A member of the studio audience wrote:
[Y]ou seem to share a lot of ground (indeed, all the ground except the "ineliminability" of certain things from moral language) with what I would regard as one of the standard moral anti-realist positions.
There is another thing I do not share.
I do not share the idea that "categorical" was ever a part of morality as practiced. There is nothing to eliminate. Morality was adopted and embraced as a technique for fulfilling desires. At some point some theorists came along and asserted that its principles are categorical, but that never made it into the meaning.
Desirism accurately describes morality as practiced. That makes it a realist theory.
Look at the elements that it can account for:
The central role that rewards (such as praise) and punishments (such as condemnation) play in the institution of morality. This is something that very few competing theories even address. Let us say that moral claims are categorical - how does this account for the practice of responding to virtue with praise and vice with condemnation?
It accounts for 'ought' implies 'can' because it only makes sense to apply praise and condemnation where it can cause the reward-learning system to effect a change in desires.
It accounts for the fact that moral claims act like truth-bearing propositions. They are truth-bearing propositions in that they make claims as to whether an act is indicative of desires that people have many and strong reasons to promote or inhibit through praise or inhibit through praise or condemnation.
It also accounts for the emotive component of moral utterances - because they often contain the praise or condemnation that the truth-bearing component says that people generally have reason to present.
It accounts for the types of evidence that people bring to moral debates - evidence supporting or refuting the truth of the proposition that people generally have reason to apply rewards such as praise or punishments such as condemnation in particular ways.
It accounts fit the three categories of moral claims - obligation, non-obligatory permission, and prohibition.
By the way, it also accounts for a fourth moral category - supererogatory action or acts above and beyond the call of duty. Some actions exhibit desires that people generally have many and strong reason to promote. However, praise and condemnation cannot be expected to bring about a desire of such strength in the public at large. We have reason to praise these people and call them heroes. But we recognize that most people can never acquire such a virtue. By virtue of 'ought' implies 'can' we do not hold them to be obligated to do so.
It fits moral claims into a general theory that can be applied to all value-laden terms. It says that all value-laden terms relate states of affairs to desires. They differ in the objects of evaluation, the desires that are relevant, the nature of the relationship (direct or indirect), and whether the relevant desires are fulfilled to thwarted.
Beauty. This term is applied to things seen and heard based on whether the experience of seeing or hearing directly fulfills the desires of the seer or hearer.
Illness and Injury: These terms evaluate changes or deviations in physical or mental functioning according to whether they tend to thwart (or give others reason to thwart) - directly or indirectly - the desires of the agent whose functioning is being examined. Furthermore, if the cause of the change is a macro cause that primitive people can see (getting trampled by a horse) it is an injury. If it is a micro cause (such as cancer or poisoning) it is an illness.
Useful: This term can refer to just about anything, but never in terms of its ability to fulfill desires directly. It is always used to identify the object of evaluation as something with the capacity to fulfill desires indirectly - in virtue of its ability to bring about something else that can fulfill desires directly.
Dangerous: This term is also used to evaluate just about anything according to its potential to thwart desires indirectly.
Virtue: This term is applied to malleable desires - desires that can be learned through triggering the reward-laerning system. A virtue tends to fulfill other desires - giving others reason to use rewards (such as praise) or punishments (such as condemnation) to facilitate the learning of that desire.
There is no categorical theory that can come close to accounting for so much of the actual use of value-laden terms in general, and moral terms in specific, such that it makes sense to claim that a "categorical" component is built into the meaning of these terms. Not only can this element be eliminated, it never existed to start with.
"Anti-realism," to most people, means the loss of moral restraint. It means passion unconstrained by the effects of praise and condemnation so that everybody does what they please whenever they please no matter what they please. The fact that this is what "anti-realism" with respect to morality means to most people tells us something about what "morality" means to most people. It tells us what they take to be "real" by telling us what they think anti-realism says is not real.
What is real is the institution of using rewards (such as praise) and punishment (such as condemnation) to promote desires that tend to fulfill other desires, while inhibiting desires that tend to thwart other desires. The "categorical" nature of morality is a mistake. "Categorical" was never a part of the meaning of moral terms. This is just a theory that some people adopted - a mistake, to be discarded.
If a non-categorical theory does a better job of accounting for the elements of morality (and can fit it into a broader theory that also handles a wide variety of non-moral value-laden terms - and can fit the theory in with what is known about biology; specifically, the reward-learning system and the effects of desires on choosing actions), then I am more than comfortable with saying that this claim of "categorical" values was never there to be eliminated.
Having said this, I do not think that the meanings of moral terms are worth debating. I have little interest in convincing somebody who holds that moral terms contain some claim about categorical value that cannot be eliminated that they are wrong. If they are right, this means that all of their moral claims are false and irrelevant anyway. The debate over whether desirism is the best account of morality as practiced, or the next-best alternative to an account that renders all moral claims false and irrelevant - is only of passive interest.