Comment on Michael Slote: Moral Sentimentalism Thomas Schramme Almost everyone who has discussed Michael Slote's recent book Moral Sentimentalism complained about his lack of explicitness regarding the core notion of empathy (D'Arms 2011, p. 136; Quigley 2011, p. 484; Smith 2011, p. 198). Although I believe this is a fair point, hence that he better had been more definite about this central idea, I also happen to agree with Michael on the foundational importance of empathy for morality. I should be explicit though, right from the start, that my interpretation of this role differs from Michael's, most importantly in that I don't speculate about empathy's function in relation to (most) meta-ethical and normative issues; my concern is empathy's role in being a moral person and I see this as a problem in moral psychology. Still, my comment should be seen as a constructive or supportive one. I will first focus on the concept of empathy and then on the question in what way empathy might be foundational for morality. Although I wholeheartedly support Michael's project and intend to add to it by hopefully clarifying some issues, I will probably not assist him in his case for moral sentimentalism either. Sentimentalism seems to rely on a reading of empathy as an emotional phenomenon. But as will be seen there are cognitive elements in empathy and since I will undermine the very distinction between cognitive and emotional elements of morality, the theory I defend is arguable not a clear-cut version of moral sentimentalism. 1 I. The concept of empathy In the last few years there has been a lot of debate about, and attempts to distinguish between, different aspects of empathy. I will use some of these distinctions in order to clarify what aspect might eventually be pertinent to moral issues. According to Michael, empathy is "a motivating psychological mechanism" (Slote 2010, p. 5) that involves "having the feeling of another (involuntary) aroused in ourselves" (ibid., p. 15). It is to be distinguished from sympathy, where a person feels for someone. 2 To experience empathy, I take it, for Michael therefore means to have a shared feeling or emotion, usually one that has a negative quality, such as pain and anguish. Obviously, we may also share a positive emotion with others, but it seems plausible to use examples of negative emotions, since these are usually related to pro-social, beneficial behaviour. Empathy might also be felt vicariously (in addition to the vicariously shared feeling of other examples of empathy), for instance when we learn about a friend being left by her partner before she herself gains knowledge about it. 1 It might still be an argument for moral sentimentalism, after all, if it is understood as merely demanding some role for sentiments in morality. For instance, Justin D'Arms says that "[a] reasonably generic statement of the sentimentalist idea is that moral or evaluative concepts and/or properties depend on the sentiments (2011, p. 135). If that is the correct description of its purpose I agree with moral sentimentalism, though I doubt that many moral philosophers would disagree with it, hence the label might lose its distinctiveness. 2 In the following, I completely ignore another phenomenon occasionally referred to under the rubric of 'empathy', namely "emotional contagion" (Coplan & Goldie 2011, p. xxxiv). This refers to a mechanism of emotional convergence where, however, the focus is not another person's mind. Indeed, this mechanism is caused subconsciously and many animals are, in this respect, empathetic, because they can be affected by the feelings of other animals. 1
We then might feel shocked and tormented when thinking about her and how she might feel. 3 This seems like a straightforward description of a common and well-known phenomenon, but one might ask whether we always, or even normally, feel what someone else feels, when we empathize with them. Indeed, do we need to feel anything at all when empathizing? It has, for instance, been pointed out that we can empathize with a depressed person without ourselves being or becoming depressed (Darwall 1998; see Slote 2010, p. 16). Someone might of course respond to this particular example that we indeed do not feel empathy towards a depressed person but sympathize with her, i.e. feel sorry for her. But, again, we might, in all honesty, say to a depressed person, "I know how you feel", because we have had similar experiences before, say, phases of feeling low and being disheartened. This seems possible without necessarily feeling for the depressed person, i.e. without sympathizing with her. Yet to know how someone else feels does not require being in any particular emotional state at all, it simply needs some kind of acquaintance with the particular emotional state the