of the gods. No doubt, his Agamemnon depends for much of its dramatic force on the conception of hubris. On his return to Mycenae, Agamemnon is deliberately enticed by Clytemnestra to tread on a carpet suitable only for the gods and this prepares the way for his murder. But a different doctrine is also invoked, according to which human suffering has a moral purpose. “Man,” intones the chorus, “must suffer to be wise.” Prosperity, it is suggested—as an admittedly unorthodox doctrine—need not provoke calamity unless the prosperity is accompanied by sin: “sin, not prosperity, engenders grief.”9

But the attempt merely to reform the Olympian religion was, in the very long run, destined to fail—although it is worth observing that its general effect was to make Zeus a figure more readily assimilable to, or replaceable by, Jahweh. Jahweh, like Zeus, made no pretence to metaphysical perfection, but he is depicted as being a righteous as well as an all-powerful God, even if his righteousness, as in his dealings with Job, sometimes takes a form quite incomprehensible to man. For our immediate purposes, however, the revolutionaries were more important than the reformers.

One important source of the religious revolution was the new speculative cosmology, traceable back to the teachings of Thales at the beginning of the sixth century. Precisely what Thales meant when he pronounced that “All things are full of gods,” it is now next-to-impossible to determine. But this consequence, at least, seems to follow: things can be divine which are not at all like human beings, which are not “of the same race as man.” They can be divine, in the sense of being the supreme powers, ultimately responsible for things happening as they do. So, in order to understand, for example, what supports the earth, it is not necessary to suppose that it is held in position by a god; Thales’ own conjecture was that it floats on water.

In Thales’ successor, Anaximander, divinity is ascribed particularly to to apeiron—the “Boundless” or “Indefinite.” The Boundless, indeed—except in respect to its material extension—begins to look very like what Pascal was to call “the God of the philosophers and the scientists,” as distinct from “the God of Abraham, God of Isaac, God of Jacob.” Aristotle sums up Anaximander’s view thus: “Of the Boundless there is no beginning . . . but this seems to be the beginning of the other things, and to surround all things and steer all. . . . And this is the divine; for it is immortal and indestructible.”10 The Boundless, that is, takes over from the Homeric gods their power, their immortality, their indestructibility. But it is everlasting, not merely immortal—it has no beginning, whereas the gods were born. All things begin from it, and it controls all things; it has the power, that is, of a monotheistic God, not of a polytheistic god who must share his power with other gods. In short, it is like “the God of the philosophers” in its freedom from what philosophers were later to think of as “metaphysical defect.” If it is not simple, it is at least everlasting and uncaused. On the other hand, it does not in the least resemble the God of Abraham, Isaac and Jacob. There is no point in praying to the Boundless, or making sacrifices to it; if the Boundless is free of moral defect, this is only because moral predicates have no application to it. It is neither righteous nor sinful, good nor evil.

In the satirical poetry of Xenophanes the impact of these new cosmologies on the old Olympian religion is made explicit. Xenophanes, no doubt, was first and foremost a moralist. As we have already seen, he condemned Homer and Hesiod for having ascribed to the gods “everything that is a shame and a reproach amongst men.” It is not unlikely that he influenced the “reformed theology” of Aeschylus. But Xenophanes was a revolutionary, not merely a reformer. Men everywhere, he says, have constructed gods in their own image: “the Ethiopians say that their gods are snub-nosed and black, the Thracians that theirs have light blue eyes and red hair.” If horses were capable of representing their gods in the form of images, he sardonically remarks, they would no doubt make their images look like horses, giving the gods bodies like their own bodies.

