Passions and Emotions

NOMOS LIII

James E. Fleming

• Publisher: NYU Press
• Year: 2012
• ISBN:
  9780814760215
• Topic: Politics & International Relations
• Subtopic: Politics
• Index: Politics & International Relations

Throughout the history of moral, political, and legal philosophy,many have portrayed passions and emotions as beingopposed to reason and good judgment. At the same time,others have defended passions and emotions as temperingreason and enriching judgment, and there is mountingempirical evidence linking emotions to moral judgment. In Passions and Emotions, a group of prominent scholars inphilosophy, political science, and law explore three clustersof issues: “Passion & Impartiality: Passions & Emotions inMoral Judgment”; “Passion & Motivation: Passions & Emotionsin Democratic Politics”; and “Passion & Dispassion:Passions & Emotions in Legal Interpretation.” This timely,interdisciplinary volume examines many of the theoreticaland practical legal, political, and moral issues raised by suchquestions.

PART I
PASSION AND IMPARTIALITY:
PASSIONS AND EMOTIONS IN
MORAL JUDGMENT

1
CONSTRUCTIVE SENTIMENTALISM:
LEGAL AND POLITICAL IMPLICATIONS

JESSE J. PRINZ

There is mounting empirical evidence linking emotions to moral judgment. Though open to competing interpretations, this evidence is best interpreted as supporting the kind of sentimentalist theory associated with philosophers of the Scottish Enlightenment. Recent findings also allow us to update sentimentalism by specifying which emotions contribute to moral judgment, and research in history, anthropology, and cross-cultural psychology is providing richer insights into the origins of our emotionally grounded values. If values are emotionally based and culturally diverse, there may be moral conflicts that have no rational resolution. This would have implications for normative ethics, for politics, and for law. Here I review the empirical case for sentimentalism and then draw attention to some of these implications.

1. THE PLACE OF EMOTIONS IN MORAL JUDGMENT

The empirical turn in ethics has been fueled, in part, by the emergence of moral neuroscience. In 2001, Joshua Greene and his collaborators published a paper showing neural activations as people reflected on trolley dilemmas.¹ Since then, scores of other studies have appeared. Brains have been scanned as people make judgments of wrongness, engage in reciprocal exchanges or charitable giving, play morally significant video games, and look at morally meaningful photographs. Throughout the many studies, one common denominator has been emotion. Again and again, areas of the brain associated with emotional response are active when people engage in moral cognition. These areas include the posterior cingulate, temporal pole, insula, ventromedial prefrontal cortex, amygdala, and ventral striatum. Authors of the studies interpret these results in different ways, but one common refrain is that emotions contribute to moral judgment.

The exact nature of that contribution, however, is difficult to assess using neuroimaging, which is a correlational method. We can see this by considering the wide range of models that are compatible with the finding that emotional activations regularly co-occur with moral judgment.

The rationalist model says that some kind of reasoning—either conscious deliberation or unconscious rules—drives our moral judgments, with emotions arising as a consequence.² The dual-process model says that emotion drives moral judgments some of the time but that reason can also, depending on the case.³ The intuitionist model says that emotions constitute intuitions about what is right or wrong and that we use these intuitions to make our judgments; reason then follows to provide post hoc rationalizations.⁴ Neo-sentimentalists claim that moral judgments are judgments about whether emotional responses are merited (“I should feel guilty/angry about this”), and we can imagine emotions that typically arise in conditions where we deem them appropriate.⁵ The constitution model, which was endorsed by the sentimentalists of the Scottish Enlightenment, says that emotions are components of moral judgments: to think that something is morally wrong is to have a negative emotion toward it.⁶ Both the neo-sentimentalists and the old-school sentimentalists agree with intuitionists that reason often plays a post hoc role in the moral domain, but, as we will see in a moment, reason can also make a more substantive contribution.

