The questions concerning both what we need to be motivated to act morally and what is required to carry out moral actions are certainly extremely difficult ones. To say that I will offer a unitary and complete theory of moral agency would be an intolerable act of hybris on my part. Intolerable, because, first, there are already enough theories of that kind in the long history of (moral) philosophy by philosophers much more skilled and experienced than myself and, second, because this effort would require more than one section of a book dedicated primarily to empathy. Instead, what I will attempt to offer is the following:

  1. (1)

    A (hopefully) convincing answer to the criticisms made by anti-empathists about the role played by empathy in moral agency and motivation.

  2. (2)

    A rough outline of a theory of moral agency in which empathy plays a crucial role, which although undoubtedly close to some sentimentalist positions, is actually more complex than those and can accommodate typical concepts of rationalism as well as virtue and caring ethics. This will be accomplished based on a reinterpretation of the Kantian theory of moral agency and especially of Kant’s doctrine of the categorical imperative and on the outline of (at least to the best of my knowledge) an original form of virtue ethics in which empathy plays a fundamental part.

Let us begin with the first point. We have seen that anti-empathists, like Paul Bloom and Jesse Prinz, see the use of empathy within the field of moral agency and motivation as objectionable, to say the least. Prinz, in particular, believes that empathy is both useless and ineffective for moral conduct. It is useless, because other emotions are very much apt to motivate us to act morally; and it is ineffective, because these other emotions are also much more efficient than empathy when it comes to instantiating moral behaviour. Remember, in fact, that for Prinz, moral judgements are all based on emotions such as shame, guilt, anger, and disgust, which are by their essence motivational states with a precise behavioural profile. Notably, they are negative emotions, which are elements that we typically work to avoid in our lives. Therefore, if we anticipate that the instantiation of the behaviour X will make us feel ashamed, or sad or something similar, then we will feel a strong inclination to avoid X.

Personally, I do not have anything against this view. I also think that moral conduct can be motivated by emotions which must not be empathic. To deny this would mean to ignore what is simply a matter of fact. Nonetheless, the point is that Prinz needs more than this to deem empathy as useless in this context. In fact, there is nothing which rules out the possibility that empathy can be a strong motivator for moral behaviour, as other (moral) emotions are. In fact, there is a certain triviality in asserting that an action (be it moral or not) can be heteroelicited, that is, caused, activated by different stimuli. Much more interesting would be to test whether empathy always and necessarily is at the base of our moral motivation and, along with it, our moral conduct. The simple answer is that it is not. Pace committed pro-empathists, such as Michael Slote or Simon Baron Cohen, anti-empathists, including Paul Bloom and Jesse Prinz, are right in affirming that moral conduct does not always rest on empathic grounds. Instead, where they fail is in their argumentative choices. Prinz chooses, in fact, to criticise empathy by using a few experiments which, while they undoubtedly serve to prove some inadequacies on the part of empathy, equally, they cannot be taken as an overwhelming evidence of the general inadequacy of this phenomenon. Ironically, I find Bloom’s attempt to fuse a rather utilitarian approach to morality with his conception of a ‘rational compassion’ as more philosophically intriguing, but on the whole, he also proves to be blind to certain advantages of empathy for morality. For instance, similarly to Prinz (to whom he is much indebted), he asserts in his book Against Empathy that, although we need what he calls a ‘motivational push’ to be a good person and acting morally, this push does not need to rest on empathy: other emotions are apt to suffice.Footnote 1 The problem here is that Bloom seems to ignore the fact that these emotions can also be felt as a result of empathy with other people.

1 How the Shortcomings of Empathy Are Also Its Strengths

In effect, what kind of strategy could be implemented in order to respond appropriately to the criticisms of the anti-empathists and to better strengthen my position? Of course, I could employ a similar technique and use cherry-picked experiments and data that confute the results of the ones cited by Bloom and Prinz. By way of example, the vast amount of research undertaken by Daniel Batson on empathy may be fruitfully employed to show that empathy has considerable potential when it comes to moral (altruistic) motivation. However, in doing so, my answer would be a mere commentary of the theses of other scholars.

