The explicit formulation of practice-dependence as a methodology in political philosophy, at least under that name, is relatively recent, dating back to Andrea Sangiovanni’s Justice and the Priority of Politics to Morality in 2008. As he formulates it, its central commitment is that “[t]he content, scope, and justification of a conception of [a given value] depends on the structure and form of the practices that the conception is intended to govern.”¹ It is contrasted with practice-independence, which rejects “the idea that the contingent, practice-mediated relations in which we find ourselves should change or affect the justifying reasons and premises sustaining a particular conception of the content and scope of [a given value].”² For example, luck egalitarianism, which requires the elimination of all and only inequalities which cannot be traced back to individual choice,³ is for Sangiovanni a practice-independent view since it sees that requirement as universally applicable. On the other hand, an account of justice which relies on “the notion of a ‘fair system of social cooperation’” is practice-dependent since it ties its principles to a particular practice.⁴

As his reference to a fair system of cooperation and so justice as fairness suggests, Sangiovanni is one of a number of self-identifying practice-dependent theorists who are clearly inspired by Rawls.⁵ Elsewhere, Sangiovanni has quoted Rawls making the apparently strongly practice-dependent claim that “[t]he correct regulative principle for a thing depends on the nature of that thing.”⁶ This raises the question of the distinctiveness and interest of practice-dependence as a methodology. Is it anything more than a piece of Rawls interpretation? So far, its explicit invocation does not seem to have spread much beyond debates on global justice, where it is typically used to deny that the principles of egalitarian justice appropriate for the domestic sphere apply to relations across state borders. Practice-dependence is certainly well suited to this task. If the content of principles of justice depends on the practices they govern, then the fact that domestic political authority is different from any practice in the international arena makes it likely that the principles of justice governing the two areas will also be different. However, beyond these debates, the only explicit use of practice-dependence I am aware of is to criticize G. A. Cohen’s advocacy of a particular moral vision for a society by Miriam Ronzoni in her Life is not a Camping Trip.⁷

In this paper, I consider the novelty of practice-dependence and its relation to various other methodological claims about moral and political philosophy and theory. In doing so, I hope to raise and illuminate various questions about method in moral and political philosophy and theory. In particular, I will argue that while the methodological stance adopted by explicitly practice-dependent theorists is appropriate, and superior to those relied on by their critics, that does not immunize them against criticism on at least partly methodological grounds. Contemporary practice-dependent theorists see themselves as operating mainly in a tradition inaugurated by Rawls, and when looking for antecedents or fellow travelers, tend to look either to the master himself or to those who have followed him.⁸ This is understandable, given that the debates they have been most active in are in a certain sense structured around arguments over which of the available positions has the best claim to be the guardian of the commitment to liberal equality Rawls made the centerpiece of his domestic theory.⁹ Yet by limiting their frame of reference in this way, practice-dependent theorists have closed themselves off from other work in moral and political philosophy and theory, and so avoided confronting some pressing interpretive issues. On the one hand, I defend practice-dependence as a method, drawing on other pieces of moral and political philosophy which seem to me to share a methodological stance with explicitly practice-dependent theories to help explain the appropriateness of that stance. On the other, I suggest that when adopting that stance, more needs to be said about the practices it makes its normative recommendations depend on. These can be interpreted in a variety of different ways, many of which practice-dependent theorists have ignored, despite the implications those different interpretations would have for the sort of principles which would be appropriate. Practice-dependence’s opponents in the global justice literature are, I argue, methodologically mistaken, but that does not mean we should uncritically adopt practice-dependence in the forms we have thus far seen, under that name at least. Rather, we need to ask questions about what the practices we are theorizing for are actually like, and what kind of principles can appropriately govern them.

This paper will proceed as follows. First, I will define practice-dependence, using examples of the method being deployed to make clear the otherwise potentially obscure claim that ‘the content, scope, and justification of a conception of a given value depends on the structure and form of the practices that the conception is intended to govern.’ In the course of defining it, I explain part of its rationale and forestall a number of potential objections. Once I have defined practice-dependence and defended it against various objections prominent in the literature, I move on to consider a range of possible but largely so far ignored antecedents.¹⁰ Here, I discuss P. F. Strawson’s Freedom and Resentment and Bernard Williams’ work in both moral and political philosophy, as well as Christopher Kutz’s Complicity more briefly.¹¹ I draw on Strawson and Williams to indicate why I think practice-dependence is an appropriate methodology, and then use Williams to try to undermine self-identified practice-dependence theorists’ confidence in their own interpretive skills. When the content, scope, and justification of a conception of a given value depends on the structure and form of the practices that the conception is intended to govern, we need accounts of the structure and form of practices, and how that shapes the conceptions of value that govern them. Williams and others like him seemingly have a different account of the structure and form of practices around us than self-identified practice-dependent theorists, and accordingly end up with different accounts of how they ought to be governed despite their acceptance of the basic methodological tenets of practice-dependence. This raises the question of whether Williams and other realists, or the Rawls-inspired, self-identified practice-dependence theorists, are right. If explicitly practice-dependent theorists are to defend their theories fully, they need to find ways to deal with the challenge raised by theorists, like Williams, who are more skeptical about the role of morality in political life.

