Moral Psychology: An Exchange

To the Editors:

Psychology originated as a branch of philosophy, and for centuries the two areas of inquiry have richly informed each other. Some of history’s greatest moral philosophers have written on the moral implications of “human nature,” “moral sentiments,” and “the crooked timber of humanity,” while psychologists have probed how concepts elucidated by philosophers are implemented in ordinary minds. As psychologists we have continued to profit from this dialogue, particularly since we have found philosophers to be the best reasoners in the academy: the scholars most likely to read carefully, summarize accurately, respect distinctions, and expose fallacies.

As such we were surprised to read a long essay in the February 25 issue of The New York Review by Tamsin Shaw (a professor of European and Mediterranean studies and philosophy at NYU) which lumped five recent books in psychology (including one by each of us) with two reports on the CIA’s 2001–2006 program of torturing detainees. In defiance of the best philosophy, Shaw asserts that psychological and biological facts are “morally irrelevant” and “can tell us nothing” about moral propositions. She insinuates that psychologists, corrupted by their current theories, lack “a reliable moral compass” that would equip them to oppose torture. And she prosecutes her case by citation-free attribution, spurious dichotomies, and standards of guilt by association that make Joseph McCarthy look like Sherlock Holmes.

The trouble begins with Shaw’s repeated assertion that psychology claims “special authority” over morality. In fact, Shaw can cite no psychologist who claims special authority or “superior wisdom” on moral matters. In projecting this ambition onto five books which merely discuss recent research on the moral sense, Shaw apparently cannot conceive of any contribution of psychology to moral philosophy that falls short of outright takeover: the only acceptable contribution, on her reckoning, is zero.

Shaw thus repeatedly asserts that researching the moral sense is tantamount to claiming to be an oracle of moral truth. She then educates us: “It is a fallacy to suggest that expertise in psychology, a descriptive natural science, can itself qualify someone to determine what is morally right and wrong.” It is indeed a fallacy. That is why Pinker wrote, in a section on morality in the book Shaw claims to have read:

The starting point is to distinguish morality per se, a topic in philosophy (in particular, normative ethics), from the human moral sense, a topic in psychology.

It is why Haidt, near the end of his book, wrote:

Philosophers typically distinguish between descriptive definitions of morality (which simply describe what people happen to think is moral) and normative definitions (which specify what is really and truly right, regardless of what anyone thinks). So far in this book I have been entirely descriptive.

Haidt then offered a definition of “moral systems” that he said “cannot stand alone as a normative definition,” but that might be useful as an “adjunct” to philosophical theories.

In a similar vein, Damon and Colby write:

A moral psychology cannot entirely avoid questions of what should be (questions that philosophers call prescriptive or normative)…. This kind of analysis…relies on philosophical argumentation rather than empirical validation.

Joshua Greene’s book, for its part, has a section entitled “Does Science Deliver the Moral Truth?” which lays out his answer: No. And when we asked Paul Bloom whether his book contained a pithy quote that acknowledged the distinction, he replied, “The fact that one cannot derive morality from psychological research is so screamingly obvious that I never thought to explicitly write it down.”

In fact it is Shaw’s insistence on the complete irrelevance of moral psychology to normative questions that is a departure from centuries of moral philosophy, and from the practice of the many contemporary philosophers who are avid consumers of and contributors to psychology. It has long been recognized that findings about moral sentiments, even if they don’t determine the truth of moral propositions, are highly relevant to philosophical inquiry about them. Utilitarianism alludes to the capacity of humans (and animals) to reason, suffer, and flourish. Virtue ethics hinges on traits of character. Kant, in explicating his deontological theory, famously wrote that “the action to which the ‘ought’ applies must indeed be possible under natural conditions.” As each of these theories is elaborated, it will necessarily make contact with assumptions about cognition, affect, personality, and behavioral flexibility—the subject matter of psychology.

Recent discoveries in moral psychology offer another point of contact. Many ethical convictions are underpinned by strongly felt intuitions that some action is inherently good or bad. Sometimes those intuitions can be justified by philosophical reflection and analysis. But sometimes they can be debunked and shown to be indefensible gut reactions, without moral warrant. Historical examples include outrage over heresy, blasphemy, and lèse-majesté, revulsion against homosexuality and racial mixing, squeamishness about medical advances like vaccination and blood transfusions, callousness toward slaves and animals, and indifference or hatred toward foreigners. Any news reader will confirm that some of these historical examples are all too modern.

