Medicine, Health Care and Philosophy

, Volume 15, Issue 2, pp 103–113

The desired moral attitude of the physician: (I) empathy


    • Institute for Ethics, History and Theory of MedicineUniversity of Muenster
    • Department of Health and SocietyInstitute of Medical and Health Sciences, University of Linköping
Scientific Contribution

DOI: 10.1007/s11019-011-9366-4

Cite this article as:
Gelhaus, P. Med Health Care and Philos (2012) 15: 103. doi:10.1007/s11019-011-9366-4


In professional medical ethics, the physician traditionally is obliged to fulfil specific duties as well as to embody a responsible and trustworthy personality. In the public discussion, different concepts are suggested to describe the desired underlying attitude of physicians. In this article, one of them—empathy—is presented in an interpretation that is meant to depicture (together with the two additional concepts compassion and care) this attitude. Therefore empathy in the clinical context is defined as the adequate understanding of the inner processes of the patient concerning his health-related problems. Adequacy is scrutinized on behalf of the emotional and subjective involvement of he physician, and on the necessary dependence on medical—moral—goals. In the present interpretation, empathy alone is no guarantee of the right moral attitude, but a necessary instrumental skill in order to perceive and treat a patient as an individual person. The concepts of compassion and care that will be discussed in two forthcoming articles are necessary parts to describe the desired moral attitude of the physician more completely.


Clinical ethicsEmpathyCompassionCareDetached concernVirtue ethicsMoral motivationProfessionalismMoral attitude


Traditionally the physician’s role is interwoven with moral1 responsibilities and obligations towards the patient. In Western culture, the ancient Greek Hippocratic tradition still has some claims to validity: the Hippocratic Oath embraced concrete practical tasks as well as obligations concerning general behaviour, character and attitudes. Modern codes such as the declarations of Geneva (1948, last revision 2006) and Helsinki (1964, last revision 2008), or the much-discussed physician’s charter from 2002, also require of the doctor, implicitly or explicitly, an adequate inner background of concrete rules of conduct. Edmund D. Pellegrino states:

“The way in which principles, rules, caring, hermeneutics, casuistry, or any alternative theory of ethics is conducted will depend on the kind of persons carrying out the moral acts and their analyses. Intention, moral psychology and the “story” of the agent’s life cannot be separated from the agent’s moral behavior.” (Pellegrino 1995, p. 266)

Though modern medicine seems to focus on science and technology and has gained remarkable successes this way, good medical practice encompasses more than this knowledge and these skills. During the last 60 or more years, medical education has focussed largely on mastering the ever expanding scientific knowledge. Since the 1970s, however, an increasing unease with the application of technical medical facilities has been expressed, leading to a rise of bioethics and the patients’ rights’ movement, and urging physicians increasingly to re-evaluate the ethical and humanistic basis of their profession. The result were not only more or less formal requirements to respect the autonomy of the patient better, but also a reinforcement of the traditional ethical demands on the trustworthiness of physicians. For example, it was emphasized in German medical curricula that students need to learn the relevant knowledge, skills and the rightattitude (Biller-Andorno et al. 2003). In many medical schools efforts are made in order to broaden the spectrum of education from theoretical knowledge to practical skills, and ethics courses are nearly everywhere obligatory. Unfortunately, appeals for appropriate behaviour or attitude on the part of a physician usually remain vague, and thus it is difficult to teach it, even in the most ambitious settings of problem-based learning and student hospitals with actors as patients. (Strassmann 2007) If the right attitude of a doctor is an ethical obligation that is among the basic requirements of medical education, it would be helpful to understand what it is. Pellegrino supports virtue ethics in this regard (Pellegrino and Thomasma 1988; Pellegrino 2007) and though he admits that psychological and emotional elements are necessary parts of morality, he still focuses on rational or cognitive virtues that predispose the good physician to act morally. He definitely criticizes moral emotivism and intuitionism as too relativistic. Still, an interesting question remains—what does the emotional moral side, or the more emotional moral virtues, let’s say, look like? A great many terms have been suggested as appropriate in order to satisfy the conception of the good, adequate attitude of the physician: sympathy, empathy, compassion, pity, benevolence, beneficence, care, love, charity and others. All of these concepts include a more or less emotional state that is understood as a necessary part of moral behaviour of the physician.

In a series of three articles I want to examine three of the most popular candidates—empathy, compassion and care—and interpret them in a way that can fill the variable ‘moral attitude of the physician’ with a more concrete and consistent content. My intention is not to give a general overview of the concepts in question, or a distillation suitable for all purposes, but rather a reconstruction within these terms that will describe more closely and hopefully more concretely the desirable moral attitude of physicians. For this purpose I suggest special (and limited) interpretations of the concepts empathy, compassion and care.

