What's real in political philosophy?
- First Online:
In the Beginning Was the Deed: Realism and Moralism in Political Argument Bernard Williams Princeton University Press, 2005
Philosophy and Real Politics Raymond Geuss Princeton University Press, 2008
In this review essay I ask what, exactly, is thought of as ‘real’ in two recent books that wear their realist hearts, as it were, on their sleeves: Bernard Williams’s posthumous collection of essays (some published for the first time) In the Beginning Was the Deed, subtitled by editor Geoffrey Hawthorn Realism and Moralism in Political Argument, and Raymond Geuss's Philosophy and Real Politics.
This essay also aims to make a modest contribution to the wider debates about reality and why it matters in political theory and philosophy. Developing debates about realism are taking on a bewildering quality, and any enquiry into what's at stake raises the usual difficulties with the construction or criticism of an ‘ism’. Typically a range of methodological, programmatic, strategic, empirical and value positions are gathered together, and because these are logically separable from one another the validity of any characterisation is inevitably contestable. Debates range over many dimensions. In a notable paper on moral realism, Peter Railton finds 13 with very little trouble – arguments cover cognitivism, truth, logic, objectivity, reductionism, naturalism, empiricism, determinateness, categoricity, universality, relativism, pluralism and the evaluation of existing moral norms (Railton, 1986, pp. 164–165). There are debates about realism in ethics, in philosophy of science, and in philosophical logic, and realism defines more or less distinctive positions in disciplines such as international relations, political thought and sociology.
Realism is not a clearly delineated position in Geuss or Williams; nor are they very clear about what reality claims they make, or why. Nevertheless, some themes are visible enough. They both argue that political philosophy must attend to the real motivations and reasons of real individuals in real settings, and the real effects of their actions (Geuss, pp. 9, 59; Williams, p. 37). Both are thus realist about politics itself, in a twofold sense. First, they deny that politics is merely epiphenomenal, an effect or distortion of something else. The major target here is moralism, the idea that politics is the application (or misapplication) of ethics (Geuss, pp. 7–8; Williams, pp. 2–5, 12). But the target could equally be the view that politics and political action are simply one class of economic exchanges and interactions. Second, they both characterise politics in terms of the causal operation of power (Geuss, pp. 27, 90; Williams, p. 13). As with the emphasis on real reasons and motivations, this amounts to a kind of explanatory realism, a commitment to the view that to explain or to come to an understanding of something must involve grasping the particular detail of an event or circumstance, and not constructing a formal model, nor coming up with a universal law. Also relevant here is that, third, both focus on the importance of historical circumstance: what is possible at one time would not have been possible at an earlier time. An implication drawn by both is that if liberal values have any significance in the world, the explanation for this lies in a concatenation of historical circumstance (Geuss, p. 60ff; Williams, pp. 25, 62–63). Again, this meshes with explanatory realism. However, it also conduces to realism in the different sense of a kind of clear-sighted, no nonsense attitude to how the world really is.
This realism is a particular moral and cognitive attitude, eschewing fantasy about how the world might be (Geuss, p. 59; Williams, p. 13). However, this does not mean that the position is wholly descriptive or normatively neutral. In both works realism is connected with normativity of the critical kind. Geuss particularly emphasises the combat against ideological illusion (Geuss, p. 53) Williams endorses what he calls the ‘critical theory principle’: that we cannot accept as legitimate any situation of power in which the parties’ own acceptance of it as legitimate is the outcome of the workings of the very power that calls for legitimation (Williams, p. 6). That is, realism prompts a question about what of reality we should seek to dismantle, and what to rebuild (although this theme is not as prominent in Williams’ work as we might expect).
These similarities between the two books are of course accompanied, and perhaps, from a general political philosophy point of view, outweighed by, significant differences – in intellectual sources and antecedents, in philosophical and disciplinary preoccupations. Geuss professes to be indebted here mainly to Lenin, while in this work Williams references Weber and positions himself in relation to Habermas (Geuss, pp. 23, 99; Williams, pp. 9–10, 15–16). Williams’ philosophical contribution to humanist ethics, of course, contrasts with Geuss's work in intellectual history. But both, and this is significant for the realism here, begin their books with Hobbes (Geuss, p. 21; Williams, p. 3).
