Abstract
Can one give an account of a perfectly just society without invoking principles governing our responses to injustice? My claim is that addressing this question puts us in a position to reveal ambiguities and problems with the way in which Rawls draws the ideal/nonideal theory distinction that have so far gone unnoticed. In the first part of my paper, I demonstrate that Rawls’s original definition of the ideal/nonideal theory distinction is ambiguous as it is composed of two different conceptual distinctions, before clarifying the distinctions involved, paying particular attention to the unfamiliar distinction between primary and secondary principles. I then show that we can best account for what Rawls is actually doing at the level of ideal and nonideal theory by invoking this distinction between primary and secondary principles. This result sets the stage for my argument in the second part. I first explain why Rawls does not have access to an understanding of the strict compliance condition that can account for the irrelevance of secondary principles for a complete account of the principles regulating a perfectly just basic structure. I then point out that there is a tension between what Rawls claims to be doing at the level of ideal theory and what he is actually doing at the level of ideal theory. On this basis, I argue that Rawls’s ideal (domestic and international) conceptions of justice are incomplete because they do not encompass secondary principles. The Conclusion unpacks the contributions this article makes to the ideal/nonideal theory debate.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
See Rawls (2001, pp. 42–43) for a formulation of the two principles.
See Nozick (1974, pp. 150–153) for the three principles of his entitlement theory.
For the most comprehensive reconstruction of Rawls’s views on ideal and nonideal theory, see Simmons (2010). There are a wide variety of questions that are discussed under the heading of the ideal/nonideal theory debate. For instance, some participants in this debate simply argue that more work needs to be done on questions of nonideal theory and that there is a certain urgency to this (Robeyns 2008, pp. 347–348; Farrelly 2007). Others critically engage with Rawls’s specific version of ideal theory. They either criticise it for being too utopian to be useful (Farrelly 2007; Mason 2010; Mills 2005), or they criticise it for not being utopian enough—insisting that in clarifying the value of justice one should not be constrained by any feasibility considerations (Cohen 2003, 2008; Swift 2008). Others again call into question Rawls’s assumption that only ideal theory can provide us with a ‘systematic grasp’ (Rawls 1999a, p. 8) of how we are to deal with existing injustices (Sen 2006, 2009, pp. 1–27, 87–113; Wiens 2011, pp. 1–26). In reply, some try to provide a full-fledged defense of Rawls’s position on this issue (Simmons 2010, pp. 30–36). The issues I am discussing in this paper are not straightforwardly addressed in any of these debates, but they nonetheless bear (indirectly) on these debates.
With the notable exception of his discussion of civil disobedience, Rawls deals in his A Theory of Justice almost exclusively with questions of ideal theory. See Rawls (1999a, pp. 319–343).
To maintain that the principles of justice as fairness suffice to give a complete account of a perfectly just basic structure is compatible with Rawls’s view that his two principles only characterise one aspect of a well-ordered society. According to him, ‘principles of justice are but a part’ of a ‘social ideal’ that also encompasses ‘other virtues’ (Rawls 1999a, pp. 8–9) besides justice, like efficiency or excellence in the arts and sciences. I will not concern myself with all aspects of a well-ordered society, only with its justice-related aspects. With regards to the virtue of justice, Rawls makes it clear that a basic structure that fully conforms to his two principles would qualify as perfectly just.
I am deliberately not saying that primary principles are those principles that establish normative standards, and secondary principles are those principles that are concerned with meeting noncompliance with these normative standards. The problem with this alternative definition is that secondary principles (e.g. principles of compensation) are also normative standards. To be a normative standard is, therefore, not a distinctive feature of primary principles that sets them apart from secondary principles. Note also that principles of the secondary type are not only concerned with responses to noncompliance with primary principles, they also tell us how to deal with noncompliance with (other) secondary principles. For example, if someone did not get the compensation to which she or he was entitled, we have to take recourse to secondary principles to identify the right response to this noncompliance with a secondary principle.
See Nozick (1974, pp. 150–153) for the three principles of his entitlement theory.
Attempts to distinguish primary from secondary principles on the basis of substantive criteria are doomed to fail. For instance, one could try to argue that what sets apart primary from secondary principles is that only the former ‘provide a way of assigning rights and duties in the basic institutions of society’ and ‘define the appropriate distribution of the benefits and burdens of social cooperation’ (Rawls 1999a, p. 4). However, secondary principles assign rights too, for example, the right to a fair trial; and they also deal with the distribution of benefits and burdens, for instance, in the case of reparation.
