From this viewpoint, it becomes difficult to deny the moral importance of any other person or group of persons, for this is not a denial that one could endorse from the targetted position itself. For example, the consistent White Supremist must hold that, if (contrary to fact) he were black, then he ought to be discriminated against. But it seems unlikely that a rational person could truly endorse this conclusion. They would not want to be harmed by others if they were black. This casts doubt on whether they can consistently endorse racist behaviour in the form of a universal moral prescription.
It will often happen that no single conclusion is deemed acceptable from every possible viewpoint. We thus require some way to adjudicate between conflicts of interest, from the moral point of view, in a manner that meets the requirements of consistency. Hare suggests that we treat the conflicts between people as we would a conflict within a person, allowing trade-offs between costs and benefits so as to reach the optimum result for the collective. In other words, consistency leads to utilitarianism.
But is it really true that consistency alone forces us to accept utilitarianism? There are three major arguments against this conclusion. We might deny that consistency requires us to adopt a universal or collective viewpoint. But even if we accept Hare's argument that consistency requires us to imagine ourselves in the position of others, it isn't clear that his "maximizing" method is the appropriate way to resolve conflicting preferences. We might think that such a method overlooks the distinction between persons. Or we might deny that "all preferences are created equal" -- that is, we might hold that there are moral facts that exist independently of human desires, such that not all preferences contribute equally towards the moral verdict.
This final objection rests on assumptions that a non-cognitivist would not be willing to grant. McNaughton (pp.169-170) tries to bolster the intuitive appeal of this objection by examining Socrates' decision to take the hemlock against his friends' wishes:
It looks as though his action will not maximize the satisfaction of preferences. He is satisfied that, having been sentenced to death by a properly constituted court, he is required to accept the verdict. Moreover, he believes that death is not fearful, and that his friends' distress at the thought of his death is due to their failure, exacerbated by their grief, to see the situation aright. He admits that, if he were in their position, he would not endorse the decision to take the hemlock. But why should this realization affect his present moral judgement when he believes that their opposition is due to an inadequate appreciation of the situation? He is quite consistent in sticking to his original decision... If, after putting myself in the other people's position, I remain convinced by the reasoning that led me to believe that the action was right in the first place, then I need not withdraw it.
There are two responses the utilitarian can make here. Firstly, he can grant that misinformed preferences need not contribute towards utility. If the friends' preferences are due to an intrinsic desire that Socrates not be harmed, and the mistaken belief that death harms him, then the utilitarian can grant that utility is in fact best served by allowing Socrates to take the hemlock. If Socrates is truly not harmed by it, then the friends' intrinsic desires are not thwarted by this course of action after all. So the utilitarian can take someone's "inadequate appreciation" of the non-moral facts, at least, into account.
Moreover, the non-cognitivist will simply deny that there are any desire-independent moral facts for us to be mistaken about. Socrates' moral conviction is simply one preference among many, and he cannot presuppose it to be well-grounded, for that is precisely what is at issue here. As Hare writes, "To insist on the prior authority of the moral intuitions that one starts with is simply to refuse to think critically." In questioning the validity of our moral intuitions, we must be prepared to go beyond them.
The second anti-utilitarian argument mentioned above is the "separateness of persons" objection, which I have discussed before.
The first argument suggests that the egoist need not be inconsistent. This poses a greater challenge to the non-cognitivist utilitarian, for they recognize no independent moral facts which may be pointed to in order to rebut the egoist, and it is not obvious that egoist exhibits any internal inconsistency.
But let us distinguish two forms of egoism. The 'ethical egoist' holds that each individual ought to pursue their own self-interest. But it would seem inconsistent for the egoist to universally endorse other people's selfishness, as their selfishness would be to his own detriment. Alternatively, the 'personal egoist' holds that everybody ought to promote his interests, no matter their own. Of course, this judgment could never survive universalization -- he would not be willing to accept it were he anybody else -- but the egoist might simply hold this as a personal preference, and refuse to make any universal claims at all. That is, he could become an 'amoralist'. I have previously argued that even the amoralist may be criticized for inconsistency, but that need not concern us here. We may merely conclude that, if he adopts any moral viewpoint at all, consistency will lead the non-cognitivist to utilitarianism.
Or will it? We have so far been applying the test of consistency only at the 'formal' level, of mediating between conflicting desires. But it might also be used to yield substantive judgments about which intrinsic desires are more rationally supported than others. Suppose we live in a society full of racists. We have seen that consistency would prevent us from being racist ourselves - we would give the preferences of black people equal weight to those of whites. But what if all the white racists prefer to see the black man suffer? The collective weight of their preferences might outweigh his lone opposition. If we merely consider ourselves in the position of each, the most preferences will be satisfied by endorsing racist behaviour. But suppose that we instead consider what it would be consistent to prefer from the position of each. We would (seemingly) then have to disregard the racists' preferences, for they are supposedly inconsistent.
But this argument goes too fast. We have in fact only established that racist moral judgments are inconsistent, as they cannot be universalized. But personal preferences need not be universalizable: I can prefer butter to margarine without thereby committing myself to the universal judgment that everyone ought to do likewise. Similarly, the racist might prefer to see black people worse off, despite recognizing that he could not universalize this into a moral judgment.
What the anti-utilitarian requires is some grounds for criticizing the consistency of personal intrinsic preferences. The non-cognitivist will refuse any move to appeal to desire-independent moral facts, for he denies the existence of such metaphysically "queer" entities (as Mackie would put it). But the non-cognitivist might follow Michael Smith in adopting a richer conception of rationality that goes beyond mere means-ends reasoning, instead allowing a desire-set to be rationally assessed on grounds of unity and internal coherence. It may be that a set of specific desires (e.g. for the good of certain people) could be better explained and justified through the addition of a more general desire (e.g. for the good of all persons). This conception of rationality enables us to rationally criticize the arbitrary distinctions drawn by racists and other bigots. We might then conclude that their preferences ought to carry less moral weight, to the degree that their desire sets are not maximally coherent.
The utilitarian might grant all this, but simply redefine his notion of utility such that it comprises the satisfaction of rational desires. This would yield a theory quite different from how utilitarianism has traditionally been conceived, but it might better capture the fundamental utilitarian ideal of treating everyone with equal concern, as it could prevent selfish or bigoted preferences from justifying the worse treatment of unpopular individuals or groups. We thus find that the requirement of consistency can have a great impact on our moral reasoning, forcing non-cognitivists towards some form of utilitarianism.