The ‘dialectical’ theory of conservatism

Recently, James Alexander has proposed a ‘dialectical definition’ of conservatism which, he believes, goes beyond ‘dispositional’ definitions, such as those proposed by Brennan and Hamlin, and by Martin Beckstein, which are ‘incomplete’. Alexander argues that, by focusing on conservative responses to ‘ruptures’ of continuity, his expanded account exposes the ‘fundamentally contradictory’ nature of conservative thought. This article offers a critique of Alexander’s ‘dialectical definition’ of conservatism, highlighting its inconsistency with the ideological content long agreed by conservative political thinkers, and with the historical realities of conservative political practice. But it also shows that there is a valuable and rightful place for a political ‘dialectic’ as part of a theory of conservatism that is more consistent with the history of conservative thought and practice. It is a dialectic with many historical precedents in political theory, two of which are examined in detail: (1) the earliest, found in Plato’s Statesman; and (2) an innovative and particularly useful formulation of it to be found in the political philosophy of R. G. Collingwood. Conservatism: disposition, political practice, ideology In the search for the meaning of conservatism, it has long been common to cite a disposition of mind. The best-known example probably remains Michael Oakeshott’s essay ‘On being conservative’ (1956), the subject of which the author tells us quite explicitly is a ‘disposition’ or ‘temperament’, and not a creed or a doctrine. More recent examples include Martin Beckstein’s ‘What does it take to be a true conservative?’, and Geoffrey Brennan’s and Alan Hamlin’s ‘Analytic conservatism’. Generally, these ‘dispositional’ theories agree that conservatism is defined by an orientation, or ‘bias’, in favour of the status quo. Accounts vary in the justifications they offer for such an orientation: it may be simply the belief that the status quo is worth defending; or it may be that the proposed innovation is likely to backfire, or will at least introduce hidden costs, perhaps due to an abundance of ignorance; or it may reflect a general scepticism about our ability to understand political and social affairs in all their complexity, variety and diversity. Brennan and Hamlin have pointed out that conservatives might also justify their disposition ‘nominally’, by pointing to the ‘particular value’ that we should assign to existing social conventions and prevailing practices, but which we cannot assign to planned or hypothetical conventions and practices. They have also described how CONTACT Christopher Fear C.Fear@hull.ac.uk JOURNAL OF POLITICAL IDEOLOGIES 2020, VOL. 25, NO. 2, 197–211 https://doi.org/10.1080/13569317.2020.1750760 © 2020 Informa UK Limited, trading as Taylor & Francis Group conservatives might appeal to the ‘practical’ belief that established conventions allow us to predict the actions of others, and therefore help us to coordinate our own. All sorts of considerations can support the conservative disposition, then, and some of them are social and political in character. Alexander’s ‘dialectical’ theory In his recent article, ‘A dialectical definition of conservatism’, James Alexander argues however that political conservatism cannot be reduced to a mere disposition. This is, in fact, a common observation among analysts of conservatism: Viscount Hailsham (Quintin Hogg), Noël O’Sullivan, Sir Roger Scruton and Anthony Quinton have said the same. For Alexander, ‘Beckstein’s, or Oakeshott’s, type of theory cannot be taken to be an entire theory of conservatism. It can only be considered an element of such a theory: a part of such a theory but not the whole; in fact, no more than the first moment of a dialectical definition of conservatism.’ The ‘dialectic’ Alexander describes plays out between two contradictory attitudes to what he calls ‘a rupture with the past’, typified by revolution. One conservative attitude is to recognize that a rupture establishes a new order which, to be conservative, one should accept and henceforth conserve. The other attitude is not to accept the rupture and to act in order to reverse it. The contradiction that is fundamental to conservatism, Alexander thinks, is that it ‘cannot sanction revolution and yet has to sanction the order which follows revolution’. These two attitudes to the rupture are not only apparently contradictory. They are irreconcilable. So this ‘dialectic’ that Alexander posits as the essence of conservatism is not constructive: it does not result in some higher harmony, some kind of ‘synthesis’. Neither do these contradictory ‘moments’ cancel each other to reveal some common idea or positive principle contained more deeply. ‘In fact’, Alexander writes, conservatism is not only situational, reactive, concessionary and inertial. It is empty. If the conservatism of the first stage is the inclination to have and hold, then at this second stage it is the negation of all attempts to change what is had and held for the sake of an idea. It is a via negationis. This means that it is impossible to say what conservatism conserves. In fact, it can conserve anything. Conservatism is therefore ‘incomplete, unconscious, criterionless – something which, lacking a criterion for anything, can only point to “what is”, as if it supplies its own criterion’. Thus, characterized forever by two mutually exclusive responses to ruptures of continuity, and with no firm content of its own to determine consistently when one response or the other is correct, conservatism is, for James Alexander, untenable by

