The burgeoning research on the psychology of empathy has clustered around its lack of expression in psychopathy and provided the important insight that the absence of the ability to feel emotions, including empathy, opens the door to immorality (Schleim ). Michael Spezio () takes a more positive approach to empathy by stressing its role in character formation and the development of love and compassion. He pushes against the view that reason in the moral life works in opposition to the emotions. He also resists Paul Bloom's () thesis that empathy is a negative rather than a positive moral attribute.

Bloom's argument is deliberately provocative: the phrase against empathy attracts public attention. He argues that empathy damages policy making by preventing a cool and rational approach to caring for others on a global scale. For Bloom, “Empathy is biased; we are more prone to feel empathy for attractive people and for those who look like us or share our ethnic or national background. And empathy is narrow; it connects us to particular individuals, real or imagined, but is insensitive to numerical differences and statistical data” (Bloom ).

But his alternative utilitarian calculus for policy making carries its own ethical difficulties. Why should the needs of minorities, for example, be supported if by ignoring their demands the sum of happiness is increased? Supporters of empathy do not claim that only empathy is needed to the exclusion of other moral virtues. Bloom's focus on the emotional aspect of empathy includes psychological problems of those who are hypersensitive to other's feelings. But is this true empathy or a distortion? He is clearly not aware of the classic tradition of practical wisdom as mediating between the intellectual and moral virtues. The intellectual virtue of practical wisdom allows its holder to perceive what virtue must entail. If the disposition veers far off what could be called the “mean” judged according to a given context, then it is no longer virtue but vice. The fact that some people are off the scale in terms of ability to feel empathy does not mean, as Bloom supposes, that empathy is morally redundant. Bloom's argument is flawed, even if, as Gregory Peterson () notes, there are elements of his position that strike an appropriate cautionary note.

I am also sympathetic to Spezio's () sharp critique of Bloom on the basis of alternative psychological evidence. I believe that a more integrative approach to reason and the emotions in the moral life and a stress on human identity has resonances with classic descriptions. The cardinal virtues of prudence or practical wisdom, justice, temperance, and fortitude were distinguished from the theological virtues of faith, hope, and charity that were infused by God's grace. The passions of the irascible power of the sensitive appetite included fear and anger, alongside virtues related to the concupiscible power of the sensitive appetite, such as temperance, chastity, and honesty, all bound together in the notion of “powers of the soul” (Rhonheimer ; for a more theological approach, see Westberg ). This plethora of virtues relevant for the moral life does not split apart reason and emotion, but sees them working in dynamic relationship with each other, with love (charity) having special precedence. Prudence, as intellectual virtue, is connected with the proper exercise of the moral virtues, while all infused virtues are connected with charity. Empathy and compassion both have a mixture of emotive and reasoned components, though the former is often regarded as more visceral.

My intention in this article is not simply to rehearse Thomistic virtues and find lines of compatibility (or not) with psychology. Rather, I will argue on philosophical grounds, drawing on Martha Nussbaum, that while empathy and compassion are distinct, they work together in the moral life. Further, without practical wisdom and charity, compassion does not reach maturity either. When dealing with virtues such as empathy and compassion, moral theology resists a purely secular understanding. Theology provides, in this instance, a challenge to “thin” descriptions of the moral life, while supporting richer interpretations in psychology that take into account religious as well as emotive reasons for particular actions. Empathy has been used by a number of writers, including Frans de Waal, in order to argue for a “bottom up” approach to the more general discussion of the evolution of morality (de Waal , , ; Deane‐Drummond, Arner, and Fuentes ). This has the advantage of demonstrating the presence of empathy in other social animals, hence its strong evolutionary lineage. But there are difficulties worth noting that are relevant to the case for opposing the dismissal of the explicitly moral value of empathy in human societies.

One difficulty arises from evolutionary anthropology, another from theology. In the first case, I discuss recent research that identifies what appear to be instances of deliberative compassion far back in the Homo lineage. In the early evolutionary history of hominins, distinctive forms of compassion arose that went beyond shorter term empathetic reactions found in other animals, and they imply a deep history of sustained compassion. I also discuss classic theological notions of infused virtue and a life of grace in relation to acquired, learned virtues, as elaborated by Thomas Aquinas; I ask how far and to what extent such descriptions of elevated infused forms of virtue, in particular the virtues of mercy and compassion, shed light on philosophical, anthropological, and psychological accounts. The infused virtues could be perceived as an account of what the moral life as a perfected life is like. This article cannot attempt to trace the complex relationships between religious belief and compassion, or the evolutionary relationship between when diverse religious beliefs appear and when compassion appears in early hominins. Rather, it intends to use theological tools in order to ask further questions about the way compassion might be interpreted as significant in the moral life, and how that might be meaningful in both religious and secular terms.

WHAT IS THE RELATIONSHIP BETWEEN EMPATHY AND COMPASSION?