other person is in. Maybe here is the right moment to pause briefly and to consider the notion of empathy from the perspective of its history. As it is well known, empathy was first used in the English language as a translation of the German term "Einfühlung", which had been introduced as a word to describe an epistemic mechanism of gaining access to other minds (Coplan & Goldie 2011, p. xiii). I believe especially for a German native speaker it is fairly obvious that Einfühlung does not necessarily require any present emotional state, because we can often understand what is going on "inside" another person who is herself not in a particular emotional state (see also Slote 2010, p. 144f., where Michael introduces the term "intellectual empathy"). So we humans can "einfühlen", i.e., in literal translation, "feel into", a target without actually feeling something. Arguably, to be able to gain access to another person's mental condition requires general emotional capacities, because it seems likely that we learn the capacity of "mind reading" at least partly in virtue of our ability to feel what the other feels. In contrast, a robot does not seem able to empathize, because it does not have experience of any feelings. To be sure, a robot might be able to "read" the mind of other people by deducing certain mental states from the behaviour of a target, but this would not constitute empathy proper, because it would not involve any aspect of the target's mental state. In other words, for a robot, it is not the mental state of a target that is in focus, but the target's behaviour. The mental state of the other is a piece of information that is deduced from studying the respective behaviour. But in empathy, the mind of a target is the object of our mental state, and not simply the deduced data of studying overt behaviour. All this is of course familiar to scholars interested in simulation-theory within the philosophy of mind, which contrasts with "theory-theory". The latter is a theory, or family of theories, which argue that understanding other minds to have a theory of mind works like applying a theory; hence the unusual title. In simulation-theory we find an interest in empathy as a capacity to "step into the shoes" of others. This can, in prin- 3 There is an interesting discussion in the relevant literature whether in empathy we mentalize about the other's mind or imagining ourselves in the situation of the other. Coplan, for instance, claims that only the former is genuine empathy, a phenomenon she calls other-oriented perspective taking (Coplan 2011, p. 55). I ignore these complications, since they seem irrelevant for the purposes of my paper. 2
ciple, be a purely cognitive mechanism, hence imply a belief about a mental state of another person, though simulation-theorists regularly include sharing affect as part of this process. It does not seem surprising, then, that in (neuro-)psychology we can find a distinction between cognitive and emotional, sometimes called affective, empathy (e.g. Shamay-Tsoory et al. 2009). As regards this distinction I have just claimed that cognitive empathy does not require the presence of emotional empathy, but that is does require emotional empathy as a general capacity of the (cognitively) empathizing person. This claim is an empirical one, and it would need to be supported by respective findings. It now seems rather obvious that Michael has interpreted empathy not as cognitive, but as emotional empathy, because he requires having the feeling of another (involuntary) aroused in ourselves for empathy. An example would be to feel the pain another person is experiencing. Now this example seems a paradigmatic one as regards morality, since feeling the pain of another person, and consequently helping her, seems to be a straightforward case of moral behaviour and of the role empathy plays for morality. Yet, again, it seems unlikely that we always need present emotional empathy to be motivated to do the morally right thing, for instance to help someone. There are cases of helping behaviour, where a person cannot have the feeling of another aroused in himself at all, because the target is unconscious, hence does not feel anything. 