On the other hand the thesis merely that physicalism is compatible with objective values is not very exciting or even very new. More revealing would be a physicalistically acceptable argument that there are objective values in the first place. §§6.1-6.3 provide such an argument, by explaining how moral truth is determined by purely descriptive truth, whatever the moral truths happen to be and whatever the best way of discovering and justifying them. The existence of objective values is a matter not of extra entities but of there being a truth of the matter as regards the correctness or incorrectness of our value judgments, a truth of the matter determined by objective, natural fact. If the physicalist is right that natural fact in turn is determined by physical fact, it follows that the correctness of our value judgments is determined ultimately by truths at the level of physics. Those physicalists are wrong who insist that "no...scientific facts about the world can by themselves determine what we should do," or who believe that objective values are incompatible with persons as emergent beings.

But is there such a thing as purely descriptive truth in the first place? If not, it could hardly determine value; the fact-value distinction would be doomed. Yet we presuppose some such distinction merely in asking whether purely descriptive truth determines moral or other normative truth. Of course the fact-value distinction is doomed in any case, if it means there is a domain of discourse such that our choice or continued use of it involves no value presuppositions whatever. In that sense there is no purely descriptive or factual discourse. Not only our decision to use, but our acquiescence in continuing to use a discourse, however objective the discourse is, reflects or presupposes as well as reinforces values we may scarcely perceive. Often the values are revealed only in the domain's entrenched metaphors. The discourse comes trailing clouds not just of etymology but of value, purpose, emphasis and mood.

And yet in another sense we can distinguish between descriptive and valuative discourse, just as surely as we can distinguish between what the House of Representatives does and what it ought to do. Without some such distinction we would be defenseless against cynical remarks like "An impeachable offense is whatever the House says it is" (as then Representative Gerald Ford said with regard to Earl Warren). What the House ought to do cannot be reduced to or expressed as what it does. And this distinction, between what descriptively is the case and what would be better (or worse), gives every indication of being vital to civilized life. So in this sense there is such a thing as purely descriptive truth, even though there is another sense in which no discourse, including the pur-est description, is value-free. For our decision to use one discourse rather than another reflects and reinforces certain values, even if strictly speaking the discourse, via its distinctive vocabulary, expresses none.

Modern science chose to use descriptive, objective discourse (not overnight, but the essential commitments seem to have been made by the end of the 17th century). This made it look as though the world could contain no entelechy, no telos, no element of value whatever. Values were banished from nature, along with much else that matters to us. All efforts to get the values back in have seemed unconvincing. Indeed no such effort can possibly succeed, assuming the irreducibility of values to facts, so long as (i) the world is identified with the totality of fact, and (ii) the only sense in which the world could contain or exhibit an element of value is via derivability from or reducibility to the factual.

The flaws in this history are increasingly easy to see. One consists in construing the chosen factual discourse as expressing the way the world is, via identifying the world with what the discourse can describe or express. This is just another form of monopoly, rejected in §4.6. Another flaw consists in supposing there is no alternative to reduction and derivation as ways of showing the world contains an element of value after all, or that ought is entirely a function of is. That there is a plausible alternative is one of the lessons of the next three sections.

Of course if ought could be reduced to is, then ought would be determined by is. For in general reduction is sufficient for determination. Again reduction means a term-term relation whereby ought would be coextensive (of necessity or at least accidentally) with some construction on is, and in that sense definable by is. But determination means that a distribution of truth and falsity over whole sentences about what is the case would allow one and only one such distribution over whole sentences about what ought to be the case. In particular, given the purely descriptive or natural facts about the world (including us), one and only one account of our obligations would be correct. If in turn the descriptive facts are determined by facts at the level of natural science, as the physicalist believes, then ought would itself be determined by facts at that level. Nature and morals would be ontologically reunified, the natural phenomena determining the moral, despite their bifurcation since the 17th century, roughly when teleological concepts had begun to be eliminated from natural science.

Notoriously, no reduction of ought to is has achieved consensus, to put it mildly, and none is likely to. Anti-reductivist arguments in ethics evidently are too powerful, if not conclusive. But even though reduction is sufficient for determination, it is not necessary, as we have seen. Further, moral oughts, at least, are indeed determined by is. The argument for this determinacy, outlined in the next section, is nonreductive and surprisingly simple, granted just two premises, one of them unproblematic, the other less so. In any case the argument applies no matter what the best way of justifying or defending moral claims might be, and no matter which such claims are correct. And it applies to moral first principles, not just to particular judgments made in light of them.

Nor does the argument presuppose that ought is either derivable from or entailed or implied by is. The correctness of moral claims can be grounded in non-valuational, non-institutional brute fact, in the sense of being determined by it, even if no ought or imperative or prescriptive can be deduced from or defined by any is or indicative. The same then proves true of the correctness of all value judgments (subject to a minor proviso). To the determinacy of moral valuation we may therefore provisionally add the determinacy of valuation generally.

The argument for the determinacy of valuation is outlined in §6.1. Some caveats about the argument, especially about one of its premises, appear in §6.2. The main caveat is that the argument as it stands in §6.1 is meant neither as a non-question-begging proof of objectivism, nor as a non-question-begging refutation of subjectivism or anti-realism in morals. Instead the argument there is meant only to undermine what has always been the final, indispensable argument for subjectivism. This is what Mackie calls "the argument from queerness": objective values would have to be very queer sorts of things, because their relation with facts is so mysterious, being a matter neither of entailment or implication, nor of derivation, nor of reduction or definability, nor even of supervenience or "resultance" (the latter two relations being either unclear or, when clear, not strong enough to capture the relation required by objectivists). Better by far to replace the alleged objective values with some sort of subjective response that can be causally related to stimulation by the natural features on which the alleged values are said to be resultant or consequential. Or so the argument goes. As against this what §6.1 shows is (only) that there is an intelligible, unqueer alternative to the listed relations. The assumption that these are the only alternatives is false.

Mackie himself explains why the argument from queerness is indispensable. Without some such argument the mere fact of widespread moral disagreement does not by itself imply there are no objective values about which to disagree, no more than disagreement in science implies there really is no truth of the matter there. Or as Nozick remarks, "It is because we do not see how an objective ethics is possible that we worry about irresolvable moral disagreements." And as Sturgeon argues, anyone who finds it plausible that objective values play no explanatory role in our account of the world must already have been persuaded by some other consideration that there are in the first place no objective values to play the role. Indeed if anything, the presumption should be that there are objective values. For this is what ordinary usage of moral terms overwhelmingly presupposes, as does most of our actual moral reasoning, including those of our explanations that appeal to moral properties (as in `Mother Theresa's goodness won her a Nobel Prize', `Hitler's depravity won him universal condemnation', and so on). Hence as Mackie sees, subjectivists like himself are compelled to advance an error thesis: our ordinary usage and reasoning, entrenched for millenia, are massively in error, for there really are no objective values. And Mackie is far more candid than most subjectivists in acknowledging that the burden of proof is on those who advance any such thesis. The argument from queerness is meant to discharge this burden.