SOURCE: "Collingwood's Ethics and Political Theory," in Critical Essays on the Philosophy of R. G. Collingwood, edited by Michael Krausz, Oxford at the Clarendon Press, 1972, pp. 296-326.

[In the following essay, Milne focuses on ethical and political ideas advanced in Collingwood's works.]

Collingwood touched briefly on ethics in Speculum Mentis and in his Autobiography had some hard things to say about contemporary British politics. But it is to his last completed book, The New Leviathan, that we must go for a systematic exposition of his ideas in ethics and political theory. These ideas had apparently been maturing long before the writing of The New Leviathan was begun soon after the outbreak of the Second World War. In the Preface, speaking of his return to Oxford after the First World War, Collingwood wrote:

It was now that I began to think out the fundamental ideas of the present book, thereafter revising and elaborating them year after year in experimental form, accumulating as time went on I will not say how many thousands of pages of manuscript on every problem of ethics and politics, and especially on the problems of history which bore on my subject; and imparting my results, when I seemed to reach any which were worth imparting, in lectures to my juniors and in manuscripts to such of my colleagues as seemed interested.

Collingwood describes The New Leviathan in the opening chapter as: 'an inquiry into civilization and the revolt against it which is the most conspicuous thing going on at the present time'. This inquiry is the context in which his ideas in ethics and political theory are expounded. He goes on:

Civilization is a condition of communities; so to understand what civilization is, we must first understand what a community is. A community is a condition of man in which are included women and children; so to understand what a community is we must first understand what men are. This gives us the scheme of the present book: Part 1, an inquiry into man; Part 2, an inquiry into communities; Part 3, an inquiry into civilizations; and Part 4, an inquiry into revolts against civilization.

And he adds: 'About each subject we need to understand only so much as we need to understand what is to be said about the next'.

All this, like Collingwood's title, suggests that he is consciously following in the footsteps of Hobbes. That was certainly his intention. He had said in the Preface: 'My own book is best understood as an attempt to bring The Leviathan up-to-date in the light of the advances made since it was written, in history, psychology, and anthropology'. An essay might well be written about Collingwood's relation to Hobbes but that is not my purpose here. I shall confine myself to his leading ideas in ethics and political theory, and I shall discuss them under two heads: (A) Freedom and Practical Reason; and (B) The Body Politic. That means that I shall be concerned mainly with the latter part of Part 1 of The New Leviathan and with Part 2. But in a brief concluding section, C. Civilization, I shall have something to say about the later stages of Collingwood's enterprise. The thesis which I shall try to develop is briefly this. Collingwood's ideas in ethics and political theory need to be criticized, revised, and reformulated, if what is of value, and I think there is much of value, in his account of civilization is to stand.

Collingwood says that: 'Civilization is a thing of the mind and a community too is a thing of the mind'. An inquiry into civilization therefore belongs to what he calls 'the sciences of mind'. A discussion of his conception of sciences of mind and of their relation to the natural sciences would also need an essay to itself. Two things however need to be said about it here. The first is that for him psychology is not a science of mind but a natural science. The second, that what he is interested in is not the human mind at large but the modern European mind.'Whatever I need to know about mind is about the modern European mind; for that is what has produced in itself the thing called modern European civilization, or civilization for short, and also the revolt against it'. I shall have something to say about this at the end of my essay, but at the outset it is enough to say that Collingwood is engaged in a conceptual inquiry of much the same sort that in the years since the Second World War has come to be regarded by many British and American philosophers as their main business. The concepts in which he is interested are those which he considers to lie at the centre of modern European thought and action in moral, social, and political contexts.

I have already said that my discussion of Collingwood's ethics and political theory will be critical. The kind of criticism in which I shall engage is best understood in terms of the Platonic distinction between 'eristic' and 'dialectic', a distinction which Collingwood himself appropriated and which plays a central role in his whole argument.'What Plato calls an eristic discussion is one in which each party tries to prove that he was right and the other wrong. In a dialectical discussion you aim at showing that your own view is one with which your opponent really agrees even if at one time he denied it; or, conversely that it was yourself and not your opponent who began by denying a view with which you really agree'. In eristic, there is disagreement, in dialectic, 'non-agreement', and the essence of the dialectical attitude is a 'constant endeavour to convert every occasion of non-agreement into an occasion of agreement'. There will be a number of occasions of non-agreement between Collingwood and myself in what follows, but I shall do my best to convert them into occasions of agreement.

(A) FREEDOM AND PRACTICAL REASON

1. According to Collingwood, men are not born free. They have to achieve freedom and this can be done only at a relatively advanced stage of mental development. What is achieved is freedom of the will.'The freedom of the will is, positively, freedom to choose; freedom to exercise a will; and negatively freedom from desire; not the condition of having no desires but the condition of not being at their mercy'. If you are at their mercy, you can act from preference but not make choices.'A man who prefers A. to B. does not choose at all; he suffers desire for A. and aversion towards B.: and goes where desire leads him'. Freedom, or freedom of the will, is achieved by an involuntary act of self-liberation. 'Liberation from what? From the dominance of desire. Liberation to do what? To make decisions'. If voluntary acts are those which issue from decisions, the act of self-liberation which makes voluntary acts possible cannot itself be a voluntary act. But while the act of self-liberation is involuntary, its occurrence is not inevitable.'This achievement of free will marks the stage at which, in modern Europe, a man is supposed to reach intellectual maturity. If anything interferes with the course of his mental development, this step may never happen; he will then become a man who is incapable of growing up; perhaps a man who hates the thing (mental maturity) he does not possess'.

But the freedom of the will which any man can achieve is always a matter of degree.'On certain questions and in certain circumstances an agent may be capable of decision, or free; on other questions or in other circumstances the same agent may be utterly unable to prevent a certain passion or a certain desire from taking charge'. Collingwood calls this breakdown of freedom 'a cracking of the will' and he adds: 'for any man, I suppose, there are conditions under which a crack of the will would happen'. Collingwood distinguishes between physical and mental force, and argues that when a man yields to mental force, he undergoes a crack of the will.'When a man suffers force, the origin of the force is always something within himself, some irresistible emotion which makes him do something he does not intend to do' and he goes on: 'If B. suffers force at the hands of A., it is A. who excites in B. this irresistible emotion'.