These similarities and differences will be revisited in what follows. In the following sections I relate the realist themes that I have identified here to philosophical realism more generally (Philosophical realism section), to a more detailed discussion of the ‘realism versus moralism’ debates in political theory (Realism and moralism section), and to the theme of ‘realisability’ that in my view is at the heart of both Williams’ and Geuss's realism in political philosophy (Realisability section). The analysis here raises questions rather than draws conclusions about associations with other realist positions in political thought and study, associated with Hobbes’ intellectual heirs.
Realism generally signals a particular philosophical analysis of existence – where existence is not a matter of our or anyone's (including, for instance, a deity's) ideas or perspectives. Realism is not synonymous with materialism – ideas or mathematical entities might be treated as real by a realist. Independence from beliefs or imaginings is quite consistent with some things being either dependent on us in the sense that their continuing existence and independence depends on the acts or omissions of someone or even of humanity collectively, or in the sense that they were made by some individual or individuals. Hence, social institutions, made objects, and even nature, are independent in the relevant sense for realism, while in other senses dependent on ‘us’. [My analysis here follows (without being directly attributable to) Harré and Madden, 1975; Newton-Smith, 1981; Lovibond, 1983; Outhwaite, 1987; Putnam, 1987; Searle, 1995; see also Scanlon, 2009.]
Debates about realism are often couched in global or universal terms, where what is at issue is the philosophical validity of understandings of reality itself – the point is to vindicate or condemn the philosophical coherence of the category reality, or the predicate real. A global realist position in philosophy is in contest with alternative positions, such as that talk of reality at all is irrelevant because philosophical analysis is a matter of the logical analysis of propositions and arguments, or that it is a matter above all of epistemology. More modestly, philosophers concede or presuppose that we can make sense of the idea of reality, and then engage in arguments about some subset of things such as moral values, scientifically knowable objects, social institutions, or mathematical or abstract entities. Arguments for realism, whether of the global or the modest kind, include reference to the best explanation, and arguments from the analysis of meaning, from the analysis of cause and from the analysis of truth.
What is the relationship between such philosophical analyses of reality and realism and the two contributions to realist political theory reviewed here? One answer to this question might be ‘none at all’.1 The thought is that realism in political theory is overwhelmingly a negative position – a rejection of abstraction, of formalism, of moralism, of reductionism in general. In contrast to these, alternative methodological principles are canvassed – for example, attention to context, or to facts, or a sociological rather than mathematical approach to power. These, it is argued, do not entail realist commitments in the sense of endorsement of the independent existence of anything. This is consistent with the view that the realism at issue in political theory is commonsense realism – which on closer analysis proves to be a hotchpotch of no-nonsense attitudes, including impatience with abstraction and utopianism, a straightforward commitment to a basic kind of materialism, and so forth.
Rhetorically, both Williams and Geuss rely on a robust scorn for the positions they reject. Williams, throughout his work, shows a penchant for brisk, and exasperated, ‘this is from all points of view idiotic’-type pronouncements (Williams, 2006/1980, pp. 167, 183, 204). In the current work he insists on platitudes that must be respected (Williams, 1985, pp. 28; 2005b, p. 13; 2005d, p. 73) in tones that are barely respectful of those who are oblivious to such platitudes. It's pleasingly witty and amusing for his readers, but from the perspective of this essay the point is that it also delivers the message that there is nothing heavily metaphysical going on here. Geuss's rejection of abstraction, formalism, utopianism and other sins of recent political philosophy also, in some places, carefully but in a lighthearted way disavows metaphysics and ontology. He rejects any ontology of ‘the political’. Rather than thinking of politics as a sphere or a domain, with the attendant risk of reification, we should think of speaking about politics as answering or asking distinctive kinds of question (Geuss, p. 23).
Nevertheless, commitments to independent existence are implicit in both Williams’ and Geuss's explanatory realism. Who does what to whom for whose benefit? What forms of power – coercion, persuasion, ideological illusion – are deployed in pursuit of domination or advantage? How are structures of inequality legitimated over time? For Geuss the answers to these questions are not to be couched in the abstractions of rational actor theory; even less should we have recourse to the idealisations of abstract systems of rights or justice. Rather, the answers have to be historically accurate. Our explanations of why matters are the way they are, similarly, have to make reference to prior states of affairs and historical transitions. Chronos – order, sequence – must be at the heart of explanation and understanding. But so too must be kairos: our understanding of the time, the historical juncture, that allows or disallows decisive intervention, or a process of transformation. At the heart of political action is the political agent who can judge the moment to act and make a difference (Geuss, pp. 31–32).
