by Jared Yarsevich
Moral realism occupies a precarious position in contemporary philosophy. It carries the notable distinction of drawing the opprobrium of both hard-nose naturalists and libertine relativists. Uniting erstwhile enemies such as these is quite the feat and is undoubtedly symptomatic of deep conceptual flaws. But then again, maybe the multi-front assaults indicate an inner strength that merits the attention. Maybe moral realism occupies that rare philosophical space that requires what Wilfrid Sellars calls “stereoscopic vision” in order to properly comprehend. Such vision takes insight from both the scientific and the manifest images of the world. The resultant image will be lost on those who cannot see the forest for the trees, and on those who can’t see the trees for the forest, but it allows the rest of us to reclaim a world in which both trees and forests exist. Less metaphorically, moral realism emerges into view once we embrace the anti-reductionist, realist lessons of post-positivist philosophy of science and apply them to insights from Kant and Wittgenstein. The non-reductionist, naturalistic moral realist asserts that just as tables and chairs are real and objective, despite not playing any role in fundamental physics or in any of the special sciences, moral principles are also real and objective. Like tables and chairs, morality is part of the Manifest Image. Yet, distinct from medium-sized furniture, morality exists as part of the social world. The metaphysics of morality, therefore, are closer to that of the metaphysics of money or nation-states than to that of tables, quarks, or genes. In this essay, I attempt to: 1) demonstrate how some popular arguments against moral realism fail; and 2) provide a positive account of what a naturalistic moral realism could look like.
Arguments against moral realism often invoke morality’s apparent mind-dependency and a distinction between objectivity and reality. I maintain that some types of mind-dependency do not disqualify things from being real, and the distinction between reality and objectivity dissolves at certain levels of description. To demonstrate how this can be the case, let’s begin by considering the case of money. Without a doubt, the US dollar did not exist prior to humans existing. And unlike bridges or buildings, the US dollar will cease to exist once humans are extinct (if not before). Therefore, the US dollar is mind-dependent in the sense that its existence depends on specific human beliefs, actions, expectations, and practices. Despite this mind-dependency there are, nonetheless, objective facts about money. It is objectively true that I currently have $1,000 in my bank account but sadly false that I have $1 million. This kind of objectivity, in turn, reveals a sense in which money is quite mind-independent too. I can’t just will a million dollars into my bank account. Money’s existence, therefore, is dependent on a collection of mental states but independent from the mental states of any particular individual.
John Searle introduces a helpful distinction between ontological objectivity/subjectivity and epistemological objectivity/subjectivity. Money, like all social entities, is ontologically subjective since its existence depends on mental states, but epistemologically objective since the standards that determine the truth or falsity of claims about it are independent of any particular individual’s mental state. Given the combination of money’s ontological subjectivity and epistemological objectivity, it makes it very difficult to sensibly deny the reality of that money. Money is real in the sense that it’s “out there” independent of any particular individual’s mental state. When we have certain complex physical and social configurations, which involve persons, their expectations, and institutions, we have money; real money. The upshot is that the reality of money, and the social world in general, is attached to its epistemological objectivity. At the level of explanation for the social and moral world, decoupling the concept of reality from epistemological objectivity makes little sense. It would be tantamount to affirming the truth that $1,000 is in my bank account, but denying that the money is real. This would be simply nonsensical. Demanding ontological objectivity as the criterion for social reality is to mistakenly apply the ontological standards of other more fundamental levels of reality to the social world. It’s a reductionist category error. Just as we don’t want to apply the standards of quantum mechanics to questions about inflation, species, or mental states, we don’t want to apply the standards appropriate for medium-sized physical entities to questions about money, nation-states, or morality.
But is money an adequate analogy for morality? Moral non-realists often favor the analogy with math and show that the objectivity of math is not sufficient to demonstrate the reality of numbers. Mathematical realism, however, is not a good comparison. Mathematical realists do not claim that numbers exist at the social level. Mathematical realists argue that numbers exist prior to and independently of human existence (i.e. numbers are ontologically objective in Searle’s sense). If something is claimed to be ontologically objective, then its reality can be decoupled from its epistemological standards. But naturalist moral realists do not argue that morality is ontologically objective (albeit non-naturalist moral realists do). We argue that morality is real yet socially constructed, like money or the boundaries of a nation, not like numbers (allegedly) or quarks.
But now for the hard part. What is the nature of morality such that it is social, ontologically subjective, epistemologically objective, and possesses categorical normativity? I contend that morality is constitutive of practical reason (a la Kant), understood as a meaningful social practice (a la Wittgenstein).  Morality refers to the rules “out there” (outside of anyone’s particular subjective beliefs or desires) that make meaningful, rational, social interaction possible. As rational agents, we understand ourselves to act for reasons. Reasons comprise our first order-beliefs and desires that are accompanied by higher-level principles with which we endorse those first-order beliefs and desires. Reasons turn mere events into actions. For the purposes of explicating my account of normativity, let’s call the activities involved in human action the reasons-giving game.