Choice is of the essence of freedom but choice in its simplest form is identical with caprice. Collingwood describes caprice as: 'Mere choice or mere decision, uncomplicated by any reason why it should be made in this way and not in that'. From this he concludes that a completely capricious act is a completely irrational act while a completely rational act is one from which all caprice has been eliminated. It follows that absence of caprice is the test of the rationality of action or of 'practical reason'. The greater the rationality the less the caprice and the greater the caprice the less the rationality. But what kinds of reason are there for choosing in one way rather than in another? What are the forms of practical reason? According to Collingwood, there are three.'It is not because three is in my eyes a magical number; but I find that people talking about practical reason distinguish various types of it, and that these types, under inspection, resolve themselves into three falling in a certain order'. The three in ascending order of rationality are: Utility, Right, and Duty. Collingwood says that: 'On any occasion when a modern European answers the question: "Why did you do that?" he will answer: 1.'Because it is useful.' 2.'Because it is right.' 3.'Because it is my duty'.

To do an act because it is useful is to do it as a means to an end. Utilitarian action therefore involves two decisions: one about the end and another about the means.

There has to be what Collingwood calls an 'ends plan' and a 'means plan'. The rationality of utilitarian action lies in: 'the abstract conformity of the means plan to the abstract specifications of the ends plan'. He calls both the ends plan and the means plan, 'indefinite individuals' and goes on: 'Everything except the conformity of these indefinite individuals to one another is, from the utilitarian point of view, irrational'. He says of an indefinite individual that it is 'required to satisfy certain specifications but free to vary so long as those specifications are satisfied'. What he has in mind seems to be this. The ends plan must specify a state of affairs in sufficient detail for it to be envisaged and brought about by human action. The means plan must specify a course of action in sufficient detail for it to be deliberately undertaken as a way of bringing about the state of affairs specified in the ends plan. The detail in each case is incomplete, being no more than is necessary for utilitarian action to get under way. With respect to the detail which it specifies, each plan is an individual plan. But with respect to what it leaves unspecified, each is indefinite or capricious. The unspecified details are free to vary, that is, are decided capriciously as action proceeds. Utility is therefore imperfectly rational owing to the element of caprice in both the ends plan and the means plan. It is also capricious in another way which Collingwood does not mention, although it is perhaps implicit in his account. This is in the decision about what state of affairs is to be the end for utilitarian action. So far as utility is concerned, there is no reason for choosing one end rather than another.

Right is the second form of practical reason. Collingwood contrasts it with utility.'A thing is useful or the opposite in relation to the end it achieves; it is right or the opposite in relation to the rule it obeys'. Practical reason in the form of utility gives rise to utilitarian action. Practical reason in the form of right gives rise to regularian action. Collingwood says that: 'A man may, and often does, make rules solely for himself: this, indeed, is regularian action in its simplest form and unless we understand this we shall never understand the complex case in which a man makes rules for others to obey'. He says that a rule is a generalized purpose which he describes as: 'a purpose to do things of a certain kind on all occasions of a certain kind'. But 'a rule is only one part of a regularian action. There is also the decision to obey or disobey it'. Two kinds of decision are therefore involved in regularian action: decisions about what rules to obey; and decisions to obey them in situations which they cover. According to Collingwood, regularian action is infected with caprice although not to the same extent as utilitarian action. This is because 'a rule only specifies some act of a certain kind. The application of it to a given occasion bids me perform one, and only one, of the acts which would conform to its specifications. The acts which so conform may be many or few; which they are, depends not on the rule but on the circumstances; if they are many, I have got to choose between them, but the rule cannot tell me how. From the regularian point of view my choice between the alternatives is a matter of caprice'. So for that matter, although Collingwood does not mention it, is my initial decision about what rule to obey.

Duty is the third form of practical reason. According to Collingwood, duty is the discharging of an obligation which the agent has already incurred. His doctrine seems to be this. The incurring of an obligation is a free act. One chooses to give an undertaking to do something. Having given the undertaking one must then carry it out. This is what duty is: the carrying out of undertakings freely given, or the discharging of obligations freely incurred. Collingwood asks what is meant by the phrase 'his duty' when a man says that something is his duty and answers: 'A man's duty on a given occasion is that act which for him is both possible and necessary: the act which at the moment character and circumstance combine to make it inevitable, if he has a free will, that he should freely will to do'. Unlike utilitarian and regularian actions, 'dutiful' actions are definite individuals. Duty is in principle free from caprice. It is the only one of the three forms of practical reason which is completely rational.'Any duty is a duty to do this act and only this, not an act of this kind'. But judgements about what one's duty is are never incorrigible. The most that a man can say is 'I have considered, X, Y and Z, as claimants for the title of my present duty; X is a better answer than Y, and Y than Z; but there may be a better answer than any which I have overlooked'.

2. In all this, there are a number of occasions of non-agreement between Collingwood and myself. To begin with his account of mental force: he says that what forces a man is always something within himself. This is not necessarily so. A man may yield to a threat not from panic, but because he judges it to be the lesser evil. His will does not crack. He makes a rational decision. Asked 'Why did you do that?' (for instance, hand over the money) his answer is: 'Because it was prudent.' Pace Collingwood, this is an answer in line with modern European ways of thinking and speaking. But although the man's action is rational, is it free? According to common sense: no, because it is done under constraint. According to Collingwood: yes, because he freely decides to yield. He might have resisted the threat. Both are partly right: Collingwood in that the man decides or wills to yield; common sense in that what he does in obeying whoever is threatening him is decided for him, not by him. He is obeying the will of another. The difficulty can be resolved if freedom is thought of in terms of 'self-determination', rather than merely in terms of 'will'. You are free to the extent that you determine your conduct for yourself. To the extent that it is determined for you, not by you, you are not free. This allows for different kinds and different degrees of freedom in a way that Collingwood's account does not. At the same time it preserves what is true in his account. A man who panics is not free. His action is determined by what he is afraid of. A man who acts from preference is free in the sense that he determines his conduct. But a man who acts from a rational decision is free in a different and wider sense. He determines his conduct in the light of an assessment of the alternatives open to him. But more about this later.