What is Practice-Dependence?

In order to see what we can learn about practice-dependence by looking at other pieces of theorizing with similar methodological commitments, we first need to get clear about what practice-dependence itself is. Here, I think it is useful to proceed by example. In particular, I want to look at Sangiovanni’s argument for his own ‘reciprocity-based internationalism,’ the examples of practice-dependence he gives in Justice and the Priority of Politics to Morality, and Miriam Ronzoni’s claims about G. A. Cohen’s socialism in her Life is not a Camping Trip. The reason I prefer to proceed by example is to avoid confusion about what exactly is meant by ‘depend’ when practice-dependent theorists claim that conceptions of a given value depend on the structure and form of the practice they are to regulate. ‘Depend’ can point to a variety of different relations. These range from a necessary and sufficient condition, as when a light switch working depends on it forming a circuit including a functioning bulb it connects to an electricity source, to at least the various combinations of necessity and sufficiency analyzed by J. L. Mackie in his Causes and Conditions, in which a fire may depend on rubbish not being cleared even though that is only “an insufficient but necessary part of a condition which is itself unnecessary but sufficient for the result.”¹² Given that the content of depend varies in this way, it seems best to understand what practice-dependent theorists mean when they say that a conception of a given value depends on the form and structure of what it is to govern by looking at how they articulate and defend conceptions of given values.

Consider first Sangiovanni’s argument for reciprocity-based internationalism. There, he limits the scope of egalitarian justice to the domestic sphere by pointing to the way that those sharing a system of law “provide the basic collective goods necessary to protect [each other] from physical attack and to maintain and reproduce a stable system of property rights and entitlements” and so have particularly demanding duties of reciprocity towards one another.¹³ Sangiovanni’s point here is that the particular character of the relation between citizens decides and gives content to the principles appropriate to govern that relation. He first puts pressure on alternative accounts of the limits of principles of egalitarian justice by showing that the features they point to are not necessary for or unique to domestic political authority, and that their presence does not track the appropriateness of egalitarian distributive norms. He then argues that what constitutes a fair return depends on the character of goods being provided, and that since fellow citizens “submitted themselves to a system of laws and social rules in ways necessary to sustain our life as citizens, producers, and biological beings,” making it possible for us to have anything at all, what they are owed is an equal share.¹⁴ The conception of justice Sangiovanni defends depends on the practices it is to regulate in the sense that the structure and form of the practice help select which abstract principles are relevant and then fill out what they actually require of us. As Sangiovanni rightly points out, the demands of reciprocity among the members of a football club are quite different from those among the members of a family.¹⁵ We need to pay attention to the relationships that the practice a principle is to govern creates in order to see whether and how it should govern.

As Sangiovanni explains elsewhere, what differentiates practice-dependence from practice-independence is that practice-dependence does not see principles to govern particular practices as simple implementations of a higher order principle in the way that practice-independence does. He uses the example of the principle that “all human beings are of equal, ultimate, and general concern.”¹⁶ Practice-independent theorists sometimes interpret this as requiring principles of egalitarian justice with global scope, since for them the principle of basic equality removes any justification for inequalities which can be traced to unchosen features of people’s circumstances, like where and to whom they were born.¹⁷ In contrast to this practice-independent cosmopolitanism, practice-dependent theorists allow the context in which they are to be applied to shape “the reasons we might have for endorsing specific principles of justice.”¹⁸ For example, we might, with Thomas Nagel, point to the way the state claims to speak in our name as giving “us the standing to demand a special justification” we lack with respect to the global order and so distinguish the principles that ought to govern the two.¹⁹ This is even though we continue to affirm the principle of basic equality. The concern we owe is simply shaped by the relations in which we stand, so that different relations ground different forms, though not levels, of concern.

The practice-independent theorist, whose “argument does not appeal, at any point, to existing institutions or practices,” instead sees context as shaping “courses of action we should adopt in implementing” principles of justice.²⁰ They deny that relations like that between state and citizen shape the principles that apply to them, rather than the mechanisms available to realize those principles. Perhaps states with egalitarian domestic political arrangements are the best way to achieve the kind of globally egalitarian distribution the principle of basic equality mandates, given the limits imposed by human sympathy and the need for political accountability. It seems unlikely, but this would at least be a possible, if probably not successful, practice-independent argument. The difference is not between those who argue that the scope of egalitarian justice is limited to the domestic sphere and those which do not. Just as many practice-independent theorists have argued against extending relational principles to the global sphere,²¹ practice-dependent theorists might insist that the international context is relevantly similar to the domestic context, or at least that it is suitable for some form of egalitarian justice.²² What matters is whether features of the context a given principle or set of principles are to govern play a role in justifying those principles beyond simply constraining the extent to which they can be achieved.