Nussbaum () discusses the potential arbitrary nature of compassion. In so far as it rests on empathy, empathy could potentially make compassion “narrow and uneven,” an argument that bears some similarities with the current objections to the moral worth of empathy, understood by its critics as leading to partiality (Nussbaum , 386). The current debate on empathy has other similarities with a much wider discussion on compassion that has been ongoing for over 2,500 years (Nussbaum , 354). While there were heated debates about what eudaimonia meant, compassion did not play a central or positive role (Nussbaum ). This rejection of compassion as of any relevance to the moral life then impacts on later writers, including Spinoza, Descartes, Kant, Adam Smith, and Nietzsche (Nussbaum , 358).

Nussbaum admits that there was a positive spirit of egalitarian cosmopolitanism at the root of the classic anti‐compassion school, so it was not intended to promote cruelty, but rather a rejection that came from the belief in the inability of compassion to make good judgments, leading to resentment and hatred when the self is impacted by misfortune (Nussbaum , 362). Mercy was still allowed, but this was a mitigating judgment based on recognition of fault, rather than its denial (Nussbaum , 366). What is interesting here is that the ancient anti‐compassion school seemed to allow for the possibility that compassion could be schooled by reason in a way that modern psychology affirms but denies in practice, rather than in theory, by antiempathy advocates such as Bloom and his supporters. For the pro‐compassion school, it is wrong to suggest that those who have compassion can never recognize that some of its judgments may be illegitimate (Nussbaum , 372). Further, it is inappropriate to claim that compassion does not lead to respect for another because it also recognizes either vulnerability or misfortune. In sum, “the pro‐compassion tradition is preoccupied with getting the theory of value right, criticizing those who attach inappropriate importance to money, status, or pleasure” (Nussbaum , 372).

Pity, sympathy, empathy, and compassion are often confusingly muddled up in the literature, both ancient and modern, which makes it even harder to come to an appraisal of the relationships between them. Nussbaum's definition of empathy is one of the clearest: empathy is “an imaginative reconstruction of another person's experience without any particular evaluation of that experience” (Nussbaum , 302). The judgment that a person is in distress, and a desire to do something about it, is a judgment of compassion rather than empathy. It is important to note that Nussbaum is not claiming that this is all there is to compassion, but that compassion, when it overcomes negative emotions such as disgust, envy, and shame, entails judgment. Sympathy is also close to the term compassion, though lacking in the latter's intensity; for Nussbaum, “If there is any difference between ‘sympathy’ and ‘compassion’ in contemporary usage, it is perhaps that ‘compassion’ seems more intense and suggests a greater degree of suffering, both on the part of the afflicted person and on the part of the person having the emotion” (Nussbaum , 302). Pity, on the other hand, has contemporary nuances of condescension and superiority to the sufferer that is rather different from the classic Greek usage, where pity is the English translation of eleos and oiktos (Nussbaum , 301). Nussbaum prefers therefore not to use the term “pity,” even while admitting that much of the literature blurs this emotion, as well as conflating sympathy with compassion.

For Nussbaum, compassion has three elements that require reasoned judgment. The first judgment of compassion is one of measurement of size—what has happened to the person or creature is a serious event. The second judgment is one of nondesert: the person/creature did not deserve this to happen to them. The third judgment is eudaimonistic, meaning this person/creature is significant in terms of my goals (Nussbaum , 321). Nussbaum also names wonder: the ability to contemplate the worth of the other even apart from our own sense of flourishing (Nussbaum , 321–22). For Nussbaum, empathy in neuropsychological terms requires cognitive recognition of the otherness of the other, rather than simply feeling as if it were my pain, which is emotional contagion. Hence, empathy means both an awareness of another's pain and yet knowing it is not mine (Nussbaum , 327–28). Empathy requires an accurate imagining of what the other person/creature is feeling and is a prelude to compassion where those feelings are associated with a bad state. Empathy, however, just as easily can result in a lack of compassion, as when enemies read the intentions of their foes and manipulate them for their own purposes (Nussbaum , 329). Contemporary neuropsychological definitions of empathy do not start with the self–other distinction in the way that Nussbaum describes, but rather begin with an initial first‐person neurological motor unity triggered in the observing agent, recalling his or her similar experiences, and often attributed to mirror neurons. This unity of response is then followed by perspective taking, at which time a self–other distinction becomes possible (de Waal ). Further, neurological research on the relationship between empathetic concern and compassion defines both as emotional and motivational states, but only compassion is also infused with feelings of loving kindness toward the other, while confusingly, in comparison with Nussbaum's definition, both are characterized by a desire to help the other (Bernhardt and Singer , 3). Magnetic resonance imaging studies show that the networks activated when those who observe another in pain are the same as for first‐hand experience (Bernhardt and Singer , 2). Even more confusingly, the authors also associate empathy with both emotional contagion and compassion, distinguishing empathy associated with the latter by naming this empathetic concern where there may or may not be shared feelings.

Nussbaum also contends that compassion may be present without empathy, but that at the same time empathy is still a good guide for compassion (2001, 330). And limited forms of empathy associated with our relationships with other animals are possible, even if such forms have been criticized as anthropomorphic (Nussbaum , 333). A question that is relevant to the specific role of empathy in the moral life is present, Nussbaum argues, even when torturers confronts their victims in awareness of their suffering; for there is a step of depravity beyond “empathy informed hatred” when there is a lack of recognition of the other as a human person, a dehumanization of the other (2001, 334).

