Monday, May 22, 2017

Evening Note for Monday, May 22

Thought for the Evening: Moral Testimony and Problems with the Asymmetry Thesis

It is often thought that there is an asymmetry between relying on testimony in nonmoral matters and relying on testimony in moral matters, so that you can get knowledge, or be reasonable relying on, the former, while there is something problematic about trying to get knowledge or relying on the latter. Pinning down what this asymmetry is supposed to be turns out to be less than straightforward, in part because in order to make it both clear and plausible everything has to be fine-tuned.

It's easy to see why this would be. In order for the asymmetry thesis to work, we must be comparing comparable kinds of testimony, where the moral testimony is taken under conditions that would not be problematic for nonmoral testimony. For instance, you can't state the thesis in a way in which we assume that your nonmoral testimony is taken from generally honest people and your moral testimony is taken from generally dishonest people, because accepting nonmoral testimony from generally dishonest people is also problematic, and the point is to find a comparison on which only accepting the moral testimony would be problematic. Likewise, people can accept testimony in all sorts of different ways, and this needs to be equalized on both sides. Likewise (and people in the literature are not always careful about this), you need to be comparing expert testimony with expert testimony and non-expert testimony with non-expert testimony, on both sides of the balance. Likewise (and this is actually somewhat difficult), you need a comparison where the stakes are not greatly imbalanced -- when the stakes are high, our standards of testimony go up. Part of the problem is that morality itself can raise the stakes, which means that you have to be quite careful to avoid comparing moral testimony on something that radically affects how one should live one's life with say, testimony about the statistical makeup of the population of Iowa. Reasonable caution in the latter case simply is not the same sort of thing as reasonable caution in the former case, and it has solely to do with the fact that the former will demand a lot more than mere acceptance.

This is a fairly serious set of problems, and one that I do not think proponents of the asymmetry thesis have really done the work to solve. But one of the more precise and careful ideas, based on the work of Sarah McGrath and discussed by Guy Fletcher, takes the asymmetry to be about deference in matters involving thin moral concepts leading directly to moral judgment that is problematic because it is without appropriate understanding.

(1) Deference is an interesting element here; our acceptance of testimony is always quite complicated, and we certainly can accept testimony nondeferentially -- for instance, we can accept it only because it agrees with what we already know. Deference requires that we accept something on authority, and this raises the question of how we are evaluating the authority to which we are deferring.

(2) The distinction between 'thin' and 'thick' moral concepts is somewhat misleading; 'thick' moral concepts are concepts that have both an evaluative and a nonevaluative component, whereas 'thin' are purely evaluative. In practice, though, the distinction is more complicated because it's not clear that there are any purely evaluative concepts, with no nonevaluative component at all, so people usually make the distinction relative, with thin moral concepts being "more purely evaluative". This, of course, makes the distinction next door to useless for the kind of work needed here. Suffice it to say that for our purposes it's really an issue of abstraction and fundamentalness, with thin concepts being more abstract and covering a fundamental portion of the moral field. So what people have in mind when they are talking about thin moral concepts are things like good, bad, right, wrong. This contrasts with what they are thinking about when they talk about thick moral concepts, which are things like generous, gracious, cruel, obscene. Thus the idea is that there's something problematic about deference to moral testimony on fundamental things like right and wrong, good or bad.

This brings us back to the equalization problem, since if we are dealing only with very general and fundamental things on the moral side, we can only compare testimony about them to testimony about very general and fundamental things on the nonmoral side. And what are the corresponding fundamental things? Presumably things like true, false, consistent, inconsistent, and (to avoid getting into the analogues of the 'thick' concepts) these concepts have to be functioning as such -- that is, if we are in a context in which dogs are relevant, the question at hand is not testimony about dogs, but testimony about what counts as truth in matters concerning dogs. If we really are operating at such a level, we are in realm where believing something out of deference to testimony is surely extraordinarily rare on either side. On the other hand, if we try to bring it down, we are getting increasingly less 'thin', and we have to do that on the moral side as well, to make the comparison legitimate.

