Ends, Means and Oakeshott

Oakeshott, ‘Rationalism in Politics and other Essays’.


Oakeshott’s criticism of rationalism in politics is marked by a characteristic disregard for academic discussions. The long-established division of philosophical thinkers into rationalists and empiricists – a division first given prominence by Kant – is only tangentially connected to Oakeshott’s use of the term ‘rationalism’. For Oakeshott the empiricist Bacon is as much a rationalist as the rationalist Descartes. And his way of explaining rationalism is to begin from the rationalist character – the peculiar attitude to self and world which Oakeshott associates with some of the disasters of modernity (of which he gives an equally eccentric list). Oakeshott’s metaphysics, as expounded in Experience and its Modes, is more rationalist than empiricist (in the traditional sense of those terms), and he is sometimes thought of as the last gasp of Hegelianism in British philosophy.

Likewise the term ‘ideology’ has, in Oakeshott’s usage, little to do either with Destutt de Tracy’s original use of the tem, or with Marx’s adaptation. For Marx and just about everyone who followed him down to Foucault and Deleuze, ideology is to be contrasted with science. An ideological system of beliefs is one adopted because it is functional in sustaining some given social order or system of power relations. Scientific beliefs, by contrast, are adapted because they are thought to be true, and whatever functionality they might have is the result of their truth.

For Oakeshott an ideology is something like a system of goals, together with the reasons for aiming at them and the techniques for achieving them. Rationalists are addicted to ideologies, since they seem to justify giving priority to ‘reason’ (in a certain use of that term) over tradition, custom and experience. For Oakeshott, however, an ideology is a kind of intellectual disease, a riotous outgrowth of the mind, that comes about when thought is prised free from reality, and begins to dwell on itself.

The Rationalist Heresy.

As Oakeshott expresses it, the error of rationalism is to confuse the ‘modes’ of experience, by applying in one realm of experience (that of politics) a mode of thought that belongs to another (that of business, or military campaigns, or…). Clear goals, precise techniques, and rules of practice make sense on the battlefield; but they mutilate and undermine the world of politics, by forcing upon it an order that it can never spontaneously adopt.

There is not much theory given in support of that claim: instead we have an intuitive description of the rationalist character, and two important distinctions. The first distinction is between technique and ‘practical knowledge’. Technique is rule-guided, can be contained in a textbook and taught. You learn techniques as you learn recipes – and the cookery book is one of Oakeshott’s examples. Practical knowledge, by contrast, is not rule-guided, and cannot be taught but only imparted. You learn it by apprenticeship – in other words by following in the footsteps of the master, like Raphael in the studio of Perugino. It is fair to say that this ‘practical knowledge’ is left under-described by Oakeshott, and stands proxy for all those forms of knowledge and wisdom that might be necessary for the proper conduct of human affairs but which cannot be reduced to rules.

The second distinction is that between ideology and tradition. Again the first term is more clearly characterised than the second. Ideologies are recipes – goals to be achieved, techniques for achieving them, and philosophical reasons for justifying them. The simplest examples in politics are the plans and schemes of revolutionaries – though Oakeshott gives Robert Owen and Owenism as his preferred example. The contrast with tradition is not clearly drawn: sometimes it is experience, or reality or custom that is being singled out as the contrasting term.

Still, we can grasp what Oakeshott is getting at – namely his own version of the (by now, I hope, familiar) epistemological argument: that the knowledge required for successful political practice (a) cannot be contained in a plan and (b) is destroyed by the attempt to provide one. Oakeshott’s additions to this argument are quite adventurous – certainly more adventurous than anything in Burke. For example, the assertion that ‘practical knowledge’ cannot be taught or learned but only imparted – where that means something like ‘transferred through familiarity and example’ – is a striking claim, reminiscent of the ancient doctrine that virtue can be acquired only by mimesis. Indeed, in so far as one can find a precedent for Oakeshott’s ‘practical knowledge’ it is something like Aristotle’s practical wisdom or sophrosune – the wisdom that resides in the harmonious exercise of virtue.