It is true, in the ‘First Part of Ideal Theory’ Rawls provides a list of principles for the regulation of relationships between peoples. This list encompasses next to principles like ‘Peoples are to observe treaties’, other principles that are secondary principles: e.g. ‘Peoples are to observe certain specified restrictions in the conduct of war’; or ‘Peoples have a duty to assist other peoples living under unfavorable conditions that prevent their having a just or decent political and social regime’ (Rawls 1999b, p. 37). However, it would be a mistake to infer from this that Rawls considers secondary principles to be part of his ideal conception of international justice. What Rawls is doing here is to invite us to look at a list of principles that are commonly invoked in standard textbooks on international law by authors who do not make a distinction between ideal and nonideal theory. It is for this very reason that this list is immediately followed by a section entitled ‘Comments and Qualifications’ in which Rawls emphasises that ‘some’ of the principles included in this list ‘are superfluous in a society of well-ordered peoples, for example, the seventh principle regarding the conduct of war’ (Rawls 1999b, p. 37; my emphasis).
Participants in the ideal/nonideal theory debate seem to simply take it for granted that punishment only figures in nonideal theory. See, e.g., Mason (2010, p. 679n6). Robyens even goes so far as to argue that Rawls was wrong to include principles of punishment in his nonideal theory, even though it seems to be a rather straightforward case for a response to injustice. Robeyns does not deny that ‘we need institutions of criminal justice based on a theory of punishment’ but she denies that we ‘need to be terribly concerned with this kind of noncompliance, since it is reasonable to assume that there will always be at least a few citizens who will violate the laws, whether these laws are just or unjust’ (Robeyns 2008, p. 347). I do not see how the fact that noncompliance of this kind is to be expected can count as a reason for not taking it seriously. The contrary seems to be the case: because it is to be expected and because how we deal with it affects the higher-order moral interests of citizens, we have to take it seriously. For an account of these higher-order moral interests, see, e.g., Rawls (2001, p. 8).
Rawls conjectures that mild sanctions are sufficient for addressing the assurance problem in societies regulated by his conception of justice as fairness. However, he does little or nothing to underpin this assertion. What it takes to assure the members of different societies will probably vary considerably, depending on a myriad of factors, for instance, on whether the society in question has a recent history of violent conflict. See, Rawls (2001, pp. 196–198).
I will engage with the strict compliance condition and the role it plays in Rawls’s argument for the irrelevance of secondary principles for a complete account of the principles that regulate a perfectly just society in more detail in the second part of this article.
However, punishments introduced for reasons of fairness might nonetheless be unjust. I come back to this point later in the paper.
Rawls leaves open the question of how we are to determine an appropriate level of assurance, since absolute assurance is impossible. Aiming at the highest level of assurance that could be manufactured with the help of a system of punishments seems undesirable, as in many cases this could imply threating citizens with excessively harsh punishments.
Of course, it is possible to re-define ideal theory in a way that severs its connection with perfect justice, and some participants in the ideal/nonideal theory debate have suggested definitions of ideal and nonideal theory that do exactly that. These accounts fall outside the scope of this paper. Laura Valentini (2009, p. 332), for instance, defines as ideal those theories that are ‘constructed under false, that is, idealised assumptions which make social reality appear significantly “simpler and better” than it actually is’, and characterises as nonideal those forms of theorising that manage without idealisations of that kind. Zofia Stemplowska defines theories as ideal that fail to issue recommendations about what is achievable and desirable ‘for us here and now’ (Stemplowska 2008, p. 319), whereas nonideal theories are those that issue ‘recommendations that are both achievable and desirable in the circumstances that we are currently facing, or are likely to face in the not too distant future’ (Stemplowska 2008, p. 324). In both cases ideal theory is not (necessarily) about perfect justice. However, most participants in the ideal/nonideal theory debate follow Rawls in the sense that they do consider the link between ideal theory and perfect justice to be definitive of ideal theory. See, e.g., Robeyns (2008, pp. 341, 343, 347), Simmons (2010, pp. 7, 10, 12, 17–18, 22, 35), Swift (2008, pp. 381–382), Valentini (2012, p. 654).