This anthropomorphic approach to the divine, Xenophanes argues, must be totally abandoned. In place of Homer’s human-all-too-human Zeus, Xenophanes sets up “One god, greatest amongst gods and men, in no way similar to mortals either in body or in thought.” Whereas Homer’s Zeus left Olympus on a variety of missions and employed messengers and physical agents to work his will, Xenophanes’ supreme god has no need either of travel or of subordinates. “Always he remains in the same place, moving not at all; . . . without toil he shakes all things by the thought of his mind.”11 And this for a particular reason: it is not fitting for him to move about the universe.* Xenophanes’ conception of a supreme God, then, is governed by ideas about what it is “fitting” for God to do. God is no longer just a power: he is, it is presumed, morally and metaphysically perfect. He will do nothing that is “not fitting”; this includes what is not metaphysically fitting as well as what is not morally fitting. Zeus is no longer to be depicted as committing adultery, not only because adultery is, morally speaking, “a shame and a reproach,” but also because adultery implies both a degree of humanity and a degree of restlessness not appropriate to God.

Greek historians of philosophy, from Aristotle onwards, described Xenophanes as the founder of the Eleatic school of philosophy, and in particular, as the teacher of Parmenides. Modern scholars generally reject this view.12 Yet there is a kind of rough justice in the Aristotelian “placing” of Xenophanes, even if little historical exactitude. From our present point of view, particularly, it is natural to think of Parmenides as carrying further Xenophanes’ metaphysical perfecting of the divine.

Although the supreme God of Xenophanes is not like man either in bodily form or in his manner of thought, he still operates in ways which are at least analogous to human modes of action. God has neither eyes, nor ears, nor a brain, but the fact remains that “all of him sees, all thinks, and all hears.”13 He acts without toil, but he still acts. Parmenides, in contrast, ruled out all these ways of talking: the Parmenidean “Being” neither sees, nor hears, nor thinks, nor acts. So far Being is like Anaximander’s Boundless; but unlike the Boundless, it must not be described as a power, as acting. “One way only is left to be spoken of,” Parmenides concluded, “that it is.” On our path to realizing this, however, we also come to see that Being is “uncreated and imperishable, for it is entire, immovable and without end.”

Nor is Being only “imperishable” in the sense of everlasting. In Parmenides the eternal is for the first time distinguished from the everlasting. Being is eternal in the full sense that “it was not in the past, nor shall it be, since it is now, all at once, one, continuous.”14 Simple, indivisible, eternal, lacking nothing, devoid of all properties which involve negation or defect—for of negation, “what is not,” it is impossible, so Parmenides argued, either to think or to talk—the Parmenidean Being represents the ideal of metaphysical perfection in the purest form it had so far assumed.

The modern reader may find it more than a little odd, when, having described Being in these terms, Parmenides goes on to say of it that it is spherical, “like the bulk of a well-rounded sphere.”15 He is liable to complain, like the spokesman for Epicureanism, Velleius, in Cicero’s The Nature of the Gods, that he sees no reason for believing that the sphere is any more “perfect” than any other geometrical figure. However, the Greeks generally took it for granted that sphericity was essential to perfection. In the Timaeus Plato has no doubt that a Demiurge-creator, setting out to construct a universe which is as perfect as it can be, will make it “rounded and spherical, equidistant every way from centre to extremity—a figure the most perfect and uniform of all; for he [the Demiurge] judged uniformity to be immeasurably better than its opposite.”16 As the Stoic Balbus replies to Velleius in Cicero’s dialogue: “What can be more beautiful than the figure that encircles and encloses in itself all other figures [since all the solid figures can be inscribed in a globe], and that can possess no roughness or point of collision on its surface, no indentation or concavity, no protuberance or depression?”17

This is an aesthetic sense of perfection. In the manner we normally take to be typical of classicism Parmenides and Plato and Balbus all presume that completeness, uniformity, evenness, regularity are obviously perfections, and that their opposites, the cragginess, irregularity, diversity so loved by romantics, are obviously aesthetic flaws.* To ascribe sphericity to Being, therefore, is to suppose it to be not only metaphysically but aesthetically perfect.

