Medicine, Health Care and Philosophy

, Volume 15, Issue 4, pp 397–410 | Cite as

The desired moral attitude of the physician: (II) compassion

Scientific Contribution


Professional medical ethics demands of health care professionals in addition to specific duties and rules of conduct that they embody a responsible and trustworthy personality. In the public discussion, different concepts are suggested to describe the desired implied attitude of physicians. In a sequel of three articles, a set of three of these concepts is presented in an interpretation that is meant to characterise the morally emotional part of this attitude: “empathy”, “compassion” and “care”. In the first article of the series, “empathy” has been developed as a mainly cognitive and morally neutral capacity of understanding. In this article, the emotional and virtuous core of the desired professional attitude—compassion—is elaborated. Compassion is distinguished from sympathy, empathy and pity. Several problems of compassion as a spontaneous, warm emotion for being a professional virtue are discussed: especially questions of over-demand, of justice and of concerns because of a possible threat to the patient’s dignity and autonomy. An interpretation of compassion as processed and learned professional attitude, that founds dignity on the general idea of man as a sentient being and on solidarity, not on his independence and capacities, is developed. It is meant to rule out the possible side effects and to make compassion as a professional attitude and as professional virtue attractive, teachable and acquirable. In order to reach the adequate warmth and closeness for the particular physician-patient-relation, professional compassion has to be combined with the capacity of empathy. If appropriate, the combination of both empathy and compassion as “empathic compassion” can demand a much warmer attitude towards the patient than each of the elements alone, or the simple addition of them can provide. The concept of “care” that will be discussed in a forthcoming article of this sequel is a missing necessary part to describe the active potential of the desired moral attitude of the physician more completely. The reconstruction of the desired professional attitude in terms of “empathic compassionate care” is certainly not the only possible description, but it is a detailed proposal in order to give an impulse for the discussion about the inner tacit values and the meaning of medicine and clinical healthcare professions.


Clinical ethics Empathy Compassion Dignity Care Detached concern Virtue ethics Moral motivation Professionalism Moral attitude 


In the search of the right attitude of physicians—that one that might be demanded in professional codes both of the past and present—I have examined the concept of “empathy” in a foregoing article. I have conceptualized it as a necessary cognitive-emotive precondition for the right attitude and as a necessary means, but not as the right attitude itself. Another promising candidate is the concept of “compassion”. The Ethics Committee of the American “Society of Academic Emergency Medicine (SAEM)”states:

“Compassion is a part of professional competence and is perhaps as important as technical competence, because both are required to effect meaningful healing.” (Adams et al. 1996, p. 964)

Though usually not explicitly reasoned, all professional codes of ethics for physicians include elements that are focused on the person (the character) and the intentions of the agent rather than only to the rules of conduct and the protection of the patient. It is the wish to understand more closely the content of these demands that has inspired the following deliberations of whether compassion could be the moral attitude in question, and in which way it can be interpreted in order to understand it as a professional virtue. Some clarifications about the notions of virtues, character, attitude and emotion are necessary before really approaching the scope of “compassion”:

A virtue is understood here in accordance with Alasdair MacIntyre as an acquired trait of character which tends to enable us to achieve the goals of a certain practice with excellence (MacIntyre 2007, p. 191). In the case of medical professional virtues, the respective goals are the intrinsic goals of medicine—one paradigm case of MacIntyre.1 I will understand the central goal of medicine as to help sick people (in itself a paradigm case of a morally good action), so we can, here, do without the broad discussion of whether moral virtues in general are only instrumentally or inherently good. With regards to medicine I regard them as inherently good, which does not imply that they are sufficient to make an action morally good.

A character is the stable core of a person, not unchangeable, not expressed in every utterance or action, but it describes how a person tends to be, to act, to value—and also to feel. MacIntyre (2007, pp. 204–225) describes the character as narrative personal identity—a perspective with several advantages, as it can easily be related to moral traditions this way; but I prefer a more ontological idea of a person, even if no person is imaginable without a context, a history, social bonds etc. A person also is unimaginable without a body, a brain and a psychological basis which have crucial influence on the character. From a moral point of view, however, it is more remarkable that the character, though a stable centre of a personality, can be altered to a certain degree by training, education and maturation (Aristotle 2002). Therefore, it is not only possible to judge a person morally, but also to educate her and to put certain demands on her self-education.

I understand an attitude also as a stable part of the personality, though not as resistant to changes as a trait of character. The more important difference between an attitude and a virtue is, however, that the focus on a virtue means an emphasis on the agent, whereas the attention to an attitude focuses on the relationship of the agent to something or someone. Attitudes can be cognitive, emotive or (often) a combination of both. Usually, they imply a valuation, though it is also possible to have an explicitly neutral attitude towards something. That is, an attitude is either positive or negative, or it refuses to take a position. It is not disregarding of valuations.

Pellegrino and Thomasma have pointed out the importance of virtue ethics for the clinical self-understanding of physicians as well as for a morally sound interaction with patients (Pellegrino and Thomasma 1988, 1993; Pellegrino 1995). They have listed several virtues (e.g. fidelity to trust, justice, fortitude, temperance, phronesis, etc.), including more emotional ones (also compassion), but they have focused on the cognitive parts of decision-making. Though Aristotle, the very founder of virtue ethics, can be read as if emotions are what have to be conquered by virtues, modern virtue ethics has shown that this impression arises if one chooses bad emotions as example (Foot 1990). From a logical perspective, purely rational deliberations have no practical outcome at all, but only hypothetical relevance (Foot 2002, p. 78). Current findings in neuroscience underline the importance of affective intentional states and their irreducibility to merely cognitive ones (de Sousa 1987; Damasio 1994; Goldie 2002; Helm 2001; Slaby and Stephan 2008). It is outside of the scope of this paper to go into the complex discussions of action theory and philosophy of mind in order to conceptualise the role of feelings and emotions in human agency. Suffice it to say that—no matter how exactly they are linked and which aspect is guiding in which situation—affective as well as cognitive aspects are important in the motivation of human behaviour, and in contrast to many ethical deliberations I will focus on the affective sides of virtue ethics.2 This is no commitment to moral emotivism, only of being persuaded that moral sentiments3 (though they may be not decisive or alone decisive for ethics) are a non-negligible part of moral agency which may take their place in virtue ethics.4

For the purpose of testing the scope of the concept of “compassion” as professional virtue for physicians, I will first distinguish it from “empathy” and “sympathy”. In comparison to the concept of “pity”, some problems inherent in “compassion” will arise. These problems will lead us to the question of whether compassion is desirable at all and for whom, and which elements of it are morally undoubtedly good and which are not. I will distinguish compassion as spontaneous emotion (CSE)5 from an interpretation of compassion as professional medical attitude (CPA) that is meant to remediate the unwished-for aspects of compassion and make it suitable as a professional virtue (CPV). The professional attitude of compassion still is regarded as an emotional attitude, but different from spontaneous compassion: calmer and influenced by the goals and duties of medical practice and relating to an idea of man as a vulnerable and solidary being. By integrating the professional attitude into one’s character, one has developed the virtue of professional compassion. There will be some remarks about if and how compassion as professional virtue could be learned and taught, moving from (CSE) over (CPA) to (CPV). Finally I will give an outlook to the more complete attitude of “empathic, compassionate care” that I propose to recommend in professional moral guidelines instead of the existing more unclear and unnecessarily demanding appeals.

