Anti-Skepticism Against Evolutionary Debunking of Ethics

Originally written for PHIL 1640 The Nature of Morality with Prof. Jamie Dreier.

Sharon Street’s paper “A Darwinian Dilemma for Realist Theories of Value” presents a case against moral realism based on the fact that our moral attitudes, like all other human features, are byproducts of our evolutionary history. The dilemma turns on the question of whether moral knowledge is adaptive, that is, if evolution naturally selects for individuals whose moral intuitions more closely resemble the person-independent truth that moral realists claim exists. If a realist claims that moral knowledge is not adaptive, then evolution has acted as a purely distorting force on the content of our moral beliefs, while if she claims it is adaptive, then she is making a scientific claim about evolution (specifically moral psychology) that Street argues is not supported by biological evidence. This paper will focus on the first horn of the dilemma and, using arguments inspired by Katia Vavova’s “Debunking Evolutionary Debunking,” argue that moral realists need not affirm a connection between evolution and moral truth to defend against skepticism. We will attempt to forge a connection between metaethical skepticism and other forms of skepticism, and adapt classical anti-skeptic lines to the case of Street’s argument in particular.

Street compares the position a realist takes in the first horn of the dilemma, in which no connection is positively affirmed between evolutionary pressures and moral truth, as “analogous to setting out for Bermuda and letting the course of your boat be determined by the wind and tides.” In other words, if we make no positive claim that evolution directs our moral faculties towards the truth, then we have no reason to believe that they end up anywhere near the truth, since we must accept that evolution did indeed direct them. Street believes this account so implausible, in fact, that she spends the vast majority of her paper defending against the second horn of the dilemma.

According to Street, a Darwinian line of attack against non-moral knowledge does not produce the same conclusions, because the second horn of the dilemma is much more open. For example, Street believes it is much more obvious that being able to accurately see the world is adaptive, since being able to visually identify dangers using one’s sense of sight is an evolutionary advantage, and thus we have no Darwinian reason to be skeptical about visual sense perception data.

What would it look like, however, if we did try to carry out a Darwinian argument against the trustworthiness of sight all the way through? A “visual realist” might try to embrace the second horn of the dilemma and claim that evolution has directed our sense of sight. However, how could she present any evidence of such a claim without reference to any visual perceptions? Suppose we, at the same time, launch a Darwinian attack against the other senses so as to block a move that compares visual perceptions against, e.g., auditory perceptions. Then, a realist has no perceptual beliefs at all that we are allowing them to use as part of her argument, since they are all simultaneously being called into question. The Darwinian skeptic is challenging so many beliefs at once that no beliefs remain “untainted” with which a realist might reply.

Robust philosophical skepticism of any sort is usually extremely hard to beat back on its own ground. Instead, we must try to tie this metaethical skepticism to a more pernicious form of skepticism, and thereby force the “debunker” to either give up her line of attack on moral realism or accept the additional baggage of, for example, Humean inductive skepticism or perceptual skepticism about the senses. Keeping this goal in mind, we can use ideas from Vavova to attack Street’s argument.

Vavova distinguishes a skeptical argument that claims that we have “no good reason to believe” some attitude (“no good” skepticism) from one which gives us a “good reason to doubt” that attitude (“good” skepticism), where a “good” reason is one independent from the skeptical line. As shown in the example of the visual skeptic, simply calling into doubt so many fundamental beliefs that the realist does not have enough remaining “good” reasons with which to respond is not a satisfactory form of skepticism, since it leads to perceptual and inductive skepticism as well. In fact, the traditional lines of Humean skepticism about induction resemble “no good” moral skepticism very strongly. Regarding an inductive belief such as “the sun will rise tomorrow” a skeptic can simply call into doubt all principles regarding induction at once, and one is left with no possible “good reason” for believing in tomorrow’s sunrise, since such a reason will by necessity involve some sort of inductive reasoning.

Thus, the skeptic must themselves present a good reason for believing that our current attitude regarding morality is mistaken. Recall that “good” here refers to sufficiently independent from the line of attack. However, the line of attack includes all of our methods that we could previously use to judge moral attitudes, and so such a “good” reason is impossible to provide. To clarify, consider what such a good reason to doubt our moral attitudes might look like. It would need to argue that the direction of evolutionary pressures is different from the independent truth of evaluative propositions (should it exist.) But how could such an argument possibly be constructed if we are suspending all beliefs regarding evaluative truth? That is, if all our moral attitudes are suspect, then we no longer even have a way of discussing moral truths. Thus, the skeptic here is susceptible to the same problem that befell the realist in our discussion of “no good” style arguments: if we call an entire body of beliefs regarding some subject (including beliefs about how we might accommodate new beliefs on that subject) into question at once then we are left insufficiently equipped to discuss the subject at all. This move from “no good” style skepticism to “good” style skepticism thus places the burden onto the skeptic in this stripped-down environment, allowing us to not accept the force of the “debunking.”