Even after this explanation we may still be puzzled by the sphericity of the Parmenidean Being. Is not to call Being a sphere at once to think, as Parmenides forbids us to do, of something as lying outside Being? But the Greeks, in a manner which was to have disastrous consequences for astronomy, were greatly struck by the fact that the circle is a continuous line. “The circle is of all lines,” as Aristotle puts it, “the most truly one, because it is whole and complete.”18 And they were no less struck by the fact that the sphere is the only solid body which can move—by revolving—without requiring any space outside of itself to move into. Aristotle calls the sphere “perfect” just in virtue of this fact. So the spherical nature of Being is presumed not to detract from but rather to emphasize its perfection, in the sense of its all-embracing completeness.

Now, if this perfect, spherical Being is identified with God—and Parmenides seems to think of it in this way, to judge from the religious atmosphere of the poem in which he formulates his philosophical ideas—men cannot possibly imitate God or take him, or it, as a moral ideal, any more than they can take Anaximander’s Boundless as an ideal. This conclusion was, of course, a consequence even of Xenophanes’ theology. For how can men imitate a God “in no way similar to mortals, either in body or in thought”? But it is even more obvious that no man can set out to imitate Parmenides’ Being. As for praying to Being or seeking its aid, that would be wholly absurd.

It might therefore be supposed that the Parmenidean Being would be automatically rejected, as obviously irreligious, by the adherents of any kind of religion for which God is an object of prayer. In fact, however, particularly by way of Plato’s much-misunderstood discussion of Parmenides’ Being—as “the One”—in his dialogue Parmenides, Being came to be regarded as a sort of philosophical preview of the God of Christianity. So the Christian Boethius, writing his Consolation of Philosophy in the fifth century ad, felt free to describe God in what are explicitly Parmenidean terms: “For such is the form of the Divine Substance,” he writes, “that it is neither divided into outward things, nor receives any such into itself, but as Parmenides says of it: “in body like a sphere well-rounded on all sides,” it rotates about the moving orb of things, while keeping itself immovable.”19 It is one of the most extraordinary facts in the history of that most extraordinary phenomenon—Christian theology—that Parmenides’ Being and Plato’s “One” should be thus identified with the Christian God.* But the identification arises naturally enough, in a sense, out of the attempt to set up a God who would be metaphysically, as well as morally, perfect, when this is combined with a definition of perfection as freedom from any sort of negation (evil, as we have already seen, was defined as a special variety of negation). Of such a God we are obliged to say not only that he is characterizable by all those attributes which do not involve negation, but that these characteristics themselves are ultimately identical. For to distinguish is to negate, to say of one property, or of one part, that it is not another. (It is interesting to observe that Hindu religious thinkers reached the same conclusion: “As a unity only is it to be looked upon,” according to the Brihadāranyaka Upanishad, “—this indemonstrable, enduring Being.”)20

So Aquinas, having begun by arguing that God has to be immobile, eternal, simple, simultaneous in his existence—it is interesting to observe that he quotes Boethius on this point—not possessed of a body, infinite, and possessed of all perfections, goes on to maintain that all these perfections must, in the end, be the same perfection. “God is simple,” Aquinas writes, “but where there is simplicity, there can be no distinction among the perfections that are present. Hence, if the perfections of all things are in God, they cannot be distinct in Him. Accordingly, they are all one in Him.”21 No doubt men ascribe various names to God, but that is only because, Aquinas argues, they are in this life restricted to what Parmenides called “the Way of Seeming.” When men come to see God as he really is, when they attain to the vision of the Divine Essence, their idea of God will be as simple as is God himself. And the essence of this metaphysically simple God is nothing other than his existence. It will be obvious how close this brings Aquinas’s God to the Parmenidean Being, of which we can ultimately say only that “it is.”22 Of course, not only the great figures of Plato and Aristotle but Philo, Plotinus, Proclus and that ardent expropriator of neo-Platonism in the name of Christian theology, Dionysius the Areopagite, have intervened between Aquinas and Parmenides. Aquinas is quite convinced that, with Aristotle’s help, he can take over a Parmenidean-type God while avoiding a Parmenidean-type monism. But that he can in fact do so, not all his critics have been convinced.