Collingwood says that Right and Duty are different forms of practical reason. But in terms of his account, they are not really different. According to him, to say 'I did X because it was right', is to say: 'I did X because there is a rule which prescribes that X should be done.' To say 'I did X because it was my duty', is to say: 'I did X because I had freely undertaken to do it.' But why should I do what I have freely undertaken to do? Collingwood does not say, but there can be no doubt about the answer. Not to do so would be wrong. It would be to break the moral rule which says that people ought to keep their word. It follows that on Collingwood's account, Duty is a special case of Right. A duty is a certain kind of right act: namely an act of the kind which meets the requirements of the moral rule that people should always do what they have freely undertaken to do. This means that both the higher forms of practical reason are regularian in character. But Collingwood's account of regularian action is unsatisfactory. Although he distinguishes between 'willing the rule' and 'willing the particular act which obeys it', he does not seem to have grasped the significance of this distinction.'Willing' or making a rule is not itself a case of regularian action: not, that is to say, a case of obeying a rule. Obeying a rule means deciding to do an act of the kind prescribed by it in a situation of the kind which it covers. You do not have to decide what to do. The rule tells you that. You only have to decide that you will do it. Making a rule means deciding what kind of situations are to be covered and what kind of act is to be done in them. You have to decide not only what is to be done, but when.

According to Collingwood, the simplest case of regularian action is that of a man making rules for himself. Making rules for others to obey is a 'more complex case'. This betrays his failure to grasp the significance of the distinction between making rules and obeying them. According to his own definition, in regularian action, acts are either right or wrong. They either obey or disobey a rule. But making a rule means deciding what rule to make and while this decision can be better or worse, it cannot be right or wrong. There cannot be a rule telling you what rule to make, although there can be a 'second-order' rule conferring on you authority to make one, but leaving it to you to decide what rule it should be. The simplest case of regularian action is that of obeying rules which you do not make but find already there. This begins early in life. A child learning to talk learns to obey linguistic rules. About the same time he begins learning how to behave which means learning to obey elementary rules of good manners and moral conduct. This priority of regularian action is logical as well as temporal. You must learn how to obey rules before you can make them either for yourself or for other people. The conclusion which follows is that the higher forms of practical reason are not regularian in character. They include the more sophisticated activity of making rules but this is not regularian. Regularian action properly so called belongs to a level below practical reason.

Collingwood arrived at the three forms of practical reason by considering how modern Europeans answer the question: 'Why did you do that?' He has neglected one kind of answer: 'I did it because it was prudent.' For instance, obeying the orders of a gunman. There is also another kind of answer he has not considered 'I did it because it was wise.' 'Why did you change your job?' 'For the sake of better prospects and more interesting work.' 'Why did you take up sailing?' 'Because it would add to my enjoyment of life.' It is wise for me to better my prospects and make my life more interesting and enjoyable. It is prudent for me to provide for my safety and security and avoid unnecessary danger. The common factor here is the idea of personal well-being. Modern Europeans think that it is rational for a man to act wisely and prudently because they think that it is rational for him to do what he can to maintain and develop his personal well-being. There are good reasons for this conviction. They lie in the fact that while circumstances play a large part in shaping the course of a man's life, he can himself influence it through his choices and decisions. What sort of life he has depends upon what he does in the face of the opportunities and limitations confronting him. He therefore has good reason to make the maintenance and development of his personal well-being a leading consideration in the determination of his conduct. If he does not concern himself with it, why should anyone else? If Collingwood had reflected on the meaning of acting wisely and prudently he might have seen that personal well-being is a form of practical reason. To say 'I did it because it was wise', or 'because it was prudent', is to say: 'I did it because in some way, it contributed to my personal well-being.'

Personal well-being must be distinguished from utility. Utility is a rational basis for the choice of means but so far as utility is concerned, the choice of ends is capricious. Personal well-being is a rational basis for the choice of ends. If a new job has better prospects and more interesting work, getting the job is a rational end of utilitarian action. But personal well-being is also relevant to the choice of means. A particular course of action may be expedient as a means, but tedious or dangerous. It may then be wise or prudent to find some other means, or, failing that, to abandon the end to which the tedious or dangerous course of action is the means. But this is not all. The perspective of utility is confined to action done as a means. The perspective of personal well-being is wider. It includes not only utilitarian action but leisure activities like sailing which are undertaken not as the means to further ends, but because they are interesting or enjoyable in themselves. As a form of practical reason, personal well-being is on a higher level than utility which is properly subordinate to it.

3. According to Collingwood, the third form of practical reason is Duty. But on his own showing, Duty is only a special case of Right. Doing my duty means obeying the moral rule that people should always do what they have freely undertaken to do. This is not an adequate account of duty, much less of the wider idea of morality. Consider the moral virtue of honesty. If there are any moral duties at all, there is certainly one to cultivate and practice this virtue. But it does not fit Collingwood's account of duty. My duty to be honest does not arise from any undertaking I have freely given. Rather it is presupposed in all such undertakings. Instead of: 'I ought to keep my word because I ought to be honest, ' Collingwood is in effect maintaining: '1 ought to be honest because I ought to keep my word.' But why ought I to be honest? More generally: why does there have to be any morality at all? Collingwood ignores this question. But it must be answered if the claim that Duty is a form of practical reason is to be justified.

Morality has a rational basis in certain conditions necessary for human co-operation. People who co-operate together become dependent upon one another. Each must do his part if the common enterprise is not to break down. This means that there can be co-operation only if there is trust, and trust presupposes honesty. Other moral commitments besides honesty are also necessary. They include self-control, justice, and responsibility. A man who cannot control himself cannot be trusted. People who co-operate must deal justly with each other if destructive disputes are to be avoided. Each must be responsible to the rest for making the best contribution he can to the common enterprise. But is there an inescapable commitment to co-operation? Yes, if there is an inescapable commitment to social living, because co-operation is of the essence of social living. I can rationally reject the commitment to co-operation and therefore to morality only if I am prepared to withdraw from social living altogether. That means either committing suicide or becoming a hermit, if the latter is still a live option in the modern world. The answer to our question then is this; there has to be morality if there is to be social living, because without it there could be no co-operation and therefore no social living.

