Abstract
The recent introduction of the concepts of tenderness, vulnerability, and parental care in the field of moral emotions has brought us closer to an understanding of the underlying mechanisms, but has not yet resulted in a systematic evolutionary and proximate analysis. Applying such an analysis, the present article proposes a hypothetical care mechanism that produces different motivational states or moral emotions (e.g., tenderness, sympathy, guilt, moral anger) in response to individuals perceived as vulnerable. The mechanism consists of a care system automatically triggered by vulnerability cues, a system for defense and aggression, and a process of causally attributing changes in the vulnerable object’s well-being. The mechanism may also be useful in explaining moral responses in domains other than harm prevention.
Conceived very broadly, morality is about the desirability or goodness of improving others’ well-being, health, maturation, or genetic fitness. Complementarily, it is about preventing people, by means of condemnation, threat, or punishment, from reducing each other’s well-being. Identifying the psychological mechanisms underlying morality is difficult due to the different reasons or motives that people may have to benefit others, and the different kinds of responses that are associated with these motives. For example, I may benefit others by obeying them, to maintain a reciprocal or cooperative relationship with them, or to improve their well-being per se. The first motive may derive from fear of punishment for not delivering the benefit, the second from wanting to ensure a future return benefit (perhaps in combination with feelings of distrust and anger at potential free riders), and the latter from a disposition to care about the well-being of others and feelings of sympathy. Reflecting this complexity, many theorists distinguish between different domains of morality such as hierarchical relationships, reciprocity, and harm prevention or reduction (e.g., Haidt, 2007; Janoff-Bulman & Carnes, 2013; Rai & Fiske, 2011).
However, morality in the sense of promoting others’ well-being per se (without a hidden agenda) has received the most attention from philosophers and psychologists, presumably because it is both highly valued and hard to exercise. There are at least two reasons why it is so difficult to be sure of the existence of a specific psychological foundation for promoting others’ well-being. First, although several theorists have postulated the existence of “other-regarding sentiments” (Gintis, Bowles, Boyd, & Fehr, 2003), a disposition to care for strangers (e.g., Gilligan, 1982), or an “approach” system prescribing people to help others (Janoff-Bulman & Carnes, 2013), they have not yet described in detail the mechanisms associated with these dispositions. Mammalian parental care too, has been associated with a care disposition, probably because it so vividly and beautifully illustrates self-sacrifice, altruism, and nonaggressive or tender feelings towards vulnerable others. Indeed, many 18th- and 19th-century thinkers associated morality with parental care and its characteristic emotion of tenderness, and many 20th-century psychologists continued to make the same association (see Dijker & Koomen, 2007). However, while its relevance to morality makes great intuitive sense, it is difficult to explain why a parental care mechanism first needed after reproductive age can have an adaptive function outside parent–offspring relationships. This generalization not only seems necessary to explain spontaneous helping and moral emotions in young children (Eisenberg, 2000; Warneken & Tomasello, 2009), but also among adult strangers.
A second problem associated with trying to identify a specific mechanism for care is that, even if something similar to parental care would exist in generalized form, it may be difficult to recognize among the other motives that may accompany true care and that may be necessary to make it effective. For example, caregivers may have to punish aggressively their object of care in order to teach it to take better care of its own health, or they may primarily devote themselves to protecting it fearfully and/or aggressively against others with harmful intentions. In these cases, there seems little room for tenderness and sympathy.
The goal of the present article is to provide a more precise description of a hypothetical care mechanism and its interaction with motivational systems that are involved in self-preservation (e.g., flight and aggression), to show that such a description helps to explain in a parsimonious way the different moral emotions that have been associated with the domain of harm prevention, and to give evolutionary reasons for its generalization beyond parental care and kinship. An additional goal of the article is to speculate about the relevance of the proposed mechanism for explaining moral responses in other domains than harm prevention. The next section starts with examining why it is necessary to identify a mechanism specifically devoted to caring for others’ well-being.
Why Care about Other’s Harm and Suffering?