Parmenideanism was the culminating point in the revolutionary transformation of the Olympian religion by pre-Socratic philosophers. The Olympian religion, however, by no means exhausted the religious life of Greece. The cult of Dionysus, uniting God and man in the ecstasy of orgy, had only a loose association with the Olympian pantheon; the mystery rites, especially at Eleusis, involved a relationship with the divine much more intimate than the Olympian religion allowed. The initiated were promised, after death, entrance to the world of the gods. However widely it may have been believed that human beings ought not to aspire to be godlike, not everyone was prepared to accept so modest an estimation of what was humanly desirable.

In Empedocles’ fifth-century poem, significantly entitled Purifications, a most un-Olympian view prevails. Man, according to Empedocles, is a demi-god, who, at the beginning of human history, committed a crime, involving the shedding of blood, for which all men since have had to pay the penalty.* Banished from their proper home among the gods, men must live, as a consequence of their guilt, in cycles of reincarnation, life after life defiled by sin until, by the exercise of purifying virtues, they finally return to earth, like Empedocles himself, as “prophets, bards, doctors or statesmen.” Then at last they escape from the cycle: “I go about among you all,” Empedocles therefore writes, “an immortal god, mortal no more.”23

To what extent is Empedocles expressing views which circulated generally in Greece? This is a much disputed question. According to some scholars, there arose in fifth-century Greece a distinctive religious movement, the Orphic religion, with sacred books and dogmatic teachings, whose adherents, setting themselves apart from the Olympian religion, lived in Orphic communities, much as the Christians were to live in Rome—although with a greater willingness to make the traditional formal sacrifices to the Olympian gods. More sceptical scholars have argued that what is commonly called “Orphism” is at most an inchoate mass of popular beliefs, corresponding roughly to that body of superstition which persists in Christian communities and may, in some degree, affect the beliefs and the conduct of the most ardent Christians. Its supposed doctrines, they suggest, are in large part the invention of a much later day, of Christian Platonists and of neo-Platonists, constructing poetic myths for their own polemical purposes.24

Let us therefore say only this, that there circulated in fifth-century Athens, more or less widely diffused, a set of beliefs of which the most important ran somewhat as follows: the human being is latently immortal and divine; his latent divinity can be made manifest only if he purifies himself by ritualistic and ascetic practices. With this belief may have been associated a mythology, according to which man’s nature is flawed and his divinity obscured as a consequence of his origins. The Titans dismembered and ate the god Dionysus; Zeus struck them dead with his thunderbolts and lightning; out of the ashes sprang the race of men.

So man is at once divine and earth-born. It is now his responsibility to free himself, by purifying rites, from the element of earth he owes to his Titan origins and from the guilt arising from his share in the crime the Titans committed. Men can attain to such freedom from mortality and guilt only through a cycle of reincarnations, somewhat in the manner of Buddhist transmigration, from which they can finally escape only when they are sufficiently enlightened to see the need for purifying themselves. Then they become immortal gods. Whereas the Buddhist sects disagree amongst themselves about whether immortality or annihilation is the final consummation, for Orphic-type religions it was clearly the first. Godlike perfection, then, by no means lies beyond the reach of man: and purification is the path to it.

