Joshua May's (Reference May2018) new book is a tremendous and much-needed intervention in the field of moral psychology. May argues that, notwithstanding a voluminous literature on the various foibles, “bugs,” and biases to which we are prone, our moral capacities are basically sound. He walks us carefully through the many halls of this literature, executing a suite of reliability checks that show these findings to be less sensational than they appear in headlines. May performs a great service by systematically cataloging these pessimistic challenges all in one place, and his comprehensive, ably argued, many-sided defense will no doubt reinvigorate key debates over rationalism versus sentimentalism, consequentialism versus deontology, and egoism versus altruism.
In this commentary, however, I consider the perspective of an ordinary, self-reflective moral agent with no philosophical horse in these races – just a “concerned (moral) citizen,” as it were, wondering how worried she should be. I offer what might be called a critical pessimism: critical in the sense of being explicitly concerned with ameliorating injustice, and in the tradition of “normative reflection that is historically and socially contextualized” (Young Reference Young1990) – that is, sensitive to how moral reasoning is always embedded within a specific sociopolitical and historical moment. Writing on the edge of the 2018 U.S. mid-term elections, I suggest there is still reason for a certain pessimism about our moral capacities, even if they are not fundamentally arational and egoist.
I begin with two worries about the twin centerpiece of May's case for optimism: the Debunker's Dilemma (for skeptics of moral cognition) and the Defeater Dilemma (for skeptics of moral motivation).
May uses the Defeater Dilemma to show that motivation can be genuinely moral, that is, it is not fundamentally egoistic. He concedes that a particular form of self-interest – desiring to see oneself as morally good – is indeed a main basis of wide-ranging moral behavior, as demonstrated by studies of motivated moral reasoning, moral licensing, and moral hypocrisy. Yet he argues that we could just as well consider such a desire to be a form of moral integrity, because it reflects an agent's authentic concern for morality. Thus, rationalizations of bad moral behavior are themselves evidence of moral integrity, because self-serving outcomes are made to seem appealing via the enlistment of bona fide moral reasoning. I find this a compelling and ingenious argument against the egoist who believes true moral motivation is rare or non-existent.
However, accepting the widespread influence of this species of moral integrity seems to cut against arguments defending the general reliability of moral cognition. May writes, for instance: “You can criticize a politician who passes anti-discrimination legislation only to acquire votes, but not if she does it primarily because she wants to see herself as doing the right thing and would otherwise feel guilty” (p. 204). But here May is only assessing moral motivation, in a case where the desire to view oneself as moral works in the right direction. If we consider a moral rationalization case where it leads a self-righteous politician to pass voter-suppression laws despite initial qualms, we can and should criticize her – precisely because her moral integrity has corrupted her moral judgment.
May might respond by pointing out that this is not unique to moral cognition; indeed, he uses desire's distorting effect on prudential reasoning to illuminate his account of rationalization. He only wants to demonstrate that moral reasoning is not inherently different from non-moral reasoning in being dependent on emotions, as sentimentalists would have it. However, there is an important disanalogy here. Prudential rationalization runs up against natural limits: those oft-cited indulgent desserts will actually prevent weight loss, and too many splurges add up to an empty bank account. In other words, bad prudential reasoning will as a matter of course work against an agent's prudential interests, because it represents a mistaken theoretical understanding of the world. This is far more tenuous for moral rationalization, however, because moral reasoning is practical rather than theoretical: it concerns the world as it ought to be, not as it actually is. Hence, there is no natural “check” on bad moral judgments, making it possible for the moral rationalizer to dig in his heels against deleterious consequences. Others’ suffering, for example, can be rationalized by painting them to be sufficiently unlike oneself, be it less competent, less deserving, or less human. Thus, moral cognition can plunge into horrific abysses, as with May's (and today's) sincere Nazis (p. 176), because it is uniquely susceptible to distortion by the very desire to be moral – that is, by moral integrity itself. The more widespread moral rationalization is, the less reliable our moral cognition.