A more promising line of reasoning might instead be that of identifying the contradictions in the anti-empathic claims and, from those contradictions, building up an argument in favour of a certain role that empathy can play. Bloom, for instance, has plenty of such contradictions. Consider, for example, pages 87 to 89 of his above-mentioned book. Here, he attempts to make a case against the spotlight nature of empathy. For Bloom, this characteristic of empathy makes its focus narrow and its outcomes unreliable, because spotlights illuminate only a portion of the space and leave the rest in darkness, thus making empathy a biased process. To substantiate this claim, Paul Bloom chooses to cite a few interesting experiments.Footnote 2 In one of them, subjects were given ten dollars and then told that they could give as much as they wanted to another person who had nothing. This person was anonymous and identified by a number that the subjects of the experiments drew at random. However, a number of the subjects drew the number first and then decided how much to give, whereas another group decided first how much to give and then drew the number. Interestingly, this prima facie small and apparently insignificant difference seemed to play a surprising role. In fact, people who drew the number first gave much more (60% more) than those who decided first. The researchers hypothesised that this difference had to be attributed to the fact that the focus on the number (and not on the money) helped the subjects of the experiment to imagine a specific needy person, as opposed to just an abstract individual.

The same kind of mechanism was observed in another study, in which people were asked to donate money to help build a home for a family. One group of them was told that the family has already been selected, whereas another one was told that the family would be selected. Again, this subtle difference resulted in a conspicuous discrepancy: the first group actually gave significantly more money than the second one, and here it also must be presumed that the discrepancy has to be ascribed to the shift between a concrete and a more abstract target. Bloom claims that this spotlight nature presupposes empathy to be ‘innumerate’ and to prefer the individual at the expense of the multitude, as can be noticed in other experiments comparing how we usually respond to the suffering of the one versus the suffering of the many. Bloom uses, in particular, a study by Tehila Kogut and Ilana Ritov,Footnote 3 where psychologists asked a number of subjects how much money they would be ready to give to financially support a drug that would save the life of one child, and another group how much they would give to save eight children. These two groups gave approximately the same. However, there was also a third group of people that were told the child’s name and even shown their picture. In this case, the donations rose dramatically and exceeded even the donations for the eight. For Bloom, this is exactly one of the shortcomings of empathy we should be wary of. After all, for his consequentialist mentality to prefer the one to the many is a serious mistake, and the problem with empathy (and with sentiments in general) is that they are innumerate. Bloom puts it provocatively with a mental experiment of Humean taste: ‘[…] imagine reading that two hundred people just died in an earthquake in a remote country. How do you feel? Now imagine that you just discovered that the actual number of deaths was two thousand. Do you feel ten times worse? Do you feel any worse?’

The argument is simple, but thought-provoking. We are generally sensitive to differences in number, and if we learn that an accident ‘A’ (or a natural cataclysm) has caused the deaths of ten times more people than accident ‘B’, we are ready to deem A to be a far more serious accident than B. The fact is that sentiments do not follow this intuitive rule and, for this reason, they should be denied a role in moral judgement, education, and conduct. I confess that in my past, as an anti-empathist, I was very sensitive to these kinds of arguments: they were so convincing in their clarity and simplicity that it took me some time to understand that under this apparent cover of good sense and rationality, other biases were hidden. Here, the bias expressed by Bloom is a rationalist one. It is one thing to affirm that morality cannot be exclusively based on sentiment; it is another to assert that morality can be based on reason alone. As we have seen, my model contemplates that empathy and other sentiments play a role together with rational considerations. The true mistake is to conceive the rational and the sentimental sphere as two separate and opposed dimensions of which one has to be preferred over the other. On the contrary, if we depart from this mechanistic perspective, we will discover what feelings can do for our rational considerations, if they are well balanced in the character of a morally virtuous person. Unwillingly and indirectly, Bloom offers us exactly the opportunity to argue in favour of this position when he quotes Mother Teresa to show how biased and innumerate our emotions are: ‘If I look at the mass’—so Mother Teresa affirms—‘I will never act. If I look at the one, I will.’Footnote 4

Reading this citation, it is odd to believe that Mother Teresa had no idea of the sheer number of people needing help in Calcutta. It is absurd to think of her falling prey to sentimental biases as the ‘spotlight’ and the ‘innumerate’ ones and helping the one individual to the detriment of the multitude. In light of this, how should this quote be interpreted? I think that what Mother Teresa wanted to say here goes beyond the dichotomy between reason and sentiment(s), and, in so doing, she displays the same concern I have. As I read it, Mother Teresa wanted to emphasise the centrality of day-to-day progress, of taking one step at a time. In other words, if you look at the totality of needy people there are (no matter if in Calcutta, in India, or in the world) you will find it difficult to feel any motivation to help. Most likely, you will be overwhelmed by the thought that you are just one person with scarce and inadequate resources and that the people you would like to help are too many to even be counted. Therefore, the best you can do in such a situation is to avoid looking at the mass and focus only on the hic et nunc, on the person(s) you can help at a given moment, in a given place. The same applies to mountain-climbing or trekking: if you keep gazing at the top of the mountain, you will hardly reach it as it will always seem too high and too far away. On the contrary, if you only concentrate on where to put your foot for the next step, then, little by little, inch by inch, you will reach the top almost without realising it.