If it is a controversial question exactly in what Orphism consists, it is even more puzzling to know how to relate Orphism to the teachings of Pythagoras, which Herodotus explicitly linked with Orphism.25 The Pythagoreans formed, it would seem, a kind of secret society; there were penalties for revealing Pythagorean doctrines to the uninitiated. To make matters worse, the followers of Pythagoras had the amiable, but historically disconcerting, habit of ascribing their own teachings to their master. Pythagoras himself wrote nothing, nor did his successors for some generations. Not surprisingly, then, it is difficult, almost to the point of impossibility, to determine what the historical Pythagoras taught.26

Some of the teachings which have come down to us as “Pythagorean” suggest the leader of a mystery religion rather than a philosopher—rules of prohibition, for example, which forbid his followers to step over a cross-bar or to allow swallows to nest under their roof. Yet there are also contemporary references to Pythagoras as a scientist, particularly a mathematician, with the strong suggestion that for him religion and mathematics were somehow allied. Amongst his successors, certainly, men proud to call themselves Pythagoreans, there were great mathematicians and great astronomers.

The two sides to Pythagoreanism, religious and philosophical, were linked by means of a set of ideas which were to be of very great importance in the subsequent history of perfectibility: the idea of contemplation, the idea of order or harmony, and the idea of purification by wisdom. Before the time of Pythagoras, sophia (wisdom) had no particular connexion with contemplation. It was possible to talk, as Homer does, of the sophia of a shipbuilder, as a way of referring to his practical ability. For Pythagoras, in contrast, sophia consists in looking on at the world: to him is ascribed—and although the ascription is doubtful, the general sentiment is certainly Pythagorean—the comparison of life to the festival-games at which “some come . . . to compete, some to ply their trade, but the best people come as spectators.”27 Here are the roots of the doctrine, which Aristotle was later to espouse, that the life of God is a life of pure contemplation—not a form of activity to which the Olympian deities were conspicuously devoted. The contemplation the Pythagoreans thus extolled is a contemplation of the order of the universe, and especially of its mathematical order. With the help of such contemplation, the soul, identifying itself with the order of the Universe, could purify and perfect itself and thus emerge from the cycle of transmigration.

If this account of Pythagoras—based on evidence, it must be confessed, which is both insubstantial in quantity and dubious in quality—is anything like the truth, then one sees how Pythagoreanism, Orphism and the teachings of Empedocles at once converge and diverge. For all three, the soul is involved in cycles of transmigration, from which it can escape only by purifying itself. But for the Orphics and for Empedocles, as not for Pythagoras, that cycle originates in a primeval deed, the guilt of which all men bear. All three agree that to escape from the cycle some form of bodily abstinence is essential. But for Empedocles and Pythagoras, as not for the Orphics, one must also become a person of a rather special sort—according to Empedocles, a poet, or prophet, or statesman, according to Pythagoras, a philosopher. And for Pythagoras alone, men are purified by contemplation of an ordered universe, achieving through that contemplation an orderly perfection in their own mind.

Now we are in a better position to understand the first systematic theory of perfectibility. This was worked out by Plato under Orphic-Pythagorean influence. But it was greatly influenced, too, by the Parmenidean analysis of metaphysical perfection and the Socratic identification of goodness and knowledge—the Socratic presumption that no one does evil willingly. As a deduction from Plato’s general metaphysics, we should expect to find him arguing that no particular man can perfect himself, that only the ideal form of humanity can be perfect, just as he argues that no particular work of art, but only the ideal form of beauty, can be perfectly beautiful, or that no particular triangle, but only the ideal form of triangularity, can be perfectly triangular. But in the Phaedo Plato avoids this conclusion by making a sharp distinction between body and soul. The soul, he suggests, is more like a form than it is like a particular; in virtue of this fact a human being is perfectible.

What does “being more like a form” involve? The forms, so far as this is compatible with there being more than one of them, exhibit that metaphysical perfection which Parmenides took to be peculiar to Being. Whereas particular works of art—a particular painting, for example—may fade away, the form of Beauty, Beauty-in-itself, Plato tells us, can never be destroyed, nor can the form of Triangularity-in-itself or the form of Goodness-in-itself. The forms, that is, are eternal—or, at the very least, everlasting—and this, above all, is why Plato does not hesitate to call them divine.