This can be further illustrated if we look at Miriam Ronzoni’s explicitly practice-dependent critique of G. A. Cohen’s advocacy of socialism. Cohen argued that socialism’s appeal consisted of its realization of two principles, the first being a luck egalitarian principle he called socialist equality of opportunity whose capacity for generating inequality was to be constrained by the second, a principle of communal reciprocity.²³ This principle of community required that “I serve you not because of what I can get in return by doing so but because you need or want my service, and you, for the same reason, serve me.”²⁴ Cohen argued for both principles through an extended example he thought was both morally appealing and met the demands of the two principles, that of a (particularly egalitarian) camping trip. Ronzoni objects to Cohen’s principle on the grounds that though perhaps Cohen’s principles are attractive when thinking about how to organize camping trips, they are not appealing ways of organizing the relations structured in other ways.²⁵ Camping trips are limited in time, voluntary, and aim at a shared enjoyment of “communal life under conditions of frugality and proximity to nature.”²⁶ Families, though also marked by strong bonds of affection, are not in the same way limited in time, voluntary, or arranged around a rich communal conception of the good, and so the very strong norm of reciprocity Cohen advocates is not appropriate for them. Because they endure for whole lives rather than days, have high exit costs, and more extensive and deeper conflicts of interest, they need to give people the space to resist the demands of others on their time and resources in a way that might be destructive of the good of camping trips. In a family, we need a room of our own.²⁷ What is appropriate for the camping trip depends on “its point and purpose and specific features of its constraints and participants,” just as what is appropriate for a family does, and so close attention to the structure of a practice is needed when formulating and defending principles intended to govern it.²⁸

Practice-dependence is in this sense a justificatory and epistemic strategy, a way of finding principles to govern a practice and drawing attention to the features that make them more appealing than their competitors as rules for that practice. It does not deny that there are more abstract principles whose details in a particular context are filled out by that strategy, though it may deny that those principles are of much interest themselves.²⁹ Just as Sangiovanni points to a general principle of reciprocity whose demands are shaped by the specific features of the practice of domestic political authority it governs, Ronzoni points to the independent value of “the creative, inventive, multifaceted, and diverse nature of human undertakings” when criticizing Cohen’s socialism.³⁰ Cohen’s requirement in his Facts and Principles that chains of argument must end in principles which are not argumentatively dependent on facts can easily be met by practice-dependent theorists then, as it can by any competent theorist.³¹ Fact-dependent principles can be made fact-independent by making them conditional, or general principles of the sort Ronzoni and Sangiovanni refer to can be used to explain the significance of facts. Indeed, rather than insisting on Cohen’s empty formal requirement, it seems more useful to focus on the way that practice-dependence makes use of facts. Paying attention to the details of the camping trip Cohen uses to argue for his two socialist principles and seeing how that may shape the principles appropriate to govern it shows how limited their scope may be, for example.³²

Interpreting principles in light of the practices to which they are to apply might raise the worry that practice-dependence lacks the critical distance necessary to subject the world as it is to proper moral scrutiny. Rather than demanding the abolition of practices plainly contrary to justice, the practice-dependent theorist suffers from a kind of adaptive preference formation, reversing the proper order and fitting principles to the ugly facts of the world as it is instead of demanding that those facts are changed to fit the principles. Practice-dependence becomes a kind of acquiescence in injustice, or at least a failure to explain what is just about the practices it cannot help but endorse.³³Beginning with an interpretation of a practice gives that practice an unwarranted priority over the demands of justice. Yet we might wonder how much and what sort of distance critique requires, and what kind is available.

Practice-dependence certainly can critique practices, since they can always point to basic moral failures that need no real interpretation to condemn. For example, Sangiovanni claims that institutions which “depend on systematic and unmediated coercion to reproduce and sustain” themselves cannot be just, because justice requires that there “be some sense in which the terms of the institution are at least capable of being justified to all participants,” for example.³⁴ Nor are practices necessarily coherent, either internally or with other practices. There was always a tension between the official egalitarianism of the United States and the racial hierarchies it inscribed into its legal system, for instance. Practice-dependent theorists can use these tensions to argue that practices demand their own or other practices’ abolition. Ronzoni draws on Rawls’ arguments about the importance of the basic structure to suggest that the current international order’s lack of regulation places excessive burdens on domestic political authorities, for example, and so should be reformed.³⁵

It begs the question to assume that a practice’s vindication requires more distance than this. The practice-dependent theorist is not claiming the facts about the structure of a practice vindicate themselves, but that the set of rules which are to govern it will draw on its structure as part of their justification. They try to show, for example, that the rationale for a particular conception of justice depends on features of the practice it should govern, and that those features are absent elsewhere. Justifying the practice itself may require showing that the practice is permissible in terms of the larger practice it is part of, so that the justifiability of state sovereignty and the bounds of distributive justice it may impose depend on how practices of international and transnational cooperation ought to be governed. Providing a theory of how to govern international and transnational cooperation is well within the capability of practice-dependent theorists though.

