In Defense of Moral Non-Cognitivism, a Response to Neil Shenvi
I found Neil Shenvi’s interesting article entitled “Do Objective Moral Values Exist?” while reading Dangerous Idea. In his article, Shenvi argues for the position that objective moral values exist, which he defines as “moral values that are true independent of the beliefs of human beings”. This piece started out as a comment on the blog, but I quickly realized that in order to properly address it, I would need quite a bit more space. In arguing for this position, Shenvi puts forth five pieces of evidence:
- Nearly universally across human cultures, there exist the same basic standards of morality. In addition, there exist in all cultures truly altruistic acts which lead to no personal or genetic benefit.
- The majority of people who explicitly deny the existence of objective morality still act as if objective morality exists.
- There exists a nearly universal human intuition that certain things are objectively right or wrong.
- The majority of philosophers recognize the existence of objective moral facts.
- Many naturalists (like Sam Harris or Shelley Kagan) affirm the existence of objective moral facts, despite the problems inherent in grounding these facts in the natural world.
Before I set out to evaluate Shenvi’s argument, I’d like to say that I’m not settled one way or another on the question of whether or not moral facts exist. Presently, if I were forced to give an answer, I would argue towards a non-cognitivist understanding of moral statements which is motivated by other epistemological and metaphysical commitments. This is important because I think Shenvi only recognizes moral relativism and moral realism without taking seriously other anti-realist positions, such as error-theory, quasi-realism, and non-cognitivism.
Shenvi asks us to consider the five points above against the theoretical backdrop of two possibilities: (1) moral facts exist and we have immediate, intuitive apprehension of their existence, and (2) moral values do not exist and any belief that they do exist is therefore an illusion. If we take seriously the other types of moral anti-realism that exist, we can see that these two possibilities are not exhaustive of the moral positions, and in fact Shenvi’s second possibility is a weak and unappealing caricature of what many anti-realists believe.
Many anti-realists don’t think that people expressing moral claims are simply uttering nonsense based on an illusion. There are, of course, a few other possibilities. Perhaps objective moral values don’t exist, but our proclamations of “murder is wrong” are actually prescriptive statements meaning “Do not murder” (prescriptivism, a version of non-cognitivism). Or, maybe our moral sentiments are expressions of our own attitudes towards moral rules, or norms. (Norm-expressivism) These other varieties of anti-realism still give us a way to interpret moral claims, but they argue that moral statements are not propositions that are either true or false.
This may seem like philosophical jibber-jabber, but it’s an important distinct. Relativism is fraught with meta-ethical problems, and has been a traditionally difficult position to hold. Anti-realist positions of other sorts, however, often are more consistent within themselves, and with the data that we observe regarding moral sentiments and attitudes.
This may seem like philosophical jibber-jabber, but it’s an important distinct. Relativism is fraught with meta-ethical problems, and has been a traditionally difficult position to hold. Anti-realist positions of other sorts, however, often are more consistent within themselves, and with the data that we observe regarding moral sentiments and attitudes.
In this essay, I’d like to argue that a non-cognitivist understanding (informed by naturalism) better explains Shenvi’s five pieces of evidence than does moral realism. Additionally, I’d like to introduce difficulties that may have previously gone unnoticed with these pieces of evidence in relation to his own position.
Let us turn to address Shenvi’s evidence. His first point is that “Nearly universally across human cultures, there exist the same basic standards of morality. In addition, there exist in all cultures truly altruistic acts which lead to no personal or genetic benefit.” Shenvi argues that across human history, cultures around the world have largely agreed on basic moral truths. Actions like murder, theft, stealing, lying, and some others have been widely declared wrong. However, I think we can easily understand how these things would also flourish under an evolutionary perspective. Human cultures and societies that promoted these actions as laudable would, as Shenvi notes, perish rather quickly.
Shenvi further argues that “true altruism” (which Shenvi defines, a la Coyne, as “behavior that will not even indirectly confer benefit to oneself or one’s relatives”) is surprising given the facts of biological evolution alone. It’s easy to agree with this assessment, prima facie, as we wouldn’t expect true altruism to increase evolutionary fitness. It would be far too lengthy of an assessment to speculate on possible evolutionary explanations here, but Coyne does a good job of laying out some possibilities here.
The most glaring difficulty with Shenvi’s first piece of evidence is that he papers over the vast majority of moral claims that human history has shown we conclusively do not agree upon. Some examples: slavery, war, genital mutilation, the equality of men and women, abortion, gay marriage, incest, honor-killing, euthanasia, racism, jihad, blood sacrifices, monogamy, polygamy, sexual activity in general, etc. We can look to human history and find an abundance of widespread and vehement disagreement over the moral status of these ideas. Even religions that claim to have found a system of objective values still have widespread disagreement over what exactly those values are.
