Two Contemporary Approaches to Political Theory

It is well-established and acknowledged by Rawls himself that he is a Kantian. I will not labour this point here. On the other hand, the accusation of Platonism is one which is less common, and one to which I will return below.
I thus leave the other most influential realist, namely Bernard Williams (2008). For discussion of his work, see Sleat (2010), Forrester (2012), Bavister-Gould (2011), Hall (2013), and Larmore (2013). I leave aside other clearly realist work as well, such as Wright (2010), Wolff (2011) and Mészáros (2011).
Note that I use the term 'realist' to refer to approaches to approaches in line with realism in political theory as introduced by Raymond Geuss, rather than 'realistic', which I use in its regular, everyday sense.
Note that, for Geuss, questions of timing are notably less important than the others.
Geuss himself has often clarified these points and others, but somehow many readers still manage to confuse his position.
For more critical discussion of Sen's critique of the Rawlsian approach, see esp. Valentini 2011, Thomas 2014, and Gilabert Forthcoming.
For one example of this, see Geuss 2003. Geuss also mentions the geneaologies of Nietzsche and Foucault as paradigmatic examples of this approach.
For more discussion of related themes, see Newey 2010.
On the issue of realism and ideology, see Williams 1980, 2002, and 2006, along with Williams, Newey 2010, Bavister-Gould 2013 Rossi 2014, and Rossi and Sleat Forthcoming 2014. Geuss (e.g. in his 2014), of course, focuses much more on opposing realism to wishful thinking.


Two Contemporary Approaches to Political Theory

Pfuel was clearly this kind of man. [The kind of man who 'imagines he knows the truth through a branch of science that is entirely his own invention, though he sees it as absolute truth.'] He had his science – the theory of oblique movement – which he had deduced from the wars of Frederick the Great, and everything he came across in today's military history seemed to him the most preposterous barbarity, a series of ugly confrontations with so many blunders on both sides that these wars were not worthy of the name of war because they didn't conform to his theory, did not lend themselves to scientific study.
Tolstoy, War and Peace

Introduction
Broadly speaking, the form that contemporary Anglo-American political theory has taken since the 1970s, with its reinvigoration in the seminal work of John Rawls (1999, 2001, 2005), can largely be described as Platonic and Kantian, focusing on the formulation, comparison, and evaluation of abstract principles of justice, and only subsequently looking to their application in the real world. From this point-of-view, it might appear that many of the greatest and most influential past philosophers did not do political theory proper – or at any rate did not spend very much of their time doing it. Instead of the precise formulation, comparison, and evaluation of abstract normative principles – be they of equality, principles of distributive justice, or what have you – these thinkers spent a lot of time on seemingly extraneous activities: historical investigation; detailed descriptive investigations that we might identify as social or human sciences; rhetorical appeals; practical advice; and so on. This is not to say that contemporary political theories do, or think one ought to do, nothing but formulate, compare, and evaluate abstract principles. Minimally, many of them also pay at least some attention to their application once formulated and adequately defended, to the epistemology through which they think such principles could be discovered or developed, accounts of how these principles legitimate certain institutions or policies, exceptions which might justify some (usually fairly restricted forms of) civil disobedience, and other things. However, the 'philosophical' focus of their work remains, in general, squarely on abstract moral principles. I suggest that this runs the risk of implying that a great deal of the work of many of the world's most important and influential political thinkers was not, or not really, political philosophy or theory, on account of the fact that the other elements they focused on to such a large extent, and the ways in which they are interwoven with their more recognisably normative components, are no longer seen as being within the proper domain of political philosophy or theory. In the same way that Tolstoy's Pfuel judges (what was for him) contemporary warfare, these thinkers risk claiming that past thinkers like Adam Smith and Karl Marx – among the most influential political thinkers of the modern world – were not really doing political theory at all, simply because their work does not fit certain very restrictive and abstract conceptions of what political theory really is, or ought to be.
Alert to these issues, a number of critics of the contemporary paradigm in political theory have emerged to offer their own alternatives. This article will discuss two of them: Raymond Geuss (2001, 2003, 2005, 2008, and 2010) and Amartya Sen (2010, 2012a and 2012b). Raymond Geuss's realist approach draws principally on Marx, the Frankfurt School, and Foucault, whilst Amartya Sen's comparative approach draws explicitly from the traditions of the enlightenment and political economy, which he takes to include, inter alia, thinkers like Adam Smith, Condorcet, John Stuart Mill, and Karl Marx. These alternative approaches seem to have been developed independently of one another, yet they have much in common. This article does three things. Its first section lays out the realist approach to political theory advocated by Raymond Geuss and the comparative approach advocated by Amartya Sen, respectively. Here I argue that Sen's comparative approach can, once some of its implications are drawn out, best be understood as a kind of realism – i.e. that comparative approaches to political theory form a proper subset of realist approaches. In the second section of this article, I explore what I take to be an underlying concern behind the work of both Sen and Geuss, and which, although it is rarely explicitly mentioned (and certainly never fully developed), I think we should bear in mind: namely a strand of anti-Platonism in political theory. Third and finally, I present three arguments for why a realist approach to political theory is likely to be superior to those of their contemporary adversaries: the argument from efficacy claims that realist approaches are likely to be more able to impact real politics; the argument from cognitive distortions claims that realist approaches are better equipped to notice and critically evaluate and revise undue assumptions and distortions of various kinds; and the argument from abstractness claims that since such distortions are likely to increase with a theory's degree of abstractness, and since realist theories tend to be less abstract than their competitors, a realist approach will likely be less susceptible to such distortions to begin with.