This analysis is neither a moral psychological inquiry nor a confession of moral emotivism.2 However interesting and important moral development and its required mental capacities are, empirical anthropology is not my issue. I want to re-construe the more emotional elements in the normative image of a good physician. No claims for a primacy of these elements in an encompassing moral theory are included. Rather, it can be seen as a contribution to a certain field of medical virtue ethics, with a focus on the intentional and motivational side, such as Michael Slote has also framed (Slote 2001).3

Undoubtedly, there is a need for something in addition to medical scientific knowledge, practical skills and experience that ensures that the physician does the right things with the right attitude in order to reach the goals of medicine. Many authors attribute a leading role to empathy in this context. According to the physician and medical philosopher Eric Cassell, for instance, empathy safeguards the emotional understanding and nearness to the patient, while the medical training is responsible for the necessary distance and objectivity. Medical art is exactly to negotiate between these poles (Cassell 1976, 1991). Empathy is an often-mentioned source of moral attitude and motivation, and a skill necessary to achieve a good relationship between physician and patient for the task of “humanizing medical practice” (Halpern 2001) Jodie Halpern in her influential work “From Detached Concern to Empathy”, argues engagedly for the need to improve the emotional skills of doctors:

“Emotions are seen as impinging on medical judgment in two distinct ways. First, certain emotional states are seen as disruptive to thinking. (…) Second, and more fundamentally, even calm emotions are seen as unreliable sources of information about the world because they are so subjective.” (p. 30)

Therefore, her goal is

“(…) to argue for the skilled inclusion of emotional reasoning in medical judgment by showing that emotions influence even seemingly detached beliefs and decisions.” (p. 34)

This goal I share. Halpern criticizes the “detached concern” movement that aims at keeping the person of the physician out of the clinical encounter, seeing him as a neutral observer rather as an agent. What is necessary for a fulfilled and good medical practice is neither “detached reasoning nor unskilled sympathy” but “something distinct, straddling cognitive and affective capacities”, namely empathy (p. 61). Unfortunately, this concept is far from being clear and unequivocal, and while supporting Halpern’s project I think one can profit from a distinction between different aspects in the desired capacity she describes.

Empathy, feeling inside the other, is often taken as synonymous or nearly synonymous with the attitude of sympathy, compassion, benevolence and care. This amount of synonymy can only lead to confusion. I shall try in what follows to make distinctions between these concepts and specify them because, however closely they may be linked in successful and morally good clinical conduct and action, they have different tasks. I will focus especially on distinct and separate understanding of empathy, compassion and care, and on how they are linked to each other. I will distinguish between the capacity (or skill) of empathy, the adequate professional inner attitude of the doctor (compassion) and the active side of this attitude (care). The combination of the three will fulfil the function of what Jodie Halpern attributes to empathy alone, and by developing the triad I hope to contribute to more clarity. This might even reconcile and explain seemingly competing positions of “detached concern” and humane closeness in form of “empathy”, “compassion” or “care”. At the same time, my suggestion may give a more detailed insight and therefore an additional way to foster the acquisition of this professional virtue(s). In a first step in this article, I will develop a definition of “empathy” in the clinical context in order to delineate the different meanings of “sympathy”. Then I will scrutinize the scope of this concept in detail, with respect to the depth of emotional involvement of the physician, the role of his subjectivity and the moral impact of “empathy”. In conclusion I will summarize my results and take a short look at the elements necessary for a morally sound attitude in physicians.

Sympathy and empathy

In the literal meaning, “sympathein” means “suffering with”, while “empathein” means “suffering inside”. Nowadays, both notions usually not only encompass feelings of suffering, but all kinds of feeling. Thus “sympathy” can be understood as feeling with another person, while empathy means feeling inside the other one. There are important differences. In sympathy, there is a (nearly) symmetrical situation, the same kind of inner state is to be reached; you try to feel the same thing as the person you are sympathizing with (Olinick 1987); yes, you are identifying with her. The first step on this way is sometimes called “emotional contagion” (Hoffman 2000). The difference between contagion and sympathy in this sense is that contagion is unconscious and unequivocally limited to the own person. Sympathy, on the other hand, is directed to another person, though it tends to merge identities. (Wilmer 1968) It is as if you were the other person.4 In empathy, you try to understand the feelings of the other person; you are aware that you are outside and have to reach inside the other one. You cannot understand these feelings without your own emotional experiences, and understanding in this context means more than rationally detecting feelings. Probably your own emotions will be evoked, and sometimes empathy can lead to sympathy. This is exactly the purpose or benefit of empathy, for example on behalf of a friend. But it is possible to distinguish between the terms.

There is also a second, derived, meaning of sympathy that should be even more clearly distinguished from empathy: sympathy as a positive, warm feeling for another person. Understanding, even fully understanding the feelings of another, does not necessarily imply liking this person or these feelings. Understanding does not imply approving. Sympathizing does. For our goal of describing the appropriate attitude of physicians towards their patients, sympathy in this second meaning can be nice, but it would be very bad if it would be a necessary condition of treating a patient. A professional and conscientious physician is exactly one who also treats a very unsympathetic person according to all rules of the art, including his own customary moral attitude. It would not be reasonable to demand sympathy in this sense as a precondition for treating patients. But also sympathy in the first sense is not an appropriate emotional basis for the clinical encounter. If the doctor would get into the same emotional state as the patient in a very severe situation, he would lose his capacity to judge the situation and the medical options in a reasonable, ‘objective’ way.5 So sympathy is not the right moral-emotive background for the doctor-patient-relationship. But what can be the role of empathy?