Here it is worth pointing out that Oakshott’s Rationalism in Politics appeared after the Second World War and was heavily influenced by his experience of that war. There is a very important historical point to be made about twentieth-century wars, and indeed all wars since those launched by the French Revolution, which is that they are fought by conscripts . The entire body of citizens is involved in a modern war, either as combatants or as targets. And this means that the entire body of citizens has to be brought under some kind of central quasi-military command if the war is to be won. Modern warfare requires the socialist state, or something like it; and when the weary population elects its first peace-time government it is on the unspoken assumption that the military order on which it has come to depend will remain in place, taking charge of all those things which were once ordered by the ‘invisible hand’. Hence, to the great surprise of many patriots, the British people elected a Labour government in 1945, notwithstanding the fact that Churchill was leader of the Conservative Party. And the post-war intellectual establishment was more or less uniformly socialist in tendency. Harold Laski had become chair of political thought at the London School of Economics, actively recruiting the new generation of socialist political scientists. However, he encountered internal opposition in the presence, at LSE, of two Viennese refugees, Popper and Hayek, both of whom succeeded in producing anti-socialist disciples; so it was not too surprising that Oakeshott succeeded Laski on the latter’s retirement. Henceforth LSE, which had been the core of the left establishment, shifted to the right. Hayek, Popper and Oakeshott were all in the business of articulating an anti-collectivist philosophy, at a time when, on account of the war, collectivism had become a kind of background assumption in political practice.

Civil Association.

In subsequent writings, notably On Human Conduct, Oakeshott develops another distinction that he regards as vital for the understanding of political order – that between enterprise association and civil association, as he calls it. Again, the negative term of the distinction is much clearer than the positive. An enterprise association is one entered into and maintained for some independent purpose – as in a business partnership. In such an association the terms are justified in terms of the purpose served, and the rules are to be understood as techniques: prescriptions which help to advance the purpose intended. The purpose provides a single and overarching criterion in terms of which to assess the value of any activity and the utility of any particular rule.

Now, on Oakeshott’s view, it is the great error of socialism, and also of the managerial kind of conservatism, to think of political order as a kind of enterprise association. Whatever the goal – be it equality, social justice, the instigation of the five-year plan, economic growth, or whatever – an enterprise association on this scale will be quite incompatible with the kind of order demanded by political coexistence. It will not provide the core parameters of citizenship, of free association, of negotiation and the free exercise of individual choice; everything will be subservient to the goal, which will exert its oppressive influence over every aspect of social life. Furthermore, for reasons spelled out in the criticism of rationalism in politics (but rather better spelled out in Burke’s criticism of the Revolutionary conception of natural rights as the goal of government and Hayek’s criticism of the tyranny of the plan) the goal will prove incapable of realisation – always present, never achieved, a standing rebuke to the renewable failure of society, kept alive only by the punishment of the citizens for their inability to achieve it.

Now we probably have little difficulty in accepting that as a criticism of communist-style collectivism, of which it is an accurate portrait. But Oakeshott’s target was not communism, except tangentially. He was far more concerned with tendencies that he discerned within Western democracies: tendencies to replace civil association, as he called it, with some kind of managerial, technocratic, or at any rate essentially goal-directed substitute. The error here is manifest, for example, in the field of education. A manager, brought in to advise an educational establishment, will first ask ‘what is its purpose?’ And he will not be satisfied with the obvious, and only, reply, which is ‘education’. For he will want to know ‘what is the purpose of that?’  To say that it has no purpose, other than itself, is to defeat the managerial enterprise, which is to make the educational institution more efficient, better at providing the goods that it was established to provide. So the manager will still want to know: what are those goods? The socialist might respond by saying that the purpose of education is to create equality, by extending new opportunities to the working class. The manager then can get to work on shaping the school accordingly: equality is the goal, and the school is a means to achieving it. The result of this, as we know, is a process of ‘dumbing down’, the first casualty of which is education. Ditto for the other fashionable goals, such as economic efficiency, multicultural inclusiveness, national identity, etc. All such causes direct our attention away from the fundamental truth, that the purpose of education is itself, that it is an end and not a means, and that schools must therefore shape their curriculum and their methods accordingly, as ways of making that end into a present reality.