The question of how realistic an ideal theory of justice should be is one of the key issues in the ideal/nonideal theory debate. Since the point and purpose of my paper is an immanent of Rawls’s account of ideal theory, I need not provide a defence or critique of Rawls’s position on the feasibility question. I return to this issue in the conclusion.
Transitional principles can, of course, also be part of partial compliance theory since in generating transitional principles we need not imagine that everyone will be willing, or able, to act in accordance with the selected principles.
The following argument is, like the entire paper, targeted at Rawls and those who share his basic assumptions. It is not designed to convince those who, following Cohen, think that we are not constrained by any facts when clarifying a fundamental value like justice. See, e.g., Cohen (2003); Swift (2008).
For the same reason, Rawlsians cannot just assume away the assurance problem in ideal theory. Rawls cannot just stipulate that the level of mutual trust between all the members of societies is similar to the one you and your best friend have because he acknowledges that it is a general sociological fact that in ‘a large community the degree of mutual confidence in one another’s integrity that renders enforcement superfluous is not to be expected’ (Rawls 1999a, p. 237; also p. 212).
For example, even our best efforts to realise the difference principle might go wrong because we do not know, and have reasonable disagreements about, which economic policies would be the most suitable to implement it. Recent studies suggesting how difficult it is to achieve fair equality of opportunity provide another case in point. Even members of hiring panels, who are committed to impartiality and who take themselves to be making unbiased hiring-decisions, do, in fact, discriminate against candidates on the basis of implicitly held prejudicial views about them. See, e.g., Wilson and Brekke (1994). The upshot of this is that even if all members of a society are willing to act in accordance with, and uphold, a basic structure regulated by the right primary principles, this does not guarantee that these principles are, as a matter of fact, fully realised. Instances of noncompliance, that have to be dealt with in the right way, are the most probable outcome of political life given its complexities and common sense facts about human nature and human social life (Rawls 2001, p. 101).
See for Rawls’s remarks on the finality of the agreement, e.g., Rawls (1999a, pp. 11, 153, 165–166); and Rawls (2001, pp. 105, 119). The finality condition is another reason why Rawls cannot argue that we do not need to choose principles for meeting (past or future) noncompliance with primary principles in the original position, but can deal with them when and if they occur. However, Rawls fails to acknowledge this (Rawls 1999a, p. 126).
Imagine, for example, the position of those who have been discriminated against in a basic structure that enforced Apartheid. One cannot expect them to forget about this in order to facilitate cooperation with those who supported and benefited from this Apartheid regime, once it has been replaced by a basic structure conforming to the right primary principles. Without any public acknowledgment of the injustice done to them and attempts to address it, this society arguably threatens to undermine the social basis of self-respect of these victims.
Even if it was the case that dealing with such a history of injustice in the right way requires nothing else than to bring about a society that conforms with the two principles that Rawls’s conception of justice as fairness articulates, this case has to be made; and making the case that the right secondary principles are identical with the right primary principles implies acknowledging that there is an issue at stake that involves secondary principles, an issue that could potentially call into doubt the claim that the society in question is perfectly just.
This point is connected to my critique of Rawls’s treatment of punishments in connection with the assurance problem.
Furthermore, the distinction between a justificatory and a factual understanding of strict compliance also allows us to identify what is misguided about claiming that ‘by assuming full compliance, ideal theorists’ like Rawls ‘violate the constraints of a realistic utopia’ (Farrelly 2007, p. 845). Rawls would only violate ‘the constraints of a realistic utopia’ if he were to adopt the factual understanding of strict compliance. But the more sensible reading of Rawls is that he subscribes to a justificatory understanding of strict compliance, that is, he merely imagines for the purposes of selecting principles that these principles will be strictly complied with; and this is compatible with his notion of a realistic utopia.
Simmons is referring to Mills (2005, pp. 168–169) who has levelled this criticism against Rawls.
