Robust moral realism is the view that moral facts exist, but that they are not reducible to non-moral or natural facts. According to the robust realist, when I say something like ‘It is morally wrong to torture an innocent child for fun’, I am saying something that is true, but whose truth is not reducible to the non-moral properties of torture or children. Robust moral realism has become surprisingly popular in recent years, with philosophers like Derek Parfit, David Enoch, Erik Wielenberg and Russell Shafer-Landau all defending versions of it.
What is interesting about these philosophers is that they are all avowedly non-religious in their moral beliefs. They don’t think there is any connection between morality and the truths of any particular religion. Indeed, several of them are explicitly atheistic in their moral outlook. In a recent paper, however, David Killoren has argued that robust moral realism is a kind of religious belief: one that must be held on faith and that shares other properties with popular religions. At the same time, he argues that it is an ‘excellent’ kind of religious belief, one that could be attractive to the non-religious and religious alike.
I want to look at the argument Killoren uses to defend this point of view.
1. Three Features of Robust Moral Realism
Before we get into the meat of Killoren’s argument, we need to have a clearer characterisation of robust moral realism. Go back to my earlier example of the proposition ‘It is morally wrong to torture an innocent child for fun’. Suppose you and I are discussing this proposition one day. You happen to agree with it. Killoren argues that if we were robust realists then we would share three key beliefs about the nature of our agreement about this claim:
Non-naturalism: We would both believe that the statement ‘it is morally wrong to torture an innocent child for fun’ is a non-natural fact. I explained this above but I can be more precise here. The key thing is that we would believe that it is an irreducibly normative fact. It may supervene on natural facts, but it is distinct from and not identical with those natural facts.
Objectivism: We would both believe that the truth of the statement is not dependent on our moral attitudes. In other words, its being true does not depend on our believing it to be true. It is mind-independent. Killoren argues that objectivism in this sense entails that moral facts are independent from two distinct moral attitudes: (i) our moral beliefs and (ii) our moral seemings (i.e. the fact that it seems like X is true).
Optimism: We would both believe that we do in fact know that the statement ‘it is morally wrong to torture an innocent child for fun’ is true. We are consequently optimistic about the truth of our moral beliefs. Killoren is, again, more precise here in saying that optimism is the view that our deepest moral beliefs are true. So we may disagree at the margins (e.g. ‘we should give 10% of our income to charity’), but we agree about more fundamental moral claims (like claims about the torturing of innocent children).
Two things are worth noting about this triptych of beliefs. The first is that the commitment to non-naturalism comes with a significant cost. Killoren calls it the ‘non-naturalist’s handicap’. If non-naturalism is true, then it means that ‘moral facts do not play a contributory role in the best explanation of any natural facts’. This is troubling for otherwise non-religious naturalists because it means that all their moral beliefs and attitudes are not best explained by the existence of true non-natural moral facts. If our minds are ultimately best explained by natural facts, then non-natural moral facts cannot feature in the explanation of the content of our minds. This ‘handicap’ is something that non-realists like Sharon Street have long complained about.
The other point is that the commitment to optimism is the only thing that saves robust realism from moral nihilism. As Killoren puts it, the first two commitments are really the standard features of robust realism. Everyone who calls themselves a robust realist will agree that they are committed to non-naturalism and objectivism. But those two commitments are ontologically neutral. One could accept them and still believe that no moral truths actually exist (e.g. because one is a metaphysical naturalist) or that we can never know what they are. Of course, no robust realist tends to accept this nihilistic view. They all think that moral truths are knowable and that we have a good grasp of the basic ones. So optimism is, implicitly, a feature of their view.
3. Robust Realism is a Type of Religious Belief
Now that we have a clearer sense of what robust realism entails, we can look at Killoren’s main argument. The first thing Killoren does is argue that robust realism requires a kind of faith. He has a lengthy discussion of faith in his paper. He notes that some accounts of faith are non-doxastic in nature, i.e. they hold that faith has nothing to do with our beliefs and everything to do with our desires and aspirations. On these non-doxastic accounts, faith is like hope (or some other positive attitude). But most accounts of faith include a doxastic element. Those are the accounts in which he is interested.
He then distinguishes between two types of doxastic faith:
Blind Faith: Belief in P in the absence of any evidence that P (or in the face of countervailing evidence that not-P).