Williams insists that political convictions and ideals, on the basis of which actors (might) act, should be understood as the upshot of causes. Among these, of course, as we appreciate from his wider work, are established ways of life that, among other things, support and validate particular interpretations, particular instances in which things make sense or not, particular judgements of rightness or rationality (Williams, pp. 11, 13). Again, in our political philosophy, there is no point engaging in wishful thinking about reasons, events and effects, nor in applying an abstract grid-like contract theory, or Habermasian discourse ethics, which can only misrepresent reality (Williams, p. 16). Williams’ philosophical analysis of reason for action insists that the agent's adopting or internalising a reason for him or herself is a psychological and cultural, as well as an ethical, matter, and it is of great ethical importance – of importance to the agent's integrity – because one who acts for reasons that are not his or her own is being coerced, or is in some other way suffering from disintegration (Williams, 1981/1980, p. 106ff; 2006/1996, pp. 117–118).
The political aspect of this analysis is not spelled out in Moral Luck but in the volume in hand, continuous with this thought, Williams insists that application of principles or norms to action is likely to pose a political problem (Williams, p. 16). The exact nature of this problem, and ethically and philosophically acceptable solutions to it, are not elaborated very clearly in what we might think of as political theoretic terms. We can infer that decisions are always in tension with principles; the rightness of decisions is only judgeable post hoc. Furthermore, politics enters with the thought that the fact that, for us, now and around here, as he is inclined to put it, the process of the autonomous taking on of reasons that are integral with one's personality and considered cultural life, and hence that finesse this tense relationship between principled reasons and the quality of decisions, will be subject to criteria that are recognisably liberal in character. Williams emphasises that this is not because liberal values transcend history or have abstract, universal, validity. It is rather, following Weber, because there is a concatenation of circumstances in history in which states are required, on pain of illegitimacy, to justify themselves to each and every individual who comes under their authority. It's not that these justifications found liberal states: liberal states have been produced by the same historical forces that produce the circumstances in which legitimation has to have this liberal character (following Weber: of universal rights, a degree of bureaucratic equality in treatment of persons, protections from state power and so on) (Williams, pp. 8–9).
The causal processes, the motivations in relation to actions, the circumstances and distributions of power, and the histories of transitions and transformations that are adverted to in these philosophical strictures are treated by Williams and Geuss as real – they exist or existed, and their existence is or was independent of what any agent thought or believed, wished or fantasised, about it. This independent existence of circumstances, forces and motivations was a constraint on what actually happened: liberal institutions can come imperfectly into existence only because definite agents have certain motivations, only because certain technologies interact with certain social formations to enable laws and legal judgement, bureaucracies and rational procedures, autonomous choice among options in markets and so on.
It is tempting to say here, as a number of philosophers who contribute to the debates about realism in political philosophy are inclined to, that the point is that our norms are or must be constrained by the facts (or not, in the case of critics of realism) (Cohen, 2003; Estlund, 2008; Miller, 2008; Swift and White 2008). It is notable that neither Williams nor Geuss is inclined to invoke facts, and in my interpretation this is because they both endorse some version of critical theory – and critical theorists are notoriously suspicious of any assertion (or, rather, allegation) of facts. The relationship between facts and reality is notably difficult to specify. In some philosophical traditions facts are linguistic entities, typically taking a propositional form, and with a particular truth value (true). Most philosophers, though, now treat them as non-linguistic, and as mind-independent, distinct from propositions that are the bearers of truth value. Indeed, many philosophers use the concepts ‘fact’ and ‘truth’ interchangeably (cf Scanlon, 2009). Other formulations define facts as what make propositions true, or not, or facts as states of affairs (things done, circumstances). For some philosophers all facts are contingent: they might have been otherwise. For others, facts can be either contingent or necessary. (My account here is based on Hume, 1748 Sec iv pt1; Frege, 1997[1892, 1918]; Russell, 1912 esp. Chapter 12; Wittgenstein, 1955/1922, Ss1–3.)