“Compassion” between “empathy”, “sympathy” and “pity”

As all notions that refer to the emotional side of the attitude of healthcare professionals also “compassion” is an oscillating concept. It is sometimes used synonymously with sympathy, empathy, commiseration, feeling-with, pity, care, etc. The differences between these notions—if they are seen at all—don’t give a clear tendency in one direction: for one, sympathy is what for another is empathy, or compassion, or care. In order to go deeper into the moral impact of this family of words, it is inevitable to propose some conceptual clarification. From its linguistic Latin background, compassion means “suffering with”, exactly as the Ancient Greek sympathy. In the current use, however, sympathy often is used much more symmetrically than compassion, but less exclusively focused on suffering. As developed in more length in the first article of the series, I propose to understand sympathy as sharing the same feelings with the other, as far as this is possible for another person, while having a positive attitude towards him.6 Empathy in the clinical context, in contrast, can be seen as 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 (See previous article). Empathy can be accompanied by a positive, but also by a neutral or even malevolent attitude.7 Compassion, as an immediate, still non-processed affect, could be understood as a certain emotional response to the experienced suffering of another person. Characteristic for a compassionate reaction is
  1. (1)

    recognition of suffering

  2. (2)


  3. (3)

    a feeling of being personally addressed and

  4. (4)

    an inclination to relieve the suffering.


Ad (1) Certain recognition of the situation of the sufferer is necessary; otherwise the response could not be evoked. However the extension of recognition has not to equal the one implied in empathy. A compassionate reaction can be misplaced (because of lacking understanding) and still remain compassionate, while misunderstanding empathy would be an oxymoron.

Ad (2) On the other hand, malicious compassion is hardly imaginable, while empathy might be used with bad intentions. Thomas Szasz uses this as an eye catcher in his provocative title “Cruel Compassion” (Szasz 1994).8 Arguing (as usual) against compulsive psychiatric treatment and hospitalisation of psychiatric patients, because he regards social harms and misinterpretations as the reasons for the label “psychiatric disease”, he uses the benevolent implications of compassion to contrast them with the malevolent adjective “cruel”. This only is no real contradiction in words because cruelty does not only depend on evil intentions, but also on the effect of the action. Cruelty as an attitude necessarily encompasses bad intention, and therefore it is impossible to have the attitudes of cruelty and compassion at the same time (Comte-Sponville 1996, p. 144). As an effect on a sentient being, however, cruelty can also result from indifference, or even—as in this case according to Szasz—of good but misdirected intentions like compassion, evoked by the mistaken label of disease. Compassion, on the other hand, is unequivocally bound to the intentional side of an action. If someone interprets an act of help as compassionate that has not come out of the attitude of compassion, he is simply mistaken. If compassionate behaviour—against its intentions—does not help, it is unsuccessful but not uncompassionate. Thus, benevolence is a decisive element of compassion, but compassion is not identical with or easily replaceable by the concept of benevolence, as it is a special kind of benevolence. Though “good will” embraces by definition a “good” element, it is not sufficient for morally good agency.9 Compassion encompasses valuable elements in addition to benevolence: the identification of a bad situation (1), a more personal feeling of responsibility, of being addressed and demanded (3) and a clearer focus to the one who is in need with an inclination to do something to help him (4). Admittedly, compassion is as well not enough to ensure good agency; not even sufficient to describe the complete desired morally-emotional attitude we are looking for. In comparison to benevolence, however, it is more encompassing as well as more specific, and thus better suited for the reconstruction.

Ad (3) Compassion involves the person of the agent. It is more than just feeling “this person should get helped by someone”. I feel moved by the suffering of the other. His need addresses me in a way that I feel responsible to react, at least emotionally. An emotional relationship to another person arises.

Ad (4) Unlike sympathy, however, compassion is no affection that aims at symmetry; the asymmetry of both persons involved is an essential element of this concept.10 I do not expect of somebody for whom I feel compassion to answer with the same feeling or to respond at all. Compassion is already the answer to a need; it does not require an action or reaction from the other person. The compassionate person is the one willing to help the other; the other one is perceived as needy and perhaps as weaker. If we want to emphasize this asymmetry, we often speak of “pity”. In pity, the aspect of distance is emphasized. Pity still is well-meaning, acknowledging the misfortune, feeling the inclination to help. As H. A. Wilmer puts it: “Pity, at best, is compassion; at worst, contempt.” (Wilmer 1968, p. 244) Pity can also imply a feeling of “Thank god, it’s not me” or even “This could never happen to me”. Very often it entails a bit of self-righteousness, of enjoying one’s own superiority and benevolence. That is the reason why, usually, it is not very attractive to be a recipient of pity, to be pitiable or pitiful. While in the case of compassion one’s own feeling of doing-well and accordingly of one’s own superiority may be much weaker, still the asymmetry persists as an element of the motivation to help. Therefore, compassion always implies a danger of a condescending attitude towards the one who suffers, or at least the latter could fear this or perceive the attitude as such. If we try to get totally rid of the asymmetry, however, we also lose the inclination to help, the motivation to act. Compassion lets me see myself as addressed helper, it asks for my engagement. Then again, compassion is not yet a direct impulse to help actively. I could respond to the other’s need only with my compassionate attitude and the wish to help, without getting active. This is the reason why for the full reconstruction of the desired professional attitude we will also need the concept of care, which I will present in the third article of this series.

Compassion as professional virtue

Thus, is compassion a desirable attitude at all? Is it desirable for health professionals? Is it a virtue? Of course, the answer to these questions depends crucially on what exactly is meant by “compassion”. If it is—as I have described until now—a spontaneous, emotional benevolent response to the suffering of patients (CSE), it seems at first view desirable to have compassionate healthcare professionals around if one is ill. They will be personally engaged and willing to help one, understand that one is not well, and they will not expect something from the patient in return. They will not look at the watch when a patient needs consolation and will not use the back stairs in order to avoid contact with weeping relatives. As long as patients do not feel humiliated by the asymmetry of the attitude it is certainly good and helpful to have compassionate doctors and nurses, and there are hints that the treatment of the patients is better if a compassionate attitude is involved (Mills et al. 1994, p. 584). Certainly, it is not a vice for a healthcare professional to be compassionate. On the other hand, it is less clear if it is in fact desirable for the healthcare professionals themselves to have an attitude of compassion. There are justified concerns that a compassionate person might become burnt out, or at least more liable to place excessive demands on themselves (Benner and Wrubel 1989). For if compassion is perceived as a professional ideal, but cannot– as a spontaneous, warm and personal feeling—be produced at will, and not towards every patient, responsible healthcare professionals will tend to feel overstrained. Too much emotional involvement is demanded all the time, qualitatively as well as quantitatively. This danger increases if compassion is understood (different from our conception) as a sympathizing, complete “feeling-with” the patient, even though tempered by an attitude of “equanimity” if the suffering cannot be diminished, as Patrick Boleyn-Fitzgerald suggests (Boleyn-Fitzgerald 2003). Continuous nearness, even to the severely suffering or dying person, is only feasible if the health professional has an equanimous attitude to her own possible suffering and death, otherwise she will have to draw a line between the patient’s and her own fears. This seems, to say the least, an unrealistic demand to make of healthcare professionals; it may even have a bad effect and overstrain them or force them to be untruthful about their emotions. Is it hence better to do without these demanding emotions, and to strive for a generally sober and unaffected state of mind? In the stoic philosophy it was regarded as better not to feel compassion, because that adds sorrow to sorrow, but rather to help without affective reaction.

However, as Nancy Sherman has shown in her work on stoic virtues for soldiers, even the Roman stoic sages were not untouched by the death of their near and dear. Sherman argues that for a mild, humane stoicism; emotions like friendship, love and compassion are no menace to resilience, but on the contrary necessary for it, because the denial of them would make a person unprepared and more vulnerable (Sherman 2005). Adam Smith has argued convincingly (on behalf of sympathy, which can be understood as a mixture of empathy and compassion in this context), that sharing grief and sorrow does not aggravate but alleviates them (Smith 1759, ch. II) and the French philosopher André Comte-Sponville points out that it is better to feel sad love (i.e. compassion) than happy hatred. However, if compassion is understood as a limitless emotional closeness to be shared with all patients all the time, that only can be sustained by an incredibly fearless attitude towards suffering and death, this seems not to be the desired attitude of healthcare professionals.

