1. What is Contextualism?
At the heart of the contextual approach is a commitment to addressing specific, concrete problems currently faced in the public sphere.  The guiding idea is that we can develop insight into the demands of justice in the process of addressing these problems. In other words, a contextual analysis does not proceed by constructing a complete theory of justice and then applying it to whatever issues arise, but instead views theory development as something that grows out of the consideration of contextualized problems. There is a resonance here with the pragmatism of John Dewey, at least in methodology if not in ultimate conclusions. The pragmatists, particularly Dewey, saw philosophical inquiry as a tool for solving problems or overcoming difficulties faced in the course of one‘s life. Adequately resolving the problem may lead to far ranging philosophical investigation, but the ensuing philosophy ought always to be somehow tied back to the resolution of some difficulty faced in the world.
In this pragmatic spirit, the first defining characteristic of the contextual approach is that it is narrow, which is to say that attention is initially focused on concrete, well-defined problems. In contrast, on a wide approach one might take the entire institutional structure of a nation, or perhaps even the global order as a whole, as the primary subject of evaluation. Rather than doing this, a contextual approach will pick out a single institution or set of institutions as the subject of discussion and possible change. Other features of society are, at least initially, held fixed. It may of course turn out that the investigation needs to be widened. In other words, what at first appear to be unrelated features of a social system may turn out to be inextricably linked with the chosen problem, and in some cases resolving the problem may lead one to call into question the entire structure of society. But a contextual approach begins under the defeasible assumption that a given problem can be resolved without radically altering the social structure.
The second defining characteristic of a contextual approach, closely related to the first, is that contextualism is minimalist, which is to say that theoretical issues about justice are addressed only to the extent necessary to solve the problem at hand. The aim is not to find sweeping principles that settle all questions of justice, but instead to identify the principles or considerations that appropriately govern a particular issue. In other words, the contextual approach will generally not require one to develop a complete hierarchy of political values, with all of the trade-offs and relationships between them fully worked out. Given that the central aim is to evaluate the justice of a particular set of institutions and find ways to make those institutions more just, it will be sufficient to identify and work with the considerations of justice relevant to the specific problem under examination.
The third defining characteristic of contextualism is that the conclusions that result from a contextual analysis are to be understood as incremental improvements. When a contextual examination of a current set of practices generates a set of recommended changes, this does not involve the claim that the new policies or institutions that would result from adopting the recommendations would be perfectly just. Rather, the claim is that the changes would minimize or eliminate particular injustices in the current system, and that they would do so more effectively and efficiently than any alternative that has yet been considered. This does not rule out the possibility that some unimagined alternative would have been even better, or that the proposed changes will lead to new forms of injustice that were not foreseen.
It is, of course, part of the contextual process to seek out and consider various alternatives and strive to anticipate the full consequences of each. Nonetheless, the aim is not to find a perfectly just solution to whatever problem is under consideration, but instead to work towards increasingly just institutions and policies by making incremental changes. At any given moment one hopes to identify a step in such a series of changes, a step which will make society more just, but there is no expectation that one will identify the final step.
To summarize, a contextual approach focuses on a narrow problem, which it investigates in a minimalist manner, yielding judgments of how to make incremental improvements. Before saying more about how I think this can be done, let me first motivate contextualism by contrasting it with two widespread approaches that dominate political philosophy, which I call the idealist approach and the schematic approach. In each of these approaches, the primary task of political philosophy is taken to be something other than addressing a specific, concrete problem faced in the public sphere. For an idealist, the primary aim is to develop an account of an ideally just state, whereas for a schematist the primary aim is to formulate a complete set of principles of justice. This typology of idealists, schematists, and contextualists is admittedly imperfect, and there is work that combines elements of different approaches, but I am not particularly concerned with sorting specific projects in political philosophy into one of these categories. Rather, my aim is to highlight problems faced by idealist and schematic approaches, and motivate the adoption of contextualism as an alternative.
2. The Idealist Approach
On a pure version of an idealist approach, one‘s analysis is focused on defending a particular vision of an ideally just society. Historical work in political philosophy often falls into this category. Plato‘s Republic, for instance, is an excellent example of what I am calling an idealist approach. On at least some interpretations, Hobbes‘ Leviathan, Locke‘s Second Treatise, and Rousseau‘s Social Contract all share this generally idealist bent.
So what is at issue between idealists and contextualists? That depends on how we understand the idealist program. One possibility is that we might construe the idealist as working towards a recommendation for wholesale societal change. The idea would be to reform society all at once in the image of the ideally just state as described by the idealist. Interpreted this way, the idealist project is potentially helpful and valuable, but only in times of revolution or state formation. I would not want to downplay the importance of those moments, but in most cases institutional and societal change must proceed piecemeal. There are enormous transition costs, often in the form of war, associated with radical, wholesale change. A contextual approach has the advantage of being able to identify ways to make a society more just that can be implemented without entirely restructuring the social order.