In the clinical encounter, empathy can be described roughly as the adequate understanding by the physician of what happens inside the patient in relation to his complaints. This involves feelings, sensations (as pain, e.g.) as well as conceptions, causal and evaluative convictions, hopes and fears about the disease and of the role and options of the physician and the therapy. This description leaves many questions open of course. One of the crucial questions is: what is an adequate understanding?6 The disagreements as to the scope of adequate understanding are comprehensive. The range passes from ‘feeling the same things as the patient’ (Katz 1963; Gladstein 1983; Rushton 1980) over ‘a morally good, hermeneutic way of understanding another human being’ (Pedersen 2008, p. 332) to a ‘value-neutral and objective clinical skill in order to achieve indispensable facts about the inner proceedings of the patient’ (Zinn 1993) At least three different aspects of “adequate understanding” can be distinguished: (1) the emotional involvement of the physician and the degree of emotional parallelism with the patient that is sufficient for understanding, (2) the role of the physician’s subjectivity in empathy, and—crucial for this investigation—(3) the involvement of morality. These three aspects are related to each other; and it may not be possible to introduce neatly distinct limits for each of them, but it is useful to emphasize the different focuses.

Emotional involvement

A fundamental question to be answered is the amount of emotional participation by the physician that can be prescribed by and is necessary for empathy. Before addressing this question, I should clarify how I use the terms “emotional”, “cognitive” and “rational”. Medicine as well as philosophy share the tradition of distinguishing sharply between emotions and feelings7 on the one side, and rationality, logics, and cognitive capacities on the other side. Emotions are thus understood as immediate, spontaneous and mainly involuntary affects of the subjective individual, while rationality is the exceptional capacity of human beings to surpass subjective perspectivism and—by means of language—understand disinterestedly objective facts and causations. The rational side has been understood as the basis of natural science and of scientifically understood humanities (“Geisteswissenschaften” in German). Modern neurology and neuropsychology question these sharp distinctions. The rationality of emotions has been discovered (de Sousa 1987; Solomon 2006), and there are fascinating discussions as to the scope, meaning and influence of emotions in (rational) human decision-making and agency in the philosophy of emotions.8 Actually, also this series of articles is an example of dissolving this hard distinction, but I will use the terms “emotional” and “cognitive” in the traditional way. It is still prevalent in medicine as well as in the main stream of ethics, and in these areas the burden of proof (or at least of arguments) seems to be carried by the proponent of a graduated system of mental capacities, and of subjectivity and objectivity.

After these clarifying remarks it is time to consider the question about the adequate emotional involvement of the physician. Is a doctor only empathic if he is able to feel the same things as the patient (making empathy a synonym for sympathy, in the literal meaning)? The odds for this are poor: the background and the experiences of the average physician tend to be quite different from those of the average patient (concerning age, education, social status etc.) (Spiro 1992)9 And the task cannot be to use empathy only on those patients who are near enough to me to be able to identify myself completely with. Adequate understanding cannot imply total understanding on an emotional level. How much parallelism is necessary to enable adequate empathic understanding? Psychiatrist Harry A. Wilmer, who tends to emphasize the complete internal perspective of empathy in the sense of a full identification, is still aware of the instrumental use of empathy and of its necessary limits:

“If there is empathy there is real understanding of the other as another person. Here we understand his suffering in relationship to his personal and social world. We share, we feel for him and with him; psychologically, we get inside him for the purpose of understanding how he feels. In empathy it is as if: ‘If I were he.’ To achieve an empathic relationship, we use ourselves as the instrument for understanding, but by the same token we keep our own identity clearly separate. (…) In this situation the observer guards against his biases and misperceptions, and must thereby understand himself.” (Wilmer, 1968, p. 245)

This position seems peculiarly oscillating: the observer “enters into the equation and is then removed”. Empathy in this sense requires full, internal understanding. Wilmer gives an example of understanding what it means for someone having lost a child. Empathy would mean (according to Wilmer himself) to share the desperation up to a point of needing comfort oneself and not being able to help anymore. Only the realization that it is another person’s grief allows acting in a helping way at all. And only distancing oneself from one’s own subjectivity allows one to obtain any adequate empathic knowledge. Charles D. Aring, citing an earlier article of Wilmer, even goes as far as using this awareness primarily in order to dissociate from the patient:

“A subtle and significant feature of a happy medical practice is to remain unencumbered by the patient’s problem.” (Aring 1958, p. 452)