In the search for the meaning of conservatism, it has long been common to cite a disposition of mind. The best-known example probably remains Michael Oakeshott's essay 'On being conservative ' (1956), 3 the subject of which the author tells us quite explicitly is a 'disposition' 4 or 'temperament', 5 and not a creed or a doctrine. More recent examples include Martin Beckstein's 'What does it take to be a true conservative?', 6 and Geoffrey Brennan's and Alan Hamlin's 'Analytic conservatism'. 7 Generally, these 'dispositional' theories agree that conservatism is defined by an orientation, or 'bias', in favour of the status quo. Accounts vary in the justifications they offer for such an orientation: it may be simply the belief that the status quo is worth defending; or it may be that the proposed innovation is likely to backfire, or will at least introduce hidden costs, perhaps due to an abundance of ignorance 8 ; or it may reflect a general scepticism about our ability to understand political and social affairs in all their complexity, variety and diversity. Brennan and Hamlin have pointed out that conservatives might also justify their disposition 'nominally', by pointing to the 'particular value' that we should assign to existing social conventions and prevailing practices, but which we cannot assign to planned or hypothetical conventions and practices. 9 They have also described how conservatives might appeal to the 'practical' belief that established conventions allow us to predict the actions of others, and therefore help us to coordinate our own. 10 All sorts of considerations can support the conservative disposition, then, and some of them are social and political in character.

Alexander's 'dialectical' theory
In his recent article, 'A dialectical definition of conservatism', James Alexander argues however that political conservatism cannot be reduced to a mere disposition. 11 This is, in fact, a common observation among analysts of conservatism: Viscount Hailsham (Quintin Hogg), Noël O'Sullivan, Sir Roger Scruton and Anthony Quinton have said the same. 12 For Alexander, 'Beckstein's, or Oakeshott's, type of theory cannot be taken to be an entire theory of conservatism. It can only be considered an element of such a theory: a part of such a theory but not the whole; in fact, no more than the first moment of a dialectical definition of conservatism.' 13 The 'dialectic' Alexander describes plays out between two contradictory attitudes to what he calls 'a rupture with the past', 14 typified by revolution. One conservative attitude is to recognize that a rupture establishes a new order which, to be conservative, one should accept and henceforth conserve. The other attitude is not to accept the rupture and to act in order to reverse it. 15 The contradiction that is fundamental to conservatism, Alexander thinks, is that it 'cannot sanction revolution and yet has to sanction the order which follows revolution'. 16 These two attitudes to the rupture are not only apparently contradictory. They are irreconcilable. So this 'dialectic' that Alexander posits as the essence of conservatism is not constructive: it does not result in some higher harmony, some kind of 'synthesis'. Neither do these contradictory 'moments' cancel each other to reveal some common idea or positive principle contained more deeply. 'In fact', Alexander writes, conservatism is not only situational, reactive, concessionary and inertial. It is empty. If the conservatism of the first stage is the inclination to have and hold, then at this second stage it is the negation of all attempts to change what is had and held for the sake of an idea. It is a via negationis. This means that it is impossible to say what conservatism conserves. In fact, it can conserve anything. 17 Conservatism is therefore 'incomplete, unconscious, criterionlesssomething which, lacking a criterion for anything, can only point to "what is", as if it supplies its own criterion'. 18 Thus, characterized forever by two mutually exclusive responses to ruptures of continuity, and with no firm content of its own to determine consistently when one response or the other is correct, conservatism is, for James Alexander, untenable by definition.

Problems with Alexander's theory
The old observation that political conservatism cannot be reduced to a disposition seems fair, even when the reasons given for that disposition might be, as in Brennan and Hamlin's work, ostensibly social or political. 19 Indeed, it is questionable whether a disposition that refers to a reason is actually simply a 'disposition' in the first place, and not already something like an argument.
Further, there are not only conservative dispositions in political life, but also conservative political parties and governments. And although they certainly might contain dispositionally conservative people, the histories of conservative parties reveal more complexity in their activity than could be reduced to a psychological disposition.