It may be, however, that a defender of an idealist approach would claim that idealist political theorizing can also generate conclusions about how to make society more just that do not require radical change. I take it this to be an implicit aim of idealism, even though it is rarely articulated by those engaged in such an approach. The idea here would be that developing an account of the ideally just state is only the first step in the idealist project. Once this has been done, we would then be in a position to evaluate any potential change to the social structure.
Notice, though, that on this understanding of the idealist approach, any conclusions about how to make actual institutions more just can be reached only by means of an indirect evaluation. A proposed change counts as an improvement if and only if it brings us closer to the ideal. But how exactly do we measure proximity to the ideal? There is a real threat of circularity here, in that an idealist cannot simply assert that whichever of two possibilities is more just is therefore closer to the ideally just state. According to idealism, we need to make judgments of proximity to the ideal in order to determine which is more just, so we cannot base conclusions about proximity on claims of greater justice.
The attempt to make such proximity judgments, and use them to assess the justice of various non-ideal social structures, runs into two related problems. The first is that in many cases, particularly cases in which the current system is vastly different from the proposed ideal, it may not be possible to settle the question of which alternative is closer to the ideal. Suppose, for instance, that ideal theory tells us that in the ideally just state children are raised communally. How, given this, are we to judge custody or child support laws in a society like ours in which children are raised in families?  Is there any way to resolve the question, which set of custody laws is closer to communal child-rearing? In many cases this sort of question will have no determinate answer.
Moreover, there is a second, deeper problem lurking here. Namely, even when we can answer questions of this sort, these answers will not always provide the guidance we are seeking. This is because it is not always the case that closeness to the ideal tracks justice, which is to say that the fact that one possible social structure is more similar to the ideally just state than another does not necessarily mean that the former is more just than the latter.
This divergence between closeness to the ideal and justice is possible because the relevant proximity judgments must be made in a way that avoids circularity. As indicated above, these proximity judgments must not presuppose assessments of relative justice, given that they are intended to be used to determine such assessments. If the proximity judgments are to avoid illicit normative premises, they must depend on the number and size of the differences between a given non-ideal state and the ideal. There are many ways this could be worked out, but at the very least it will require that a social structure with two deviations from the ideally just state is judged to be less just than an alternate structure with only one of those deviations.
This may appear to be an innocuous conclusion, but it turns out to be problematic. Consider, for instance, the case of affirmative action. One common defense of affirmative action relies on the premise that minorities are unjustly denied opportunities through systematic de facto segregation and inferior educational systems. Preferential admissions to universities are, on this view, a means of overcoming this systematic injustice. Those who defend affirmative action in this way often explicitly agree that in an ideally just society there would be equality of access to primary education and impartial admissions to colleges and universities. The system with inequalities in early education and preferential admissions is therefore two steps from the ideal. As a result, if we attempt to assess proximity to the ideal in terms of structural similarity rather than deriving proximity judgments from antecedent evaluations of justice, we are forced to conclude that this system could be made closer to the ideal, and therefore more just, by eliminating the affirmative action programs.
This simple proximity calculation, however, is clearly not sufficient to settle the issue. Even those who reject affirmative action acknowledge the need to develop an argument that goes beyond pointing out that two deviations from the ideal are more than one deviation from the ideal. Instead, the debate turns on direct assessments of and comparisons between social structures with affirmative action and those without it. These direct evaluations, which involve shifting away from an idealist approach and into a contextual approach, are what settle the matter. Idealism, it turns out, will not on its own enable us to evaluate non-ideal societies.
3. The Schematic Approach
In what I call the schematic approach, which has become increasingly prevalent in recent years, the primary task of political philosophy is taken to be the identification of a complete set of principles of justice. These principles can then be used either to develop an account of an ideally just state, one which reflects them perfectly, or to evaluate and criticize non-ideal states.
This schematic approach is exemplified both by Rawls‘ A Theory of Justice and Nozick‘s Anarchy, State, and Utopia. Both share the defining characteristic of schematism, which is the assumption, implied or overt, that once we have done the hard philosophical work of identifying principles of justice and determining how to make trade-offs between them we will be able to solve specific problems more or less automatically. The fundamental problem faced by the schematic approach, or so I will argue, stems from the need to make trade-offs between principles or to find a balance between competing political values.
The difficulty I have in mind here is an epistemological problem. I am not denying that some ways of balancing competing political values or trading off between conflicting principles are better than others. There may even be, in any given instance of conflict, a single best way to resolve the tension. The problem is that our only epistemological access to the solution will typically be through a consideration of the details of the situation in which the conflict occurs. When such a conflict arises, the appropriate response is therefore to engage in a contextual investigation of the problem rather than appealing to some previously settled schema or system of trade-offs.