(3) We could use moral testimony in multiple ways; what the asymmetry proponents want to argue is that we cannot use it to form moral judgments about good or bad, right or wrong. For instance, if you don't know if something is right, but defer to someone else's judgment that it is right. This we can leave as straightforward enough, although, if we are going to be serious about the 'thin moral concepts' part, we have to be careful that we don't accidentally turn our thin moral concepts into indirect ways of talking about thick moral concepts.

(4) If we're going to say that accepting moral testimony is more problematic than accepting nonmoral testimony, we have to have some notion of what 'more problematic' means. There are two ways that one can go here. Either (i) the moral case does not give you the right kind of understanding and the nonmoral does; or (ii) neither the moral case nor the nonmoral case give you understanding, but morality by its nature requires understanding in a way that other things don't. The latter seems to be favored; but obviously it would depend on what your account of morality was.


In reality, I think the asymmetry thesis fails totally, if we understand it in something like this way. Think of how testimony works. I tell you something, witnessing to its truth; on authority of my testimony, you come to believe it to be true. One way to characterize this is to say that I am giving you evidence about what you ought to believe, or at least treat as true, that I am putting it forward as right to believe, or at least good to believe. So:

(A) You are very skeptical of modern biology -- biologists talk so much about everything that it's sometimes difficult to take them entirely seriously. But, I insist, "At least some theories put forward by biologists are very credible." "Well, I couldn't say for myself," you say, "but you know more about it, so I guess I should indeed defer to you on this point and accept that some biological theories are very credible."

(B) You are very skeptical of modern biology -- biologists talk so much about everything that it's sometimes difficult to take them entirely seriously. But, I insist, "At least some theories put forward by biologists are probably true." "Well, I couldn't say for myself," you say, "but you know more about it, so I guess I should indeed defer to you on this point and accept that some biological theories are probably true."

(C) You are very skeptical of modern biology -- biologists talk so much about everything that it's sometimes difficult to take them entirely seriously. But, I insist, "Believing in some of the theories biologists put forward is good ." "Well, I couldn't say for myself," you say, "but you know more about it, so I guess I should indeed defer to you on this point and accept that belief in some of the biological theories is good."

In (A) I explicitly mention a 'thick' moral concept, credibility; in (B) I just mention probable truth; in (C) I explicitly mention a 'thin' moral concept. It is difficult to find cases in which we can't treat (A) and (B) as interchangeable, at least with commonly true assumptions. But there also doesn't seem to be any way in which (C) could possibly be problematic if (A) and (B) aren't; (C) seems interchangeable with (A), at least on commonly true assumptions. But it seems like you can run similar translations in any particular case.

It's possible that one can resolve these problems, but not, I think, at this level of precision. Proponents of the asymmetry thesis need a great deal more clarity than has yet been put forward. (My own view is that testimony is testimony; the cases are symmetrical.)

Various Links of Interest

* Helen Andrews has a sharply critical article on John Stuart Mill and Hariet Taylor (in part talking about Hayek's research into the subject). It seems right enough, although it should be noted that the acid tongues of the Carlyles are not always to be trusted.

* Julie D. notes a podcast currently running on Boethius's Consolation of Philosophy.

* Rebilius Crusoe, Francis William Newman's paraphrase of Robinson Crusoe into Latin. Francis William was the younger brother of John Henry; where J. H. went Catholic, F. W. went very liberal, eventually becoming a sort of religious agnostic. He was an enthusiast for languages, and was professor of Latin at University College, London. [I misstated his university earlier; I think I was muddling him up with someone else.]

* Manuel Vargas, If Free Will Doesn't Exist, Neither Does Water (PDF)

Currently Reading

Craig Williamson, The Complete Old English Poems
Vladimir Soloviev, The Justification of the Good
Satischandra Chatterjee, The Nyaya Theory of Knowledge
Roger Scruton, The Soul of the World
St. Romanos the Melodist, On the Life of Christ: Kontakia