Self-control is the most elementary moral virtue. The duty to cultivate and practice it must be inculcated in early childhood because it is a necessary condition for any coherent activity: regularian and rational, moral and non-moral. Being honest means obeying certain rules: keeping your word, not lying, deceiving, or stealing. This suggests that over and above the elementary duty of self-control, morality is regularian in character. But justice and responsibility suggest that it is also something more. They are principles rather than merely rules. Acting on a principle involves deciding how best to implement it in the particular circumstances of a given situation: deciding what is most just or least un-just, having regard to all the interests involved; deciding how best to fulfil your responsibilities in the face of particular problems and exigencies. All this points to a distinction between two levels of moral development: a lower regularian level and a higher rational level. The lower level is that of childhood. Being moral consists of obeying established moral rules and in cases which these do not cover, or in which they conflict, relying on the guidance of parents or others to whose authority children are subject. The rational level is that of mature adult life. Being moral now means acting responsibly and justly, obeying moral rules in the spirit rather than the letter and, if doing so on a particular occasion is the lesser evil, breaking them. An important part of the process of growing up consists in making a gradual transition from the lower to the higher level. But there is nothing inevitable about this transition and, for most of us, it is never wholly completed.

At the lower level, morality is a form of regularian action, at the higher level a form of practical reason. But as a form of practical reason, it is best described as Social Morality. This brings out its rational basis. The drawback of the term 'duty' is that it is not confined to practical reason. There is a duty to act responsibly and also to act justly. But self-control and honesty are duties, and there is a general duty to obey moral rules. The question; 'Why did you do that?' can be answered in terms of social morality: 'I did it because it was the best way in which I could fulfil my responsibility', or: 'I did it because in the circumstances, it was more just than any alternative open to me.' It follows that, pace Collingwood, the three forms of practical reason are: Utility, Personal Well-Being, and Social Morality. Regularian action belongs to a level below practical reason but is an indispensable preparation for it and continues within it, being, however, subordinate to its requirements. To answer the question: 'Why did you do that?' by 'Because it was right', assumes that the rule which the right act obeys has a rational justification and that on the particular occasion there are no special circumstances which would justify breaking it.

Like personal well-being, social morality is a rational basis for the choice of ends in utilitarian action.'It is my responsibility to see that the safety precautions are adequate.' 'If we are not to be unjust, we must give him a chance to state his case and make the necessary arrangements.' Social morality is also relevant to the choice of means. No matter how expedient an action may be, if it is unjust or socially irresponsible, it ought not to be chosen as a means. Social morality and personal well-being can sometimes conflict. There is no pre-established social harmony which guarantees that the maximum personal well-being of each is always and everywhere compatible with the maximum personal well-being of all. Acting responsibly or justly can sometimes mean personal sacrifice, hardship, or even death. When this happens, there are good reasons for giving priority to the demands of social morality. Not to do so would be to abandon the commitment entailed in social living. I am not entitled to the benefits of social living unless I am prepared to pay the price and the price is meeting the demands of social morality, not that social morality and personal well-being are always in conflict. What is responsible and just is for the most part usually what is wise and prudent, but not invariably. The perspective of social morality is wider than that of personal well-being. It takes into account the nature and significance of the social dimension of human life. As a form of practical reason, it is on a higher level than personal well-being, which is properly subordinate to it.

According to Collingwood, a completely capricious act is completely irrational. He concludes from this that the less the caprice, the greater the rationality. Right is a higher form of practical reason than Utility, because it is less capricious. Duty is higher than Right, because in principle it is devoid of caprice altogether. But since Duty is only a special case of Right, and since his account of Right is vitiated by his failure to distinguish between regularian action properly so called, and the more sophisticated activity of making rules, his exposition of practical reason gets no further than Utility. I agree that a completely capricious act is completely irrational, or perhaps better, non-rational. I also agree that utility is infected with caprice, but what makes personal well-being a higher form of practical reason than utility is not that it is less capricious, although it is, but that its perspective is wider. The same is true of social morality in relation to personal well-being. One form of practical reason is higher than another, that is to say, because it embodies a better understanding of the character and conditions of human action. According to Collingwood, a rational act is always an individual act. But the point which seems to have eluded him is that in practical reason you have to deal with an individual situation, not as in regularian action merely with a situation of a particular kind. The problem is to decide what is most responsible, prudent, or expedient, in this situation now confronting you with its peculiar circumstances and features. A just or wise act is likely to contain elements of caprice. But that does not matter so long as they do not impair its justice or wisdom. A rational act must be as devoid of caprice as the situation being dealt with requires. It does not have to be a completely 'definite individual' but only as definite as is necessary for it to be responsible, wise, or expedient.

I suggested earlier that thinking of freedom in terms of self-determination, rather than merely in terms of will, brings out more clearly the sense in which freedom is a matter of degree. The relation of regularian action to practical reason bears this out. Regularian action is the first step towards the fuller self-determination involved in practical reason. A child who can follow rules is no longer completely dependent upon adults to tell him what to do. But he is still dependent on rules. He emancipates himself from this dependence and becomes more self-determining to the extent that, as he grows up, he becomes capable of acting in terms of practical reason. A man who is capable of pursuing his personal well-being wisely and prudently achieves personal freedom. Practical reason in the form of social morality is moral freedom. A man who acts responsibly and justly is self-determining as a moral agent. Finally, a brief word on dialectic and eristic. The thinking involved in regularian action is essentially eristical. A regularian act is either right or wrong. There is no room for discussion, for reconsideration and modification of decisions about what to do. The rule must be obeyed and that is the end of the matter. The thinking involved in practical reason is essentially dialectical. A rational act is better or worse, rather than simply right or wrong; for instance, more or less just, more or less prudent, more or less expedient. There is room for discussion, re-consideration, and revision of provisional decisions. The question is what would be the most responsible, the wisest, the most efficient course, and second thoughts are always possible. As we shall see later, this has a bearing on Collingwood's ideas about civilization.

(B) THE BODY POLITIC

1. Collingwood's theory of the body politic has two components: a theory of community and a theory of ruling. Of central importance in the theory of community is a distinction between a social and a non-social community. The theory of ruling is the key to this distinction. According to Collingwood: 'Ruling is either immanent or transeunt. It is immanent when that which rules, rules itself, the same thing being both agent and patient in respect of the same activity. It is transeunt when that which rules, rules something other than itself: when in respect of the same activity of ruling there is one thing which is agent, the ruler, and another thing which is patient, the ruled'. Immanent rule is an aspect of freedom. A free man rules himself. Those who have not yet reached the stage of freedom or who are incapable of reaching it, are incapable of immanent rule and, if they are to be ruled, must be subjected to transeunt rule at the hands of those who can exercise immanent rule over themselves. Transeunt rule is a form of mental force. It operates by means of reward and punishment. This is the basis for the distinction between a social and non-social community. A social community or society is made up of free persons and rules itself through the immanent rule of its members. A non-social community is made up of persons who, being not free and therefore incapable of immanent rule, have to be subjected to transeunt rule. But before saying more about this distinction, let us see what Collingwood's general theory of community is.