Not surprisingly it is difficult to say what that involves. To say that something is not a means is easy, just as it is easy to criticize those who treat as a means what is not a means at all. But to describe something as an end, and to say what is involved in so treating it, is always difficult. All attempts to vindicate such a description seem to have a fatal tendency to undermine themselves, to show that the thing alleged to be an end is really a means after all. This is one reason why Oakeshott’s description of ‘civil association’ is so much less clear than his criticism of the view that political order is really a kind of enlarged enterprise association. Civil association is not association for a purpose, not even for the purpose of civil association. It lies in an important sense beyond purpose, in a realm of pure personal fulfilment, like love. If someone asks ‘what is the purpose of love?’ you will know that he does not understand what he is talking about. Of course, from the sociological point of view, love has a function: it is one part of the social cement on which societies depend for their survival. But it doesn’t have a purpose: it is beyond purpose, a feature of the ‘kingdom of ends’, to use the Kantian language. To say that you love for the sake of love is wrong too: that describes a narcissistic perversion of love, not the thing itself.

Ends, means and friendship.

Oakeshott’s discussion takes us back to Aristotle’s account of friendship. For Aristotle political order should be conceived on the model of friendship: the mutual affection that binds people together so that they to a certain measure share each other’s fate. But Aristotle distinguished three kinds of friendship: the friendship of pleasure, the friendship of utility and the friendship of virtue, corresponding to a threefold division among practical reasons. (Three kinds of answer to the question: ‘Why do that?’.) A friendship of utility, like a business partnership, is dominated by a common purpose, and permeated by ‘instrumental reasoning’, as someone like Habermas would put it. This kind of friendship corresponds to Oakeshott’s enterprise association, and was regarded by Aristotle as in some way an incomplete or unfulfilled form of friendship. The fulfilled kind occurs between virtuous people, who love each other for their virtues. This is exaggerated, of course, and would be better put in the Kantian idiom, by saying that fulfilled friendship exists between people who appreciate each other for what they are, (as ‘ends in themselves’), and not for what they do or produce. But what they are is also what they give to the other in friendship. Put in a more Christian way: fulfilled friendship occurs between people who give to each other, and who do not seek a return. The return comes only because they do not seek it, since it comes in the form of a gift. There is a complex phenomenology here, but the logic of the point is not, I think, difficult to grasp. Fulfilled friendship comes about only between people who not treat each other instrumentally. And the benefit of such a friendship is available only to the person who does not pursue it – who values his friend for his own sake, and not as a means to the benefits attendant on the friendship.

That, I think, is the way in which Oakeshott is envisaging civil association. Those joined in a civil association receive an enormous benefit from their membership – all the benefits that Hegel summarizes in his theory of civil society, including the benefits of shared citizenship under a single rule of law, which provides the all-surrounding guarantee of safety and civil order. However, they do not associate for the purpose of these benefits, for the reason that these benefits come about only as the by-product of civil association and not as its goal. If ever these benefits should become the purpose (as in a ‘social contract’) then they would disappear. They would disappear because the civil association would collapse into an association of another kind – not an enterprise association, exactly, but nevertheless something approaching to Philip Bobbitt’s ‘market state’ (The Shield of Achilles). In such an association the bond of civil society would be absent: each would be ‘in it for what he can get’: the important attitudes which hold society together (public spirit, self-sacrifice, the benevolence which, worked upon by convention, produces, according to Hume, the ‘artificial virtue’ of justice) would all be at risk, subdued by the kind of instrumental reasoning that fails to see the point of them. (To see how the details of the argument might go, we might try to develop the parallel with friendship step by step; and it is fruitful to consider Hume’s thoughts about justice too. For Hume social order comes about only because we find reasons other than self-love for joining with our neighbours – in other words, only when we leave purely contractual reasoning behind.)

Now this does not mean that civil association is an ‘end in itself’ – whatever that does mean in this context. It means only that it must not be instrumentalised, not even in relation to the benefits that it brings. As with friendship, if we treat civil association as a means to the benefits that it confers on us, then we undermine the social relations from which those benefits flow. We enjoy those benefits, as we enjoy love and friendship; but our ability to do so is predicated on our ability to give, as people give in convivial company, in conversation, in a holiday gathering, or when making music together. Indeed, in several of his later writings, Oakeshott seems to adopt conversation as his preferred model of a non-instrumental association, and to imply that civil association is in some way composed of such innocent activities, in which people are at ease with each other, receiving benefits only because their purpose is to confer them. This is a very attractive model of society – who would not opt for it if he could? But of course it seems to overlook exactly what Hegel was drawing attention to, when he distinguished civil society from state. The joys of civil society are possible, Hegel argued, only if civil society is ‘transcended’ (aufgehoben) into the state – in other words into the sphere of law, which imposes its coercive order on our relations, and demands obedience, even from those who see no reason to obey. It was precisely in order to justify this kind of coercion that theorists of the social contract attempted to trace the origins of the state to the citizens’ consent.