However, this is not the only message that this quote suggests. Indirectly, in fact, this citation also reminds us that any element capable of focussing our attention on the hic et nunc, without making us lose sight of the general situation and of our goal, is useful for the purposes of moral conduct. In other words, if we agree on the fact that it is easier to help others and instantiate acts of altruism and solidarity if we focus on the next step and not on the mountain, then everything helping us in that sense will, at least, have an instrumental necessity. Once we acknowledge this, reading the opinions of Paul Bloom again can be instructive, in fact, one becomes aware of the implicit paradoxes contained there. It is true: our emotional capacities are not able to conceive of the death or the suffering of thousands (or even more) people adequately and, in that sense, our emotivity (as well as our empathy) certainly is innumerate. Nevertheless, the cure to this potential shortcoming of empathy is contained in the shortcoming itself, and it is strange to observe that Bloom seems to ignore this. The secret is personalising the victims, giving them a face, an identity, an individuality. In this way, the aspect observed and showed in the above-mentioned studies ceases to be a weakness and becomes a strength. Apparently, even Paul Bloom seems to contradict himself when talking about this feature. On page 45 of his book Against Empathy,Footnote 5 he in fact wrote that as a young graduate student, he was deeply influenced by the rational, consequentialist arguments of Peter Singer. Thanks to the Australian philosopher, he convinced himself that spending money on luxury products, expensive meals, or fancy clothing was really no different than seeing a child drowning in shallow water and doing nothing for fear of ruining his own expensive shoes. The argument was extremely rational and utilitarian and Bloom loved to repeat that in front of friends, colleagues, and students alike, until the day when one of his students simply asked him how much of his money he actually gave to poor and needy people. Bloom, embarrassed, had to tell the truth: nothing. Thus, the rational argument that managed to persuade him to believe that enacting behaviour X was morally right was apparently not enough to motivate him sufficiently to instantiate that behaviour. He was, that is, rationally convinced, but not emotionally moved and not adequately motivated. Aware of this failing, Bloom decided that it was time to give to a charity. Accordingly, he asked an international aid agency information as to how he could support their cause. He expected to find graphs, statistics, and data about their operations, but, instead, received a small photograph of a young boy from Indonesia and a short letter in which they wrote that the child in the photo was the life he would save by contributing to the agency. As a result, what did our utilitarian champion do? Let us hear it in his own words:

I’m not sure if the feeling this prompted was empathy, but it was certainly a sentimental appeal, triggering my heart and not my head. And it worked: Many years later we were still sending money to that child’s family.Footnote 6

As you can see, even Paul Bloom, with all his vis polemica against empathy, had to admit that by giving a kind of ‘face’Footnote 7 to the target of our potential moral actions (such as acts of altruism or similar) the motivational trigger may increase. This should be taken as good news, even by who is maybe the staunchest critic of empathy as, in fact, it means that empathy is able to motivate appropriately, it just needs to work with particulars. However, this is exactly Bloom’s problem with empathy. Remember that his main attacks on empathy—back to which all other criticisms can be traced—regard two characteristics: empathy’s narrowness and its biased/parochial nature. The big problem for Bloom is not that empathy is unable to motivate, but that the way in which it motivates is unsuitable for morality. Empathy, in fact, can only motivate to help ‘that precise person’ at that given moment, but it is useless (if not deleterious) when it involves making decisions in order to help ‘the most’ effective way possible, which, for Bloom, means nothing else than to help the greatest quantity of people in the most efficient way. Why is that? Paul Bloom believes the answer to this question to be fairly obvious. On the one side, he thinks, exactly like Jesse Prinz, that empathy is incapable of widening its scope in order to embrace humanity itself and not just that precise human being one is observing. On the other, he combines this criticism with the idea that empathy is inherently partial, that is, the fact that we are more inclined to feel empathy for people we know and love, than for those we do not know or consider hateful or just different from us. Showing the same taste for hyperbolic metaphors as Prinz, Bloom boldly affirms: ‘Asking people to feel as much empathy for an enemy as for their own child is like asking them to feel as much hunger for a dog turd as for an apple – it’s logically possible, but it doesn’t reflect the normal functioning of the human mind.’Footnote 8

Now, this assertion by Bloom may seem prima facie quite undeniable: how could one person possibly feel as much empathy for their enemy as for their own child? However, this is a strawman: even the most convinced pro-empathists would not go so far as to affirm that we can feel the same amount of empathy for any individual. Thus, it is perfectly normal to feel more empathy for, say, one’s own family members than for enemies. It is so normal that this can, in fact, be taken without question as the reason why we find it easier to help our family as opposed to people we do not like.Footnote 9 Nonetheless, this cannot constitute, per se, a criticism of the motivational role of empathy: at best, it only serves to highlight the fact that the strength of empathic moral motivation is directly proportional to the strength of the empathy we feel. This is a much more modest assertion that few would be ready to deny, but that also does not seem to be the argument Bloom needed to stab empathy at the heart.