Briefly consider the two major ways in which social psychologists tend to explain an emotion like sympathy. First, sympathy may be seen as a consequence of attributing perceived suffering or harm to situational causes (Weiner, 2006; see also Goetz, Keltner, & Simon-Thomas, 2010), thereby highlighting the lack of responsibility or victim status of the suffering person. A recent variant of such a primarily cognitive explanation of sympathy arousal is the idea that perceivers represent victims as “patients,” contrasting them with “agents” responsible for their harm (Gray & Wegner, 2011). A second explanation of sympathy assumes that it is partly based on the ability of perceivers to take over or mirror others’ suffering, either directly in response to certain expressive signs of distress or sadness (Preston & de Waal, 2002), or as a result of a more effortful cognitive process of perspective taking or empathy (Batson, 1987, 2011).
Both explanations, however, fail to mention a mechanism that would explain why perceivers who normally also experience fear or distress at the sight of suffering (Batson, 1987; Eisenberg, 2000), would care about alleviating or removing the other’s suffering; an essential motivational aspect of sympathy (Wispé, 1991). Why do they not avoid or reduce attention to the victim? It is clear that the altruistic aspect of sympathy is not sufficiently explained by alluding to processes of attribution, mirroring, increased attention to the target, or perspective taking. A logical step in trying to explain it is to assume that people are disposed to care for others (see previous lines). However, without specifying the mechanisms associated with such a disposition this assumption does not contribute much to a better explanation of sympathy and prosocial behavior.
Sometimes, the altruistic element of sympathy is explained by invoking a role for acquired positive attitudes towards the victim, associated with feelings of attachment, familiarity, and liking (for a review, see Park, Schaller, & van Vugt, 2008). However, these attitudes only help to explain why the victim is selected or relatively favored as a recipient of care. For example, the “right” recipients of sympathy and care are kin-related individuals (Hamilton, 1964; Park et al., 2008) or those who have helped the perceiver in the past (Trivers, 1971). Yet, this selectivity leaves unexplained how need states are recognized and can motivate helping behavior that is both relevant and effective in improving the situation of needy individuals (whether positive attitude objects or complete strangers).
A recent introduction of additional explanatory concepts has brought us closer to a description of the mechanisms underlying moral emotions such as sympathy. The concepts at stake here are tenderness, vulnerability, and parental care (Batson, 2011; Batson, Lishner, Cook, & Sawyer, 2005; Brown, Brown, & Preston, 2012; Dijker, 2001, 2010, 2011; Kalawski, 2010; Lishner, Batson, & Huss, 2011; Numan, 2012; Preston & Hofelich, 2012). Although there are differences in theoretical interpretation, authors seem to converge on the following relationships among these terms. First, tenderness is a distinct emotion, responsible for motivating true care, protection, and different prosocial behaviors.
Second, perceived vulnerability is a perceptual or cognitive antecedent of tenderness. Importantly, vulnerability is not the same as harm or suffering (Dijker, 2001). Most theories of morality focus on the perception of harm as a causal antecedent of moral emotions, thereby failing to recognize (a) that the perceived vulnerability of unharmed and healthy individuals can trigger tenderness, an essentially pleasant emotion; and (b) that the mechanism underlying the arousal of this emotion may also be responsible for the arousal of moral emotions typically experienced when actual suffering is perceived or expected.
Third, the mammalian parental care system is a plausible candidate mechanism for vulnerability perception and the arousal of tenderness, also explaining why perceived vulnerability and protectiveness are strongly influenced by age-related physical features (Berry & McArthur, 1986; Dijker, 2001; Lishner, Oceja, Stocks, & Zaspel, 2008).
Fourth, rather puzzling, the pleasant emotion of tenderness can also be aroused while observing others in acute need and suffering, and hence can co-occur with distress and fear. One explanation of this phenomenon would be that harm and suffering can be seen as evidence for vulnerability as harm makes one vulnerable to more harm. Alternatively, tenderness could have been aroused by certain vulnerability cues that are unrelated to suffering, allowing one to perceive the suffering as inconsistent with the protective tendencies that are associated with tenderness and to experience other emotions as well.