Given the ambiguity of imaging evidence, how do we decide among these models? I think there are good reasons to be dubious about all but the constitution model. The evidence is behavioral and philosophical. The rationalist model neither explains nor predicts the well-established fact that emotions can influence moral judgments. For example, people assess vignettes as more wrong than they otherwise would if they are hypnotized to feel disgust,⁷ exposed to noxious smells,⁸ or asked to imbibe bitter beverages.⁹ People make more utilitarian judgments when amused¹⁰ and more deontological judgments when feeling elevated.¹¹ Clearly emotions are not just an effect of moral judgment.

The dual-process model has been supported by appeal to the fact that consequentialist judgments show less activation in emotion centers than deontological judgments in trolley dilemmas, and individuals with ventromedial brain injuries—known for disrupting emotion-based inferences—are more likely than others to make consequentialist judgments. These findings are intriguing, but the fMRI data also clearly show that consequentialist judgments show more emotional activation than nonmoral judgments. Likewise, ventromedial patients do not lack emotions; the very fact that they engage in reward-seeking behavior¹² shows that they are at least capable of experiencing and acting on appetitive emotional states. Their deficit principally involves an inability to curtail reward seeking in light of negative feedback. When confronted with a trolley dilemma, we must normally decide between two moral injunctions: it would be good to help the people in need, and it would be bad to harm someone in the process. These can be regarded as a positive and a negative norm, respectively. VM patients seem to be motivated by the former and indifferent to the latter. They don’t lack emotions; they simply lack the ability to regulate positive in light of negative emotions. The dual-process model is undermotivated.

The intuitionist model advanced by Jonathan Haidt is an improvement but remains, in an important way, obscure. Haidt suggests that emotions precede moral judgments. That implies that they are not components or parts. What, then, are moral judgments? It can’t just be that they are sentences of English, like “Cannibalism is wrong,” because one can make a moral judgment without verbalizing it. Also, if moral judgments are the effects of emotions, then they should be able to occur without emotions, just because most effects can come about in different ways. But Haidt offers no evidence that we can make moral judgments without emotions, and that would go against the spirit of his approach. A further issue concerns Haidt’s very strong skepticism about the role of reason in morality. He gives the impression that reasoning never contributes to moral deliberation. That seems implausible. We often need reason to determine whether something is morally significant. This is especially clear in policy decisions. Is inheritance tax unjust? Is late-term abortion permissible? Should factory farming be regulated? Should we fight to stop vaginal circumcision? Should we assist in foreign wars? Haidt would have us believe that such cases are rare or that the reasoning here always involves some kind of blind social conformity, but there is little reason for such a cynical view.

Let’s turn from these psychological theories to an account that has gained currency in philosophy: neo-sentimentalism. This turn covers a range of positions, but they share in common what can be called the meta-move. They say that moral judgments are judgments about the merit, warrant, or appropriateness of an affective response. The meta-move is designed to improve on traditional sentimentalism, which says that thinking something is wrong is a matter of having a negative emotion toward it. Clearly, we sometimes have negative feelings toward things that we would not, on reflection, view as wrong. The wrong is not simply that which causes our disapproval; it is that which warrants it. Or so the story goes. But the view faces some serious worries. First, it seems to mislocate the object of moral judgments; when we say that killing is wrong, we are saying something about killing, not about our feelings.¹³ Second, it is hard to define merit without circularity. Depending on circumstances, a murder might merit fear (prudentially speaking) or forgiveness (if we aim for reconciliation). Saying that killing merits anger is to say that it merits it morally, even if these other considerations make different emotions more appropriate overall. But it would be circular to define a moral judgment as a judgment that certain emotions are morally merited, because that is just another moral judgment.¹⁴ There is also an empirical worry: some individuals (children and some people with autism) make moral judgments easily but lack the capacity to form beliefs about emotional states.¹⁵

