Symposium Introduction

Empirical Assumptions and Philosophical Ethics: On Mark Alfano’s Moral Psychology

Mark Alfano’s Moral Psychology is written with the hopes of “putting empirical flesh on the conceptual bones” of normative moral theories (21). Normative ethics, in its various forms, is posited on empirical assumptions about how human beings work. In particular, most theories depend on truths about how we manage our preferences, whether our disagreements are as fundamental as we sometimes feel, whether we have stable moral characters, and how we identify responsibly performed actions. The findings of empirical psychology are relevant to those assumptions, because they can identify how demanding a norm is and what means we have to live up to it, and they can identify whether some of our moral concepts track how we actually make decisions. Alfano modifies the synthetic Kantian dictum, then, to show the mutual relevance of psychology and moral philosophy: “Moral philosophy without psychological content is empty; whereas psychological investigation without philosophical content is blind” (1).

Alfano considers five central components of moral life in light of current psychological literature: preferences, responsibility, emotion, character, and disagreement. In each of these domains, Alfano makes the case that the research in empirical psychology shows that some revision of our normative theories is in order. The issue, then, is the extent of the revision. Alfano’s defaults are to make minimally invasive changes, but these vary depending on the domain and the findings from psychology.

In some cases Alfano suggests minimal revision. For example, with preferences, Alfano’s review of the literature of choice blindness and preference reversals shows that our ordinal rankings of what we take to be goods can be influenced by factors irrelevant to the choice. This has significant consequences for moral theories if the good life and what we owe to each other depends on ensuring preferences and their ordering is respected—that is, if we assume good lives depend on one’s preferences being met or that we respect the preferences of others. But this variance phenomenon has relatively limited scope, as there is only a “penumbra of indeterminacy and instability” (39). (For example, preference reversals can happen between owning a nice book and having $50, but it won’t happen between the book and $50,000.) In similar fashion, Alfano holds that facts about how we can be subject to implicit biases should influence our assessments of responsibility, and features of emotional responses to particulars of situations can influence our moral judgments.

In other cases, Alfano proposes more significant revisions to our ethical concepts. With regard to moral character, Alfano argues that experiments such as the Good Samaritan Study and examples of unresponsive bystanders show that our notions of stable and substantial character traits (like virtues and vices in particular) are more situational than we are tempted to think. That is, what is more predictive of people’s good behavior in the Good Samaritan Situation is whether they are in a hurry, and with bystanders, whether there are others who could help. In light of these findings, Alfano holds that “character should be recast in interactionist terms, as partly due to features of the agent, partly due to features of the situation, and partly due to the interaction of the two” (130–31).

Additionally, Alfano proposes that findings in contemporary social science yield revision to a widely held thought that there are truly fundamental moral disagreements between cultures that exhibit psychological depth, are not defusable, and are sincere. Most are cases of different weightings of shared values or are cases where one party merely is ignorant of non-moral facts of the situation.

While superficial moral disagreements are fairly easy to identify, modally robust fundamental disagreements between epistemically responsible agents are much harder to pinpoint (173)

As a result, the worries about deep disagreement and clashes of intelligibility between cultures are, as Alfano takes it, not well supported by the data about convergent values.

Mark Alfano’s Moral Psychology is an exciting and challenging introduction to the interface between philosophical ethics and the empirical sciences that bear on the assumptions about humans that programs in ethics must make. Given that Alfano’s proposals are all modest (to various degrees), both sides to these debates will have things to say. Traditional ethicists will insist on considerably less revision, and others will call for greater divergence from the older programs. But it is in the arguments that we find were we must go, and so I am pleased to open this symposium on this important book.



Method in Moral Psychology

“Moral psychology” is a recently-invented term of art with a somewhat disputed reference, which reflects disagreement about its proper method(s).1 Some use the term to designate venerable philosophical inquiry into the nature and moral significance of psychological states. These inquiries are carried out through armchair reflection based in common sense, everyday observation, or intuition about, for instance, what moral responsibility is and whether it requires volitional activity; what emotions are and when, if ever, they are appropriate; and (the old holy grail of ethics) what happiness is and whether ethical virtue always delivers it. The Socratic paradoxes—such as that virtue is knowledge, that weakness of the will is impossible, that a good person cannot be harmed—comprise perhaps the earliest recorded instances of this style of moral psychology in European philosophy.

Some recent practitioners—dissatisfied with the impoverished results of what we might call reflective moral psychology—have instead carried out their inquiries largely through humanistic study, using historical, literary, and sociological results and the interpretative methods of those disciplines. We may call this humanistic moral psychology. In this category we find Bernard Williams’s use of Sophocles’s Ajax to provide an account of shame and moral incapacity,2 Martha Nussbaum’s interpretation of Aeschylus’s Oresteia in service of her accounts of anger and forgiveness,3 Gabriele Taylor’s analysis of Shakespeare’s Coriolanus in her monograph on vice,4 and Charles Mills’s study of John Hearne’s Voices Under the Window5 in aid of his analysis of the phenomenology of race and class, all of which make use of hermeneutical and analogical reasoning to help illuminate moral experience via these artworks.

Still other philosophers use “moral psychology” to refer to the scientific study of the determinants of well-being and moral life, often aiming to validate or falsify the presuppositions of reflective and humanistic moral psychology. The spectacular advances of the last fifty years in the cognitive and behavioral sciences fuel this work, which we might call experimental moral psychology. There is also ample historical precedent for it in the work of Hobbes, Hume, and Dewey, who all introduce experimental methods of reasoning into moral philosophy. Proponents of experimental moral psychology often can be found enlivening the question-and-answer period of reflective and humanistic conference presentations with pointed questions about the empirical adequacy of various assumptions. (After one recent talk of mine on the topic of emotional self-knowledge, a prominent proponent of this methodology raised his hand to ask, simply, “You know that your conclusion contradicts the science on this matter, right?”)

These approaches are, on the whole, largely complementary. Nussbaum’s work on emotion, for example, appeals alternatingly to Aeschylus and to neuroscientist Joseph LeDoux. Likewise, Nietzsche combines literary and historical reflection on Greek tragedy, Christianity, and Wagner with appeals to contemporary (nineteenth-century) psychology, physiology, and race theory (though he rejects certain tendencies of British empirical psychology).

Alfano’s introductory text for advanced undergraduate students brings experimental moral psychology into the classroom. It has many virtues. For example, it sketches solutions to numerous controversies in the field, including: (1) whether empirical findings about the indeterminacy and instability of preferences undermine assessments of rightness and well-being, (2) whether findings in neuroscience related to the “dual-process theory” of cognition undermine assessments of rightness that depend on the doctrine of double effect, (3) whether “situationist” findings in social psychology undermine character-based assessments of moral worth and rightness, and (4) whether anthropological findings about the extent of ethical disagreement undermine some versions of moral realism. It also serves as an excellent annotated bibliography of recent experimental moral psychology.

Alfano insists, and I agree, that interdisciplinary confrontation is necessary for doing moral psychology well, since “moral philosophy without psychological content is empty, whereas psychological investigation without philosophical insight is blind” (1). From where should moral philosophers source our psychological content? One answer, which we might call the broad experimental view, prioritizes equally the results of all relevant natural and social sciences, including sociology, criminology, anthropology, archaeology, and linguistics. Even though Alfano does not discuss most of these fields—and it would be practically impossible for any short introductory text to do so—he mentions all of them, at least in passing, in a way that makes clear that he endorses broad experimentalism. Elsewhere Alfano has outlined a research program that is “naturalistic” in the sense that it aims to “employ only methods consonant with those used in the [‘hard’ and ‘soft’] sciences and refer only to entities countenanced by the sciences.”6 Alternately, we might accept a humanistic view, according to which one necessary component of the psychological content needed for moral philosophy can be acquired only by humanistic study. Such study involves making sense of (for instance) our relation to the past, our aesthetic practices, and our membership in political communities, where the project of sense-making depends crucially—in ways that distinguish it from some scientific projects—on our distinctive cultural, ethical, political, and aesthetic experiences and values.

One point in favor of the humanistic view, it seems to me, is that the study of moral psychology must attend to what our moral experience is like, shaped as it is by our interpretations of that experience. Much moral psychology aims to help us to make sense of our moral experience by using interpretive methods proper to humanistic study, not merely to diagnose the network of causes that make morality “work” (1) or not. Alfano’s exclusion of humanistic moral psychology makes me wonder about the place of such reflection in the work of experimental moral psychologists. In their view, must we deny that humanistic reflection provides necessary psychological content to moral philosophy? Or must humanistic reflection proceed by methods consonant with those of the sciences?

I worry that by ignoring humanistic moral psychology, this introductory text erases much interesting research in moral psychology in that it encourages students to think that there’s nothing to see there. Perhaps this is an inadvertent result of, say, marketing pressures from the publisher. Alfano does suggest, though, that the text’s design is to offer “a comprehensive survey of contemporary moral psychology” (ix). If so, then Alfano has concealed from the introductory student the fruitful pluralism of the field.

Perhaps Alfano will respond that the book doesn’t need to “give the other side,” because it is the other side. Or perhaps Alfano denies that humanistic moral psychology is intellectually reputable. I look forward to his clarification on this matter.