He says that: 'By a community I mean a state of affairs in which something is divided or shared by a number of human beings. This state of affairs I call the Suum Cuique of the community. What matters to the existence of a community is that it should have a suum cuique. Its taking one form rather than another makes no difference to the things being a community; though much to what kind of a community it is'. In addition to a suum cuique, something else is also necessary.'Thus any community must have a home or place in which corporately it lives'. He goes on to say that: 'A community must be ruled if it is to exist'. This is because: 'A community depends for its existence on something which makes it a community and keeps it a community; that is, alots to its members their respective shares in whatever is divided between them, and causes them to remain faithful to this alotment: maintains the suum cuique which is the essence of its communal character. The establishment and maintainance of the suum cuique is called ruling'. The difference between a society or social community and a non-social community is that: 'A society is a self-ruling community. A non-social community needs for its existence to be ruled by something other than itself.

What Collingwood has in mind can be seen in the case of the family. The parents, being adults, are in some measure free persons. They exercise immanent rule over themselves and maintain between them the life of the family as a social community. But for their children who are not yet free persons, the life of the family is the life of a non-social community. This non-social community of the 'Nursery' is kept in being by the parents who exercise transeunt rule over their children who are its members. Since all human beings begin life as children, all human beings begin by being members of non-social communities. They become fit for social life properly so called only when they develop mentally to the point of achieving freedom. Collingwood finds the key to social living in the idea of partnership. Becoming a free person means inter alia becoming capable of entering into partnerships with other free persons for the sake of carrying on joint activities with them.'People become partners by deciding to behave like partners. A society or partnership is constituted by the social will of the partners, an act of free will whereby the person who thereby becomes a partner decides to take upon himself a share of the joint enterprise'. Partnerships or societies are originated and kept in being by the 'joint will' of the partners or members. There are as many forms of partnership or society as there are kinds of enterprise which free persons can undertake together. One distinction however, is of special importance for the theory of the body politic.'This is a distinction between two kinds of enterprise, one intended to terminate within a length of time, planned to reach a conclusion at some definite period in the future; this I call a temporary enterprise; the other intended in Stevenson's words to travel hopefully but not to arrive; no time of termination being stated or implied; this I call a permanent enterprise. Every society is formed for the prosecution of some enterprise. Where it is a temporary enterprise, I call the society a temporary society; where permanent, a permanent society'. Two men going for a walk together constitute a temporary society. A society for studying the antiquities of a district is a permanent society.

According to Collingwood, a body politic is like a family 'writ large'.'The simplest body politic differs from the simplest family only at one point. Each is divided into a social part and a non-social part; but whereas the family society is a temporary society, the political society is a permanent society'. The members of a body politic consist of two classes.'The first class is a society and rules itself. Its members are persons or agents possessed of free will. It also rules the second class which is a community only because it is ruled. Members of the second class are devoid of free will'. He goes on: 'Let us call the first class the Council of the body politic; the second, its Nursery. It recruits the council by promotion from the nursery; it recruits the nursery by breeding babies and taking the consequences'. The body politic qua society is permanent because: 'In a body politic new babies are always being born; the nursery is always being replenished and the work of imposing order upon it is never concluded. Equally the work of establishing relations between it and the council is never concluded, nor the work of ordering the council itself, for that, too, is constantly being recruited'. The council is therefore faced with three never-ending tasks. It must define and maintain a way of life for itself, define and maintain a way of life for the nursery, and maintain the relation between itself and the nursery.

All this is summarized by Collingwood in three propositions which he calls the 'three laws of politics'. Of them he says: 'They are meant to hold good of every body politic without exception, irrespective of all differences between one kind and another. All good political practice is based on grasping them, and most bad political practice is based on failing to grasp them as rules of political activity'. They are normative, not descriptive laws, comparable in status to the laws of thought of traditional logic. The first law of politics is that: 'a body politic is divided into a ruling class and a ruled class.' The second that: 'the barrier between the two classes is permeable in an upward direction.' The third that: 'there is a correspondence between the ruler and the ruled whereby the former become adapted to ruling these as distinct from other persons, and the latter to being ruled by these as distinct from other persons'. To this, however, he adds: 'But the third law also works inversely from the ruled class upwards and determines that whoever is to rule a certain people must rule them in a way in which they will let themselves be ruled'.

There are two other things in Collingwood's theory of the body politic about which a word must be said: his idea of political action, and of the 'dialectical' character of political life. Of the former he says: 'Political action pure and simple is will pure and simple; but differs from will as such in being, first, the joint will of a society, the rulers of a body politic; secondly, that will exercised immanently upon those who exercise it as the self-rule of that society; and thirdly, the same will exercised as force in transeunt rule over a non-social community, the ruled class of a body politic'. Political action, that is to say, is the action of free persons in their capacity as members of the council of a body politic not in their capacity as private persons. Like the action of an individual free man, political action can be capricious. It can also be rational in any one of the three forms of practical reason. Capricious political action is expressed in decrees.'A decree is the simplest form of political action because it represents the simplest form of will, namely caprice, transposed into the key of politics'. He goes on to say that political action in the form of Utility is 'policy', in the form of Right, it is 'Law', while in the form of Duty, it is what he calls 'Political Duty'. The trouble with this attempt to 'transpose practical reason into the key of politics' is that the defects of his account of practical reason reappear in his account of political action. But more about that later.