I am not, however, trying to develop this into a sweeping argument subject to the Debunker's Dilemma, which demonstrates that there is no single factor – emotions, evolutionary influence, or framing effects – that threatens our moral reasoning as a whole. But note that skepticism does not require there be some “One True Debunker,” that is, the wide-scope claim that there exists some debunker universally undermining justification for moral judgments across all contexts. We would also be in trouble if we accepted the narrow-scope claim that for all contexts, there lurks some distinct debunker undermining justification in that particular context. I do not think our concerned citizen need be a global skeptic of this sort, because as May demonstrates, it is doubtful that the skeptic will be able to substantiate the strong narrow-scope claim. However, this disambiguation points us toward a different flavor of skepticism that can still be quite concerning. The worry here is not: “Are all of my moral judgments unjustified because they are influenced by some universal debunker”? but rather: “Does this particular token moral judgment involve a specific context that is prey to a narrowly debunking influence”? May does endorse such targeted debunking arguments. But in the absence of systematic principles for determining whether or not a given moral judgment falls within the restricted range of these “danger zones,” we might lose justification for it.
Now, May does specify one kind of context where we should be wary: peer disagreement can serve to debunk controversial moral judgments. But the obvious problem here is that these are contexts where we might be most worried about the veracity of specific token judgments (“Am I right to believe that violent resistance is never justified? that it is right to protect the spotted owl at the expense of loggers’ jobs?” etc.). This is where our concerned citizen would most want to depend on the reliability of moral cognition: yet it is precisely here that May's optimism reaches its limits.
On a very different note: Is it really advisable to be less confident in controversial moral beliefs (p. 128), given that these are often cases in which it is most important to stand up firmly for them? Consider “abnormal moral contexts” (Calhoun Reference Calhoun1989), where controversy arises precisely because moral knowledge within a specialized community – obvious examples include trans and disabled communities – develops faster than it can be disseminated externally. Here, standing in solidarity with an oppressed group might require me to defer to their judgment even when I disagree (Kolers Reference Kolers2016). The fact is that when it comes to contested moral terrain, there is usually more at stake than just the accuracy of my beliefs, that is, whether I end up with a correct or incorrect moral judgment. What we believe in controversial cases is not merely an epistemic but also a political matter: of whom we trust, to whose testimony we are willing to defer, and whose interests we value.
This brings me to a wider critical pessimism about the way that moral cognition is typically understood. In the actual world, moral reasoning is never really about the socially abstracted, stick-figure characters that feature in experimental vignettes. This means that we should be careful about optimistic results obtained through generating moral intuitions about underspecified individuals. Studies of gender attribution, for instance, demonstrate that in the absence of substantial cues, people assume the “default” person to be male (Hamilton Reference Hamilton1991; Merritt & Kok Reference Merritt and Kok1995). Similarly, some fans’ outrage over fictional characters being revealed as Black or gay suggest that their default is to imagine them as White and heterosexual (Hetter Reference Hetter2015; Holmes Reference Holmes2012). Such simplifying abstractions might be necessary for scientific investigation, but moral cognition in the wild always concerns full-blooded individuals replete with traits attributed on the basis of multiple social categories. It would be naive and dangerous, for example, to think that differences between liberals and conservatives come down only to different emphases on moral foundations (cf. Graham et al. Reference Graham, Haidt and Nosek2009), rather than being deeply intertwined with racial resentment, sexism, and reactionary White male rage. Yet social categories are rarely manipulated in studies of moral cognition, where “liberal” and “conservative” function as the primary contrast classes. May discusses only two (from the same publication), to which I will return shortly.
This disconnect between research on moral versus social cognition is, of course, not unique to May. But I worry that May underestimates the influence of social difference on moral cognition as a result of an overly narrow construal of moral judgment as explicit beliefs concerning the rightness and wrongness of acts. May circumscribes the threat of implicit biases, for example, by using evidence that they only weakly predict measured behaviors (p. 217). Elsewhere, he distinguishes between judgments about the wrongness of an act versus judgments ascribing blameworthiness (pp. 60–61) or negative moral traits to agents (p. 35), excluding the latter from his arguments. But why shouldn't these count as genuine and important instances of moral cognition? Much moral thinking and acting proceeds through attunement to morally relevant properties other than rightness/wrongness. So I am not sure we can so cleanly separate out moral from social cognition, as May seems to do. (Discussing social psychological effects of group membership on moral cognition, he writes that “these flaws can largely be attributed to cognitive biases present in other domains, not to something particular about moral cognition itself” [p. 129].)