That ordinary things—particulars—pass away is, he suggests, a necessary consequence of the fact that they are complex, composite, changeable. The forms, therefore, as eternal, must be simple, indivisible, unchangeable. The soul is “like the forms” precisely in respect to these metaphysical perfections. “The soul,” Plato says, “most closely resembles the divine and immortal, intelligible and uniform and indissoluble and ever-unchangeable, while the body most resembles the human and mortal, the unintellectual and dissoluble and ceaselessly changing.”28 When Plato speaks in such contexts of “the soul” he does not mean the form of the soul—the soul-in-itself. It is each individual soul which possesses these perfections, or “something like them.”

The conclusion might seem to follow that there is no question of perfecting the soul; it is already perfect, by its very nature. But the soul, Plato argues, does not ordinarily exhibit, in this life, its full perfection. This is because it is tainted by the body in which it is “imprisoned”—a metaphor he ascribes to Orphic origins.29 The body, itself metaphysically imperfect, impermanent, mutable, corrupts the soul and drags it down to its own level.

This is a view which has obvious logical difficulties, partly concealed by Plato’s use of the phrase “like the forms.” In order to deduce the soul’s immortality from its likeness to the forms, he has to treat this likeness as rather more than likeness, and as a likeness which applies equally to all souls. If souls have only some resemblance to forms, then it would not follow that they share with the forms the particular property of not passing out of existence—this could well be one of the points on which they differed. And if only perfected souls are like the forms, then the conclusion would seem to follow that only perfected souls are immortal. So Plato’s interest in demonstrating the immortality of the soul leads him to insist on the close resemblance of each and every soul to the forms—indeed, to describe souls in the very same language he applies to the forms themselves.

On the other hand, his desire, as a moralist, to set before men the care of their souls as a moral task leads him to describe the soul not as unchangeable but as capable of being corrupted, not as simple but as capable of being “interpenetrated by the corporeal.” More than likely, it was Plato’s own uncertainty about the soul’s nature which led Renaissance Platonists to deny that the soul has any fixed metaphysical nature, to think of it as capable of becoming either something like the divine or something like a material object, but as in itself not belonging to either category. For Plato, however, the soul is naturally divine, even if it can only too easily be corrupted. So perfection is a task, but a task in which success can be hoped for—it is a task because most souls are not in fact morally perfect, it is a task in which success can be hoped for because the soul’s metaphysical nature allies it with the perfect.

Since the unwillingness of the body to submit to its authority is the principal cause of the soul’s imperfection, the path to perfection involves some measure of asceticism. The philosopher, so Plato argues, will not be greatly concerned with “such so-called pleasures as eating and drinking,” or “the pleasures of love,” or the adornment of the body.30 This asceticism, however, by no means involves the mortification of the body; there is in Plato no suggestion that in order to achieve perfection the philosopher should, as Diogenes the Cynic was to do, live in a barrel, or that like the Christian hermits he should cut himself off from society, or that he should fast to the point of malnutrition, or that he should entirely eschew all sexual relations. The Socrates who in the Phaedo is depicted as bidding the philosopher not to devote himself to bodily pleasures is the very same Socrates who is a convivial guest in the all-night drinking party described by Plato in the Symposium. Simply, the philosopher is not “much concerned” with eating and drinking, or sexual relations, or personal adornment.

The crucial point for Plato, furthermore, is that the soul frees itself from the body not by asceticism as such—let alone by ritual incantations or ceremonial rites—but by acquiring knowledge. And knowledge, for Plato, consists in the apprehension of relationships between ideal forms; it is not to be acquired by any kind of empirical observation, involving the use of the bodily senses. “She [the soul] reasons best,” he writes in the Phaedo, “. . . when most she is alone and apart, paying no heed to the body.” When the soul tries to make discoveries by means of sensory observation “she is torn away by the body into the region of constant fluctuation, and she herself wanders about in confusion, reeling like a drunken man, because of her contact with things in similar confusion.” When, on the contrary, the soul relies entirely on her own resources, forgetting the body, she “remains ever constant and changeless with the unchanging, because of her contact with things similarly immutable.”31