But conservatism is also more than the historical vicissitudes of conservative politics: it is a political-philosophical tradition, or an 'ideology'though this has often been denied. 20 By this I mean only to say that conservative ideology is something different from the words and actions of any political actor or party in particular, consisting not (simply) in a descriptive compilation of those political actions and speech acts, but (also) in the normative thoughts that motivate and/or determine them and/or the demonstrative language that justifies them. 21 And, as Joseph Femia has pointed out, these cannot be simply inferred from a psychological disposition either. 22 So although the meaning of conservatism might certainly include the conservative disposition, as Alexander points out, it is not exhausted by it.
There are however a few things that are not quite right about Alexander's 'dialectical definition'. First, it is unconvincing that there are not and never have been any consistent rules, principles, standards, or commitments to guide conservative political action or argument. Rather than 'anything', conservative authors have typically agreed on the importance of conserving certain things in particular: the establishment 23 and the balanced constitution, 24 autonomous institutions, 25 the organic society, 26 authority/ leadership, 27 discipline and allegiance, 28 law and social order, 29 the nation state and patriotism, 30 hierarchy, private property and the market, 31 the family, 32 classical education, established religion and morality, 33 tradition, 34 pragmatism, the piecemeal solution of problems, 35 and so forth. A theory of conservatism should, I think, accommodate or explain the major long-running themes and arguments of conservative political theory. Alexander's theory instead dismisses them.
Second, there is one traditional referent of conservative thought in particular that is conspicuously absent from Alexander's account and which, I think, cannot really be omitted from any political ideology, including conservatism. It is liberty. (Or freedom, or autonomy, which I will treat here as synonymous.) Alexander thinks of liberty as one of the necessarily abstract, secular ideals that belongs to liberalism and therefore not to conservatism. And yet liberty has obviously been of central importance to conservative thinkers and politicians, who have usually been hostile to socialist government because of what seems in practice to happen to the freedom of the individual under it; and who have viewed transnational government sceptically because of its perceived tendency to erode the autonomy of the nation state. Burke's Reflections on the Revolution in France is full of references to the ancient rights and liberties of English subjects 36 and to the 'manly, moral, regulated liberty' which he professes to 'love'. 37 Hailsham thinks that Conservatives 'regard themselves in the twentieth century as the true champions of liberty'. 38 Roger Scruton toothe modern conservative philosopher par excellenceexplains the central importance of freedom, liberty and autonomy in The Meaning of Conservatism, 39 and has very recently narrated 'the roots of British freedom' in some detailan important part (as he sees it) of understanding Where We Are. 40 Conservatives would, I think, be surprised to learn that liberty is, in fact, no true part of their thoughts and actions as conservatives, but merely a buzzword they have stolen from liberalism. 41 And third, Alexander's definition relies very heavily on the concept of a 'rupture' in continuity, since it is in responses to rupturesand only in responses to rupturesthat he identifies the defining contradiction of conservatism. 42 There are three problems with this, each highlighting the poor consistency of Alexander's theory with the practical history of conservative politics. First, the same variability might also be identified among the responses of liberal and socialist actors and parties to ruptures, so there is nothing characteristically 'conservative' about it. Second, conservatives have on occasion not only accepted ruptures, but also pushed for and effected them. The work of Margaret Thatcher and the 'New Right' has been thought rather radical, for example, particularly with regard to the de-nationalization of the management of some natural resources and public services. 43 An adequate theory of conservatism should be able to accommodate such episodes without taking refuge in the 'no true Scotsman' fallacy. But third, it is fallacious to categorize practical changes that conservatives have accepted together with those that they have rejected and overturned, and to claim that it is the responses that are inconsistent, rather than the conceptual category itself that is too broad. Such a range of responses is only inconsistent if all of the events that constitute 'ruptures' are essentially the same. But of course, in practice there are relevant differences between them. There is, for example, cautious reform and reckless populist legislation; there are manageable upheavals, major improvements, major retrogressions and downright political disasterseven collapses of a civilization. This is why Burke recognized the importance of distinguishing between reform based on precedent and reverence for antiquity, and fundamental change based on 'metaphysic sophistry' and 'presumptuous speculations'. 44 Though Burke's distinction is also rather crude, he did at least recognize the need for one. After all, a state 'without the means of some change is without the means of its conservation'. 45 These problems with Alexander's dialectical definition may be attributable to the fact that although he has sought to augment the 'dispositional' theory, he has also tried to do so with another purely 'conceptual' account of conservatism. 46 This means that his theory does not seek consistency with ideological tradition or past practice, which are contingent and mutable. I do not wish to argue that 'conceptual' questions cannot be useful, or indeed that we can only settle for some fully comprehensive master definition which captures the complete history and future of conservative action, language and thought. After all, conservative theory and conservative practice are different things and are discussed differently. But they are not separable things 47 ; in reality, they continuously modify and delimit each other. I am arguing only that a putative theory of conservatism, as of any ideology, ought to be at least partly historical; it should be generally consistent with conservative ideological tradition and historical conservative practice, and incompatibility with these should be rare.