To see this, consider the following toy example of a debate over campaign finance reform. On one side of the debate, let us say, are those who claim that campaign contributions are a form of speech. Freedom of speech, they insist, must be protected, and so campaign contributions ought not be regulated. On the other side, imagine, are those who point out that unlimited campaign contributions allow the wealthy to have far more influence over the political system than the poor. These proponents of campaign finance reform insist that concern for the fair value of our political liberties requires us to impose limits on campaign contributions. Granted, this is at best a caricature of any actual debate over campaign finance reform, but it will serve to illustrate the problems faced by a schematic approach.
Assuming that our schematic approach endorses both freedom of speech and fair value of political liberties as genuine political values, it now must rely on some predetermined method of balancing these values to settle the matter. What could that balancing principle possibly be? One possibility would be for one of the considerations, let us say freedom of speech, to trump the other. We might express that by asserting that freedom of speech is lexically prior to the fair value of political liberties.
This sort of view would be very convenient on a schematic approach because it allows the schematist to avoid formulating a more complicated principle for weighing competing considerations against one another. Nonetheless, although the idea of trumping or lexical priority would constitute a tidy side-stepping of the balancing problem, it is not a particularly satisfying solution. If we really do count both freedom of speech and fair value of political liberties as genuine political values, how can we accept a relationship of lexical priority between the two? Such a relationship entails that even the slightest gain in freedom of speech is sufficient to justify a complete sacrifice of the fair value of political liberties. Surely that is not right, at least not if we take both values seriously.
A more sophisticated version of lexical priority might endorse the weaker claim that freedom of speech trumps fair value of political liberties up until a sufficient threshold of free speech has been secured. But then the trade-off problem simply recurs. Once this threshold has been reached, are any further advances in terms of free speech irrelevant? Are they automatically trumped by considerations of the fair value of political liberties? This too seems implausible, but it is what the sophisticated version of lexical priority would require.
At some point we must admit that genuine balancing between the two political values is needed. Neither trumps nor is lexically prior to the other; the question of how large a gain or sacrifice is being made is critical to finding a resolution.
Once we recognize that we need a richer balancing principle than is afforded by assertions of lexical priority, the difficulty with a schematic approach becomes apparent. The problem is that there is no way to articulate a neat formula for making trade-offs. Although we can talk generally about, say, large gains in the fair value of political liberties outweighing small sacrifices in freedom of speech, I see no way to move from this level of abstraction to a precise formula that will tell us which trade-offs to make. This is exacerbated by the fact that only by confronting an actual, contextualized problem can we imagine possible solutions and understand what sort of trade-offs are available. Without engaging with the details of the case, in a contextualist manner, we won‘t be able even to identify the possible trade-offs, let alone judge which is best.
4. How to Be a Contextualist
I have been arguing that a contextual approach has certain advantages over both idealist and schematic approaches. The idea has been that by beginning with the consideration of specific problems and gradually generating theoretical insights in the process of addressing these problems, a contextualist can avoid resorting to either the proximity judgments required by an idealist approach or the balancing principle required by a schematic approach. At this point, though, one might wonder how a contextual analysis is supposed to work. How will a contextualist be able to make the direct comparisons between alternative possible solutions that generate incremental improvements in the justice of a social structure? Will not these judgments have to presuppose a theory of justice like the ones offered by idealists or schematists?
My first response to this sort of worry is to point out that if there is a difficulty for contextualism here, presupposing an idealist or schematic theory will not be of much help. This, at least, is what I hope to have shown in the discussion of those two approaches. What is needed instead is a way of engaging with the problem from within the contextualist framework. This is, admittedly, not an easy task, but there are at least some examples of how it might be accomplished. For instance, Marx‘s critique of capitalism, insofar as his argument is intended to reveal the injustice inherent in the capitalist system and then identify a way of eliminating it, proceeds along roughly contextualist lines. More recently, the approach taken by Walzer in Spheres of Justice has similarity to the form of contextualism I am endorsing. 
In any case, though, the real challenge for an advocate of contextualism is to demonstrate the effectiveness of the approach by engaging in it. I have attempted to do this elsewhere, focusing on the case of welfare policy in the U.S. The details of that analysis are beyond the scope of this paper, but perhaps a quick summary can help illustrate the potential effectiveness of a contextual approach.
