Skip to main content

Robust moral realism: an excellent religion


According to robust moral realism, there exist objective, non-natural moral facts. Moral facts of this sort do not fit easily into the world as illuminated by natural science. Further, if such facts exist at all, it is hard to see how we could know of their existence by any familiar means. Yet robust realists are not moral skeptics; they believe that we do know (some of) the moral facts. Thus robust moral realism comes with a number of hard-to-defend ontological and epistemological commitments. Recently, Sharon Street has claimed, in light of these commitments, that robust moral realism requires a kind of faith and “has become a strange form of religion.” I believe that Street is right. I argue at some length that robust moral realism does require faith, and is a religion. However, I further argue that it is an excellent religion. I argue that it has three principal advantages: it is avoids wishful thinking, is guaranteed not to contradict the results of natural science, and is profoundly simple in its ontological commitments. Further, robust moral realism may be rationally defensible on evidentialist grounds. Consequently, even if the standard arguments for traditional religions are not compelling, there might still be compelling arguments for robust moral realism.

This is a preview of subscription content, access via your institution.


  1. 1.

    Street (unpublished manuscript, pp. 22–23).

  2. 2.

    (Mackie (1977), p. 37), says that moral properties are “of a very strange sort, utterly different from anything else in the universe.” It is unclear, to many of us, precisely how Mackie supposes these properties to be strange; but one possibility is that he regards them as fundamentally different from the sorts of properties that are familiar in a modern scientific worldview, as opposed to a prescientific religious worldview. (Mackie does not indicate whether he thinks these properties would seem particularly strange to a medieval Scholastic theologian, for instance; but one suspects not).

  3. 3.

    Annette Baier argues (unsuccessfully, in my view) that morality, science, and knowledge acquisition, in the absence of faith in God, all require “faith in the human community and its evolving procedures” (1980, p. 133). I’ll be arguing for a far more specific faith requirement.

  4. 4.

    (Parfit (1986), p. 454).

  5. 5.

    Schellenberg suggests that we develop a new, skeptical kind of religion—one that is appropriate for the level of scientific and philosophical knowledge that we now possess. So, he is engaged in an activity that we might call religious engineering (borrowing a term from Frank Herbert’s 1965 novel, Dune). See his (2013) for an overview of his project.

  6. 6.

    (Shafer-Landau (2003), p. 59), (Enoch (2011), p. 103), and (Parfit (2013), p. 109) understand the natural in a similar fashion.

  7. 7.

    The core idea behind this assumption is that (ideal) science is maximally open: it embraces absolutely anything that substantially contributes to best explanations of natural phenomena. This means that even God’s existence, for example, would be a scientific (hence natural) fact, if God’s existence played the right kind of explanatory role.

  8. 8.

    David Enoch, for example, wiggles on this point: he says he is both “neutral” and “pessimistic” on the question of the explanatory utility of moral facts (2011, p. 53).

  9. 9.

    Here and throughout this paper, I’ll assume Michael Huemer’s account of seemings. Huemer refuses, I think rightly, to analyze “seems” (2013, pp. 328–9). Seemings are non-doxastic content-bearing experiences of the sort that we typically have in mind when we say, “It seems to me that...” Beyond that rough and imprecise characterization, the best way to give an idea of seemings in general, and moral seemings in particular, is simply to cite illustrative cases. It seems to me, e.g., that torturing an innocent child is morally wrong.

  10. 10.

    A moral belief that P is deep for a subject S iff (a) P is believed with very high confidence by S, and (b) P is highly coherent with S’s other moral beliefs.

  11. 11.

    I assume that a belief that P is true iff it is in fact the case that P. So, if our deepest moral beliefs are mostly true (as moral optimists say) then moral facts exist (as non-nihilists say).

  12. 12.

    (Audi (2011), chap. 3) contains a very useful discussion of the relations between faith, belief, and hope.

  13. 13.

    Some might object [as Caleb Ontiveros does, in conversation; and as (Schellenberg (2005), chaps. 5, 6) argues at length] that faith is voluntary, whereas belief is involuntary, and therefore faith and belief are mutually exclusive. But, for one thing, it’s not obvious that all belief is involuntary. There is a deep literature on doxastic voluntarism; see Weatherson (2008) for a limited defense. Secondly, it’s not obvious to me that all faith is voluntary. (I suppose that a person who is brought up in an environment steeped in religious faith might well lack voluntary control over her faith, just as a person brought up in a household of geographers might have no voluntary control over her beliefs about the location of Asia.) My guess is that some but not all belief is voluntary, and that some but not all faith is voluntary.