Also, it might pave the way to a revival of the image of the physician or nurse as a supernatural being, a demi-god who is morally superior in essence. The following quote from Ferrel illustrates an extreme position from an extreme situation, but is included, since more nuanced forms of it occur in the nursing literature.

“A powerful contribution to feminist scholarship is the book by Melissa Raphael […] “The female face of God in Auschwitz” (Raphael 2003). […] Raphael describes the touch of these women and their attempts to cleanse and comfort the dying and suffering as kindnesses of an ethic of care but also symbolic of women’s restoration of the human and, therefore, the divine. Perhaps nurses are the face of god in pain.” (Ferrel 2005, p. 88)

Of course, I do not want to mock heroic efforts to hold humanism high in unbearable circumstances, nor heroic self-experiments or self-sacrifice in the face of dangerous infectious diseases, for example. But it would overcharge the concept of compassion to put the ideal this high, and it could lead to overestimation of the own professional role and to forgetting that the asymmetry in the situation is due to the patient’s vulnerable situation, rather than to one’s own moral superiority. This self-perception may be in conflict with a genuinely moral attitude and behaviour.

“And a romanticised description of a nurse’s practice that does not correspond to actual interaction between nurse and patient can aggravate the situation. It can mean that nurses’ actions are put into words in a new way, and that nurses have an idealised idea of themselves and their practice that does not correspond to their patients’ experience. […] A romanticised idea of nursing practice can result in worse patient care, not better, and thereby violate patients’ human dignity.” (Hem and Heggen 2004, p. 28)

It is hardly necessary to mention that this applies to all healthcare professions.

In addition to the danger of hypocrisy and self-adulation, compassion also challenges the ideal of good care in another respect: as a spontaneous direct emotion it probably cannot be directed to all patients to the same extent. There are factors that enforce the eagerness to be compassionate towards someone, as for example young age, severity of the disease, a sympathetic or worthy character, similarity of any kind to the compassionate person.11 If compassion is a necessary or at least important and effective part of good healthcare, how can it be justifiable not to give it to all patients to the same degree? Even considering that–perhaps–compassionate behaviour may be imitated and serve as an example (Schantz 2007, p. 50), a just and equally good treatment of all patients seems to be rather endangered than ensured by compassion. Feminist ethics often emphasizes the particular and relational aspect of ethics, (Fry 1989; Jecker and Self 1991; Carse 1998) and authors from nursing sciences feel endorsed by articles like these in their support of nurses’ advocacy of the patients (Winslow 1984). However, the awareness of a personal, relational and motivational side of ethics cannot legitimate a neglect of other persons who are in the same relationship as the one I feel responsible for. Many strains of feminist ethics involve moral emotions, take them seriously and reject a superiority of abstract and generalized rules and principles, but even they do not rely solely on immediate spontaneous feelings of sympathy (as opposed to reflected feelings). Nobody claims an obligation to give preference to the most loved child over its siblings, for example; and concentration on one patient who has aroused a caregiver’s compassion at the expense of all others might be a problem for the concept of compassion. If compassion is the right attitude of a healthcare professional towards a patient, all patients who need and want it should profit from it. Feminist ethics is not the appropriate background to rationalize treating adequately only patients who are lucky enough to have accessed my immediate compassion. Thus, considering all desirable and morally good aspects implied in compassion, there remain three doubts about its desirability as a professional attitude: the fear of overstraining, of condescension and of injustice. In the following two sub-chapters I will address these concerns and propose an understanding of compassion that avoids them.

Prescribing emotions

Generally, it does not seem to be prudent to prescribe certain emotions for any kind of role, not even if it is usually strongly linked with it (like love for a parent). If the desired emotion is not there on its own, the demand of it can only lead to devaluation of the agent, but not to an improved attitude or behaviour. A parent who does not really love his or her child still might be a dutiful parent,12 and though we might wish that he or she would rather love the child (and that would make the tasks of dutiful care easier by far), it would simply be useless or even harmful to demand this. It seems to be similar for physicians and nurses. Though we might prefer persons as healthcare professionals who are talented for compassion and feel it rather often for their patients, it is useless to demand it in the cases when it is absent. As a source of reliable moral behaviour, it seems to be much too arbitrary and contingent.

Therefore, though compassion for a patient usually will be no vice and may be of use for him and give motivation to healthcare professionals, and so encompasses many morally good aspects, it seems very problematic to recommend it as a professional virtue or a moral demand. Admittedly, virtues do not demand to act always, without exception, upon them. They describe a character that typically acts in the desired way. They are an ideal that can be reached by getting ever more used to a morally good behaviour. So we could very well argue that we start with showing compassionate (as morally good) behaviour to the persons we feel close to in this way, and try to get so used to it that we also can—step by step—be compassionate to all others. By imitation and “acting as if”, the virtue of compassion could be acquired, learned and even taught (Aristotle 2002). André Comte-Sponville emphasizes that though having a feeling cannot be a duty, there can be a duty to cultivate it, as feelings are neither passive nor unchangeable and determined by fate (Comte-Sponville 1996, p. 148).

However, is there a way to keep the valuable aspects of the emotion of compassion (CSE) for a professional ethics and get rid of the problematic implications? I think there is if we do not think of compassion as the usual spontaneous unprocessed emotion. The way of developing gradually a general virtue of it may be a step on this way. But I suggest not aiming at feeling the spontaneous emotion of compassion for every single patient. Rather, it would be advisory to internalise the moral goods that compassion shows us and motivates us to act upon: We all generally recognize the bad and unwished-for in the state of illness, and compassion is a response to this and inclines us to do something against it and help. I think it is no excessive or extremely difficult demand on healthcare professionals to orientate their doing and the meaning of their profession in this direction. Compassion in this meaning (CPA), as a general attitude of a healthcare professional to feel the inclination to help suffering people, can very well serve as a general moral value and guideline that can and ought to be demanded as the right attitude, directed to the central goals of medicine. If the spontaneous, warm feeling of compassion is not demanded in every case, it does not seem to overstrain the capacities of average persons to internalise this attitude. Even extremely unsympathetic, disgusting or vicious patients can be treated with compassion in this sense (CPA), as it does not necessarily imply a warm emotional closeness and personal sympathy. As desired professional attitude, it is not necessarily supposed to be the only or the strongest emotion a physician should have, but it should be that emotion that guides his professional behaviour.

Compassion (CPA) is no very passionate (or in terms of Hume “violent”) but a rather calm emotion (or “passion of the soul”). Calm passions “produce little emotion in the mind and are more known by their effects than by the immediate feeling or sensation.” Therefore they are easily “confounded with reason by all those who judge things from the first view and appearance.” (Hume 2004, II 3,3)13 According to Anthony Kenny, the imperceptibility of these calm emotions may be invisible for the one who is moved by them, not for the observer (Kenny 2003, p. 15). The question at hand is if it is necessary at all to look for an emotional basis. Does it not suffice to demand a stability of adequate behaviour towards the patients, rather good manners than a certain emotional background? The answer may be found in Kenny’s deliberations about the explanatory strength of emotions as motives of behaviour:

“The kisses of Judas need more explanation than the kisses of Romeo” (Kenny 2003, p. 65)