Even so, we might accept that there is an important point underlying Oakeshott’s idea, and one that engages with some of the fundamental intuitions of modern conservatism. The point is this: political obligation is non-defeasible, an absolute obligation which we neither undertook nor can easily escape from. It does not have the character of the normal contractual obligations of the marketplace, and cannot be understood in instrumental terms. Moreover, we come to accept it alongside many other equally non-instrumental obligations, such as those of family, erotic love, conversation, club and team – all of which flesh out the obligations of citizenship with the content of civility . All these relations are undermined or corrupted by instrumental reasoning, and instrumental ways of thinking (whether socialist or capitalist, it should be added) are the natural enemies of civility in all its forms. They stand to civil association as a functionalist factory stands to a Gothic cathedral or Walmart to Princeton University.

In this way we can see Oakeshott’s account as picking up something that Burke was referring to, when he traced the obligations of civil society to the ‘little platoons’ which are our apprenticeship in social affection. And the purpose in both cases is to distinguish peaceful forms of association from belligerent forms. The easiest form of association to create and to understand is that of the army, united by a chain of command, and dedicated to the single purpose of victory over enemies. All collectivist projects end (even if they don’t begin) as bands of ‘citizens on the march’, as Burke pointed out in his analysis of the French revolution (which was held together, according to Burke, by an ‘armed doctrine’). The hardest form of association to create is one that has no single chain of command, which is protected by the offices and laws of a peaceful state, and which is dedicated to the mutual accomodation and well-being of its citizens. In such a state belligerence is marginalized in favour of peaceful associations. And these associations are not merely peaceful in the way that herds are peaceful. They give a moral content to the idea of peace, as the forum in which people enjoy each other. Friendship is not their purpose: it is the way in which they are.


Readers of Oakeshott often complain about a certain Bloomsbury-ish attitude to ordinary people in his prose: a sense that, deep down, ordinary people ought to be entertaining themselves with Socratic conversations, gentle affairs, and dinner parties. The harshness of ordinary life, the business of survival in times of need, the distresses of failure, jealousy, resentment and so on: all this sits uneasily with the ‘civil association’ that Oakeshott advocates – even though, as an acknowledged expert on Hobbes, Oakeshott was familiar with the argument from human nature in the raw.

Among the features of our social life that are seldom mentioned by Oakeshott, by no means the least important is religion. Where does this fit in to his kind of conservative vision, or to any conservative vision for that matter? This question takes us back to a consideration of the Enlightenment, and the view of culture that emerged from it. Oakeshott’s theory of civil association is downstream from, and distantly related to, Schiller’s argument in Letters on the Aesthetic Education of Man, which I discussed in the class on culture. Both are looking for a way of separating activities that are ends in themselves from activities that are means to some external end; and both hold that our social and political nature is fulfilled by the first kind of activity but not by the second. For Schiller this meant that aesthetic education, as he called it – in other words culture in its singular meaning – is a necessary preparation for the life of society. And Oakeshott too, in one essay (The Voice of Poetry in the Conversation of Mankind) seems to be consciously echoing Schiller’s romantic effort to ‘save the appearances’ of community, through its aesthetic epitaph. But we should remember that Schiller’s Letters , while explicitly Kantian in inspiration, were also in part a reaction to the Enlightenment view of religion. God had retreated from the world; the rituals and texts in which he had made his presence known had lost their authority; and with the retreat of religion had come a loss of the sense of intrinsic value – the sense that attaches to ritual and sacred texts, to acts of worship and the ‘life in obedience’ that religion demands. And Schiller looked to culture in order to recapture that sense of intrinsic value. In a way Oakeshott’s encomium to civil association is attempting something similar – though with far less conscious recognition of the fact that, what is being reconstructed here, is an outlook that, in less self-conscious times, is the child of religion.