  14. 14.

    Audi considers the possibility of content restrictions on faith (2011, p. 60) but seems to leave open the question of whether such restrictions might actually exist.

  15. 15.

    It bears noting, however, that some philosophers have argued that faith can be non-religious; e.g., see Baier (1980).

  16. 16.

    Paul Draper (2005) suggests that genuinely supernatural phenomena must be causally efficacious. I think this is wrong. Intuitively, God would be no less supernatural if God never engaged in any creative activity and remained causally inactive. More moderately, Stephen Ingram argues (in conversation) that the supernatural can be causally efficacious (whereas the non-natural cannot be). If Ingram’s more modest suggestion turns out to be correct, it would still be plausible to suppose that the non-natural is a causally and explanatorily inert subtype of the supernatural, which is all I need to claim here.

  17. 17.

    Relatedly, J. L. Schellenberg says that the following features are “defining features [of religiosity], if any are”: “(1) frequent thoughts of a transmundane reality; (2) an emphasis on a significant good, for oneself and others, that may be realized through a proper relation to this reality; (3) the cultivation of such a relation; and (4) a disposition or tendency, when attending to matters in which they are implicated, to—as I shall put it—totalize or ultimize in some way the central elements of features (1) to (3)” (2005, p. 12). It seems to me that robust realists satisfy at least the first three conditions. (Moral facts, in the robust realist worldview, are transmundane just because they are non-natural.) Whether or not robust realists satisfy the fourth condition is unclear to me; this depends on what it means to totalize or ultimize something. Schellenberg sometimes sounds as if he thinks that the ultimate, for a given individual, is just whatever that individual considers to be of greatest importance.

  18. 18.

    Many robust realists—in conversation, if not in print—will say that they defend robust realism because they hope it’s true, or wish it were true, or because they can’t bear the thought that it is false, etc. But I, for one, don’t wish or hope that robust realism is true. When I blow out my birthday candles, I wish for many things; the truth of robust realism is not among them. I endorse robust realism for the reasons I’ve already given: (a) I think objectivism is a conceptual truth, (b) I think non-naturalism is a sound empirical generalization, and (c) I think moral optimism is true because I believe that moral seemings are trustworthy. It is not unlike the way in which I believe that I am surrounded by tables and chairs. If it turns out that I’m not surrounded by tables and chairs (because, e.g., ontological nihilism is true, or because I’m in a simulation of the sort that Nick Bostrom imagines) then I would certainly be surprised; I might begin to doubt my sanity; even so, I am not emotionally invested in the presence of tables and chairs around me.

  19. 19.

    (Shafer-Landau (2007), p. 323). Here it is also worth noting that robust realism’s lack of empirical implications might be taken as a disadvantage for it—because its lack of empirical implications might be thought to make it unfalsifiable, and the unfalsifiability of a theory might be considered to be a disadvantage of that theory. There are several ways to respond to this concern; here is my favorite one. First, it is not the case that robust realism is unfalsifiable, if that means that no evidence against it is possible. On the contrary, plenty of evidence against robust realism is possible. For example, the phenomenon of moral disagreement might be taken to be an important piece of evidence against moral optimism, which is one of the three components of robust realism. It is true that robust realism lacks empirical predictions, and thus cannot be tested by empirical evidence (at least not in the way that scientific theories can be). But I do not see why the mere fact that a theory cannot be tested by one type of evidence is a problem for that theory, provided that the theory in question is still susceptible to evidence of some sort.

  20. 20.

    Mavrodes (1986).

  21. 21.

    In other words, the Christian can posit the Divine Command Theory. To be sure, many Christians reject the Divine Command Theory. But I am unconvinced by the standard objections, including the Euthyphro objection. I think that the Divine Command Theory, or some variation on it, is plausible—if God exists (c.f. (Carson (2000), p. 250)).

  22. 22.

    (Wielenberg (2014), p. 38).

  23. 23.

    Anscombe (1958).

  24. 24.