Certain behaviour gives good indications of an emotional background, even if this is not infallible. Typically, kisses indicate love. If they are used in order to hide other feelings and intentions, i.e., if they are inauthentic, it needs extra explanation. The kind and caring behaviour of a physician is not only desired because that is nice for the patient. It is taken as a sign for an authentic attitude which underlies the trust of the patient. It may be difficult to distinguish it from an only faked attitude, but it will be most reliably shown if it is really internalised; and even if a doctor never fails in showing the desired behaviour but only lacks the underlying attitude something would be missing in the image of a good doctor (Gelhaus 2011a, b). Thus, though it may be hopeless to prescribe and reliably test an emotional attitude, we may still wish it to be there, and regard it as the adequate and authentic background of the behaviour we demand: the usual one, which does not demand further explanations like “being obedient to rules” or “believing that this contributes to professional success” or even “hiding the underlying cynical and selfish attitude for fear of bad consequences”. Being an irresponsible cog in the wheel, being rather ambitious or greedy than wishing to help the patient, despising the patients14—these are typical reproaches of being a bad doctor, to be watched in all kinds of TV films. Being a good doctor requires the opposite, and having the fitting emotional attitude makes the right behaviour an authentic behaviour. Having the right kind of attitude which is implied in the inner logic of one’s tasks bears an important advantage: it implies appreciating the inherent values of one’s doing, and understanding and endorsing the inner meaning of one’s profession. As such, compassion as an authentic professional attitude can protect against over-demand and burn-out, as the demands correlate with one’s own values and convictions.15 It is crucial, however, to demand an emotion that cannot only be reached by saints (like parental or passionate love for all humans). Spontaneous, warm, personal compassion (CSE) for each patient is unrealistic, the kind of compassion implicit in medicine (CPA) as internalised recognition of suffering and motivation to help is not. As a state of mind, it does not cost extra time or spare resources, and it is the adequate reason of trust which improves the patient’s compliance. Therefore, it has the potential to use the resources more effectively and avoid unnecessary double investigations and avoidable side effects. On the other hand, it might bring the doctor in conflict with social and organisational shortcomings and prioritisation with a focus different from the individual patient. I will address these practical consequences a bit more in the third article on “care” (though also not comprehensively). I see however good reasons to keep the traditional ideal that physicians ought to have a privileged attention to and responsibility for suffering patients, even if that has to be balanced with other duties, demands and incentives.


Still, one serious doubt about the moral value of compassion seems to be untouched: the asymmetry and its possibly condescending implications. There is no chance to guard the morally good aspects of compassion and at the same time to make it something symmetrical that focuses on direct mutuality. On the contrary, compassion is a concept that conquers precisely this basis of morality and gives an alternative orientation for situations that are not and cannot be reciprocal. The German philosopher Käte Hamburger in her encompassing study on compassion also comes to this conclusion (Hamburger 1985). According to her, this is the reason for understanding compassion as a morally neutral feeling which in essence distances the seemingly altruistic person from the other (compared to love). She elaborates convincingly how disappointing the attitude of compassion instead of love is in a lover (or a spouse).16 The situation between physician and patient is however essentially different from the one of lovers. It is deeply asymmetric, and it is a meeting not of relatives, friends or other ‘near and dear’ persons, but of strangers. It is the patient who suffers, who is anxious and existentially endangered. He is the one who needs and wants help. The physician, on the other hand, has got the available knowledge and measures to help, and he is in his usual professional setting with no unusual personal involvement.

In many situations where persons meet, respect and justice and appreciation depend on their equality—not factual equality, but equality of rights and entitlements. Most ethical theories rely on some kind of equality as basic pattern, of which one of the most prominent examples is the foundation of human dignity on the capacity of autonomy by Immanuel Kant. For these theories, a basically non-reciprocal situation as the usual one in the medical encounter is a challenge. There are several methods to ensure a morally good behaviour of the stronger one in this situation, for example the emphasis on patient’s rights and patient’s autonomy, the focus on rules, codes and principles and the like.17 The concept of compassion as a professional virtue is one of these approaches that works not only in spite, but exactly because of the asymmetry of the situation. It is only problematic if it hurts or could hurt the dignity and autonomy of the patient. The danger for this is bigger if the spontaneous emotion of compassion (CSE) is concerned. If compassion is no emotion for one specific patient but a general attitude towards each patient that might appear in this typical situation, the chance to conceive it as personally humiliating is accordingly lesser. In fact, compassion as a general attitude that founds the basic setting of the health care system (CPA) relies on an idea of man that is different from the free and unrelated autonomous, rational being we are used to think of since the Enlightenment. Instead, human beings are seen as sentient and possibly suffering beings, all of them. This constitutes a general neediness and relatedness on which society has answered with the healthcare system in order to guarantee help.

“The basis for every kind of help is the acknowledgement of the fundamental features of human existence: vulnerability, dependency, fragility and mortality. […] Dependency and vulnerability are what make a person human, and care is directed towards these fundamental aspects of the human condition, about which we have no choice.” (Hem and Heggen 2004, p. 21)

This institutionalised compassion (CPA) does not create a symmetric situation between patient and physician; it does neither make the patient a partner of the healthcare professional, nor a normal customer who chooses freely from some offers.18 It equalizes in another sense that is in minor danger of disregarding the considerateness of the patient in his special situation, though it still safeguards his dignity and the respect for him as a person. Not in spite, but because of exerting compassion in this generalised sense, the dignity and social affiliation of the patient is confirmed (See also Teuber 1982). He profits from compassion not because he, individually, is weak, but because he is a member of a society of vulnerable human beings that cares for this typical neediness. Therefore, compassion does not endanger, but ensures the respect of the patients’ dignity:

We care for people in crisis who are exceedingly vulnerable to humiliation and harm at the hands of caregivers who lack compassion. (Adams et al. 1996, p. 964)

The typical, asymmetric situation of a physician face to face with a patient therefore is not something to be overcome or reinterpreted in another way, but it ought to be accompanied by the appropriate attitude of respect towards a sentient and suffering being with his or her human dignity, no matter how sympathetic the individual patient may be or may not be. Of course, it is not forbidden to feel specific, spontaneous compassion (CSE) at times, it may give life and fresh motivation to the professional, and even fill the moral ideal with new energy. The moral agency however is not diminished if the compassionate action is done without this personally directed feeling. On the contrary, one must even be able to act in a compassionate way (CPA) if there is a deep antipathy towards a patient. This much of justice must be demanded of a morally valuable professional attitude that should guarantee that the individual person of the patient is helped, regardless of his merits and his social status.

Closeness and warmth

Another remark is necessary on this rather detached idea of compassion (CPA). An important function of the feeling of compassion which is often demanded especially in the nursing literature is to ensure the individuality and personality in the treatment of the patient. Compassion safeguards that the patient is not only seen as a number or an example of a certain state diagnosed according to the knowledge of scientific medicine. Is this function of compassion endangered by seeing it as an ideal of benevolence no matter to which patient? Each patient suffers from his illness in his own and individual way; this knowledge is not diminished by the acceptance that this applies to each patient. My motivation to act upon this may be more genuine by the spontaneous emotion (CSE), but the focus of a compassionate response on individual suffering should not be given up by its generalisation. Still, the degree of necessary compassionate nearness is debatable. Howard Brody strongly emphasizes the importance of being near to the patient:

“True compassion requires vulnerability. If I am going to suffer with the patient, I must at some level be open to feeling at least a part of the patient’s suffering, as a result, I am likely to be changed as a person myself. If this prospect frightens me too much, then I can erect psychological defences to protect me from that suffering. The old notion of “detached concern” is probably a good descriptor of that protective barrier. But if I erect that barrier, then I will have lost my ability to be a compassionate presence for my patient. Patients will experience the psychological wall behind which I am hiding, not the company of a fellow human who cares enough about them to remain despite the suffering. […] True compassion, which has the power to heal because it reconnects the sufferer with the human community, is not easy to achieve.” (Brody 1998, pp. 591–592)

At first reading, he nearly seems to demand an attitude of sympathy (according to my understanding), a symmetrical attitude of thoroughly sharing the feelings of the patient. Especially his distance to “detached concern” as a “psychological barrier” seems to indicate this. Looking closer, the sharing only is required “at some level” and of “a part of the patient’s suffering”. Elsewhere, Brody speaks of a duality of the physician (or being a bridge) between being a fellow human being and having access to superior knowledge (Brody 1998, p. 591). Thus, even Brody seems to see that a balance between nearness and objective distance has to be found. Elizabeth Latimer formulates:

“In essence, health professionals must be prepared to enter into a relationship of considerable professional intimacy and compassion with dying patients while encouraging and respecting their personal strength and individuality. The psychologic separateness of patient and health professional is maintained.” (Latimer 1991, p. 862)

And the Ethics committee of the SAEM states:

“[…] compassion is a mean between indifference and involvement to the point of ineffectiveness. It is possible to overidentify with the suffering and to become overly involved. Perhaps the more common problem is cold indifference.” (Adams et al. 1996, p. 964)

In order to act from an attitude of compassion, it seems important to act from the position of a fellow human being, as a person, not only as an incarnation of the professional role. Therefore the health professional has to take a certain risk of meeting the patient as a real person. It does not seem to be useful or desirable that she completely identifies with the patient. On the other hand, this required openness cannot imply that she reveals too much of her own state. The physician’s own headaches or personal problems are not in the focus of the clinical encounter. She meets the patient as a fellow human being, but at the same time as a doctor, not as an equal partner or friend. Her openness and integrity is a physician’s authenticity, not the one of a parent, a friend, or a lover. She can be altered and touched by the meeting with the patient, and in order to be authentic she must be prepared for this; she does not meet the patient as a medical robot. She meets him as a physician, which implies being a whole person, but in a specific relationship to the other person. It implies neither complete identification nor detachment, but a deeply internalised (role-associated) sense of benevolent inclination to help every single patient, or it could result in insufficient treatment or even rejection of less sympathetic fellow human beings.

Is this attitude of compassion (CPA) warm enough? I have characterized it referring to Hume as a rather calm emotion which makes it easier to demand it, learn it and apply it in a just way; but have I cooled it down too much? Can calm compassion, which I am prepared to give to each patient, move me to accompany a severely suffering patient with sufficient inner engagement? Compassion alone may be not sufficient, but if it is combined with empathy, the situation looks differently. Compassion recognizes needs and moves the doctor to react in a kind and respectful way in order to help. Empathy refines the recognition of the needs and adds a specific interest in the person in his special situation. Empathy, i.e. the skill of understanding the mental states of a patient with regard to his health needs, can tell us how much warmth and consolation a special person wants and needs—or how much professional distance in order not to intrude in personal spheres where such intrusion is perceived as humiliating or disturbing. If a physician has to give her patient the diagnosis of a severe chronic disease, and if the patient is in a vulnerable state of mind (which is usually the case in this situation) this may imply very much warmth and nearness, and a neglect of these can be hurting and disrespectful. For a gynaecologist empathy may help to keep a—respectful—professional distance touching very private parts of the patient’s life and body. On the other hand, it may be necessary to create an atmosphere of closeness and understanding in order to be allowed to touch and ask. The required warmth neither is implied in the calm emotion of professional compassion nor in the skill of empathy alone, but results from the combination of both, which may explain why all related concepts are so often mixed and confused in moral appeals to healthcare professionals. In an insightful evaluation of desired affects of young doctors from the 1960s, based on interviews with 46 interns, Morris J. Daniels describes the approved type of affective involvement and expression as follows:

“In its intermediate form, compassion is a mixture of impersonality and warm human interest. In the relationship between intern and patient, characterized by a minimum of uniqueness and particularity, there is still, ideally, a profound sense of the common-human.” (Daniels 1960, p. 260)

This is a good summary of the ideal state of mind (also for physicians in the beginning twenty-first century) that I have tried to explain in more detail, in terms of professional compassion, empathy, and solidarity with a fellow-being.

A virtue or even more than a virtue?

Conceived as the described generalised, attitude (CPA), compassion can and should be judged as a professional virtue (CPV), especially if it is combined with the skill of empathy. It is not only a desired and valuable attitude; rather it is implicit in the professional role and as such necessary for any good physician, nurse or other healthcare professional. It depends on anthropological premises that are constitutive for our healthcare system and other institutions of solidarity in our society. On the other hand, it is not clear if it is necessary to demand of healthcare professionals to have implemented this valuable and desired attitude in their character. Must a doctor be a compassionate person (CPV), or is it enough if he has a compassionate attitude (CPA) to his patients? When we arrive to the idea of how compassion could be learned (see next sub-chapter), it will seem that there is not too much of a practical difference here. It is hardly imaginable how a professional who works a lot of his time in a compassionate attitude should be untouched by this in his character. Nevertheless it is the authentic attitude (CPA) that we wish for the patients, and though compassion qualifies as virtue, i.e. as a character trait an admirable doctor-personality owns, it should be sufficient to ask for the attitude (CPA) in professional codes and recommendations, and to leave it open if the demands of professional life also have a good influence on the whole person and thus also to his private life.

Is compassion perhaps even more than a virtue? Is it a fundamental value source of our society, a basis for ethics in general? Though not being the main stream of ethics, there is a long tradition of founding ethics in compassion (Nussbaum 2001). In Buddhism, it is generally the attitude other beings ought to be perceived with. For Schopenhauer, compassion simply is the elementary basis of any ethics, because of its motivational power as well as because of its intuitive self-evidence as a guiding ethical principle (Schopenhauer 1977). For some phenomenologists, compassion is even more fundamental: the very means to get out of empirical solipsism and to take seriously the other in an equal way (Scheler 1970; Housset 2003). Though the importance of compassion as a moral element for the healthcare system hopefully has been shown, the role of it as fundamental element of any human and moral development and any ethical approach shall not be discussed here. However, one may fear that this overcharges the conception, at least as it is described here. Compassion alone certainly is not sufficient for an encompassing ethics, and though it should not be neglected as a moral element, its role as most important, fundamental source of morality is and has been contested.

Learning compassion?

Medical education is an important and complex subject on its own, and I will not go into detail about its functions, requirements, and desiderata. The reason for mentioning it here is to understand more closely the role of compassion as a professional virtue, in contrast to skills, rules, principles and knowledge.19 The French philosopher André Comte-Sponville in his “A small treatise on the great virtues” has presented 18 virtues that represent the most important admirable character traits (Comte-Sponville 1996). The first one—politeness—is rather a first developmental step for children in order to learn real virtues later. The last one—love—is what as a general ideal would make all duties superfluous and what would embrace the virtues naturally. Compassion, however, as a virtue and emotion on a medium stage between these two, is an ideal that cannot be directly initiated by will or order, but that still can be admired, be strived for, and be learned.

If compassion is a medical professional virtue, how should we make sure our physicians are sufficiently compassionate? Is there a need for a character analysis for selecting medical students on the basis of which we leave the less compassionate out? Or is there a way of acquiring compassion?20

Carlo Leget and Gert Olthuis, who even endorse Housset’s mentioned fundamental compassion-based approach of ethics (see “previous sub-chapter”) as a basis for medical ethical education, state:

“Since morality is based on human freedom and maturity, a portfolio build on compassion can succeed only in a climate of confidence and respect, where good role models are willing to invest in personal contact. Only then can the originary experience of compassion develop into a virtue that, as a moral attitude, is a part of the personal and professional identity of the physician.” (Leget and Olthuis 2007, p. 620)