Meeting the demand for critical distance seems to require ignoring any details about the practices the principles being theorized are to govern. This is what avoiding granting priority to the unfortunate facts of the matter would presumably involve, since it is the shaping of the content of principles by the structure of their objects that appears objectionable, an unwarranted compromise with the world. Assuming nothing about the object of principles may make the principles that are arrived at inappropriate though, and poorly defended principles can hardly be a sound basis for critique either. After all, Cohen’s example of the camping trip fails to do the justificatory work he intended it to because it is too far from the facts about our collective life, not because it is too close to them. The principles appropriate for camping trips depend in part on what camping trips are like, just as the principles appropriate for collective life more generally depend on what that is like. Strong norms of communal reciprocity are appropriate for camping trips because the voluntariness and shortness of camping trips allows many conflicts of interest and moral disagreements to be smoothed over or set aside. Conflicts of interest and moral disagreements, however, are endemic to the experience of family life, let alone shared citizenship, and so norms of reciprocity need to be more circumscribed. In failing to pay sufficient attention to what the contexts he was theorizing were actually like, Cohen deprived his position of any critical potential by failing to adequately defend it. It is not clear then that ignoring the world gives theories purchase on what is wrong with it.³⁶ Given that some practices will fail to meet the demands of the most minimal interpretation of whatever value it is proposed to govern them with, it is not clear that practice-dependence’s integration of features of the world into its justificatory structure means it does worse at articulating complaints about how things are than the available alternatives.

Fellow Travelers and Forgotten Forebears

The distinctive feature of practice-dependence is then that it apparently rightly pays close attention to the details, to the structure, of whichever practices its theorists are trying to formulate principles for. There is, as advocates of practice-dependence emphasize, good reason for thinking that Rawls was in this sense a practice-dependent theorist. This is true not only of his later work,³⁷ but also in A Theory of Justice, where he criticizes utilitarianism for “extending to society the principle of choice for one man,” for treating “the decision made by the ideal legislator” as not “materially different from that of an entrepreneur deciding how to maximize his profit.”³⁸ Utilitarianism’s problem is that it has not understood the content of the practice it is offering principles to regulate. Rawls and those who consciously adopt distinctively Rawlsian positions in contemporary debates can be safely assumed, at least in the absence of contradictory evidence, to have practice-dependent theories. Indeed, this makes it clear why so many of the objections standardly made against explicitly practice-dependent theorizing were earlier made against Rawls. The echoes of Cohen’s complaints about Rawls’ compromises with human imperfection and Pogge’s demands that his principles govern global, and not just domestic, interactions in contemporary criticisms of practice-dependence should hardly be surprising.³⁹

This certainly makes sense of the way in which practice-dependent theorists have concentrated on Rawls and those who can in some sense be called his followers. That narrowness of focus, though, may have under-emphasized both practice-dependence’s strengths and its weaknesses. On the one hand, there may be resources in the canon of moral and political theory and philosophy that could help explicitly practice-dependent theorists to explain why their methodology is appropriate. On the other, by concentrating on debates which take it as read that political life has certain kinds of features and so presents certain kinds of problems with certain kinds of solutions, they may have not confronted the challenge posed by different accounts of politics in general and our politics in particular. In this section, I try and show that self-identified practice-dependent theorists and their Rawlsian forebears were not the only or even the first to adopt the view that which principles are appropriate to regulate a practice depends on the details of the practice. Showing that there are other, if not explicitly, practice-dependent theorists then allows me, in the next section, to show what we can learn about practice-dependence as it is from those theorists.

We can begin with the source of the quote which forms the first part of the title of this paper, P. F. Strawson’s Freedom and Resentment. Freedom and Resentment argues against what Strawson calls pessimism and optimism about moral responsibility by trying to show that both misinterpret the practice they draw their motivation from. They “over-intellectualize the facts,” making demands of the practices associated with our reactive attitudes of holding each other and ourselves responsible that those practices cannot meet without destroying themselves and, with them, the very basis of the demands both the pessimist and optimist make.⁴⁰ The optimist is a victim of “a characteristically incomplete empiricism, a one-eyed utilitarianism.”⁴¹ By adopting an objective stance towards reactive attitudes like indignation and treating them as if they were mechanisms for bringing about good effects rather than responses warranted by particular forms of agency, they deny that the ideas of agency those attitudes depend on and constitute make sense or have purchase in the world. The pessimist, rightly feeling that “that to speak in terms of social utility alone is to leave out something vital in our conception of these practices,” searches for “some general metaphysical proposition … verified in all cases where it is appropriate to attribute moral responsibility.”⁴² Just like the optimist, they search for an “external, ‘rational’” justification for our practices of moral responsibility, not seeing that such a thing is impossible because of the mutual dependence of our sense of how it is appropriate to treat others and the very attitudes whose existence is being questioned by that sense.⁴³ For Strawson, it is only possible to recover “a sense of what we mean … when … we speak of desert, responsibility, guilt, condemnation, and justice” by “attending to that complicated web of attitudes and feelings which form an essential part of moral life as we know it.”⁴⁴ Doing more is not necessary. We do recover that sense “from the facts as we know them” and “do not have to go beyond them.”⁴⁵ Again, attention to the practice is central to generating the correct principles for regulating it.