What Plato says at this point obviously rests on the presumption that the human being becomes like what he perceives. There is, one might object, no reason why an orderly soul should become disorderly merely as a result of contemplating the disorderly, or a disorderly soul become orderly merely as a result of contemplating the orderly. But Plato, like the censors of our own time, thought otherwise. This presumption is even more important in the Republic where Plato introduces a supreme form, the form of the good, the knowledge of which he takes to be a sufficient and necessary condition of all wisdom. That it is a necessary condition, he explicitly argues: “This form of the good must be seen by anyone who hopes to act wisely, either in public or in private.”32 That it is a sufficient condition he takes for granted: to know the form of the good is to be good, indeed to be perfect.

It should be observed, however, that in the Republic he speaks of the philosopher not only as a knower, but as a lover: philosophers are, he says, “lovers . . . of that reality which always is.”33 And this is suggested by the etymological root of the word “philosopher,” its derivation from philein, to love. He reaches the conclusion that by “associating with what is divine and ordered [a philosopher] becomes ordered and divine as far as mortal may,” by relying on the principle that a man naturally imitates “that with which he lovingly associates.”34 Love and knowledge are, in Plato’s thinking, closely associated, even if sometimes his stress is on knowledge and at other times, as in the Symposium, on love. So his fundamental presumption might be more plausibly put by saying that we become like what we lovingly-know; that presumption we shall observe, too, in Christian mysticism.

Plato described the form of the good in terms which readily lend themselves to a mystical interpretation. It is, he says, “the cause of knowledge and truth . . . but while knowledge and truth are both beautiful, you will be right in thinking it other and fairer than these.” “It . . . transcends,” he also writes, “even Being in dignity and power.”35 These are among the most influential statements in the whole history of human thought—the most disastrously influential, we might say, considering how often they have been used to lend philosophical respectability to beliefs which are in spirit totally anti-philosophical.

In neo-Platonic thought, and in the writings of such Christian theologians as were influenced by it, the form of the good was not uncommonly identified with God. This is not altogether astonishing, considering that Plato described the form of the good as “the cause of all that is right and beautiful in all things.”36 Nor is it, I suppose, wholly surprising that the form of the good was also identified with the supreme Beauty of the Symposium. But what is surprising, on the face of it, is that the form of the good was also identified with “the One” of the first Hypothesis of Plato’s Parmenides. For the context makes it perfectly clear what Plato was doing when he wrote: “The One neither is one nor is at all. . . . It cannot have a name or be spoken of, nor can there be any knowledge or perception or opinion of it. It is not named or spoken of, not an object of opinion or of knowledge, not perceived by any creature.”37 He was constructing a reductio ad absurdum, not defending “negative theology.”

Yet if, as a scholar, one finds it surprising that Plato should have been thus interpreted, from another point of view it is not at all surprising. For what resulted from the conflation of the form of the good, the One, and Beauty was the conception of a Being in all respects perfect: perfect in beauty, perfect in goodness, perfect in divinity, perfect in every metaphysical property. The Platonic forms, as they were represented in the Phaedo, had, from the point of view of the perfectibilist metaphysician, one great disadvantage: they were a multiplicity. Even in the Republic, where Plato came as close as he ever did to a Parmenidean monism, it is not suggested that the form of the good has the kind of absolute metaphysical superiority possessed by the Parmenidean One.38 In none of his dialogues, indeed, does Plato show himself fully prepared to put all his metaphysical eggs in one basket, by subordinating, for example, the perfection of the forms to the perfection of God. Only by taking that step for him, by identifying God, the form of the good, Beauty and the One, could his authority be claimed for so absolute a metaphysical amalgam.