Psychology, neuroscience, and evolutionary biology, though they cannot by themselves debunk moral intuitions, are highly relevant to evaluating them, since our naive intuitions about the ethical warrant of our gut feelings might themselves be gut feelings and cannot be taken at face value. Haidt’s research on moral diversity within and across cultures suggests that moral sentiments can attach to a wide range of concerns, including authority, purity, loyalty, and conformity, which can spark the same self-righteous conviction and punitive sentiment that liberal Westerners expend on the ideals of beneficence, freedom, and fairness. Greene has shown that some of these moral sentiments are rooted in automatic and ancient circuits of the emotional brain and that the related intuitions are sensitive to factors that nearly everyone regards as morally irrelevant. Discoveries about human nature like these are far from obvious, and only those imprisoned in a disciplinary silo could insist that they are utterly irrelevant to the challenge of assessing which moral intuitions we should discount.

What about normative moral reasoning itself? Shaw writes that “all of those I discuss here are making claims about which kinds of moral judgments are good or bad by assessing which are adaptive or maladaptive in relation to a norm of social cooperation.” Actually, none of them do. Shaw appears to have misunderstood the technical term cooperation, which biologists define as “the process where two or more organisms act together for mutual benefit.” Cooperation gets a lot of attention in psychology because it is rare among nonrelated animals but common in humans, and because the delicately balanced risks and advantages of cooperation help explain how moral sentiments such as trust, sympathy, gratitude, guilt, shame, and anger might have evolved. But the term is a rubric for a family of morally relevant behaviors, not the criterion by which behavior is judged to be truly moral.

In any case, Shaw soon contradicts herself by noting that Pinker in fact invokes “the interchangeability of perspectives” as a source of morality. She dismisses this, too, as being “oddly restrictive” and distinct from “any higher-order philosophical theory.” Yet the sentence following the one she quotes elaborates:

This foundation of morality may be seen in the many versions of the Golden Rule that have been discovered by the world’s major religions, and also in Spinoza’s Viewpoint of Eternity, Kant’s Categorical Imperative, Hobbes and Rousseau’s Social Contract, and Locke and Jefferson’s self-evident truth that all people are created equal.

In Shaw’s understanding, these “oddly restrictive” principles must not count as “higher-order philosophical theories.” She also fails to note that the nature and history of the ideal of human rights is a major theme of the book.

From these slapdash summaries Shaw segues to her innuendo about psychologists’ acquiescence to torture:

So in addition to questioning whether psychological research can vindicate moral norms, we also have to ask whether the minimal moral norm of cooperation employed by psychologists is sufficient to provide them with a reliable moral compass.

Recent developments in the profession of psychology have been discouraging in this respect.

Yes, that is the transition: psychologists write a lot about cooperation, so perhaps they can’t figure out that torture is wrong. The next few pages review the Hoffman report, initiated by the American Psychological Association to investigate the role the APA played in the years after the September 11 attacks, when a few members of its leadership cooperated with the US military to widen the scope of permission for psychologists to advise the military on “enhanced interrogation,” otherwise known as torture. These members do indeed seem to be guilty of subverting normal process to get the APA to approve wider permission than its membership would ever have knowingly allowed. Two psychologists worked closely with the military on these programs. Such a lapse in accountability and ethics may be an organizational hazard of a guild as sprawling as the APA, with its 80,000 members and associates, five hundred staff, and fifty-six divisions (embracing everything from mathematical psychophysics to couples therapy). When the news broke about this collusion, most APA members, together with its current leadership, were horrified. Nonetheless, Shaw smears the entire “profession of psychology” as being implicated in the collusion.

Shaw’s intimations about the authors of the five books she discusses are even more tenuous.

1. Guilt by imaginability. Shaw quotes a passage from the Hoffman report about how some psychologists, upon learning about the APA’s collusion, felt physically sick. She then writes:

It is easy to imagine the psychologists who claim to be moral experts dismissing such a reaction as an unreliable “gut response” that must be overridden by more sophisticated reasoning. But a thorough distrust of rapid, emotional responses might well leave human beings without a moral compass sufficiently strong to guide them through times of crisis, when our judgment is most severely challenged, or to compete with powerful nonmoral motivations.