Other writers on responsibility have come to similar methodological conclusions. For example, Christopher Kutz’s work on complicity requires us to “discard a picture of accountability that sees judgments of accountability as external outputs, or verdicts, of a process of moral evaluation, which are then used as premises in a separate process of justifying liability.”⁴⁶ Instead we should see such judgments as “elements of a unified, dynamic system of social life, themselves constitutive of the goods and relations they protect.”⁴⁷ Unless we switch to such a view, the “evaluative solipsism” of some of our “commonsense principles of accountability” will prevent us from seeing anything more than a gap in our understanding of accountability where, instead, a way of dealing with involvement in others’ wrongdoing should be.⁴⁸ Like Strawson, Kutz begins his attempt to formulate principles to regulate a given practice by interpreting the practice itself, believing that unless we have a proper grasp of the relations the practice creates and is constituted by, theories will succeed, if at all, more by luck than by judgment. In this, like Strawson, he is in agreement with the central thesis of practice-dependence, that ‘the content, scope, and justification of a conception of a given value depends on the structure and form of the practices that the conception is intended to govern.’

Nor is this a special feature of theorizing about responsibility. Consider Bernard Williams’ From Freedom to Liberty: The Construction of a Political Value.⁴⁹ There, Williams claims, among other things, that a conception of liberty in and from a political authority has to satisfy constraints which are both different from those which apply to conceptions of liberty wanted for other purposes and historically variable. For example, Williams denies that anarchism is a “political position” because as liberty is “a political value, no complaint is a complaint in liberty if it would apply to any political system or any state whatsoever, so the existence of the state is not itself an offense against or limitation on liberty.”⁵⁰ Anarchism, whatever else is true about it, has a conception of freedom which is not appropriately sensitive to the distinctive character of political arrangements, centered around the idea of opposition, and so is not political.⁵¹ Equally, in order to make sense of the idea of a claim in liberty here and for us, a conception of liberty needs to rely on an idea of modernity to identify which forms of intentional action can count as constraining our liberty. Consequently, “competition in something like a zero-sum game” is not a limit on liberty now, “because competition is integral to the social system” as a whole and resenting it implies an alternative world so distant from ours as to be irrelevant.⁵² Yet we were once not modern, and when we were not, competition may well have been a limit on liberty.

Williams argues that in order to provide ourselves with a serviceable idea of liberty, we have to pay attention to the details of the practice which the idea of liberty is to be applied to. Anarchism fails to see what politics does, that it organizes who has which freedom in a way that increases freedom, and so necessarily misleads us about what liberty is and how we might get more of it. Equally, the specific social and political structures which this idea of liberty will be applied to need to be taken into account. If we want to regulate a “modern commercial society,” we had better not treat one of its fundamental features, that of competition for scarce positions, as a loss of liberty.⁵³ If we do treat competition as resulting in a loss of liberty, we will end up calling for the abolition of the world we live in, which might be in some sense wise or even inevitable, but is hardly a useful way of thinking about how to try and structure it now we have it. Williams, like Kutz and Strawson, is urging us to make ‘the content, scope, and justification of a conception of liberty depend on the structure and form of the practices that the conception is intended to govern.’

This is not an unusual feature of Williams’ work on liberty or other political concepts. Williams consistently insisted that philosophical discussion take its orientation from the practices it aims at regulating. His work on ethics and its foundations, for example, emphasizes that asking ‘how one should live’ is a first-personal question to which answers must make sense to the actors who ask them, and that these answers will have to be different for people in different circumstances. Just as he insisted that a Kantian rationality is too thin, too abstract and denatured, to commit us to impartiality in practical reasoning,⁵⁴ he also insisted that we cannot sustain the integrated accounts of human nature and ethics which made it possible for Aristotle to see living the ethical life as “a harmonious culmination of human potentialities, recoverable from an absolute understanding of nature.”⁵⁵ Neither of these answers to the question of how one should live can work for us, given that they are addressed to agents with commitments and a past and a future, and whose commitments include a rejection, in the forms of life they lead, of an Aristotelian biology. Again, attention to the practice, to the facts of how it is, that is being theorized – here, practical reason under the conditions of modernity – is for Williams central to getting the right answers to how that practice ought to operate.