Nussbaum's analysis is highly illuminating in clarifying the relationships between empathy and compassion, as well in her interrogation of the classic literature in order to define more clearly what compassion means. However, her suggestion that empathy is merely a guide for compassion seems somewhat thin. So while theoretically, according to her definitions, it might be possible to show compassion without empathy, compassion is more meaningful and fuller as a virtue if it is inclusive of empathy rather than excluding it. She is also sensitive to historical trajectories in the meanings of different terms that are commonly confused, and she demonstrates in a helpful way important elements of the historical roots of heated debates between what could be called the compassion deniers and compassion protagonists regarding its importance in the moral life. This debate is parsed out in slightly different terms in the present debates within psychology on the role of empathy in the moral life, as represented, for example, in the work of Bloom and Spezio.

My own view is that the most common interpretation of the meaning of the term “pity” also fails to include an eschatological element in the way that compassion clearly does, that is, through hope that an alternative might be possible. Nussbaum admits that in some cases sympathy lacks eudaimonistic judgment (2001, 302, n.9). This is correct, although the contrast in contemporary usage is rather more extreme. Furthermore, pity also implies a more detached state, rather like common understandings of mercy, which Nussbaum deals with at some length (, 364–68). Mercy is, for Nussbaum, linked with justice; it is the reasoned choice not to punish to the extent that is deserved, and in that sense mercy can be accommodated in compassion‐deniers' theory of the moral life.

It is worth exploring further the more emotive content of compassion. In my view, Nussbaum fails to incorporate this emotive aspect in her understanding of compassion in her desire to stress cognitive reasoned judgments. In this respect, compassion is also complex insofar as it is intricately linked with the capacity for agape love for another in a way that pity is not. It seems to me that, although Nussbaum has defined clearly the judgments of compassion, placing the emotion of wonder in the role of paying attention to the worth of the other is barely sufficient. Her focus on the role of the judgments of compassion in overcoming natural impediments to care, such as disgust, tends to weaken the more common presupposition that compassion is linked with a more positive emotion of love and care for the other. This limitation may also be related to her somewhat truncated treatment of empathy noted above. One of the structural difficulties in her method is that she separates her extensive discussion of compassion from that of love, and while she alludes briefly to the relationship between them in her statement that “compassion pushes the boundaries of the self further outward than many types of love” (Nussbaum , 300), it is not immediately clear how love and compassion might be related, or precisely what this statement means. While there is a poetic link between love and sympathy in a section on Walt Whitman's poetry (Nussbaum , 646–55), influenced by a Christian understanding of love, the lack of a sustained treatment of the relationships between compassion and love in the volume as a whole is disappointing, given the sheer breadth of the material covered.

Hence, modifying her account, any definition of compassion needs therefore to include in a primary sense a positive affirmation of the judgment of compassion toward the other informed by love for the other, rather than, in the way that Nussbaum sets up her argument, a response to large negative events that have happened to another that were undeserved. Such responses can be included, but her notion of ordering toward “wonder” needs modifying to one of active compassion directed by love. A more expansive understanding of compassion also fits with “predictive” theories of the brain. According to such theories, the brain and mind are actively driven toward a particular vision of the other person. For compassion to be operative, this means at least implicitly envisaging the suffering other in a healthy state, along with a prediction of any role that the active compassion‐filled agent might play in bringing this about.

As I will develop further below, in Thomas Aquinas the Aristotelian position that Nussbaum also draws on is elaborated in the light of Augustine's focus on charity, so, just as practical wisdom or prudence is essential for the acquired moral virtues, charity is essential to the infused virtues. Practical wisdom is another way of conceiving the particular judgments of compassion insofar as they become, through mercy, capable of being moral acts. But before developing this thread, it is worth considering archaic human communities that, arguably and in a fascinating way, show glimpses of the first evidence for distinctly human forms of compassion.

THE EVOLUTION OF HUMAN COMPASSION

How did humans begin to express this complex mixture of empathy and reasoned compassion? Empathy is arguably one of the basic building blocks in the evolution of human (and sometimes other animal) morality (de Waal ). Evolutionary biology, however, generally avoids taking account of the psychological states of the actors involved, loosely using terms like selfishness or altruism without proper consideration of the intentions of the actors (de Waal ). For de Waal, the fundamental core of the moral life is necessarily an emotive one; so, incipient forms of justice become instead measures of “inequity aversion,” and compassion is parsed out in terms of directed altruism or intentional altruism. I am not intending to go into more detail of the work of de Waal here, for the reason that it has already received fairly copious attention, apart from suggesting that his theory has been very influential in discussions on the evolution of morality, especially among scholars of virtue and the moral life (see, for example, Porter , 210–15; Deane‐Drummond ; Deane‐Drummond et al. 2016).