An older 'dialectical' theory
Alexander deploys the idea of a 'dialectic' or 'contradiction' to make conservatism incoherent and to dissolve whatever concrete doctrinal content has hitherto been claimed for it. But the idea of a political dialectic has actually been used before to make sense of conservatism and to give it concrete doctrinal content. We find this political dialectic in the work of many conservative writers, from Burke to Scruton. But its formulation at the hands of two non-conservatives is particularly illuminating. I will outline these as briefly as I can and go on to explain how they might help us to understand conservatism (in part) 'dialectically', without falling foul of the same problems that attend Alexander's account.
The earliest discussion of this political dialectic (as far as I am aware) is to be found in Plato's Statesman, where it is explained that there is 'something astonishing in the relation' between those on the one hand who are 'especially orderly' and 'always ready to live the quiet life, carrying on their own private business by themselves', who are 'ready to preserve peace of some sort in any way they can', and those on the other hand 'who incline more towards courage' and desire a life 'which is more vigorous than it should be'. 48 Good statesmanship, Plato explains, requires 'intertwining' or 'weaving together' these two different types of character, to 'draw together a smooth and "fine-woven" fabric out of them', 'uniting parts of virtue that are by nature unlike each other, and tend in opposite directions'. 49 He goes on: For the dispositions of moderate people when in office are markedly cautious, just, and conservative (σωτήρια), but they lack bite, and a certain sharp and practical keenness . . . And the dispositions of the courageous (ανδρεία), in their turn, are inferior to the others in relation to justice and caution, but have an exceptional degree of keenness when it comes to action. Everything in cities cannot go well, either on the private or on the public level, unless both of these groups are there to give their help. 50 Plato's term 'souteria' (σωτήρια) might just as easily be translated 'preserving', 'saving', as 'conservative'. 'Conservative' in the political sense is, of course, a 19th-century innovation, so Plato's discussion long predates the 'conservative' political traditioneven as Anthony Quinton understands it. Whichever translation we take, it is clear that Plato is anyway not naming a political faction or a set of ideals, but (again) a group of citizens sharing a personal disposition. But it is a disposition with political significance. Good statecraft for Plato involves establishing a relation between the 'moderate' (or 'conservative') and the 'courageous', such that the overall result is harmonious and productive, rather than dissonant. This is a 'dialectical' theory of statecraft, then, in which the positive value of the 'conservative' is understood to derive from its role in some sort of process that requires another force in addition: here, 'courage' (andreia, ανδρεία). The meaning or significance of each dispositional group refers dialectically to the meaning or significance of the other.
This idea is not restricted to the writings of Plato. We can find some prominent conservatives, from Burke and Coleridge onwards, 51 recognizing a complementary role for something like 'order' in an ongoing political dialectic with an opposing force that represents something like 'progress'. Lord Cecil, MP for Oxford University from 1910 until 1937, writes that 'The two sentiments of desire to advance and fear of the dangers of moving, apparently contradictory are in fact complementary and mutually necessary', 52 and that 'though conservatism seems at first sight to be the direct opposite of progress, it is an essential element in making it safe and effectual'. 53 Something similar is also true of the complementary relationship between what Oakeshott calls 'the politics of faith' and 'the politics of scepticism'though Oakeshott tries to avoid the simple identification of the politics of scepticism with conservatism. 54 In general, then, this theory of statecraft that posits conservatism in a dialectical relationship with liberalism (or something like it), where the conservative has a moderating, cautionary, or deceleratory function, has significant historical pedigree, including among conservatives themselves.
One variant of the idea can be found in R. G. Collingwood's New Leviathan (1942). Collingwood calls his attitude to politics 'what in England is called democratic and on the Continent liberal'. 55 He does not seem to consider himself a conservative. But it is worth examining his version of this political dialectic in detail, not only because his framing of it in terms of 'recruitment' is very innovative, but also because it can better accommodate much of the practical political and ideological history of conservatism from which we found Alexander's theory to be rather distant.