Conceptual Cousins and Family Disagreements

Practice-dependence is then more widespread than its advocates’ references suggest. Kutz, Strawson, and Williams are not part of any Rawlsian school of political philosophy that might be drawn up and then claimed as practice-dependence’s natural home. Kutz and Strawson are concerned with principles of individual responsibility, not of social or political justice, and while Freedom and Resentment was anyway written before the publication of A Theory of Justice, neither that text nor Political Liberalism, Justice as Fairness: A Restatement or The Law of Peoples are referred to once in Kutz’s book-length exposition of his views.⁵⁶ Williams did deal with some of the topics associated with Rawlsian political philosophy, but, despite his apparent endorsement of practice-dependence, was scathing about what he identified as its methodological presuppositions. His Realism and Moralism in Political Theory uses Rawls as the exemplar of one of the two models of political theory he rejects, for example, while the only discussion of Rawls in From Freedom to Liberty similarly disparages Rawls as overly moralizing.⁵⁷ Nor does it seem likely that Kutz, Strawson, and Williams are the only non-Rawlsian practice-dependent theorists. The self-described ‘realist’ school in contemporary political philosophy of which Williams has posthumously become one of the founding fathers, for example, holds that politics must be taken “seriously as a particular field of human endeavour”, and that this constrains the kinds of recommendations that can be given about it.⁵⁸ This seems strikingly practice-dependent. What then, if anything, does discovering these particular three apparently practice-dependent arguments outside of what its advocates see as its natural home tell us about the method and the explicit uses it has been put to?

First, in two senses, it adds force to the rejection of complaints against practice-dependence. The first sense in which it strengthens practice-dependence against its critics is by adding to the number of pieces of apparently successful theorizing seemingly guided by practice-dependence’s prescriptions. Rejecting practice-dependence now means rejecting not just Rawls and the Rawlsians, but also Kutz, Strawson, and Williams. At least Williams and Strawson occupy significant enough places in contemporary philosophy that this ought to give critics some pause. Having to condemn, for example, Williams’ ‘integrity’ argument against utilitarianism on the same grounds and in the same terms as practice-dependent arguments for restricting the scope of principles of distributive justice might not be a very attractive prospect.⁵⁹ By showing that apparently practice-dependent arguments can be successful, demonstrating that they have been deployed outside of debates about global justice gives us reason to accept them both there and elsewhere. One obvious way of assessing a method, after all, is to check whether the results it provides are appropriate.

It helps to defend a method, though, if we can also explain why it is successful, not merely point to examples of its success. The other examples of practice-dependence are also useful in this regard. Part of Strawson’s point about pessimists and optimists about responsibility is that they rely on misinterpretations of judgments drawn from an internally complex, normatively loaded practice, that of attributions of responsibility. The optimist sees that we sometimes do not hold people responsible because there seems to be no point to it, and then infers that the justification of the whole practice of attributions of responsibility must be instrumental. The problem with this is that the practice as a whole is constitutive of the goods it protects and they have fixed on a particular part whose rationale depends on the relations it has to the other parts of the practice. The pessimist instead sees that the practice is not instrumental and so tries to find an external warrant for it. A single external warrant though is very unlikely to match the variety of related justifications the many different parts of the practice require though, and will make it impossible for the practice to create the goods it currently does. Strawson’s account is practice-dependent because if the practice of responsibility attribution works as a whole to give us part of our sense of ourselves as agents, then we need to pay attention to the practice itself, not least because our judgments about that practice themselves rely on the normative resources it provides.

Assuming Strawson is right about how practices of responsibility attribution work, this gives us a reason to adopt practice-dependent theories of responsibility attribution. Responsibility attribution is an internally complex, normatively loaded practice which constitutes the good it protects. The successful operation of one part of the practice may depend on another part of the practice which operates in a quite different way, and so any one part’s normativity may easily be misunderstood if considered in isolation. Recall that earlier I suggested that practice-dependence can be useful in identifying the features of a case to which a judgment about it is responding. Observing that camping trips have a certain form, that they are limited in time, voluntary, and presuppose commitment to rich, specific values for their duration, helps us understand why certain principles governing them is appealing and why those principles may not have purchase outside that context. Internally complex, normatively loaded practices like responsibility attribution not only require attention to the appropriate characterization of the situation that a judgment is drawn from, but also to the other situations the practice in question might generate. It is not just that when we adopt Strawson’s objective attitude to someone that we are responding to the pointlessness of holding them responsible here and now, but that we are also marking the way that in a different situation, it might be appropriate to hold them responsible. The point of something may not be just what it is, but also what it is not, and we may need to understand both in order to draw appropriate conclusions about it.