However, perhaps there is another way to read this argument: the point is not that empathy is variously limited, and so we feel different amounts of it for different categories of subjects, but that it is also, under many different aspects, biased. In other words, Bloom wants to emphasise what he takes to be an impossibility for us as human beings: to empathise across the parochial boundaries that surround us and, in some cases, define us, such as, inter alia, national identities, political ideologies, family- or friendship-ties.

However, even granted this explanation, Bloom’s argument does not cease to appear as a strawman to me, and I claim that this is what it fundamentally is. In fact, it is one thing to say that empathy is intrinsically biased and quite another to state that we tend to have prejudices stemming from different sources (social, political, ideological, for example) and that empathy seems to prove unable to overcome them. Even though Bloom wishes to assert the former (as he often does in the course of his book), it appears that he merely offers a redundant series of (pseudo)evidence for the latter. In other words, all the limits that Bloom attributes to empathy are not limits integral to empathy as such, but to the social, cultural, political, economic, religious, and historical structures within which empathy (not differently from any other moral attitude) exerts its influence.

What is more, Bloom’s arguments are almost never based on conceptual grounds; they are, on the contrary, of an empirical, experimental nature. For example, you will often find in his work arguments that sound like the following: ‘You cannot empathize with more than one or two people at the same time. Try it.’,Footnote 10 all with the intention of undermining the credibility of the thesis that empathy is capable of overcoming the biases set by parochialism, narrow-mindedness, proximity, and others. Nevertheless, you will struggle to find even a single piece of evidence that it is conceptually and consequently inherently impossible for empathy to overcome prejudices. Yet, this is exactly what he needs to uphold his position. If the only arguments Bloom presents are of a practical, empirical nature, then one only needs concrete, applied examples of the contrary to confute his theses. The fact is, there are plenty of examples of that kind. Consider the supporters of Peace Now, for instance, or think of the story that more than any other—at least in the occidental, Christian world—has influenced the image we have of the good empathic person: that of the Good Samaritan. A person whose capacity for empathy and compassion managed to overcome differences in culture, religion, ethnicity, to effectively help an enemy. History itself—among many horrors—is full of bright examples in which one’s own empathy is the primary cause of the acknowledgement of a common humanity that ties the Self with the Other and that breaks down every barrier. Think, for instance, of Oskar Schindler, who, although a member of the Nationalist Socialist Party, saved 1200 Jews after the shock and the horror the brutal persecutions in Cracow caused him. Or consider the example of Giorgio Perlasca, a committed Italian fascist who fought as a volunteer in the Second Italo-Ethiopian War and also in the Spanish Civil War (siding with the general Francisco Franco), but who refused to give his support to the racial laws promulgated by Benito Mussolini. Perlasca hid his aversion for these laws which threatened to make Jews appear as inferiors and worked with the Spanish Chargé d’Affaires, Ángel Sanz Briz, saving the lives of thousands of Jews.Footnote 11 This was achieved, in particular, to the provision of safe-conducts (“protection-cards”), and, according to what he told,Footnote 12 to the thwarting of plans for the arson and extermination of the Budapest ghetto, which would have caused the death of around 60,000 Hungarian Jews.

2 Generalising Empathy

Stories like these are possible because human beings possess an extremely powerful capacity that can be efficaciously used in tandem with empathy, and that is the capacity of induction, generalising from the particular to the universal. Generalisation-plus-empathy leads to a we-perspective starting from a me- or a you-perspective. The first is a very well-known method that has been mentioned in the section on moral education and development. Basically, if I know, from experience or imagination, that action X, for me, would be unpleasant and unnecessarily cruel and that I would try to avoid it, then, empathy for individuals who are undergoing X may easily lead me to judge that X is wrong and give me reasons and enough motivation to try to stop X. The second method is similar: I see, for example, that someone is suffering and this consideration brings me to the awareness that I would suffer, too, were I to experience the same circumstances. This gives me a reason to consider, for the first time, that we are not so different and that if I wished to be helped in that situation, maybe that other person might want or need help, too; which of course gives me some motivation to engage in helping behaviour.