Factuality is relevant to realism by way of the matter of givenness and its implications. That something is given, cannot be changed and must be accepted as factual is consistent with its contingency, and it is consistent with the thought that the future need not be like the past. Givenness is not equivalent to necessity. But this quality of independence, and to-be-facedness, meshes with the realism of theorists like Weber. A number of philosophers explicitly couch analyses of reality and realism in terms of truth (for example, Sainsbury, 2001, p. 330). On the other hand, factuality can also be understood to be either orthoganal to or antithetical to realism. Orthoganal because analyses of facts and truths can be quite free of the concepts of independence and existence that I began with, and conversely philosophical analysis of independent existence can be quite free of consideration of truth or facuality. Antithetical because focus on facts doesn’t discriminate among different kinds or degrees of independence. A fact is a fact is a fact: facts about the dream I had last night have exactly the same philosophical status as facts about the distribution among some population of a dream of justice, or facts about the causes of the end of armed conflict in Ireland.
Now, what I want is, Facts. Teach these boys and girls nothing but Facts. Facts alone are wanted in life. Plant nothing else, and root out everything else. You can only form the minds of reasoning animals upon Facts: nothing else will be of any service to them. This is the principle on which I bring up my own children, and this is the principle of which I bring up these children. Stick to Facts, sir! [p. 1]
Insofar as it is permissible to take Dickens’s and Woolf's texts as political theory (and I think it must be permissible) we can clearly observe that their satire consists in mockery of any simple-minded focus on facts as opposed to systems of justice. Further, it is plausible to say that both Dickens and Woolf disparage facts in comparison with focus on what is real. For Woolf, death and horror in war, differences in income between men and women really matter; for Dickens, the Gradgrinds of the world are a reality and a dangerous one.
Some of the same lines of reasoning apply I think to Williams. There are facts about legitimacy. There are facts about attributions of thick concepts like courage, and there are facts that stem from such attributions. There are facts about causes, and about the conditions of causes. It does not help to run these all together – to homogenise them, flatten them out, formalise them. It might well be a fact that a certain proportion of people accept as factual the description of rational economic man. The critical issue is the historical process by which this comes about (both the proportion and the acceptance). Certainly Geuss would always want to pay attention to the power processes by which such facts are constructed. For both of them the critical theory principle does the work in properly determining our attitude to facts, and realities.
Realism and Moralism
For Williams and Geuss realism seems to consist, in part, in the rejection of moralism, by which they both mean first, the idea that political philosophy is no more than a branch of moral philosophy or ethics, and second, the idea that political philosophy consists in the application of moral principles to political questions, whatever they are. This negative position suggests several things. First, that the realism/anti-realism division is made between political theory (or philosophy) on the one hand and moral philosophy or ethics on the other. The implication is that political theory should be the domain of realism, and we can leave the people in the ethics department to get on with something else – idealism or utopianism, abstraction or formalism, normativity as opposed to prescription, or ideal prescription as opposed to theoretical description. The distinction echoes the commonplace that politicians are or must be realists; the philosophers, poets, ethicists and artists can be dreamers. There are clearly a number of problems with this scheme, and a question as to whether Williams or Geuss would concur with it.
To begin with and most obviously the realism/anti-realism distinction is also drawn within moral theory and ethics. Williams himself is a notable contributor to debates about questions of cognitivism and objectivism in ethics, developing a position that is characteristically reluctant to be counted in any particular camp, but that is more in sympathy with realist themes than with the opponents of realism (Williams, 1995/1985, 1995/1993). Second, at the risk of stating the obvious, the realism/anti-realism distinction is also drawn within political theory and political philosophy. However, sometimes it is that distinction – theory versus philosophy – that serves to mark the realist/non-realist distinction.
In this article, I am using the terms political theory and political philosophy interchangeably. For some thinkers, by contrast, a good deal turns on the distinction between these two. For instance, Hannah Arendt rejected the term political philosophy in application to her own work because she associated it with ontology, and she favoured a focus on action. She also asserted the view that only political action has any relevance vis-à-vis political evils (that is, philosophical ethics does not). More recently, those who favour the term political theory often do so in order to distance themselves from what is known as analytic philosophy. The indeterminacy of this term notwithstanding (ask any three analytic philosophers what they analyse and what they analyse it into, and you will likely get three different answers if any at all), it is quite clear that analytic philosophy is a target for both Williams and Geuss. Williams does, though, favour the term political philosophy, not least because among his commitments is a rejection of the view that philosophy should be in the business of constructing and applying theories (Williams, 2006/2000, pp. 182–190; Mendus, 2007, pp. 368–369). Geuss uses the terms political philosophy and political theory interchangeably, and for him this is an aspect of his rejection of ‘Kantian’ political philosophy (Geuss, p. 16).