Given the lack of evidence through historical writing, tracing the origins of empathy from the ancestral record requires some imaginative interpretation, and it may be one reason why so few evolutionary anthropologists discuss this. Finding evidence for consolation behavior in rodents as well as great apes implies some basic neural pathways may possibly be conserved (Burkett et al. ). Anthropologists, however, are rather more prepared to talk about the distinctive human capacity for symbolic thought, commonly confining such capacity to anatomically modern humans (Deacon ; Tattersall ). The thesis that symbolic thinking defined the emergence of cognitively modern Homo sapiens has begun to be challenged by cumulative evidence of complex artifacts existing in association with much earlier hominin remains, including the use of ochre, stone tools, blades, and so on. These glimpses of some sort of symbolic capability in the material record imply a far deeper history, even if their fullest development came much later (Kissel and Fuentes ). The challenge, however, of adjudicating what might be taking place in the mental processes of our earliest ancestors from material remains is similar to the challenge of tracking empathy and compassion: in both cases, the evidence is indirect. Nonetheless, it is possible to derive a probabilistic account based on the most plausible scenario from a preponderance of available evidence. Such accounts are still controversial, due to very small sample sizes and difficulties interpreting the evidence, but they are worth careful consideration nonetheless. British evolutionary anthropologist Penny Spikins and her colleagues have done some fascinating work on the possible reconstruction of psychological emotions in the prehistory of the Homo lineage (Spikins, Rutherford, and Needham ). The most common attitude among anthropologists, Spikins claims, is to ignore all emotions in human prehistory on the basis that they are far too hard to detect. What kind of evidence could possibly point to changes in the mental lives of these very early humans? Spikins recognizes the sociomoral role of empathy, shame, and remorse, but she believes that long‐term compassion is the most distinctly human characteristic marking humans out from other social animals; compassion was also highly significant in the earliest evolutionary history of highly cooperative human societies.

Thus, Spikins explains that

understanding the evolution and role of compassion in past human species entails recognizing that compassion is more than just a feeling that we recognize as personal, but also in a wider analytical perspective, it is a biological response, a ‘motivation to act’ whose roots lie in the hormonal and neuronal working of our mind. (Spikins et al. , 305)

Spikins argues that compassion involves an initial step of empathy and then a strong motivation to help the other in distress. While recognizing that compassion in other social animals is possible, including our nearest primate relatives, drawing on the work of de Waal and others, her evidence for expressions of compassion in these very earliest human societies indicates that it is far more sustained and long‐term, as in other primates.

Spikins cites a number of examples, which are intriguing to consider more carefully:

The most well‐known early example of long‐term support for an incapacitated individual comes from KNM‐ER 1808, a female Homo ergaster dated to around 1.5 mya.… Examinations of the skeletal remains of this individual have led to suggestions that she was suffering from hypervitaminosis A, a disease caused by excessive intake of vitamin A. (Spikins et al. , 309)

These symptoms can be tracked in human remains through reduction in bone density and the development of coarse bone growths. The symptoms for sufferers are known from current medical studies to include “abdominal pain, nausea, headaches, dizziness, blurred vision, lethargy, loss of muscular coordination, and impaired consciousness” (Spikins et al. , 309). This pathology would have taken many months to develop, which shows that caretaking in this case must have been long‐term as the individual could not have survived on her own without the intensive care of others. The point is that this requires long‐term and sustained care of a type that has not yet been found in primates not in the Homo lineage.

This kind of evidence is not direct evidence for sustained compassion; it is inferred, but it seems reasonable to make that inference. For example, an alternative explanation might be that long‐term care of such debilitated individuals was somehow forced through the hierarchical status of that individual in that society, rather than involving genuine caring emotions. But such a suggestion goes completely against what is known about the social structure of early human communities and their growing capacity to cooperate with one another (see Novak ).

A second example comes from even further back in history—1.77 million years ago, from the well‐known Dmanisi archaeological site in Georgia. As Spikins explains, “One of the Dmanisi hominins had lost all but one tooth several years before death, with all the sockets except for the canine teeth having been reabsorbed. Since it could only have consumed soft plant or animal foods, it seems likely that it would have needed support from others” (Spikins et al. , 309).

A third example is relatively recent, and it is a Neanderthal lineage concurrent with the Homo sapiens lineage:

Shanidar 1, the “Old Man of Shanidar,” dating to around 60–80,000 BP is perhaps one of the best‐known examples. This individual suffered multiple fractures across his body, with the right side being particularly badly affected, the right arm being described as completely “withered” (Klein , 333). The individual also received a “crushing” injury to his cranium, possibly causing blindness in his left eye due to the deformity of the skull. (Spikins et al. , 309)

Yet, a close study of the bones revealed that the injuries happened during adolescence, with death at a relatively advanced age (for a Neanderthal) of 35–50 years. Further, evidence for a young child in the Middle Pleistocene era aged 5–8 who suffered severe birth defects of the Cranium (craniosynotosisi) showed that, in this case at least, compassion also extended to babies.