There are at least two lines of empirical research that offer preliminary support for these relationships. First, exposure to vulnerable stimuli (e.g., an infant, puppy) not in acute need may arouse tenderness but not sympathy, while exposure to those in acute need may evoke both tenderness and sympathy (Lishner et al., 2011), or both tenderness and sadness (Kalawski, 2010). The latter two findings should not be surprising because since the early 1980s, Batson (1987, 2011) has demonstrated repeatedly that feelings that are very similar to tenderness such as softheartedness and warmth typically are aroused by increased attention to needy or suffering others. Yet, the relationship with tenderness may have been overlooked for such a long time because Batson used the label empathy to describe these feelings, as did the many researchers who used his empathy measure (for critical discussions, see Dijker, 2001; Wispé, 1991).
Second, there is increasing evidence that tenderness and other emotions motivating prosocial behavior correlate with the neural and hormonal activity associated with the mammalian parental care system (Brown et al., 2012; Numan, 2012; Panksepp, 1998; Preston & Hofelich, 2012). Especially the role of oxytocin, a hormone released by a paraventrical part of the hypothalamus and acting on the parasympathetic nervous system, is worth noting since its physiological effects (e.g., a fall in blood pressure and cortisol levels, inhibition of flight and fear; Panksepp, 1998) seem responsible for certain subjective aspects of tenderness (e.g., calmness, feelings of relaxation) in both parents and nonparents.
Although introducing the concepts of tenderness, vulnerability, and parental care allows important theoretical distinctions to be made, it has not yet resulted in a comprehensive mechanistic analysis of different moral emotions or an evolutionary explanation for the usefulness of these concepts outside parent–offspring relationships.
This article proposes that a more fundamental analysis of the property vulnerability may not only resolve this evolutionary puzzle, but also suggest the general properties that a psychological mechanism for the effective provisioning of care should have. The next two sections use an evolutionary perspective to identify this mechanism. A subsequent section will show how a task analysis (Tooby & Cosmides, 1992) of such a hypothetical care mechanism reveals different subgoals, subtasks, and motivational states (moral emotions) that, dependent on the perceiver’s assessment of the situation, would be required to promote the fitness or well-being of individuals perceived as vulnerable.
A Psychological Adaptation for Responding to Vulnerability
With respect to living things, vulnerability can be defined as the property or disposition of objects to change into a state of lowered fitness (i.e., a state that is inconsistent with genetic “design specifications”) when exposed to certain conditions. For those concerned with the fitness and well-being of others (see following lines), an assessment of vulnerability would be crucial as information about this property can help perceivers to predict and thus prevent actual harm. After all, especially in ancestral environments, it would have been much better to prevent injury than to try to relieve harm already inflicted and likely resulting in death (Dijker, 2011).
Like other properties or dispositions, vulnerability cannot be directly perceived. Instead, perceivers may form sensorimotor, what-leads-to-what expectancies by observing the perceptual consequences of interacting with objects (e.g., Gallese & Lakoff, 2005). For example, an object’s vulnerability can be internally represented by the expectancy that the object will deform or break when manipulated in certain ways, and its heaviness by the expectancy that a particular intensity of motor output or force is needed to lift the object to a certain height.
Fortunately, organisms do not always need to form experience-based internal representations of properties in order to respond adaptively to objects. In particular, objects and their properties may be reliably correlated with certain cues that can automatically trigger adaptive behavioral mechanisms. One can imagine that early organisms primarily responded to fitness-relevant properties by means of a relatively simple nervous system directly connecting sensory neurons to effector cells; with the first firing, for example, when exposed to light patterns that were correlated with the presence of food, and the latter causing the body to move into the direction of the light source. Similarly, an object’s vulnerability may be correlated with certain cues such as relatively small size, small weight, transparency or other signs of fragility, lack of muscles, visual or vocal signs associated with young age, and environmental cues signaling imminent danger.
Although the genetic cost–benefit model of kin selection or inclusive fitness (Hamilton, 1964) has been primarily used to study how kinship cues and kin recognition mechanisms positively affect prosocial behavior (Park et al., 2008), this model also allows the prediction that a reactive psychological mechanism will evolve that specifically responds to vulnerability cues and helps to prevent the harmful treatment of vulnerable kin.