This is an important difficulty, because Shenvi is arguing that our comprehension of moral values is both immediate and intuitive. However, if we take the issue of slavery to be wrong, it is quite clear that for a great deal of human history, this claim was neither immediate nor intuitive, and in fact most people (even relatively recently!) thought
the practice of slavery was an honorable tradition.
These disagreements and agreements are quite easily understood under the banner of non-cognitivism. If moral claims are merely expressions of prescription or attitude towards norms and rules, we would expect wide-spread disagreement over these matters, especially in lands that are quite disconnected from each other and will have understandably different cultures. As the world becomes increasingly interconnected, we might expect a shared system of wants, desires, and needs to become evident and perhaps our moral claims might further begin to coalesce (think about the Enlightenment’s impact upon western culture generally).
Another difficulty for Shenvi’s first piece of evidence is that it seems that basic moral intuitions, such as fairness positively correlate with norms and institutions rather than persist consistently across the board. From the abstract of a study in Science by Joseph Henrich (emphasis mine):
“Using three behavioral experiments administered across 15 diverse populations, we show that market integration (measured as the percentage of purchased calories) positively covaries with fairness while community size positively covaries with punishment. Participation in a world religion is associated with fairness, although not across all measures. These results suggest that modern pro-sociality is not solely the product of an innate psychology, but also reflects norms and institutions that have emerged over the course of human history.”
If prosocial behaviors were informed by immediate and intuitive objective moral values, it seems we should expect these behaviors to be relatively stable across the board. If, as non-cognitivists might suggest, expressions of moral values are reflections upon norms or prescriptions, it would be easier to understand why societies with pro-social norms and institutions will reflect greater prosocial behaviors. (Compare, for example, early American history with contemporary America regarding slavery, racism, sexism, and other aspects closely related to fairness.)
Let us move to Shenvi’s second point: “The majority of people who explicitly deny the existence of objective morality still act as if objective morality exists.” Here, Shenvi argues that moral relativists still seem to act as if objective moral values exist. Why don’t moral relativists, or moral anti-realists, cheat on their taxes? Or shoplift? He argues that the “vast majority” of relativists lead moral lives, and this doesn’t make sense if they truly believe that cheating on their taxes is neither right nor wrong.
Of course, we can do all of these things without needing to believe that they are objectively right or wrong. While I know Shenvi was intentionally avoiding “You can’t act moral without believing in God”, this response dangerously treads upon it. It may be that our psychology from birth has endowed us with an inclination towards these behaviors. Or, as the above study in Science indicates, our culture’s structure may have influenced our pro-social behavior. We may fear punishment or retribution, or not wish to disappoint those around us who expect moral behavior. There is no shortage of responses here, and we need not posit moral objectivity to explain these behaviors.
His third point is almost identical to his first, and suffers from the same weakness. He argues that individual human beings have an intuition that certain things are objectively right and wrong. Of course, we can also explain this from an evolutionary perspective without any need for objective moral values. As we covered in the discussion of (1), this is not at all clear or obvious. Human beings regularly disagree about a great deal of moral statements, and even ones that seem fundamental.
Shenvi’s fourth and fifth point are made of the same cloth. His fourth piece of evidence argues that a “majority of philosophers accept moral realism” and his fifth argues that even naturalists have argued for moral realism despite problems within the general conception of naturalism. I’m not generally interested in appeals to authority/popularity, but I do feel it’s pertinent to correct his inflated estimation of consensus among philosophers. He argues that philosophers, at a rate of two to one, “favor” moral realism to moral anti-realism. The poll he’s using is a Phil Papers survey, and here are the percentages: 56.3% accept or lean toward moral realism, 27.7% accept or lean toward moral anti-realism, with 15.8% declaring other. This is a pretty weak consensus/majority, as we can see much stronger ones in other domains (atheism – 72.8%, scientific realism 75%, ‘switch’ on the Trolley problem – 68.2%). We also don’t have a notion of how many “accept” versus how many “lean towards” (which is pretty ambiguous language), so I think that using this individual survey as evidence is tenuous at best.
This doesn’t mean that we should totally ignore these points, as obviously many people feel that they have justified reasons for believing these to be true. We should investigate the evidence and reasons they have for believing in moral facts, not merely appeal to the fact that they do.
In summary, I think that non-cognitivism gives us a better explanation of the evidence that Shenvi argues for than does his proposition of moral realism. The “nearly universal” basic standards across cultures is over-stated, while Shenvi largely passes over the vast and immense disagreement over most moral matters. This disagreement is better explained under non-cognitivism, whereas if there was an immediate and intuitive understanding for objective moral law, we wouldn’t expect large scale disagreement about such a vast array of moral values. Finally, an appeal to philosophers and to naturalists who are inclined toward moral realism may be indicative of sufficient epistemic warrant, but it won’t suffice as evidence itself.