The point that principally needs to be made here is that the whole conflict between liberalism and conservatism (a conflict that lies deeper than that between either of them and the various socialisms that have recently clouded human vision) is, in the end, a conflict that is rooted in religion. Liberalism created the Enlightenment; conservatism tried to undo it – to fill this unnaturally illuminated space with soothing shadows. As children of the Enlightenment, Burke, Hegel and de Maistre were conscious of its anti-religious impulse, and all reacted against this. Only Maistre reacted with a whole-hearted commitment to the old religion, though it was a commitment infused with irony and disdain. Burke’s attitude was one of melancholy affirmation of the contours of the Christian faith, while avoiding any defence of the theology; Hegel’s attitude too was essentially anthropological, seeing in the Christian vision an extended metaphor of man’s spiritual journey, which could be ‘cashed out’ in terms of the institutions and laws of civic life.

In contrast the attitude of the liberal has been to reconstruct the entire underpinning of social order without reference to religion – to relegate religion to the sphere of private choice, while founding political obligation and the rule of law on purely secular principles that have no religious endorsement. This is clear enough in Locke’s social contract, and in the writings of Rousseau. But it becomes even clearer in the writings of modern liberals. For Rawls religion belongs, if at all, in that sphere of ‘conceptions of the good’ which have to be hidden behind the veil of ignorance, when it comes to choosing the basic structure of social distribution and the constitutional provisions which offer each of us the guarantees associated with social membership. For Dworkin religion appears as a mere vestige of the old order, a way of protecting the irrational and obstructive attitudes of the rednecks whom the Constitution is designed to subdue. In his writings on abortion, for example, Dworkin tries to borrow the religious value of the sacred, to give it a universal and secular content, and then to offer the ‘pro-choice’ camp the benefit of the belief that the sovereign choice of a woman in matters of reproductive life is something that it is legitimate to regard as sacred. This is a long way from the religious view, of course: but it is a strong affirmation of the liberal conception that religion, after all, is not what real disputes in politics are about. Even if religious concepts seem to capture the ideas of intrinsic worth over which we are contending, they can be voided of all religious connotation and made available to the liberal in their acceptable and sanitised versions.

Internal and External Perspectives.

In considering the place of religion in political life it is necessary to distinguish the internal from the external perspective. Someone can believe that religion is necessary to human communities, can even offer his endorsement to one particular form of it, without having any religious belief of his own. Such an ‘external’ perspective, which sees religion from outside, as a social phenomenon with costs and benefits of its own, is common among conservatives – especially among conservative anthropologists like Sir Henry Maine and Sir James Frazer. It goes hand-in-hand with the kind of broad utilitarian picture that we associate with Hume, though he himself was indifferent to the fate of religion as he knew it. From the internal perspective, however, religion offers a stark choice: either believe or disbelieve. And religious belief is not like scientific belief, in leaving room for doubt and scepticism: it is not the conclusion of an argument, but the result of upbringing or conversion, an existential predicament, whose main result is membership within the religious community. Rituals, sacred words, sacred texts and observances are all to be performed meticulously: get them wrong and you fall out of ‘communion’ with the believers, and the consequences of this can be lethal. This is an immensely puzzling yet wholly familiar feature of religions, and points to the role of religion in overcoming ‘difference’, and binding people to a common posture of obedience. The more mysterious and inexplicable the ritual, the more powerfully does it command us to join: for our joining then shows our submission to something greater than ourselves, something that we do not understand because it is literally unintelligible. That is why we make pilgrimages to Mecca, and do all the weird things there required of us. But this also means that religion is ambivalently connected to the political realm. If it is providing the core experience of social membership, what room is there for political organization, and how can politics compete with it? If, on the other hand, a community is organized politically, what room is there for religion among its members, and how can the absolute commandments of religion be reconciled with the compromises required by political order?