The mechanism at stake here is best referred to as care mechanism. Its activation is most urgent when vulnerable children compete with each other or are responded to by more mature and stronger individuals. In these cases, the most important thing the mechanism should accomplish is to reduce aggressive tendencies in the perceiver that are directed at the vulnerable object. The mechanism may explain self-handicapping and other means of aggression inhibition during rough-and-tumble play (Fry, 2005) and spontaneous helping among siblings (Warneken & Tomasello, 2009). At a later developmental stage, this mechanism may also be involved in asymmetrical parental care for young and immature offspring.
Figure 1 proposes that reduction of aggression and flight can be realized through reciprocal neural inhibition between a fight-or-flight system that is activated by threat cues, and a care system that is specifically activated by vulnerability cues (for a general treatment of neural competition between behavioral systems or “decision making,” see Enquist & Ghirlanda, 2005). There is considerable neurophysiological evidence for a competition between a fight-or-flight and a care system in mammals, including humans. The former is associated with activation of the sympathetic adrenomedulary system, the latter with, among other things, effects of oxytocin (see previous lines). (Figure 1 greatly simplifies matters, leaving out feedback and learning processes and the formation and use of internal representations of object properties, let alone the role of symbolic mental activity.)
A care mechanism consisting of a motivational care system specifically responding to vulnerability cues and a fight-or-flight system specifically responding to threat cues.
However, instead of inhibition of aggressive or fearful tendencies, changes in attention may also result in “cooperation” or quick alternation between activity of the fight-or-flight and care system, resulting in a fearful but concerned and defensive attitude (see Figure 1). This cooperation is also suggested by neurophysiological evidence. For example, when lactating rats are injected with oxytocin, both maternal behavior toward pups and aggression toward intruders are intensified (Bosch, Meddle, Beiderbeck, Douglas, & Neumann, 2005).
In addition to defense against third parties, it is conceivable that aggressive output of the fight-or-flight system can also be used to punish and teach the target to promote its own health and fitness, or to motivate perceivers to punish themselves and change their behavior, as in guilt.
Vulnerability Is All Around
In humans, an exceptionally strong care system may have evolved, easily activated by the slightest evidence of vulnerability and immaturity. Complementarily, humans may have evolved a rich variety of ways to signal vulnerability and immaturity in order to trigger a care system in others, thereby ensuring aggression reduction, diverse prosocial behaviors, and different moral emotions to be discussed in the next section. For example, vulnerability displays can be recognized in rough-and-tumble play (Fry, 2005) and politeness rituals (Eibl-Eibesfeldt, 1989). In addition, a person’s vulnerability can be derived from attributing the individual’s behavior to external or situational causes.
The evolutionary reasons for strong sensitivity to vulnerability and development of vulnerability signaling in humans may be sought in a combination of giving birth to altricial babies, exceptionally long childhood, and a cooperative breeding system in which group members respond in nonaggressive ways to immature individuals and often take part in (allo)parenting (Hrdy, 2009).
A Task Analysis of the Care Mechanism: Explaining Moral Emotions
When the care system in Figure 1 is sufficiently activated, the mechanism should produce goal-directed behavioral and mental activity to promote the fitness or health of a vulnerable object. In particular, the activated mechanism should motivate perceivers to detect changes in the object’s current fitness or well-being, evaluate them, and determine their cause (for convenience, it will be assumed that another’s fitness can be reliably estimated on the basis of perceived changes in health or well-being). It is further assumed that perceived changes in the object’s well-being may also activate self-preservational mechanisms (e.g., those dedicated to flight or aggression), resulting in a particular motivational state that fine tunes the perceiver’s prosocial tendencies. These motivational states can be described with emotion terms that psychologists use to refer to typically human moral emotions (for reviews and discussions, see Eisenberg, 2000; Tangney, Stuewig, & Mashek, 2007).
