The View from Within: Normativity and the Limits of Self-Criticism

Part 1. Through Thick and Thin

PART I Through Thick and Thin Introduction The special character of moral diversity and moral disagreement has long been considered the basis for a normative irrealism that has led several writers to deny in general that normative discourse is factual. In their opinion, our diverse normative commitments are to be explained by appealing to noncognitive mental states such as desires rather than to belief in independent moral facts. And since desires, unlike beliefs, do not (of themselves) track truth or propriety,1 disagreement, deliberation, and change of commitment in the normative realm are not expected to yield to rational consideration. From such a perspective, our commitment to the diversity of normative outlooks would seem quite at odds with the very idea of rational normative revision that animates this study. The two chapters comprising Part I propose a less crude notion of normative irrealism with regard to which the possibility of rational normative revision remains at least an open question. But first a word about the opposition. Such a combination of antirealism (regarding the factuality of moral properties) and noncognitivism (regarding the source of moral commitments ) as described above is typical, for example, of Gilbert Harman’s approach.2 Harman’s version of ethical relativism comprises three main claims:3 1. If a person fails to comply by a moral demand D that applies to him, it will be due either to ignorance of the relevant facts, to a failure to reason something through, or to some mental defect or disorder like irrationality , stupidity, confusion, or illness. 2. Nonetheless, not all are necessarily subject to D. For it is possible that for two people A and B (both subject to some moral demands) D applies only to A but does not to B, because B does not accept D. 3. There exists no more basic moral principle D* such that D* applies to A and B and explains why D applies to A but not to B. The difficulty seems straightforward: if A is subject to D and B is not, there must be something true of A but not of B that accounts for the difference. There should be a way of arguing that due to certain features lacking in B, A is subject to D while B is not. A description of these 32 Through Thick and Thin features along with their moral implications would constitute the moral principle D* that applies to A and B and explains their different relationships to D. To this Harman replies as follows: 4. A but not B may be subject to D because it is rational for A to accept D, but not rational for B to do so, and this is sometimes . . . due simply to the fact that A and B start out accepting sufficiently different moral demands in the first place.4 Harman’s use of “rational” notwithstanding, his answer calls for further elaboration. How do A and B view each other in this respect? Is it not the case that if A regards D as morally compelling, he must also believe that B should accept D as well, otherwise he is liable not to conceive his own subjection to D as moral. But if Harman’s relativism is valid, A must be wrong. The source of his mistake, suggests Harman,5 is that he conceives his acceptance of D as being due to moral facts; he believes that certain modes of conduct are good and others bad as a matter of fact. But this is not so. According to Harman, there is no such thing as a moral property or a moral fact. In insisting that D applies to B, A fails to realize that the source of his own commitments is noncognitive.Accepting D is the right thing to do for A, not by virtue of objective facts, but due to his particular set of desires. For B, who lacks this set of desires, it is not. It is easy to see how such a combination of noncognitivism and antirealism applies to all manner of normativity. For the sake of brevity, we shall call this position “standard normative irrealism.” Many find this combination of premises too high a price to pay. We agree. But is there a way of being more selective without surrendering Normative Diversity altogether? To take normative diversity seriously obviously requires premising some form of normative antirealism.On the other hand, the source...