The definition of sound morality cannot be philosophically straightforward or, prephilosophically, more than the recital of honest prejudice. Answers in accord with the first form of definition cannot be pertinent if they don’t fit, perspicuously, a significant run of convictions of the second kind, and no reasonably large collection of answers of the second kind can possibly be sufficiently consistent or coherent or convergent to justify any confirmably objective proposals of the first kind. Definition cannot but be equilibrative and never entirely free of partisan conviction. I call any such effort “second-best,” that is, the best we can do -insuperably disputatious.
Efforts of the second kind I call sittlich, meaning by that no more than the anthropologically-minded collection of the apparent, operative moralities of different societies, including their own disputes about the normative standing of their own moralities. I cannot imagine that the radical Taliban and conventional American Christians could possibly be reconciled, though I see no reason why, at the present time or in the near future, American and Taliban negotiators could not find common ground for a not insignificant modus vivendi that each side could validate along its own moral lines without agreeing to the same moral principles - instantiated by their own least disputed specimens.
This may already seem an unnecessarily complex way of beginning. Let me explain. Most of the classics of moral philosophy - not all, it should be said -suppose that
moral norms and values can be read off from an analysis of human nature or from the right application of a faculty made suitably apt (somehow) for the task (Plato’s functional model of man, in the Republic, say, or Kant’s conception of the power of practical reason to legislate suitable rules for one’s own conduct). I believe there will always be an argumentative gap in any such account that cannot be closed by any would - be sensibility or intuition or cognitive power or form of inference regarding normative values, such that the supposed conditions of true morality can be shown, by its exercise, to be indeed true or valid. I think there is a vast difference between theories of these sorts and others like Aristotle’s account (for the most part) in the Nicomachean Ethics or Hume’s account in which he expresses his easy conformity with what he takes to be the general convergence of civilized sensibilities.
I don’t deny that, here and there, Aristotle and Hume may also be rightly viewed as allies of Plato and Kant in the regard already mentioned. But they do acknowledge, one way or another, that they simply favor the dispositions of the societies they admire and belong to (Aristotle, living among the Athenians; Hume, at home in London and Paris). In any case, I take it that ethics is not a science of the normative in any sense at all; also, conformity with custom and conviction is not, in itself, evidence of having captured moral truths of any substantive sort. I’m persuaded that the idea of a moral science is an oxymoron and that mere conformity with any sittlich morality falls far short of the requirement of a proper moral science. There is none, of course. Conditionally, then, the conceptual gap cannot be closed.
Now, I believe there is a very powerful premise, largely neglected, that secures all that I’ve already said about the prospects of a moral science of any kind and that, accordingly, explains the import of the difference between a philosophical or scientific and a sittlich morality. For, as far as the evidence goes, there is no known society that does not possess a code of conduct that would be recognized as a moral code in at least the sittlich sense, no matter how alien or unacceptable we might find it (actual cannibalism among the Inca, say): that is, a code answering, in first-order terms, to what is deemed right or wrong, forbidden or obligatory or permissible, in the way of conduct within the society affected. It’s a problematic but ineliminable premise, because it sets an empirically confirmable condition that, without possessing any moral force of its own, utterly precludes the possibility of ever discovering the true morality of mankind. It’s simply this: that, on post-Darwinian or paleoanthropological grounds, the functionally unique powers that the acquisition of a true language makes possible (exclusively among humans), and only thus makes possible, the original ability to conceive of moral questions, to reflect on moral choices, and to choose to commit ourselves conformably.
In short, on the pertinent argument, the human self or person (as distinct, though also inseparable, from the members of Homosapiens) are (hybrid) artifactual transforms of the species’ primate members. But if so, then morality, like language and the arts, is also artifactually constituted within the living history of human societies -in the sittlich way; furthermore, there is more than enough evidence to confirm the immensely diverse and endlessly changing forms of the sittlich (first-order) moralities of the human race.
I take the human self to be indissolubly incarnate, individually, in each of the biological creatures we call human beings, a functionally emergent competence that does not otherwise appear among living creatures: that is, otherwise than by internalizing the mastery of a true language and the cultural competences that that makes possible. The question of incipient selves among the higher (non-human) mammals remains intriguing but unresolved: I have no prejudice against the idea. Indeed, perfectly reasonably questions arise regarding gifted bonobos and elephants; but whatever incipience may attract us here still seems too primitive to justify regarding them as moral agents in their own world or ours. (I may be wrong.) Hence, it follows (from the paleontological evidence) that philosophically defensible moralities cannot but be constructed artifacts that, not unlike the hybrid artifacts we ourselves are, we have (somehow) found it reasonable to fashion and alter over time.
The question still remains: If a would-be objective morality cannot be confirmed as true for the whole or any part of humanity, then in what sense can it be reasonably endorsed as a “second-best” morality? You realize that its champions would wish to show that, at the very least, it compared favorably with any presumptively scientific or philosophically valid (or “best”) morality (for instance, utilitarianism or virtue ethics or Kantian rational autonomy or any religiously revealed code of conduct deemed to be rationally defensible or the disclosures of moral intuition or anything of the kind). But in what way?
I can think of three second-best strategies which, on their face, yield moralities at least as reasonable as anything advanced in the name of privileged access - without pretending to overcome the conceptual gap (already mentioned) in any illicit way. First of all, if the post-Darwinian argument is conceded, then there cannot be any natural norms of the moral sort directly confirmable from details regarding human nature (at whatever emergent level functional selves first appear - a fortiori, at the biological level). But if so, then the standing of would-be moral injunctions regarding matters like abortion, homosexuality, divorce, the use of drugs, the care of children and the like cannot be read off human nature.
This is not to claim that there are no reasonable constraints to impose on second-best moralities: there are, but promoting them does not oblige us to suppose that our proposals are not themselves constructions within our own artifactual world - and that they must earn their way there. Secondly, every society that lives by a generally accepted sittlich code is bound to have some conception -partially formed by the appraisal of operatively comparable proposals among other societies (in addition to its own) - to regarding “the greatest evils” to be shunned and “the least goods” to be secured, consistent with the first constraint: these will be valid in second-best terms, if any proposals are, almost without regard to how they vary from society to society.
We are bound to appraise them in terms of the viability of actual societies in accord with material conditions, historical orientation, technological capacities, sheer imagination and self-discipline: for instance, regarding the care of children, personal security, adequate nourishment, medical provisions, education, and the like. And, thirdly, provisions of the second sort (just mentioned) will require a reasoned grasp of the social organization of responsibilities and services that answer to what we may call our prudential needs, which are not moral values themselves but very general motivational concerns that cannot be entirely absent from our sittlich life or from whatever we envision as a candidate second-best morality. Our prudential needs do not entail moral goods. But it is implausible, given our original premises, to suppose that an objective second-best morality will not go some distance along the lines fen joining the fulfillment (in some realistic measure) of selected prudential needs cast in terms (say) of our second sort of constraint. But to concede this sort of defense is to grasp its unsuspected resources without violating morality’s artifactual nature.
An objective morality, then, is a second-best morality that satisfies the constraints of the constructivist sort just sampled or of others of a similar kind But it does so under the shadow of being profoundly self-deceptive.
-  2 Joseph Zalman Margolis — Temple University, Philadelphia — radical historicist,he has published many books critical of the central assumptions of Westernphilosophy, and has elaborated a robust form of relativism.