Although Street presents her argument as a “debunking,” it is important to note the parallels with classic skeptical arguments. How is a Humean skeptical argument against induction or a solipsistic Brain-in-a-Vat style argument structured? In each case, the skeptic takes some large class of beliefs (scientific in the former case and perceptual in the latter) that we usually take for granted and calls them all into question simultaneously. Then, the realist is left without any recourse, since all arguments attempting to justify induction or perception will themselves have some inductive or perceptual character. An even more dramatic example is a rational skeptic, who asks a question like “why must I be rational?” If the usual state of affairs in which beliefs and actions are based on reason rather than unreason is itself called into question, then what could a response to the skeptic even possibly look like? If the rational realist is not permitted to use reason in her response, then no response is possible, since philosophical arguments are by nature rational. That is, there is no way to disabuse a rational skeptic of her skepticism on her own terms, since those terms do not include reason. This non-response as response to skepticism is particularly compelling in this case, but such a move exists for inductive and perceptual skepticism as well. The most well known idea of this sort in perceptual skepticism is G. E. Moore’s “Here is one hand” argument, in which skepticism is refuted from the commonsense proposition that, after putting a hand in front of my face, my hand is in front of my face (as “my hand is in front of my face” is inconsistent with a brain-in-a-vat style skepticism.) The inductive analog of this style of response is the most promising for adaptation to metaethical skepticism. Namely, an inductive skeptic cannot square their purported skepticism with the way they actually act. Sure, they can launch an evolutionary attack on our inductive tendencies, or else simply directly call into question the logical bedrock of induction, but they are by their own nature forced to act inductively.

To clarify, suppose we have a glass of bleach and a glass of juice, and are forced to select one to drink. Since the only arguments for the position that bleach is dangerous and juice is not are rooted in an inductive generalization from past instances of bleach and juice, a genuine inductive skeptic should resist this conclusion, and instead be agnostic about this choice. We could then employ a sort of Moorean argument that goes: “then drink the bleach.” In other words, we have made explicit the conversion from second-order inductive skepticism to first-order skepticism about individual inductive conclusions. This move can be applied to error-theoretic metaethical skepticism—and, as we shall see momentarily, to the particular position advanced by Street. Suppose some metaethical skeptic (an error theorist) affirms a position like “there is no such thing as moral truth” or “there are no moral properties.” Then, suppose we put her in a position in which she is forced to choose between killing and not killing. She is forced to be agnostic about such a decision because endorsing the idea that “killing is wrong” would be a tacit admission of the existence of moral truth or moral properties. Then, our argument could go: “then, commit murder.” The metaethical skeptic has one recourse available to them over the inductive skeptic. They may declare something along the lines of “I would prefer not to, because of my evaluative attitudes, but I do not endorse the wrongness of doing so.” This avenue can be blocked by considering a third party faced with the dilemma. Then, a metaethical skeptic cannot resort to essentially declaring an opinion.

Does this argument require a particularly thoroughgoing skeptic as an adversary? Although Street tries to present her view as non-skeptical, it seems that she can be forced to make similar first-order claims. Street says that “antirealist views understand evaluative facts or truths to be a function of our evaluative attitudes… [which] leaves room for the possibility of evaluative error.” How, then, could we possibly have any confidence in a first-order ethical proposition? For example, suppose that P is “murder is wrong.” Street can defend P in the terms of her version of “evaluative truth,” but this amounts to nothing more than some particular way of according P with her, or perhaps the collective’s, evaluative dispositions. She cannot ascribe P any sort of independent truth (the meaning typically ascribed to “true.”) This is, after all, the whole point of her “debunking” argument: everyone’s evaluative beliefs may be gravely mistaken, and so the fact that P is “true” in a sense that is only measured by accord with evaluative beliefs should give no comfort. Thus, this antirealist position is susceptible to some kind of “performative” counterargument as described above.

Of course, the foregoing argument is very unlikely to be convincing to a committed antirealist, but that is an unreasonable expectation of any such response. Instead, this performative counterargument exposes the weakness of the skeptic’s position in application, and also ties the most powerful anti-skeptic responses in other branches of philosophy into a metaethical response.

In conclusion, if we are not prone to widespread skepticism throughout our philosophy, we should not be particularly worried about Street’s skepticism. Vavova rightly identifies that “the potential strength of a debunking argument is inversely proportional to its ambition.” That is, by attacking such a wide swath of our beliefs at once, metaethical skepticism like Street’s fails to be convincing. We can see this through our analysis of either side’s inability to provide good (independent) reasons once all ethical beliefs are suspended and through our idea of a performative counterargument in which second-order skepticism is transferred to an untenable first-order skepticism.

Works Cited

Moore, G. E. (1939). “Proof of an External World.”

Street, Sharon (2006). “A Darwinian Dilemma for Realist Theories of Value.”

Vavova, Katia (2013). “Debunking Evolutionary Debunking.”