Table 1 describes how the different goals and tasks of a perceiver, in whom the care mechanism is activated, depend on detecting and causally attributing four different types of changes in the care object’s well-being. In particular, the two types of changes that are undesirable from the perceiver’s point of view are a threat to well-being and an actual reduction in well-being. The two types of changes that are desirable for the perceiver are a restoration of well-being to an acceptable level after it has been reduced, and a general improvement in well-being or fitness due to maturation. Furthermore, the perceiver is expected to attribute the changes in well-being to three potential causes: The care object itself, the self (the individual in which the care mechanism has been activated), and a third party.
Fitness-promoting tasks, sample behaviors, and relevant human emotions from the perspective of a perceiver in which a care mechanism has been activated, as a function of a perceived change in well-being of a care object and its perceived cause
Undesirable Changes in Well-Being
When the perceiver observes cues associated with a threat to, or actual reduction in, the object’s well-being, it will be the perceiver’s general goal to prevent (further) reduction in well-being or to restore reduced well-being. First consider the situation in which the nature of a threat is still unclear (last column of Table 1). A typical motivational state here (concern or worry) would be one that motivates the perceiver to gather more information about the possible presence of agents or conditions that could be dangerous to the object of care, and to engage in mental simulation to assess the nature and relevance of these events. Indeed, concern and worry seem to be especially cognitive or mindful emotions (Ortony, Clore, & Collins, 1988).
However, the motivational state may transform into sympathy when the three potential causes of the object’s lowered well-being are less relevant and attention is exclusively focused on the object’s need state. It is argued that it is the perceived vulnerability of the object and the resulting activation of the care system that makes the actual suffering appear undesirable and as something to be removed (Wispé, 1991), for example, with efforts to heal or comfort. When the care object is close to death or has died, sadness and grief may take over, associated with desperate attempts to bring back the lost object or to ensure its well-being in the afterlife (Archer, 1999). Fultz, Schaller, and Cialdini (1988) found strong relationships among sympathy, concern, and sadness.
When the care object is under threat or actually harmed, aggression too may be involved in benefitting the object and in the arousal of the moral emotions. Indeed, aggressive or angry responses at a causally responsible agent seem to be involved in both guilt and moral anger. In particular, guilt can be conceived as anger directed at the self, motivating self-punishment and behavioral change to prevent or undo harm (Baumeister, Stillwell, & Heatherton, 1994; for evidence of aggressive elements in guilt, see Nelissen & Zeelenberg, 2009; Roseman, Wiest, & Swartz, 1994). Most theorists recognize that guilt is based on previous harmful behavior toward relationship partners (Baumeister et al., 1994; McCullough, Kilpatrick, Emmons, & Larson, 2001). Involvement of the care system in guilt explains the relation between guilt and helping behavior and between guilt, sympathy, and tenderness (Baumeister et al., 1994; Tangney et al., 2007).
Guilt may also be accompanied by shame when the harm being done is associated with fear of (public) punishment by dominant others or a powerful crowd (implying a strongly activated fight-or-flight system in Figure 1). However, when experienced in isolation and in the absence of a care object or an activated care mechanism, shame may result in mere submission, avoidance, or outward aggression (Combs, Campbell, Jackson, & Smith, 2010) rather than in intrinsic interest in changing one’s own behavior and repairing the harm (Gilbert, 2003; Tangney et al., 2007).
Yet, when the salient cause of a threat or an actual fitness reduction is a third party, moral anger or indignation may be experienced, resulting in a relatively unforgiving and sometimes even self-sacrificial form of punishment (Fehr & Gächter, 2002) in order to make sure that the third party will not produce further harm.
Note that aggression may also be directed at the object of care itself, in order to induce it to care more about its own fitness. An object that is thought to be unnecessarily activating the perceiver’s care mechanism (e.g., by endangering itself repeatedly, refraining from self-grooming, or laziness) may be viewed as a social parasite, thus inviting disrespect, a milder form of moral anger (Dijker & Koomen, 2007).