However, one might still contend—as Paul Bloom does—that this can be a rather common behavioural pattern in interpersonal relationships, but it can hardly become a central one with respect to making the right, moral choice in contexts where the lives at stake are so many, so far away, so anonymous, that they present themselves as data, as numbers in a newspaper article or graphics on the news. This is indeed true, but the question here is: why should it be that way? I claim that there are numerous examples which attest to the fact that the display of a certain, particular detail—for instance, a single person with whom one can empathise—is sufficient to elicit empathy, and this empathy, in turn, becomes the motivational spark for moral judgement and agency. Take into consideration the immense empathic power present in photos. Who can possibly erase from their memory, for example, the image of poor Alan Kurdi, the three-year-old child found drowned on a shore in Turkey in 2015? The photo, so harsh in its merciless, but direct realism, portrays the lifeless little body of the child, with his arms at his side and his face down on the sand. This heartrending picture was the key to reawakening the conscience of many politicians and decision-makers in Europe and beyond, to attracting the interest of the mass media, and to explaining, better than any graph or data analysis could possibly do, what really happened in the Mediterranean Sea every day as a result of the European Migrant Crisis. That touching image, the inescapable awareness that that poor child might have been our child, that he deserved better, that no one should suffer the same fate, in short, the empathy we felt for a child we did not know personally, and for his family, made us understand that we are not special, that migrants deserve to have the opportunity to build a future that we also would wish for ourselves. We can really say that there was a European policy on migrants before and after Alan Kurdi, so powerful was this photo.

Nevertheless, the power of empathy is not limited exclusively to the medium of photos or to a target composed of human beings. Its capacity to elicit thoughts that focus on the fact that basic emotions, such as joy and suffering, make all living and sensible beings similar is very well known by animal activists as well. Consider, for instance, various sentences that are commonly used in videos by animal rights or vegan associations: ‘Imagine being a chicken. The first day of your life on earth you are castrated and then forced to live your entire existence in a small place overcrowded by other chickens like you, while literally dragging yourself over your own excrement, waiting to be slaughtered’ or ‘Imagine being a pig and seeing the butcher stunning your mate and then killing them in front of your eyes, knowing that you would be the next.’ It is crystal-clear that this method is employed precisely in order to elicit empathic distress in us, in the hope that this distress will lead the viewer of the film or the reader of the article, to not only feel with animals, but to feel for them and, eventually, do something for them. Indeed, this actually works. When asked, most people who have become vegan answer that they did so out of respect, empathy, or love for animals.Footnote 13 Therefore, not only can empathy lead to compassion (which should not be surprising), but even when it leads to an experience of empathic distress, this distress can often be preliminary to altruistic behaviour. Personal distress does not necessarily bring to the instantiation of egoistic behaviour.

Paul Bloom does not seem to realise that what could prima facie seem a serious shortcoming of empathy—its being partial and limited, its ‘spotlight nature’—can in fact be turned into an incentive for moral conduct. Primo Levi rightly affirmed that there is no proportion ‘between the pity we feel and the extent of the pain by which the pity is aroused: a single Anne Frank excites more emotion than the myriads who suffered as she did but whose image has remained in the shadows’.Footnote 14 I agree with that statement, but why is the Diary of Anne Frank so important? Why do we read Anne Frank and Primo Levi? It is because they are archetypes. They have become paradigms, symbols of the discriminated, of the persecuted, of the unjustly oppressed. Moreover, as symbols, they stand for all the persecuted in any part of the world, at any time, they represent them. In this way, the empathy we feel for the young Anne Frank or for Primo Levi, who faced with a rare dignity his imprisonment in Auschwitz, despite all the humiliations and deprivations he had to suffer, becomes the cornerstone for, inter alia, a rejection of violent and intolerant ideologies and a resistance against discrimination, oppression, and persecution of any kind. Empathy tells us that Anne Frank and Primo Levi were not ontologically different from (or even worse than) us qua Jews; on the contrary, in spite of the Nazi propaganda, they shared our same humanity. They had the same desires and hopes we have, they loved and suffered as we also do. Yet, here is the deception: it was not a deviant, perverted use of empathy that led to the horrors of Nazism, but a distorted, corrupted use of reason. Philosophical and evolutionary-biological arguments were used to spread racism and to defend the Endlösung, the Final Solution. In other words, it was not reason that was perverted by the biases of empathy, but the natural empathy that people have that was corrupted by insidious and dangerous rational arguments. Thus (pace Bloom and Singer), reason cannot per se constitute a guarantee for morality. On the contrary, if we accept that the capacity to empathise with others is as much essential to our human nature as it is that of reasoning, we will find ways to coordinate these two with the aim of reaching moral excellence. Therefore, when Bloom affirms—rather trivially—that policies in general, as well as social, economic, political decisions, should be taken on the base of reason and utilitarian calculations and not on the base of empathy, we can simply respond that empathy must not be employed in the same amount at every stage of the process of deliberation. Hence, for instance, the example of Primo Levi might empathically motivate me to fight against racial prejudices, but then it would be my reason that would suggest the best way to do it, such as campaigning for or against a certain law or donating my money to a certain ONG.