Unscientific Belief: Belief in P in the absence of any scientific evidence that P (or in the face of countervailing scientific evidence that not-P).
Obviously, blind faith is much broader than unscientific belief. The idea is that there might be some evidence to support an unscientific belief and this evidence might make belief that P rationally defensible, even though P is not supported by or consistent with scientific beliefs. Blind faith involves belief in the absence of even unscientific evidence.
Killoren’s first argument is that robust realism requires faith because it requires unscientific belief. In defence of this he introduces something we can call the argument from the ‘explanatory superfluity entails faith’ principle. It works like this:
- (1)If one believes that P even though one accepts that P does not play any contributory role in the best available explanations of any natural facts or phenomena, then one believes that P on faith. (the ESEF principle)
- (2) Commitment to robust realism requires that one believe in moral truths even though one accepts that moral truths do not play any contributory role in the best available explanations of any natural facts or phenomena.
- (3) Therefore, robust realism entails faith.
It is trivial argument in many ways. The conclusion follows from the earlier characterisations of faith and robust realism. If you find those earlier definitions persuasive, then you’ll find the argument persuasive. Interestingly, Killoren devotes a good portion of his paper to defending the ESEF principle from competing accounts of faith. I found this section of the paper unnecessarily distracting, but if you are keen on learning more about the nature of faith then it might be worth reading. I was quite happy to accept the argument as it stands.
Obviously, to say that robust realism requires faith is already to suggest that it is a type of religion (since, for most people, ‘faith’ is practically synonymous with ‘religion’). But Killoren goes on to enumerate three additional properties shared by religions and robust realism.
The first property is belief in the supernatural. Not all religions require this, but most do. They believe that there exists a realm of facts that lies beyond the natural world. Usually, this realm is taken to consist of supernatural agents, i.e. gods, angels, demons and so on. Robust realism also requires a form of supernaturalism. The moral facts so beloved by the robust realist exist in a non-natural realm. This realm does not consist of supernatural agents, but it doesn’t thereby lose any entitlement to the ‘supernatural’ label. Or at least that’s what Killoren argues. I have a somewhat different view. I tend to think that supernaturalism is a mind-first (or agent-first) ontology. Following the likes of Paul Draper, I tend to think of supernaturalism as being equivalent to the view that ‘mental entities have explanatory priority’, i.e. that the natural world is ultimately explained by some mental entity; and I tend to think of naturalism as the opposite, i.e. that the mental is ultimately explained by the natural. On this definition of naturalism and supernaturalism, robust realism would actually end up falling under the scope of naturalism. Felipe Leon refers to this as ‘broad naturalism’. But I admit that this is, to a degree, definition-mongering. As long as you don’t believe that moral facts are ultimately explained by supernatural agents, then I’d be happy enough to say that robust realism is a species of supernaturalism.
The second property is guidance on how to live. Most religions purport to provide their believers with some set of principles about how they ought to live. Sometimes these principles are very detailed. Robust realism provides something similar to its believers. It is optimistic about the possibility of finding out how we ought to live. And the moral beliefs that are at the core of that position do attempt to provide some guidance on how to live.
The third property is organisation. Here Killoren is suggesting that most religions have an institutional or organisational structure that supports the belief system, ensures that it is promulgated and propagated, and defends it from attack. This seems obviously true of the religions we encounter. They all try to sustain their networks of belief through some set of social organisations (I honestly can’t think of a counterexample). He argues that robust realism also has an organisational structure. He states that robust realists organise themselves via philosophy departments, journals and conferences. In this way they sustain and defend their network of beliefs.
For all these reasons — faith, supernaturalism, guidance on how to live, and organisation — Killoren submits that robust realism is a type of religion.
3. Robust Realism as an Excellent Religion
And yet robust realism is clearly an unusual kind of religion. It doesn’t have the same eschatology or creation myths as most religions. Nor does it purport to provide us with a comprehensive worldview. But therein may lie its strength. Killoren closes his article by arguing that even if robust realism is a religious belief, it is, nonetheless, an excellent religious belief, particularly if you are normally averse to religious belief. There are three reasons for this.
