How to Derive "Ought" from "Is"

Back when I was in college and taking up philosophy, the received opinion concerning ethics claims, the standard doctrine espoused by all my teachers, was that, since Hume at least, we can all agree that one can't derive "ought" statements from "is" statements, that is claims about what we ought to do in any given case do not follow based on the descriptions of the facts of the case alone. Of course, this is moderated somewhat by the realization that some "is" statements present us with reasons to make "ought" claims to the extent that we are so inclined and that we believe others share the same inclinations that we do. Confronted with a fact that prompts us to choose X, for instance, we will naturally expect that someone else with values like ours will be susceptible to the same prompt and recognize the same reason to act as we do. To the extent moral assertions are built on that, it is possible to move in a seemingly logical way from what there is to what we ought to do about it. But the problem, particularly in the moral case, boils down to situations where the prompts themselves are in question.

If seeing someone in danger or in pain serves to prompt me to try to alleviate the conditions causing the other person pain or putting them in danger, it doesn't follow that that prompt will have the same effect on someone else. Nor does it follow that it should have that effect on me if it so happens that it doesn't. This is the problem of deriving oughts from is's. And it lies at the very heart of the moral case.

Since Hume this has been standard stuff in moral philosophy and has led thinkers interested in establishing some reasoned basis for moral judgment to look for ways to get around this apparent obstacle. In the early twentieth century it led to a subjectivist view of ethics which took the form of the logical postivitists' doctrine of emotivism (that statements about goodness amount to no more than expressions of our feelings or attitudes towards certain things, that they do not denote anything objective or in the world at all but only our own feelings, sensibilities, inclinations to favor, etc.). Of course, while this may give us reasons to do things, it doesn't support giving reasons to others or to ourselves when we are conflicted or torn between different inclinations. It doesn't provide a substantive support for a system of moral claiming.

R. M. Hare attempted to get round this by proposing that judgments of moral value played a special role in our language, i.e., a purely prescriptive one. The idea was that asserting a thing's goodness wasn't any more about finding some feature that's equivalent to what we mean by "good" in the thing nor is it just about expressing our feelings about or attitudes toward it a la emotivism. Rather, Hare suggested, it's about directing behavior. Hare moved from this to a variation of the Kantian principle of universalizability as a way of providing a basis for differentiating between prescriptive choices but, in the end, his approach faces the same problem Kant's did, namely what potency can such a claim have without a reason to think universalizability is itself important.

Kant had thought that as rational beings we must see the unavoidability of acting in a way that our acts could work if universalized and Hare finally appeared to hold the same though for Hare it was not as strong a claim as Kant thought it because Kant believed that one could not successfully turn off reason without losing one's rationality (and one could not do that, though one might deny the potency of rationality in which case one would be denying what one was). For Hare it was logical to act logically but it was finally only a matter of choosing to do so -- and one could choose otherwise.

Of course, the problem for moral valuing is to tell us what we should choose though, and to the extent we could do one thing instead of another, without reason to distinguish between them, the moral project seems to fail. Thus the is/ought problem continues to fester. In the latter part of the twentieth century some philosophers attempted to move back toward naturalizing notions of goodness as the proper solution. To the extent good could be equated with some thing or things in the natural world (rather than merely being a name for some "non-natural quality" as per G. E. Moore or for no quality in the world at all as per the emotivists and other subjectivists -- and even Hare, ultimately, seems to have been a kind of subjectivist), we can indeed expect to derive "ought" from "is", what we ought to do based on a description of some natural "good."

Anscombe and Foot, among others, made the case for equating the moral good with some natural good. In both cases they moved toward claims that what is good in the moral sense is what is good for us as the creatures we are, as human beings. Reaching back to the ancient Greek notion of eudaimonia, translated as human happiness or, as Duncan Richter and some others render it, human flourishing, the idea is that there are some things that are good just because they are a part of what it means to be human and that, to most fully realize our humanity, we must choose to do or achieve those things. For people like Anscombe and Foot, such things include the sorts of behaviors that sustain human social capabilities, that make it possible for us to live effectively in a human cultural environment.

Promise keeping, honesty, loyalty, compassion are all traits that advance human social cohesiveness which is an essential condition of living well in this world for human beings like ourselves. Therefore we are morally bound to pursue and secure such things and act in such ways that these traits, these virtues are manifested in ourselves and advanced (through recommendations and such) in others. In this way it is supposed that oughts can be derived from is's for the oughts really are just variant forms of the is's. It's a fact that we are humans and it's another fact that certain behavioral traits are human goods. Thus, the view goes, the oughts which serve as guides of behavior are derivable from the is's which state how things are.

The problem with this though is that there is still a gap. For why should anyone choose to act in a way that is good for humankind in general if, and to the extent that, such behavior doesn't serve that person in a way that he or she wants? While we may be able to agree that humans, in general, are better off in a successfully socialized group or polity (we can live longer, be more prosperous, more comfortable, more content), moral valuation is ultimately about what we, ourselves, should do -- and acting for the benefit of any group, whether in the abstract (humanity as a kind) or in the particular (this or that cultural cluster) may not be in the direct interests of the agent him or herself in any given case.

Why should the agent care about the good of the whole when he, himself (or herself), is also a stand-alone entity in the universe?

Can any of us actually be that? Well you can argue that, finally, we can't. But there is still a sense in which we are (we have our own sets of needs and interests and, finally, we die alone as it were) and no reason, in principle, that we shouldn't think of ourselves in that way in any given situation in any case.

How then can we surmount the obstacle of deriving oughts from is's? And if we can't, can we solve the moral question?