4 And there are examples of moral reactions where our feelings and the target's feelings are incongruent, as in the already mentioned case where we help someone who is depressed. Finally, there are cases of moral behaviour where we actually should attempt to suppress any shared emotion, or emotional empathy, for instance when we have to decide between two people whom to help. So emotional empathy might trigger moral behaviour but it might not be required for morality, and potentially even be detrimental to it. Surely, Michael is aware of all this, as can be seen especially in his discussion of deontological elements of morality (cf. Slote 2007, p. 27ff.). In line with this, emotional empathy does not seem to be the main aspect of morality for him, but "empathic concern". The latter term was introduced by psychologist Daniel Batson, and Michael also uses it, though in the final analysis he sticks to Martin Hoffman's developmental story of how the ability of empathy matures in humans (Slote 2010, p. 16ff.). Empathic concern seems to have an element of sympathy as part and parcel of the concept, because the concern aspect seems to be calling for a moral point of view, i.e. implying sympathy, which arguably is again to be explained by empathy. In other words, it seems to be potentially circular reasoning if we explain sympathy, i.e. a genuine moral feeling, by empathy, but define empathy as concern for the other. Michael also occasionally uses expressions that seem to imply a sympathetic element in empathy, for instance when he refers to feeling (more) empathy for someone (Slote 2010, p. 22). I suppose strictly speaking, emotional empathy is not felt for someone, but is mirroring the feeling of another person. However, it seems impossible when referring to a mature capacity to empathize to distinguish it from fellowfeelings such as sympathy. Hoffman, for instance, explicitly refers to a "sympathetic 4 The mentioned motivation might still require the presence of cognitive empathy, but I shall be noncommittal in this respect. 3
component" in empathy (Hoffman 2011, p. 234f.; see also Darwall's notion of "protosympathetic empathy", Darwall 1998, p. 271f.). So the conclusion of this first part of my comment is as follows: Michael appears to be interpreting empathy as emotional empathy, but when he considers the role of empathy for morality, he seems to imply a reading of empathy that is close to the notion of sympathy. I don't see this as a defect of the theory, but rather as an unavoidable outcome of a perspective in moral philosophy that focuses on empathy as basic notion. Empathy, I take it, has an important role for, and cannot be completely separated from, sympathy, although they are surely distinct phenomena. Sympathy, again, i.e. feeling for someone else, is arguably a necessary requirement of being able to take the moral point of view. This way of putting it entails a particular thesis about the role of empathy and sympathy for morality that I believe is at least implicit in Michael's theory, namely that both capacities are necessary for morality, but that they don't always need to be present as actual feelings in order for a person to be morally motivated (see also Slote 2010, p. 64, 76, for an account of empathy as capacity). Through the development of (our capacities for) empathy and sympathy we are able to take the moral point of view, but once we are moral persons, we don't have to actually empathize and sympathize in order to appreciate the morally right thing (see Slote 2010, p. 88, for an analogous statement that feeling of direct empathic concern need not be present in order to be disposed to act morally). In other words, we should see empathy and sympathy as capacities or skills that can be developed, i.e. generally and potentially universally applied. In the next part I want to elaborate on the foundational role of empathy understood as empathic concern, which has a sympathetic component in morality, i.e. in respect to taking the moral point of view. I will also argue that cognitive empathy has an important role to play in this development and that therefore the distinction between moral sentimentalism and moral rationalism gets blurred. II. Empathy as foundational for morality In what way is empathic concern the "cement of the moral universe" (Slote 2010, p. 14)? In this section I wish to argue that empathy, here interpreted as a capacity, not as a present feeling, is indeed a necessary requirement of morality. Morality, again, is understood as taking the moral point of view, or being a moral person. 5 In virtue of gaining access to the minds of others via empathy, we also develop the capacity to care about others. But this is not the end of the story. I believe in the final analysis to be moral means to care about morality, or to care about being moral, not to care about others (though we probably have to care for some people in order to become moral persons). To care about morality does not imply to care about particular moral standards. In fact, we often do not care about particular moral requirements that others do care a lot about. But a person who cares about morality does take moral considerations whatever they might be into account, hence a moral person is not an amoral person. 6 Now, an amoral person, in the sense of being a person who does not care 5 A moral person does not always need to act morally, but is able to act morally. This description is therefore analogous to calling persons rational, which also does not exclude irrational behaviour. 6 Again, an amoral person is not an immoral person. A person, who (regularly) does morally bad things might be a person who does indeed appreciate moral standards, but for whatever reasons does not 4