Obviously, in teaching professional ethics, compassion (CSE) can serve as a starting point and has to be modified as a basic emotion and completed by several additional aspects in order to safeguard a desirable moral attitude of the physician. However compassion is no moral principle that can be intellectually recognized and applied to factual situations. The way of application is (even) more complex. To learn compassion we need examples (“role models”). It is not enough to watch them carefully and imitate them; they also have to be prepared to have contact with the students. A thorough ‘infection’ with compassion (CPV) works better in the interaction of teacher and student and is not independent of the persons and their character. The learning of compassion is no matter of knowledge,21 but of maturity and developing one’s character. That needs time and continuous training, it cannot be attained through one ethics course sometime during medical studies. Virtues, though they can be learned by imitation and training like skills, are rather the basis of medical skills. Their acquisition is not only a habit but a special kind of habit that transforms the person. They are a tendency to act in a certain way, but it is not an unconscious or inborn tendency, but one acquired with the end to fulfil the role tasks in a good way (von Wright 1963). Insofar, they include knowledge, but not knowledge of scientific facts and theories. Today, the attitudes and virtues of a good doctor are usually taught as a part of tacit knowledge, (Polanyi 1983; Turner 1994) transferred to the practice of medicine by the demands of the patients and of the healthcare system and hidden behind some rules and codes of ethical behaviour. Admittedly, learning knowledge about virtues is not the same as learning to be virtuous, but I can see no reason why the attitudes and virtues in question could not be identified and explicitly addressed. On the contrary, they ought to be explicitly known, because this knowledge opens the possibility to distinguish ‘good’ internalisation of the observed role from less than good or even harmful such. For example a caring and responsible attitude is rather frequently caricatured by a self-righteous and overly paternalistic behaviour that misses the crucial points: the attention to and interest in the patient (empathy) and the sincere wish to help him in his suffering (compassion). Alasdair MacIntyre shows how the inner rules and goals of a practice (like medicine) can and must be distinguished from and defended against the institution (e.g. a hospital) which necessarily has to mix goals inherent to the practice (directed to its meaning and inner logic) with external motivations and influences (MacIntyre 2007). This defence work as one of the major tasks of professional virtues is much easier if the central goals of the practice (according to Cassell (1982, 1991) helping suffering patients) and the virtues in order to pursue these goals (in my approach empathic compassionate care and responsibility)22 are explicitly and not only tacitly known. This, in addition to the knowledge of rules and principles, could be the task of an ethics course, while becoming compassionate is especially fostered by contact with and by example of good physicians and trained in the meeting with patients.

Returning to my initial question of choice or education of medical students, it seems plausible that there are persons more or less gifted to become a good physician and to develop the professional attitudes and virtues. Insofar as there is an image in the broader society (though not too explicit) what a good doctor is like, it is probable that those who are more gifted are already attracted (among others who respond to rather external motivations). Making the professional virtues more explicit might even improve the choice, and facilitate further education and (self-) development. People who are not capable of developing sufficiently the demanded attitudes should be excluded from this profession as well as those who are unable to learn sufficient knowledge and skills.23


I have tried to specify the morally valuable aspects that are implied in the spontaneous compassionate attitude (CSE) and the way in which it can be understood as a professional virtue (CPV). Compassion combines the recognition of a situation of suffering with benevolence and the motivation to help. Thus it combines many aspects that are desirable for a healthcare professional. Unfortunately, it also implies aspects that are problematic: mainly the difficulty of how to create emotional attitudes at will, and related to this the fear of over-demand or self-overestimation and injustice. In addition to that, the patient might feel threatened in his dignity and his autonomy by the asymmetry implied in the concept. I have suggested an interpretation of compassion not as the spontaneous emotion, but as an internalised attitude (CPA) towards patients based on a general respect of and solidarity with all human beings, exactly because of our shared vulnerability and finiteness. Compassion understood this way could serve as professional virtue and evade the problematic aspects of the spontaneous, warm and personal emotion. The emotional closeness professional compassion implies is that of openness and authenticity, combined with respect, rather than a more sentimental, passionate closeness. Acquiring this professional emotional attitude is not an additional exhausting demand, but it reconciles the physician with his tasks and their meaning and is thus a personal gain of satisfaction and a protection from burn-out. It is not a small amount of spontaneous compassion but a full engagement in a qualitatively modified, calmer emotion. Many positions on the professional ideals of detached concern, or the right balance between closeness and objective distance could be understood this way.

Empathy, understood as primarily cognitive skill to understand the needs of the patient should combine with compassion. Firstly, this ensures that the demands of the patients are realized adequately. Secondly, though empathy alone is conceptualised with little emotional and moral content, the combination of both empathy and compassion can describe the way of closeness and warmth that is desirable for a healthcare professional: a constructive and well-meaning tuning-in with the patient without getting lost in emotion, and a meeting of one person with another neither hidden behind role masks nor with the illusion of equally capacitated and equally strong partners.

What is still missing in the description of the desired, basic, emotional (or also emotional, not exclusively cognitive) attitude of the physician? It is the aspect of coming into practice, of activation and realization of the still too sterile state of mind. For this purpose, the concept of “care” will be introduced and interpreted in a forthcoming article. Care will be shown as embracing an activity as well as a benevolent attitude. Empathic compassion (CPA) will be understood as adequate underlying attitude of that kind of care that is expected from healthcare professionals, and so the triad of ‘empathic compassionate care’ will be identified as the desired morally-emotive attitude at the basis of the physician’s role.


  1. 1.

    There are several suggestions which goals medicine actually pursues (The goals of medicine1996; Nordenfelt and Tengland 1996). Sometimes they are connected with a position in a special contested question, e.g. euthanasia; in stating a special goal of medicine in helping patients to die in dignity (Miller and Brody 2001). The German physicians’ association (Bundesärztekammer) accepts the goals of alleviating suffering, prolonging life, treating diseases and improving health (BÄK 2011). I hold that these (sometimes conflicting) goals are not much contested and describe adequately enough the internal orientation of medicine. If we try to condense it even more to the very core, we will get close to the position of Eric Cassell. According to him, helping suffering patients is the central goal of medicine (Cassell 1982, 1991).

  2. 2.

    It depends strongly on the kind of ethical theory if there is room for this intentional state within the theory, or if it is understood as something outside ethics as a philosophical discipline. Virtue ethics as agent-focussed theory has room to integrate moral motivations (Slote 2001).

  3. 3.

    As moral sentiments I do not only understand the feelings of guilt, shame, remorse etc., that is, of retrospective evaluation, but also those that have a benevolent (or sometimes called “altruistic”) implication like love, friendship and compassion.

  4. 4.

    It is necessary to distinguish clearer between affects, feelings, sentiments, emotions, perceptions, sensations etc., in order to get an encompassing picture of the role of affective states in agency in general, and moral agency in particular (Nordenfelt 1974; Kenny 1963, 2003). For the purpose of identifying and describing those professional virtues of physicians that involve affective states, however, most of these clarifications can be postponed. So the term “emotional” is used in the wide sense as mental state as opposed to “cognitive”. “Emotion” , however, is understood in the narrower sense of an intentional affective state with an identifiable object (Kenny 2003), that is, a dispositional attitude (Wollheim 2000).

  5. 5.

    The point about “spontaneous compassion” is not its spontaneity but its specific emotional content—I do not exclude totally that also “professional compassion” could occur spontaneously, though typically it is the result of a successful professional socialisation. All presented interpretations of “compassion” understand it as a dispositional attitude with inclusion of an emotional element. As the attitude of spontaneous compassion is not quite what seems to be expected as a professional attitude, particularly with regard to its emotional content, I focus on these emotional aspects of spontaneous compassion, in order to make clearer in which way compassion has to be developed in order to become a professional attitude.

  6. 6.

    “Identification” would be a further step in the direction of leaving the own personality, for a short time while acting or reading, or even permanently in pathological cases (Scheler 1970).

  7. 7.

    In moral psychology, it has become common to distinguish quite sharply between cognitive empathy and emotional empathy: the intellectual capacity to imagine what happens in other people’s minds (having a “Theory of mind”), and the emotional identification with the other person. In my reconstruction I understand the needed empathy as mainly cognitive one, while I try to describe a more specific morally-emotional disposition than the broad complex of so-called emotional empathy, which embraces all kinds of “altruistic” emotional dispositions).

  8. 8.

    According to M. Nussbaum, for Stoics like Seneca, cruelty and compassion are no opposites at all, but fruits of the same tree. Compassion is then the answer on behalf of others, where we would feel anger for ourselves, and react perhaps with cruelty in the extreme (Nussbaum 2001, pp. 361–363). For Stoics, compassion is, of course, no morally good attitude at all. However, if we would generally share this intuition, Szasz would have lost a good book title.