This actually moves the focus of empathy into the opposite of the usual sense. Though being aware of the limits of empathy, its use generally is understood to guarantee certain nearness, not a useful distance between patient and physician, “a middle way between emotional contagion and overintellectualization” (Agosta 1984, p. 59).10 So our question now is less the possible parallelism between the physician and the patient but the depth of the physician’s emotional involvement. As mentioned above, Eric Cassell suggests using empathy as a nearly unlimited emotional pole of medicine that has to be balanced with the objective, scientific side of medicine.11 The physician accordingly has to split into two: one empathic, thoroughly emotional and understanding part (which we are investigating now), and a different, rational, objective part that has been trained by the usual medical education. The “art of medicine” is to negotiate between both sides, the person of the physician is who “has to put Humpty Dumpty together again” (Cassell 1991). So on behalf of empathy, full emotional involvement is demanded at first, though it has to be regulated by an instance of judgement later on. In contrast, the Swedish physician and philosopher Rolf Ahlzén proposes not approaching the patient with both extreme poles of medical skill: full emotional understanding and absolute objectiveness; and leave it to medical art to find a way between them. He understands empathy as a mixed cognitive-emotional conception, generally12; and in the clinical context with an emphasis on the cognitive aspects. He thinks that it is possible (and advisable) to understand a person’s feeling without having it oneself at the moment, or even having felt it at all, or only to a very slight degree. Otherwise, we could never understand a dying person in such a way as to be able to treat him adequately, and a female doctor would not be able to understand a male patient (or vice versa) concerning sex-related sufferings. Ahlzén even holds that a too strong involvement could be dangerous, as it could tempt the physician to mistake his own feelings and preferences for the ones of the patient. Therefore, too deep emotional involvement can even stand in the way of adequate understanding, and mislead the physician into being strongly paternalistic. (Ahlzén 2010, pp. 278–287) Nevertheless, a certain emotional resonance will be an advantage for an adequate understanding. Without any experience of and reference to one’s own feelings it will be impossible to understand the feelings of another person. And if signs of respect, interest, understanding and good will are shown, the chances are much better that the patient will make understanding easier and show and explain what he feels. So it is not possible to make totally clear distinctions between the physician’s emotional capacities and involvement and a mere objective (or at least inter-subjective) understanding of the feelings of the patient. But though I consider basic experiences in emotions are necessary for any kind of understanding, for reasons of conceptual clearness I follow Rolf Ahlzén, who holds that even the initial access to the patient’s feelings should be seen under a largely non-emotional, goal-directed aspect. He distinguishes between empathy as an instrumental skill of understanding and the additional aspects of good-will (or benevolence) and the practical capacity to judge (“phronesis”) and perform what is necessary for attaining the goal of helping the patient.13

In the words of Ellen S. More:

“Empathy as I understand it does not stand in opposition to objectivity; it is not an attitude such as sympathy, compassion or concern. Rather, it is a form of relational knowledge. Its manifestation is not “concern” but “presence”. The empathic physician is neither objective nor subjective, neither detached nor identified, but dialogically linked to the patient in a continuing cycle of reflexive interpretation that integrates the objective and the subjective.” (More 1996, p. 245)

Sometimes it is questioned if empathy is necessary for all physicians. Does a surgeon need empathy at all? Is not the only imperative that he should possess all the required knowledge and is a skilful operator? Often this question is even more pointed: would you prefer a skilful or an understanding surgeon? It is clearly not a choice between the one and the other. Both capabilities are necessary for a good surgeon, if he doesn’t want to dispense with all contact with the waken patient and renounce all history-taking, diagnostics and detection of indications. It may be true that a surgeon usually needs less empathy for his work than an oncologist or a psychiatrist, but this is more a quantitative than a qualitative difference—he doesn’t need it actively only while the patient is unconscious. As a capacity it is indispensable.14 Swedish physician Carl-Edvard Rudebeck has emphasized the fundamental role of “bodily empathy”, that is understanding of the symptoms of a lived body, for any clinical decision-making. (Rudebeck 1992) Even a very basic and seemingly banal history-taking is not feasible without a considerable amount of empathy. “From time to time I drink a glass” can mean anything from being a water drinker to severe alcoholism. Nearly every single expression of the patient needs interpretation, and without empathy it is not possible either to understand, or to ask the right and relevant questions, in order to guide further diagnostics and find out about indications and contraindications. So also the choice of the right therapy depends on sufficient empathy, not to mention the capacity to explain the indicated measures to the patient and get his consent and compliance. If a patient does not take his life-sustaining pills according to the therapy plan (and that is not unusual), he is not just stupid or stubborn; something has gone substantially wrong with the physician-patient-relationship. The physician has failed to understand the patient sufficiently concerning his priorities and failed to make him understand (or at least to rely on the physician’s trustworthy knowledge about) the absolute importance of the therapy. So empathy is a skill of understanding that is necessary for every physician who works in a clinical setting.

We have already rejected the thesis of total emotional parallelism, but given our thesis that empathy should be a predominantly cognitive skill of understanding the inner processes of the patient, the open question is still: how much understanding is adequate? It is not possible to answer this question without a link to the intentions of the physician. Adequacy is not a matter of objective measurement but of the goals pursued by the physician. The physician understands the patient’s emotions sufficiently when he is acquainted with all the inner processes of the patient that influence what the physician needs this knowledge for. What is necessary for every physician in the encounter with a patient, in order for him to obtain this knowledge, is interest in the inner processes of the patient, emotional experiences that enable him to understand and interpret the patient’s emotions, and an adequate skill in responding non-verbally and by skilful and sensitive dialogue (Zinn 1993); but not necessarily sympathy or benevolent intentions. (Ahlzén 2010, p. 281) Usually, the physician will have the intention of following the goals of medicine, to help the patient, treat his illness, relieve his suffering and save his life, if possible and will usually answer to the needs and expectations of the patient: this makes even clearer how important the skill of empathy is in fulfilling these intentions. It is possible, however, to distinguish these morally good intentions and goals from the instrumental skill of empathy. As Annette Baier (2007) appeals to physicians:

“Must an understanding of the causes of pain be left to torturers, as understanding trust was largely left to con-artists and terrorists?” (p. 141)

That implies a generally good intention of physicians, but a lack of necessary understanding and nearness. On the other hand, it is easy to see that there can be too much closeness. The doctor is allowed to enter usually private and intimate parts of the patient’s life, and too much physical touch or an overly sentimental way of consolation may be perceived as intrusive. Though this is a rare complaint with regard to doctors, in principle the adequate closeness is a matter of balance between two extremes. Empathy as capacity of understanding should allow perceiving how much closeness the patient wants and needs. Can there also be an excess of understanding? It happens quite often that the doctor gets to know details of the life that the patient is not proud of, and that he would prefer nobody to know, or at least not a stranger like the doctor is. The more empathic a physician is, the more of these insights she will gain. It needs trust on the professional distance and objectivity and on the abstention from personal or moral judgment, not to perceive this over-understanding as intrusive and menacing. Empathy is the prerogative capacity to detect the right balance.