In examining welfare programs in the U.S., I begin with the idea that such programs are best understood as aimed at discharging mutual obligations owed between fellow citizens, which implies that an analysis of U.S. welfare policy requires determining the content of these obligations. Rather than attempting to provide a fully worked out view of the demands of fellow citizenship in the abstract, I appeal to contingent facts about the U.S. to substantiate what I take to be a set of genuine reciprocal obligations relevant to welfare policy that are owed between fellow U.S. citizens. These are, on the one hand, obligations to minimize one‘s claims for assistance and, on the other hand, obligations to provide assistance when necessary and to do so in a way that does not undermine the recipients‘ status as fellow citizens. Appealing to further contingent facts about the U.S., I then argue that these obligations translate into a duty to work and a right to work, respectively. This leads me to endorse what I call “modified welfare reform,” which is a set of programs that includes work requirements similar to those instituted in the U.S. welfare reform of 1996, but also incorporates several significant modifications to current welfare programs in the U.S. In particular, I advocate increasing the scope of welfare programs to include all those in need of assistance, tailoring sanctions to facilitate future compliance, eliminating time limits, and, perhaps most importantly, creating a publicly funded program that provides jobs of last resort.
What is noteworthy about this analysis, for present purposes, is that it addresses a specific, concrete problem, offers a recommendation for how to reduce injustice or promote justice, and does this without appeal to an idealist or schematic theory of justice. This is not to say that I abstain from theoretical or philosophical analysis, but rather that I attempt to engage in this analysis in a minimalist fashion, directly addressing the problem at hand rather than treating any theoretical issues as mere instances of more general, abstract problems. Moreover, it is the appeal to contingent features of the situation that allows me to make progress at key junctures of the analysis. As a result, insofar as my analysis can be characterized as constructing a theory of welfare policy or distributive justice, it is at most a partial theory. This illustrates one of the differences between my contextual approach and an idealist or schematic one. Importantly, this difference reveals one of the traps that beset idealist or schematic approaches, which stems from the fact that defending a general theory of welfare policy, or of the obligations of fellow citizenship, is an extremely difficult task. If one operates under the, perhaps implicit, assumption that this sort of task needs to be accomplished before one can evaluate welfare policy in the U.S., it is possible that one will never get to the point of being able to draw meaningful, convincing conclusions.
It is, of course, also possible that my own analysis is unconvincing, and perhaps the way in which it fails exposes a vital flaw in the contextual approach. The only way to address that possibility would be to work through the details of the analysis and attempt to demonstrate its success, which is beyond what I can do here. My hope, though, is that describing the analysis helps illuminate the differences between contextualism, on the one hand, and idealism or schematism, on the other, and that the analysis is sufficiently promising to motivate further exploration and development of the contextual approach.
Daskal, Steven (2008). “Fellow Citizenship and U.S. Welfare Policy.” International Journal of Applied Philosophy 22: 281-301.
Dewey, John (1998). “The Logic of Judgements of Practice.” In Larry A. Hickman and Thomas M. Alexander, eds., The Essential Dewey, Volume 2: Ethics, Logic, Psychology, 236-271. Bloomington: Indiana University Press.
Dewey, John (2002). Human Nature and Conduct. Amherst, NY: Prometheus Books.
Herzog, Don (1985). Without Foundations: Justification in Political Theory. Ithaca: Cornell University Press.
Hobbes, Thomas (1994). Leviathan. Indianapolis: Hackett Publishing.
Lichtenberg, Judith (1987). “Foundations and Limits of Freedom of the Press.” Philosophy and Public Affairs 16: 329-355.
Locke, John (1980). Second Treatise of Government. Indianapolis: Hackett Publishing.
Marx, Karl (1978a). Capital, Volume One. In Robert C. Tucker, ed., The Marx-Engels Reader, Second Edition, 294-438. New York: W. W. Norton.
Marx, Karl (1978b). Manifesto of the Communist Party. In Robert C. Tucker, ed., The Marx-Engels Reader, Second Edition, 469-500. New York: W. W. Norton.
Nozick, Robert (1974). Anarchy, State, and Utopia. New York: Basic Books.
Plato (1997). Republic. In John M. Cooper, ed., Plato: Complete Works, 971-1223. Indianapolis: Hackett Publishing.
Rawls, John (1999). A Theory of Justice: Revised Edition. Cambridge, MA: Harvard University Press.
Rousseau, Jean-Jacques (1997). Of the Social Contract. In Victor Gourevitch, ed., The Social Contract and other later political writings, 39-152. Cambridge: Cambridge University Press.
Sunstein, Cass R. (1999). One Case at a Time: Judicial Minimalism on the Supreme Court. Cambridge, MA: Harvard University Press.
Walzer, Michael (1983). Spheres of Justice: A Defense of Pluralism and Equality. New York: Basic Books.
Wood, Allen W. (1999). Kant‘s Ethical Thought. Cambridge: Cambridge University Press.