Spikins compares this evidence, revealing what she believes to be a deep commitment to care, with cemetery evidence for the abandonment of babies suffering from exactly the same conditions in modern human societies. There are other examples of early upper Paleolithic individuals suffering from conditions such as acromesomelic dysplasia that leads to severe disabilities. Spikins believes that compassion, which finds expression in the human ability to extend care and commitment in a sustained sense to others, can include commitment to animals, or even objects and even ideas. While she does not mention religion, it is an intriguing possibility that the human ability to exercise compassion provided the psychological machinery for religious belief. This might lead to a clash between commitments, or even the possibility of exploitation. Spikins also speculates that there were four stages in the evolution of compassion: the first stage fleeting and still found in other primates; the second stage showing sustained investment in others; the third stage revealing more deliberative, committed, widespread, and long‐term caring; and finally, in the fourth stage—from 120,000 years ago in Africa and 40,000 years ago in Europe—compassion starting to be extended in a more abstract way to objects and more remote others.

Her evolutionary model fits an older though commonly accepted evolutionary account of human evolution in more popular literature that poses a linear developmental trajectory, focused on an out‐of‐Africa origin model with a later European expansion. This model is now becoming outdated and challenged by alternative theories that suggest a much more complex geographical origin and timeline. There is also growing cumulative evidence of much earlier symbolic capacities that appear and then disappear at a range of geographical sites (Kissel and Fuentes ; 2017). It is possible therefore to interpret Spikins's results rather differently. If compassion, like symbolic thought, is part of a slow, more sporadic process that includes flickerings of compassion rather than a single process, as has been described for complex cognition (Marean ), then it might be possible to track compassion alongside symbolic thought in general. In fact, if symbolic thinking and compassion are interrelated, then each would influence the other, since compassion for another encourages the accumulation of artifacts that act as proxies for the other and also express symbolic thought that, in turn, fosters deeper, longer term compassion, and so on. This would further be consistent with the thesis that the evolutionary lineage Homo was a slowly evolving community niche. This complex dynamic system included cognitive, social, and ecological components interacting with each other and with the genotype in a complex feedback system (Fuentes ).

Spikins and her colleagues concur with de Waal that a study of the great apes provides a reasonable basis for a direct consideration of the evolution of compassion in human beings. Arguments for evolutionary homologous neurological empathetic pathways in humans, primates, and other social animals based on consolation behavior (Burkett et al. ) remain somewhat speculative and the precise relationships between empathy and compassion are unclear. For example, when comparing Nussbaum's account of compassion with that of Spikins, it seems that Spikin's understanding is less well defined, even if it provides a “placeholder,” to use biological terminology. It is not clear, for example, at what stage the different judgments of compassion that Nussbaum identifies surfaced in human communities. For example, when did the judgment of compassion in terms of how serious an event might be, or its nondesert, or the idea that your flourishing matters to me arise? How might any of these judgments relate to evolutionary or “fitness” requirements of compassion? It is likely that such subtler elements within compassion in archaic communities will be extremely difficult, if not impossible, to discern, which means that from a philosophical perspective there will always be a gap between the different accounts.

The compassion shown toward some severely incapacitated babies is also interesting, given the range of attitudes shown toward babies with disabilities in hunter‐gatherer societies (Hrdy ). As Sarah Hrdy notes, a strong commitment to babies is common in wild apes and “no matter how deformed, scrawny, odd, or burdensome, there is no baby that a wild ape will not keep. Babies born blind, limbless, or afflicted with cerebral palsy—newborns that a hunter gatherer mother would likely abandon at birth—are picked up and held close” (2009, 70). According to Hrdy, this contrasts with hunter‐gatherer societies where there is a more complex process of discrimination, so “a newborn perceived as defective may be drowned, buried alive, or simply wrapped in leaves and left in the bush within hours of birth” (2009, 71). The differences among apes, other primates, and human hunter‐gatherer societies may not be as stark as Hrdy assumes here, given that across many primate species there are at least some occasions when primate mothers will also discard, ignore, or mistreat offspring (Campbell et al. ). What is not clear, however, is the relationship in the archaic mind(s) of the kind of attachment that Spikins suggests arises from a strong attachment to young, as in the majority of apes, to some other more cognitively aligned process that is more distinctly human. Hence, Spikins may have found evidence for strong empathy toward newborns in very early hominin societies, but how far this is a distinctly human form of compassion is much harder to judge. The fact that this was extended over many years does imply that this was distinctly human behavior, though the reasons for that behavior are also rather harder to adjudicate. The size of the samples used makes this research more speculative than evidence that has more robust statistical evidence. It is comparable, perhaps, to the more anecdotal evidence for empathy in primate groups that was collected by ethologists prior to more rigorous scientific studies. In addition, Spikins's research should not leave the romantic impression that such societies were necessarily virtuous in each and every respect, a distant memory of an idyllic time when human societies lived and worked together in peace. Alongside the evidence for forms of human compassion that Spikins has elaborated, there is also good evidence for its opposite, namely, the ability to be violent toward each other that is expressed eventually in peculiarly human forms of cruelty (Fuentes ).