Nonetheless, using empathy mainly instrumentally in the context of the moral attitude of the physician does not mean we only need its rational, cognitive parts. The doctor is not simply a detective who identifies the patient’s feelings from an objective standpoint. Empathy necessarily relies on emotional interactions, on resonating and attuning to the mood of the patient. It is a complex experiential way of understanding the other one, not merely a cognitive skill. (Halpern 2001, pp. 18, 73, 75, 84) Using empathy instrumentally only means that the empathy required is necessarily dependent on its intentional and practical background, and so its adequate range differs from the empathy that is required in other contexts, such as care for children, love of a partner, hearing music or reading a book.

Empathy, as I have conceptualized it, does not specify the adequate kind of nearness, however. It is a mainly receptive capacity; it should not imply getting into the same emotional state as the patient; it shows how much emotional closeness is adequate for the patient between the medical necessities and his needs and wishes, but it leaves open which inner state of the physician is desirable. Sociologist Talcott Parsons describes as one of five patterns of role orientation “affective neutrality”. (Parsons 1991, pp. 58–67) Undesired affectivity (like anger) should be inhibited and “evaluatively” neutralized. Not all emotional affectivities are to be inhibited, however. (Daniels 1960) Nancy Sherman elaborates “good emotions” as virtues even for stoic sages and soldiers, neither of whom ought to be a “robotic creature, devoid of all feeling” (Sherman 2005, p. 109). Role-specific active involvements are rather encouraged and supported. Parsons focusses on the limits of emotional involvement; proponents of emotional virtues emphasize the required warmth and nearness to the patient. (Gelhaus 2011a, b) With regard to the right form of inner engagement of the physician, the concept of compassion seems to offer more specific insights than the concept of empathy. I will interpret it in the second article of this series.

The subjectivity of the physician

The instrumental and rational use of empathy should not foster the illusion that it constitutes objective and encompassing knowledge, or facts that are established through its use. As Reidar Pedersen remarks in his thorough overview of different understandings of empathy, it is hopeless to eliminate the empathizing subject from the concept of empathy. (Pedersen 2008, p. 327) Attempts have been made to understand empathy as direct, accurate or telepathic access to the patient’s feelings, but for Pedersen this is to abstract empathy from the empathic subject. He also rejects the reliance on universal human traits, mimicry, “mirror neurons” or similar constructions that seem to make superfluous the analysis of the shaping influences of the empathic subject on the empathic experience. As Ahlzén also warns, it is crucial to be aware of the limits and pitfalls of empathic capacities, exactly in order to guarantee its accuracy as far as possible. It would be presumptuous to assume owning the objective and totally appropriate “facts” about the patient’s illness experiences. It is important in any encounter with patients, but especially in circumstances that are focussed on his inner states and feelings, to be aware of the attitudes, emotions and biases that shape one’s own perceptions. Phenomenological attention to one’s own historical and social constitution shows that it would be a mistake to forget one’s own subjectivity. In addition, fundamental psychological and psychoanalytic knowledge show that a careful awareness of one’s own perception of the patient’s inner processes is necessary: I am referring to knowledge about phenomena such as transference and counter-transference, projection and repression, as also biological, anthropologic conditions of every human being (including the physician). The way of adequate understanding encompasses a self-critical analysis on the part of the physician of his own attitudes, emotions, biases and experiences as related to this unique situation and this unique patient. Thus far Wilmer is right in his demand that the physician must know himself, his own weaknesses and biases, his fears and rationalisations. (Wilmer 1968) However, is he right in asserting that the physician can know himself well enough to efface himself from the calculation? If Wilmer sees the physician as being his own instrument for understanding the patient, it is quite clear that this is impossible. Even as a mere spectator, the doctor is a part of what happens in the clinical encounter, and obviously he is much more than a passive spectator. How can a self-effacing person be empathic? If she only mirrors the mental states of the other one without having an own personality, the benefit of empathy would be negligible. Would the mirrored person just feel less alone, if the empathic person had no own personality? Furthermore, how would it be possible to mirror the inner states of a person without own personality? Could not at best the behaviour be imitated, which would be no empathy at all? In a meaning of “understanding”, as I have defined empathy, being empathic presupposes an own mental sphere that recognizes another person. Rather than effacing himself as a result of empathic understanding, a physician should determine what his feelings are, not only towards the illness experiences of the patient, but also with reference to his own relation to the patient, his own feelings. Is he sympathetic with him, does he pity him, or does he dislike him and detest everything he stands for? Empathic understanding means not only realizing the internal emotional and conceptual reality of the other, but also analyzing one’s own nearness or distance to the other’s experience. Without this, the supposed understanding will not be adequate. But one’s own authentic person cannot be totally removed. Rather, one must find ways to dispose of one’s deleterious reactions and control them in a way that makes a fair treatment of the patient possible. In a phenomenological understanding, inter-subjectivity cannot be reached as simple face-to-face encounters, “rather, the three regions ‘self’, ‘others’, and ‘world’ belong together; they reciprocally illuminate one another, and can only be understood in their interconnection” (Zahavi 2001, p. 151).