Take into consideration another crucial point. When Paul Bloom asserts that empathy is innumerate, and for this reason it cannot be taken as a solid and reliable guide for moral agency, he seems to ignore that this is more a bias inherent to human beings per se, than to empathy. By this I mean that having to deal with large numbers always pose difficulties. For instance, learning that every year in the world about nine million people die from hunger does not only constitute a challenge for our empathy, but even also for our cognitive abilities, like our imagination. How can one picture such a vast number? It is literally overwhelming. Nowadays, the sheer quantity of information we can acquire by simply browsing on the Internet is astonishing and the truth is that our moral sense is not ready to deal with such large-scale numbers. Our morality, with all its baggage of rational principles and moral sentiments, such as a sense of guilt or compassion, evolved over thousands and thousands of years in small communities, namely, the family and the village. We are, therefore, shaped by evolution to deal with much smaller numbers. As we have seen, the solution proposed by Bloom is simply to stop relying on empathy: it is untrustworthy. My view on the matter, instead, cannot be the one suggesting the complete eradication of empathy (which, since empathy is a capacity with such a long evolutionary history, would simply be nonsensical and conceivably impossible) but, rather, I think that empathy can be of help in the sense I have described earlier: as a motivational incentive acting, inter alia, by means of the generalisation of paradigmatic cases.

Empathy, in other words, can be a very useful tool for morality. However, there is more to be considered. It has been mentioned previously how the criticisms that Bloom advance towards empathy are flawed, since he accuses empathy of biases that are inherent to us or to our culture and society, not to empathy per se. Narrow-mindedness, for instance, is a flaw of intelligence (possibly stemming from a bad education), not of empathy. As we have said, only the empathy of parochial people is parochial. The empathy of a person who is not racist or parochial embraces, instead, any person, independent of ethnicity, culture, religion, and other differences. Yet, this is not the only problematic criticism by Bloom. In fact, he insists various times in his book that empathy should be abandoned because focusing on ‘personal tragedies’ intensively covered by the media steals away resources which could be more fruitfully and productively employed in other more serious matters. For instance, earlier, in his 2013 article and before his famous book Against Empathy, he asserted that the 2005 Natalee Holloway case (which involved the disappearance and probable murder of a high-school graduate while on a trip to Aruba) used far more television time in the United States than the concurrent genocide in Darfur. In cases such as these, Bloom argues that empathy induces us to identify with the suffering of a single individual even one that we do not know but whom we can, thanks to the media, readily identify with and who seems more ‘like us’. This phenomenon inevitably leads to an immoral misallocation of priorities, as well as resources and attention that could and should be granted to other issues.

If I understand the argument correctly, the charges against empathy are twofold: on the one hand, Bloom accuses empathy of being easy to manipulate (mass media can exploit our empathic capacities and make us focus on the one as opposed to the many), and, on the other hand, he accuses empathy of being inherently biased towards the easy identifiable ‘singular tragedy’ than to the less ‘appealing’ everyday tragedies, such as the number of people starving everyday in the world. These are astute criticisms, but unfortunately for Bloom, I believe that they totally miss the target. In fact, it is not—or at least not directly—our empathy, but rather our attention that mass media manipulate. Of course, it is easier to empathise with a young girl than with people dying in Darfur if her story is everywhere to be found, if you can see and hear the interviews with her parents, if you learn about her background, and more, whilst the terrible troubles of a far foreign land are given little coverage. However, to say that empathy should be blamed for that situation is to mistake the consequence for the cause. In fact, by simply changing the focus, and by giving attention to the matters that deserve it the most, we would soon discover that the intensity of our empathy would also vary accordingly. Remember the famous old saying: ‘we are what we eat?’ In the same spirit we can say that ‘we empathise only with what we focus on’. This is also the reason why younger generations are on the whole much more open-minded and much less racist than their parents or grandparents. It is due to the fact that they are more used to contact with people of other cultures and ethnicities and because they might have had the opportunity to grow up in an educational system emphasising empathy and respect for people from, for example, an immigrant background.