These questions are of concern equally to liberals and conservatives, especially now. Roughly speaking there is an Enlightenment consensus within Western communities, which is shared by liberals and conservatives, and which holds that religious freedom is the premise of political order. Unless each citizen is free to worship as his conscience directs, or not to worship at all, then there is not a truly political order, but order of another kind, theocratic or totalitarian. (It is important to see that the communist and fascist forms of totalitarianism are also derogations from political order, which, even if atheistical, are also insisting on a kind of religious rather than political unity among their subjects – unity founded in a shared observance, and a forbidding of the fundamental questions. Hence the enraged pursuit of heretics in communist societies.) The essence of political order is that it is negotiated, founded in compromises, based in membership of another kind from that implied by religious observance – the kind for which ‘citizenship’ is the name. Here is a quote from my book The West and the Rest:

‘… the contest between religion and politics is not in itself a modern one. This we know not only from the Bible, but also from Greek tragedy. The action of Sophocles’ Antigone hinges on the conflict between political order, represented and upheld by Creon, and religious duty, in the person of Antigone. The first is public, involving the whole community; the second is private, involving Antigone alone. Hence the conflict cannot be resolved. Public interest has no bearing on Antigone’s decision to bury her dead brother, while the duty laid by divine command on Antigone cannot possibly be a reason for Creon to jeopardise the State.

A similar conflict informs the Oresteia of Aeschylus, in which a succession of religious murders, beginning with Agamemnon’s ritual sacrifice of his daughter, lead at last to the terrifying persecution of Orestes by the furies. The gods demand the murders; the gods also punish them. Religion binds the house of Atreus, but in dilemmas that it does not resolve. Resolution comes at last only when judgement is handed over to the city, personified in Athena. In the political order, we are led to understand, justice replaces vengeance, and negotiated solutions abolish absolute commands. The message of the Oresteia resounds down the centuries of Western civilisation: it is through politics, not religion, that peace is secured. Vengeance is mine, saith the Lord; but justice, says the City, is mine.

The Greek tragedians wrote at the beginning of Western civilisation. But their world is continuous with our world. Their law is the law of the city, in which political decisions are arrived at by discussion, participation and dissent. It was in the context of the Greek city-state that political philosophy began, and the great questions of justice, authority and the constitution are discussed by Plato and Aristotle in terms that are current today.’

I go on to argue that Christianity, unlike Islam, has gone on to perpetuate and amplify the idea of the city, as a place of secular jurisdiction in which freedom of conscious prevails. The resulting idea of citizenship dates back to the Greek polis and the Roman idea of the civis romanus. But it came into its own with the Enlightenment, and it is in the name of citizenship that the French Revolution set out on its subversive path. To cut a long story short, the dispute between liberals and conservatives over the place of religion in civil life exists – if at all – as a dispute over the extent to which political order, even though by its very nature emancipated from religious authority, nevertheless draws on religious belief.

If civil association is sufficient in itself to foster intrinsic values and life-enhancing goals, then it has effectively taken over the traditional role of religion, which can exist henceforth merely as a support for those who need that kind of thing. This is the view that seems to be emerging in the tradition that begins with Schiller and ends with Oakeshott. Their vision differs from the usual socialist position in recognizing the purposeless, the intrinsically meaningful, and the ‘end in itself’ as the aim of political order, and in marginalising all utilitarian thinking. And it differs from the usual liberal position in recognizing limits to state action in the formation of civil association. In all probability it is a position incompatible with the normal liberal ideas of ‘social justice’, for reasons given in our class on justice. Its Aristotelian emphasis on friendship sets it apart from the nanny state, and delivers a vision of human society in which individual virtue, autonomous institutions and educational excellence play a role of the kind that is not clearly acknowledge in modern liberalism. Of course, you could say that classical liberalism is not so very different in these matters, and that is so. But of course, classical liberalism and conservatism co-exist in thinkers like Burke, Smith and Hume, and it is only in recent times, following J.S. Mill, that liberalism has taken its statist direction.

The external perspective may still contain a strong endorsement of religion, as providing some of the background of socialization on which even the most secular constitution may need to draw: this is the view that you find in Burke and Disraeli, for example, and maybe even Maistre. But this endorsement raises a serious question of how, exactly, the state should respond to the demands of religion and also to the demands of those who attack it. We see the conflicts here in the rival interpretations of the ‘no establishment’ clause in the Firest Amendment – conflicts between conservatives, who believe that this clause does not exclude religion from public life, and who believe that the state should acknowledge, even if it does not impose, the demands of religion, and liberals who take the clause to authorize a policy of complete and radical secularization. (We have discussed this earlier in the course.)

This is the point where I propose to leave off, since it connects to the problem of constitutions, and the topics introduced under this heading by Maistre.

Return to the Topics
Read Constitutions