This brings us to the constitution model, which says that moral judgments contain emotions. To judge that something is wrong, on this view, is to have a negative emotion toward it. This seems to be the kind of view Hume and his contemporaries had in mind, and it overcomes all the difficulties that confront the other models. The most obvious objection to the constitution model is that people often seem to have negative emotions while withholding moral judgments. A person raised in a homophobic community might later in life experience disgust when seeing homosexual affection while insisting that the observed activity is morally acceptable. Doesn’t this show that moral judgments are not constituted by emotions? An alternative explanation is that such an individual does in fact think homosexuality is wrong at some level but also thinks it’s permissible and identifies with the latter conviction. Compare the person who exhibits implicit racism but also believes in racial equality. We should say in both cases that there is an automatic, bigoted appraisal that happens to get outweighed by a considered appraisal. In a questionnaire study, I was able to show that this, in fact, is how ordinary people interpret such cases.

Another worry about the constitution model pushes in the opposite direction, pointing out that we can make moral judgments without emotions. Empirically, this has not been explored, but cases are easy enough to imagine. When we speak in generalities (cruelty is wrong) or about complex policies (sin tax is wrong) or during episodes of numb depression, we might not feel strong moral emotions. But here I caution that we must recognize that there are at least dispositions to emote. Someone who did not shudder at a case of cruelty could not be credited with truly believing the generalization that cruelty is wrong. I like to distinguish emotions, which are occurrent states, from sentiments, which are emotional dispositions. I think moral judgments usually contain emotions, but our long-standing moral values are sentiments. In some cases, when we say that something is wrong, we are communicating that we have a certain value, not making a judgment based on that value. Compare: I can declare, “Sushi is delicious” at 6:00 A.M., when I have no desire to eat it. But this statement of value would be empty were I not disposed to experience sushi as delicious.

In summary, I think the constitution model can withstand objections and account for the data better than its alternatives. But the model still needs some fleshing out. Moral judgments contain negative feelings, but which ones? Clearly, not every bad feeling is a case of moral judgment.

2. THE NATURE OF EMOTIONS IN MORAL JUDGMENT

To read the literature on sentimentalism, one might think there is a single emotional state called disapproval. But the empirical literature suggests that moral disapproval is felt differently in different contexts. We can speak of a family of disapproval emotions.¹⁶

One important emotion in this family is anger and its variants, indignation and outrage. Anger arises most typically in cases in which we learn that someone has intentionally harmed or sought to harm another. Another form of disapproval is disgust. This arises when we encounter crimes against nature, such as violations of sexual taboos, even when no one is harmed (necrophilia, for example). Corresponding to these other-directed emotions, we also experience self-directed disapproval when we ourselves misbehave. Guilt arises when we harm others, and shame arises when we violate sexual taboos.

I think all of these emotions originate outside the moral context. Anger is a feeling of the body’s preparation to aggress, and that can occur when we are under threat, even if no norm has been violated: consider the anger mustered by two boxers in a bout. Disgust is a feeling of the body’s preparation to expel contaminants, and it can arise when seeing or tasting rotten foods. Even shame and guilt may have nonmoral variants. Shame is a kind of unpleasant embarrassment, manifested as a feeling of the body as we try to conceal ourselves from others. Guilt may be a blend of fear and sadness—a feeling in the paradoxical state of flight preparation even as we reach out dolefully to those we have harmed. Sadness characteristically arises when we become separated from those we care about, and this is precisely the risk we incur when we harm someone.

Given that emotions of disapproval can arise in nonmoral contexts, one might wonder what distinguishes the moral cases from these others. The answer, I think, has to do with the distinction between self- and other-directed emotions just adduced. If you see me eating rotten food, you will feel disgusted; but if you eat rotten food yourself, you will not feel ashamed, you will feel disgusted—an outward emotion for both cases. If you are boxing, you might feel aggressive irritation when your sparring partner hits you, but you won’t feel guilt when you hit back. The moral domain is distinguished from the nonmoral by the pairing of disgust and shame on the one hand and anger and guilt on the other. A judgment ...