The present two-system account of sympathy and moral anger helps to explain in mechanistic terms how the word responsibility can have such a strong and opposite influence on the experience of these emotions (Dijker & Koomen, 2003; Weiner, 2006). In particular, presenting people as accident victims or suffering patients makes them appear vulnerable and triggers the care system. Emphasizing that this person is lacking in responsibility, more strongly activates the care system because it stresses the victim’s immaturity, “innocence,” or vulnerability to situational forces. In contrast, emphasizing that the (care-arousing) victim is responsible for his or her own condition or is unmotivated to get out of it, activates the aggression component of the fight-or-flight system, generating anger and aggression, which may in turn inhibit activation of the care system. As in the case of shame, a too strong activation of the fight-or-flight system may reduce the role of care-based moral emotions. For example, a hierarchical relationship between perceiver and target may primarily result in contempt and stigmatization of vulnerable individuals (Dijker & Koomen, 2007).
Desirable Changes in Well-Being
The two desirable changes in the care object’s well-being are expected to produce positive emotions such as tenderness, admiration, pride, and gratitude. Tenderness is defined here as an emotion that is caused by perceiving that the level of fitness or well-being of a vulnerable individual is acceptable, and motivates the perceiver to keep it at this level, at a minimum by simply staying near the object and careful handling (Table 1). Keeping well-being at an acceptable level may also be interpreted as reducing the object’s vulnerability by facilitating and stimulating its maturation through nurturance, feeding, or play. The positive and pleasant aspects of a tender response to vulnerable but fit individuals seem to be captured by the appraisal term cuteness; a term that recently has been associated with morality (Sherman & Haidt, 2011). Interestingly, tenderness experienced while perceiving a vulnerable but healthy object may also be associated with aesthetic experience. Such a relation was proposed by Burke (1759/1990) who suggested that beautiful things tend to be relatively small, smooth, soft, lightly colored, and appearing delicate, weak, and dependent.
Another emotion closely related to tenderness is being moved (to tears) by watching the affectionate and protective responses of third parties toward a care object. Important for this emotion is the expression (crying, shedding of tears) but not the unpleasant experience of sadness, probably due to remembering or imagining one’s own vulnerability, together with evidence for protection and consolation (Miceli & Castelfranchi, 2003).
Finally, consider responses to a salient restoration of a lowered fitness (see Table 1). Gratitude is typically considered an emotion that is felt by an object of care itself after having been helped, and that motivates the object to help the benefactor in return, thereby contributing to establish patterns of reciprocal helping (e.g., Bartlett & DeSteno, 2006; McCullough et al., 2001; Trivers, 1971). However, gratitude may also be felt by caregivers when a third party is seen as causally responsible for restoring the care object’s lowered fitness. Then the expression of gratitude would motivate the third party to benefit the object again in the future; which would be especially necessary if the benefactor is a stranger and hence provides help unexpectedly (McCullough et al., 2001). Furthermore, awareness that the self has been responsible for improving others’ well-being may also arouse gratitude or pride. Indeed, a recent study found that gratitude can be felt when individuals expect to posthumously donate their organs to needy patients (Dijker, Nelissen, & Stijnen, 2013). When the object itself is seen as responsible for a restoration of well-being or fitness, admiration for its competence may be felt. Expression of this emotion in terms of praise may increase the object’s self-confidence, thereby ensuring future success in protecting its own well-being. Finally, when improvement in the object’s well-being is attributed to mysterious or supernatural interventions, admiration may be tinged with fear or awe (Keltner & Haidt, 2003).
A relatively complex but morally important motivational state, presumably also related to activation of a care mechanism but not included in Table 1, is forgiveness. This state can be interpreted as a reduction in anger at harmdoers through the inhibitory activity of a care system. The latter mechanism may be activated when harmdoers express remorse (e.g., by saying things like “I feel terrible about what I did” or by crying or begging for forgiveness), thereby giving evidence for their vulnerability and evoking sympathy or guilt in angry perceivers who expect to punish and harm them (Gold & Weiner, 2000; McCullough, Worthington, & Rachal, 1997).