about morality, is usually a hypothetical character in moral philosophy. Similarly to debates in epistemology this person might also be called a moral sceptic. In moral philosophy, it is sometimes claimed that we need to give an answer to the most basic question, "why be moral?", i.e. to the question that the moral sceptic raises. But it seems that real exemplars of persons not caring about morality, psychopaths, are not capable of taking the moral point of view. At least this is what we seem to learn from research findings. So a psychopath is different from the amoralist in philosophical debates, who might, but does not, take the moral point of view although, to make things more complicated, occasionally an amoralist is also described as a person with a lack of moral sense, which, again, sounds more like a lack of capacity. Psychopaths are arguably not able to take this point of view because they lack certain capacities that are necessary in order to be a moral person. So the argument I want to make is to point out an empirical route to making progress in a philosophical dispute: When we consider what is the cement of morality, as Michael does, and when we understand morality as a particular point of view of persons, i.e. as an aspect of their being, not their doing, then we might be able to learn what this cement is by studying individuals who are incapable of taking the moral point of view; incapable of being a moral person. Psychopathy is the negative test case for establishing a theory about the required capacities for morality (see also Slote 2010, p. 51, 54f.). The most important finding in relation to our concerns here is that psychopaths seem to lack empathy (see, e.g., Baron-Cohen 2011). This apparently speaks in favour of moral sentimentalism. But now we need to be wary of terminology. I have already described the notion of empathic concern in a way that brings it close to what I have called the moral point of view, in virtue of including a sympathetic component. So, again, there is a potentially circular reasoning: i) Taking the moral point of view means to have empathic concern. ii) Psychopaths are unable to take the moral point of view. iii) Psychopaths lack empathic concern. However, I don't believe that here we have a vicious circle; since this line of reasoning seems more like an empirical backing of a conceptual thesis, established in the first claim. After all, at least the third claim is partly an empirical one, and it might prove false in the future. Accordingly many philosophers, not just me, study psychopathy as an interesting phenomenon for basic issues in moral philosophy. Indeed Shaun Nichols has argued that the findings on psychopathy threaten moral rationalism (Nichols 2002). Yet again, based on similar evidence, Heidi Maibom (2005) points out rational deficits of psychopaths, which seems to strengthen the case for moral rationalism. I believe this discussion is intriguing and worth pursuing, but it might be questionable whether it will in the end decide the philosophical quarrel between moral sentimentalism and rationalism. Indeed, I would now like to establish why I believe that the phenomenon of psychopathy actually undermines this very distinction and how it shows that morality (i.e. being a moral person) both requires rational and emotional capacities. If we follow many researchers and consider empathy to be (one of the) most basic moral capacities, or in any case to be one capacity that is missing in psychopaths, we first need to be aware of the pitfalls of the notion of empathy. In the first section I act accordingly, hence is substantially not a moral person and, in severe cases, an immoral person. An amoral person, in contrast, does not care about moral standards at all. He nevertheless might never act immorally, simply because he might not happen to violate any moral requirements. The difference can be described as the difference between being a non-moral person vs. not being a moral person. 5