Within political theory, there is a debate between ideal and non-ideal theory. Non-ideal theory covers a range of positions. Theory of the second best emphasises that in case the best or ideal solution is not attainable, fallback positions can be worse than the status quo. Non-ideal theory can involve responding to the question of what we should do or aim for, all things considered. It can involve considering what is permissible to do or aim for, rather than what we ought. It can also involve considering the gap between what we ought to aim for and what we can do. And so on (Swift, 2003; Swift and White, 2008). Neither Geuss nor Williams says anything that lends significance to these particular distinctions.
The realism/anti-realism division is also sometimes connected with the distinction between analysis of fact-independent principles (which can be political principles, in the sense that they have the form of ought-statements with respect to matters like justice – matters of state and social policy rather than matters of interpersonal conduct) and the analysis of fact-dependent principles (such as principles of recognition or membership that are dependent on facts regarding the existence of relevant organisation) (Cohen, 2003; Miller, 2008, pp. 5–7). Sometimes it is made between the analysis of fact-independent principles and the application or realisation of those principles, given the facts (Miller, 2008, p. 38).
All these ways of drawing the distinctions can conduce to a view of an acceptably handy division of labour in which each sticks to his last if not his current university department, and a world in which there is room both for theories that have a good chance of being realised, and in that sense are realist, and those that have zero chance but are valid nonetheless (Estlund, 2008, pp. 264–267). A principle of laissez-faire is implied when Geuss suggests in a downbeat way that his realist political philosophy might be treated as ‘orthoganal to the mainstream of contemporary analytic political philosophy’, and in an equally downbeat way suggests that political philosophy will be ‘much more interesting’ when it breaks out of the straitjacket of a clear is versus ought distinction (Geuss, 2008, pp. 17–18). Williams also wants to stick with the broad school of analytic philosophy while hoping that it will be more reflexive, more responsible, more subtle about the fact-value distinction (2006/1980, pp. 167–168).
However, realism implies scepticism about the image of either ethics or idealist political philosophy as independent of reality, facts or contexts. So any laissez-faire principle is not without implied censure. Elsewhere in his book Geuss attacks any presumption that ‘ethics can be studied alone without locating it within the rest of human life’ (Geuss, 2008, p. 7; see also Geuss, 2005/2002, 2005/2003). The need for moral philosophy to ‘touch reality’ is a repeated call throughout Williams’ work and it should deal with concepts that ‘have a grounding in reality’ (Williams, 2005/1999, p. 21; 2006/1980, p. 159). Normative notions ‘cannot be forced on to a recalcitrant world’ (Williams, 2005/1999, p. 23). That is, Williams’s and Geuss's realism in political philosophy is concomitantly a realist critique of ethics and moral philosophy. It is not simply that they deny that abstract moral philosophy is sufficient for the analysis of political philosophy questions. It is that they consider abstract moral philosophy to be insufficient qua ethics too.
One realistic strand of political philosophy centres on the criterion of realisability. The idea is that realist theorists take prescription seriously. This involves, first, that theory should be prescriptive, not simply normative. The contrast here is that theory can be normative in the sense of reasoning about what ought to be the case, without prescribing that anything in particular be done. The philosopher is focussing only on ought, and not on what is, or is possible, or is probable (Estlund, pp. 264–265). The realist riposte of prescriptivism ties political theory to action and intervention. It gives full weight to the role of political theory itself, and attention to political power and institutions, in specifying what values might be pursued and realised, rather than simply giving to political actors and institutions the task of realising values that are justified independent of the political process (Frazer, 2006, p. 19; Philp, 2007, p. 4). It implies, second and accordingly, that prescription should be based on valid analysis of how things are – structures, norms, motivations, causal connections. Prescription involves strategy. This point relates realisability to descriptivism. It implies that a valid line of criticism of normative, prescriptive, or even simply formal, political theory might be aimed at the implied descriptive theory of the human condition, or subjectivity and agency, or social structure and economy.