This is an important reason for having certain demands on the character and attitude of the physician. It is not morally desirable to have a cynical, misanthropic, malevolent doctor who really would be in need of effacing his personality during his contact with patients, nor would such a person have any genuine, intrinsic motive to do so.15 Identifying as a whole person with one’s professional goals and tasks is also desirable for the physician himself. It makes the profession more rewarding and fulfilling and thus can protect from burn-out and help to overcome difficulties and frustrations. Empathy, however, at least as understood here, is too weak a concept to guarantee that. It only encompasses the understanding, not the inner benevolent attitude to the other person.16 Additional concepts that influence the motives and goals of agency (such as compassion and care) seem to be needed, in order to describe the professionally benevolent attitude in a more specified way.


Having conceptualized empathy as mainly cognitive capacity and as mainly instrumental skill could imply that I simply exclude it as a candidate for emotional virtues of physicians. This is not the case, and empathy will show to be a necessary part of the complex of empathy, compassion and care that I propose to be a detailed description of the desired attitude of the physician. But before introducing empathy as necessary modifying aspect in the adequate attitude of compassion in the next article of the series, it is worthwhile to consider the moral implications it contains on its own.

As Rolf Ahlzén (and also Zinn 1993) show convincingly, even emotional parallelism does nothing to guarantee a positive attitude towards the patient:

First, I may very well feel what you feel and do nothing about it exactly because I empathize in this sense. My “feeling-with” may not evoke sympathetic responses at all, on the contrary, I may feel revulsion and disgust at your weakness and disgust at the ugliness of your predicament. (Ahlzén, 2010, p. 279)

A perfect empathic understanding may very well be used for selfish or even malicious intentions (as in market research and advertising, or in some sectarian contexts, for example). Especially because the possibility exists, however uncommonly, of being at the same time empathic and depreciative or even malevolent, it does not seem to be useful to equate empathy with a morally good attitude. Moral neutrality is even possible if the concept of empathy is understood as a strong emotional parallelism. Even if I can understand another person very well, and really sense the same things; even identify myself strongly with him, I still may not accept my own feelings or my own value. The imperative “Thou shalt love thy neighbour as thyself” presupposes a healthy self-love in order to be a reliable moral guidance. And in this case even the right attitude (love) is prescribed. Empathy even in its largest extension only implies the same feeling, no positive attitude towards it. If it is only understood, as here, as a necessary but not sufficient tool in order to achieve a certain goal, the automatic connection with morality seems even less convincing.17

But we have to be aware that the possibility of misusing a good thing does not necessarily rob it of its moral status. For example, my general inclination to help people may be used to support people with bad intentions. If I save a known murderer who is eager to kill again, he might be able to do so only because of my helpfulness. That does not make helpfulness a morally neutral or even bad attitude.18 It is a morally good attitude that is perhaps misused but not disvalued as a concept or as an attitude. Is it the same with empathy?

Is empathy as such a moral source, a core of what makes a good attitude good? In this analysis, empathy is understood as a form of cognitive-emotional understanding. The attitude and the intentions with which this understanding is used will here be conceptualized as compassion and care, interpreted in a specific way. If empathy is conceptualized this way, it is not the internal value of empathy that makes a malicious, deceitful use of empathy bad. Empathy only demands not to be satisfied with an insufficient (e.g. only rational or theoretical) understanding, lacking curiosity and interest. Usually, empathy is used with benevolent intentions in the clinical setting—and that is how it ought to be. But the concept of empathy used here does not include good intentions and the appropriate attitude. In spite of this there are three ways in which empathy is an important part of the moral action:
  1. 1.

    Current neuropsychological findings support the idea that empathy is a precondition for developing moral behaviour. (Hoffman 2000) These findings fit well with the theories about moral-psychological development of Kohlberg (Kohlberg et al. 1983), and even more so with Gilligan (Gilligan 1982). Without a real cognitive-emotive understanding of the other (different from me, but still understandable and somehow like myself) I will not be able to learn the very basics of respecting others nor develop the inner necessity treating them in a good way (e.g. according to the Golden Rule).

  2. 2.

    In the clinical context, empathy is an instrumental necessity for setting good intentions into practice. It is, in fact, indispensable for getting the information needed for diagnosis and therapy, for without empathy we cannot communicate about the somatic and personal afflictions and predicaments at all. For a meaningful history-taking, for identifying and understanding the relevant symptoms, for explaining the options of biomedicine and for making correct diagnostic and therapeutic decisions (especially in the form of shared decision-making), the physician needs empathy, and needs it even more for the very basic condition of a good patient-doctor-relationship, for gaining the trust of the patient, for getting his “compliance”.

  3. 3.