Indeed, I would argue that, in the real world, moral cognition is in some sense social cognition, and vice versa, because we live in a world so deeply and unjustly stratified by social difference. Our biggest moral quandaries today are not so much difficult exercises in moral reasoning as they are failures to unite masses of people who should be making common cause against systems of domination that benefit a tiny minority at their expense – but who instead remain divided by race, gender, class, and so on. In the two studies that did consider the effects of racial difference on consequentialist versus deontological reasoning, the results were significant: liberals were less willing to sacrifice a Black (vs. White) man in a trolley scenario, and conservatives were less willing to sacrifice innocent American (vs. Iraqi) civilians (Uhlmann et al. Reference Uhlmann, Pizarro, Tannenbaum and Ditto2009). May and the authors interpret these findings as demonstrating the role of ideology in motivated moral reasoning. But they also highlight the pervasive (albeit unmeasured) entanglement of moral and social cognition. Social psychologists have long demonstrated a host of cognitive phenomena – fundamental attribution error (Ross Reference Ross1977), defensive attribution (Shaver Reference Shaver1970), just world hypothesis (Lerner Reference Lerner1980), systems justification theory (Jost et al. Reference Jost, Banaji and Nosek2004), to name a few – that underwrite people's rationalizations of a deeply unjust status quo. Thus, May's optimism might be dampened if we kept in view how these factors are always ineluctably in play within actual moral cognition.
To sum up, I think that there is reason for critical pessimism about how our moral-social capacities fare in a deeply unequal, highly segregated, hyper-partisan society. Under such conditions it is unsurprising that we continue to commit or allow moral atrocities against others, despite sharing the same, basically sound systems of moral cognition and motivation. I do not think that this critical pessimism should lead us to defeatism. But it should remind us, as others have argued (e.g., Vargas Reference Vargas2013a), that our moral agency is profoundly dependent on scaffolding from our social contexts. We cannot reason or act morally well in a badly ordered society – and none of us, moral psychologists included, should rest easy with this.
You have
Access
Joshua May's (Reference May2018) new book is a tremendous and much-needed intervention in the field of moral psychology. May argues that, notwithstanding a voluminous literature on the various foibles, “bugs,” and biases to which we are prone, our moral capacities are basically sound. He walks us carefully through the many halls of this literature, executing a suite of reliability checks that show these findings to be less sensational than they appear in headlines. May performs a great service by systematically cataloging these pessimistic challenges all in one place, and his comprehensive, ably argued, many-sided defense will no doubt reinvigorate key debates over rationalism versus sentimentalism, consequentialism versus deontology, and egoism versus altruism.
In this commentary, however, I consider the perspective of an ordinary, self-reflective moral agent with no philosophical horse in these races – just a “concerned (moral) citizen,” as it were, wondering how worried she should be. I offer what might be called a critical pessimism: critical in the sense of being explicitly concerned with ameliorating injustice, and in the tradition of “normative reflection that is historically and socially contextualized” (Young Reference Young1990) – that is, sensitive to how moral reasoning is always embedded within a specific sociopolitical and historical moment. Writing on the edge of the 2018 U.S. mid-term elections, I suggest there is still reason for a certain pessimism about our moral capacities, even if they are not fundamentally arational and egoist.
I begin with two worries about the twin centerpiece of May's case for optimism: the Debunker's Dilemma (for skeptics of moral cognition) and the Defeater Dilemma (for skeptics of moral motivation).
May uses the Defeater Dilemma to show that motivation can be genuinely moral, that is, it is not fundamentally egoistic. He concedes that a particular form of self-interest – desiring to see oneself as morally good – is indeed a main basis of wide-ranging moral behavior, as demonstrated by studies of motivated moral reasoning, moral licensing, and moral hypocrisy. Yet he argues that we could just as well consider such a desire to be a form of moral integrity, because it reflects an agent's authentic concern for morality. Thus, rationalizations of bad moral behavior are themselves evidence of moral integrity, because self-serving outcomes are made to seem appealing via the enlistment of bona fide moral reasoning. I find this a compelling and ingenious argument against the egoist who believes true moral motivation is rare or non-existent.