Realist and Comparative Political Theory
The realism of Raymond Geuss starts from the basically Hobbesian insight that social interaction and organisation is not simply given by nature; but something which must be produced and reproduced over time. Real politics is, fundamentally, about how these actions and interactions are organised, coordinated, and carried out. Ultimately, this approach centres on three questions. The first question is inspired by Lenin's 'who, whom?', and is expanded to: who does – and who can do – what to whom in whose interests? In other words, the first set of questions is about the agents of political action, and the contexts which enable and constrain them in various ways. The second set of questions concerns the timing of political actions: whether, for instance, a law should be introduced this year or next; or whether November 1917 is, or is not, the time to storm the Winter Palace. This also includes issues of the temporal priority of different actions. For example, invading Iraq directly after 9/11 would likely be a much harder sell to the American public than invading Afghanistan and then moving on to Iraq. The third set of questions involves issues of motivation, justification, and legitimation. After all, one of the things political agents very often need to do in order for their intentions to be carried out is to convince a certain number of other people that what they are doing is somehow beneficial, socially, morally and religiously acceptable, the just, right or good thing to do, and so forth.
The role of the political theorist, then, is built around this conception of politics and agency. The role of the political theorist is, minimally, to seek to understand this domain of real politics and political agency, and then to contribute, in some way, to improving the actions of the agents involved. This approach is contrasted with what Geuss calls the 'ethics-first' view of political theory, on which the work of political theory centres on the articulation, comparison, and evaluation of different individual or sets of abstract normative principles and which only thereafter seeks, if at all, to apply them to the world of real politics. The basic reason Geuss objects to this has been ably summarised by one of his commentators:

If political action is fundamentally influenced by existing power structures, by the heterogeneity of agents' evaluative rankings, and by historically and geographically diverse criteria of legitimacy, then it is not possible to understand such action simply or largely as a transparent implementation of normative principles. (Menke 2010: 140)

If it is not possible to understand political actions as the transparent implementation of normative principles, then an 'ethics-first' approach will make it difficult for such theorists properly to understand real politics. If this is the case, then it is plausible to suppose, further, that their theoretical contributions to political concerns – including their normative aspects – will suffer considerably as a result. Galston (2010: 394) rightly remarks that the conjunction of the realists' rejection of the ethics-first approach (or the 'priority of morality') and their 'focus on institutions leads realists to recommend a shift in the agenda of political theory'; a shift, that is, to questions of real political practice, centred on the aforementioned three questions, and the ways in which a theorist can contribute to it.
There is a growing literature on realism in political theory of various kinds (including Baderin 2014, Bavister-Gould 2011, Finlayson 2014, Floyd and Stears 2011, Forrester 2012, Frazer 2010, Galston 2010, Gledhill, Hall 2011, Honig 1993, Jubb 2012, Larmore 2013, Menke 2010, Mészáros 2011, Mouffe 2006, Newey 2010, Philp 2012, Rossi 2010, 2012, and 2014, Rossi and Sleat Forthcoming 2014, Sangiovanni 2007, Sleat 2010, Valentini 2012, Williams 2008, Wolff 2011, and Wright 2010). The discussion in this literature ranges over a wide variety of different, and only sometimes overlapping, topics: the importance of taking politics seriously as a particular form of human endeavour; the importance of social and historical context; the importance of stability and civil order; the evaluation and comparison of concrete institutions and policies; the persistence and ineliminable nature of political disagreement; the importance the complexity and heterogeneity of human motivations (including non-moral ones); about the influence, if any, of facts and practices on normative principles and ideals, and many more (for surveys, see esp. Frazer 2010 and Galston 2010, Philp 2012, and Rossi and Sleat 2014). Instead of exploring this literature in detail, I will stick to my stated focus on the works of Geuss and Sen. The rest of this section will thus clear up some of the misunderstandings about Geuss's realism specifically, before moving on to my discussion of Sen, arguing that his comparativism, rightly understood, forms a subset of Geuss's realism. The following sections will discuss their common theme of anti-Platonism and three arguments for favouring a political theory along realist and comparative lines.
Realist political philosophy in Geuss's sense is often misunderstood and confused with other, and very different, theses. First of all, realism in this sense ought not to be confused with 'realism' as the term is employed in the international relations literature. 'Realism' in international relations is essentially a thesis about how states or governments act in international affairs, and more particularly about how these acts are driven or motivated (for more detailed discussion, see Rossi and Sleat Forthcoming 2014). Realism in Geuss's sense is a thesis about how political theory can be, and ought to be, carried out; it makes no specific commitments to particular theories of the motivations or actions of any individual, set of individuals, or institutions. Geuss's realism does, however, demand that we take these entities and their motivations seriously, whatever they may be.
Secondly, realism in Geuss's sense should not be confused with calls – all too common in everyday political discourse – for critics of various existing social arrangements to be 'realistic' in their wishes and goals, in the sense of calling for their restriction and limitation. Alternatively, it can (as per Baderin 2013, Floyd and Stears in the introduction to their 2011, Honig and Stears 2011, Galston 2010, Kelly 2011, and Valentini 2012), be read as calling for less demanding normative standards, in order to avoid the charge of ineffective utopianism. Such calls often amount to demanding that people (political theorists included) and groups limit their goals to what is possible within certain existing structural parameters of the world as we know it – e.g. the existence of the state, or the existence of a capitalist economy. This reading is mistaken: Geuss's realism in no way calls upon people or organisations to limit their wishes or goals in any sense. It does call upon political theorists to take existing realities seriously when considering the various options available, but it does not insist that any particular components or aspects of these realities be taken as an unquestionable boundary that should restrict the normative or other ideals and standards that any individual or group may legitimately wish for or hold as a goal to work towards (for more on this, see Finlayson 2014). In fact, reflection on past and present realities may have the opposite effect: by enlightening us about past actualities, they can expand the scope of the possibilities w we are capable of envisioning, making us more ambitious as a result (see Lane 2011).
Thirdly, realism in Geuss's sense should not be confused with the use of 'realist' in various metaphysical senses found throughout different branches of philosophy, such as realism about the external world, about universals, or about moral properties. Briefly put, Geuss's realism does not commit him or any of its adherents to any such metaphysical theses about the existence of any such things or entities. With a basic understanding of Geuss's realism in place, I can now turn to Sen's comparative approach.
