Naturalism and Nonnaturalism
6
What Is Moral Realism?
This chapter will spell out what moral realism is and identify various formulations of what it might look like. Then it will point to and motivate some of the prima facie reasons to accept it and some of the salient reasons to reject it, or at least retain skepticism about it. The chapter will then briefly introduce a range of alternatives to moral realism to be considered in some detail.
The Nature of Moral Realism
Moral realism can be defined as requiring (1) a semantic or cognitivist thesis, (2) an alethic or nonnihilist thesis, (3) an ontic or metaphysical thesis, and then, for our purposes, (4) a nonskeptical thesis. Some familiarity with this language and these ideas will facilitate understanding moving forward. A way to get a handle on the first condition—the semantic or cognitivist thesis—is to ask if moral sentences are truth-apt. That is, do they express propositions? Another way to ask the question is to ask whether moral judgments have a truth value; in other words, Are they either true or false? Moral realism affirms that at least some of our moral judgments are indeed truth-apt. They express beliefs, and they have a truth value. Moral sentences are what they seem: statements about morality that either correspond with reality or do not. If they do, they are true; if not, they are false.
Such cognitivism distinguishes moral realism from expressivism that, at least in its earliest versions, denies that moral judgments are cognitively meaningful. Moral judgments, according to such expressivists, are neither true nor false but are rather something like an expression of positive or negative emotions. Expressivism obviously contrasts with the sturdily realist view that moral statements represent moral facts and that they express actual propositions that are true or false.
What about the second condition—the alethic or nonnihilist thesis? This requirement insists not just that moral judgments are true or false but that at least some of our moral judgments are true. In affirming the alethic thesis, moral realism distinguishes itself from moral error theory, a view historically espoused by J. L. Mackie and, more recently, by Richard Joyce, Jonas Olson, Bart Streumer, and others.1 Error theorists hold that all moral claims are false: Moral statements express propositions, contrary to what the expressivists say, and possess truth values, but every one of them is false. In contrast, moral realists think that at least some of our moral judgments are true.
More is needed, however. Moral relativism affirms that some moral propositions are true, but it is not a realist account. One version of such a “constructivist” or subjectivist view is that the prevailing moral view in a society renders that view true for that society. Such relativism is just one variant of constructivism, which we will have occasion to discuss later, but the present point is that, like moral realists, moral relativists speak in terms of true moral claims or judgments. Unlike moral realists, however, moral relativists make moral truths depend on the prevailing attitudes or preferences of individuals, cultures, or subcultures. Moral realists are not relativists. They instead believe that there are objective facts of the matter when it comes to what is right or wrong, good or bad, virtuous or vicious.
Moral realism can thus be said to affirm a third requirement of objectivity that we can call the ontic (or metaphysical) thesis. Whereas the semantic thesis pertains to the nature of moral discourse, and the alethic thesis pertains to the truth of moral claims, the ontic thesis pertains to the character or nature of moral facts. Typically, realists of either naturalist or nonnaturalist stripe underscore something like the stance-independent nature of moral truths. This is to say that their truth does not depend on the attitudes, inclinations, or preferences of human beings. If someone has a duty to perform an action, for example, the person’s resistance to discharging his obligation does not diminish the authority of the obligation, nor does it sanction noncompliance.
J. L. Mackie characterized moral realism by suggesting that (at least some) moral truths feature objective prescriptivity. Both features are important here: objectivity and prescriptivity. Morality is not purely subjective—relative to one’s culture or to individuals within that culture. It applies across the board and is not subject to the winds of popularity, caprice of majority rule, or tyranny of individual preference. Morality is not only objective but also at times altogether prescriptive—imposing binding and authoritative prescriptions and proscriptions on our behavior.
This union of objectivity and prescriptivity may well be the most interesting feature of morality. It is also the hardest part of moral realism to defend. Objectivity alone is not very difficult to generate; nor is moral prescriptivity in isolation. But defending their conjunction poses a significant challenge for moral realists. Mackie was not wrong to issue this very challenge. His strategic move was exactly right, and what is at stake is nothing less than the issue of moral authority.
Many might stop right here in their characterization of moral realism: the semantic thesis, the alethic thesis, and the ontic thesis. Morality is cognitivist, sometimes true, and objective. Affirming all three conditions, however, is consistent with a denial of any moral knowledge. To us this seems problematic. Of course, to believe, concerning some field of discourse, that there is an objective fact of the matter is distinct both conceptually and in practice from claiming to know exactly what that fact is. In ethics, as in other areas, we see through a glass darkly and remain susceptible to enough confirmation biases, blinders, and distorting influences that retaining an attitude of epistemic humility is paramount.