In order for anyone to participate in the reasons-giving game, one must first presume that the interlocutors, including oneself, are agents (i.e. reason-responsive). My utterances and bodily movement only convey meaning under the presumption of a shared linguistic community of reason-responsive agents. If I fail to make this presumption of agency, reasons disappear, and we’re left with flailing and noise. Therefore, within any meaningful, rational linguistic practice lies the implicit rule to treat oneself and others as capable, reason-responsive agents. In other words, if I’m participating in the reasons-giving game, I must treat myself and others as ends-in-themselves (creatures with the reflective capacity to endorse certain beliefs and desires and not endorse others). Failure to do so undermines reason-giving itself. Just as moving one’s pawns all over a chess board willy-nilly undermines the movements as chess-moves, disregarding others as having reasons and ends of their own undermines my own behavior as meaningful, reasonable action. If one is committed to continuing a game of chess, one must acknowledge objective rules that give meaning to certain movements but also constrain one’s movements; if the rules were not objective then each player could use their own subjective rules; but that wouldn’t be chess, nor would it be much fun. Likewise, if one is committed to reason-responsive rationality (personhood), then one must acknowledge even deeper, more general, objective rules that constrain behavior but provide for the logical possibility of meaningful action in the first place.
But for moral realism to obtain, the moral principles “out there” must have categorical normative force; if they aren’t categorical, we merely have hypothetical, culturally relative truths. One might agree that the rules of chess objectively apply to and constrain behavior for those assenting to play chess, but also observe, correctly, there’s no rule that compels one to play chess in the first place. The normativity of chess, therefore, is hypothetical. Its force only applies when one chooses to bind oneself to the rules of chess. For morality on the other hand, the force, allegedly, applies to everyone regardless of their particular desires or commitments. I maintain that it’s not because there’s an external force that compels one to play the reasons-giving game; rather, the categorical nature of morality arises out of the difficulty of opting out of the reasons-giving game. Anyone participating in the reasons-giving game, like chess, is subject to its constitutive rules. The key difference is that for someone to opt out of the reasons-giving game is for someone to opt out of one’s personhood, understood as rational, meaningful agency (i.e. acting from reasons). The normative force of morality, then, comes on pain of irrationality and meaninglessness.
I suppose one could say this account still merely gives us a hypothetical imperative: individuals are only compelled to follow moral rules insofar as they commit themselves to the reasons-giving game. But that retort misunderstands the categorical claim made by morality. No one suggests that moral obligations are truly universal in the sense that they apply to rocks, trees, gerbils and galaxies. They’re categorical in the sense that they apply to all rational agents. For individuals to truly opt out of the reasons-giving game, their lives would need to mimic the characters from Lars von Trier’s The Idiots. These individuals are not responsive to reasons; their utterances and motions do not convey meaning; they’re pretending not to be rational agents. If they were truly successful in their idiocy, they would fall outside the kingdom of ends (i.e. they would not be moral agents). But the pretense is difficult to maintain. Fully functional humans are, indeed, rational agents capable of responding to reasons. We are, therefore, already subject to the normativity of morality because the rules of morality are the most general rules constitutive of reason-giving.
Now this account of morality is quite abstract and may appear to be too far removed from practical decision-making to actually serve as a practical guide (although it rules out genocide, slavery, Jim Crow, etc). But this is where actual politics and Rawls’ idea of public reason enter the picture. In order to fill in more precise moral obligations, we actually have to express our agency and deliberate together, constrained by the principles of our moral equality established by the categorical imperative already discussed. The results of deliberations under these (approximate) conditions then become objective, real obligations. Moral truths are constructed (like all of the social world) out of reasonable agents attempting to find principles that all similarly situated agents can endorse, or at least that no other similarly motivated agent could reasonably reject (Scanlon). Despite their construction, moral principles are as real as money and the rules of chess.  But unlike chess, moral rules apply to all similarly situated rational agents. In essence, morality is really “out there” and discoverable through a combination of observation, rational reflection and deliberation.
Jared Yarsevich has a MSc in Philosophy from the London School of Economics, a Masters of City and Regional Planning from Georgia Tech, and a BA in Economics from North Carolina State University. He lives in Atlanta, Georgia.
 My account of normativity is highly influenced by Christine Korsgaard’s The Sources of Normativity, yet, is distinct from it. For one, she doesn’t claim to be a moral realist (although I think she should).
 If one denies the rules of chess are real, then one must also deny that touchdowns and balls and strikes are real. It would be a commitment to claiming that Nolan Ryan hasn’t ever really struck anyone out or that Deep Blue never really beat Kasparov.