The theme of realisability is present, in rather different ways, in Geuss and in Williams. In both cases it is connected with descriptivism and with pessimism and these three are the subject of this section. Geuss is very clear. The difficulty with abstract and formal political philosophy is that its significance for real life can lie only in its playing an ideological part in the world (Geuss, 2008, p. 96). By this, he means that its role will be connected with the fostering of illusions. In this connection, realist political philosophy will certainly concern itself with puncturing the pretensions of abstractions like St Just's or Bentham's ‘happiness’ (Geuss, 2005/2002). Realist political analysis, however, which pays proper attention to what actually motivates people, to the workings of power, and which aligns its evaluative vocabulary with history, can certainly aim at conceptual innovation, or at the dismantling of illusion (Geuss, 2008, pp. 49, 53, 59).
We have seen Williams’s repeated scepticism about any image of political philosophy as the construction of theory that is then applied. Rather, he seems to ask for a more realistic consideration of causes and consequences. For example, in his essay ‘Humanitarianism and the right to intervene’ his point is to get past any abstract consideration of moral right or duty, obligation however that is curbed by considerations of capacity, proximity or opportunity, to the practical consideration of who has the capacity to solve coordination problems, and who has the necessary resources (2005e, pp. 144, 149, 152). As he puts it in ‘Realism and Moralism’ we (we being political philosophers) must have a more realistic view of power, opportunities and limitations (2005b, p. 12). He emphasises the point that Geuss echoes: that powerful political discourses can be proleptic, and help create the conditions they foresee (2005/1999, p. 26). Of course, understanding of likelihoods regarding other people's actions, reliable knowledge of who the relevant other people are, and how they are situated, certainly can decrease the level of uncertainty. But these are not matters for philosophy as such.
These emphases on realisability, involving realistic consideration of causes and consequences, obviously presuppose a good description and characterisation of how things are. It is important to insert some clarifications and caveats here. First, of course, description does not have to be of things that are real – it is perfectly possible to describe an ideal, or a figment of the imagination, or any number of other things from which one would wish to withhold the predicate real. Second, description itself, even if it is of something indubitably real, can consist in part of idealisation; it can be particularised or include an element of abstraction; it can have the quality of Proustian detail, Woolfian impressionism, or Dickensian caricature. Third, we can add here that description can be critical and constructive – critical redescriptions of existing phenomena can reconstruct them as something new, or new descriptions of new phenomena can institutionalise those phenomena under that description. The theme of construction reintroduces to our analysis the idea of independence discussed earlier. We don’t say that all that is real is independent of any human action and understanding whatsoever, but real things are independent of what any one person or group of people think about them (it is possible to be mistaken).
Three elements of Williams’ philosophy can be pointed to here in connection with realism. First, his distinction between philosophical conceptions of reasons as internal or external. Second, his distinction between thick and thin concepts in political theory and ethics. Third, his suggestions about the relationship of political philosophy to social science. In each case, it is a pressing implication of his analysis that ethics and political philosophy need to engage in (critical) description and understanding.
In his analysis of internal and external philosophical construals of reasons, Williams argues that reasons, properly speaking, must be reasons for and of the relevant actor. An external imperative, or a judgement by one actor or group about what another actor ought to do, should not be presented as a reason for the second actor. This analysis verges on realism in two ways. First, there is the view that philosophers should focus on agents’ actual motivations – we can’t work out reasons a priori or in the abstract or without reference to how the world actually is. Second, it raises the matter of power and how it works. Obviously, to say that there is reason for an actor to act in a certain way has a rhetorical and power effect – in particular, it enables observers to judge that an actor who fails to act in the way they judge he should is irrational (Williams, 1981/1980, esp. p.111; 2006/1996).
Thick concepts – Williams deploys a range of examples including cowardice – are philosophically significant in that they are related to actors’ reasons for action in a way that thin concepts, such as utility, or justice as that is theoretically constructed in liberal political philosophy, cannot be (Williams, 1985, p. 140). The point is that because thick concepts are those that are properties of actual cultures and lives, institutionalised and understood by concrete actors, they are the ideas that guide understanding and action, that enable us to judge one another and ourselves. The philosophically constructed concepts are technical and abstract. Either they are not part of real lives or, if they are, they must be related to other concepts and values in such a way that they are, effectively, thick concepts after all. This means that any philosophical account of these real moral concepts has to be, at least partly, descriptive. At least, any very clear distinction between description and prescription, between fact and norm, is and ought, is problematised because, to take the case of cowardice, to describe someone as a coward is to evaluate and deploy a norm. Conversely, to analyse the norm, to question its validity, to ask whether its place in a culture or a life is justified, to ask whether its application to a case is justified is to engage in description, or redescription, of a culture (Williams, 1985, pp. 139–145; 2005c, pp. 48–49; 2006/1980, p. 158).