Still, any moral realism worth our allegiance requires a nonskeptical constraint. This is a fourth condition for the sort of moral realism we will be discussing. We can know at least some moral propositions. If we are wrong, for example, about the moral atrocity of burning people alive for fun, then our moral realism does not mean much. If we cannot have the requisite confidence to rule anything out from the realm of moral propriety, then the line between moral realism and antirealism practically disappears. Moral realism in such a case is rendered an effete intellectual conjecture with little to no practical punch, purpose, or point.
This means that moral realism that is watered down or easily domesticated will not be our present concern. We will not be defending a version of moral realism consistent with complete moral skepticism, nor a variant of moral realism that rejects any of the conditions we have discussed. Weaker forms of moral realism—sometimes called minimal or moderate—pick and choose among the conditions specified, but we are interested in categorical oughtness, objective prescriptivity, and robust moral knowledge.
Moral realists come in different stripes, some more “full blown” than others. This book will be especially interested in moral realism in a quite strong sense. To a good first approximation, then, strong moral realism affirms that moral judgments are truth-apt, that some of them are true, that their truth does not depend on the attitudes of human beings, and that some clear moral truths are known. This is to say that strong or robust moral realism affirms the semantic thesis, the alethic thesis, the metaphysical thesis, and the nonskeptical thesis.
Such realism, if it can be defended, arguably provides the relevant evidence for a theistic hypothesis. If morality is reducible to one version of antirealism or another, it becomes much less likely that it can do appreciable evidential work for theism.
Naturalism and Nonnaturalism
Before discussing how we will go about making the case for moral realism, we need to identify an additional distinction among moral realists in order to make clear the variant of moral realism of particular interest to us. G. E. Moore, in his classic Principia Ethica, distinguished between naturalist morality and nonnaturalist morality.2 His favored term for the latter was metaphysical morality, but we will employ the term “nonnaturalist.” Realism, then, can come in naturalist varieties and nonnaturalist varieties. Our focus in this volume will be on the latter, but let’s quickly discuss naturalistic ethics. Giving it its due would require a volume of its own; here we will simply describe it a bit before setting it to the side and pointing to some additional resources for those who wish to explore it further.
Although this section of the book will argue for robust moral realism, and there are naturalist moral realists, naturalistic ethics does not really help our cause. It would not show that God does not exist, but it would show that morality, even if it is real, does not give us reasons to think that God exists, at least not directly. Perhaps God would be the indirect source of morality in virtue of having created the material world, but that’s not the sort of argument that we will advance in this book. Of course, if the naturalist ethics under discussion presupposed naturalism, then that would show that God doesn’t exist, but that would be problematically circular. We are starting our analysis by assuming neither that God exists nor that God does not exist.
Moral naturalists, as we are characterizing them, are simply those who adhere to the idea that morality is somehow reducible to what we find in the natural world. The moral realists among them, our focus here, affirm mind-independent, objective moral facts. They are cognitivists who see moral judgments as belief-like mental states that describe the way the world is. They affirm that moral facts exist and that they obtain independent of our understanding them or our stance toward them. Usually, moral naturalism affirms not just the metaphysical component that moral facts are natural facts but also an epistemic commitment that we know moral claims are true in the same way that we know about claims in the sciences. Such realism can come in different varieties; one important salient example is Cornell realism, which accepts the view that moral facts are natural facts, falling within the province of the natural and social sciences.3
When discussing moral naturalism, Moore lodged his famous naturalistic fallacy against it. On Moore’s analysis, any effort to equate any natural property or combination of natural properties with what is morally good is a mistake. Perhaps there is coextension between the categories, but that does not entail sameness. Moore’s so-called open question argument is designed to show the distinction or disconnect between any proposed naturalistic account of goodness and goodness itself. Suppose we say that pleasure is good. Whereas the question “Is good good?” is a closed question, the question “Is pleasure good?” is not. That the latter, but not the former, remains an “open question” reveals, he thought, that goodness and pleasure are not the same. The point is generalizable to any naturalistic proposal for the good. The good, he thought, is rather something beyond garden-variety description, something ineffable and undefinable, analogous in certain respects to something like the color yellow.