For Collingwood, the 'life of the body politic' is to recruit or promote individuals into the 'ruling class' from the 'ruled class'. 56 It is essential not to confuse these with 'classes' in the economic sense that now prevails (clarification is provided below). 57 A body politic, Collingwood explains, consists of a 'social nucleus and a non-social circumnuclear body' 58 : the 'social' part comprises autonomous members between whom there is social consciousnessthat is, the realization of 'joint will' 59 ; while the 'non-social' community comprises them that have not (yet) awoken to 'consciousness of their own and each other's freedom', 60 and who are therefore not ready to join their individual wills with others in social enterprise. But importantly, the whole body politic is to be understood as 'a non-social community in process of turning into a society'. 61 Members of the ruled class 'must be susceptible of promotion into the ruling class', 62 not least because the ruling class 'must not be allowed to die out', 63 and must, therefore, replenish itself by recruiting from the non-social community. And because the ruling class must always be replenished, the question always arises of how it can do so to make itself as 'strong' as possible. 64 To this question, Collingwood says, two answers are always given. The first is to enlarge the ruling class 'so far as is possible. By recruiting into it, to discharge one or other function, every member of the ruled class who may constitute an addition to its strength'. The second is to restrict it 'so far as is needful. By excluding from its membership everyone who does not or would not increase its strength'. 65 The first answer Collingwood calls 'democracy', the second answer 'aristocracy'. Again it will prove important not to confuse this idiosyncratic use of 'aristocracy' with other, commoner meanings, such as an economic class, or a network of families with heritable titles and estates.
'There is', Collingwood adds, 'no quarrel between these answers', 66 i.e. 'democracy' and 'aristocracy'. The ruling class is strengthened by recruiting as widely as possible, but no more widely than is needful. If it recruits less widely than it could, it will be weaker than it could be. But it will also be weakened by recruiting more widely than it needs. The rate of recruitment must, then, be regulated in order to achieve the 'optimum velocity'. 67 And for Collingwood, it is this function, maintaining the 'optimum velocity', that the Liberal and Conservative parties between them perform, on account of the interaction, mutual checking and alternating rule of the acceleratory, 'democratic' principle embodied in the Liberals and the deceleratory, 'aristocratic' principle embodied in the Conservatives.
Because recruitment and promotion bestow liberties and duties upon persons, the process Collingwood describes is also the process of 'percolating' 68 liberty: 'To hasten the percolation of liberty throughout every part of the body politic was the avowed aim of the Liberal party; to retard it was the avowed aim of the Conservative party', 69 he explains; 'Both held it as an axiom that the process of percolation must go on', as both recognize the necessity of perpetual recruitment. But both also recognize that 'if it went too fast, and equally if it went too slow, the whole political life of the country would suffer'. 70 The role of the Conservative party is to be a 'brake' on the vehicle of progress. And it is necessary that vehicles have brakes, 'not to stop the vehicle, but to slow it down when it seemed likely to go too fast'. 71 'Nothing', Collingwood adds, 'could be a plainer statement of the Conservative's essentially dialectical function'. 72 Any account of conservatism that abstracts it from this dialectical process and supposes that conservatives and liberals are 'mutually opposite and hostile entities' is, he says, a 'false abstraction'. 73 For Collingwood then, Conservatives are not really committed to halting recruitment and the percolation of liberty, or to maintaining the status quo for its own sakethough we can understand how Liberals could mistakenly see it that way. Rather, Conservatives are committed to decelerating the rate of 'recruitment' and the percolation of liberty when it seems likely to go 'too fast'.

Provisional summary and analysis
So far we have an ancient political dialectic in which (1) the conservative (or moderate) is understood in relation not to changes, or 'ruptures', but (2) to another temperament, the courageous, the liberal. (3) The dialectic by which conservatism is defined is not (therefore) an internal conflict, as Alexander thinks, but a wider process of which the conservative is a part. (4) That wider process is constructive, the mark of good statecraft, and not only because it contains a liberal 'engine', but (5) also because it has a conservative 'brake'. And from Collingwood we have the further specification that (6) the constructive process is that by which the ruled are recruited into the ruling class; (7) which is also the process by which liberty is percolated through a body politic; (8) the two temperaments are here realized as two distinct and specifically political answers to the question of recruitment, the 'democratic' and 'aristocratic' principles, but which (9) are not in fact incompatible; and (10) the 'aristocratic', deceleratory principle is to be identified in conservatism.