Implicit Bias

Let’s consider one of the many stimulating discussions in Alfano’s text—a discussion about blaming people for being implicitly biased that is representative of the mix of empirical and conceptual claims offered in the text. In anti-racist communities there has long been a suspicion, to say the least, that even individuals who sincerely disavow racism might still harbor racial bias. Recent empirical research seems to confirm this suspicion, and there is a burgeoning philosophical literature about how to understand moral responsibility for implicit bias, which Alfano ably highlights.7 How should we react to people with implicit biases who sincerely avow a commitment to the moral and political equality of all? There are numerous dimensions of analysis that are relevant to answering this question, including (1) whether holding one responsible would be useful (the pragmatic factors), (2) whether the biased person is in a position to know about their bias (the epistemic factors), and (3) whether the biased person is in a position to control their bias (the control factors). Alfano discusses them all, but let us focus on the pragmatic factors.

Alfano presents the following argument that we should not think of implicitly biased though explicitly egalitarian persons as racist or sexist. Studies suggest that one’s self-conception is often self-confirming: if I conceive of myself as a racist, say, then I am more likely to act like a racist than I would otherwise be. Likewise, accusing others of racism risks making them even more likely to act like a racist. Alfano takes these claims to support what he calls “the factitious, interactionist framework” (132) of virtue, according to which behavior is explained in terms of, among other things, “the ongoing feedback between the individual and environment” (187). On this view, social expectation-signaling and one’s self-conception help to bring about and sustain dispositions to think, feel, and act that are similar to traditional Aristotelian virtues or vices. Therefore, conceiving of people as racist is “dangerous” (71). Instead, Alfano suggests, perhaps one should think of such a person “as someone who strives to be fair to targets of negative stereotypes but who suffers in his human, all-too-human, way from various biases” (71). In doing so, one would ascribe lack of ill will (or perhaps even good will) to the implicitly biased person, which might help to bring about personal improvement. Lest one think that Alfano is urging emotional calm in the face of injustice, he adds that there is reason for victims of implicit bias to “angrily denounce people who are trying their best, despite implicit biases. . . . Even if it ruffles a few feathers” (78). For implicit bias can cause “immense harm” (78).

Several aspects of this argument would make for interesting classroom debate. First, there is some tension between the claims that (a) we should treat such implicitly biased people as if they have genuinely good will (or, at least lack ill will) towards the targets of their bias and (b) we should sometimes angrily denounce such people. For anger and blame are, plausibly, responses to ill will. Pragmatic arguments to be angry in such cases then seem to counsel emotional dishonesty or confusion. Perhaps, as Nussbaum has recently argued, other emotional attitudes are both more fitting and more productive (though perhaps not).8

Second, if even angry denunciation is compatible with constructive efforts to improve the offender, then I see no reason why the same cannot be said of calm and supportive communication that the offender embodies some form of racism. Indeed, George Yancy has recently hypothesized that bringing a person’s racism to their attention might be crucial for facilitating improvement.9 Yancy urges well-intentioned whites to consider themselves precisely as people who harbor racism despite their anti-racist intentions, beliefs, and actions. He conceives of such direct communication as a kind of gift, designed to help well-intentioned whites escape from the racist lies and self-deceptions that cloud their minds. No doubt such a direct intervention requires tact and perhaps even, as Yancy insists, a kind of love. So, on what empirical grounds should we reject Yancy’s model of direct anti-racist (and anti-sexist) intervention in favor of Alfano’s model?

This question leads to a third concern. It is not clear from the text what empirical support, if any, there is for the claim that the relevant virtue of justice is best characterized as factitious and interactionist. Alfano states that there is some evidence that the virtues of tidiness, charity, cooperativeness and competitiveness, helpfulness, eco-friendliness, and scholastic motivation can be understood in this way. In general, though, “we currently lack evidence one way or the other about which virtues” (132) can be thought of in this interactionist way. So, it seems, we don’t know whether the virtue of justice at issue is among those that can be inculcated “by fine-tuning your self-concept and the social expectations directed at you” (132). Given this lack evidence, the status of these pragmatic claims about redressing bias is unclear. Does Alfano present them as hypotheses or as justified by particular empirical research? I worry that the reader will receive the false impression that it is a robust finding that we should not call (merely) implicitly biased people “racist” or “sexist,” and that such a reader will be motivated to adjust their behavior in a problematic direction, for instance away from Yancy’s proposal, because of an inadequately supported claim.

Alfano’s discussion of the pragmatics of blaming implicitly biased people would make for stimulating classroom debate. That said, this text would work best when supplemented by secondary literature that addresses Alfano’s conceptual and empirical assumptions. To this end, Alfano helpfully provides suggested further readings at the end of each chapter. With respect to the discussion about implicit bias, supplemental readings might engage assumptions about the nature of good and ill will, the ethics of expressing anger and blame, the nature of racism, and the empirical status of Alfano’s interactionist theory of virtue.10

  1. According to the following Google Ngram, “moral psychology” rapidly rose in popularity starting in the late 1970s:

  2. Bernard Williams, Shame and Necessity; Williams, “Philosophy as a Humanistic Discipline”; Williams, “Nietzsche’s Naturalistic Moral Psychology.”

  3. Martha Nussbaum, Anger and Forgiveness: Resentment, Generosity, and Justice (New York: Oxford University Press, 2016).

  4. Gabriele Taylor, Deadly Vices (New York: Oxford University Press, 2008).

  5. Charles Mills, “Better Dread (if Still Dead) than Red: High-Brown Passing in John Hearne’s Voices Under the Window,” Journal of Applied Philosophy 34 (2017) 519–40.

  6. Character as Moral Fiction (Cambridge: Cambridge University Press, 2013), 5.

  7. I should note that the psychological theory of implicit bias and the evidential import of Implicit Association Tests (IATs) to dispositions to discrimination are controversial. See Oswald, Mitchell, Blanton, Jaccard, and Tetlock, “Predicting Ethnic and Racial Discrimination: A Meta-analysis of IAT Criterion Studies,” Journal of Personality and Social Psychology 105 (2013) 171–92; and Greenwald, Uhlmann, Poehlman, and Banaji, “Understanding and Using the Implicit Association Test: III. Meta-Analysis of Predictive Validity,” Journal of Personality and Social Psychology 97 (2009) 17–41.

  8. Martha Nussbaum, Anger and Forgiveness: Resentment, Generosity, and Justice (New York: Oxford University Press, 2016).

  9. George Yancy, “Dear White America,” New York Times, December 24, 2015,

  10. Thanks to Rachel Fredericks for her valuable comments on a previous draft.

  • Avatar

    Mark Alfano


    Reply to Fischer

    Thanks to Jeremy Fischer for his thought-provoking comments on Moral Psychology: An Introduction. I welcome this opportunity to reflect on the methodology of moral psychology.

    Fischer distinguishes three methodologies in moral psychology, which he dubs reflective, humanistic, and experimental. Each of these provides its own perspective on the topic. Reflective moral psychology employs armchair pondering of everyday experience, supplemented by common sense. As Fischer notes, reflective moral psychology faces several challenges, and we could easily add more challenges to his list. Which everyday experiences should one reflect on, and why? How universal is common sense? Whose experiences are liable to be ignored in this process? If someone disagrees with you about a philosophical intuition, does that mean at least one of you lacks common sense?

    These are age-old questions, and they have rightly prompted philosophers to seek a wider, more diverse range of experiences on which to reflect and a process that is—in the ideal case, at least—reproducible and intersubjectively valid. This leads us to humanistic and experimental moral psychology. In the former, the range of experience is broadened by going back in time to exemplary historical and literary cases, which furnish rich portraits of people’s conduct and inner lives. In this connection, Fischer points to examples such as Bernard Williams’s interpretation of Ajax and Martha Nussbaum’s interpretation of Oresteia.

    By contrast, experimental moral psychology works with data to seek out trends, identify effects both weak and strong, and construct causal or computational models of moral psychological processes. Instead of uncovering the deep meanings embedded in great historical and mythical exemplars, experimental moral psychology aggregates and analyzes data from ordinary human animals. Fischer questions whether this approach is sufficient. Can an experimental moral psychology reveal everything that a humanistic approach would illuminate? To get a firmer grip on this question, it’s helpful to ask what, exactly, distinguishes the humanistic approach. Fischer primarily associates it with interpretation or sense-making. Science can tell us what there is and how it works, but humanistic inquiry excels in telling us (or helping us tell ourselves) what it means. I would add that, in engaging our imaginative capacities, humanistic inquiry may also prompt us to consider possibilities and prospects that have hitherto remained unrealized. Reading Ursula K. Le Guin’s stories, for example, is likely to lead to this phenomenon.

    Because it can be hard to engage fully with the statistics, humanistic inquiry can be uniquely instructive. Humanistic interpretation explores stylized exemplars, even as it elides the full range of variance in people’s psychologies. This has the advantage of focusing our attention on particular cases that make various processes, experiences, and attitudes salient. Just as it is often pedagogically, cognitively, or communicatively helpful to do geometry with drawings on paper, chalkboards, or computer screens, so it is often pedagogically, cognitively, or communicatively helpful to do moral psychology with narratives from books and film. As Andy Clark (2002) has argued, human minds operate best when they are able to iteratively alternate between cognitive processes such as deciding, inferring, and evaluating and perceptual and agential processes such as seeing, feeling, and manipulating. Shifting back and forth between the analysis of data and artistic or hermeneutical representations of idealized types enables us to take advantages of both our cognitive powers and our perceptual and storytelling capacities.