Shaw is right: it is easy to imagine that. But does that make it true? We prefer consulting the world rather than our imaginations, so we polled the psychologists mentioned in the article on whether they would dismiss a visceral reaction to torture as an unreliable gut response. The results: seven out of seven said “no.” None of these psychologists believes that a reaction of physical revulsion must be overridden or should be thoroughly distrusted. But several pointed out that in the past, people have felt physically sick upon contemplating homosexuality, interracial marriage, vaccination, and other morally unexceptionable acts, so gut feelings alone cannot constitute a “moral compass.” Nor is the case against “enhanced interrogation” so fragile, as Shaw implies, that it has to rest on gut feelings: the moral arguments against torture are overwhelming. So while primitive physical revulsion may serve as an early warning signal indicating that some practice calls for moral scrutiny, it is the “more sophisticated reasoning” that should guide us through times of crisis.

2. Guilt by association. With evident relish Shaw traces out a web of connections among moral psychologists and Martin Seligman. Seligman developed “learned helplessness theory” in the 1970s and went on to found Positive Psychology in the late 1990s. The two psychologists whose consulting firm advised the military on torture drew on Seligman’s ideas about learned helplessness. The Hoffman report found no evidence that Seligman knew his research was being used in this way. But for Shaw that doesn’t matter; by her standards, Seligman is associated with torture. And as her careful research shows, each of us is associated with Seligman. Bloom and Haidt teach one weekend a year in a Positive Psychology course. Pinker served on an advisory committee for a Seligman project. Seligman blurbed Colby and Damon’s book. Seligman is one of the most prolific and creative psychologists of all time; we are pleased to be associated with him, as are thousands of people inside and outside of psychology who have used his ideas, sought his counsel and support, or collaborated with him on projects. What exactly is Shaw trying to say by plotting a web of associations, one of whose nodes is “torture”? This style of insinuation is, to use her words, “not altogether appealing.”

3. Guilt by tweet. Shaw quotes extensively from the Senate Select Committee on Intelligence report, including sections so disturbing that some readers might request a trigger warning at the front of her essay. It’s not clear what any of this has to do with moral psychology until Shaw unveils her smoking gun:

When the Senate Select Committee on Intelligence published its extensive report on official torture in December 2014, Jonathan Haidt tweeted a link to an article by Matt Motyl, his former Ph.D. student, claiming that the report would not change anyone’s views on the morality or effectiveness of torture, owing to the phenomenon of cognitive bias, which distorts people’s assessment of the relevant evidence. Motyl warned that none of us should assume that our beliefs about torture are based on facts. Nevertheless, there are established facts. One of them is that psychologists secured enormous financial gains by collaborating in official torture, while also having clear evidence that it was ineffective.

It is true that Haidt sent out a tweet with the word “torture” in it, directing readers to Motyl’s Time essay which explained how, in our politically polarized country, the report was unlikely to change many minds on either side. Though Motyl did explain how people on all sides of a controversy are apt to believe that only their side has the facts, he did not, as Shaw implies, try to deny the factual basis of the report. And if Shaw wants to trawl through tweets to expose her defendants’ underlying moral beliefs, she could have quoted this retweet from Haidt (originally by Hend Amry): “‘Does torture work?’ as a question needs to be put in the bin with ‘Is slavery commercially feasible?’ and ‘Can genocide help overpopulation?’”

4. Guilt by shared methods and standards. Shaw has uncovered two damning coincidences involving Haidt and the US military. She points out that the military has used cognitive behavioral therapy and resilience training, and that Haidt has advocated for these skills, too (in an Atlantic essay with Greg Lukianoff). She adds that “the military is concerned about a ‘very liberal if not extremely leftist’ orientation among academic psychologists,” and so is Haidt! As she puts it: “Jonathan Haidt has repeatedly decried the lack of conservatives in the profession of social psychology…. His priorities appear to align closely with those of the Department of Defense.”

Align closely? Haidt, together with five other social psychologists, has documented a loss of political diversity in social psychology and the well-known dangers that this kind of orthodoxy can produce, including bias, error, and groupthink. Shaw misreads the article as being about a “lack of conservatives.” In fact the authors are not making a political argument (none of them is conservative) but a methodological one: that any field of social science with a monoculture of political belief is intellectually unhealthy. By tendentiously aligning this concern with the “priorities” of the Department of Defense, Shaw is implying that diversity of opinion is somehow suspicious, something that only the military cares about, as if everyone else should be content with unexamined conformity.

Shaw accuses moral psychologists of being “dismissive of philosophical theories,” but she herself seems dismissive of philosophical norms and standards. Her tissue of misattributions, confusions, and smears amounts to conduct unbecoming a philosopher.