The comparative approach that Amartya Sen advocates holds that political theory ought to focus on particular sets of feasible alternatives and evaluate these in terms of their social realisations. This approach is distinguished from what Sen terms 'transcendental institutionalism', which instead aims to develop a conception of what a morally perfect or perfectly just society, or set of social institutions, would look like, and only subsequently seeks somehow to apply these conceptions to evaluate existing social institutions and to guide their replacement or reform. 'Transcendental institutionalism' sounds a lot like the 'ethics-first' approach recently discussed, but there is at least one major point of contrast: whereas they both focus on the formulation and evaluation of abstract normative principles, the former, namely transcendental institutionalism, further attempts to specify what a set of perfect real institutions constructed on the basis of such abstract principles – with or without some attention to certain basic facts about the state of the world, human nature, and so on – would look like. The 'perfect' institutions constructed in transcendental institutionalism need not be 'perfect' across all social and historical contexts; they may, like Rawls' justice as fairness, only aim at presenting what the perfectly just institutions appropriate to a modern society, however specified, would be like. Again, however, Sen's main problem with transcendental institutionalism is similar to Geuss's problem with ethics-first theories: they neglect the nature and importance of the options available, and their relative merits, in favour of a focus on abstract normative constructs. In so doing, they run a similar risk of rendering themselves unable to both understand and to remain theoretically relevant to influencing real political processes.
Now, it is important to realise that one can evaluate competing alternative institutions, legislation, and policy in light of these alternatives' social realisations and the values we deem to be appropriate without having any kind of conception of either what a perfect set of values, ranks, weights etc. would be, or what a perfect set of social institutions embodying such a perfect set would be like. In other words, having such a conception of a perfect society is neither necessary nor sufficient for making comparative judgments. Such a conception of perfection is not sufficient for making comparative judgments, since comparative judgments require additional information about what the alternatives available are: which are possible; which are feasible; what their consequences and inbuilt tendencies are; and so on. Without this additional information, no sensible comparative judgment is even possible. Nor is having a conception of perfection necessary for comparative assessments: we can perfectly well have a list of values – whether ranked and weighted, or not – and evaluate competing alternatives and their realisations in light of these without knowing what any perfect configuration of principles or institutions would look like. For example, we can judge policy-option A to be preferable to policy-option B on the grounds that, out of any number of values – all unranked and un-weighted – the two options only differ with respect to two, a and b, and option A does better than option B according to both.
Having just rejected one of the typical misunderstandings that the comparative approach tends to encounter, I have assumed, for the sake of argument, that the evaluation in terms of social realisations takes place on the basis of explicit, determinate values of some kind. This is not, however, strictly speaking necessary. It is entirely possible to evaluate alternatives in light of the information available and their foreseeable social realisations without making explicit use of determinate evaluative criteria or desiderata. An experienced political agent might be able, in light of a wide range of available information about the relevant alternatives (and their social realisations), to choose between them in view of this knowledge, without explicit reference to any determinate values. People – experts in particular – make such judgments all the time: the expert butcher knows when to apply pressure to the knife, when to hold back, and how to go with the grain of the muscle for each carcass; the expert javelin-thrower ineffably estimates how to take into account wind, rain, and how they feel on the day in order to make her throw; and the expert martial artist knows, without thinking, when to duck, weave, and counter-strike. Obviously all of these agents have goals and intentions, and likewise they have trained and practiced with a number of different values in mind; but the action in question, and the judgment on which it is based, does not itself explicitly utilise or make reference to these values when it is, in fact, made. If expertise or mastery is possible in politics, as it is in other aspects of human life and endeavour, we should be open to the possibility of political judgments which are made without explicit reference to determinate values. Nor should we suppose that action is always necessary or useful. If something is basically well-ordered, or if all the available actions can be expected to result in worse social realisations, an agent's best available alternative might be to do nothing at all.
Clearly Sen's comparativism and Geuss's realism are closely connected approaches to political theory. In fact, I think the former can best be described as a special case of the latter. If political theory is taken to start from questions of political agency and its contexts, and the task of political theory is to inform and improve it in some way, i.e. if one organises one's theoretical contributions around questions of concrete political actions and their contexts, questions of the timing of actions, and questions of the ways in which these in fact are motivated, justified and legitimated, then there are, as Geuss readily points out, a wide variety of ways in which one can proceed. One can write a genealogy that will cause political actors to think differently about certain taken-for-granted concepts and presuppositions, thereby influencing their actions; one can examine and lay out the structure of the contexts of actions that agents face, thereby enabling them to act more appropriately; one can develop new concepts, thereby enabling new modes of thought, which in turn might influence action in various ways; and many more. Another thing one can do is to examine the various feasible alternatives available to an existing agent when it comes to dealing with a particular issue or concern in terms of those alternatives' foreseeable consequences. Clearly, some understanding of the actions available to the agents and forces in any given society, and the contexts within which these operate, is essential to any conception of which alternatives are feasible, and which alternatives are not. Moreover, any estimation of an alternative's social realisations requires descriptive information or presuppositions about what the alternative's effects are likely to be. Since the effects of, for instance, an institution, a piece of legislation, or a particular policy are highly dependent on the context within which they are introduced, any estimation of its consequences will further depend, in part, on knowledge about these contexts – i.e. on the contexts of the relevant kinds of political action. On the other hand, in order to be feasible a candidate an alternative must not only be possible in the sense of being able to survive and maintain itself over time in light of certain basic parameters to do with human nature, planetary conditions, available resources, and so on, it must also be the case that there is some factor, agent or process, within an appropriate kind of context, which really can bring about the alternative in question. This, in turn, presupposes a descriptive understanding of the actors or agents available, their context of action, issues of timing and so forth, and how the relevant actions are, or can be, motivated, justified and legitimated.