    Empathic understanding is often helpful (good) in itself. Even (or especially) if there is not much to do about a disease any more in a biomedical sense, the fact of being recognized and understood by a doctor may be very important, and consoling, and catalyse finding a way to cope with the disease. An effective available therapy does not exclude the value of an empathic understanding but they reinforce each other. It is remarkable that the simple action of understanding can make so much of a difference. (Berger and Mohr 1967; Cassell 1976 and 1991)


The concept of empathy as understood here implies some aspects of moral value: an anthropological precondition for morality, a necessary skill for good clinical action and decision-making, and also a moral-pragmatic activity. But mainly, in the clinical context and for the purpose of describing the moral attitude of the physician rather than the good outcome or the good action, it should be seen as a morally neutral conception in itself. It is, of course, not morally neutral to possess this capacity in a morally good context: if empathy is necessary for a good clinical encounter, it is of course a valuable part of the good action. It does not hide or imply, however, an intrinsic value with respect to the desired attitude, or, in order to put it differently: the value of the morally good behaviour towards the patient is not nourished by the necessary empathy in it (in this it is similar to the essential scientific knowledge that also is indispensable for the goals of medicine, but not the genuine source of the morally good).


Summarising my analysis of empathy in the clinical context I have defined empathy as adequate understanding by the physician of what happens inside the patient in relation to his health problems. This involves feelings, sensations (as pain, e.g.) as well as conceptions, causal and evaluative convictions, hopes and fears about the disease and of the role and options of the physician and the therapy.

Clinical empathy is a mainly instrumental skill, necessary for any physician in her encounter with the patient. I have examined the adequacy of empathic understanding in three directions: the emotional involvement of the physician, the role of the physician’s subjective state of mind and the moral impact of the concept of empathy. The first two points focus on near-lying mistakes the physician could make in his attempt of understanding: mistaking his own feelings for those of the patient or meddling both. The third point refers to the goals of medicine for which the physician’s empathy ought to be adequate.

Too much emotional involvement is generally counterproductive for the best outcome for the patient. It would also make fair treatment of all patients regardless of their sympathetic personality (or of their gender, skin colour, age etc.) impossible. The emotional involvement of the physician must therefore be limited, though not limited to the point that makes adequate understanding impossible. Empathy demands of the doctor to be sensitive, open, interested, and near, but not too near, to the patient. The physician has to understand the emotional state of the patient, but she must retain the emotional state that is appropriate for her tasks. The careful distinction between the subjectivity of the physician and that one of the patient is also necessary to guarantee one of the most important functions of empathy in the clinical context: taking seriously the patient as a complete, distinct, unique human being, not only as a carrier of diseases and symptoms to treat.

Even if we reject empathy as a genuine and necessary source of a moral motivation and a moral attitude, we still have to admit that a good empathic capacity will make it more probable that these can both be sustained and be filled with life. One’s ability to understand the inner processes of a patient has a good potential for bringing one nearer him. It also has the potential to foster whatever moral motivation one may have, as also one’s positive attitude, good intentions and general identification with one’s professional duties. That empathy has many different and interconnected potentials conducive to good moral agency is evident.

Empathy can be accompanied by a positive, but also by a neutral or even malevolent attitude. In order to understand more closely the moral sources of the desired attitude of physicians, we will also have to examine the concepts of compassion and care in two forthcoming articles. These concepts will give us a more detailed idea of the inner attitude and the activating impulse that professional ethics may have in mind when it demands “integrity” and “altruism” as in the Physicians’ Charter (ABIM et al. 2002); or the “purity of my life and my arts”, which are to be dedicated to the good of the patient, as in the Hippocratic Oath; or “complete loyalty” to the patients and the consecration of one’s life “to the service of humanity” in the WMA International Code of Medical Ethics (1949) and the Code of Geneva (WMA 1948).


In the following, I will use the terms „moral“, „good“ (respectively „inherently good“) and „right“ in the meaning of referring to the idea of a special type of normativity that is not easily defined in other terms as it belongs to some basic categories (Moore 1903). Sometimes, „good“ might also appear in the sense of „instrumentally good“ or „useful“. Whenever this happens, it should be clear from the context. I will not take position in these articles between moral objectivism and relativism (though I tend to the first). For my purposes, it suffices to refer to the historical and cultural practice of medicine and its internal values. Like Alasdair MacIntyre (2007), I see however reasons to take medical practice for an excellent example of an objectively good practice. I will not defend that here. Ethics is understood as the philosophy of morals.


Moral emotions (or sentiments) are a wide field. In this series, I am not dealing with those emotions that refer to moral judgments, such as guilt and remorse, and I abstain from a stance on the role of emotions in ethics in general. I do, however, take position by regarding morals not exclusively as a means to restrict emotions. I take for granted that there are morally adequate emotions, and indeed that it is an impoverished view on ethics reserving it to purely rationalistic judgments.


In his later work, Slote tries to found morals generally on empathy. That is a very different project from mine (Slote 2007).


Wilmer emphasizes that this full emotional identification doesn’t imply good understanding: it neglects the unique history and experiences of the other persons. For understanding, the more appropriate way of identification is the empathic “feeling as if I would be he”. (See below).


That is the reason why physicians are warned to treat close relatives, friends, or even themselves.


In the concept of empathy in medical practice, the borderline between epistemology and ethics becomes diffuse. Being defined as a mainly (not thoroughly) cognitive way of understanding, empathy appears to be a merely epistemic tool. As a necessary part of medical practice, it is at the same time subjected to moral analysis and judgment.