Collingwood's version has some historical pedigree of its own. A dialectical relationship between an 'aristocratic' and a 'democratic' party is also found in Tocqueville's Democracy in America, where these two principles are 'as old as the world and all free societies display them under different names and forms. The first aimed to restrict popular power, the second to extend it indefinitely'. 74 It also has one recent exponent in the one-time editor of The Daily Telegraph, Peregrine Worsthorne, who writes: 'To the Conservative, the problem is to ensure enough stability and continuity to prevent tomorrow's aspirants to power pushing themselves upward so fast that nobody can rule in an orderly and civilized fashion. In other words, Conservatism is about resisting over speedy renewal of the ruling class, to the point where evolution descends into revolution'. 75 But Collingwood's version raises three questions: (1) What is the 'ruling class'?
(2) What does it mean for liberty to percolate 'too fast'? (3) Is it really true that conservatives (must) share with liberals the belief that liberty must percolate?
These three claims are central to Collingwood's theory, but much of his reasoning is implicit. By offering my own answers to these three questions I am explaining what use we can make of these ideas to understand conservatism, especially given that we are seeking consistency with ideological tradition and political practice. I am not offering the historical claim that this is what Collingwood really thoughtthough I think that what I have to say happens not to be very far from that mark either.
What is the 'ruling class'?
First, then, it is important not to mistake the 'ruling class' for an economic class, a network of families, or simply central government. Persons become members of the ruling class by virtue of the functions that they perform in society. 76 In practice, when persons are 'recruited' they do not (usually) join some unitary central authority, though they may, of course, elect representatives to it. Rather they are recruited into positions of responsibility, 'ruling', within the various institutions and associations that facilitate social activity: schools, businesses, courts of law, banks, research institutions, clubs, the armed forces, emergency services, trustswhich in Scruton's view are particularly important 77and so forth. For example, a schoolboy's 'recruitment' might begin with his being given minor administrative work to do for the school; further institutional liberties and duties may be awarded as he proves that he is willing and able to do more important workwhich is how his autonomy advances. He might later be awarded a teaching role in a university, and, in another field, the financial liberty of a mortgage so that he might buy his own house. He might meanwhile be a major 'ruler' in a trust or charitable organization. Importantly then, recruitment is by degree. It is a process by which one goes from being completely ruled in life, determined by people and things that are not oneself, to being increasingly self-determining, responsible, decisive, powerful, in and through one's institutions. So although Collingwood's account appears to be about merely the recruitment of personnel, it really concerns the basic question of politics in practice: how best should a society continuously bestow and delegate the liberties and obligations that it requires for its progress among persons who are equally mortal but not equally mature or competent?
The 'ruling class', then, is a classification of the various roles throughout society which involve some element of 'ruling'; its power is not concentrated in government, but is diffusedas conservatives have long said it should be. 78 Of course, a person might also become a cabinet minister. But all social roles are 'political' in so far as they pertain to institutions in a polis, even when they are not institutions of the state. 79 What does it mean for liberty to percolate too fast?
This process of recruitment is the process by which the individual becomes freer. No-one will ever be awarded all the liberties pertaining to every institution in society, so liberty can never be 'perfectly' realized. It is a matter of degree and always realized unequally. 80 Any such involvement and promotion furthers a person's liberty. But importantly, liberties only remain possible so long as the institutions which facilitate and give them meaning continue to exist and function, and these are forever changing as new people must be continuously recruited and promoted.
Collingwood says that the rate of this process in a society can be too fast. Now, it is obvious why vehicles must have brakes, but why must liberty? The notion is, as far as I know, unknown in contemporary political theory. Political philosophers theorizing liberty usually define it, and assess a state, a party, an ideologyor, more commonly, other political philosophersaccording to the provision made for liberty so defined. They are typically uninterested in whether the rate at which liberty is percolating is 'too slow' or 'too fast'.
The answer for Collingwood is that, if liberty percolates too fast, the ruling class is weakened. Part of what he has in mind by this is clear from what he identifies as the practical role of 'aristocracy': namely, to exclude from the ruling class 'persons not capable of ruling, granted the political problems calling for solution'. 81 The 'only permissible grounds for exclusion', Collingwood writes, echoing Burke, is inability to do the required work. 82 But it is vital to the welfare of the body politic that they be excluded; that they not be admitted until they are ready; that people be dismissed when someone better is available 83 ; that the ruling class be 'constantly purged' of 'incompetents', 84 that 'every member of the ruling class not up to that work should find his level'. 85 We should take this as follows. When institutional liberties and obligations are hastily given out to people who are not yet willing or able to perform them properly, institutions become less able to perform their functionto solve the problems they are set up to solve, to provide whatever it is they are supposed to provide, etc. If the functions of ruling are performed by complete incompetents, the institution may even be destroyed altogether. Of course, it might be very gratifying for incompetent people to be granted liberties as soon as they demand them, and for a while their personal sense of freedom may indeed be furthered. But the institutional cost could obviously be severe. And again, liberties are only possible so long as the institutions which facilitate them continue to exist and function. This is why children are not awarded the liberty of running their own schools; why people without training are not allowed to practice law; why people with no realistic prospect of repaying large loans are (or should be) excluded from the liberty of borrowing money from financial institutions. Institutions and the liberties associated with them might also need to be defended from other institutions, including those of the state. This is why, in practice, conservatives have cared about institutions including the family, property, the law, education, the Church, and about the maintenance within them of 'hierarchy'that is, a form of organization that facilitates the promotion and ruling of the competent and the exclusion and obedience of the incompetent.