Furthermore, I do not share Bloom’s belief regarding the fact that empathy is responsible for the misallocation of resources. This is an opinion of Singerian taste: basically, when you help someone under the incentive of empathy, you are, with all probability, guiltily ignoring the fact that you could help many other suffering individuals. Perhaps, this is the reason why Bloom labels empathy as ‘morally corrosive’.Footnote 15 Since empathy, for Bloom, always leads to doing less good, then we should stop making use of it. He goes so far as to defend a position which is not only counterintuitive, but also extremely fundamentalist, namely: ‘If you are struggling with a moral decision and find yourself trying to feel someone else’s pain or pleasure, you should stop.’ Thus, no matter how morally good your act might be, if you do it while trying to empathise with the target of your act, you are already doing wrong. It is easy to see what kind of peculiar consequences this principle entails: any single act made out of empathy would be deemed morally bad, or at least less good, because the individual carrying it out would have missed the opportunity to do better somewhere else and/or to someone else. There is a strange logic behind the thought that, say, paying for the lunch of a homeless person you meet on the street turns out to be bad or less good if, with the same money, you could have paid for the lunch of ten children in Burkina Faso. Even if we were ready to accept that logic, as Singer, Bloom, and other Utilitarians appear to do, we would still have to admit that calling an act less good is not the same as calling it bad or morally corrosive. Moreover, we should also be careful in ruling out the possibility that someone may make the best decision in utilitaristic terms (e.g. paying for lunch for ten people in Burkina Faso) precisely under the influence of empathy. In actual fact, this takes us to another question.

3 Is Empathy Necessarily Biased?

I think we have shown with enough clarity the partiality of Bloom in his treatment of empathy: many of the criticisms he makes cannot be directed towards empathy per se, and the examples he uses only show that empathy does not always lead to instantiating moral actions. However, this is merely contingent: there is a plethora of other cases in which empathy effectively acts as a wonderful motivator for moral conduct. Notice, also, that all the criticisms expressed by Prinz and Bloom have the same two presumptions at their core, the assumption that empathy is intrinsically biased and the assumption that empathy is a kind of fixed biological trait, not amenable to modification. Bloom, in particular, affirms that empathy is ‘bred in the bone’ and that occurs automatically and involuntarily.Footnote 16 Nevertheless, in a preceding article in 2013 he writes rather affectedly and dramatically: ‘But empathy will have to yield to reason if humanity is to have a future.’Footnote 17 The only way to interpret his theory, according to which (1) empathy is a fixed trait because it is bred in the bone, (2) it is uncontrollable because it occurs involuntarily, and (3) it nonetheless must yield to reason, is to think that although we cannot control our empathic reactions, we can control what we do with them. It goes without saying that—according to Bloom—what we should do with them is simply to silence them, so that they cannot interfere with our purely rational moral decision-making.

The notion of the fixity of the moral emotions (and of emotions in general) is a perspective Bloom and Prinz share, at various levels, with some important names in the history of philosophy, such as David Hume,Footnote 18 Immanuel Kant,Footnote 19 and Henry Sidgwick.Footnote 20 In what follows, I would like to oppose this view in two different ways: the first is to show that empathy is not the fixed trait that Jesse Prinz and Paul Bloom believed it to be. Notice, in fact, that their criticisms hold only if the negative impacts of empathy cannot be regulated while retaining the positive impacts, otherwise it would make little sense to be ‘against empathy’. If I can show that it is possible to modulate our empathic responses, then the anti-empathic position will be confuted. The second line of thought I would like to develop concerns an aspect of empathy, which is often downplayed, ignored, or even denied by anti-empathists, is its ethical valence. I am going to argue, indeed, that empathy is essential to our flourishing as ethical individuals and that it can even be considered (pace its critics) a moral virtue.

Let us start for the moment with the first argument: that about the supposed fixity of empathy. What does it mean that empathy is seen as ‘fixed’ by anti-empathists? It means, grosso modo, that, when activated, it constantly produces identical (or approximatively identical) effects in identical (or, again, approximately identical) situations. Which of course also means that empathy is not susceptible to voluntary behavioural modification, neither individually nor societally: if certain conditions are met, empathy will occur as a result, and its outcomes will always be predictable given the circumstances that triggered it. In order to test whether this line of thought is correct, there is no better way than to turn to the scientific literature on the matter. Does the scientific evidence clearly show that empathy is fixed (and biased) as anti-empathists hold to be true? Well, let us consider some of the most prominent experiments cited by Bloom and Prinz and which for them patently demonstrate the bad influence empathy has on moral agency/conduct. Take, for instance, the experiment by Batson et al. (1995) cited by Prinz (2011).Footnote 21 In this experiment, subjects watched a (fictitious) interview of a terminally ill child, and they were then given the opportunity to move that child up on a waiting list for a medicine that would improve the child’s quality of life, though not reverse the terminal condition. Then, subjects were given a further opportunity: they could decide to whether or not to move the child up on the waiting list, at the cost of displacing other children who had been on the list longer and whose condition was more serious. The experimenters divided the subjects into three groups: to one of them, no further communication was given, whereas the other two groups received, in one case, a communication to be objective and, in the other, that of taking the perspective of the child: ‘imagine how the child who is interviewed feels’.Footnote 22 After listening to the interview, subjects had to fill out an emotional reaction questionnaire, rating themselves on such emotions as ‘sympathetic’, ‘warm’, ‘compassionate’, or ‘tender’.Footnote 23 When given the opportunity to help the child by moving them up the list, 73% of those who were primed with the ‘high empathy/perspective-taking communication’ chose to do so, while only a minority of those in the control conditions did not.