In other words, any comparison of different alternatives in light of their social realisations must presuppose some conception of political agency and its context for determining what the different feasible alternatives are and what these alternatives' social realisations can be expected to be. As I have noted above, this is precisely where Geuss's realism insists that political theory should start from. This means that a comparative approach like Sen's is, in fact, a special case of Geuss's realism.

A Common Theme
Perhaps equally important, however, is the deeper intuition that both Geuss and Sen seem to be informed by. Neither thinker tends to render this intuition fully explicit (but see Geuss 2005, ch. 13), but I think their views can be much better understood if we do so. This is an intuition that there is a certain continuity between different kinds of human activities: that practices such as moral and political theorising and the political practices they ought to inform are not fundamentally different in kind from other, more familiar, practises such as building houses, constructing tools, and organising village agriculture on the one hand, and the reflections we make about these practices in order to inform and improve them on the other. I have already mentioned one aspect of this in my discussion of the possibility of political expertise. This has brought out an important difference between the abstract theoretical knowledge in which much academic discourse trades and the nature of the practical powers and capacities which are so central to a great many human practices and endeavours. Political agency and efficacy are questions concerning agents' practical powers and capacities – what they really can and cannot do. The function of realist political theory is to contribute to these powers and capacities and to the ways in which they are exercised. In other words, as I have repeatedly stressed, political theory along these lines aims to contribute to political practice, and such political practice is in no way fully determined by, or in any way subservient to, any recommendations a political theorist might attempt to make. In this sense, realist approaches – and, trivially, comparative ones as well – object to a certain Platonism that is tacitly implied in much of contemporary (at least Anglo-American) political thought.
The Platonism in question is the assumption that correct political practice requires, and is best guided by, some sort of theoretical and systematic account of the good, the right, the just, and so forth, which is supposed to determine – either fully or in part – any correct course of political action. Plato has a general argument to the effect that, as Geuss puts it, 'you cannot be performing an activity well unless you can explain why you are doing every component part of it in the way you are, and you can't do that unless you have the correct general theory' (Geuss 2014: 7). In the Ion (541e-542a), for instance, Plato's Socrates chastises Ion for his claims to expertise on the grounds that he can give no general account of what it is a rhapsode such as himself has knowledge of, as a result of which he cannot really be said to have knowledge about it at all. Since he has no such general theory, he cannot really know what he is doing, and so succeeds in doing anything well only by accident or by divine favour. On this way of thinking, it makes sense, at least ceteris paribus, for a society to be ruled by philosopher-kings, as Plato famously advocates in the Republic (5.473c-4a). Ruling, like everything else, is an ability that is performed correctly (and unaccidentally) only on the basis of solidly founded knowledge; in order to rule well non-accidentally one needs knowledge of the good in the form of a certain theoretical and systematic account; knowledge in this sense (especially of the good) is the domain of philosophers; therefore philosophers will be the best rulers. Philosophers' knowledge of forms or ideas, in particular of the good, means that they have the kind of expertise needed to steer a city well, just like the navigators' knowledge means that they have the right expertise to steer a ship well (Republic 6.488d). Naturally, Plato's view affords a privileged place not just to philosophers, but more specifically to the theoretical accounts or theories of the good, the right, the just, and so on, that they produce – or, more accurately with regard to Plato's views, which they get access to. Since practical activities do or teach correctly only accidentally unless based on an antecedent and true, theoretical and systematic account, the formulation, comparison, and evaluation of such competing accounts becomes a vitally important prerequisite for any kind of reliably good political action: only once we know what the good is can we be expected to do good; only when we know what justice is can we subsequently set about making the world more just.
This sense of Platonism in political theory should not be confused with a different conception of Platonism – sometimes attributed to G. A. Cohen – consisting in a 'basic contrast between fact-insensitive principle and mere rules of regulation' (Thomas 2014: 247, cf. Cohen 2008, p. 269 and 276ff, and, in a slightly different vein, Rossi 2014). Call this (Fact-Insensitive) FI-Platonism. Rawls' and other Rawlsians' principles of justice are not, strictly speaking, fact-insensitive, since e.g. they apply only under the circumstances of justice and take certain pragmatic concerns into account. Rawls is thus not a FI-Platonist, a point on which Cohen, who is one, criticises him. Accepting that at least some facts are prior to, and relevantly constrain and/or determine, the normative political theory one develops and advocates is clearly an anti-FI-Platonist position, and this arguably makes Rawls more 'realist' than some of his opponents such as Cohen and Nozick in the sense that it allows some facts about reality to inform his normative theorising (and here I agree with, inter alia, Gledhill 2012, Jubb 2014, Rossi 2014, and Sangiovanni 2007; cf. also Rossi 2010).
However, none of this obviates my charge that Rawls was a Platonist in the sense I have just outlined. The kind of Platonism I have accused Rawls of only holds that political action requires some sort of theoretical and systematic account of the right, the good, the just, and so on, in order to be reliably good. Call this S-Platonism. S-Platonism requires no absolute contrast between fact-insensitive principles and mere rules of regulation; consequently, the limited fact-sensitivity of Rawls' normative theorising, although it can save him from the charge of FI-Platonism, cannot defend him against S-Platonism. How are Geuss and Sen different from Rawls in this respect? In what sense to they reject S- Platonism in political theory?