As we saw earlier, the concrete content of conservatism is commonly identified as the will to maintain the status quo and/or order. But this is too imprecise and is inconsistent with the history of conservative political practice, where exceptions are very common. It is I think more accurate to say that what conservatives really want to preserve is the ability of institutions to execute their functions, and that this includes protecting institutions from the weakening effects of having their peculiar liberties too rapidly and indiscriminately distributed.
Must conservatives share the belief that liberty must percolate?
For Collingwood, the Conservative and Liberal parties of the past agreed that 'the process of percolation must go on', 86 and that liberty must percolate at the right rate. 87 They agreed on this so completely, he says, that 'one party could steal the other's thunder' 88as in the case of Disraeli's Reform Act of 1867, which doubled the number of enfranchised adults. At some point in the nineteenth century, Collingwood explains, this understanding was replaced by the belief that liberalism and conservatism are 'mutually independent and hostile entities', Liberals crying 'recruit them all!' and Conservatives 'don't recruit any!'. 89 Instead of recognizing theirs as one of two important roles in a dialectical process, the Liberals 'pictured themselves as dragging the vehicle of progress against the dead weight of human stupidity', believing the Conservatives to be 'part of that dead weight'. 90 The irony that Collingwood highlights is that, though the Liberals always prided themselves on 'having a philosophy', it was eventually only the Conservatives who understood their role correctly and recognized that there must be a party to represent the principle opposite to their own. 91 Readers can evaluate for themselves Collingwood's interpretation of these historical episodes, though it is worth noting that he doubts that the Liberal and Conservative parties were ever fully aware of their proper function. 92 But as I have already said, the understanding of a dialectical relationship pertaining between conservatism and liberalism has not been entirely lost. Scruton hopes for what he calls a 'reflective equilibrium', for example, 93 and that equilibrium is, for him as for Collingwood, based on a shared presupposition: what conservatives and liberals must all presuppose is 'that the perspective of the autonomous agent is inescapable, and that it is one of the first tasks of political existence, to ensure that it may flourish, as best it can'. 94 But of course the conservative, or a theorist of conservatism such as Alexander, might reject such 'shared' presuppositions. Either might argue that abstracting the meaning of conservatism from presuppositions shared with liberals is not a false move, but a more faithful one, since conservatives are committed to the real liberties and flourishings of the status quo, and need profess no commitment at all to liberty in the abstractthe potential, promised liberty of an 'enlightened' future. It seems to follow that conservatives do not and perhaps cannot share the characteristically liberal aim of percolating liberty among those who do not currently enjoy it; that any conservatives who have professed it have done so by expressing a 'liberal' streak in their thinking, or have done so disingenuously.
Again, though, this is inconsistent with conservative political history and ideology, where examples of liberty being percolated, in policy and in ideological narrative, are abundant. This is partly, I think, because in practice the choice between present and future liberties is a false one. Conservatives from Burke to Scruton have preferred present liberties to promised ones where the options have appeared mutually exclusive. But it is also clear that they want to preserve the real liberties of the present so that they might also be enjoyed by future people. What Collingwood calls 'percolation' should include what Burke means by 'transmitting' and 'perpetuating' 95 the liberties that become the 'inheritance' or 'patrimony' 96 of future generations. So to deny that the percolation of liberty is a fundamental aim for conservatives as well as for liberals is to restrict conservatism to the cry, 'don't recruit any!'.
