Epistemological arguments and the Dependence Thesis

Linville has been arguing that AEN provides an epistemological argument for moral skepticism, to show that our moral beliefs lack warrant because the mechanisms responsible for our moral beliefs appear to be fitness-aimed, rather than truth-aimed. If our best theory of why people believe P doesn’t require that P is true, then we lack good grounds to believe P is true. This much resembles an argument by Gilbert Harman.

Harman’s so-called “problem with ethics” is that moral facts, if such there are, appear to be explanatorily irrelevant in a way that natural facts are not. According to Harman, we need not suppose that over and above such natural facts about Hitler as his monomania and anti-Semitism there is a moral fact of Hitler’s depravity. Nor must we appeal to his actual depravity in order to explain our belief that he was depraved. Harman may thus be viewed as arguing in his own manner that we have no reason to believe that the best explanation for our moral beliefs involves their truth. We have no good reason to think that the causes of those beliefs are dependent on whatever would make them true.

Sturgeon has replied first by noting that moral facts are commonly and plausibly thought to have explanatory relevance. Both Hitler’s behavior and our belief that he was depraved are handily explained by his actual depravity, and this is in fact the default explanation. Sturgeon follows the method of reflective equilibrium, a method employed in both science and ethics, which begins with certain considered judgments, and with the assumption that our theories, scientific and otherwise, are roughly correct, then moves dialectically in this way between plausible general theses and plausible views about cases, seeking a reflective equilibrium. Sturgeon notes that, whereas he allows for the inclusion of moral beliefs among the initial set, Harman does not. But he argues there’s no non-question-begging justification for singling out moral beliefs as unwelcome in the initial set while allowing those of a scientific or commonsense nature.

Sturgeon’s approach invokes the supervenience of moral properties on natural properties. On standard accounts, if some moral property M supervenes on some natural property (or, more likely, some set of natural properties) N, then it is impossible for N to be instantiated unless M is also instantiated. In all worlds in which Hitler believes and acts as he did, his depravity would supervene on such properties and be instantiated; he couldn’t have had those properties without being depraved. Harman, by denying this, tacitly assumes there are no moral facts or properties, which is of course the point at issue.

Sturgeon’s appeal to reflective equilibrium is crucial in his reply to Harman. Brink goes to some length to argue that Harman fails to demonstrate any explanatory disanalogy between the scientific and moral cases. Linville finds Sturgeon’s reply successful. Sorley once said the true beginning of metaphysics lies in ethics. He thought that holding off on ethics until the task of worldview construction was complete would result in an artificially truncated worldview, and that moral ideas would be given short shrift. The exclusion of moral experience seemed arbitrary. Harman seems to be following in the tradition Sorley criticized. Harman’s results are achieved only by begging the question against the moral realist.

But even Sorley would in principle admit that the initial “ethical data” must prove to be compatible with everything else that is included in our final interpretation of reality. In fact, the same year Sorley delivered the Gifford Lectures, George Santayana published Winds of Doctrine, in which he complained that Bertrand Russell’s then-held moral realism was the result of Russell’s “monocular” vision. Santayana said Russell didn’t look and see that our moral bias is conditioned and has its basis in the physical order of things. Eventually Russell abandoned his moral realism, crediting these very arguments. AEN suggests following Santayana’s advice, and bearing in mind Sharon Street’s worry: “If the fund of exhaustive judgments with which human reflection was thoroughly contaminated with illegitimate influence…then the tools of rational reflection were equally contaminated, for the latter is always just a subset of the former.” What we require is some assurance that our original fund is not contaminated. So, what reason have we for supposing that the mechanisms responsible for those judgments are truth-aimed, that the Dependence Thesis is true?

Santayana suggested that if God exists and has fashioned the human constitution with the purpose of discerning moral truth, then we have reason to embrace the Dependence Thesis. But neither Russell nor Santayana was a theist. Moral realists need to give an account of moral beliefs that would lead us to suppose that they are reliable indicators of truth. Quine offers such a story with a Darwinian spin to inspire confidence in our ability to acquire knowledge of the world around us. Natural selection is unkind to those whose behaviors stem from either false beliefs or profound stupidity. We should expect our cognitive faculties to be truth-aimed and generally reliable given such selection pressures.

Plantinga has challenged such stories with what he calls “Darwin’s Doubt.” The connection between fitness-conferring behavior and true belief might not be so certain as Quine suggests. If Plantinga is correct, then evolutionary naturalism is saddled with a far-ranging skepticism that takes in much more than our moral beliefs. Despite Plantinga’s many ingenious examples in which adaptive behavior results from false beliefs, many people just find the link between true belief and adaptive behavior plausible. And in any event the moral and nonmoral cases appear to be significantly different.

The core of Street’s paper is her “Darwinian Dilemma” she poses to value realists like Sturgeon. Our moral beliefs are fitness-aimed. Are they also truth-aimed? Either there is a fitness-truth relation or there is not. If not, and evolution has shaped our basic evaluative attitudes, moral skepticism is in order. If there is a relation, then it is either that moral beliefs have reproductive fitness because they are true (the “tracking” relation), or we have the moral beliefs that we have simply because of the fitness that they have conferred (the “adaptive” link account). Adaptive link leads to constructivism. The moral realist needs a tracking account, but Street thinks fitness following mind-independent moral truths is implausible. A tracking account of paternal instincts would have to say more than that the behavior tends toward DNA preservation—something like the instincts were favored because it’s independently true that parents ought to care for their offspring. Nonnaturalists have the worst deal in light of the causal inertness of moral properties on their view. Ethical naturalists have a better time at it, but why not just eschew realism and go with an adaptive account?

A dilemma similar to that urged by Street comes from another consideration of Darwinian counterfactuals. Sturgeon thinks moral terms rigidly designate natural properties. If justice picks out some natural property or properties, we might expect an ethical naturalist to conclude that moral judgments if true are true in all possible worlds. But Linville writes that to insist that our moral terms rigidly designate specific earthly natural properties to which human sentiments have come to be attached appears to be an instance of what Judith Thomason has called metaphysical imperialism.

Sturgeon dialogued with Gibbard, who argued for expressivism. Sturgeon’s reply is that perhaps our ancestors called bargaining outcomes just because they really were. But is this so? The bargaining situation Gibbard had envisioned involved a cast of characters who were self-interested individualists. In such a situation, there was pressure in the direction of equitable arrangements. But imagine a different set of initial conditions—like lupine bargainers. If justice supervenes on certain natural facts, these will essentially include facts about the psychological constitution of the respective bargainers. It seems to Linville that the most plausible explanation is that such counterfactual moral beliefs are formed as the result of selection pressures that are themselves in place due to the contingencies of the evolutionary landscape—contingencies that are morally indifferent. While ethical naturalists in those worlds no doubt argue for the supervenience of the moral on the natural, the efficacy of moral explanations, and the existence of corresponding moral facts, we should, Linville thinks, regard them as mistaken. If the moral beliefs of the actual world have also taken their cue from predispositions that were fitness-conferring, then it is hard to see why our own ethical naturalists are in any better position so to argue.

Photo: "Darwin Divergence" by Jwyg. CC License.