According to Collingwood: 'The world of politics is a dialectical world in which non-social communities (communities of men in what Hobbes called a "state of nature") turn into societies'. Whether or not he is right to interpret Hobbes in this way, what is there which is dialectical about this world of politics? The answer lies in his view of the body politic as a mixed community made up of a social and a non-social element.'Such a community might be described by attending to the positive element as a society; by attending to the negative element as a non-social community; yet it might be one community which was being so described; the difference being only a difference in point of view, a dialectical difference'. The non-agreement between these points of view is converted into agreement when it is realized that the body politic is a mixed community, that it is always in process of becoming a society, and therefore ceasing to be a non-social community but that this process, because the nursery is always being replenished, never ceases. According to Collingwood, things whose existence is a process of transition from one state to another, a process which can never be completed and which never begins, but is always already under way, must be thought of in dialectical terms. To think of them eristically, that is as either being in the state from which they are turning or as being in the state into which they are turning, would be to fail to understand their nature. They are never wholly in either state but are always in the course of transition from one to the other, and this from of 'process' existence is best described as dialectical.

2. Collingwood regarded his theory of the body politic as an amendment to what he called the 'Classical Politics', that is, the political theories of Hobbes, Locke, and Rousseau. The amendment was the replacement of 'the doctrine of the State of Nature' in these theories by his theory of the non-social community. But there is also a clear Platonic strain in his theory of the body politic. A central thesis of The Republic is that those who are to rule others must first be able to rule themselves, and this is also the central thesis in Collingwood's theory. The resemblance between his ruled class and Plato's class of 'Producers' is obvious. So too is the resemblance between Collingwood's theory of mind and Plato's 'doctrine of the Soul'. But that is a matter which I cannot pursue here. Granted that those who are to rule others must first be able to rule themselves, is Collingwood's account of ruling in terms of the distinction between immanent and transeunt rule acceptable? I do not think that it is.

Immanent rule, like freedom, is a matter of degree. It is part and parcel of the capacity for self-determination. Its most elementary form is self-control which is a necessary condition for regularian action. For practical reason, something more than merely self-control is necessary, namely the ability to control and determine one's conduct in the light of purposes which have been freely chosen. Transeunt rule need not necessarily be a matter of force. People can and do obey rules which have been made by others for them, and which they could not have made for themselves, without having to be forced to do so. Civil law largely consists of such rules. People obey them for the most part because they recognize the need for them, although they could not have made them and do not understand in detail the particular reasons for their taking one form rather than another. Collingwood is maintaining that only those who can make rules for themselves can freely obey rules made by others, a view which repeats the error in his account of regularian action discussed in the last section.

All this has implications for his theory of the non-social community. Its members are said to be incapable of immanent rule and hence of freedom. They are at the mercy of their desires and passions, and act only from preference, not from choice. Could such people constitute a community at all? Any community must be a regularian community in the sense that its members must be capable of obeying rules and must therefore be capable of enough immanent rule to enable them to do so. They cannot therefore be wholly at the mercy of their desires and passions. No doubt there are cases, for instance a community of prisoners, in which the threat of force plays a large part in securing obedience. But the effective operation of such a threat presupposes some degree of prudence on the part of those who are threatened, not simply uncontrolled fear. According to Collingwood, non-social communities are always dependent on something else which rules them. In fact, he is not wholly consistent about this. In the course of his discussion of political action, he refers to international law as 'the customary law of a very ancient non-social community', and likens this customary law to the law of the Iceland of the Sagas. Presumably a community which maintains a customary law is not dependent on some-thing else to rule it, and is best described as a regularian community. What would he say about the so-called 'traditional societies' studied by social anthropologists, that is tribal communities? These are regularian communities, or at least predominantly so. There may be the rudiments of practical reason in the form of utility, but little or nothing of free partnership or 'society' in his sense. Yet tribal communities are not dependent on something else for their existence. They maintain their own largely customary life from generation to generation. They can hardly be called non-social communities in his sense.

A tribal community may not be a social community, but it is a self-sufficient community in the sense of being sufficient for the maintenance of human life. Human life is and always has been carried on in self-sufficient communities. At different times and in different places these have taken very different forms. Tribes, kingdoms, city-states, the principalities and duchies of Medieval Europe, and the nation-states of the modern world, are instances. Collingwood's theory of the body politic is a theory of the essential characteristics which every self-sufficient community must possess. He is saying, in effect, that if a community is to be self-sufficient for human life, it must be a body politic. But as the case of a tribal community shows, he is wrong. A regularian community can be self-sufficient for human life. It does not have to be a social community as well. Moreover, since there can be no community without regularian action, every self-sufficient community must at least be a regularian community. But it need not be merely a regularian community, it can also be something more. It can become a social community to the extent that its adult members are able to make the transition from regularian to rational morality, that is, become capable of practical reason in the forms of personal well-being and social morality. After making this transition, they are still members of the self-sufficient community, but rational, not merely regularian members. They have become free persons and free moral agents in the sense of being self-determining in both capacities.

This suggests that something can be salvaged from Collingwood's theory of the body politic. The theory of the non-social community must go and be replaced by the theory of the regularian community, which unlike the non-social community, can be a self-sufficient community. But his theory of the social community based on the idea of partnership is compatible with the revised account of practical reason of the last section, and can be cormbined with the theory of the regularian community to yield a theory of the typical form of self-sufficient community in modern Europe: the nation-state. It is after all with modern Europe that he is primarily concerned. But while this is all right so far as it goes, there is another objection to be reckoned with. In what sense is this revised theory a theory of modern European bodies politic, as distinct from modern European self-sufficient communities? It may be acceptable as a theory of community and society, but in what respects, if any, can it be regarded as a political theory? The short answer is that with the removal of the theory of the non-social community, the political element, what I called earlier the Platonic strain, in his theory of the body politic has gone.

Modern European self-sufficient communities are political communities. They are not only nations but 'States'. Their political character comes from the institution of government. This institution is not essential to the existence of a self-sufficient community as such. Tribes have existed without it and it is absent from the Icelandic community of the Sagas. But it is clearly indispensable to modern European nations. Their ways of life would not be what they are without it. The key to its role lies in distinguishing between the 'public' and 'private' aspects of the ways of life of modern European nations. From one point of view, the life of a modern European nation is a life of 'private enterprise', not merely in economic activity, but in domestic and cultural spheres as well. It embraces many varieties of 'partnerships' in Collingwood's sense, or voluntary associations, undertaken by people who in some measure at least are free persons and free moral agents. But from another point of view it is a public life, that is a life subject to public regulation, supervision, and direction, the purpose of which is to establish and maintain conditions under which the members of the national community who may number many millions, can all of them participate in the life of private partnerships and voluntary associations. The role of government is to be the custodian of the public life of the nation. While this is only a hint of the complex character of the institution of government, it is enough to bring out the inadequacy of Collingwood's theory of the body politic as a theory of the modern European nation-state. His concept of the 'council' of the body politic throws no light on the distinction between 'public' and 'private'. To think of the body politic as a family 'writ large' is to miss the significance of the institution of government for the life of free partnership and voluntary association.