    I hope that these irenic remarks go some way to reducing the distance between Fischer and me. However, I don’t want to pretend that we agree about everything, for I also see two substantial dangers in the kind of humanistic moral psychology he enjoins. First, the range of exemplars on which to reflect is liable to be at least as cramped as one’s own experience. The store of literary and historical exemplars is almost laughably narrow. In works of humanistic moral psychology, how often does one encounter exemplars who are not “the ancients,” characters in Jane Austen or George Eliot novels, or Huckleberry Finn? If one of the deficiencies of reflective moral psychology is that it does not draw on a sufficiently diverse range of actual and possible experience, then humanistic moral psychology is liable to exacerbate the problem. This is not essential to humanistic moral psychology, but one is hard-pressed to find counterexamples. Experimental moral psychology does better on this score (if and only if it uses large and diverse samples of participants—a desideratum that is sometimes neglected).

    In addition to (typically) providing a worm’s-eye view of a very small number of exemplars, humanistic moral psychology unconstrained by data is prone to lead to misimagination. As Adam Morton (2006) points out, it’s only possible to learn from our imaginings if they have the possibility of being either correct or incorrect. Fictional, mythical, and idealized historical exemplars are, in this context “an invitation to illusion . . . because when we respond to fiction we react to the characters in many of the ways we do to real people, and so if a way of reacting makes sense with respect to a fiction we tend to think that it makes sense with respect to real people.” As Morton goes on to argue, this can lead us to think that the characters, motives, and experiences we attribute to fictional characters are possible (perhaps even desirable) characters, motives, and experiences in real life. It can also lead us to expect causal or conceptual connections in real life that only exist in fiction.

    To illustrate: Bernard Williams moves from shame in Sophocles’s Ajax to shame in actual human communities. Martha Nussbaum moves from anger and forgiveness in Aeschylus’s Oresteia to anger and forgiveness in real human encounters. Gabriele Taylor moves from vice in Coriolanus to vice in everyday life. And Charles Mills moves from the experience of class and race in Hearne’s Voices Under the Window to the experience of class and race in contemporary society. We must ask, though: does shame really work that way? Can real people forgive in the way Nussbaum imagines forgiveness to work in the Oresteia? What is vice like for the fancy apes that we are, and how similar is it to the vice of fictional characters in Shakespeare’s plays? Do victims of our current racist and classist society experience their own lives as Mills imagines the characters in Hearne’s novel to experience them? Maybe. But then again, maybe not. The only way to answer these questions is to employ the methods of science. If my arguments here are on the right track, then humanistic moral psychology may help us make sense of our experience, but it may also help us make nonsense of it. That does not make humanistic moral psychology useless. It does, though, show that the truth-values of the insights it promises need to be corroborated by science. Or, as I put it in the book, “moral philosophy without psychological content is empty, whereas psychological investigation without philosophical insight is blind.”

    I turn now to Fischer’s remarks about bias and responsibility (about which I have more to say in my responses to Trujillo and Radke). Fischer points out that there is a prima facie tension between treating people who embody implicit but not explicit bias as if they have good will, on the one hand, and sometimes angrily denouncing them, on the other hand. I believe that this tension can be resolved by appealing to the distinction I articulate in more detail in chapter 4 between ascriptions of traits (e.g., “You are such a sexist pig!”) and evaluations of actions (e.g., “That was a sexist thing to do!”). The former, but not the latter, tend to function as self-fulfilling prophecies. For this reason, it can be dangerous to accuse other people of harboring biases (even if the accusation is correct), but it is still perfectly possible to get angry and express that anger by calling out bad behavior. Fischer rightly points out that the extant evidence for self-fulfilling prophecies relates to ascriptions of traits other than being a racist or being a sexist. So my tactical advice here and in the book is based on the speculation that the same effect is liable to crop up in this context as well. Further research could corroborate or falsify this speculation. Further research could also shed light on whether my precautionary advice is better tailored to the case than George Yancy’s suggestion.

    I conclude by noting that all of this may turn out to be moot if the implicit bias paradigm is overthrown. The implicit association test itself may be unreliable and not useful for predicting behavior. I raised this possibility in a tentative way on page 66 of Moral Psychology, but recent research has made me even more worried (e.g., Forscher et al. 2017). If implicit bias turns out not to exist or not to have a serious influence on people’s conduct, then the conversation we need to have is not about the unicorns who embody implicit-but-not-explicit bias, but about individuals who harbor and express good old-fashioned explicit bias. Indeed, we need to have that conversation no matter what.



    Alfano, M. 2016. Moral Psychology: An Introduction. Malden, MA: Polity.

    Clark, A. 2002. “Towards a Science of the Bio-Technological Mind.” International Journal of Cognition and Technology 1: 21–33.

    Forscher, P., et al. 2017. “A Meta-Analysis of Change in Implicit Bias.” Open Science Framework, October 5. Retrieved from

    Morton, A. 2006. “Imagination and Mmisimagination.” In The Architecture of the Imagination: New Essays on Pretence, Possibility, and Fiction, edited by S. Nichols. Clarendon.



The Backfire Effect and Political Psychology

Mark Alfano’s Moral Psychology: An Introduction palatably samples topics in moral psychology. It could naturally accompany systematic works like Walter Sinnott-Armstrong’s multivolume edited series on moral psychology or John Doris’s Moral Psychology Handbook. Alfano’s book, though, sets itself apart in its accessibility. Students and non-academics will gravitate toward his examples—police brutality, workplace discrimination, mundane monstrosities, and cross-cultural disagreements; he makes philosophy and social science relevant to everyday life. Philosophers and psychologists will marvel at how effortlessly he integrates research from centuries of philosophy and decades of social science. He’s just as comfortable glossing Plato, Aristotle, and Kant as Daniel Kahneman, Anthony Greenwald, and Paul Ekman. Throughout, Alfano resists obvious conclusions and extravagant claims, thus exhibiting a calm, fluid confidence in his restrained theses. This book is as smooth as academic introductions get.

I want to criticize Alfano’s book; critique is the highest form of praise for our peculiar profession of philosophy. But I found his arguments expertly qualified and well-situated in contemporary ethics and psychology. Even as someone who disagrees with his conclusions, I know by the structure of his arguments that Alfano listened carefully to his opposition. I also realize the task of an introductory book isn’t to shore up every argument. So, here I’ll offer expansions on Alfano’s arguments. Specifically, I’ll argue that recent literature on the “backfire effect” complicates some of Alfano’s practical advice. But if we apply Alfano’s ideas to his own solutions, we might surmount their practical shortcomings.

In Moral Psychology, Alfano advocates that a person who is oppressed (or speaking up for the oppressed) should directly confront explicit or complicit oppressors. For example, he advises that people can combat implicit biases by confronting whoever has the bias (74, 78). Also, when discussing emotion, he highlights Lisa Tessman’s observation that people sometimes face a dilemma when regulating injustice-responsive anger: they can (a) recognize any injustice and get angry at it, which leads to all-consuming anger in an unjust world, or (b) get angry in a more measured way, but risk enduring degradation (2005, 124–25; cited by Alfano, 91). He extends this point in his chapter on character to take a dig at Aristotle. Aristotle argues that a virtue of mildness helps us to regulate our anger (2002, 1108a5ff.). But Alfano, following the work of feminist and anti-racist scholars like Macalester Bell and Zac Cogley, argues that anger responds to harms; it’s about more than mere temper. Alfano writes:

Aristotle restricts his theory to adult male citizens in a polity built around satisfying their and only their needs. For such people, injustices are bound to be rarer than for people in oppressed groups. When such privileged citizens do experience an offense, they are likely to have the power and perceived authority to make their anger acknowledged and the fault redressed. They have the luxury to be good-tempered. Someone who systematically faces injustices and whose anger about them is ignored, dismissed, or met with further injustice is in a very different position. For them, being or appearing good-tempered may even contribute to their oppression since it suggests that nothing is seriously wrong. (117)

Alfano’s point on Aristotle is well-taken, even though I think he should also have discussed Aristotle’s treatment of righteous indignation (1108b1ff.), and nowhere does Alfano explain that virtues and vices are inextricably linked to the lifelong project of human flourishing in a community. I even think Aristotle would agree with this advice in the abstract. But the devils of injustice lurk in the details. It matters which person confronts the other and how. It also matters what results from such confrontations, as confrontations aren’t intrinsically good.