  9. 9.

    If not for a very strong deontologist. But for a practical focus like ours, it nearly needs no explanation why such a position would lead to dangerous and undesired results. If I as a totally inexperienced surgeon would operate patients in the best intention, the outcome would be foreseeable disastrous, and I defend the position that this would perhaps be benevolent, but no morally good action.

  10. 10.

    It is also possible to suffer oneself and be compassionate for a person in the same situation. Perhaps this should even be the paradigm case which makes the understanding and the good will and the wish to change the situation most genuine. It is not necessary in order to be compassionate, however. The focus in compassion is on the suffering of the other. The own suffering may help to understand, but it also can distract from concentrating to the other one’s suffering. So one may alternate between suffering, self-pity and compassion. I regard compassion as an asymmetric inclination to help, even in this situation (and I admire the compassionate person even more).

  11. 11.

    Surveys about preferences for prioritisation in healthcare show, that e.g. the German population supports nearly everything that can be done for children, while even severe psychiatric disease would be neglected (Müller and Groß 2009).

  12. 12.

    On behalf of the helpful replacement of a fully moral performance by means of dutifulness, Michael Stocker mentions in a footnote: “Such “feelings” are at times worn thin. At these times duty may be looked to (…) to get done at least the modicum of those things love would normally provide. To some rough extent, the frequency with which a family member acts out of duty, instead of love, toward another in a family is a measure of the lack of love (…). But this is not to deny that there are duties of love, friendship, and the like.” (Stocker 1976, p. 465).

  13. 13.

    In Hume’s philosophy, different modern streams of virtue ethics (Foot 2002; Baier 2007) and philosophy of emotions (Kenny 2003) meet, though they critically modify the historical ideas.

  14. 14.

    And, of course, being incompetent, but this reproach is not in the focus of this article. However, it is worth remembering that good will and being the right kind of person is necessary but not sufficient for the image of a good doctor.

  15. 15.

    Having more tasks than one can manage in the given time, more psychological challenges than one can handle, and meaningless work with little responsibility for the outcome and thus little inner reward and satisfaction in one’s work are typical reasons for burn-out. Attitudes do not need extra-time, and appropriate attitudes can help handling dilemmas by offering orientation and meaning for one’s actions, and thus have the potential to feel more satisfaction and value in one’s profession.

  16. 16.

    On the other hand, a lack of compassion in a balanced partnership may also show a lack of respect and attention to the situation of the other. If only one of the partners is in a painful situation, compassion may be the appropriate additional feeling, not a contradiction or replacement of love.

  17. 17.

    This is also mirrored in approaches of human rights and human dignity. Both may follow one and the same moral intention, but they have a different focus. Compassion is related to dignity, and the obligations in an individual physician-patient relationship mirror the larger scale obligations in global humanitarian engagement.

  18. 18.

    That does not mean that the patient’s choices should be overruled. It is of course his right to decide for his own life, and if he does not need special consideration of his in comparison with the physician more vulnerable state, it should make medical decisions much easier, not more demanding.

  19. 19.

    Eric Cassell describes convincingly how different necessary aspects of modern medicine compete for attention and make a caring, compassionate attitude difficult. A good medical education should prepare students in all necessary aspects of good clinical practice (Cassell 2001).

  20. 20.

    For those who are interested in the archetypical description of good and evil in fiction, and the role compassion can play here, it might be telling to have a look at J. R. R. Tolkien’s “Lord of the Rings” and “Silmarillion”. Here, it is not only possible to learn compassion, by means of being with a person who is excellent in this and by exposing oneself to beings who need and deserve compassion. It is also the decisive distinction between the ultimate and non-corruptible good and the evil, between Gandalf and Sauron. Gandalf learns compassion from Nienna, and his sympathy and understanding for all kinds of beings that he gained and improved this way protect him from being corrupted by the fascination of power—his own power as well as the additional power of the evil ring. It does not seem far-fetched to see a potential in compassion of also protecting physicians from misusing their power or from being corrupted by secondary incentives.

  21. 21.

    Rosalind Hursthouse describes the learning of virtues as follows: “(…) many of the goals appropriate to scientific knowledge—universality, consistency, completeness, simplicity—are not appropriate to moral knowledge; the acquisition of moral knowledge involves the training of the emotions in a way that the acquisition of scientific knowledge does not (…)” (Hursthouse 1993 p. 32).

  22. 22.

    Until now I only have presented two of the three parts of the morally-emotional attitude of “empathic compassionate care”. In order to combine this virtue with medical knowledge and skills a further quality is needed which by several authors refering to Aristotle is called “phronesis”, or also “clinical judgement”. I suggest to describe the attitude from which this capacity is used as “responsibility”.

  23. 23.

    I am aware of the difficulties of examining attitudes and the danger of hypocrisy and rigidity. Probably it would be wise only to exclude extreme cases of psychopaths and habitually disrespectful people, like it is already done. Nevertheless it seems important to me today (more than in the past) to make unequivocally clear that the whole person is demanded in the profession of a physician, not only a splintered and exclusively cognitively-educated part. Nevertheless, I think it is sufficient and more appropriate to focus on demands on the right attitude, not the character of a health care professional.



I wish to thank Bettina Schöne-Seifert, Münster (Germany), Lennart Nordenfelt, Linköping (Sweden), Kristin Zeiler, Linköping (Sweden) 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), and to three anonymous reviewers.