COMPASSION AND WISDOM

The right judgment of compassion and, in this sense, the channeling of empathy toward moral rather than immoral ends, requires the ability to navigate complex relationships successfully. Thomas Aquinas is well aware of this in his treatment of the moral virtues interpreted as the correct alignment of concupiscible and irascible appetitive powers to that of reason (Aquinas, 23, 1a2ae, Qu. 58.1). The intellectual virtue of practical wisdom, or prudence, connects with all moral virtues. Yet, this does not mean that practical wisdom could replace affective capacities such as empathy as requisite for the moral life. When Bloom (), for example, argues that it is more compassionate to use a utilitarian calculus in order to conclude that preventing a hundred deaths is morally better than preventing one, he is not, it seems to me, really talking about compassion at all, but rather a particular form of social justice that rejects compassion as relevant for the moral life, namely, that we need to give each person his or her due.

For Aquinas, there is always the possibility of misalignment between the appetitive powers and reason, but that does not mean they are split apart. There are passions more properly associated with the will, though bearing a relation to intellect, which amount to pseudopassions (King ). Both passions and pseudopassions (also known as affectiones) have a role in human flourishing. There is therefore always a bodily element in a Thomistic understanding of the passions that is important in distinguishing passions as such and the rather more detached affectiones for those acts that were not associated with the sensitive appetite (Miner , 35). Problems arise when passions are confused with affectiones. The way such appetitive powers are regulated by reason has some analogies with Nussbaum's insistence that emotions also have cognitive content. Aquinas also keeps the virtues together, rather than apart, hence justice as virtue existed alongside the capacity to show compassion, and the theological virtues of hope and charity were morally significant virtues along with justice and gratitude. The theological virtue of faith is allied with the intellectual virtues of speculative reason that include science, understanding, and wisdom, as well as the intellectual virtues of practical reason, namely, practical wisdom and art. But the virtues of the will, such as justice, work together with the virtues of the intellect, such as practical wisdom. Just as empathy is in general a prerequisite for the basic expression of compassion, so practical wisdom is needed for more deliberative forms of compassion.

Nussbaum argues that the assertion that empathy is necessary for compassion is “dogmatic,” and names some examples, such as compassion for animals, where it is difficult for us to imagine what the other might be feeling (2001, 330). She also rejects the idea that the imagination entering empathetically into the other's mental state is always required to perform acts of compassion. But while there may be occasional exceptions, it seems to me that, in general at least, empathy and compassion are associated, as they seem to be in evolutionary history. Further, in her attempt to demonstrate how emotions such as compassion have “reasonable” elements, she has neglected the more implicit aspects of compassion that would allow an empathetic human imagining of how an animal other than a human might be feeling. Aquinas found a way to link everything together through what he terms the powers of the soul and a stress on the human will, even if that term now seems quaint for those schooled in psychology (Blasi ).

Practical wisdom or prudence is also necessary in order for a moral virtue to be guided by right reason and in order to make a disposition habitual. So, “Right judgment in matters of prudence is included in the definition of moral virtue, not as part of its essence, but as something belonging by way of participation to all the moral virtues, insomuch as all fall under the direction of prudence” (23, 1a2ae Qu. 58.2, ad. 4). Rather than opposing natural inclinations to reason, Thomas argued that reason provided a training ground for them, so “the natural inclination to the good of virtue is a kind of beginning of virtue, but is not perfect virtue…it also needs to be joined with right reason, as Aristotle declares” (23, 1a2ae, Qu. 58.4). Reasoning power can, of course, go awry and be used for evil as well as good ends.

Given the enormous capacity of humans for deliberative cruelty, of the most vicious and premeditated sort, Nussbaum () claims that social animals, lacking deliberative judgment, are at times at a moral advantage. The particular distortion in human abilities to show compassion arises, she suggests, by the judgment of fault rather than, as in compassion, that of nonfault. She does not go as far as George Pitcher, for example, who contends that the judgment of fault is always a defect and makes animals—in his case domesticated dogs—better off morally than humans. Yet, Nussbaum admits that the judgment of fault is often incorrect, such as blaming illnesses on guilt for sinful acts. Further, very rudimentary judgments of fault may be present in, for example, sophisticated social animals such as canids, where individuals who fail to conform to societal norms are excluded.

TRANSCENDENT COMPASSION

This article would not be complete, however, without at least mentioning the relationship between transcendent compassion and the possibility of infused virtues, including that of compassion, a claim that naturally endowed human capabilities to love others and make judgments of compassion is capable of being exceeded through special acts of divine grace. This makes both the evolution of and a neurobiological study of compassion still yet more complicated, since it now is coupled with a religious sense. In the first place, it is appropriate to point out that empathy and compassion are not used in the Summa Theologiae; rather, Thomas uses the terms charity and misericordia, a term that when translated into English means variously compassion, pity, and mercy. Where misericordia denotes being so affected that it leads to action on behalf of the other it can refer to compassion, whereas when misericordia means removal of another's pain through forgiveness of harm, it is best translated as mercy as that given by the one who forgives. Robert Miner, in a helpful article that considers the place of misericordia in Thomistic thought, focusing on the questions in the Summa Theologiae on charity, interprets compassion as “the affect that Thomas regards as a necessary but not sufficient condition for mercy as a human virtue” (Miner , 71 n.1). In other words, mercy is the developed virtue, even if compassion is its precondition. The relation between empathy and compassion is analogous to that between compassion and mercy, in that mercy cannot exist without compassion, but compassion alone is insufficient for misercordia as virtue. Thomas's placing of his question on mercy in the middle of his section on the acts and effects of charity that itself is placed in the central section on charity, leads Miner to conclude that “mercy lies at the very center of Thomas’ treatment of charity” (Miner , 73). Although translations of misercordia often use the word “pity,” it does not carry the connotations of condescension that Nussbaum alludes to in her discussion; given that pity as understood by Thomas also includes action on behalf of another, it can be translated as compassion in the present context.