Of his three laws of politics, only the third contributes anything to the understanding of modern European political life.'A people can only be governed in the way in which it will let itself be governed'. The first two turn on his distinction between the ruling class and the ruled. But this is only a distinction between those who are and those who are not capable of practical reason, a distinction which is a matter of degree, and which blurs the distinction between the public and private aspects of a nation's life. His account of political action fails to distinguish between political and social action. So far as 'transposing practical reason into the key of politics' is concerned, all that needs to be said is this. If the work of governing is to be well done, it must be in the hands of persons capable of practical reason in the form of social morality, that is, persons who can act responsibly and justly as 'custodians of the public interest', as well as wisely and prudently in determining in detail what must be done to promote it.

Collingwood argued that the world of politics is a dialectical world because a body politic is always in the course of a transition from being a non-social community to being a society. It is never wholly the one nor wholly the other. There is a sense in which this might be applied to the public and private aspects of the life of a modern European nation. Its life is never wholly private nor wholly public but always partly one and partly the other, and the line between the two is never fixed. But that is not all. There is another sense in which political life can be dialectical; when the form of government is democratic. However, more about that later. What is false in Collingwood's contention is that the life of a self-sufficient community is dialectical, simply in virtue of its being a self-sufficient community. A tribal community with a predominantly regularian way of life is not in any meaningful sense always in the course of turning from one state into another. The fact that there is always a new generation in the course of growing up makes no difference. This is one of the most stable characteristics of the community. On the other hand, in a self-sufficient community like a modern European nation which is a social as well as a regularian community, the idea of a dialectical character is more plausible. Whether much is to be gained by thinking of modern European communities in this way is another question. Perhaps more important is the sense in which their lives are historical in character, that is, change from generation to generation through the cumulative and largely unforeseen actions in each generation. Collingwood would not have denied this, although he did not discuss it in The New Leviathan. But the matter cannot be pursued further here.

3. The Modern European Mind has produced 'the thing called Modern European Civilisation'. It has also produced the Modern Democratic State. While Collingwood has something to say about democracy and aristocracy in The New Leviathan, it is in connection with his doctrine of the ruling class in a body politic. He does not undertake any sort of examination of democracy as a form of government specially characteristic of Modern European Civilization. The omission is surprising in view of his declared intention to 'inquire into civilization and the revolt against it'. I shall try briefly to fill the gap. The modern democratic state, or 'Western Democracy', rests upon four main principles: the rule of law; representative government; constitutional opposition; and equality of citizenship. I do not say that these principles are fully implemented in self-professing Western democracies: far from it. But I think that thoughtful adherents of Western democracy would agree that they are morally committed to them. Let us look more closely at them.

The rule of law involves three things: the supremacy of law, that is, that legal obligations are paramount; equality before the law, that is the equal subjection of all including the government to the law, and the equal protection for all of the law; and freedom under the law, that is that where the law is silent, all are free to act according to their own volition. A representative government is one which is chosen from and accountable to a wider citizen body, this being secured by periodic free elections. Hence the necessity of the rule of law: electoral procedure must be provided for by law. But without the opportunity for choice and criticism, there can be no free elections: hence the necessity for constitutional opposition. The composition of the citizen body is determined by the fourth principle. Its membership must be co-extensive with the entire adult population. This guarantees not only 'one man, one vote', but the equal right of all to participate in politics. The principle of constitutional opposition means what it says. Opposition must be within the limits of the law. You are free to criticize and oppose the government of the day but not to undermine its authority to govern. This presupposes that the great majority of the citizen body had a clear practical understanding of the difference between opposition and rebellion. But they must not only understand the difference. They must be able and willing to act on it. That means acknowledging and fulfilling the paramount obligation to obey the law, and eschewing all forms of revolutionary political action.

Two preconditions are necessary for the successful working of democracy: one social, the other cultural. The social precondition is the absence of fundamental conflict. The members of the democratic body politic must be in broad agreement about the fundamental character of their way of life. It must not matter too much who wins the next election. People must be willing to accept the verdict of the ballot-box. Where this precondition is absent, for instance, in a self-sufficient community deeply and painfully divided along racial or religious lines, the best that can be managed is a government drawn from and accountable to the members of the dominant racial or religious group. Members of the subordinate group will be 'second-class citizens' whose voice can become effective only through revolutionary political action. The cultural precondition is a well-established and widespread tradition of discussion and argument, based on the recognition that to most questions there is more than one side, and that those who are on the opposite side are not for that reason either knaves or fools. This cultural precondition presupposes the social precondition. Both are necessary for the effective working of democracy.

In view of this, it is hardly surprising that democracy has taken root and lasted in only a minority of the nations of the modern world. It is only in Western Europe, North America, and Australasia that both the social and cultural preconditions are to be found. The majority of mankind have never known democracy, and for the foreseeable future will have to manage without it. It is not the best form of government if that means best for everyone always and everywhere. But there is no one form of government which is best in this sense. What is true, however, is that where the social and cultural preconditions are present, democracy is better than any other form of government. This is so for two reasons: one negative, the other positive. The negative reason is that democracy more than any other form of government takes account of and provides protection against human fallibility. It recognizes that no individual and no group has a monopoly of virtue or wisdom. The positive reason is that through the principles of constitutional opposition and equality of citizenship, democracy gives more scope and encouragement to more people to advance from regularian to rational morality than any other form of government. It enables practical reason 'to be transposed into the key of politics', which is another way of saying that it makes possible dialectical rather than merely eristical politics. Under a democratic form of government, occasions of political non-agreement have at least a better chance of being converted into occasions of political agreement, instead of hardening into disagreement. To be fair to Collingwood, an understanding of this is implicit in his discussion of aristocracy and democracy although he did not relate it to an examination of the principles underlying democracy.

