David Chalmers has been posting chapters of a draft manuscript entitled “Constructing the World”. The project has to do with the idea of “scrutability”: given some set of base truths and ideal reasoning, can all truths be known? He thinks so, and the work is mainly about fleshing out (in much detail) variations on this thesis.
Chalmer’s arguments engage the more technical side of analytic philosophy, which makes it more difficult reading for me. But, there is good stuff here to be sure (I thought the chapter which confronts Quine’s critique of the analytic/synthetic distinction was very valuable by itself).
In Chapter 6, Chalmers briefly discusses “hard cases” of truths which seem difficult to derive from a more limited set of truths: these include some mathematical truths, philosophical truths, and moral truths. The case of moral truths was on my mind given the recent discussion of Sam Harris’ remarks on developing a science of morality. Chalmers seems to think that if moral truths are real, then they should be scrutable from the set of non-moral truths.
What follows are some quotes from Chalmers (Ch.6, Sect.2). Please note his brief discussion on moral truths needs to be considered in the context of the much broader work; I’m just using this excerpt to further provoke my own thinking on the topics of moral realism/moral knowledge. Also note that for Chalmers, non-moral truths are not just physical truths, but would typically include phenomenal, or experiential truths as well.
“On the face of it, there are good grounds to hold that insofar as there are moral truths and they are knowable, then they are scrutable from non-moral truths.
Perhaps the best reason to deny that moral truths are a priori scrutable from non-moral truths arises from the possibility of moral disagreement even among ideal reasoners who agree on the non-moral truths. It is not obvious that this sort of disagreement is possible: perhaps apparent moral disagreement always involves empirical disagreement, or nonideal reasoning, or merely verbal disagreement. But it is also not obvious that this sort of disagreement is impossible. If this sort of disagreement is possible, then the truth of moral claims that are the object of disagreement will not be scrutable from non-moral truths.
However, if this sort of disagreement is possible, then it is natural to hold that there is no fact of the matter about who is correct. That is, one will then naturally embrace a form of moral anti-realism, according to which there are no moral truths. If so, there will be no inscrutable moral truths. Alternatively, one might embrace some form of moral relativism, so that moral utterances are adjudged true insofar as they are true according to an appropriate standard (that of a speaker, or an assessor); but then one can argue that according to that standard, the claim that the utterance in question is implied by nonmoral truths will also be adjudged correct. Either way, there is no trouble for scrutability here.
The meta-ethical view that is most obviously incompatible with scrutability of the moral from the nonmoral is a hardline form of moral realism on which there are moral truths that are not knowable even on full knowledge of nonmoral truths and ideal reflection. Such a view is unattractive, though. The best reason for being a moral realist stems precisely from our apparent knowledge of moral truths. If that knowledge is denied, moral anti-realism seems much the more natural option.”
My reading of these passages tends to reinforce my prior opinion: a consistent moral realism implies one can know moral truths and can in principle derive these from non-moral truths (I fully concede the practical difficulties may be extreme – Chalmers invokes ideal reasoning and full access to non-moral truths).
I don’t think naturalists who deny this in principle possibility can also be moral realists. I continue to think the problem here is that most subscribers to naturalism, inspired by the paradigm of physical science (which seeks to emulate objectivity), limit their conception of nature to non-experiential truths, and thus end up with no foundation in reality for human (and animal) intentions and values. I think naturalism as a worldview can and should expand to incorporate first-person truths, and this brings moral truths within reach. Perhaps some would argue this stretches the concept of “naturalism” too far, meaning we need a new label; nonetheless, there is plenty of room for such an expansion without embracing traditional supernaturalisms.