Jonathan Haidt
Stern School of Business
New York University
New York City

Steven Pinker
Department of Psychology
Harvard University
Cambridge, Massachusetts

Tamsin Shaw replies:

Moral psychology is an invaluable aspect of human understanding insofar as it sheds light on the moral capacities and limitations of human beings. And this fact has indeed long been appreciated by philosophers (perhaps by none so much as Friedrich Nietzsche, whose writings have been the primary focus of my own scholarly work). The findings of moral psychology have also begun to find a place in the public imagination, via prominent editorials and more popular psychology books. But current research by psychologists in this area has risen to prominence at the same time as an extraordinary moral crisis in their profession, a fact that inevitably lends their reflections a special significance that requires scrutiny.

The recent Hoffman report, commissioned by the American Psychological Association, which relates to ethics, interrogations, and torture, revealed both devastating moral failings by a few individual psychologists and a more general vulnerability in broader sections of the profession. Although the vast majority of the APA’s members were not involved in the defense, design, or implementation of torture and several had worked strenuously to expose what was going on, it was clear that the leadership of their professional association had betrayed its fundamental values, including the principle of doing no harm. Two CIA psychologists responsible for actually inflicting torture on detainees, James Mitchell and Bruce Jessen, have been the principal focus of the media’s coverage. But the Hoffman report draws attention to the involvement of several very senior academic psychologists.

Since Jonathan Haidt and Steven Pinker insist that former APA president Martin Seligman was not himself found to have participated knowingly in the torture program, it is worth setting out briefly what the report in fact tells us about his involvement. Although Seligman was one of only three witnesses out of 148 who refused to speak directly with Hoffman’s investigators, demanding instead that they send him questions in writing, they were able to establish the following.

In December 2001, Seligman convened a meeting at his home to discuss the participation of academics in national security efforts following September 11. Among those present were CIA psychologist James Mitchell and the chief of research and analysis in the CIA’s Operational Division, Kirk Hubbard.

Unlike Seligman, both Hubbard and Mitchell agreed to be interviewed. Seligman claimed to remember meeting with Hubbard on one subsequent occasion at his home, in April 2002, to discuss his theory of “learned helplessness” with Hubbard and a female lawyer, and that on this occasion he was invited to speak on the theory of learned helplessness at the Survival, Evasion, Resistance, and Escape (SERE) school, sponsored by the US government. Hubbard, however, recalled meeting Seligman at his home several times after the initial meeting, including a meeting in April 2002 at which, according the Hoffman report, “he, Mitchell, and Jessen met with Seligman in his home to invite him to speak about learned helplessness at the SERE school.”

Seligman later conceded that the meeting must have taken place and that it was on this occasion that he was invited to speak at the SERE school. According to the timeline set out in the extensive report by the Senate Intelligence Committee on torture, the meeting with Mitchell and Jessen must have taken place shortly before Mitchell flew to Thailand to assist in the interrogation of Abu Zubaydah using torture techniques derived, according to the Hoffman report, from Seligman’s theory of learned helplessness. The extent of Seligman’s further involvement has not been established, but in an e-mail sent by Hubbard in 2004, he expressed gratitude for Seligman’s help “over the past four years.”

The Hoffman report twice states that Seligman’s denial of any suspicion that the CIA’s interest in his theories was for use in interrogations is not credible. Plausible deniability here is strained by the fact that the person in question is an expert on human behavior employed by the military. A failure to imagine the likely consequences of his actions would seem to call into question his qualifications for that position.

But of greater significance than individual moral failings is, for many APA members, the fact that the Hoffman report exposes the way in which, more broadly, financial incentives generate troubling conflicts of interest within the profession. In response to the report, the organization Psychologists for Social Responsibility made this statement:

In light of the evidence compiled by the Hoffman investigative team, psychologists have a duty to recognize and examine how decades of dependency on the military and intelligence agencies for employment, funding, and stature have influenced our entire profession, in ways that have never been adequately examined.

It is therefore in the light of these concerns that the dominant discourse on morality within the profession of psychology, laying claim to that profession’s expertise, must be examined.