However, accepting the widespread influence of this species of moral integrity seems to cut against arguments defending the general reliability of moral cognition. May writes, for instance: “You can criticize a politician who passes anti-discrimination legislation only to acquire votes, but not if she does it primarily because she wants to see herself as doing the right thing and would otherwise feel guilty” (p. 204). But here May is only assessing moral motivation, in a case where the desire to view oneself as moral works in the right direction. If we consider a moral rationalization case where it leads a self-righteous politician to pass voter-suppression laws despite initial qualms, we can and should criticize her – precisely because her moral integrity has corrupted her moral judgment.
May might respond by pointing out that this is not unique to moral cognition; indeed, he uses desire's distorting effect on prudential reasoning to illuminate his account of rationalization. He only wants to demonstrate that moral reasoning is not inherently different from non-moral reasoning in being dependent on emotions, as sentimentalists would have it. However, there is an important disanalogy here. Prudential rationalization runs up against natural limits: those oft-cited indulgent desserts will actually prevent weight loss, and too many splurges add up to an empty bank account. In other words, bad prudential reasoning will as a matter of course work against an agent's prudential interests, because it represents a mistaken theoretical understanding of the world. This is far more tenuous for moral rationalization, however, because moral reasoning is practical rather than theoretical: it concerns the world as it ought to be, not as it actually is. Hence, there is no natural “check” on bad moral judgments, making it possible for the moral rationalizer to dig in his heels against deleterious consequences. Others’ suffering, for example, can be rationalized by painting them to be sufficiently unlike oneself, be it less competent, less deserving, or less human. Thus, moral cognition can plunge into horrific abysses, as with May's (and today's) sincere Nazis (p. 176), because it is uniquely susceptible to distortion by the very desire to be moral – that is, by moral integrity itself. The more widespread moral rationalization is, the less reliable our moral cognition.
I am not, however, trying to develop this into a sweeping argument subject to the Debunker's Dilemma, which demonstrates that there is no single factor – emotions, evolutionary influence, or framing effects – that threatens our moral reasoning as a whole. But note that skepticism does not require there be some “One True Debunker,” that is, the wide-scope claim that there exists some debunker universally undermining justification for moral judgments across all contexts. We would also be in trouble if we accepted the narrow-scope claim that for all contexts, there lurks some distinct debunker undermining justification in that particular context. I do not think our concerned citizen need be a global skeptic of this sort, because as May demonstrates, it is doubtful that the skeptic will be able to substantiate the strong narrow-scope claim. However, this disambiguation points us toward a different flavor of skepticism that can still be quite concerning. The worry here is not: “Are all of my moral judgments unjustified because they are influenced by some universal debunker”? but rather: “Does this particular token moral judgment involve a specific context that is prey to a narrowly debunking influence”? May does endorse such targeted debunking arguments. But in the absence of systematic principles for determining whether or not a given moral judgment falls within the restricted range of these “danger zones,” we might lose justification for it.
Now, May does specify one kind of context where we should be wary: peer disagreement can serve to debunk controversial moral judgments. But the obvious problem here is that these are contexts where we might be most worried about the veracity of specific token judgments (“Am I right to believe that violent resistance is never justified? that it is right to protect the spotted owl at the expense of loggers’ jobs?” etc.). This is where our concerned citizen would most want to depend on the reliability of moral cognition: yet it is precisely here that May's optimism reaches its limits.
On a very different note: Is it really advisable to be less confident in controversial moral beliefs (p. 128), given that these are often cases in which it is most important to stand up firmly for them? Consider “abnormal moral contexts” (Calhoun Reference Calhoun1989), where controversy arises precisely because moral knowledge within a specialized community – obvious examples include trans and disabled communities – develops faster than it can be disseminated externally. Here, standing in solidarity with an oppressed group might require me to defer to their judgment even when I disagree (Kolers Reference Kolers2016). The fact is that when it comes to contested moral terrain, there is usually more at stake than just the accuracy of my beliefs, that is, whether I end up with a correct or incorrect moral judgment. What we believe in controversial cases is not merely an epistemic but also a political matter: of whom we trust, to whose testimony we are willing to defer, and whose interests we value.