The resulting inconvenience—a supreme Being which “is not named nor spoken, not an object of opinion or knowledge”—was easily borne by mystics, and even by theologians whom one might have expected to rebel against it. So just as, by an exaggeration of a certain tendency in his teachings, Plato came to serve as a principal authority for ascetic doctrines of perfectibility, he also came to serve as a principal authority for mystical doctrines of perfectibility, according to which perfection has to be sought by rising above the body, above any sort of knowledge or truth or being, to a God describable only in negative terms. Mysticism and asceticism were, indeed, often allied: asceticism was conceived of as the first stage on the process towards mystical perfectibility. The renunciation by the ascetic of his material possessions, his desires, his attachments to the world, served as a symbol of his final renunciation of his individuality, of all those predicates which made of him a distinctive human being.

For Plato himself, however, the situation was very different. For one thing, mysticism is essentially a doctrine of individual perfection. Individual perfection is no doubt the kind of perfection principally emphasized in the Phaedo, concerned as it is with the situation of a man confronting that essentially solitary experience—death. But even in that dialogue, Socrates is depicted as delighting in, and learning from, the company of his fellow-men. There is no sign whatever of the ascetic misanthropy of the early Christian Sayings of the Fathers; it is impossible to imagine Plato writing: “A man who avoids men is like a ripe grape. A man who companies with men is like a sour grape.”39 In the Republic, furthermore, Plato makes it perfectly clear that on his view no man can achieve perfection except by way of a perfect society, a society ruled by philosopher-kings. “Neither city nor constitution,” he writes, “and not even any individual man, will ever be perfect, until philosophers . . . are in some way compelled to take charge of the state . . . or until those who now have charge of dominions and kingdoms are inspired by the love of true philosophy.”40

Granted that the ideal city does not exist, the man who tries to perfect himself will nevertheless, so Plato suggests, try to live his life by its laws: the ideal city “is laid up in heaven as a pattern for him who wills to see, and seeing, to found a city in himself.”41 But only if the city can be actualized, it would seem, can men find the absolute perfection they are seeking. At the very least, this approach to perfection requires that men should think of themselves as citizens, not as engaged in a solitary pursuit of perfection.

No less important was Plato’s insistence on education. This, above all, distinguishes him from the Christian mystic. Even if the final contemplation of the form of the good is, on Plato’s view, a direct vision which defies all description and all analysis, the fact remains that in order to achieve that vision men, on Plato’s view, have first to undergo a long and rigorous philosophical education. Education, not prayer and self-mortification, is the path to perfection.

Such an education, Plato adds, is not for all men. “A multitude cannot be philosophical.”42 Most men cannot achieve what Plato significantly calls “philosophical” goodness, for him the only true goodness. The most they can hope for is perfection in “civic” goodness, the sort of goodness they can acquire by obedience to the laws laid down for them by the philosopher-kings, by those rulers, that is, who in virtue of their knowledge of the good are capable of philosophical goodness. Men who live civically good lives, Plato suggests in the Phaedo, may return to earth, in a subsequent life, as “bees . . . or ants”—symbols of industry and order. They have no prospect, however, of joining “the company of the gods.”43

Both Plato’s emphasis on education as a means to perfection and his distinction between a perfectible élite and a non-perfectible multitude were to assume, in the centuries to come, a variety of shapes, religious and secular. Not that they were wholly peculiar to Plato. The Greek mystery-religions commonly drew a distinction between “beginners” and those who were, in virtue of a prolonged initiation, fit to be admitted to the highest mysteries. It is interesting to observe, too, that the highest mystery, at least at Eleusis, seems to have involved the vision of some object—perhaps a corn-ear. Plato, one might say, replaces initiation by education and the corn-ear by the form of the good.44 But, by so doing, he greatly reduces the accessibility of the highest vision to the ordinary man.

In Plato’s later thought, the theological impulse is more naked. In his earlier dialogues, the forms, in an extremely abstract sense, are rather like the Homeric heroes.