(C) CIVILIZATION

Collingwood distinguishes between the generic and the specific meaning of 'civilization'. According to the generic meaning: 'Civilization is a process of approximation to an ideal state. To civilize a thing is to impose on it or promote in it a process; a process of becoming; a process in something which we know to be a community; whereby it approximates nearer to an ideal state which I will call civility and recedes farther from its contradictory, an ideal state which I will call barbarity '. No community is ever simply barbarous, or completely civil. It is always in a condition of turning from the one into the other. To discover the specific meaning, it is necessary to determine the specific character of the process, to know what happens to a community as it becomes more civilized. Discussing the specific meaning, Collingwood says: 'According to the view I find expressed in books I have looked at, and in the mouths of persons I have questioned to find out what the thing called civilization is commonly thought to involve, civilization has something to do with the mutual relations of the members within a community; something to do with the relation of these members to the world of nature; and something to do with the relation between them and other human beings not being members of the same community'. The clue to what this something is, lies in the idea of civil behaviour.

'Behaving civilly to a man means respecting his feelings: abstaining from shocking him, annoying him, frightening him, or (briefly) arousing in him any passion or desire that might diminish his self-respect; that is, threatening his consciousness of freedom by making him feel that his power of choice is in danger of breaking down and passion or desire likely to take charge'. People who treat one another civilly act in a dialectical, not an eristical spirit. 'Being civilized means living so far as possible, dialectically, that is, in constant endeavour to convert occasions of non-agreement into occasions of agreement.' What Collingwood means is clear. He is equating the process of civilization with the process by which a community becomes a social community. Civility equals sociality.'Civilization is the process in a community by which the various members assert themselves as will; severally as individual will, corporately as social will (the two being inseparable)'. So far as relations with other communities are concerned, civilization means acting civilly towards their members. So far as the natural world is concerned, civilization means 'a spirit of intelligent exploitation'. This is possible only in a community whose members already treat one another civilly since that is necessary for development, conservation, and transmission of the knowledge and experience which is itself necessary for the intelligent exploitation of nature.

What about 'revolts against civilization', the fourth part of Collingwood's inquiry? It is important to distinguish between lack of civilization and the repudiation of civilization.

I distinguish two ways of being uncivilized. I call them savagery and barbarism and distinguish them as follows. Savagery is a negative idea. It means not being civilized, and that is all. In practice, I need hardly say, there is no such thing as absolute savagery; there is only relative savagery, that is, being civilized up to a certain point and no more. By barbarism I mean hostility towards civilization; the effort, conscious, or unconscious, to become less civilized than you are, either in general or in some special way, and, so far as in you lies, to promote a similar change in others.

Relative savagery is equivalent to relative barbarity in his account of the generic meaning of civilization. Barbarism is the attempt to reverse the process of civilization and move back towards the ideal state of barbarity, and away from civility. It is the paradoxical attempt by members of a social community to revert to non-social community from which they have emerged, paradoxical because free persons are choosing to abandon freedom, choosing to give up choosing. Collingwood describes the will to barbarism as: 'a will to acquiesce in the chaotic rule of emotion, which it began by destroying. All it does is to assert itself as will and then deny itself as will'.

If my argument in this essay is sound, Collingwood's account of civilization needs amending. Civilization is a process which is undergone by a self-sufficient community. Such a community is always at the very least a regularian community. It becomes civilized to the extent that it becomes more than merely a regularian community, that is to the extent that practical reason, not merely regularian action, plays a part in its way of life. But practical reason means utility, personal well-being, and social morality. Collingwood's account of it gets no further than utility. The thinking involved in practical reason in terms of my revised version is dialectical, so Collingwood is right to say that 'being civilized means, so far as possible, living dialectically'. But he has failed to give a coherent account of how this manner of living is embodied in practical reason. To say that civilization is a process towards an ideal state of civility and away from an ideal state of barbarity is misleading. It fails to allow for the case of tribal communities which are not in any meaningful sense social communities, and which are untouched by civilization so far as the higher forms of practical reason are concerned. Finally, Collingwood's idea of the body politic leaves no room for a distinction between civilized and uncivilized forms of political life, since he is committed to holding that every self-sufficient community must be a body politic without differentiating between the social and political aspects of its life. He has failed to see that democracy, because more than any other form of government it makes dialectical politics possible, is more civilized than any other form of government.

His conception of barbarism is really a conception of irrationality, as distinct from non-rationality and imperfect rationality. To what extent the historical examples of barbarism which he discusses in Part 4 of The New Leviathan are really examples of irrationality I am not competent to judge. But he is certainly right that irrationality, in the sense of the repudiation of practical reason by those who are capable of rational conduct, is a revolt against civilization. Whether he is right in contending that this always takes the form of 'acquiescing in the chaotic rule of emotion' is another matter. It may take the form of ideological commitment in the face of rational misgivings which are repressed or 'rationalized' away. It may take the form of a reaction against what Popper called 'the strain of civilization'. This might be described in Collingwood's terms as a relapse from dialectical into eristical thinking. Collingwood limited himself to revolts against civilization. Equally important is an inquiry into its 'strains' and into the obstacles which impede or arrest it. The social precondition for democracy, the absence of social conflict, is not something which can be created by good will. Dialectical thinking cannot end social divisions since it can flourish only where they are absent.

Collingwood regarded his inquiry into civilization as an essay in the science of mind. I suggested at the beginning of this essay that what he was engaged in was a conceptual inquiry of much the same sort as about a decade later were undertaken by many British and American philosophers. Both Collingwood and his postwar successors seemed to have thought of their task as essentially descriptive rather than critical. Collingwood was led to this conclusion by his ideas about the relation of philosophy to history, postwar philosophers by their views about philosophy and language. In the end perhaps these two views are not really different, since language is essentially a historical phenomenon. Be that as it may, a philosophical study of concepts which eschews criticism is failing to carry out its proper task. Not only 'What can be meant by …?' but 'How should we think o f …?' concerns the philosopher. The defects in Collingwood's account of freedom and practical reason, and in his account of the body politic, stem as much from lack of critical appraisal as from incomplete descriptive analysis. But this is not the note on which to end. The impression given of The New Leviathan in this essay is misleading because of its concentration upon defects in certain parts of the argument. To do justice to the book as a whole would need more space than is available to me. In his autobiography Collingwood expressed his attitude towards potential critics by saying: 'Let them write, not about me, but about the subject'. The measure of his achievement in The New Leviathan is that if the subject is civilization and the writer a philosopher, he can hardly avoid writing about Collingwood.