Internally complex, normatively loaded practices like responsibility attribution then give us a further reason for using practice-dependence when generating theories about how they should operate. Whenever we rely on a judgment about a case as evidence for or against some other judgment or principle, we may need to think about the features of that case and the practice that it is embedded in, so as to see how those features may constrain that judgment, and how that judgment responds to them. When using a judgment about cases from internally complex, normatively loaded practices, it is not only the features of that case which the judgment responds to, but also the role that the case plays in the practice and its relations to other parts of the practice. These also constrain the judgment, making it context-relative. Practice-independence’s failure to consider the contexts of the judgments it must also use to support its more abstract claims is then doubly damned when those judgments concern internally complex, normatively loaded practices. It is not just that they risk misapplying those judgments, but that misapplication will undermine the very basis of the judgment itself.

How significant this is depends on how many of our practices are internally complex and normatively loaded. Given that humans are cultural animals, who in groups of varying size create various interlocking systems of meaning that shape their interpretations of the world, it seems likely that many, if not most, of them are. To see this, consider again Cohen’s camping trip example. It seems likely that our reasons for going on camping trips, for regarding camping trips as worthwhile, depend in part on how the rest of our lives is. Medieval peasants did not, so far as I know, deliberately set out to spend their spare time ‘living communally under conditions of frugality and proximity to nature.’ It is the contrast between an urban life, marked by comfort sustained in a certain sort of mutual indifference, and the need to work together in circumstances where we are no longer shielded from our natural smallness and insignificance that makes camping appealing.⁶⁰ The attraction of camping and of the principles that ought to govern it depend on what it is not as much as on what it is. The relation between camping trips and other activities structures the reasons for going camping and how we ought to behave when we go camping, just as the availability of going camping presumably structures the reasons for doing other things and how we ought to behave when we do them. Another way of putting this would be to say that going camping is part of an internally complex, normatively loaded practice we could call ‘leisure activities in conditions of (Western) modernity.’ This practice seems itself part of a broader practice, ‘life in conditions of (Western) modernity.’ Consequently, if we are going to use judgments about camping trips or indeed many other activities we often engage in to illustrate or support our arguments, particularly if the judgment is being generalized, we should consider the way that camping trips fit into a broader pattern of relations.

The fact that humans are reflective as well as cultural creatures suggests not just that many of the practices of which our activities are part are likely to be internally complex and normatively loaded, but that they are likely to be part of one particular practice, the practice of constituting ourselves as agents. By reflecting on our choices, by understanding what we are doing in a certain range of ways, we shape our own agency.⁶¹ This is Strawson’s point in Freedom and Resentment. Of course, we do not shape it from nothing, any more than parents shape a child’s personality from nothing. Still, operating within various constraints, the way our cultural practices shape our choices and their meaning also shapes the kinds of agent we are and can be by describing the range of choices available to us in various particular ways. In Williams’ example, it seems there was a time when it was straightforward to think of nature itself as teleological and humans as occupying a privileged place within it. The fact that we struggle to understand ourselves in that way is a change in our conception of ourselves related to how we understand what our reasons for doing certain things are, itself produced by changes in what we think is possible for beings like us, particularly in comparison to other beings.⁶² If our activities are standardly implicated in our constitution of ourselves as agents, then they will standardly be part of an internally complex, normatively loaded practice. If we want to use judgments about our activities to support claims we make in political and moral philosophy, then, we will often benefit from being clear about the role those activities play in our self-constitution. This is only truer of moral judgments, which are likely to be particularly centrally connected to our conception of ourselves since they govern our relations with other, apparently similar, agents. We may be vulnerable to the same treatment, after all.

Let us set aside what the non-Rawlsian examples of practice-dependence teach us about our reasons for adopting that method, though, and move on to what they might tell us about the uses to which it can be put. Here, Williams’ work is particularly helpful. As we have already seen, Williams was notably hostile to much Rawlsian political philosophy despite sharing a general methodological stance with it. He and other advocates of the realist critique of contemporary analytical political philosophy may not have been or be fully aware of the extent to which they share methodological perspectives with Rawlsians, and so there may be some confusion on both sides about what differentiates them. Still, it would be odd to see them as in agreement. There are often significant differences between what they prescribe for political actors and institutions. It is difficult to imagine Williams condemning welfare state capitalism in the strongly moralized terms that Rawls did, for example.⁶³ Those differences call for explanation, given the apparent similarity of methodological commitments, and in doing so, suggest one of the weaknesses of contemporary Rawlsianism.

