The Charm of Harm

Earlier in the 21st century, I asked two PhD students, who were defending their work on moral judgment, what their descriptive findings might tell us about normative models of morality. They both dodged the question, seeking shelter behind a veil of ignorance. We don't care, they seemed to say, because we cannot know. Normative morality is the province of the Rawlses and the rabbis of the world. We empirical researchers adhere to the categorical imperative of not telling people what not to do.I was unconvinced. Work on rationality got off the ground by setting up normative models of how humans should think and then studying how they fail to do so (Dawes, 1988; Krueger & Funder, 2004). Normative models of rationality promise to improve human judgment; without them, descriptive research on human judgment floats in a vacuum. The study of moral judgment poses an analogous challenge. Purely descriptive research faces the so-what question. Any morality appears to be as good as any other. Without normative benchmarks, why bother?In Outraged: Why We Fight About Morality and Politics, and How to Find Common Ground, Kurt Gray declares his commitment to description. Let the moral philosophers struggle with the normative questions, as they have for 2,500 years. They have not found consensus, and it seems they never will. This view suggests that one reason for the commitment to the descriptive project is that all normative models are suspect.Yet the normative impulse is hard to banish from psychological work. In moral psychology especially, normative inference and insinuation lie close to the surface. And so it is in Outraged. Readers learn that their morality is deficient and that help is on the way. Gray offers ways to improve moral judgments; ergo, he makes normative claims.Gray grounds his approach in evolutionary psychology, which treats our moral sense as an adaptation, that is, a set of psychological states and behaviors that has been favored by natural selection. Nature, to be sure, has no moral intent, but the humans she has produced do, and we can therefore ask whether our evolved moral judgmental apparatus is equal to the task in the contemporary world. Gray is among the many who see trouble. Much moral judgment, as it emerges in descriptive research, has degenerated into moralistic judgment. People these days are outraged and ready to punish at the drop of a hat. Gray wishes for a gentler humanity, which is of course a normative matter, if he could only admit it.If it is the case that we, as a society and perhaps as a species, have become prone to outrage, then we must seek an explanation, and perhaps a remedy. Gray finds the root cause of moral outrage in the perception of harm. Indeed, he claims that perceived harm is the master key to all of morality. Outraged is written to defend this claim. Gray's argument is that his harm-only hypothesis is right because Haidt's multiple foundations theory is wrong. According to multiple foundations theory, some moral judgments do not depend on harm (Haidt, 2007). People are disgusted by and they condemn incest even when all harm is carefully removed from the description of such an episode. Gray objects that people will perceive at least a potential for harm, no matter how diligently objective harm has been removed.The move from objective to subjective harm is a master stroke, but it is also the Achilles heel of Gray's theory. The elevation of subjective harm is a master stroke because it grounds all moral judgment in one common currency. If, and only if, there is perceived harm will the moral module be activated, outrage be felt, and punishment be sought. This if-and-only-if proposition rises to the status of an axiom; it becomes true by assertion and immune to falsification. And there is the Achilles heel.Axioms cannot be falsified. They can only be abandoned when other axioms come along that do a better job grounding accurate and useful theories. Meanwhile, an axiomatic if-and-only-if proposition does more for explanation than for prediction (Krueger, 2017, 2020). In the case of Gray's harm theory, a person's or a group's moral outrage and thirst for punishment raise the question of where the perceived harm lies. When it is found, we can explain moral affect and action. For example, we may find that people find safe-sex incest repugnant because they cannot believe that the probability of later memories turning traumatic is exactly zero. Incidentally, a complication of this scenario is that the permissibility of self-harm, and especially suicide, is a hotly debated normative question. If Gray does not distinguish self-harm from harm to others, he is taking a normative position, if only implicitly.The harm theory's Achilles heel is that although it can claim to explain all outrage once it has occurred, it cannot predict who will be outraged by what event without referring to a wealth of psychological, sociological, and anthropological extratheoretical background knowledge. Consider an episode that involved Gray himself. In 2022, the editor of Perspectives on Psychological Science was forced to resign three days after 1,375 psychologists called for his dismissal in an open letter. Gray was signatory number 954. The tone of the letter was moralistic, ad hominem, and not based on evidence beyond the self-reported distress of the purportedly injured party (see Jussim, 2022, for a summary and the time line of the episode). The point here is not to comment on the merits of the accusation or the defense, which also came in the form of an open letter signed by a multitude, but to note that the author of Outraged, like many others, joined an agitated collective in a state of moralistic fury, when there was a countercollective that was just as outraged. If the point of Outraged is that there is too much polarizing outrage because people rely too readily on their evolved capacity to moralize, then the moral imperative is to search for ways to lower the psychological temperature. This, indeed, is the tenor of many moral teachings. Just as a point of reference, imagine His Holiness The Dalai Lama as an outraged member of a lynch mob. It does not compute.Outraged has merits. It opens with the claim that humans