Electorally, an 'anti-recruitment' form of conservatism could appeal to none but those already enjoying liberties which they did not wish to extend. But more significantly it would really be an ideological dead end, since it would have to deny the necessity of replenishment, and this would clearly be a mistake. The cry 'don't recruit any!' might be coherent if the 'structure' of a body politic is imagined in static terms; and it might be sustainable so long as institutions could continue to function with their current personnel. But as soon as we remember that societies exist in time and that they are made of people who were born, who mature and who die, leaving behind social functions that must still be performedthen a 'static' ideology, like any that treats a populace as structured but not also as undergoing processes, becomes practically unsustainable. A ruling class that refuses to replenish itself must shrink and weaken from starvation of human resources. Far from conserving present liberties for the ruling class's future members to inherit, such a society would bury them with its dead. 97 Conservatives, then, do not simply try to maintain the status quo. They recognize that the bestowing of liberties upon new recruits is necessary. But they see it as someone else's job to accelerate recruitment and the percolation of liberty when it is possible. Their own function is to decelerate when necessary: to restrict institutional recruitment to what is needed and to slow the distribution rate of liberties when it seems likely to go too fast.
Historical examples of conservative theory and practice consistent with this political dialectic lend themselves very easily. Burke's reaction to the French Revolution obviously expresses his concern that the autonomy of the French ruling class would be further weakened as novice revolutionaries extended their power unchecked throughout French civil society. Disraeli's Reform Act of 1867 has already been mentioned: it doubled the number of enfranchised adults, partly in recognition that such an extension was no faster than was now necessary and would anyway not weaken the 'ruling class'. The attempt to block the repeal of the Corn Laws 98 could also be interpreted as an attempt to prevent any weakening of the ruling class resulting from the hasty redistribution of the liberties governing the corn market. This dialectical theory is even, I think, consistent with the apparent ideological contradiction of ostensibly 'Conservative' governments embracing classical ('liberal') economics in the 1970s, 80s and 90s: those governments that broke up their own monopolies in gas, coal and oil, railways, aviation, banking, telecoms, etc. and sold the assetswhich Harold Macmillan memorably compared to selling off the family silver. Other motivations are not to be excluded, especially the desire to reduce borrowing by raising revenue in other ways. But nevertheless, it is quite consistent that conservatives might sometimes believe that the only way to preserve a socially valuable institution is to grant it the autonomy to manage its own recruitment. In a case of denationalization, conservatives might understand themselves (partly) as returning autonomy to an institution threatened by the effects of government control: the intrusion and promotion of incompetents, for example, the imposition of popular but impractical plans and rules, the burden of public 'transparency', complacence, inefficiency and possible obsolescence. For the conservative, freedom is 'percolated' when liberties, duties and the power to decide are relinquished by government and diffused. The chief concern of the conservative is, as ever, that the process be necessary and not proceed too fast. Reforms remain risky, of course, and misjudgements remain possible.

Conclusion
The aim of this essay has been to draw attention to a 'dialectical' theory in conservative political thought that is older than that recently proposed by James Alexander and, I think, preferable to it. I have not, however, claimed that conservatism is only this, and I have not sought to undermine distinctions between varieties of conservatism. 99 It has not been my intention to attempt a complete master theory of conservatism, or to reduce it to a single argument. Rather I have sought to explain the significance of one 'dialectical' theory of politics, which defines conservatism in relation to liberalism (or some sort of 'acceleratory' force)rather than in relation to Alexander's all-inclusive 'ruptures', or by reference to a problematic distinction between 'change' and 'reform'. 100 But as well as being older, this theory is, I have argued, also preferable and more generally useful in the study of conservatism. First, it accommodates the 'conservative' disposition which in some form has been discussed since Plato, but is not exhausted by it.
Second, it is consistent with the historical actions and words of the politicians and parties that are normally classified as 'conservative'. Relative to their opponents, conservative actors and parties really have typically sought to restrict institutional recruitment to what is necessary, and to decelerate the percolation of liberty when it appeared to proceed too fast. Third, it is consistent with the traditional concrete content of conservative political thought, for which Alexander has sought to substitute emptiness. Fourth, it restores to conservative ideology a central place for liberty, and particularly the liberties which are the real political content of what ought to be conserved. Indeed, the belief that liberties must be preserved and transmitted is one good reason for conserving institutions. And finally, it is also consistent with the traditional conservative suspicion of political philosophy and 'feverish ideologies'. 101 The characteristically abstract methods and nomenclature of 'philosophy' have more commonly been deployed to accelerate the awarding of liberties, in the name of social justice, liberty, equality and so on, but very rarely to explain the dependence of people's liberties upon their recruitment and promotion within social and institutional hierarchies. For many conservatives, the understanding of the relationship between our lives and our institutions, let alone how our innumerable institutions work and what they are for, is better acquired through practical experience of them than through the writings of political philosophers.