Yet, it is worth dwelling on the specific way in which Aquinas refers to misericordia in relation to God's acts, since this also gives some indication of its overall importance. In speaking of divine omnipotence he claims: “Then again, as we have seen, the carrying out of divine mercy is at the root of all God's works; we are entitled to nothing except on the basis of what has come from God in the first place as sheer gift” (5, 1a, Qu. 25.3). It seems to me that there is no reason why misericordia cannot have connotations of both compassion and mercy here, given that it is at the root of God's acts, and given that misericordia is also elemental to God's love, for amor is the sole reason for God's miseretur (34, 2a2ae, Qu. 30.2). This becomes even clearer in the passage where Thomas claims,

Above all misericordia is to be attributed to God, nevertheless in its effect, not in the effect of feeling. By way of explanation, we note that a person is called misericors because he has a heart with misery, and is affected with sadness for another's plight as though it were his own. He identifies himself with the other, and springs to the rescue; this is the effect of misericordia. (5, 1a, Qu. 21.3)

Aquinas then explains that to feel sad about another's plight is not a divine attribute, but rather that God is capable of driving out every kind of defect. The movement from identification with another to springing into action is exactly the trajectory expected for empathy and compassion to work together. In other words, the narrower meaning of mercy understood simply as forgiveness of sins does not seem to apply in this context. Further, as Peterson () has pointed out, the empirical work of contemporary research in psychology (Klimecki et al. ) is aligned with the idea that empathy can be positive and not just negative in its emotional state.

In the second book of the Summa, Thomas cites Augustine when identifying the close identification between mercy and compassion: “since mercy is compassion for another's wretchedness, mercy is properly shown to another and not to oneself” (34, 2a2ae, Qu. 30.1). He then goes on to discuss the play on words between the Latin for misericordia and miseriae. Mercy also includes one of the judgments that Nussbaum names for compassion, namely, that of nondesert. So mercy is “strictly speaking…compassion for the misery of another” (34, 2a2ae, Qu. 30.1).

It is through the movement of misericordia beyond sensing the pain of the other toward action regulated by right reason that shows misericordia is a moral virtue rather than simply an emotion (34, 2a2ae, Qu. 30.3). There is no doubt in Aquinas's mind that, even though charity is that virtue that unites humanity to God, the greatest virtue with respect to our neighbor is that of mercy, so “of all the virtues which have to do with our neighbor, however, misericordia is the greatest, even as its acts surpasses all others, for to relieve the wants of another is, as such, the function of someone higher and better” (34, 2a2ae, Qu. 30.4). While the last phrase implies superiority for the one showing the compassion, this should not imply pride or condescension in that giving, but rather an indication that the person is closer to God; showing misericordia “is a sacrifice more acceptable to God” (34, 2a2ae, Qu. 30.4). He goes further by saying that misericordia “sums up the Christian religion as to outward activities,” but at the same time, the inward affection of charity, which unites the believer to God, is “something which outweighs both love and mercy for our neighbors” (34, 2a2ae, Qu. 30.4).

Thomas's explicit discussion of virtues infused by divine grace has been the topic of much scholarly debate with respect to their relationship with acquired virtues (see, for example, Mattison ; Pinsent ). Evidence for little distinction between infused and acquired virtues comes from texts such as the following: “The actions which are exercises of an infused disposition do not cause any new disposition, but merely strengthen the disposition which already exists” (22, 1a2ae Qu. 51.4). Even in this case, however, infused dispositions aim at a goal that he describes as “superhuman” and which “exceed the power of human nature” (22, 1a2ae, Qu. 51.4). Hence, infused virtues, even when the disposition is merely strengthened relative to acquired virtues, are still ordered to different purposes or ends (23, 1a2ae, Qu. 63.4). Other texts, however, imply that there is a more significant break between the infused and acquired virtues. For example, the infused theological virtues of faith, hope, and charity are always bestowed on the believer in a way that is “parallel to… natural principles,” which implies separation from acquired virtues (23, 1a2ae, Qu. 63.3). In the discussion that follows this Aquinas suggests that to be prepared for our supernatural end there needs to be “additional springs of activity” (23, 1a2ae, Qu. 63.3), drawing on Augustine's rather than Aristotle's definition of virtue, namely, “God works in us without us” (23, 1a2ae, Qu. 63.4). The difference between acquired and infused virtues, however, is the manner of the way they operate, so that while acquired virtues arise through the proper exercise of human reason, infused virtues respond to divine “rule,” so that temperance that is inspired by God leads to fasting and self‐denial in a way that would not be the case for the acquired virtue, which would lead to moderation in eating, for example, in order to sustain good health. This means that “infused and divine temperance differ in kind; the same reasoning applies to the other virtues” (23, 1a2ae, Qu. 63.4).