The use of the term “feelings” and “emotions” is varying: sometimes they are used synonymously; often feelings are understood as a broader category that includes emotions as specifically directed feelings; or emotions are taken for the objectively observable phenomena of which feelings are the subjective experience (Damasio 1995), etc. For my purposes a clear distinction is not necessary, but as I focus on directed, conscious affective states that can be influenced in a desired way, I usually prefer the term “emotion”, as including the subjective experience of it.


A thorough deliberation about rational emotions in medical practice can be found in Maier and Shibles (2010), pp. 137–160.


Howard Spiro illustrates it this way: “We doctors are selected by victories: We reached college because we were bright and competitive in high-school, and we reached medical school through competition and hard-edged achievements. (…) No wonder we have little empathy for the defeated, the humble, the dying, those who have not made it to the top of the heap, and even for the sick.” p. 844.


Aring is one of the authors who support an attitude of „detached concern“ for doctors; he warns of losing the objectivity and capacity to act by indulging in emotions. Jodie Halpern recommends empathy as a remedy for too much distance which is, according to her, the more common problem of physicians. Nonetheless also Halpern recommends a balance between nearness and distance (Halpern 2001).


In fact, there is much more for Cassell in understanding the patient as a person than empathy alone. He reminds us of many different aspects in which a person can be inflicted by an illness (Cassell 1982), he emphasizes the meaning of the patient’s narratives and the importance of his values (Cassell 1991). But he always describes the attention to the emotional and subjective side of the patient as competing with the scientific, medical knowledge and technical know-how, and the main task of a physician as a person to unite two opposite poles.


This goes very well with the current neuropsychological findings about psychopathic delinquents where different brain areas are brought into connection with cognitive and emotional aspects of empathy as a precondition to really understanding the implications of the own deeds for others. (Damasio 1995) In a useful review Grit Hein and Tania Singer summarize: “Accumulating evidence has put forward the view that there are at least two different ways to put us in the shoes (the mind) of the other person. One route is to share the other person’s feelings in an embodied manner, known as empathy (…). The other route is to cognitively infer about the state of the other person, known as ‘theory of mind’, ‘mentalizing’, ‘mind-reading’, or ‘cognitive perspective-taking’. Although often occurring in concert, findings from functional magnetic resonance imaging (fMRI) studies suggest that understanding others on the basis of cognitive perspective taking and empathy recruit different neural networks”. (Hein and Singer 2008, p. 153) Mark that the ‘empathy’ concept in this article is different from our definition which encompasses both emotional as well as cognitive aspects, i.e. the full capacity to “put us in the shoes of another person”. (See also Bauer 2005) In fact the difference between empathy and emotional contagion (which is made by Hein and Singer) is not easily to detect if awareness and cognitive modulation (see Lamm et al. 2007) are taken out of the concept of empathy.


It is important to keep in mind that this special meaning of empathy is directed on the clinical encounter between physician and patient, in order to safeguard the capacity of the physician to help efficiently and not to lose himself in pity and compassion. In other contexts, the adequate extent of empathy may be much more emotional and nearer to “feeling-with”, of emotional resonance.


As the psychiatrist deals with very special inner states, namely pathological ones, he also may need a qualitatively different kind of empathy. But the transfer from normal over exceptional to unequivocally sick mental states is even more fluent than in somatic contexts. So the qualitative step in psychiatry might also result from the overlarge quantity. Another example for the context-sensitivity and necessity of empathy is veterinary medicine. Here empathy must work beyond species limits and without help of language. The life of a veterinary surgeon who lacks this capacity is short, and his successes will be few.


This means in no way that empathy is only an inborn talent that cannot be taught or learned; on the contrary (Platt and Keller 1994). According to Aristotle, virtues that are important elements of character can and should be learned and fostered, and habituation to right behaviour will lead to a better character on the longer run (Aristotle 2002).


In chapter II of his „Theory of Moral Sentiments“, Adam Smith emphasizes the „pleasures of mutual sympathy“ (sympathy in a meaning very close to contemporary „empathy“). Though at first view it seems to be the sharing of the same emotions, it becomes clearer and clearer that it is the combination of understanding and the kind attitude that is the source of the pleasure (Smith 1759, pp. 12–18).


„While empathy is believed to have high social value, in actuality it is a neutral human facility whose value is derived from the manner in which it is used“ (Zinn 1993, p. 309).


In order to be precise, I must admit that it depends on the theory of morality: a strictly consequentialist approach implies that the bad result does make the helpfulness in this case a part of bad agency. But even a consequentialist, if he scrutinises the general attitude of helpfulness, probably would come to the conclusion that, generally, helpfulness leads to morally good results. So it is not the attitude that makes the act of helping bad, and as it normally leads to good consequences, it is rather a good attitude, though it is not decisive for judging the whole act.



I wish to thank Bettina Schöne-Seifert, Münster (Germany) Sabine Müller, Berlin (Germany) and particularly Rolf Ahlzén, Karlstad (Sweden) for reading and commenting earlier versions of this article. I am also indebted to the teams of the Institute for Ethics, History and Philosophy of Medicine in Münster (Germany), the Department for Health and Society in Linköping (Sweden) and the Centre for Applied Ethics in Linköping (Sweden). Special thanks to three engaged anonymous reviewers. I am responsible for the remaining mistakes and lack of clarity, of course, and I hope they will be reason for further discussion.

Copyright information

© Springer Science+Business Media B.V. 2011