As mentioned, Geuss and Sen start from a position which takes certain practices and their contexts for granted, and tries to contribute to how some agents act in light of them, whether this be to maintain, alter, destroy, or replace them altogether. Geuss and Sen, I believe, reject the S-Platonism in political theory I have been accusing Rawls (and others) of, because it is clear from their work that they believe that it is possible to act well in real politics, and reliably so, without any theoretical and systematic account of the good, the right, the just, and so on. My aforementioned point about political and other expertise brings out the contrast: no S-Platonist can accept the possibility of a political expertise such that it is possible reliably to act well without making explicit use of determinate values or principles of any kind; realist and comparative political theorists, however, can accept such a possibility. Note that, pace some of Jubb's (2014) comments (though he seems not fully to commit to them), this is not to claim that normative considerations are in any way irrelevant, or that political theorists should not concern themselves with such issues. Some information about what is just or unjust (in this context or in general), about what is good or bad (in this context or in general), about causal connections between courses of action, about the likely consequences of competing alternatives, and so on, may well contribute to improving political actions by making them more well-informed on a realist view; and political theorists may well be able to contribute to political actions via their contributions to such debates. None of this, however, requires such normative elements to come in the form of a theoretical and systematic account of e.g. justice. As we have seen in the discussion of Sen, no antecedent theoretical and systematic account of what things like perfect goodness and justice are in themselves is, strictly speaking, necessary for political tasks such as making comparative assessments between different feasible alternatives.
But if good politics and political theory is possible without developing a theoretical and systematic account of the good, the right, the just, and so on, this still doesn't give us reason to think that the realism of Geuss and Sen is better than their prevalent S-Platonist competitors. Even if we accept that reliably good politics is possible without a theoretical and systematic account of e.g. justice, and that political theorists can do normatively interesting work apart from developing, debating, and applying such accounts, we still lack good arguments for saying that the non-S-Platonist approach is preferable to the S-Platonist one. The next section will present three arguments to just this effect – i.e. for preferring realist and comparative political theories of the more S-Platonist approaches of thinkers like Rawls. I will argue that focusing the energy of political theorists on such S-Platonist tasks will be detrimental to their ability positively to affect real politics because thinkers who follow this approach (1) are likely to produce theories which will be less efficacious in the world of real politics, (2) are less likely and able to deal with certain local cognitive distortions of various kinds, and (3), as a result of their theories' necessarily highly abstract character, tend to be more susceptible to such distortions than their realist competitors in the first. I turn now to these three arguments.

Three Arguments for Realism
Here I will present three arguments for why we should prefer realist approaches to political theory: the argument from efficacy; the argument from cognitive distortions; and the argument from abstractness. These are all arguments to the effect that ethics-first approaches fail, and are likely to keep failing, the practical vocation of political theory compared to their realist and comparative competitors. The argument from efficacy claims that, in light of past and present experiences, realist approaches are more likely to be effective and influential in the real politics than their adversaries. Prima facie, this is just what we would expect from looking at what realism is: since realism, in Geuss's sense, consists in starting from an understanding of the nature and context(s) of political agency, and then attempting to intervene or influence political action in some appropriate way – including the writing of genealogies, descriptive analyses of economic and political institutions, making normative arguments for and against alternative courses of action, and so on – it would be very surprising if they tended to be less influential in real politics than approaches which instead began from highly abstract inquiries into the right and the good, and which only far downstream from that, if at all, deal with questions of application. Although any sort of statistical evidence in support of this claim is absent, I think a couple of points show it to be plausible. First, we should note that the political theorists who have been the most influential throughout history were, broadly speaking, realist in their approaches to the subject. Obvious candidates in this regard include Adam Smith's (1976) contributions in the Wealth of Nations, which greatly influenced economic and political life in Europe and the US (and still does, although to a lesser extent); John Stuart Mill's (2010) On Liberty, with its wide-ranging impacts on legal thought and reform; Karl Marx's work, which for generations has influenced political movements around the world; and many more. We might also note that, in addition to this, many prominent contemporary political theorists began explicitly advocating realism after and presumably partly in response to actual involvement in real politics. Both Bernard Williams and Jonathan Wolff (see esp. 2011), for example, advocate something much closer to a kind of realism in part as a result of the lessons learned whilst involved with various commissions in the UK; and both Geuss and Sen were experienced in real politics (albeit of very different kinds) well before they developed their methodological reflections. It would be hard to argue that all of this is all purely coincidental.
The argument from cognitive distortions claims that cognitive distortions of various kinds are likely to be present not only among a society's lay population and their common beliefs and preferences, but also among its political theorists. These cognitive distortions include a variety of subtly different phenomena, including those of adaptive preferences, ideology, and wishful thinking. These are distinct phenomena, but they share a common feature in that they are all ways in which the contexts within which people live somehow distort what they want and value, and what they believe to be necessary, possible, and even actual. Realist and comparative approaches, by openly and explicitly building various descriptive claims into their theories, allow for their claims and underlying presuppositions to be more open to various forms of empirical testing and contestation. Such underlying descriptive presuppositions are, I contend, woven into and present throughout political theories of any kind; the acknowledgment and openness to this fact allows realist and comparative approaches to be more amenable to eliminating cognitive distortions through empirical investigation, debate, and dispute.
The argument has the two following premises. First, explicit empirical testability makes something more amenable to eliminating cognitive distortions of various kinds. Second, realist and comparative approaches build in more, and more explicit, testability into their theories. The second premise is obviously correct. For instance, Adam Smith's and Karl Marx's thoughts about the virtues and vices of capitalism, and the prospects for working class organisation within it, are clearly (i) critical parts of their political theories, (ii) opposed to one another, and (iii) empirically testable. Since their resulting visions for the best feasible arrangement for a future society – a kind of free market capitalism for Smith, communism for Marx – are explicitly based, in part, on these aforementioned descriptive premises, their overall conclusions are consciously based on components which, no matter how distorted or one-sided their inspiration and motivation might be, at least can be subjected to empirical argument and testing.