Preliminary Evidence for Vulnerability as a Common Basis of Moral Emotions
As mentioned earlier, evidence for the existence of a care mechanism comes from studies showing the influence of oxytocin on a wide variety of prosocial behaviors, and from studies revealing the motivational aspects of different vulnerability cues such as infantile features that probably can activate the mechanism. However, thus far only one study has demonstrated that different moral emotions may have a common basis in the perception of vulnerability. In a study by Dijker (2010), participants reported different moral emotions in relation to different full-body photographs of male and female target persons widely differing in age, and in perceived vulnerability and felt protectiveness. Using variation between stimulus objects, it was found that the different moral emotions not only were highly intercorrelated but also strongly related to perceived vulnerability and aroused protective tendency; with children and elderly arousing the strongest and adult males the weakest, emotions. Importantly, correlations among these emotions largely disappeared when vulnerability and protectiveness were controlled.
Studies making use of individual differences additionally suggest a common mechanistic basis for several of these emotions. For example, individuals prone to feel tenderness or sympathy are also more likely to feel guilt (Tangney et al., 2007) and forgiveness (McCullough et al., 1997).
Future tests of the present theory should employ more sophisticated methods such as a priming paradigm to show how activation of a care system helps to arouse a variety of moral emotions and to establish the boundary conditions of this phenomenon.
Conclusions
The hypothetical psychological mechanism proposed in this article provides an integrative explanation of a wide variety of moral emotions that until now have been treated independently from one another. This explanation combines a more traditional, cognitive-appraisal or attribution account of emotions (e.g., Ortony et al., 1988; Weiner, 2006) with a description of underlying motivational systems or “instincts” that are linked to the different goals that organisms may strive at under different circumstances. This account assigns a central place to the emotion of tenderness in morality, as was advocated earlier by many 18th- and 19th-century thinkers on the subject. It also highlights certain complex issues in explaining moral emotion and behavior that must be addressed in future research, in particular the sometimes cooperative relationship between a care and a fight-or-flight system. As the analysis has shown, many situations require this cooperation, for example, in order to aggressively defend vulnerable individuals or teach them in a mildly aggressive manner to care more about their own fitness and health. This suggests that even in the absence of acute harm and suffering, vulnerable targets do not always trigger tender and pleasant feelings in perceivers (an assumption made by, for example, Kalawski, 2010; Lishner et al., 2011), or require the inhibition of a fight-or-flight system (as suggested by, for example, Brown et al., 2012; Preston & Hofelich, 2012).
The proposed mechanism may also help to address the role of perspective taking in causing prosocial behavior. According to the present view, an emotion like sympathy requires that a care mechanism is automatically activated by perceived vulnerability cues, producing an immediate readiness to pay full attention to changes in the well-being of others and their determinants. Rather than being necessary for the arousal of this emotion (Batson, 1987), perspective taking may accentuate the vulnerability of suffering individuals and may help perceivers to recognize effective ways of responding. Interpersonal attitudes too, do not seem necessary for sympathy arousal. Instead, they may help to lower the activation threshold of the care system (see Figure 1) and as such explain variation in intensity of moral emotions and selectivity in helping.
To what extent can the proposed care mechanism help explain moral responses in other domains than harm prevention? First note that moral responses in certain domains may have little to do with the proposed mechanism. For example, valuing obedience in the context of a hierarchical interpersonal relationship may be primarily driven by fearful submission and contempt for those who are disobedient and violate norms. Other domains, however, may allow room for a care mechanism. For example, as Trivers (1971) recognizes, reciprocity not only requires a felt obligation to reciprocate help and a sensitivity to free riding, but also the arousal of a motivation to help others when needed. Furthermore, even in the absence of apparent need states and suffering, perceptions of vulnerability can motivate cooperation. For example, framing a social interaction in terms of a prisoner’s dilemma may heighten people’s awareness that both the self and the other are dependent on each other and hence are vulnerable, thereby inducing different moral emotions that may prevent cheating.
Young and Dungan (2012) recently pointed out that neuroscience has found little evidence for the existence of a specific module for morality in the brain and that morality may primarily depend on general mechanisms for emotion arousal and social information processing. They advise researchers to step back and perform a functional analysis to determine what the different moral domains have in common and what exactly distinguishes moral from nonmoral issues. The present article may contribute to such a theoretical enterprise by describing how a large part of morality may depend on a single care mechanism while also identifying the conditions that may obscure its presence or that depend more on other mechanisms and thus represent truly different moral domains.