It seems then that in those cases where an infused virtue leads to the kind of activity that is different either in its end or the good sought, then the acquired and infused virtue will be different, even if having the same name. Mercy as that which is at the heart of the theological virtue of charity is also “a gift of the Holy Spirit and a new infused virtue” (Overmyer , 365). Yet, mercy is also a secular virtue. The question would then be, how might misericordia be transformed if it were subject to divine command, rather than acquired through the proper exercise of reason? In this case, we might expect a superabundance of the capacity to show compassion and mercy. Just as prudence guides the acquired virtues, so an elevated and infused prudence, quickened by faith and charity, guides infused moral virtues. Hence, the ability to show compassion and mercy toward one's enemies, or those to whom one might find naturally repulsive as well as one's kin, would be through the workings of divine grace in the manner of an infused virtue. Such grace may or may not be recognized as such by the one who is endowed with it.

The discussion so far then raises the intriguing possibility that religious belief, in this case belief in God, could foster forms of compassion and mercy that are over and above what might be expected otherwise. Research on the moral psychology of those who are considered highly experienced carers in L'Arche communities shows interesting aspects of the relationship between particular dispositions and the self‐identity of those involved (Reimer et al. ). The results imply that those who are drawn to show a high degree of empathy/compassion to others with severe learning disabilities, by living among them as one of them rather than in detachment, also develop a self‐identity that scores high on the empathy‐compassionate scale as well as on the religious scale. How far and to what extent the virtues expressed in such communities are either acquired or transformed into infused virtues requires more research on the motivations of those involved, their faith commitment, and perhaps reported experiences of grace.

TENTATIVE CONCLUSIONS

I have argued so far that the most recent attempts by some psychologists to dismiss the value of empathy for the moral life have their ideological roots in ancient philosophical debates about the value of compassion. Drawing on the work of Martha Nussbaum, it is clear that the definitions of terms such as empathy, compassion, pity, mercy, and sympathy are often blurred or muddled in the literature. I also argue that her own definition of compassion that focuses on the particular cognitive judgments required in compassionate acts fails to consider adequately the priority of a relationship between love and compassion, namely, that part of compassion that is actively engaged, rather than just in response to another's distress. While empathy strictly speaking means the ability to feel what the other feels, and therefore is potentially either passive or even negative in the moral sphere, compassion is more positive and aside from some rare exceptions requires the capacity of empathy for its operation.

Compassion is also a distinctly human ability and there is reasonable evidential basis for something like compassion being operative in at least some of the earliest Homo lineages. Although archeological evidence is always somewhat speculative, the ability of Homo neanderthalis to show long‐term compassion toward the severely disabled in a way that later modern humans clearly did not raises intriguing questions about the extent to which compassion was enhanced in these very early human communities. Does the story of Eden point to a distant memory of intense cooperation and compassion? Inasmuch as compassion existed alongside violence, the case for any sequential movement from peaceful to violent beginnings seems reasonably unlikely, although the specific instances of compassion and its apparent expression prior to the Homo sapiens lineage is remarkable, as is the evidence for the flickerings of other cognitive abilities in humans long before anatomically modern Homo sapiens. So far the evidence is not sufficient to make any definitive claims, but theologians are bound to be particularly interested in the way this evolutionary anthropology unfolds.

An analysis of compassion and its limits is also integral to the classic Christian tradition, insofar as misericordia is one of the moral virtues central to the life of charity. Although there have been heated debates about the place of misericordia in Thomas Aquinas's moral theology, it is capable of being infused as well as acquired. Given that infused virtues point to capacities beyond what could be thought of as natural endowments, this raises the interesting question of the relationship between religion and compassion. Perhaps the dawn of religious belief, which itself was somewhat dependent on a prior capacity for compassion, acts like a prism through which the virtue of compassion comes to be both highly regarded and more intense, as well as being capable of extension to others outside one's immediate kinship group, and perhaps even other species.

ACKNOWLEDGMENTS

This is an expanded and revised version of a paper, “Morality's Deep History: Empathy and the Evolution of Compassion and Wisdom,” presented as part of the International Society for Science and Religion (ISSR) panel session “Against Empathy? Theological Reflection on Neuroscientific Opposition to the Existence of Mirror Neurons and the Moral Significance of Empathy,” Atlanta, GA, November 20, 2015. I am grateful to Ron Cole‐Turner for the invitation to present and for comments from the audience and comments on the working draft from my fellow panelists Michael Spezio, Gregory Peterson, and Charlene Burns. I am also very grateful for helpful comments on the final draft from William Mattison and my colleague Agustín Fuentes, as well as editorial support from Craig Iffland, my research assistant at Notre Dame, and Rebecca Artinian‐Kaiser. This research was made possible in part through support from the following grants: Celia Deane‐Drummond and Agustín Fuentes, Human Distinctiveness, Grant number 57496, funded by the John Templeton Foundation; Celia Deane‐Drummond, Darcia Narvaez, and Tom Stapleford, Developing Virtues in the Practice of Science, Grant number 0088, funded by the Templeton Religion Trust.