An unfortunate truth for academics is that social justice isn’t only about epistemic, moral, or political justification. If it were, history wouldn’t be a “slaughter-bench,” as Hegel once called it (1988, 24). Bad, vicious, and evil people would yield to overwhelming justification against wrongdoings if it were only about reasons. Instead, we must also confront the empirical realities. And if we’re empirical about moral psychology and social justice, then we must admit what works and what doesn’t. I can’t help but hear Aristotle replying to Alfano with something like: “Virtues must be nested within a framework for human flourishing, as we need to consider the effects of emotions and character traits on the overall well-being of a person and community. After all, Nicomachean Ethics and Politics are a continuous project because ethics and politics are inseparable.” In other words, social justice shouldn’t exclusively focus on means or ends, and it shouldn’t separate moral psychology from political psychology. Rather, to actualize justice, we must scrutinize both the goals we want to reach and the winding, forked paths we take to get there.

One recently explored area in politics and psychology is the “backfire effect.” Many experiments document a curious phenomenon when some people are presented with new information that undermines their extant beliefs. Rather than change their minds, some people believe their original opinions more deeply. The backfire effect turns humans reasoning ironic; people in its grip believe more firmly their original opinions in the face of strong countervailing evidence (see Nyhan and Reifler; Trevors et al.; Chong and Druckman; Lewandowsky et al.). American politics has demonstrated this since the 2016 primary elections. Trump boasted that he could shoot somebody and not lose voters, and nineteen women alleged sexual misconduct suffered through interactions with Trump. Yet in November 2017, among Republicans, Trump had an approval rating as high as 85 percent (Marcin 2017). And this is not a single-party problem. Democrats failed to confront Hillary Clinton’s career moves that exacerbated American poverty and extended campaigns in Iraq and Afghanistan that murdered (and still murder) civilians. Note, the backfire effect is different from and more insidious than Alfano’s mention of confirmation bias (cf. 119).

Why does the backfire effect matter? Sometimes experience gives counterintuitive results. And academics should yield to this experience in ethical and political matters rather than save theoretical integrity in systematic projects. This is especially the case when academics don’t come from the backgrounds of the people they’re talking about. A recent study by Justin P. Brienza and Igor Grossmann suggests that people from higher social classes are less wise than people from lower classes; they’re worse at finding solutions to practical problems. Additionally, philosophers have long understood how slaves can have deeper philosophical understanding than their masters (see Hegel 1977, §§178–96), and when people live in this world as a person of color, their consciousness splits into self-conception, awareness of how others perceive them, and the complex interplay between the two (see DuBois 1986, 364ff.). Social justice is more than mere academic curiosity and deserves as much rigor as any reputable science or philosophical discipline. For these reasons, any academic should read widely before weighing in on the philosophy of race, gender, or sexuality, and they shouldn’t be surprised that theory might not work out in practice.

So, what do we do with research on the backfire effect, and what can we learn from the recent election in America? I don’t know. But if scholars like Alfano (and me) want to combat biases, psychological distress, and social injustice, then we must offer smarter advice than mere confrontation of the oppressor. Alfano does this for biases; his systemic recommendations—such as reforming application procedures, disarming most police, and getting healthcare workers to treat diverse populations—deserve praise (74–79). But we don’t see this level of informed practicality in discussions on emotion, character traits, or disagreement. That is, how can we regulate emotions (and which ones?), how can we train character (selecting for which traits?), and how should we argue with other cultures, especially when we can’t agree on a policy where one side will be harmed? Even leaving these questions aside, we can complicate Alfano’s explicit advice of confrontation of oppressors. It should be emphasized that, when considering confrontation, we must regard who’s doing the confronting and in what ways. Confrontation might not only be unhelpful; it might actually make things worse for the oppressed or victims of violence (see McKiernan 2016, 150–51). We need to be suspicious of proceeding unstrategically or without tried and tested methods of organizing (perhaps by expanding our catalogue of psychological evidence to include historical accounts of resistance). This is especially the case if we garner any form of privilege or power.

Alfano deserves praise for confronting these issues in his book. But he (and I) need to be more aware of counterintuitive results in the public sphere. Sometimes we intend well and act earnestly on behalf of others, but sometimes it backfires, inflicting the burns of the backfire on the groups that academics advocate for. Maybe we face another Tessmanian dilemma: (1) we combat oppression vociferously, albeit with the risk of worsening conditions, or (2) we quietly resist but face the inevitable cruelty of public indifference to suffering and squandered opportunities to improve. Maybe part of the messiness of politics is that sometimes we need both strategies, and in either case, single solutions won’t work in all contexts. Maybe we also sense imagination’s unfortunate absence in political strategy. Alfano confesses that character—due to its complex interactions with patiency, agency, reflexivity, sociality, temporality, preferences, responsibility assessments, and emotions—is the most complicated phenomenon he investigates (136–37). This can’t be overemphasized, especially since he should include political character traits and group dynamics in the discussion.

As Alfano repeatedly implores, researchers of the moral psychology of biases, emotions, and character traits need to take sociality more seriously. And, as much as anything, this means finding psychological tools to combat oppression, even if they diverge from contemporary academic or political strategies. Academics should also shun overly simplified advice that obscures detail crucial to solutions, and we must acknowledge our privilege in writing pieces that almost no one in public policy or the general public reads. And at some point, we must deal with ugly realities that Aristotle observed two millennia ago: “For the person who lives according to emotion will not listen to talk that tries to turn him away from it, nor again will he comprehend such talk; how will it be possible to persuade someone like this to change? And in general it is not talk that makes emotion yield but force” (1179b26–29). He goes on to add, “And perhaps if someone wishes to make people better—whether in large numbers or in small—by exercising supervision over them, he too should attempt to become an expert in legislation, if it’s through laws that we become good” (1180b23–25). That is, if academics want to change harsh realities, we can’t stay exclusively on manicured campuses. Politics, and maybe force, may be our best means, and both will soil our oxfords.

Alfano already does much of what I advised. For example, he confronts counterintuitive results. He shows how preferences aren’t completely stable or determinate, but how they can nonetheless inform behavioral science (ch. 1). He argues that lack of control and knowledge undermine responsibility, but how we nonetheless can be culpable for certain forms of non-capacity or ignorance (ch. 2). He exhibits how emotions and reasons aren’t completely separable but that some emotions, like disgust, can nonetheless be discounted as providing good reasons (ch. 3). He reveals that it can be good to tell people they’re virtuous because they’ll act virtuously, but cautions us on the dangers of attributing vices (ch. 4). And he explains how cross-cultural studies oversimplify values both when they ascribe to different cultures fundamental disagreement on deep values and when they erase important differences in the ways people value (ch. 5). My cautionary note only extends this pattern to the solutions we find to facing social problems. Alfano has gifted us with a lucid exploration of the effects of agency, patiency, reflexivity, sociality, and temporality on moral psychology. But there is more work to be done to apply the same lucidity and rigor to the practical upshots that Alfano suggests and to the psychological problems in politics, which are inseparable from the lessons Alfano presents in Moral Psychology.



Alfano, Mark. 2016. Moral Psychology: An Introduction. Cambridge: Polity.

Aristotle. 2002. Nicomachean Ethics. Translated by Christopher Rowe. Commentary by Sarah Broadie. Oxford: Oxford University Press.

Brienza, Justin P., and Igor Grossman. 2017. “Social class and wise reasoning about interpersonal conflicts across regions, persons and situations.” Proceedings from the Royal Society B. December.

Chong, Dennis, and James N. Druckman. 2007. “Framing Public Opinion in Competitive Democracies.” American Political Science Review 101: 637–55.

DuBois, W. E. B. 1986. The Souls of Black Folk. In Writings: The Suppression of the African Slave-Trade, The Souls of Black Folk, Dusk of Dawn, Essays and Articles, 357–547. New York: Library of America.

Hegel, G. W. F. 1977. Phenomenology of Spirit. Translated by A. V. Miller. Oxford: Oxford University Press.

———. 1988. Introduction to the Philosophy of History. Translated by Leo Rauch. Indianapolis: Hackett.

Nyhan, Brendan, and Jason Reifler. 2010. “When Corrections Fail: The Persistence of Political Misperceptions.” Political Behavior 32: 303–30.

Lewandowsky, Stephan, et al. 2012. “Misinformation and Its Correction: Continued Influence and Successful Debiasing.” Psychological Science in the Public Interest 13: 106–31.

Marcin, Tim. 2017. “Do Republicans Like Trump? Latest Poll Shows Least Popular President Ever Is Starting to Lose His Base.” Newsweek, November 18.

McKiernan, Amy L. 2016. “Standing Conditions and Blame.” Southwest Philosophy Review 32: 145–51.

Tessman, Lisa. 2005. Burdened Virtues: Virtue Ethics for Liberatory Struggles. Oxford: Oxford University Press.

Trevors, Gregory J., et al. 2016. Discourse Processes 53: 339–70.

  • Avatar

    Mark Alfano


    Reply to Trujillo

    Thanks to G. M. Trujillo for his forceful commentary on Moral Psychology: An Introduction. Trujillo builds a case for complicating the ways in which people confront the individuals and systems that contribute to oppression. Like Fischer, he is skeptical of the details of my recommendation to approach such confrontation with caution. Whereas Fischer suggests that it may be tactically or strategically better to follow Yancy’s advice, offering biased people knowledge about themselves as a gift of love, Trujillo worries that any direct confrontation is liable to induce a “backfire effect.”