Two distinct things are at issue here: first, whether the components of a political theory are testable; and second, whether a political theory tends to smuggle in empirical presuppositions. With regard to the first point, my argument contends that realist and comparative political theories build in more descriptive components than their ethics-first counterparts, and that this renders the former more able to reflexively address the various cognitive distortions we all are liable to (cf. Williams 1980). My argument also contends that realist and comparative political theories tend to be much more transparent and up-front about including various descriptive premises in their argumentation, whereas ethics-first theories instead often smuggle them in. This is not to say that ethics-first approaches, such as Rawls', do not explicitly contain any clearly testable descriptive premises, or that they do, but never properly acknowledge it. Rawls' justice as fairness is, as Sangiovanni (2007: 138) shows, explicitly practice-dependent in the sense that '[t]he content, scope, and justification of a conception of justice depends on the structure and form of the practices that the conception is intended to govern'. Some testable descriptive claims about social practices of this kind are clearly stated in Rawls' work. Including descriptive components is not a weakness – quite the contrary – and arguably no political theory worthy of the name could ever do without them (on this point, see Rossi's 2014 critique of Cohen 2008, and cf. also Hall 2013 and Jubb 2014). The problem is that when these components are smuggled in they are not recognised for what they are, and thus are harder to critically test and contest accordingly. Is it true that ethics-first theories do smuggle in descriptive premises? Let us answer the question via a possible reply to my argument.
One possible counter to the argument from cognitive distortions would be to claim that the transparent and up-front inclusion of descriptive premises is true of traditional ethics-first approaches as well, once we get to applying them. Obviously, a highly abstract and idealised ethics-first theory of, say, goodness or justice will initially contain fewer (or at least less explicit) directly testable premises in the sense discussed here. However, once it becomes a matter of applying said theory a full range of such premises comes into play precisely when it is appropriate for them to do so. As a result of this, any virtues claimed for realist approaches in this regard are spurious. To answer this, I will give an example of how, in one instance, a clear case of an ethics-first approach tacitly builds in a critical descriptive premise for its normative argumentation, and then how this tacit assumption, due its falsity, aids the theorist in advocating a conclusion which is unwarranted by his own standards. This will illustrate how descriptive premises feature in purportedly purely normative argumentation in ethics-first approaches, and how a blindness to this fact both aids such theorists in developing deeply flawed theories and makes it hard for them to recognise these as descriptive premises, ignoring concerns about empirical argument and testing. The example I choose is Nozick's argument for a minimal state. The scope of this article precludes a full reconstruction of Nozick's argument, but a brief summary is possible.
Nozick famously begins his argument from the assumption that all individuals have certain rights as side-constraints – i.e. that '[i]ndividuals have rights, and there are things no person or group may do to them (without violating their rights)' (Nozick 1974: ix). This conception of rights, and the over-arching importance Nozick gives them, means that anything which violates someone's rights is ipso facto unjustified. Since states inevitably do things like monopolise the use of force within some area, or impose taxation, etc., the immediate question for a right-libertarian like Nozick becomes: can a state of any sort be justified? Nozick goes on to argue that a minimal state which monopolises rights-protection within a given territory and charges its clients more than it otherwise would to extend its protection to non-paying clients can be justified (for more detail, see Nozick 1974, Mack 2011 and Wolff 1991). Furthermore, he holds that such a state could, at least in principle, arise without ever violating anybody's rights. For Nozick, the upshot of all of this is that he is able to justify the existence of a minimal state. As Murray Rothbard, one of Nozick's right-libertarian critics, points out:

[D]espite Nozick's attempt (6-9) to cover his tracks, it is highly relevant to see whether Nozick's ingenious logical construction has ever indeed occurred in historical reality: namely, whether any State, or most or all States, have in fact evolved in the Nozickian manner. It is a grave defect in itself, when discussing an institution all too well grounded in historical reality that Nozick has failed to make a single mention or reference to the history of actual States. In fact, there is no evidence whatsoever that any State was founded or developed in the Nozickian manner. On the contrary, the historical evidence cuts precisely the other way: for every State where facts are available originated by a process of violence, conquest and exploitation: in short, in a manner which Nozick himself would have to admit violated individual rights. (…)
Since Nozick's justification of existing States (…) rests on their alleged immaculate conception, and since no such State exists, then none of them can be justified, even if they should later become minimal. (Rothbard 1977: 45)

Rothbard's criticism takes Nozick to task on precisely the point of whether or not the state Nozick attempts to justify really can be justified in his own terms at all. If it is the case that any state or state-like entity has developed through rights-violations, or if it is the case that anything like a state would require rights-violations (such as taxation) in order to survive and maintain itself, or both, then, on Nozick's own conception of rights and their importance, no state is justified. There are two slightly different anti-Nozickian arguments we can draw from this. The first goes as follows: Since no states arose without violating people's rights, no present state is justified at all on Nozick's political theory, even if it should later become minimal. That is an argument about the justification of existing states and future states that might develop from them. We can also, however, draw a broader inductive argument from these foundations: no existing states have ever developed or existed without violating Nozickian rights, therefore it unlikely that any state can ever develop or exist without violating Nozickian rights.
Recall that it is not abstract, logical possibilities of arrangements and hypothetical histories that is important for Nozick's conception of rights violations, but the concrete issue of whether rights violations in fact occur and have occurred, or not. Here I think Nozick's adherence to an ethics-first approach therefore misleads him: he thinks he can establish his purely normative conclusion on the basis of hypothetical histories and thought-experiments, but what his argument really requires is some plausible basis in facts about human nature, history, and society. The fact of this requirement is hidden, perhaps even to himself, by Nozick's ethics-first approach to political theory, as a result of which they must be smuggled in instead.
I contend that cases such as these are the norm rather than the exception in a great deal of contemporary political theory (though, I must point out, far from all of it), particularly in the more ethics-first-based approaches. If this is correct, it means that the problem is not one that can, as the aforementioned objection maintains, be expected to be solved once we move from the construction of normative theory to its application. Instead, we can expect these sorts of problems to infect and trouble the very construction of normative ethics-first theories themselves – much more so than in realist and comparative approaches.