Most moral naturalists today reject Moore’s heady conception of the good, eschewing it in favor of something less mysterious and more down-to-earth, as we see in Peter Railton’s approach. They also can employ insights from direct reference theorists like Saul Kripke, Hilary Putnam, and the like, to add sophistication to their analysis of the relationship between language and that to which language refers. Consider the analogy with the natural kind water. What we can know a priori is that whatever liquid we find in oceans and rivers, the stuff we drink and bathe with, is water, but it takes empirical investigation to discover water’s molecular composition. Its phenomenal features fix its reference but underdetermine its essential nature.
Likewise we know that what stirs our admiration and proactive attitudes in the relevant way is goodness, but, according to Richard Boyd and others, its essential nature remains to be determined, open to empirical investigation. The moral naturalists are convinced that naturalistic accounts of moral goodness are on offer that can accomplish the task. Perhaps it is not just one natural property but rather a complex combination of natural properties. The semantics of natural kinds can be deployed to answer the conceptual challenges posed by the naturalistic fallacy and open question argument. For example, “Is water water?” is a closed question, whereas “Is water H2O?” was long an open question requiring empirical confirmation. The relevant concepts are indeed distinct.
Apart from construing the challenge along conceptualist lines, though, the challenge can also be interpreted metaphysically. A metaphysical interpretation of the open question may well be justified, for this basic reason: Naturalist properties by their nature seem intrinsically inadequate to account for something like moral normativity. As Russ Shafer-Landau puts it, “There are genuine features of our world that remain forever outside the purview of the natural sciences. Moral facts are such features. They introduce an element of normativity that cannot be captured in the records of the natural sciences. They tell us what we ought to do; how we should behave; what is worth pursuing; what reasons we have; what is justifiable and what not. There is no science that can inform us of such things.”4
Naturalists who are moral realists agree that there are facts by virtue of which moral locutions (like “wrong”) have determinate content. The facts to which they point are all natural facts, though, and natural facts arguably fail to account for the sort of normativity at the heart of ethical claims. Surely moral naturalism can account for some legitimate pushback against inappropriate normative claims, but the moral naturalist intends to insist that every bit of such pushback must be natural. Advocates of the moral argument, in contrast, argue that this is not so. At least some of the resistance is nonnatural. This will be our claim in this book, and at the heart of our conviction is the difficulty that naturalism encounters in accounting for moral normativity. Natural facts strike us as ineliminably nonnormative. As we will argue, a nonnatural account, specifically a theistic one, provides the better grounding for moral normativity.
This is the metaphysical import of Moore’s reasoning from earlier. As J. P. Moreland and David Horner put it,
The central intuition of Moore’s reasoning is that every attempt to reduce moral properties to nonmoral properties fails and must fail because that reduction leaves out what is most essential to moral properties—their irreducible normativity. Indeed, because all properties (according to naturalism) are natural, scientific properties, normativity of all kinds (not just moral normativity) turns out to be impossible. Since normativity, including moral normativity, is a real feature of the world, however, naturalism fails. As Christian philosopher Alvin Plantinga puts it, “[Naturalism’s] Achilles’ Heel (in addition to its deplorable falsehood) is that it has no room for normativity. There is no room, within naturalism, for right or wrong, or good or bad.”5
At any rate, we think that the evidence of morality points in a different direction. We will leave defenders of moral naturalism to defend their own views; here we will mainly be laying out our own vision about morality and the evidential direction in which it points.
With that brief introduction to moral realism, let us now consider its main metaethical rivals.
1 J. L. Mackie, Ethics: Inventing Right and Wrong (Penguin Books, 1977); Richard Joyce, The Evolution of Morality (MIT Press, 2007); Jonas Olson, Moral Error Theory: History/Critique/Defence (Oxford University Press, 2017); Bart Streumer, Unbelievable Errors: An Error Theory About All Normative Judgements (Oxford University Press, 2017).
2 G. E. Moore, Principia Ethica (1903; repr., Dover, 2004).
3 A distinction holds between reductive moral naturalists and nonreductive moral naturalists, but both variants of moral naturalism would answer the question, “Are values facts?” in the affirmative, specifically natural facts. Cornell realism is a view in metaethics, associated with the work of Richard Boyd, Nicholas Sturgeon, and David Brink, who earned his doctorate at Cornell University.
4 Russ Shafer-Landau, Moral Realism: A Defence (Oxford University Press, 2003), 4.
5 David A. Horner and J. P. Moreland, Metaethics: A Short Companion (B&H Academic, 2024), 110–11. Horner and Moreland quote Alvin Plantinga, afterword to The Analytic Theist: An Alvin Plantinga Reader, ed. James F. Sennett (Eerdmans, 1998), 356.