    The basic idea behind the backfire effect is that, sometimes, directly telling someone that they’re mistaken can lead them to become even more confident, rather than reversing their opinion. To illustrate with an example from my own life: in 2011 I spent several hours walking my father through the evidence demonstrating that Barack Obama was born in Hawaii, not Kenya. He had jumped aboard the “birther” bandwagon and approached the conversation with absolute certainty. By the end of our heated exchange, I had cornered him into admitting that maybe Obama was really an American citizen, but within a week he had reverted to the racist canard. In 2016, he voted for Donald Trump in Pennsylvania. Clearly, my provision of evidence and arguments did not have the intended effect. I sometimes worry that I may have made things worse by arguing my case. Perhaps if I had failed to rise to the bait, things would have played out differently.

    To better understand this worry, let’s review the backfire effect. The phenomenon has been the object of scientific fascination for over a decade.1 Imagine someone who accepts proposition p. Now imagine this person is confronted with the counter-claim, ~p. Simple contradiction of this sort, if it comes from a trusted source, might lead them to change their mind. If I think the time is 7 p.m. and you tell me that it’s actually 6 p.m., as long as I have no reason to distrust you I’ll probably just update my belief. It would be strange to become even more convinced of my preexisting opinion by your counterclaim. But we can imagine ways in which this might happen. For instance, you tell me that it’s 6 p.m.; I find this puzzling; I look at my phone, check my watch, and jump on my computer to Google “current time Melbourne.” All three sources of information indicate that the current time is 7 p.m. Your disagreement thus leads me to review the evidence, which in this case confirms my extant belief, which leads me to become more confident in it.

    Plausibly, the backfire effect tends to play out in this way: a belief is challenged; the challenge is reviewed (perhaps in a biased or self-protective way) and rejected; the agent becomes more confident in the original belief (Prasad et al. 2009). Note that, unless the review of the evidence is egregiously unfair, none of this is unreasonable. We cannot accuse people who go through this process of having done anything irrational. Indeed, the similarity to the dogmatism paradox is striking (Kripke 2011; Harman 1973, 1986). If I believe that p, then I take myself to know that p. And if I really do know that p, then any evidence against p must be misleading. Since it is epistemically permissible to disregard misleading evidence, I am justified in disregarding evidence against p.

    An additional mechanism at work in the backfire effect may relate to social trust and the (re)wiring of epistemic networks. If you tell me things I know to be false, that gives me a good epistemic reason not to trust what you say on this topic—both now and in the future. So if I believe that p and you challenge my belief, I may review and reject your challenge, then conclude that you are either a liar or epistemically incompetent in this domain (or generally). If I know that p, then any testifier who contradicts p is disposed to mislead me. Since it is epistemically permissible to disregard misleading testifiers, I am justified in distrusting testifiers who contradict or offer evidence against p. Indeed, this is precisely what happened when I tried to tell my father in October 2016 that Donald Trump was an anti-Semite and a xenophobe. He wrote in an email response, “Don’t know where you get your info on him being anti Jewish, and not wanting immigration, he wants legal immigration.” I contradicted his beliefs; he chose to trust sources such as Fox News and Breitbart and to distrust me. My confrontational intervention didn’t help and may have backfired.

    What to do about this? In my recent research, I’ve tried to better understand the ways in which epistemic trust works in social networks (Alfano 2016b). In light of the dogmatism paradox and the backfire effect, how should people distribute their trust and distrust? What are the best ways to practice intellectual humility and avoid getting trapped in dogmatism, epistemic insouciance, and a variety of other epistemic vices (Alfano 2015; Cassam 2015, 2017; Sherman 2015)? Turning these questions around, we can also ask: in light of the dogmatism paradox and the backfire effect, how should people confront, cajole, and coax each other to redistribute their trust and distrust? What are the best ways to extricate people from filter bubbles (Pariser 2011) and the dogmatism, epistemic insouciance, and other epistemic vices that they foster? Trujillo is exactly right that we need to think in more nuanced and creative ways about these questions.

    In the meantime, I would be remiss not to note that the backfire effect has recently come under criticism for failing to replicate or generalize (e.g., Wood & Porter forthcoming). However, even this new research should not make us epistemically sanguine. While failing in some contexts to find an outright backfire, these researchers showed that correcting false beliefs often has no effect at all. And they did not investigate whether, as I suggest might happen, people who face corrections or contradictions redistribute their trust and distrust in problematic ways in future testimonial exchanges. Clearly, more interdisciplinary research in this domain is called for.



    Alfano, M. 2015. “Becoming Less Unreasonable: A Reply to Sherman.” Social Epistemology Review and Reply Collective 4: 59–62.

    ———. 2016a. Moral Psychology: An Introduction. Malden, MA: Polity.

    ———. 2016b. “The Topology of Communities of Trust.” Russian Sociological Review 15: 30–56.

    Cassam, Q. 2015. “Stealthy Vices.” Social Epistemology Review and Reply Collective 4: 19–25.

    ———. 2017. “Epistemic Insouciance.” Journal of Philosophical Research.

    Harman, G. 1973. Thought. Princeton: Princeton University Press.

    ———.1986. Change in View. Cambridge: MIT Press.

    Kripke, S. 2011. “Two Paradoxes of Knowledge.” In Philosophical Troubles: Collected Papers, edited by S. Kripke, vol. 1. Oxford: Oxford University Press.

    Nyhan, B., and J. Reifler. 2010. “When Corrections Fail: The Persistence of Political Misperceptions.” Political Behavior 32: 303–30.

    Pariser, E. 2011. The Filter Bubble: What the Internet Is Hiding from You. New York: Penguin.

    Prasad, M., et al. 2009. “‘There Must Be a Reason’: Osama, Saddam, and Inferred Justification.” Sociological Inquiry 79: 142–62.

    Schwartz, N., et al. 2007. “Metacognitive Experiences and the Intricacies of Setting People Straight: Implications for Debiasing and Public Information Campaigns.” Advances in Experimental Social Psychology 39: 127–61.

    Sherman, B. 2015. “There’s No (Testimonial) Justice: Why Pursuit of a Virtue Is Not the Solution to Epistemic Injustice.” Social Epistemology, 1–22.

    Skurnik, I., et al. 2005. “How Warnings about False Claims Become Recommendations.” Journal of Consumer Research 31: 713–24.

    Weaver, K., et al. 2007. “Inferring the Popularity of an Opinion from Its Familiarity: A Repetitive Voice Sounds like a Chorus.” Journal of Personality and Social Psychology 92: 821–33.

    Wood, T., & Porter, E. Forthcoming. “The Elusive Backfire Effect: Mass Attitudes’ Steadfast Factual Adherence.” Political Behavior.

    1. For details, see Skurnik et al. (2005), Weaver et al. (2007), Schwartz et al. (2007), and Nyhan & Reifler (2010).



Scaffolding Virtue in Defensive Agents

Mark Alfano’s insightful book uses findings from empirical research to uncover new ways of thinking about longstanding philosophical issues. With chapters on responsibility, the emotions, virtue, and disagreement, Moral Psychology: An Introduction offers a refreshing take on topics of interdisciplinary interest. In the fourth chapter, Alfano turns the Aristotelian conception of character on its head. In his view, we do not practice certain virtues in order to eventually gain an understanding of ourselves as virtuous. Rather, we come to understand ourselves as virtuous—with the help of others—and consequently find ourselves motivated to pursue virtue in our actions. Subscribing to an interactionist understanding of character formation, Alfano introduces a compelling idea: through feedback, expectation-signaling, and emotional cues (among other things), other agents scaffold our potentiality for virtue by shaping our self-concept. Alfano’s attention to sociality is a strength of the account, and presents a convincing picture of the ways that we rely on other agents for the practice of virtue. What remains underexplored is how we might practice virtue together in nonideal conditions. If we are tasked with scaffolding other agents’ characters, how might we do so effectively in contexts of oppression? In what follows, I engage with Alfano about a potential strategy for responding to implicit bias, expressing worries about its effectiveness in nonideal conditions.

A Possible Alternative to Angry Confrontation

In chapter 2, Alfano asks: Given that we are all perpetrators of implicit bias of one kind or another, how ought we to feel about ourselves, our attitudes, and our behavior? From the point of view of Alfano and others, one thing seems clear—even though we are guilty of implicit bias, we should not view ourselves as “one of those bad racist or sexist people” (71). And, importantly, we should avoid addressing others in those terms (71–72). Drawing on claims about self-concept and character development advanced in chapter 4, Alfano offers a pragmatic reason for avoiding attributions of X-ism:

Pragmatically, it may do no good in reforming one’s ways to think of oneself as a racist, sexist, and so on. As I argue in the chapter on virtue below, self-concept is often self-confirming: if I think of myself as an X, I’m more likely to act like an X than I would be otherwise. For this reason, it’s dangerous to think of oneself as racist, sexist, ableist, and so on. Likewise, it may not help others to convince him to reform his ways if they accuse him of being racist, sexist, and so on. As I also argue in my discussion of virtue in chapter 4, such attributions also tend to be self-confirming. (71)

On the view advanced here, we have reason to refrain from accusing others of prejudice. The thought appears to be that viewing agents as sexist or racist affects their self-conception such that they are somehow more likely to be sexist or racist; their self-concept is often self-confirming. But how exactly does thinking of ourselves as sexist or racist make us more likely to act in sexist or racist ways? For the pragmatic argument against confrontation to hold water, we need a working idea of how confrontational attributions can be self-confirming.