Both the analysis of internalism regarding reasons and the thickness of moral and ethical concepts link to Williams’ view that political philosophy must interrelate closely with the social sciences. Political philosophy should include the study of complex values, and it should deploy a reflexive sensitivity to its own contribution to the social and psychological institution of these values (2006/1980, p. 167). These themes also occur in Geuss's presentation of realism. For him, political philosophy must be realist in the sense that it starts from a concern with how institutions actually operate, with real motivation. The tasks of political theory include understanding and explanation, aspiring to go beyond a ‘low-level descriptive account’ to a level of generalisation. A reasonable generalisation about human beings, furthermore, is that they are inclined to evaluation: hence the dimensions along which states or affairs or events can be evaluated as more or less, better or worse, are also part of the describable reality (Geuss, 2008, pp. 37–39).
Realist description is consistent with critical theory. This can be put more strongly: a realist approach to ways of life, normative structures, social institutions and political capacities will be cognisant of the processes by which facts come to be constructed. This process itself is a matter of the power, the effort, the clashes that political theory has to address. As both Geuss and Williams emphasise, it is also important to recognise that ‘theories can have distinct effects’ (Geuss, p. 12). Geuss's position here is complex. Most straightforwardly, he argues that the point of political theory is, in part at least, to have effects and make differences: to orient agents and to supply representation, to contribute conceptual innovations, to combat ideological illusion (Geuss, pp. 40–53). Second, he canvasses the idea – admittedly controversial – that a task of political theory could be to foster ideological illusions. More importantly, third, he is concerned with drawing attention to the ways political theories, wittingly or unwittingly, have political impact – because aspects of them (and not necessarily the ones their producers would wish) chime with and might amplify social trends (Geuss, 2005/2003, pp. 35–37). Williams is less clear about this: he entertains both the idea that theories themselves have social explanations, and that they can also have social effects, emphasising that the merits of any particular allegation of either of these have to be determined on a case by case basis. His point, though, is that the analytic style of philosophy is too often ill equipped to examine its own impact, or to examine the impact of social reality on it, a point echoed by Geuss (Geuss, 2005/2003, p. 36; Williams, 2006/1980, p. 159).
Both Geuss and Williams can be understood as participating in a tradition of political and social thought that is pessimistic in that they both dissent from the idea of harmony as a political goal. Williams questions the desideratum of harmony between values that is common in analytic political philosophy (Williams, 2006/1980, pp. 162–163). Geuss's criticism is more straightforwardly sceptical about any philosophical enterprise that seeks harmony between principles, given the nature of ‘power, its acquisition, distribution and use’, and the inevitability of political judgement. Moral and political philosophy need to be more honest about their own participation in power (Geuss, 2008, p. 99). They both dissent from the idea that if political philosophers can get principles, concepts and inferences correct, then that correct order of things could be applied in real life. For Geuss, no political philosophy or theory can actually be applied in any neat way (Geuss, 2008, p. 96).
In his late work Williams reemphasises themes that have been present throughout his academic thinking about political theory questions. Among these are the centrality of disagreement, and the concomitant need for decisiveness (Williams, 2005b, pp. 13–14). It is also an ineliminable feature of politics that staying in office is a precondition for political action (Williams, 2005b, pp. 13–14). Accordingly, a distinctive orientation to the ethics of politics is needed, and Williams, in a Weberian spirit, focusses on legitimacy and responsibility (Williams, 2005b, pp. 12, 15). It also must presuppose the ‘critical theory principle’, by which he means that the legitimation of power over a person is no legtimation at all, if the coercive power which is supposedly being justified is used in order to secure the person's assent to it’ (Williams, 2005b, pp. 6, 27).
Now clearly the characterisation of these principles and understandings as pessimistic is questionable. There is a focus on indeterminacy, conflict, maintaining power and coercion. All of these can be understood to be politics’ downside. But these bad things about politics are also the good things about it. We want people in office to make political decisions; ineliminable disagreement means that people express dissent and this is tied to important forms of freedom; taking coercive power seriously means not pretending it is something else, or disguising it as harmony. Clearly there is a difficulty with this line of thought if it ushers in a world in which, in the words of Shakespeare's Timon ‘black [is made] white, foul fair, wrong right, base noble, old young, coward valiant’ (Timon of Athens 4.3.28–30). That political values have to be understood and evaluated in the context of the reality of how political power and its legitimation work does not commit us to the view that wrong is right.
