find easiest to forgive when they hurt us.39 Roy Baumeister, Arlene Stillwell, and Todd Heatherton reviewed the social-psychological literature on guilt and found that it went hand in hand with empathy. More empathic people are also more guilt-prone (particularly women, who excel at both emotions), and it is the targets of our empathy who engage our guilt. The effect is enormous: when people are asked to recall incidents that made them feel guilty, 93 percent involved families, friends, and lovers; only 7 percent involved acquaintances or strangers. The proportions were similar when it came to memories of eliciting guilt: we guilt-trip our friends and families, not acquaintances and strangers.
Baumeister and his collaborators explain the pattern with a distinction we will return to in the section on morality. Sympathy and guilt, they note, operate within a circle of communal relationships.40 They are less likely to be felt in exchange or equality-matching relationships, the kind we have with acquaintances, neighbors, colleagues, associates, clients, and service providers. Exchange relationships are regulated by norms of fairness and are accompanied by emotions that are cordial rather than genuinely sympathetic. When we harm them or they harm us, we can explicitly negotiate the fines, refunds, and other forms of compensation that rectify the harm. When that is not possible, we reduce our distress by distancing ourselves from them or derogating them. The businesslike quid pro quo negotiations that can repair an exchange relationship are, we shall see, generally taboo in our communal relationships, and the option of severing a communal relationship comes with a high cost.41 So we repair our communal relationships with the messier but longer-lasting emotional glue of sympathy, guilt, and forgiveness.
So what are the prospects that we can expand the circle of sympathy outward from babies, fuzzy animals, and the people bound to us in communal relationships, to lasso in larger and larger sets of strangers? One set of predictions comes from the theory of reciprocal altruism and its implementation in Tit for Tat and other strategies that are “nice” in the technical sense that they cooperate on the first move and don’t defect until defected upon. If people are nice in this sense, they should have some tendency to be sympathetic to strangers, with the ultimate (that is, evolutionary) goal of probing for the possibility of a mutually beneficial relationship.42 Sympathy should be particularly likely to spring into action when an opportunity presents itself to confer a large benefit to another person at a relatively small cost to oneself, that is, when we come across a person in need. It should also be fired up where there are common interests that grease the skids toward a mutually beneficial relationship, such as having similar values and belonging to a common coalition.
Neediness, like cuteness, is a general elicitor of sympathy. Even toddlers go out of their way to help someone in difficulty or to comfort someone in distress.43 In his studies of empathy, Batson found that when students are faced with someone in need, such as a patient recovering from leg surgery, they respond with sympathy even when the needy one falls outside their usual social circle. The sympathy is triggered whether the patient is a fellow student, an older stranger, a child, or even a puppy.44 The other day I came across an overturned horseshoe crab on the beach, its dozen legs writhing uselessly in the air. When I righted it and it slithered beneath the waves, I felt a surge of happiness.
With less easily helped individuals, a perception of shared values and other kinds of similarity makes a big difference.45 In a seminal experiment, the psychologist Dennis Krebs had student participants watch a second (fake) participant play a perverse game of roulette that paid him whenever the ball landed on an even number and shocked him when it landed on an odd number. 46 The player had been introduced either as a fellow student in the same field who had a similar personality, or as a nonstudent with a dissimilar personality. When the participants thought they were similar to the player, they sweated and their hearts pounded more when they saw him get shocked. They said they felt worse while anticipating his shock and were more willing to get shocked themselves and forgo payments to spare their counterpart additional pain.
Krebs explained the sacrifice of his participants on behalf of their fellows with an idea he called the empathy-altruism hypothesis: empathy encourages altruism.47 The word empathy, as we have seen, is ambiguous, and