Gaus begins this section by describing a Kantian proposal for how to reconcile the freedom and equality of persons with the authority of morality. First, we bracket our private ends-those things that divide us. Second, we focus on those common considerations that can serve as the basis for justifications of moral requirements. Rawls famously formalized this two-step procedure with the veil of ignorance and the idea of a list of primary goods essential to the exercise and development of the two moral powers. But because the Kant-Rawls solution involves bracketing everything that divides us–most importantly, our different comprehensive doctrines–there is no real evaluative diversity in the original position. The original position, as many critics have pointed out, is not a problem of collective choice, but a problem of individual choice since there is only a single perspective that remains after the veil of ignorance has been imposed. For Gaus, this creates two big problems.
First, how do we know that we have sufficient reason to follow the rules of morality if those rules have been determined only by abstracting away from those values (i.e. our comprehensive doctrines) that we care deeply about? Merely showing that we can get a set of justified moral rules by bracketing what divides us doesn’t establish that we have good reasons to follow these rules when we consider our full set of reasons. This is the problem of stability. Gaus argues the later Rawls recognized this problem, and this explains Rawls’s claim in Political Liberalism that the rules generating by the freestanding argument in the original position provide only a pro tanto justification–the full justification of those rules is not achieved until each person sees that he or she has reasons from within his or her wider doctrine to endorse the rules. When everyone achieves this reconciliation, we famously have an overlapping consensus on a public conception of justice.
But Rawls’s solution appears to ignore (or at least downplay) a second problem: indeterminacy. Rawls gets a unique set of principles out of the original position by eliminating all the evaluative diversity from that choice procedure. But once we let the evaluative diversity back in, either via the overlapping consensus, or in some other way, Gaus argues we are bound to get quite a bit of indeterminacy regarding the publicly justifiable rules. There just isn’t enough common ground, Gaus believes, to derive anything like a uniquely justified list of moral rules. Rawls saw indeterminacy as a potentially fatal problem, but Gaus encourages us to embrace it as the inevitable result of respecting evaluative diversity while searching for publicly justifiable moral rules. If we consider what could be justified to members of the public, Gaus suggests that what we will get is a non-empty and non-singular set of optimal eligible proposals, and of course there will be disagreement regarding the ranking of these different proposals.
So how do we solve this problem of indeterminacy? Here is where Hume provides a helping hand. Instead of following Kant or Rawls and looking for ways to exclude evaluative diversity and thereby avoid indeterminacy, Gaus urges us to draw on the resources of a more Humean approach to social morality. In particular, Gaus says we can draw on the idea that moral norms evolve and become stable as a solution to the large-scale coordination problem of how to live together cooperatively in a way that can benefit everyone. Without this kind of convergence–for example, convergence on rules over property rights and how they are to be respected–social life would be filled with costly and inefficient conflicts. So even if idealized members of the public will not converge on a unique set of moral rules due to evaluative diversity, real people will inevitably do so as a result of evolutionary pressures to reach an equilibrium. This evolutionary perspective also helps to dissolve the puzzle of mutual authority mentioned in the previous sections. Social morality is a form of decentralized social authority between many people, and not some puzzling form of mutual authority between two agents.
Gaus’s brilliant solution to the puzzle of how morality’s authority can be reconciled with our freedom and equality is thus a reconciliation of Kantian and Humean perspectives. A publicly justified morality is a set of moral rules that is both: (1) an equilibrium solution that has evolved to help individuals solve their large-scale coordination problem of social cooperation, and (2) is among the members of the optimal eligible set that could be justified to idealized members of the public. As Gaus puts it, without the evolutionary story, the Kantian approach is hopelessly controversial or indeterminate. But without the emphasis on public justification, evolved moral rules are merely authoritarian.
Obviously there is a lot going on here since this section effectively summarizes the main argument of the whole book. Here are just a couple of questions/comments:
1. Suppose Albert looks at the moral rules that emerge from the original position (or some similar device) and correctly decides that these rules cannot be justified (or their deliberative priority cannot be justified) within his own comprehensive doctrine. Gaus seems to believe this is a fatal objection to that set of rules, at least with regard to their authority over Albert. But don’t we need to know more about Albert’s doctrine, in particular about its normative content, before we can reach this conclusion? After all, what if Albert’s doctrine is racist, or is deeply repugnant in some other way? Surely the fact that racists cannot reconcile their comprehensive doctrines with a freestanding justification is no reason to doubt that the freestanding justification may nevertheless be authoritative with regard to that person? Put another way, don’t we need to restrict the constituency whose objections might generate genuine stability problems to some normatively idealized set of people? And if we do this, as I think we have to in order to avoid wildly counter-intuitive results, this is another kind of bracketing strategy. I’m not sure whether or not this is consistent with Gaus’s approach, though his rejection of the expansive view in the previous section may indicate it is inconsistent. I do think this kind of normative bracketing strategy may also help to reduce, though not eliminate, the amount of indeterminacy we find amongst members of the public.
2. Gaus supposes that there will be a fair amount of indeterminacy regarding what can be justified to members of the public. But now consider the Humean side of things–there is surely also a fair amount of indeterminacy regarding which set of rules can evolve to serve as a stable equilibrium for social cooperation–which particular set of rules we end up with is (as Gaus will argue) path-dependent. So we have two indeterminate sets of moral rules: the one generated by consulting the members of the public (the optimal eligible set), and the one that can meet evolutionary pressures (call it the evolutionarily eligible set). Here’s an obvious question: what if these sets don’t overlap at all? What if none of the moralities in the optimal eligible set can be found in the evolutionarily eligible set? Maybe this seems very unlikely, but it all depends on how demanding the test of public justification turns out to be. Suppose, for example, that there is such evaluative diversity among members of the public that the only rules that can be justified to all of them will be very minimal rules prohibiting certain kinds of killing and harm. Nothing more complex can be justified to everyone. But suppose it turns out that only much more complex moralities are evolutionarily stable–e.g. stable moralities must include all kinds of rules about sex, greetings, and property–but these sort of complex rules are far too controversial to ever be acceptable to all members of the public. Now it looks like we face a bad choice between unjustified but stable moral authority, or unstable but justified social life. One way to avoid this result might be to stipulate that if social cooperation is not possible without a moral rule to govern behaviour type X, then this provides all members of the public with sufficient reason to prefer any moral rule rather than no moral rule regarding X. If the need for social cooperation always provides decisive reasons in this way, then we ensure there will always be overlap between our two sets. But this kind of stipulation comes with a big price tag, namely, it seems like an extreme form of the bracketing strategy that Gaus disapproves of: it assumes that individuals share one overriding aim, social cooperation, and achieving this aim is always sufficient to override a person’s other values or commitments. Morality authority starts to look pretty authoritarian if we make this stipulation.