have evolved as prey animals and that our status as apex predators is a recent development, too recent to undo the deep-rooted fears of a creature that is hunted and eaten, that is, a creature that is victimized. Becoming interdependent group animals solved many of the prey animal's problems—at a price. We humans had to become wary, if not fearful, of what conspecific outgroups and even friends and family might do to us. Morality evolved as a solution to these new threats. Evolution often cuts with a dull blade, and so type I errors (false alarms) have survived in greater numbers than type II errors (misses) because mistaken perceptions of harm turned out to be less costly than overlooked harms—at least to those doing the punishing. The overgeneralization of harm perception is rewarded, but when there is no brake, morality degenerates into moralism. At the end of this road, the thirst for moral punishment not only permits acts of violence but takes them to be virtuous.Outraged captures the threat of moralism with the ideas of concept creep and moral typecasting. Gray attributes concept creep to the layperson who perceives ever more threats, by redefining what a threat is, when true threats are actually diminishing. Note that Gray does refer to objective reality here, while putting all his theoretical money on the subjectivity of fear. The presumed negative correlation between the actual probability of harm and its subjective likelihood is the product of what Gray calls the safety transition. To make an a fortiori case for the counterintuitive idea that as a category actually shrinks, it will be perceived as growing, Gray reviews research by Levari et al. (2018), who found that when respondents looked at displays of light, gray, and black dots, they categorized the gray dots as dark when there were no black dots, thereby preserving and enlarging the nonwhite category. This illustration is a fortiori because it only requires theories of psychophysical judgment to explain the findings (Parducci, 1968; Wixted, 2020). The elements unique and critical to Gray's theory—the perception of harm, the need to punish, and a temporal trend toward greater objective safety—are unnecessary.Gray's own concept creep, we suspect, lies in his determination to find perceived harm if moral outrage has occurred. If harm sits at the foundation of all morality, high moral deeds are problematic. The only harm they allow to be perceived is the harm that would occur if the moral deeds were withheld. Maimonides praised visiting the sick as the most praiseworthy act—a super-mitzvah, as it were. Failing to visit the sick leaves the world in a sadder state, hence it is immoral. But if we can find the time to visit the sick, we can find the time to visit two, or three, or more. Whatever good we do, we could do more, and thus we remain morally guilty. Schindler, who felt this way about the limited length of his list, had to be reassured. He was honored for what he had done, not condemned for what he left undone. The harm theory fails to explain such important nuance.What about moral typecasting? This is an intriguing and troubling phenomenon by which people see others—though not so much themselves—as either all good and moral or all bad and immoral. This type of simplistic categorization is also familiar from psychophysical research (Krueger, 1992; Tajfel, 1969) and certainly from the Abrahamic scriptures, which distinguish between the righteous and the wicked, the saved and the damned, the goats and the sheep. Dante Aleghieri made finer distinctions, but only among the good and among the bad; he did not seem to be interested in mixed types. If all-or-nothing categorization is a heuristic of limited use and immense risk, then perhaps another heuristic can step in to help. If people relied more on social projection when judging others, they might make more room for moral complexity (Krueger, 2007). As they themselves are bearers of both beneficent to noxious inclinations, so are—they will recognize with projection—most others.Chief among the concerns about the harm-as-the-moral-master-key theory is its claim to exhaustiveness. If there is moral judgment, there must have been harm; if there is harm, there will be moral judgment; so says the theory. Gray briefly discusses some difficult cases, but only to dismiss them. For example, many people find it immoral to disrespect the dead or the gods. According to the theory, there must be a perception of harm. If harm requires sensation, the dead cannot be harmed, and the gods’ feelings are indeterminate. But members of the same congregation, religious affiliation, or subculture can take offense and claim to be harmed because they believe that the dead or the gods can be harmed.According to the theory, this is enough to ground moral judgment, but we see the problem with the theory's elasticity. Besides, people also actively benefit the memory of the dead by giving eulogies and building monuments. These acts are often seen as morally good, but where is the harm avoidance? In Eva Luna, Isabel Allende (1987) describes in great detail the efforts of Walimai, an indigenous Amazonian man, who kills an abused young woman of his tribe out of mercy and then follows an elaborate religious protocol to set her soul free. The power of this narrative is that it impresses Walimai's moral goodness even on readers who are culturally as far removed from Walimai's world as human diversity will allow.Outraged appropriately asks how human relations can be improved by reducing moralism. Here, the master key is nonconfrontational communication. To illustrate how it is done, Gray relies mainly on Foss and Griffin's (1995) “feminist” approach to gentle persuasion. They call their method invitational rhetoric because it seeks understanding instead of domination, yet attitude change may result precisely because the other party does not feel pressured. Pascal (1658/2015) had much the same idea. He (see epigraph) knew that belief is grounded in perception. Belief change is best achieved when the target of persuasion concludes they have always seen things this way, that is, when change occurs while being imperceptible to the person. Haven't we always known this to be true?