Lastly, the argument from abstractness claims that errors and distortions of various kinds – arising from over-generalisation, the aforementioned distortion of beliefs and preferences, and other sources – are likely to increase the more abstract and disconnected political theories become from the realities of human social existence. Since realist and comparative political theories tend to be less abstract, and due to the fact that they are better able to detect and address such errors and distortions, they are superior in this respect as well. There are two different senses in which realist and comparative political theories can be less abstract and more 'down-to-earth' than ethics-first ones. On the detachment reading, the former can be less abstract in the sense that they try to avoid overly idealised and un-contextual perspectives on political questions and problems that the more ethics-first approaches often contain. On the restriction reading, realist and comparative political theories can be less abstract and more 'down to earth' in the sense of restricting their scope and concern only to issues of day-to-day politics, perhaps especially state-politics. If we adopt this latter sense of the distinction, then we face a plausible counter-argument: many important political problems and issues can only be perceived properly once we manage to attain some critical distance from them. Such distance can be important, inter alia, for discovering and challenging a variety of parochial limitations and cognitive distortions we are subject to. The sense relevant to my argument, however, is the sense of the detachment reading: realist and comparative theories are less abstract and more 'down to earth' than their ethics-first competitors only in the sense that they tend to avoid overly idealised and un-contextual perspectives on political questions and problems; and they do so by focusing on the realities of political agency and its contexts and on how best to contribute thereto.
I think this argument is more obvious that the other two, so I won't labour the point too much. Realist and comparative political theories tend to be much more local and problem-based than ethics-first approaches, and their claims to develop grand theoretical constructs articulating what the good, right or just would look like, either for the whole universe, for all of humanity, or even just for a modern society (as in e.g. Rawls), however conceived. The genealogies of Geuss (2003) and Foucault (1988, 1990a, 1990b, 1995, 1998) aim to enlighten us about the historical development of particular conceptions relevant to particular aspects of contemporary thinking and politics (e.g. about the public-private distinction, about conceptions of mental illness, and so on); Adam Smith's (1976) political economy is penned as advice to a beneficent legislator principally on economic policy for a commercial (capitalist) society with its corresponding class and state structure; and John Stuart Mill's On Liberty (1991) aims to advise citizens and legislators of a modern society on where and how, in light of the present nature of such society, the scope of the criminal law and other social sanctions ought to be restricted. It is only plausible to suppose that this less abstract and more down-to-earth character should render these works less susceptible to various distortions or parochial delusions of much ethics-first political theory, since it means that realist and comparative political theories constantly have to come back to, engage with, and relate themselves to an independent political reality. By contrast, the theorisations of ethics-first approaches, lacking these corrective driving forces as a result of their overly abstract character, are likely to be even more susceptible to the aforementioned cognitive distortions to begin with.

Conclusion
In sum, I have discussed two contemporary approaches to political theory: the realism of Raymond Geuss and the comparativism of Amartya Sen. I have argued that the latter approaches form a proper subset of the former, and that they both share an important strand of anti-Platonism in political thought. Lastly, I have argued that realism in this sense is superior to the ethics-first approaches currently more prevalent in (especially the Anglo-American) academy. Let me end on a brief note of optimism. Increasingly, political theorists, even of the kind I have implicitly been criticising, are turning to engage more closely with issues of political practice, concrete questions about the actual reform or replacement of institutions, legislation, and policy. This is very encouraging, but it still has some way to go. Even most realists, after all, spend very little time engaging in realist and comparative political theory of the of the kind that Geuss and Sen, among others, admire. Instead, they focus on often irrelevant methodological disputes and discussions among one another, rather than on contributing realist political theories of their own. A more explicitly realist and comparative orientation, along with a closer study of the methods and approaches of earlier thinkers in this tradition and engagement with contemporary political practices, would likely be very beneficial to helping this trend to grow, develop, and flourish.



References:
Baderin, Alec. 2014. Two forms of realism in political theory. European Journal of Political Theory 13(2), pp. 132-53.
Bavister-Gould, Alex. 2011. Bernard Williams: Political Realism and the Limits of Legitimacy. European Journal of Philosophy 21(4), pp. 593-610.
Cohen, Gerald A. 2008. Rescuing Justice and Equality. Cambridge, MA: Harvard University Press.
Cohen, Gerald A. 2009. Why not Socialism? Princeton: Princeton University Press.
Cooper, John M. (ed.) 1997. Plato: Complete Works. Indianapolis: Hackett.
Finlayson, Lorna. 2014. With Radicals Like These…: doom, gloom, and political 'realism' Unpublished manuscript, University of Cambridge, URL:
https://www.academia.edu/5916118/_draft_With_radicals_like_these_doom_gloom_and_political_realism.
Floyd, Jonathan, and Stears, Marc (eds.). 2011. Political Philosophy versus History? Cambridge: Cambridge University Press.
Forrester, Katrina. 2012. Judith Shklar, Bernard Williams, and political realism. European Journal of Political Theory 11(3), pp. 247-72.
Foucault, Michel. 1988a. The History of Sexuality, Volume 3: The Use of Pleasure. New York: Vintage.
Foucault, Michel. 1990a. The History of Sexuality, Volume 1: An Introduction. New York: Vintage.
Foucault, Michel. 1990b. The History of Sexuality, Volume 2: The Care of the Self. New York: Vintage.
Foucault, Michel. 1995. Discipline and Punish: The Birth of the Prison 2nd ed. New York: Vintage.
Foucault, Michel. 1998. Madness and Civilization: The History of Insanity in the Age of Reason. New York: Vintage.
Frazer, Elizabeth. 2010. What's real in political philosophy? Contemporary Political Theory 9(4), pp. 490-507.
Galston, William A. 2010. Realism in political theory. European Journal of Political Theory 9(4), pp. 385-411.
Geuss, Raymond. 2001. History and Illusion in Politics. Cambridge: Cambridge University Press.
Geuss, Raymond. 2003. Public goods, private goods. Princeton, NJ: Princeton University Press.