referred to the common distinction between cognitive and emotional empathy. So even if someone establishes evidence that psychopaths lack empathy and that they are unable to be moral, it still has not yet been established whether moral sentimentalism has won, if only because empathy might be interpreted as a cognitive capacity after all. Hence researchers need to be more specific what aspect of empathy they study. Now, it seems to me that empathy is indeed a perfect example to show why our neat philosophical distinction between moral rationalism and moral sentimentalism might not work after all. I have repeatedly stressed that there are two aspects of empathy, which are called cognitive and emotional empathy. This might still allow for a distinction between moral rationalism and moral sentimentalism by focussing on one of the aspects of empathy. However, when we consider again the role of empathy for morality, it seems difficult to think of either by itself as a sufficient element of the moral standpoint. To be a moral person seems to require both aspects. We need emotional empathy to appreciate that certain behaviour is bad for others persons, and we need cognitive empathy to appreciate that all persons have needs and interests that are protected by morality, even those people we do not emotionally empathize with, and even in relation to those needs and interests that do not raise any emotional concerns (cf. Decety & Meltzoff 2011, p. 78ff., for further considerations regarding cognitive capacities in pro-social attitudes). Although we can analytically distinguish between emotional and cognitive empathy, we need to be wary of such a distinction when talking about morality. Indeed, we might go further in undermining the very distinction between cognitive and emotional aspects of our minds more generally. I like, for instance, what Gilbert Ryle has to say about these divisions: "In our abstract theorizing about human nature we are still in the archaic habit of treating ourselves and all other human beings as animated department stores, in which the intellect is one department, the will is another department and the feelings a third department." (Ryle 1972, p. 442) Obviously, much more needs to be said and scrutinized in order to make good such a basic and general claim, but at least the phenomenon of empathy should bring us to question our distinctions of different mental realms. In conclusion, I have tried to establish that the most basic aspect of morality is to take the moral point of view. I have argued that we can only be a moral person if we have certain capacities. These capacities can be determined by studying psychopaths, who arguably are morally incapacitated. Empirical research concerning psychopathy and related research regarding empathy calls into question the neat distinction between moral sentimentalism and moral rationalism, but underlines the importance of empathy for morality. So my paper has been an attempt to make good Michael's claim that empathy is "the cement of the moral universe", and I have spent most of the time clarifying in what sense this might be true. 6
References Baron-Cohen, Simon (2011), Zero Degrees of Empathy: A New Theory of Human Cruelty, London: Allen Lane. Coplan, Amy (2011), "Will the Real Empathy Please Stand Up? A Case for a Narrow Conceptualization", The Southern Journal of Philosophy, Vol. 49, Spindel Suppl., 40-64. Coplan, Amy & Goldie, Peter (2011), "Introduction", in: Coplan, Amy & Goldie, Peter (eds.), Empathy: Philosophical and Psychological Perspectives, Oxford: Oxford University Press, IX-XLVII. D'Arms, Justin (2011), "Empathy, Approval, and Disapproval in Moral Sentimentalism", The Southern Journal of Philosophy, Vol. 49, Spindel Suppl., 134-41. Darwall, Stephen (1998), "Empathy, Sympathy, Care", Philosophical Studies 89: 261-282. Decety, Jean & Meltzoff, Andrew N. (2011), "Empathy, Imitation, and the Social Brain", in: Coplan, Amy & Goldie, Peter (eds.), Empathy: Philosophical and Psychological Perspectives, Oxford: Oxford University Press, 58-81. Hoffman, Martin L. (2011), "Empathy, Justice, and the Law", in: Coplan, Amy & Goldie, Peter (eds.), Empathy: Philosophical and Psychological Perspectives, Oxford: Oxford University Press, 230-254. Maibom, Heidi L. (2005), "Moral Unreason: The Case of Psychopathy", Mind and Language 20 (2), 237-257. Nichols, Shaun (2002), "How Psychopaths Threaten Moral Rationalism: Is It Irrational to be Moral?" The Monist 85 (2), 285-303. Quigley, James G. (2011), "Review: Michael Slote: Moral Sentimentalism", Ethical Theory and Moral Practice 14: 483-486. Ryle, Gilbert (1972), "Can Virtues Be Taught?" In: R. F. Dearden, P.H. Hirst and R.S. Peters (eds.), Education and the Development of Reason, London: Routledge, 434-447. Shamay-Tsoory, Simone G., Aharon-Peretz, Judith, Perry, Daniella (2009), "Two systems for empathy: a double dissociation between emotional and cognitive empathy in inferior frontal gyrus versus ventromedial prefrontal lesions", Brain 132: 617-627. Slote, Michael (2007), The Ethics of Care and Empathy, London: Routledge. Slote, Michael (2010), Moral Sentimentalism, Oxford: Oxford University Press. Smith, Angela M. (2011), "Review: Moral Sentimentalism by Michael Slote", Analysis 71 (1), 197-200. 7