  1. Adams, J. et al. (SAEM Ethics Committee). 1996. Virtue in emergency medicine. Academic Emergency Medicine 3: 961–966.Google Scholar
  2. Aristotle. 2002. Nicomachean Ethics. New York, Oxford: Oxford University Press.Google Scholar
  3. BÄK. 2011. Grundsätze der Bundesärztekammer zur Sterbebegleitung. Deutsches Ärzteblatt 108(7): A346–348.Google Scholar
  4. Baier, A. C. 2007. Trust, suffering, and the Aesculapian virtues. In Working virtue: Virtue ethics and contemporary moral problems, eds Walker, R.L., Ivanhoe, P.J., 136–153. Oxford/New York: Oxford University Press.Google Scholar
  5. Benner, P., and J. Wrubel. 1989. The primacy of caring. Stress and coping in health and illness. Menlo Park, California: Addison-Wesley Publishing Company.Google Scholar
  6. Boleyn-Fitzgerald, P. 2003. Care and the problem of pity. Bioethics 17(1): 1–20.PubMedCrossRefGoogle Scholar
  7. Brody, H. 1998. The family practician: What sort of person? Family Medicine 30(8): 589–593.PubMedGoogle Scholar
  8. Carse, A.L. 1998. Impartial principle and moral context: Securing a place for the particular in ethical theory. Journal of Medicine and Philosophy 23(2): 153–169.PubMedCrossRefGoogle Scholar
  9. Cassell, E.J. 1982. The nature of suffering and the goals of medicine. The New England Journal of Medicine 306(11): 639–645.CrossRefGoogle Scholar
  10. Cassell, E.J. 1991. The nature of suffering and the goals of medicine. New York/Oxford: Oxford University Press.Google Scholar
  11. Cassell, E. 2001. Forces affecting caring by physicians. In The lost art of caring, ed. L.E. Cluff, and R.H. Binstock, 104–124. Baltimore/London: John Hopkins University Press.Google Scholar
  12. Comte-Sponville, A. 1996. Ermutigung zum unzeitgemäβen Leben. Ein kleines Brevier der Tugenden und Werte. Reinbek bei Hamburg: Rowohlt (orig. 1995. Petit traité des grandes vertus. Paris: Presses universitaires de France).Google Scholar
  13. Damasio, A.R. 1994. Descartes’ error. New York: Grosset/Putnam.Google Scholar
  14. de Sousa, R. 1987. The rationality of emotion. Cambridge, MA: MIT Press.Google Scholar
  15. Daniels, M.J. 1960. Affect and its control in the medical intern. American Journal of Sociology 66(3): 259–267.CrossRefGoogle Scholar
  16. Ferrel, B. 2005. Ethical perspectives on pain and suffering. Pain Management Nursing 6(3): 83–90.CrossRefGoogle Scholar
  17. Foot, P. 1990. Von Wright on Virtue. In The philosophy of Georg Henry von Wright, ed. P.A. Schilpp, and L.E. Hahn, 271–279. La Salle, Illinois: Open Court.Google Scholar
  18. Foot, P. 2002. Virtues and vices. Oxford/New York: Oxford University Press.CrossRefGoogle Scholar
  19. Fry, S.T. 1989. The role of caring in a theory of nursing ethics. Hypatia 4(2): 88–103.PubMedCrossRefGoogle Scholar
  20. Gelhaus, P. 2011a. Robot Decisions. On the importance of virtuous judgment in clinical decision-making. Journal of Evaluation in Clinical Practice 17(5): 883–887.PubMedCrossRefGoogle Scholar
  21. Gelhaus, P.I. 2011b. Medical Robot. On the differences between a virtuous doctor and a good robot. International Journal of Person-Centered Medicine 1(2): 301–306.Google Scholar
  22. Goldie, P. 2002. Emotions, feelings and intentionality. Phenomenology and the Cognitive Sciences 1: 235–254.CrossRefGoogle Scholar
  23. Hamburger, K. 1985. Das Mitleid. Stuttgart: Klett-Cotta.Google Scholar
  24. Helm, B. 2001. Emotional reason. Deliberation, motivation and the nature of value. Cambridge: Cambridge University Press.CrossRefGoogle Scholar
  25. Hem, M.-H., and K. Heggen. 2004. Is compassion essential to nursing practice? Contemporary nurse 17(1–2): 19–31.PubMedCrossRefGoogle Scholar
  26. Housset, E. 2003. L’inteligence de la pitié. Phénoménologie de la communauté. Paris: Les Éditions du Cerfs.Google Scholar
  27. Hume, D. 2004. 1739/40 A treatise of human nature. Mineola, NY: Dover Publications.Google Scholar
  28. Hursthouse, R. 1993. Normative virtue ethics. In Identity, character and morality. Essays in moral psychology, ed. O. Flanagan, and A. Oksenberg Rorty, 19–36. Cambridge (Mass.)/London: MIT Press.Google Scholar
  29. Jecker, N.S., and D.J. Self. 1991. Separating care and cure: An analysis of historical and contemporary images of nursing and medicine. The Journal of Medicine and Philosophy 16: 285–306.PubMedGoogle Scholar
  30. Kenny, A. 2003. Action, emotion and will, 2nd ed. London/New York: Routledge.Google Scholar
  31. Latimer, E. 1991. Caring for seriously ill and dying patients: The philosophy and ethics. Canadian Medical Association’s Journal 144(7): 859–864.Google Scholar
  32. Leget, C., and G. Olthuis. 2007. Compassion as a basis for ethics in medical education. Journal of Medical Ethics 33(10): 617–620.PubMedCrossRefGoogle Scholar
  33. MacIntyre, A. 2007. After virtue, 3rd ed. London: Duckworth.Google Scholar
  34. Mills, M., H. Davies, and W. Macrae. 1994. Care of dying patients in hospital. British Medical Journal 309: 583–586.PubMedCrossRefGoogle Scholar
  35. Miller, F.G., and H. Brody. 2001. The internal morality of medicine: An evolutionary perspective. Journal of Medicine and Philosophy 26(6): 581–599.PubMedCrossRefGoogle Scholar
  36. Müller, S., and D. Groß. 2009. Akzeptanz von Leistungsbegrenzungen im Gesundheitswesen: Strategien, Kriterien und Finanzierungsmodelle unter Berücksichtigung ethischer Aspekte. In Gesundheitsmonitor 2009. Gesundheitsversorgung und Gestaltungsoptionen aus der Perspektive der Bevölkerung, ed. J. Böcken, B. Braun, and R. Amhof, 258–279. Gütersloh: Bertelsmann.Google Scholar
  37. Nordenfelt, L. 1974. Explanation of human actions. (Philosophical Studies no. 20) Uppsala.Google Scholar
  38. Nordenfelt, L., and P.-A. Tengland. 1996. The goals and limits of medicine. Uppsala: Almqvist Wiksell.Google Scholar
  39. Nussbaum, M.C. 2001. Upheavals of thought. The intelligence of emotion. Cambridge: Cambridge University Press.CrossRefGoogle Scholar
  40. Pellegrino, E.D. 1995. Toward a virtue-based normative ethics for the health professions. Kennedy Institute of Ethics Journal 5(3): 263–277.CrossRefGoogle Scholar
  41. Pellegrino, E.D., and D.C. Thomasma. 1988. For the patient’s good. The restoration of beneficence in health care. New York/Oxford: Oxford University Press.Google Scholar
  42. Pellegrino, E.D., and D.C. Thomasma. 1993. The virtues in medical practice. New York/Oxford: Oxford University Press.Google Scholar
  43. Polanyi, M. 1983. The tacit knowledge. Magnolia MA: Peter Smith.Google Scholar
  44. Schantz, M.L. 2007. Compassion: A concept analysis. Nursing Forum 42(2): 48–55.PubMedCrossRefGoogle Scholar
  45. Scheler, M. F. 1970. The nature of sympathy. New York: Archion Books (orig. Wesen und Formen der Sympathie, 1923).Google Scholar
  46. Schopenhauer, A. 1977. (1859) Über die Grundlage der Moral. Herrsching: Manfred-Pawlak-Taschenbuch-Verlagsgesellschaft.Google Scholar
  47. Sherman, N. 2005. Stoic warriors: the ancient philosophy behind the military mind. Oxford/New York: Oxford University Press.Google Scholar
  48. Slaby, J., and A. Stephan. 2008. Affective intentionality and self-consciousness. Consciousness and Cognition 17: 506–513.PubMedCrossRefGoogle Scholar
  49. Slote, M. 2001. Morals from motives. Oxford/New York: Oxford University Press.CrossRefGoogle Scholar
  50. Smith, A. 1759. The theory of moral sentiments. London.Google Scholar
  51. Stocker, M. 1976. The schizophrenia of modern ethical theories. The Journal of Philosophy 73: 453–466.CrossRefGoogle Scholar
  52. Szasz, T. 1994. Cruel compassion. New York: Wiley.Google Scholar
  53. Teuber, A. 1982. Simone Weil: Equality as compassion. Philosophy and Phenomenological Research 43(2): 221–237.CrossRefGoogle Scholar
  54. The Goals of Medicine.1996. Setting New Priorities. In: Hastings Center Report Nov-Dec. Special Supplement.Google Scholar
  55. Turner, S. 1994. The social theory of practices: tradition, tacit knowledge and presuppositions. Chicago: University of Chicago Press.Google Scholar
  56. von Wright, G.H. 1963. The varieties of goodness. London: Routledge & Keegan Paul.Google Scholar
  57. Wilmer, H.A. 1968. The doctor-patient relationship and the issues of pity, sympathy and empathy. The British journal of medical psychology 41(3): 243–248.PubMedCrossRefGoogle Scholar
  58. Winslow, G.R. 1984. From loyalty to advocacy: A new metaphor for nursing. The Hastings Center Report 14: 32–40.PubMedCrossRefGoogle Scholar
  59. Wollheim, R. 2000. On the emotions. New Haven: Yale University Press.Google Scholar

Copyright information

© Springer Science+Business Media B.V. 2011

Authors and Affiliations

  1. 1.Institute for Ethics, History and Theory of MedicineUniversity of MuensterMuensterGermany
  2. 2.Department of Health and Society, Institute of Medical and Health SciencesUniversity of LinköpingLinköpingSweden

Personalised recommendations