The first is that robust realism is devoid of wishful thinking. Unlike most religions it doesn’t provide for salvation. No one is coming to save us from our moral sins or lead us into everlasting life. To the religiously inclined, this might seem like a disadvantage, but to the usually non-religious it probably won’t. They often like the idea of facing reality and avoiding false hope.
The second is that robust realism will never conflict with the results of scientific inquiry. This makes it unlike many (but not all) other religious beliefs. Many religions are criticised because their core tenets (e.g. creation stories, historical origin myths) conflict with the best available scientific information. This often puts the scientifically inclined off. Robust realists don’t need to worry about this. It is baked into their worldview that their moral beliefs can never conflict with the results of scientific inquiry.
The third is that robust realism provides some basis for morality. Non-religious people are often criticised for not having a coherent or defensible grounding for morality; robust realism provides one. Now, admittedly, Killoren falters in his support for robust realism on this score. He thinks that it provides a grounding for morality but that this grounding is less coherent than the grounding provided by more traditional religious beliefs (e.g. traditional theistic metaethics). I happen to disagree quite strongly. I have written about it at length before. In essence, I don’t think that religion provides a coherent and defensible grounding for morality. In fact, I think that most religiously-motivated metaethical views end up collapsing into a form of robust realism. I won’t get into the arguments here, but you can read about them elsewhere on the blog.
In addition to this, Killoren thinks that robust realism is epistemically defensible. His main argument for this is an argument from phenomenal conservatism:
Phenomenal Conservatism: If it seems to be the case (to us) that P is true, then this provides evidence in favour of P’s truth.
The idea is that when we reflect on our deepest moral beliefs (like ‘it is morally wrong to torture an innocent child for fun’) they seem to us to be true. This provides evidence for their truth. That evidence is defeasible. In other words, you could introduce other evidence to suggest that our phenomenology is misleading us as to the truth of such moral claims. Killoren thinks that it is possible to show that robust realism is not defeated by other sorts of evidence. He doesn’t state this in his paper, but he has contributed to the debate in other work.
4. Concluding Thoughts
I have already offered some doubts about Killoren’s central thesis, in particular his definition of supernaturalism. Let me close with two further critical reflections.
First, I’m not convinced that Killoren’s thesis is an interesting or important one. For one thing, I have doubts about the methodology. Enumerating the key properties of religious beliefs and then highlighting how those properties apply to other sets of beliefs seems to be of dubious merit. Killoren could just as easily argue that mathematical realism (or Platonism) is a species of religious belief. It is also committed to non-naturalism, objectivism and optimism. It requires belief in the supernatural (according to Killoren’s definition). And its believers are organised in the same way that robust realists are. I think it may also require faith in the sense that it requires unscientific belief: It’s true that mathematical equations feature in scientific explanations, but they don’t feature causally (i.e. no one thinks that Newton’s laws cause gravity to exist). Mathematical facts, according to the Platonist, do not play a contributory role in the explanation of natural phenomenon; they merely help to describe them. But moral facts can do the exact same thing. Admittedly, mathematical realism doesn’t provide guidance on how to live (not directly anyway), but that means it lacks one feature among many. Should it still count as a religion? Maybe. Is that an interesting claim? I’m not sure.
Second, I left Killoren’s article still thinking that the most important challenge facing the realist is the epistemological one. The explanatory handicap is a serious one: if moral facts play no contributory role in the best explanation of our moral beliefs, then how can we be sure that our moral beliefs are rationally justified. Killoren falls back on phenomenal conservatism at the end, but I’m not sure that is enough. I think Street’s Darwinian dilemma still bites on this point. I know there are responses to it (I have covered them on the blog before) but they all seem like a stretch to me. The responses from, say, Enoch or Skarsaune tend to require a belief that the telos of evolution is in some pre-ordained harmony with the non-natural moral facts. I don’t know if I can make sense of that view. Killoren defends an another response in his paper 'Moral Occasionalism'. According to him, moral reasons are a type of natural fact and they influence our moral seemings in such a way that our moral seemings match the moral facts without themselves causally interacting with those moral facts. I haven't read the paper in detail, but again this sounds like quite an odd metaphysical view. (Which is not surprising since takes its lead from occasionalism, which is a view about the relationship between the divine and natural causation).
Still, I should probably read the paper before dismissing it.