Now, if this experiment undoubtedly shows that priming empathy boosts motivation and induces helping behaviour, it also shows that empathy does it in a biased manner, violating principles of justice and fairness. However, based on the arguments I have made previously, we know now that the only thing which is biased here is the way subjects of the experiments were manipulated. They were, in fact, only given three options and the only one which involved the use of empathy was already a biased choice. They could actually empathise, but only with the interviewed child. Nevertheless, as we have argued previously, the results would have been different if they had been given the opportunity to empathise both with the one child and with the other children waiting on the list. The result of the experiment is not caused by an inherent partiality of empathy, but by the partiality of the experimenters’ request.

This same explanation applies to most of the studies cited by Bloom and Prinz. It is not the parochialism or even the racism of empathy that is tested (let alone proved) in such studies, but the racism/parochialism of the subjects put to the test. Wittgenstein famously asserted in the Tractatus that ‘the limits of my language mean the limits of my world’; in the same spirit we can affirm that the limits of my concept of humanity and of the category’s application of ‘fellow human being’ mean the limits of my empathy. If I consider, for example, a certain ethnic group to be essentially different from mine, empathy for members of this group will hardly occur. Nevertheless it should be noted (as I have already mentioned in this work) that among the possible cures for racism one is, indeed, empathy. In fact, empathy is often the key to comprehension that the other is not so different from me, that we share a common humanity.Footnote 24

Another quite famous study which constitutes one of the cornerstones of Bloom’s argument is the experiment conducted by Klimecki et al. and reported in a 2014 article. As you will probably remember from the previous discussion, subjects here were asked to watch a combination of low-emotion and high-emotion videos, the former depicting everyday scenes with low emotional content, the latter scenes of human suffering. Specifically, they were asked to do so twice, once after what was referred to as ‘empathy training’, which required them to resonate with the plight of the other, and the second time after ostensible ‘compassion training’, which involved feelings of benevolence and friendliness in a state of quiet concentration. While watching the videos, the subjects underwent functional magnetic resonance imaging (fMRI) scanning. The researchers discovered that empathy training increased self-reports of empathic reactions as well as the strength of negative affect, while compassion training increased self-reports of positive affect. These results were interpreted by Bloom as demonstrating that (1) empathy and compassion are fundamentally different mechanisms; (2) compassion is better than empathy because it has a positive impact on the subjects, whilst empathy has a negative one. The problems with this interpretation are many. First of all, the experiment does not test what would happen if the subjects were exposed to video depicting individuals experiencing positive emotions: this is a serious shortcoming, since it is not obvious that here compassion would be ‘better’ (let alone ‘possible’) nor that the neural circuits would be different. Furthermore, the study does not engage the subjects in a behavioural task, thus it remains unclear whether empathy or compassion would be more effective as a moral motivator.

To sum up, all these experiments suffer from several critical weaknesses: either they do not test empathy, but something else, such as sympathy and compassion, or they test cognitive but not emotional empathy and vice versa, or they do not test behaviour at all. Moreover, absolutely none of these studies show that our empathic reactions are fixed; if anything, they prove exactly the opposite. For instance, the experiment by Klimecki et al. shows that our empathic responses may vary depending on whether we engage in empathy training or in compassion training. In fact, our empathic responses are sensitive to interventions (intentional or not).

Hence, empathy is far from being fixed and its effects can be controlled. Already this constitutes a significant undermining for the theses of the anti-empathists. Nonetheless, it is possible to go further. There is a theme we have touched upon a few times in this book and that we have not properly examined until now. I am speaking of what empathy essentially is. We have said in fact that empathy is a psychological mechanism (certainly it can be described as that) and that it is a capacity that people have to varying degrees (from people with high empathy, to people with low or impaired empathy). In addition, we have mentioned that it functions as an ability or a skill, one that can not only be enhanced and refined, but also reduced, depending on what kind of education one receives or which kind of training (or experiences) one undergoes. The question I am going to ask now is whether empathy can be defined as a virtue. Paul Bloom and Jesse Prinz would obviously deny that, but they are not alone. Heather Battaly (2011) is probably the staunchest critic of this view, having dedicated an entire article to the issue. In what follows, I will argue against Battaly that empathy should be considered a moral virtue and that this proves once again the centrality of empathy for morality.