Metaethics is a Mess. A Modest Attempt to Sorting It Out

I have a forthcoming paper in the Erasmus Journal for Philosophy and Economics on Ken Binmore’s theory of the social contract. This article especially focuses on the naturalistic features of Binmore’s account and discusses its implications regarding the status of morality and moral propositions. Writing the paper and especially revising it have forced me to enter into the complex literature on metaethics. Metaethics is dedicated to a vast array of issues related to the meaning and nature of moral propositions. Of course, given the topic of my paper, I have been especially interested in the debate related to the so-called “evolutionary debunking of morality”, i.e. the claim that the naturalistic origins of morality (biological evolution through evolutionary processes like natural selection) undermine the justification of our moral beliefs and judgments. On this issue, Richard Joyce’s book The Evolution of Morality figures as one of the most important recent contributions. Sharon Street’s widely cited paper “A Darwinian Dilemma for Realist Theories of Value” (Philosophical Studies, 2006) is also an important reference on this topic. The evolutionary debunking argument is also largely discussed by Derek Parfit in On What Matters (especially Part 6). In the course of my readings in metaethics, I have found out that it is often difficult to compare and to articulate various arguments and claims coming from different authors because concepts are not always consistently used. This post is a modest attempt to achieve conceptual clarification in the very specific case of the evolutionary debunking argument, and even more specifically on the basis of the three references just cited.

Part 6 of On What Matters (OWM) offers a long and sophisticated discussion of the opposition between naturalism and what Parfit calls “non-naturalist cognitivism” (Parfit has not much to say on non-cognitivism). Interestingly, Parfit eschews labels like “realism” or “moral realism” which are sometimes used to refer to the position which endorses cognitivism (moral propositions express judgments that may either be true or false). In relationship with the opposition between “objectivism” and “subjectivism” in the theory of normative reasons that he tackles in part 1 of OWM, Parfit defends what can be called an “objectivist-non-naturalist-cognitivist” account of morality and normativity. In a nutshell, reasons for actions are provided by values which are independent from the persons’ attitudes (beliefs, desires) – the objectivist part – and there exist authentic normative properties in the world that are independent from naturalistic properties – the non-naturalist part. Though Parfit treats these two features separately, he recognizes that they are highly related and indeed, what is the nature of this relationship is what will be at stake below. Parfit’s position is thus that are there are things in the world that make some actions or events “bad”, “good”, “right”, “desirable” and so on. These things are properties that do not depend on our attitudes or even on our existence. Moreover, these properties cannot reduced in any meaningful way to naturalistic properties. For instance, the property of “X being good for Y in circumstances Z” is neither identical with nor supervene on a naturalistic property (e.g. X is pleasurable for Y in Z). In Parfit’s famous example of Tuesday Indifference, it is irrational plain and simple to be indifferent to any suffering endured any Thursday but not from sufferings endured any other day of the week because there is no reason – in an objectivist sense – for such an indifference. Moreover, this lacks of reason is due to the normative properties in the world that makes events or actions good or rational.

Of course, as a corollary, Parfit rejects both subjectivism about reasons and all forms of (cognitive) naturalism.  The former holds that reasons for action are provided by one’s desires. The latter can be separated between an analytical and a non-analytical versions. Analytical naturalism holds not only that normative properties are identical with naturalistic properties, but also that all normative concepts can ultimately be defined in terms of naturalistic concepts. Non-analytical naturalism claims that normative concepts are different than naturalistic concepts, but accept the naturalistic contention that normative properties can be reduced to (or supervene on) naturalistic properties. Joyce and Street are both explicitly naturalists. Their positions with respect to the evolutionary debunking argument are not however the same. In his book, Joyce provides a forceful defense of moral skepticism, i.e. the view that the naturalistic origins of morality fail to justify our moral beliefs and judgments. On this view, knowing from where our moral beliefs come from, we do not have any positive reason to consider that they are justified. Joyce’s claim is that a naturalist must be a moral skeptic. In other words, what he calls “moral naturalism” (which can be defined as the view that there moral properties while acknowledging their naturalistic foundations) is an untenable doctrine. Street’s position, as I understand it, is quite different. In her paper, she presents the “value realist” with the following dilemma. Either the value realist contends that there is no relationship (causal or other) between independent moral truths and evolutionary influences on our attitudes. In this case, it is highly unlikely but for some extraordinary coincidence that our beliefs about moral truths are correct.  Or there is indeed a relationship between moral truth and evolutionary influences on our attitudes. However, in this case, the value naturalist must hold that natural selection are selected our attitudes for their ability to track the moral truth, a claim which cannot be supported on an evolutionary basis. Against this “Darwinian dilemma”, Street’s conclusion is however not moral skepticism. In a recent contribution in a book dedicated to discussing OWM’s sixth part, Street argues that we can be anti-realist about values while denying that skeptic conclusions are unavoidable.

I have had a hard time making sense of this whole discussion because Parfit, Joyce and Street seem not always to speak of the same thing. Parfit cautiously avoids using the term “realism” while Street makes it her major focus. Joyce opposes moral realism to moral naturalism but his use of the latter seems not to be incompatible with some form of realism. Moreover, Joyce’s notion of naturalism is broad and not fully consistent with Parfit’s use of the term. Ultimately, I have arrived at the following understanding. Parfit’s opposition between objectivism and subjectivism corresponds to Street’s opposition between realism and anti-realism. It concerns the issue of whether normative reasons are attitude-independent or attitude-dependent. The most satisfactory understanding I have of Joyce’s distinction between moral realism and moral naturalism is that it does not exclude in principle (“analytically”) that there is an overlapping area between the two. Indeed, I interpret Joyce as saying that we can have three different metaethical stances:

  1. Moral realism with attitude-independent normative reasons: since this view is not naturalist, Joyce takes it to be irrelevant.
  2. Moral naturalism with attitude-dependent normative reasons: Joyce argues that this must lead to moral skepticism.
  3. Realism-compatible moral naturalism: normative reasons are attitude-dependent and naturalistic and normative properties are ultimately identical, but normative properties are still part of our world. Joyce argues that this view is untenable because the attitude-dependent normative reasons are actually not really normative (i.e. lack a form of practical authority).

I have no space to present and discuss in details Joyce’s argument against this “hybrid” form of moral naturalism but my point is that this seems to correspond to the form of anti-realistic value naturalism that Street is endorsing. The table below summarizes the various metaethical positions with respect to the evolution debunking argument, using Parfit’s way of classifying views:


If one is worried with moral skepticism, then this table indicates that the alternatives are either to endorse a strong form of moral realism, grounded on the combination of non-naturalism and objectivism, or to defend a form of non-analytic naturalism. The combination of objectivism (thus claiming the existence of attitude-independent normative reasons) and non-analytic naturalism would answer Joyce’s skeptic critic. However, such a position seems hard to entertain, if not contradictory. The issue is then whether an anti-realistic value naturalism is substantively (rather than conceptually) possible. But this is a topic for another post.