As one of the reviewers has pointed out to me, there is not only the issue of whether secondary principles are an inherent part of a complete account of the principles that regulate a perfectly just basic structure, there is also the question of whether primary principles can be among the principles that regulate a less than perfectly just basic structure. This depends on how one conceptualises nonideal theory. I am, in this article, not defending a particular view on this issue but am concerned with pointing out a conceptual ambiguity in how Rawls draws the ideal/nonideal theory distinction and showing that his ideal theories of justice are incomplete. Therefore I do not need to resolve this question here. Still, let me conjecture what Rawls’s answer might have been. I take it that he would have argued that primary principles cannot be part of nonideal theory since he advocates a transitional view of nonideal theory (Rawls 1999b, p. 90; Simmons 2010). Rawls conceives of the principles that are adopted under less than perfectly just conditions as a set. He attempts to identify the set that is most suitable to promote a transition to a perfectly just social order. Some of the principles in this set can be, in terms of their contents, identical with primary principles regulating a perfectly just social order. But Rawls could still argue that the role they are playing in a less than perfectly just society is transitional and as such secondary (in line with my functional account of the distinction between primary and secondary principles). But there are competing understandings of nonideal theory and some of them might not look at the principles that regulate a less than perfectly just social order in that way, that is, as part of a set promoting transition to perfect justice. Such approaches could regard principles regulating a less than perfectly just social order as primary principles and, possibly, also as identical, in terms of their contents, with primary principles regulating a perfectly just social order. The question which conception of nonideal theory we should ultimately adopt is, however, outside the scope of this article.
References
Cohen, Gerald. A. 2003. Facts and Principles. Philosophy Public Affairs 31: 211–245.
Cohen, Gerald A. 2008. Rescuing justice and equality. Cambridge: Harvard University Press.
Farrelly, Colin. 2007. Justice in ideal theory: A refutation. Political Studies 55: 844–864.
Freeman, Samuel. 2007. Rawls. London and New York: Routlege.
Mason, Andrew. 2010. Rawlsian theory and the circumstances of politics. Political Theory 38: 658–683.
Miller, David. 2011. Taking up the slack? Responsibility and justice in situations of partial compliance. In Responsibility and distributive justice, ed. Zofia Stemplowska, and Carl Knight, 230–245. Oxford: Oxford University Press.
Mills, Charles W. 2005. ‘Ideal theory’ as ideology. Hypatia 20: 165–184.
Murphy, Liam. 2003. Moral demands in nonideal theory. Oxford: Oxford University Press.
Nozick, Robert. 1974. Anarchy, state, and utopia. New York: Basic Books.
Rawls, John. 1993. Political liberalism. New York: Columbia University Press.
Rawls, John. 1999a. A theory of justice (Revised edition). Cambridge: Harvard University Press.
Rawls, John. 1999b. The law of peoples. Cambridge: Harvard University Press.
Rawls, John. 2001. Justice as fairness: A restatement. Cambridge: Harvard University Press.
Robeyns, Ingrid. 2008. Ideal theory in theory and practice. Social Theory and Practice 34: 341–362.
Sen, Amartya. 2006. What do we want from a theory of justice? The Journal of Philosophy 103: 215–238.
Sen, Amartya. 2009. The idea of justice. London: Penguin/Allen Lane.
Simmons, A. John. 2010. Ideal and nonideal theory. Philosophy & Public Affairs 38: 5–36.
Stemplowska, Zofia. 2008. What’s ideal about ideal theory? Social Theory and Practice 34: 331–340.
Swift, Adam. 2008. The value of philosophy in nonideal circumstances. Social Theory and Practice 34: 363–387.
Valentini, Laura. 2009. On the apparent paradox of ideal theory. The Journal of Political Philosophy 17: 332–355.
Valentini, Laura. 2012. Ideal vs. non-ideal theory: A conceptual map. Philosophy Compass 7: 654–664.
Wiens, David. 2011. Prescribing institutions without ideal theory. The Journal of Political Philosophy 19: 1–26.
Wilson, Timothy D., and Nancy Brekke. 1994. Mental contamination and mental correction: Unwanted influences on judgments and evaluations. Psychological Bulletin 116: 117–142.
Acknowledgments
For helpful comments on earlier versions of this article, I am grateful to Paul Bou-Habib, David Heyd and the reviewers for this journal. In particular, I would like to thank Fabian Freyenhagen, my conversations with him have been—as always—invaluable. The idea to distinguish between primary and secondary principles originated in conversations with Joshua Kleinfeld. My argument in this article has also benefitted from discussions at the universities of Bochum, Darmstadt, Essex, Frankfurt, Hannover and Vienna, as well as the ‘Philosophy and the Social Sciences’ colloquium in Prague.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Schaub, J. The Incompleteness of Ideal Theory. Res Publica 20, 413–439 (2014). https://doi.org/10.1007/s11158-014-9255-9
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11158-014-9255-9