    These points do not apply to certain extremely simple forms of religion. For example, imagine a religion that posits a perfectly simple God, and posits that this simple God is responsible for both morality and for the natural world. If this religion posits nothing further—e.g., it does not posit any angels, or Heaven, or Christ, etc.—then I’m prepared to grant that it would be simpler and, arguably, more parsimonious than robust moral realism. Of course, this doesn’t rule out the very real possibility that the simple religion on offer would have other problems (e.g., the need to explain how a perfectly simple God could be responsible for a highly complex universe).

  25. 25.

    See the valuable discussion of related issues in Shafer-Landau (2007).

  26. 26.

    See my unpublished paper, “Moral Occasionalism.”


  1. Anscombe, G. (1958). Modern moral philosophy. Philosophy, 33, 1–19.

    Article  Google Scholar 

  2. Audi, R. (2011). Rationality and religious commitment. Oxford: Oxford University Press.

    Book  Google Scholar 

  3. Baier, A. (1980). Secular faith. Canadian Journal of Philosophy, 10(1), 131–148.

    Google Scholar 

  4. Carson, T. (2000). Value and the good life. Notre Dame: Notre Dame Press.

    Google Scholar 

  5. Draper, P. (2005). God, science and naturalism. In W. J. Wainwright (Ed.), The Oxford Handbook of Philosophy of Religion. Oxford: Oxford University Press.

  6. Enoch, D. (2011). Taking morality seriously. Oxford: Oxford University Press.

    Book  Google Scholar 

  7. Huemer, M. (2013). Phenomenal conservatism Uber Alles. In C. Tucker (Ed.), Seemings and justification. Oxford: Oxford University Press.

    Google Scholar 

  8. Mackie, J. L. (1977). Ethics: Inventing right and wrong. London: Pelican.

    Google Scholar 

  9. Mavrodes, (1986). Religion and the queerness of morality. In Audi & Wainwright (Eds.), Rationality, religious belief and moral commitment. Ithaca: Cornell.

    Google Scholar 

  10. Parfit, D. (1986). Reasons and persons. Oxford: Oxford University Press.

    Book  Google Scholar 

  11. Parfit, D. (2013). On what matters: Volume one. Oxford: Oxford University Press.

    Google Scholar 

  12. Schellenberg, J. L. (2005). Prolegomena to a philosophy of religion. Ithaca: Cornell.

    Google Scholar 

  13. Schellenberg, J. L. (2013). Evolutionary religion. Oxford: Oxford University Press.

    Book  Google Scholar 

  14. Shafer-Landau, R. (2003). Moral realism: A defence. Oxford: Oxford University Press.

    Book  Google Scholar 

  15. Shafer-Landau, R. (2007). Morality and theological realism: The explanatory argument. Journal of Moral Philosophy, 4(3), 311–329.

    Article  Google Scholar 

  16. Street, S. (unpublished manuscript). Objectivity and truth: You’d better rethink it.

  17. Weatherson, B. (2008). Deontology and Descartes’ Demon. Journal of Philosophy, 105(9), 540–569.

    Article  Google Scholar 

  18. Wielenberg, E. (2014). Robust ethics: The metaphysics and epistemology of godless normative realism. Oxford: Oxford University Press.

    Book  Google Scholar 

Download references


Thanks to Ben McCraw, Caleb Ontiveros, Calum Miller, Casey Woodling, Cliff Sosis, Dan Hausman, Daniel Crow, Derrick Murphy, Dylan Cordaro, Emily Crookston, John Basl, Jonathan Lang, Justin Weinberg, Justin Horn, Kate Myrna, Matthew Kopec, Molly Gardner, Nick McKinney, Nils Rauhut, Preston McKever-Floyd , Rob Streiffer, Robert Audi, Russ Shafer-Landau, Shannon Kelly, Stew Eskew, Syd Nichols, Tom Carson, Yishai Cohen an anonymous reviewer for this journal, and audiences at the South Carolina Society for Philosophy, the Philosophers’ Cocoon Annual Conference, and the Rocky Mountain Ethics Congress for valuable discussion and feedback.

Author information



Corresponding author

Correspondence to David Killoren.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Killoren, D. Robust moral realism: an excellent religion. Int J Philos Relig 79, 223–237 (2016).

Download citation


  • Ethics
  • Moral realism
  • Faith
  • Atheism
  • Phenomenal conservatism
  • Evidentialism