The natural place to look for answers is Alfano’s model of virtue, but it does not seem helpful in explaining the way that self-concept can scaffold prejudice. On the virtue acquisition model advanced in chapter 4, agents gain virtue through others’ expectation-signaling and the entrenchment of their self-concept as virtuous people (132–35). For instance, we become honest by thinking of ourselves as honest, and by knowing that others think of us as honest—and by knowing that they know that we know that they think of us as honest. Although I find Alfano’s account of virtue plausible, it’s hard to see how it can help explain the phenomenon of entrenched prejudice. How can accusations of sexism or racism shape our self-concept in a way that reinforces biased tendencies? The cognitive impairments that constitute bias are not the product of others’ view of us, but of many other factors; consequently, it seems like character attributions would do little to affect them. As it stands, it’s not clear how these attributions pose a pragmatic problem—it’s not clear why they should be avoided.

An alternative interpretation of the pragmatic problem with confrontation centers on agents’ defensive reactions to accusations. Agents who are accused of being sexist, racist, or ableist are often defensive. Recasting the pragmatic problem as one of defensiveness has two upshots. First, it allows us to get clear on how exactly accusations of X-ism can hurt reform efforts and make the offender more likely to be X-ist. Confrontation that triggers defensiveness can ultimately prevent agents from being motivated to attend to their biases. Second, it makes sense of Alfano’s proposed alternative to confrontation:

Perhaps instead of thinking of himself as a racist or sexist person, he should think of himself as someone who strives to be fair to targets of negative stereotypes but who suffers in his human, all-too-human, way from various biases. Perhaps, instead of thinking of him as a racist or sexist person, other people should think of him as someone who strives to be fair but who suffers in human, all-too-human, ways from various biases. These attitudes have the benefit of ascribing good will (or at least lack of ill will) while recognizing serious defects. (71–72)

These gentler attitudes appear to be designed to mitigate defensiveness by offering a positive self-concept for the offender. The above stance focuses on agents’ good intentions while pointing out their flaws in a way that is somewhat exculpatory. Instead of seeing herself as a racist, sexist, or ableist, the agent can understand herself as a flawed human being. The self-image on offer thus prevents defensiveness by neutralizing an attack on the agent’s self-concept. Pragmatically, these attitudes work because they circumvent defensiveness and encourage the agent to acquire knowledge and control over her biases.

A Worry

What do we make of the pragmatic strategy that Alfano entertains in the above passage? Elsewhere in the chapter on responsibility, he is clear that victims’ angry confrontation can be justified or even required, despite the pragmatic risks. On Alfano’s view, engagement with others—even the angry or resentful kind—can help us see our faults and assist in our self-improvement (74). These claims stand in tension with the prudential approach that he considers above, one that recommends a neutral attitude (one akin to Strawson’s objective attitude). When, if ever, should prudential concerns win out over agents’ apt expressions of anger? How much weight does Alfano assign prudential considerations? Alfano’s commentary on these conflicting strategies is very speculative, so it would be interesting to hear more about how he would reconcile these different approaches. In the meantime, I’ll conclude with a brief worry about Alfano’s pragmatic alternative to confrontation: in nonideal conditions, it’s not obvious that protecting the self-concept of defensive agents is pragmatic or desirable.

At first glance, Alfano’s proposed alternative has merit: a nonconfrontational attitude can do valuable work. For one thing, it disarms defensive individuals, allowing them to maintain a positive self-concept while indicating the work that still needs to be done. When executed well, firm engagement can foster agents’ self-improvement by encouraging them to interrogate themselves and work on their bad cognitive habits. However, despite these apparent benefits, protecting an offender’s positive self-concept may not always be the best route in nonideal conditions—especially if that self-concept is part of the problem.

The well-documented phenomenon of white fragility posits resistant agents’ self-concept (their understanding of themselves as “good people”) as the source of their defensiveness. If we emphasize their good intentions when they harm others, chalking up their mistakes to a universal human fallibility, then we do not complicate the “good person” complex that drives their defensiveness. Neutrally-toned attitudes risk rewarding defensiveness with face-saving. In these cases, reinforcing privileged agents’ positive and virtuous self-concept looks like it actually undermines their capacity to build virtue over time—they are prone to overestimating their own goodness and receive little correction. While shaping an agent’s positive self-concept may scaffold virtue in some cases by encouraging the agent to work, it could just as easily undermine the agent’s improvement if she does not adopt the right critical attitude.

It’s not clear then, that the proposed strategy is always pragmatically effective in nonideal conditions. In any case, the larger tension between pragmatic concerns and victims’ apt anger is still a difficult one to reconcile. In a recent paper on anger, Amia Srinivasan articulates the unfairness that oppressed people experience by introducing the concept of affective injustice, defined as “the injustice of having to negotiate between one’s apt emotional response to the injustice of one’s situation and one’s desire to better one’s situation—a conflict of responsibilities that are ‘all but irreconcilable.’” While the tension between prudential concerns and apt expression may be irresolvable, it’s worth thinking through our strategies for engaging with others. Alfano’s view of virtue gives us a framework to understand the ways that we can influence each other, offering a convincing and useful contribution to an ongoing conversation.



DiAngelo, Robin. 2011. “White Fragility.” International Journal of Critical Pedagogy 3.

Srinivasan, Amia. 2017. “The Aptness of Anger.” Journal of Political Philosophy. July.

Strawson, Peter. 2008. “Freedom and Resentment.” In Free Will and Reactive Attitudes, edited by Michael S. Mckenna and Paul Russell. New edition. Burlington, VT: Ashgate.

  • Avatar

    Mark Alfano


    Reply to Radke

    Thanks to Lyn Radke for her critically constructive comments on Moral Psychology: An Introduction (Alfano 2016). Like both Fischer and Trujillo, Radke focuses on the second and third chapters of the book. She is interested in the question of how best to foster virtues and eradicate vices in one’s social circles, especially when the moral and political context in which one operates is far from ideal. The distinction between normative theories tailored to ideal conditions and those tailored to nonideal conditions is due to John Rawls (1971). For Rawls, ideal normative theories make two crucial assumptions: first, all agents and collectives negotiate in good faith about what to do, and second, all agents and collectives have already achieved a satisfactory (if not perfect) level of prosperity and peace.

    While these conditions are obviously not met on planet Earth in the year 2018, there have been striking and encouraging trends in this direction. For example, according to the World Bank, even as the number of people alive has increased by several billion since 1990, the percentage of people living in extreme poverty has decreased from 35 percent to just under 11 percent.1 In The Better Angels of Our Nature, Steven Pinker (2011) documents the absolutely astonishing decrease in human violence over the last few millennia. And Transparency International observes a decrease in global corruption over the last decade.2 None of this establishes that the current state of affairs approximates the Rawlsian ideal. Societies around the world might be so badly off that making things less bad still doesn’t get us anywhere near the ideal. The indexes used by the World Bank and Transparency International are not perfectly valid and reliable. But it is worthwhile to bear these trends in mind.

    Radke is especially concerned by how far contemporary societies fall short on the first of Rawls’s two idealizing conditions. Essentially, she thinks that many participants in discussions about what to do about prejudice are not negotiating in good faith—and that they may even be incapable of negotiating in good faith. In this connection, Radke refers to the construct of white fragility, which Robin DiAngelo (2011) defines as “a state in which even a minimal amount of racial stress becomes intolerable, triggering a range of defensive moves” that include displays of self-directed and other-directed negative emotion, as well as behaviors such as opposition, silence, and escape. These responses tend to refocus attention and concern from victims onto perpetrators. When white fragility manifests, instead of thinking about how a racist remark, attitude, action, or policy has affected a person of color, those involved find themselves worrying about how white people feel about being accused of racism. In a recent op-ed, Jenée Desmond-Harris dubbed a related social phenomenon the “vanity sizing” of racism.3 It sometimes seems that, unless someone is currently getting a swastika tattoo while burning a cross and enslaving black people on a plantation, it is considered rude to accuse them of racism. Every winter, I encounter this sort of reaction from otherwise right-thinking Dutchies when I mention that the cultural practices associated with the mythical character Zwarte Piet are racist and colonialist. It sometimes seems that the word racist itself triggers white people in the same way that outright racial slurs and caricatures trigger their targets. And as we are seeing day by day with the #MeToo movement and the inevitable backlash it has prompted, accusing someone of sexism often provokes a similar defensive response.

    These patterns of reaction have self-reinforcing social effects. If people’s racial attitudes are only ever to be praised or ignored, if accusing someone of bias tends to backfire while praising them for their lack of prejudice tends to further entrench them in their bias, then it would seem that no matter what one does, things will only get worse. In Character as Moral Fiction (Alfano 2013, 93), I addressed this problem, pointing out that trait-labeling is likely to function as a self-fulfilling prophecy (instead of backfiring) only when the labeled person has a “roughly correct conception” of the trait in question. There, I lamented the fact that labeling “as reasonable a person with a history of unreasonable behavior who believed in his own reasonableness” wouldn’t help and might hurt. And in Alfano (2015) I pointed to John Boehner’s assertion that he was “the most reasonable, responsible person here in Washington” as evidence that he couldn’t be reasoned with. In the same way, labeling as unbiased a person with a history of biased behavior who believed in his own lack of bias likely wouldn’t help and might hurt.

