This revision is insufficient, however. Two sides of one coin indicate that it is not all that we would like, ideally, for the objectivity of morals. On the one side of the coin, we want to leave space for the voice of one calling us from our collective moral darkness into the light. If moral error is judging differently from the rest, then there is no room for the prophet. There would be no prospect for the beginning of moral awareness that can grow into a widespread awakening. The other side of this coin is that at any given time at which there is sufficient consensus, those so consenting cannot be wrong. There can be no corporate error. In the same way that the individualistic version of Protagoreanism makes the individual automatically correct, the corporate version does the same for the group. If corporate moral error is possible, then we have our first reason to reject this version of projectivism.
Our second reason is that the consensus picture fails to account for our moral knowledge, as well. If the substance of moral matters is the uniform way in which humans respond, then we do not really consult those relevant facts when we form moral judgements. We do not check to see if there is a consensus. The moral epistemology associated with this picture is most plausible when the theory focuses on the individual, which renders an account of objectivity implausible. When the account of objectivity is more plausible by placing the focus beyond the individual, the attending moral epistemology is less plausible. We must search for a yet more nuanced version of Protagoreanism for the projectivist
I assumed in my formulation that the near consensus of attitude and judgement was ipso facto correct. To make sense of a moral mistake as the mere divergence of attitude and judgement from the consensus I made the analogy with syntax and semantics in natural languages. I ignored the crucial differences between two quite different approaches to the facts of language: the platonist and the nominalist approaches. According to linguistic platonism, the facts of language are abstract, objective facts not determined by the practices of language users. Language users get the meaning of ‘bachelor’ correct, for instance, when their use of ‘bachelor’ or their mental associations with the word are appropriately related to the abstract object that is the meaning of ‘bachelor’. On this account, majority use or opinions certainly do not constitute the meaning of the term. Accordingly, it is possible for the majority to be wrong about its meaning, while the lone outlier is correct. Only with some substantial background assumptions would the majority opinion even good evidence for the proper meaning of the term, according to this platonist option. On the nominalist view, there are no abstract objects to constitute the language or the meaning of words. Though many different views are possible, let us consider only that the consensus use of ‘bachelor’ constitutes its meaning. Using a term differently from the rest just is—and not merely evidence for—making a linguistic error.
We can now see that the projectivist is in a bit of a bind. If it is agreed that the morally correct prophet is possible, projectivism should be most analogous with linguistic platonism, with its attending external facts to which our judgements are responsible and which the prophet might grasp more clearly than do the rest of us. If projectivism is the modern formulation of Protagoreanism, then it must be most analogous to linguistic nominalism, with its attending lack of external facts or standards to which our judgements are responsible. Our linguistic practices, i.e., our linguistic measurings, make meanings what they are according to the latter model, but not according to the former. Linguistic nominalism is the linguistic version of the measurement thesis on the social level. The social version of the moral measurement thesis can avoid relativism and it can secure the possibility of moral disagreement, but it renders the prophet and wide-ranging error impossible, as does linguistic nominalism render the linguistic equivalent to the prophet and wide-ranging error impossible.
But, wait! There’s more! I have mentioned, so far, only atemporal issues. Morally relevant attitudes and judgements can change over time from incorrect to correct. What we once held to be just or virtuous we might now hold to be unjust or vicious. What we once found to be grotesque and abhorrent, we now rather approve. Temporal relativism threatens. Instead of the community once being blind and now seeing, it goes from seeing one thing perfectly clearly to seeing something else perfectly clearly, perhaps with a stage of uncertainty in between. Moral blindness for the group is not possible at any particular time, which entails that there is no possibility of moral conversion. Of course, anyone hoping to make sense of the possibility of the moral prophet is likely to want to make sense of the possibility of moral conversion where moral conversion is something other than being correct in making inconsistent moral judgements regarding the same thing. That is conversion after a fashion, but not what one in pursuit of objectivity—or even moral improvement—seeks. One pursuing moral objectivity or improvement is pursuing not merely the possibility that at one time we might have attitudes and judgements that are inconsistent with those we might have at some other time. Such a one is searching for a difference in moral correctness. If change of mind can never bring error or the correction of error, then the introduction of the temporal dimension reintroduces the prospect of relativism for a single dynamic group over time, instead of over multiple groups.
The projectivist purchased the possibility of moral correctness with the coin of consensus. Of course, such consensus is rare, even on what any one of us takes to be the most blindingly obvious moral matters. Most typically, there are more than the few outliers who disagree with the majority—if there even is a majority judgement on a given issue—so, getting objectivity by consensus would be quite unlikely and cannot be the substance of the correctness of our typical current judgements. If there is objectivity in morals, consensus does not constitute it. The failure of consensus is merely a variation on the theme raised by the twin possibilities of the moral prophet and corporate conversion from error to correctness. If the vast majority can be morally defective while the smallest majority can be upright, there seems to be no theoretical barrier to many being wrong in many different ways. All acquainted with the range of contrary opinions regarding what it is morally best to do in complex circumstances and who nevertheless thinks that their own respective opinions are (likely to be) correct, cannot avail themselves of any account—projectivist or otherwise—that finds objectivity in consensus. To the degree to which anyone of the general Protagorean persuasion seeks to recover moral objectivity as we ordinarily understand it as obtaining even when there is great diversity of opinion, attitudes, and judgements must face the at least the occasional mismatch between consensus and objectivity. Any who claims that under conditions of diversity there is no moral objectivity will still face the consequence of the impossibility of conversion. Any transition from diversity to uniformity will not be any great awakening, it will be an exercise in corporate creation of the moral where before there had been nothing moral or immoral.
If moral correctness is a matter of neither consensus nor diversity, the Protagorean measurement thesis fails as a general account of moral objectivity. The modern versions, of course, take ‘consensus’ loosely to cover not so much uniformity of opinion, but uniformity of attitudes of (dis)approval and the like. No matter. What has been said applies to that uniformity as well. There are ways to link the desired objectivity of projectivism with the portions that it sometimes eschews, but those links will not provide any solution to the difficulties above.
One might think that projectivism links concerns about attitudes of (dis)approval, moral judgement, and moral epistemology with perfectly non-psychological features of the world as follows. What constitutes moral correctness is judging moral matters rightly. Judging moral matters rightly is approving and disapproving of things appropriately. Appropriateness is a matter of labelling ‘just’, for instance, all and only the just things. The just things are all and only those things that share some (relevant) characteristics independently of our attitudes of approval or our labelling them just. Precisely what those characteristics are may be opaque to us and that just things possess those characteristics may form no part of the meaning of ‘just’; nevertheless, those characteristics are what constitute justice. This seems to give all that anyone could want from a meta-ethical theory.
The linkage, however, works best when there is sufficient uniformity and stability of judgements. Making the redness of a rose and a fire engine and t-shirt a matter of non-psychological characteristics of roses, engines, and shirts that produce the right responses in observers seems plausible to the extent that observers do not vary widely in their responses to those objects and others of different colours. The application of this strategy to the moral case is problematic precisely because it is the moral domain that lacks the requisite invariance of response. We can conclude, then, that Protagoreanism, in both its ancient and modern formulations, will not deliver moral objectivity.