Subjective Reasons and Political Justification

On one family of views—which encompasses a range of consequentialist and rights-​based approaches—there are independent truths, which philosophers can seek to discover, about how people ought to treat each other. Which political orders are justified depends on these truths, and not directly on whether they are widely accepted by the members of a society so ordered. The qualifier directly is important, because on any plausible view, one of these objective truths will be that there are limits on how people may be treated without their actual, subjective consent.

Another family of views—which philosopher Kevin Vallier characterizes as “contract liberatarianism,” but which encompass contractualist views more generally—consent is not just an important feature of the content of correct political principles, but an important feature of what makes them the correct principles. On these views, justified political principles are just those that could be accepted (or on some variants, could not be rejected) by any reasonable member of a pluralistic society whose members hold wildly diverse conceptions of the good life, with equally diverse accompanying metaphysical, ethical, and religious views. Central to most of these views is the idea of “public reason”—a kind of stripped-​down justificatory language that expresses “thin” policy rationales that dispense as far as possible with contentious premises that might be rejected (even if true!) by some reasonable groups. My own inclination is toward this sort of view, but I want to problematize one way Vallier seeks to argue for this conception:

Here’s one difficulty. We know that, in addition to their deep disagreements about ethics, metaphysics, and religion, members of a pluralistic society also differ profoundly about justifiable uses of political power. Moreover, people who in fact benefit from unjust features of the status quo have reason to reject proposed deviations in a more just direction—but nobody thinks this ought to factor into our evaluation of which orders are acceptable. Contractualist theories only actually get us anywhere by making the relevant form of agreement hypothetical, and importing certain normative terms into the conditions of agreement: Principles all reasonable persons would accept or could not reasonably reject.

The claim, then, is not that I am justified in ordering you to leave me and my property alone—and calling the police or using physical force to make that order effective—because you do actually accept the legitimacy of private property in general or my claim to it under our current legal system, for perhaps you don’t. Rather, the argument is that I am justified because you would accept (or could not reasonably reject) the principles on which my claim rests if you had correct factual beliefs and deliberated in a fully procedurally rational way, regardless of your deeper metaphysical commitments or the comprehensive conception of the good you embrace. This allows the view to be genuinely normative because your current actual beliefs are subject to correction by the requirements of factual accuracy, “reasonableness,” and procedurally rational deliberation—but still distinguishable from the first family of views in that it takes your normative commitments as given. Hypothetical consent here is grounded on formally or procedurally rational deliberation from aims and values that we take as we find them—it does not assume any of these are substantively rationally required.

The trouble is that the distinction is not as clear cut as all that. For one, our aims and values are typically inextricably tied up with a variety of factual beliefs that may be false. Many people’s values flow from religious doctrines that make factual claims about the world that may be false—and indeed, at least some of which must be false. In choosing important life projects—like pursuing a career as a writer on political ideas—we rely on predictions about how likely this is to render us personally satisfied in the future, whether this project is likely to be a way of meaningfully contributing to the larger welfare of society, and so on. Even a weaker corrective—purging us only of those beliefs widely agreed to be false by the best current physical and social science—would likely take us a substantial distance from our actual selves.

Perhaps more importantly, the stringent requirements of full procedural rationality are rarely satisfied by actual flesh-​and-​blood human beings. The arguments for private property, or for a certain voting protocol, or for structuring the federal government in a certain way, may turn on complex economic, sociological, or philosophical arguments that the modal member of society “may or may not be able to grasp.” As the old cliche has it, half the population is of below-​average intelligence by definition, and there’s no guarantee that all (or even most) of us would be able to really understand the validity of Arrow’s Theorem (or Rawls’ Political Liberalism—or the arguments of Buchanan, Narveson, Lomasky, Gaus, and Tomasi) even under the best of circumstances. The most highly educated imaginable society will still contain many young children subject to political authority and affected by government policies.

When we consider more modest forms of idealization, there’s intutive force to the idea that we treat others as free and equal despite some mild amount of “correction” of their revealed preferences—and even that such correction is sometimes required to treat them as free and equal. I see you about to raise a glass of poisoned wine to your lips and, without a moment to lose, knock it from your hand. Supposing that either you were unaware the wine was poisoned (a false factual belief) or were not thinking clearly in the grips of depression, powerful drugs, or some other mental impairment (a failure of procedural rationality), I have a compelling argument that I have coerced you only in a way that respects your free and equal status: Given your own deeper values, which include a desire to live, I have only done what you yourself would most strongly want me to had you been fully informed and thinking clearly. To borrow an example from Plato, I do not treat my neighbor with equal dignity and respect when I return his borrowed weapons to him while he’s temporarily insane. But the further we travel along this road, the harder it is to claim that the hypothetically-​endorsed principle is one the actually-​embodied person already has subjective reason to accept.

What Vallier’s argument needs, then, is some account of why we treat people in ways that respect their freedom and equality when we treat them in accordance with principles they could or would accept subject to idealized deliberative conditions—regardless of whether they actually accept these principles—but only provided this idealization takes their substantive value commitments as given. Yet it’s not clear that these commitments would themselves survive such idealization, nor that they are more deeply constitutive of “the person”—the actual flesh-​and-​blood person we’re meant to be respecting—than the cognitive and information constraints we feel free to abstract away. In other words, we need an explanation for why “You would endorse this principle if you could understand the proof of Arrows Theorem” is more compelling for justificatory purposes than “You would endorse this principle if you were motivated by impartial concern for all other citizens/​humans/​sentient beings,” despite the implausibility of supposing that these idealized deliberative conditions would not also yield profound revision of people’s actual motivational sets. Moreover, we need an account that does not rely on a theory of the moral person tied to one of those controversial worldviews about which citizens of a pluralistic society reasonably disagree.

Needless to say, Rawls and other contract theorists anticipate and have answers to this line of attack—and this post is long enough without attempting to say anything about whether they are good answers. The point is just that any helpfully normative theory—one that purports to answer questions about justification that actual people currently disagree upon—is going to say something about reasons that “exist,” or that people “have,” even if they do not currently acknowledge having them. What this actually amounts to, and which forms of idealization generate these kinds of reasons, is not at all straightforward, which means that the legitimacy of Vallier’s distinction between families of theories can’t actually be made ex ante: It depends on how well the argument for a given construction procedure within the liberal contract theories actually holds up.