    I cannot resist further illustrating this problem with the fact that, in January 2018, shortly after referring to every country in Africa as a “shithole,” Donald Trump told reporters, “I am the least racist person you have ever interviewed, that I can tell you.”4 Radke elaborates on this problem by arguing, in effect, that Donald Trump is not special. If DiAngelo’s arguments about white fragility are on the right track and generalize to fragility associated with other identities, contemporary societies are structured in such a way that people who benefit from racist, sexist, and other biases tend to lack the conceptual resources to make sense of either their own advantages or the rights violations that others encounter (Zack 2015). In other words, they tend to have inadequate conceptions of the virtues and vices associated with race, gender, class, disability, and so on. We might, stretching the definition of the phrase slightly, call this a systemic hermeneutic injustice (Fricker 2007). Whereas those who have reason to make accusations of bias typically possess adequate conceptual resources to understand those accusations, the people they accuse often lack the conceptual resources to understand what they’re being charged with. This in turn leads to the defensive displays that DiAngelo associates with white fragility.

    Are there ways out of this morass? I think so. But they’re slow, laborious, and uncertain. First, as I mentioned in my reply to Fischer, it’s probably best to stop worrying about the unicorns who embody implicit bias but are explicitly anti-racist, anti-sexist, and so on. There may be a few such individuals, but they are already on the royal road to virtue. From the point of view of responsibility and blame, what demands our urgent attention are the explicitly biased people who tend to get defensive when their prejudices are pointed out. And from the point of view of well-being and justice, what demands our even more urgent attention, concern, and solidarity are the people who suffer from these prejudices. Second, given the temporal and social scale of the problem, the current generations (by which I mean, in particular, Baby Boomers) may be hopeless. If one of the background conditions ensuring that discussions of prejudice tend to be fruitless is conceptual and linguistic incompetence, then hope lies in those who are still acquiring and enriching their conceptual and linguistic capacities. Slightly more concretely, what I’m recommending is a campaign to enculturate the coming generations in a way that makes them capable of understanding accusations of bias without automatically reacting in the ways DiAngelo describes. In other words, the sort of distal control that I recommend in the second chapter of Moral Psychology may need to be extended even further to the intergenerational scale. What this would look like in practice, I am not in a position to say. But I thank Radke for prompting me to reflect further on this matter.



    Alfano, M. 2013. Character as Moral Fiction. Cambridge: Cambridge University Press.

    ———. 2015. “Becoming Less Unreasonable: A Reply to Sherman.” Social Epistemology Review and Reply Collective 4: 59–62.

    ———. 2016. Moral Psychology: An Introduction. Malden, MA: Polity.

    DiAngelo, R. 2011. “White Fragility.” International Journal of Critical Pedagogy 3: 54–70.

    Fricker, M. 2007. Epistemic Injustice: Power and the Ethics of Knowing. Oxford: Oxford University Press.

    Pinker, S. 2011. The Better Angels of Our Nature: Why Violence Has Declined. New York: Penguin.

    Rawls, J. 1971. A Theory of Justice. Cambridge: Harvard University Press.

    Zack, N. 2015. White Privilege and Black Rights: The Injustice of U.S. Police Racial Profiling and Homicide. Lanham, MD: Rowman & Littlefield.



Alexandra Plakias’s Commentary on Moral Psychology

In this piece, I will focus on Alfano’s discussion of moral disagreement in chapter 5. He begins the chapter by telling the reader that the “main conclusion will be that we do not know how much and how deeply different people and cultures disagree and . . . perhaps we cannot know, given the conceptual, empirical, and moral difficulties associated with investigating moral disagreement” (140). But along the way Alfano argues for another thesis as well, namely that fundamental moral disagreements are fewer than antirealists (and others impressed by the empirical record) have claimed. I will argue that he is too quick to dismiss the possibility that particular moral disagreements can be fundamental.

Moral Disagreement: Fundamental versus Superficial, Particulars versus Principles

In Mackie’s influential 1977 discussion of the argument from disagreement (or “the argument from relativity,” as he called it), he lays out the argument and immediately introduces what he describes as “a well-known counter” to it, “namely to say that the items for which objective validity is in the first place to be claimed are not specific moral rules but very general basic principles . . . recognized at least implicitly . . . in all society” (37). Mackie dismisses this response rather quickly, on the grounds that it doesn’t adequately explain how particular moral judgments can be objectively true. Alfano’s main concern is not to reject the possibility of objective moral facts, of course, so he doesn’t have the same stake in the outcome that Mackie does; hence he gives this response the more detailed discussion it deserves. Nonetheless, I think he is too convinced by it, for reasons I’ll now discuss.

Why does Alfano think that disagreement about particular cases is not fundamental? Here he tells us that “to count as fundamental, a disagreement should be about something deep” (150), a requirement implied, he argues, by the very term “fundamental,” with its derivation from the Latin term for “foundation.” Why are particular judgments neither deep nor foundational? To illustrate the point, Alfano offers us the case of two agents disagreeing over the morality of a particular action. Both agree that the action is wasteful and fair, he imagines, but they weight the importance of these features differently, with one placing greater importance on efficiency and one on “outcome based distributive justice.” Because they agree on the general features of the case, and on the moral worth of the values in play, Alfano argues that this is not a case of fundamental disagreement; “this case seems insufficiently deep on its own to count as fundamental.” Perhaps. But this case will not, if the description is accurate, take place on its own—not in the “real world.” For if one agent—or culture—prioritizes efficiency over justice, this difference will surely manifest itself in many, rather than a single, particular case. We would therefore expect to see a pattern of disagreements concerning cases where efficiency and outcome-based distributive justice conflict. And if, in cases where two values A and B conflict, one group consistently weights A over B while another group consistently weights B over A, we will, I think, judge that these groups have different values. Yes, they both take A and B to be valuable, but the priority they accord these values conflicts.

Alfano argues that cases of disagreement about how to implement “the same or nearly identical values” are not instances of fundamental disagreement. But such cases may be evidence against the very idea that the values in question are the same or nearly identical. Enough disagreement about implementation, and we may conclude either that the values in question aren’t shared, or that the values in question are fairly vacuous. Consider one of Alfano’s examples. In a modern-day variant on Herodotus’ story about Darius, he asks us to imagine that Obama invites a modern-day Englishman and Frenchwoman to the White House and asks each how much they would have to be paid to drive on the right and left sides of the road, respectively. The fact that the two drivers would resist any such offer tells us only, Alfano thinks, that they don’t want to kill anyone—not that they value driving on any side of the road per se. The analogy is supposed to be with values like respect: the fact that we disagree over what constitutes respectful treatment of a corpse doesn’t show us that we have a fundamental disagreement, because we all agree that we should respect the dead. But there’s something funny about this case, namely, that the value that determines what side of the road we drive on is just something like, “coordinate with other people.” And this is a fairly valueless statement. I don’t drive on the right because I don’t want to kill people; I drive on the right because everyone here drives on the right. The analogy in this case appears convincing precisely because the driving case is one in which we care very little about the act itself. However, we don’t value treating our dead as others treat their dead. If we did, we wouldn’t give much thought to eating our dead if we ended up in a country where others ate theirs, burning them in a country where others burned theirs, and so on. But this is precisely not the case: changing our treatment of the dead, or our attitude towards marrying our sisters (to use another of Herodotus’ examples) is not something we would do lightly or without real emotional turmoil—unlike changing the side of the road we drive on. It’s true that Herodotus describes such practices as matters of “custom,” but he precedes the anecdote by stating that everyone believes the custom of their own upbringing to be the best—so much so that “only a madman would mock at such things.”

A further objection concerns the claim that particular disagreements aren’t “foundational.” I don’t know whether this is intended to invoke a foundationalist picture of moral epistemology, but it certainly has the connotation, and which picture of justification one has in mind does affect how one views moral disagreement as a problem. On a foundationalist picture, we intuit a set of basic principles, values, or duties, and apply these to particular cases. But there is no hard-and-fast rule about how to apply them; instead, it’s at the level of duties that we find agreement, and intuitive justification. When duties conflict, there may be no particular right answer. But in such cases our verdicts are justified inferentially insofar as they are based on the foundational moral judgments. On this picture, particular judgments are shallow in that they are not implicated in the foundations of moral justification. But as Alfano himself recognizes by quoting Weber, this is not the only picture available; on a coherentist picture, or a reflective equilibrium model, particular judgments play just as important a role in justification as general principles, duties, or values. None is foundational; alternatively, all are foundational. That’s because particular judgments can provide a reason for rejecting or reevaluating a basic value, just as a basic value can give us reason to accept or countenance an otherwise objectionable particular judgment.