There are contrasting inflections in Geuss's and Williams’ realism, despite the fact that, perhaps surprisingly, there is also a good deal of coincidence between their respective analyses of what's real in politics and political philosophy. Both avoid realism in politics in the sense of a focus on the ontology of ‘the political’ or anything similar. For Williams, political is mainly an adjective signalling a particular kind of response to a particular set of problems, and politics the noun names a characteristic set of activities and modes of conduct. Geuss distances himself from any view of politics as a distinct domain of life and institutions, focusing instead on political actors and thinkers when they respond to a particular set of questions and problems. These questions and problems, it is implied, are irrelevant from the point of view of economics, or ethics, or established traditional cultures.
However, this shared metaphysical (or perhaps anti-metaphysical) approach is accompanied by contrasting ethical attitudes to politics. In general, Geuss adopts a more exuberant attitude to the realities of politics as power. The inequalities he brings into view are those of class and status, based on exploitation and oppression. Underpinning these are the irreducible realities of threats of disorder that are different in quality, perhaps worse, than the social inequalities that we benefit or suffer from now. Political philosophy can play a role in dismantling the ideologies that contribute to the sustaining of these structures of inequality. It will always have a role in the route out of violence and war, and the institutionalisation of some contingent order. It can also adopt a role in the maintenance of more egalitarian just structures, or even new inequalities that are in some sense preferable to the existing ones. Williams’s liberalism is characteristically less exuberant. The problem facing us is that of disorder to which a political structure of power and authority is one possible response. Given such a political structure, power holders and wielders face questions of justification and legitimation; and those who have power wielded over them face an array of options for action, from acquiescence more or less sullen, to challenge and conflict. For Williams, an important question for political philosophy and social science is that of the conditions under which liberal values can govern this process.
Geuss clearly looks for challenges to class exploitation, and can envision processes by which exploitative structures might be dismantled. One element of his realism is the clear-sighted, hard-nosed attitude that such processes themselves will involve the uses of coercive power, and success will depend on the wise (politically wise) identification of the right time, and the skilful use of available resources. Williams comes closer to a Weberian emphasis on responsibility as the twin political virtue of wisdom. He emphasises the inevitability of politics and its dilemmas, and endorses liberalism as a solution because of its status as a historically developed real possibility – a set of ideas and ideals that is broadly congruent with twenty-first century states, societies and individuals (or, some of them, at least). The differences between Williams and Geuss, then, are mainly characterised as differences in the partisanship that both of them consider to be an ineliminable (and perhaps unavoidable) aspect of political relations and action.
‘Realism’ in the context of political thought and theory is a protean and multivocal term. In this review article I have concentrated on its philosophical connotations of independence and existence. In science these themes have their counterpart in the philosophy of explanation and realist (as opposed to formal) analyses of cause. In connection with Williams and Geuss, the subjective connotations of realism as the eschewing of fantasy are also significant. How these themes connect to other meanings of realism, such as the commitment to the study of power politics, and inter-state anarchy, that are common but by no means uncontested in international study, is a topic for enquiry that would take much more space than is available here. It is suggestive that both Williams and Geuss do address the idea that disorder, and the uses of power within and to resolve disorder, come first.
This is the response of Edward Hall, see Acknowledgements above, and Hall (2009).
This review has benefitted greatly from suggestions and comments on an earlier draft by Edward Hall, whose understanding of Bernard Williams’s realism is quite different to the one set out here (he argues that it amounts to not much more than opposition to ‘moralism’), and by two anonymous reviewers for the journal. The revised version has incorporated thoughts I brought away from a hugely enjoyable reading group consisting of Lois McNay, Michael Freeden, Marc Stears, Mark Philp, me, and Bonnie Honig and Carlo Bonura (who organised it) during Bonnie′s stay in Oxford in 2009. Marc Stears read a later draft, and has helped me greatly with very insightful and constructive observations, some of which I have been able to respond to in this final version. I am very grateful also to Volker Halbach for a clarifying and interesting conversation about problems of translation in connection with the term ‘fact’ between Frege and Wittgenstein, and Russell and later anglophone philosophers. If I am able to say anything helpful about realism here it is thanks in great part to the support of my colleagues; any errors in my arguments and interpretations are entirely my own.