This brings me to a wider critical pessimism about the way that moral cognition is typically understood. In the actual world, moral reasoning is never really about the socially abstracted, stick-figure characters that feature in experimental vignettes. This means that we should be careful about optimistic results obtained through generating moral intuitions about underspecified individuals. Studies of gender attribution, for instance, demonstrate that in the absence of substantial cues, people assume the “default” person to be male (Hamilton Reference Hamilton1991; Merritt & Kok Reference Merritt and Kok1995). Similarly, some fans’ outrage over fictional characters being revealed as Black or gay suggest that their default is to imagine them as White and heterosexual (Hetter Reference Hetter2015; Holmes Reference Holmes2012). Such simplifying abstractions might be necessary for scientific investigation, but moral cognition in the wild always concerns full-blooded individuals replete with traits attributed on the basis of multiple social categories. It would be naive and dangerous, for example, to think that differences between liberals and conservatives come down only to different emphases on moral foundations (cf. Graham et al. Reference Graham, Haidt and Nosek2009), rather than being deeply intertwined with racial resentment, sexism, and reactionary White male rage. Yet social categories are rarely manipulated in studies of moral cognition, where “liberal” and “conservative” function as the primary contrast classes. May discusses only two (from the same publication), to which I will return shortly.
This disconnect between research on moral versus social cognition is, of course, not unique to May. But I worry that May underestimates the influence of social difference on moral cognition as a result of an overly narrow construal of moral judgment as explicit beliefs concerning the rightness and wrongness of acts. May circumscribes the threat of implicit biases, for example, by using evidence that they only weakly predict measured behaviors (p. 217). Elsewhere, he distinguishes between judgments about the wrongness of an act versus judgments ascribing blameworthiness (pp. 60–61) or negative moral traits to agents (p. 35), excluding the latter from his arguments. But why shouldn't these count as genuine and important instances of moral cognition? Much moral thinking and acting proceeds through attunement to morally relevant properties other than rightness/wrongness. So I am not sure we can so cleanly separate out moral from social cognition, as May seems to do. (Discussing social psychological effects of group membership on moral cognition, he writes that “these flaws can largely be attributed to cognitive biases present in other domains, not to something particular about moral cognition itself” [p. 129].)
Indeed, I would argue that, in the real world, moral cognition is in some sense social cognition, and vice versa, because we live in a world so deeply and unjustly stratified by social difference. Our biggest moral quandaries today are not so much difficult exercises in moral reasoning as they are failures to unite masses of people who should be making common cause against systems of domination that benefit a tiny minority at their expense – but who instead remain divided by race, gender, class, and so on. In the two studies that did consider the effects of racial difference on consequentialist versus deontological reasoning, the results were significant: liberals were less willing to sacrifice a Black (vs. White) man in a trolley scenario, and conservatives were less willing to sacrifice innocent American (vs. Iraqi) civilians (Uhlmann et al. Reference Uhlmann, Pizarro, Tannenbaum and Ditto2009). May and the authors interpret these findings as demonstrating the role of ideology in motivated moral reasoning. But they also highlight the pervasive (albeit unmeasured) entanglement of moral and social cognition. Social psychologists have long demonstrated a host of cognitive phenomena – fundamental attribution error (Ross Reference Ross1977), defensive attribution (Shaver Reference Shaver1970), just world hypothesis (Lerner Reference Lerner1980), systems justification theory (Jost et al. Reference Jost, Banaji and Nosek2004), to name a few – that underwrite people's rationalizations of a deeply unjust status quo. Thus, May's optimism might be dampened if we kept in view how these factors are always ineluctably in play within actual moral cognition.
To sum up, I think that there is reason for critical pessimism about how our moral-social capacities fare in a deeply unequal, highly segregated, hyper-partisan society. Under such conditions it is unsurprising that we continue to commit or allow moral atrocities against others, despite sharing the same, basically sound systems of moral cognition and motivation. I do not think that this critical pessimism should lead us to defeatism. But it should remind us, as others have argued (e.g., Vargas Reference Vargas2013a), that our moral agency is profoundly dependent on scaffolding from our social contexts. We cannot reason or act morally well in a badly ordered society – and none of us, moral psychologists included, should rest easy with this.