I want to conclude by returning to the main thesis of the chapter, namely, the prospects for deciding the argument from disagreement on empirical grounds. Like Alfano, I am dubious that this can be done. Alfano cites several challenges to resolution: identifying cross-cultural disagreements requires interpreting a culture’s beliefs and practices (which in turn may require making a claim about “the” beliefs or values of a culture—itself a difficult if not conceptually confused task, given the heterogeneity of cultures); we also face the challenge of interpreting a culture’s practices and translating their value claims into our own language. In doing so, we must employ some sort of charity—but there are multiple ways to do this, since we can apply charity at various levels such as reference, truth, or ontological commitments. I think Alfano is correct about all of these challenges, and I would add yet another one: even at the level of individuals, what we value is sometimes difficult to discern, not only to observers, but even to ourselves. Our actions fail to line up with our professed values and beliefs in ways big and small. Sometimes we feel confident in claiming that an individual has lied about their values, or is merely paying lip service, as when someone claims to value serving the community but consistently passes up opportunities to volunteer their time or money. But other cases, and I think the more common type of case, may be more complicated, and here we have to decide whether an individual’s values are revealed by her behaviors, the beliefs those behaviors suggest, the beliefs that she professes, her implicit attitudes and biases, and so on. This challenge, not incidentally, is mirrored in the laboratories of experimental philosophers and psychologists: when we ask people for moral intuitions, do we ask them in first- or third-person scenarios? If there’s a discrepancy between the answers to the questions, “what should Jack do” and “what would you do,” which do we take as indicative of subjects’ moral judgments and values? If it turns out that subjects’ professed intuitions don’t line up with or predict their behavior in actual, real-world scenarios (assuming such a discovery were even possible; surely the world lacks sufficient runaway trolleys), to which do we defer, the intuition or the action? This decision is itself a normative judgment, and is likely to be the subject of disagreement: is the real moral judgment the one made in cool reflective conditions, or the hot, affect-backed reaction that immediately moves us to action? Our decision here might itself be influenced about our intuitions about particular cases, such that a more emotional judgment is discounted because it fails to maximize utility (see, e.g., Singer, “Ethics and Intuition,” for an argument to this effect). Far from peripheral, particular cases and the moral disagreements they engender are at the very foundation of our thinking about morality and moral psychology.



Mackie, J. M. 1977. Ethics: Inventing Right and Wrong. New York: Penguin.

Singer, P. 2005. “Ethics and Intuition.” Journal of Ethics 9: 331–52.

  • Avatar

    Mark Alfano


    Reply to Plakias

    Thanks to Alexandra Plakias for her nuanced comments on Moral Psychology: An Introduction. Unlike the other three commenters, who focus in different ways on responsibility, bias, and prejudicial dispositions, Plakias addresses the fifth chapter of the book. In that chapter, I put forward a rather stringent definition of a deep moral disagreement and argue that the empirical evidence—rather than showing that there are such disagreements between cultures—is best interpreted as suggesting that there is massive moral agreement between cultures.

    Plakias offers several criticisms of this argument. First, she is skeptical about my claim that differences in the priority assigned to distinct values do not suffice for deep disagreement. This is a hard case to resolve, as it involves translating a difference in degree into a categorical distinction. If I assign slightly more weight to one value than you do, it seems silly to say that we have a deep disagreement. By contrast, if I assign almost no weight to a value and you assign it supreme weight, we might reasonably call that a deep disagreement (provided neither of us is willing to budge or compromise). What about intermediate cases? I see no principled way to draw a line. It’s a Sorites paradox of axiology.

    That said, I am comforted by the eye-popping correlation between individuals’ values documented by Fischer et al. (2010), which is as high as .98. I’m further comforted by Saucier’s (forthcoming) finding that the ordinal rankings of individual values are also remarkably similar. I review this research on page 164 of the book. One might respond to these statistics with a shrug. After all, even Saucier’s findings are only about ordinal rankings of values. You and I could both rank A above B and still disagree significantly about how much more important A is than B. The only way to handle this criticism would be to develop and use a measure of values that is sensitive to cardinal and not just ordinal values. To my knowledge, no one has done this; and if my arguments in the first chapter of the book are on the right track, it may be impossible.

    Nevertheless, Curry et al. (2018) further corroborate, using a very different methodology, my claim that there is nearly universal agreement across cultures about which values are important. Curry and his colleagues show that the same seven values (which function as solutions to non-zero-sum cooperative games) are universally venerated across sixty anthropologically well-documented societies ranging from small-band hunter-gatherers to industrialized democracies. With a total of 3,460, exactly one ran counter to the hypothesized model. That kind of unanimity is almost never seen in social science; if this paper were not on the Open Science Framework, I would assume that it was a fraud.1

    But Plakias has two further arguments in favor of skepticism: implementation-disagreement and anti-foundationalism. According to the former, stark-enough disagreement about how to implement shared values should itself count as a deep moral disagreement. As she puts it, “Enough disagreement about implementation, and we may conclude either that the values in question aren’t shared, or that the values in question are fairly vacuous.” According to the latter argument, I illegitimately presuppose a foundationalist moral epistemology, which relieves me of the burden of showing that there are no deep disagreements about particular cases. But, urges Plakias, I am not entitled to such an epistemology (indeed, I don’t accept one), and there are reasons to think that a coherentist moral epistemology would turn up deep disagreements about particulars.

    Regarding implementation, we once again face the difficulty of translating a difference in degree into a categorical distinction. How different do your and my favored ways to implement the same values have to be for our disagreement to count as deep? As before, where exactly to draw the line is arbitrary. I also suspect that, because implementing shared values is often a collective action problem, the gradients are likely to be jagged. I’ll strongly prefer my way of implementing our shared values up until enough other people signal that they prefer yours and are prepared to sanction me for acting otherwise, then switch all at once to strongly preferring your way. That’s because I, like a plurality of people, am a conditional norm-follower (Bicchieri 2005). If Curry et al. (2018) are right, then clear Pareto improvements should garner the approval of all parties to a disagreement about implementation. But, in many cases, improvements in the aggregate are likely to bring negative consequences for some people in their wake. Strident disagreements about how to implement shared values are likely to crop up when such changes are proposed or enacted. Does this make the underlying values vacuous? Only if they don’t motivate people to find a solution to such disagreements. Sometimes, I speculate, they do. Indeed, Cristina Bicchieri (2017) has made impressive advances in both measuring and effectuating such social changes related to practices like child marriage, female genital cutting, and corporal punishment of children.

    Finally, regarding moral epistemology, Plakias is right to criticize my implicit foundationalism in the book. This was, as she mentions, in part due to the fact that the debate I was engaged in is about fundamental moral disagreements. But one shouldn’t be so easily bamboozled by etymology. What would a deep or fundamental moral disagreement look like in a coherentist framework? Coherentists model attitudes as webs or networks of cognitive and affective states and dispositions (Quine 1970). A network doesn’t have a bottom or foundation; instead, it has a center. Recent methodological advances have made it possible to measure and model values as networks and to assign centralities to their various nodes (Alfano et al. 2017; Alfano and Higgins forthcoming; Brandt et al. 2018). The question then becomes: do the central value-nodes in these networks differ in significant and modally robust ways across cultures? It’s still early days, but my (self-serving) interpretation of the evidence is that they do not. If this is on the right track, then the next question is whether the morally-relevant cognitive and affective states that tend to form the core of people’s mental networks are particular or general. Are the most central nodes things like It was wrong of my father to vote for Donald Trump in 2016 or Racism is evil (or both)? Quine had little to say about ethics, but his view was that abstractions and principles, such as the law of non-contradiction, would be at the center of the web of belief. Perhaps what holds in logic also holds in ethics. More research is needed to answer this question.

    I’d like to take this opportunity to thank Plakias and the other commentators again for their time, attention, and engagement. With any luck, a second edition of Moral Psychology: An Introduction will be published in the coming years, and I will then have the chance to revise based on their feedback.



    Alfano, M. 2013. Character as Moral Fiction. Cambridge: Cambridge University Press.

    Alfano, M., et al. 2017. “Identifying Virtues and Values through Obituary Data-Mining.” Journal of Value Inquiry.

    Alfano, M., and A. Higgins. Forthcoming. Natural language processing and semantic network visualization for philosophers. In E. Fischer & M. Curtis (eds.), Methodological Advances in Experimental Philosophy. Bloomsbury.

    Bicchieri, C. 2005. The Grammar of Society: The Nature and Dynamics of Social Norms. Cambridge: Cambridge University Press.

    ———. 2017. Norms in the Wild: How to Diagnose, Measure, and Change Social Norms. Oxford: Oxford University Press.

    Brandt, M., et al. 2018. “What Is Central to Belief System Networks?” PsyArXiv Preprints. DOI, 10.17605/OSF.IO/TYR64. URL,

    Curry, O., et al. 2018. “Is It Good to Cooperate? Testing the Theory of Morality-as-Cooperation in 60 Societies.” Current Anthropology.

    Fischer, R., et al. 2010. “Are Individual-Level and Country-Level Value Structures Different? Testing Hofstede’s Legacy with the Schwartz Value Survey.” Journal of Cross-Cultural Psychology 41: 135–51.

    Quine, W. V. O. 1970. The Web of Belief. McGraw-Hill.

    Saucier, G. Forthcoming. “Value Hierarchies Within and Across Cultures: A Comparative Test of 18 Value Theories.”

    1., accessed February 4, 2018.