Geuss, Raymond. 2005. Outside Ethics. Princeton, NJ: Princeton University Press.
Geuss, Raymond. 2008. Philosophy and Real Politics. Princeton, NJ: Princeton University Press.
Geuss, Raymond. 2010. Politics and the Imagination. Oxford: Princeton University Press.
Geuss, Raymond. 2014. A World Without Why. Princeton, NJ: Princeton University Press.
Gilabert, Pablo. Forthcoming. Comparative Assessments of Justice, Political Feasibility, and Ideal Theory. Ethical Theory and Moral Practice.
Gledhill, James. 2012. Rawls and Realism. Social Theory and Practice 38(1), pp. 55-82.
Hall, Edward. 2013. Bernard Williams and the Basic Legitimation Demand: A Defence. Political Studies DOI: 10.1111/1467-9248.12070
Honig, Bonnie. 1993. Political Theory and the Displacement of Politics London: Cornell University Press.
Honig, Bonnie, and Stears, Marc. 2011. The new realism: from modus Vivendi to justice. In Floyd and Stears (eds.) 2011, pp. 177-205.
Jubb, Robert. 2014. Playing Kant at the Court of King Arthur. Political Studies doi: 10.1111/1467-9248.12132.
Kelly, Paul. 2011. Rescuing political theory from the tyranny of history. In Floyd and Stears (eds.) 2011, pp. 13-37.
Lane, Melissa. 2011. Constraint, freedom, and exemplar: history and theory without teleology. In Floyd and Stears (eds.) 2011, pp. 128-50.
Larmore, Charles. 2013. What is Political Philosophy. Journal of Moral Philosophy 10(3), pp. 276-306
Mack, Eric. 2011. 'Nozickian arguments for the more-than-minimal state' in Bader, Ralf M. and Meadowcroft, John (eds.) 2011. The Cambridge companion to Nozick's Anarchy, state, and utopia. Cambridge: Cambridge University Press, pp. 89-115.
Menke, Christoph. 2010. Neither Rawls Nor Adorno: Raymond Geuss' Programme for a 'Realist' Political Philosophy. European Journal of Philosophy 18(1), pp. 139-47.
Mészáros, I. 2011. Social Structure and Forms of Consciousness, Volume II: The Dialectic of Structure and History. New York: Monthly Review Press.
Mill, John S. 2010. 'On Liberty' and Other Writings. Cambridge: Cambridge University Press.
Mouffe, Chantal. 2006. The Return of the Political. London: Verso.
Newey, Glen. 2010. Two Dogmas of Liberalism. European Journal of Political Theory 9(4), pp. 449-65.
Nozick, Robert. 1977. Anarchy, State and Utopia. New York: Basic Books.
Philp, Mark. 2012. Realism Without Illusions. Political Theory 40(5), pp. 629-49.
Rawls, John. 1999. A Theory of Justice (Revised Edition). Harvard: Belknap.
Rawls, John. 2001. Justice as Fairness: A Restatement. Harvard: Belknap.
Rawls, John. 2005. Political Liberalism (Expanded Edition). New York: Columbia University Press.
Rossi, Enzo. 2010. Review: Reality and imagination in political theory and practice: On Raymond Geuss's Realism. European Journal of Political Theory 9(4), pp. 504-12.
Rossi, Enzo. 2012. Justice, Legitimacy and (Normative) Authority for Political Realists. Critical Review of International Social and Political Philosophy 15(2), pp. 149-64
Rossi, Enzo. 2014. Facts, Principles, and Politics. SSRN Working Paper.
Rossi, Enzo, and Sleat, Mat. Forthcoming 2014. Realism in Normative Political Theory. Philosophy Compass.
Rothbard, Murray. 1977. Robert Nozick and the Immaculate Conception of the State. Journal of Libertarian Studies 1(1), pp. 45-57.
Sangiovanni, Andrea. 2007. Justice and the Priority of Politics to Morality. The Journal of Political Philosophy 16(2), pp. 137-64.
Sen, Amartya. 2010. The Idea of Justice. London: Penguin.
Sen, Amartya. 2012a. Values and justice. Journal of Economic Methodology 19(2): 101-8.
Sen, Amartya. 2012b. A reply to Robeyns, Peter and Davis. Journal of Economic Methodology 19(2): 173-6.
Sleat, Matt. 2010. Bernard Williams and the possibility of a realist political theory. European Journal of Political Theory 9(4), pp. 485-503.
Smith, Adam. 1976. An Inquiry into the Nature and Causes of the Wealth of Nations. Oxford University Press.
Thomas, Alan. 2014. Sen on Rawls's "transcendental institutionalism": An analysis and critique. European Journal of Political Theory 13(3), pp. 241-63.
Valentini, Laura. 2011. A Paradigm Shift in Theorizing about Justice? A Critique of Sen. Economics and Philosophy 27(3), pp. 297-315.
Valentini, Laura. 2012. Ideal vs. Non-Ideal Theory: A Conceptual Map. Philosophy Compass 7(9), pp. 654-64.
Vrousalis, Nicholas. 2010. G. A. Cohen's Vision of Socialism. Journal of Ethics 14 (3), pp. 185-216.
Williams, Bernard. 2008. In the Beginning Was the Deed: Realism and Moralism in Political Argument. Princeton: Princeton University Press.
Williams, Bernard. 1980. Political Philosophy and the Analytic Tradition. Reprinted in Williams, Bernard. 2006. Political Philosophy as a Humanistic Discipline. Oxford: Princeton University Press, pp. 180-99.
Williams, Bernard. 2002. Truth and Truthfulness. Princeton, NJ: Princeton University Press.
Wolff, Jonathan. 1991. Robert Nozick: property, justice and the minimal state. Cambridge: Polity Press.
Wolff, Jonathan. 2011. Ethics and Public Policy: A Philosophical Inquiry. London: Routledge.
Wright, Erik O. 2010 Envisioning Real Utopias. London: Verso.