I noted in my article that much of the recent literature on moral psychology is influenced by the school of Positive Psychology, founded by Martin Seligman. According to Michael Matthews’s book Head Strong: How Psychology Is Revolutionizing War, the school of thought associated with Positive Psychology has come to be especially prized by the military in the “global war on terror.” And indeed Seligman has received a $31 million contract from the Department of Defense for his Positive Psychology Center at the University of Pennsylvania, as well as being placed in charge of the military’s $125 million Comprehensive Soldier Fitness Initiative. The point of drawing attention to the links between prominent figures in the field of moral psychology and the Positive Psychology movement is not to impute blame for torture where it does not belong, or to imply guilt by association with Seligman, but to ask the necessary questions about whether their approach to morality is conducive to the kind of rigorous moral self-reflection that both the authors of the Hoffman report and many APA members deem necessary.

The links between the authors I discuss and Positive Psychology are not tendentious or personal, but reflect deep intellectual debts and affinities that deserve deeper explanation than they have had so far. Jonathan Haidt, in particular, has since the inception of this movement stood shoulder to shoulder with Seligman in what they clearly both take to be a project of moral regeneration. In his 2006 book, The Happiness Hypothesis, Haidt describes an American culture that has, since the social turmoil of the 1960s, “lost its way,” and he tells us that one of Seligman’s first tasks in launching Positive Psychology was to discover the positive human qualities that promote moral development. In 2003, Haidt and Corey Keyes edited a volume with a foreword by Seligman called Flourishing: Positive Psychology and the Life Well Lived. In his own contribution, on “Elevation and the Positive Psychology of Morality,” Haidt stresses that Positive Psychology can bring about “a balanced reappraisal of human nature and human potential” by studying the roots of goodness.

Pinker and Haidt insist that they and all the other authors discussed are fully aware that the descriptive science in which they are engaged cannot in itself offer solutions to moral questions. They nevertheless claim in their letter that psychology, neuroscience, and evolutionary biology, though they cannot debunk moral intuitions, are “highly relevant to evaluating them.” But their notion of “relevance” is perplexing. They tell us:

Many ethical convictions are underpinned by strongly felt intuitions that some action is inherently good or bad. Sometimes those intuitions can be justified by philosophical reflection and analysis. But sometimes they can be debunked and shown to be indefensible gut reactions, without moral warrant.

Pinker and Haidt themselves point out that we have often, throughout history, come to distrust our gut instincts about moral questions: “revulsion against homosexuality and racial mixing” or “callousness toward slaves and animals,” for example.

But these large-scale moral corrections were made without the help of fMRIs or data from psychological questionnaires. The work was done by moral deliberation in the light of which our gut instincts came to seem inappropriate. And in any case of moral judgment this must be true: the appropriateness or otherwise of our gut instincts about a given phenomenon must be determined by our moral reasoning about that phenomenon. Revulsion seems like a bad response to homosexuality and racial mixing. It seems like an appropriate response to child abuse. It has not been shown that multimillion-dollar research projects in psychology and neuroscience can themselves contribute some additionally relevant insight, in the light of which we will be better equipped to assess our gut instincts or evaluate the truth of our moral intuitions.

This lack of clarity at the heart of the works in moral psychology under discussion in my review serves, whether intentionally or unintentionally in each case, to generate the false impression that scientific expertise has important weight in making moral judgments, and that behavioral science can provide us with a guide to moral development. In circumstances in which psychologists are being called upon by their own colleagues, among others, to address very serious moral questions, they have a special responsibility to be clear about any claims they appear to make to moral authority. My review questioned whether they have fulfilled this responsibility.

There are also many other important questions to be asked about the specific doctrines in Positive Psychology that these moral psychologists have embraced: their insistence on optimism about the future of humankind, which other schools in psychology hold to be strongly correlated with indifference to suffering; their distrust, whether “thorough-going” or not, of moral emotions such as empathy; their promotion of a psychological model of resilience that includes a lack of self-blame. There is important analytic work to be done in ascertaining the relationships between purportedly moral and nonmoral practical ends, such as “resilience,” in this movement. And that work should involve addressing the likely practical ends to which their research will be put by those who are funding it. Pinker and Haidt say they prefer reality to imagination, but imagination is the capacity that allows us to take responsibility, insofar as it is ever possible, for the ends for which our work will be used and the consequences that it will have in the world. Such imagination is a moral and intellectual virtue that clearly needs to be cultivated.

I have not addressed specifically many of the ways by which Pinker and Haidt distort and misread the claims I make in my article, since I believe that any careful reader will be able to see them for themselves. As to their claim that I have drawn connections with “evident relish,” that is a groundless psychological assumption. I can only assure them that there was not a single moment during my writing and researching of the review that I felt any such pleasurable or positive emotion.