I have argued elsewhere at length that the difference between contemporary Rawlsianism and Williams’ form of political realism is Rawlsianism’s greater confidence in the tools of philosophical reasoning as mechanisms for solving the various different conflicts which characterize politics as a distinct sphere.⁶⁴ There, I suggest that each side to the dispute might sensibly learn something from the other, that there are both authentic political questions realists tend to ignore because of their philosophical cast and properly philosophical questions Rawlsians sometimes raise when politics exists because we cannot agree on answers to them. Realists, though, presumably at least partly because they are articulating dissatisfactions with an apparently dominant paradigm, have been fairly clear about how their interpretation of the practice of politics differs from that of the moralism they see in most contemporary analytical political philosophy. Moralism “naturally construes conflictual political thought in society in terms of rival elaborations of a moral text,” its assumption that its audience is able to enact whatever it recommends revealing its disinterest in universals of politics like power and powerlessness.⁶⁵ Politics is for realists about conflict and its management, the first political question of the provision of order.⁶⁶ Order cannot be provided by philosophical arguments alone, because those arguments are neither concrete enough nor capable of resolving differences whose roots lie elsewhere. Historical, institutional, and psychological resources will instead need to be mobilized to legitimate, to make sense of, contemporary power structures and their claim to authority over their subjects.⁶⁷ The political practices contemporary analytical political theory hopes to govern have, for realists, a structure that makes it impossible for the kinds of claims Rawls and his followers make to have purchase on them. Philosophy is not the answer to the questions politics poses – or at least philosophy of the sort most political philosophers do isn’t.

This is a problem for self-identified practice-dependent theorists. It suggests that for all their apparent attention to the structure and form of the practices they are concerned with, they have made a fundamental mistake about the kinds of conceptions of values which can govern them.⁶⁸ The abstract, often philosophical, resources self-identified practice-dependent theorists point to are too remote or controversial to hold the allegiance of citizens that political explanation should, in the end, be for. It is far from clear that many contemporary Western citizens could think of their states as creating and sustaining them in the sense that Sangiovanni suggests as the basis of their distributive obligations to each other, at least not without those states becoming considerably more encompassing and less liberal, for example. Rousseau after all thought that “the total alienation of each associate with all his rights to the whole community” required “changing human nature,” and that most nations in Europe had grown so “incorrigible” as to be incapable of making that change.⁶⁹ Relying on value of ‘the creative, inventive, multifaceted, and diverse nature of human undertakings’ to buttress your defence of liberalism will not do either, since it will hardly count as an answer to the clamoring of those committed to those various undertakings for the special treatment to which they believe themselves entitled, including rights over their competitors.⁷⁰ Equally, treating realist concerns about whether moral ideals can serve as the foundation of a coercive political order as worries about whether their implementation is feasible, as both Ronzoni and Valentini do, is hardly an adequate response.⁷¹

Sangiovanni has tried to argue that practice-dependence gives pride of place to the realist question of “the exercise and command of political power,” avoiding the confusion between it and questions of moral philosophy many contemporary political philosophers fall victim to.⁷² However, not adopting the view from nowhere, accepting that the features of the practice which are being theorized are relevant for the justification of principles to govern it, is not the same as taking the particular perspective on what political problems are and how they can be solved that characterizes realism. If realism is anything, it is the denial of Rawlsianism’s most fundamental methodological commitment, that “formulating idealized … conceptions of society and persons connected” with the ideas which together characterize “society as a fair system of cooperation” are “essential to finding a reasonable political conception.”⁷³ The issue between Rawlsians and realists is not whether politics is prior to morality, but how much morality politics needs. Yet self-identifying practice-dependent theorists have done little to defend the way they integrate their interpretations of practices into their broader theories, as opposed to defending integrating interpretations of practices into their theories at all. Yet as we have seen, there are other theorists who agree that moral and political philosophy and theory should be practice-dependent, yet disagree about how it should be practice-dependent. Self-identified practice-dependent theorists are not the only epistemically respectable theorists around, and a full defense of their work will mean doing more than pointing out the justificatory difficulties of ignoring the world.

By pressing self-identified practice-dependent theorists in this way, I do not intend to suggest that they are incapable of providing the defense I claim they need. As things stand, by addressing themselves to theorists who mainly deny the relevance of practice to moral and political theorizing, they have largely side-stepped issues of what practices are actually like and which sorts of principles are appropriate for them. In particular, they seem to have said nothing that could be used to defend their theories against alternatives we could group together under the banner of realism. Observing that they have side-stepped these issues and not yet said anything about them, of course, does not mean that they must side-step these issues and can say nothing about them. It is simply that as things stand, they are exposed to a powerful critique. That critique is not methodological in the sense that it disagrees about the dependence of conceptions of values on the practices they are to govern. It is methodological in a different sense, in that it queries the canons of interpretation which underlie self-identifying practice-dependence and questions the results they generate. If we are to be practice-dependent theorists, as I have argued we should, then we need to look hard and carefully at the practices our principles will end up depending on. They may, like realists have argued, not be as amenable to being governed by philosophy as self-identified practice-dependent theorists assume.⁷⁴