“Anthropology and Medicine:
Empathy, Experience and Knowledge”.
Un texte publié dans l’ouvrage sous la direction de Els van Dongen et Ruth Kutalek, Facing Distress. Distance and proximity intimes of illness, pp. 1-20. Wien, Wiener ethnomedizinische Reihe, Bd. 4, 2007.
- Empathy in the doctor-patient relationship
- Empathy in the anthropological field
- Empathy and experience
- Empathy and knowledge
- To conclude
Even though the issue of empathy is widely debated in disciplines such as ethnology and medicine, ethnologists' relationship with their informants cannot be likened to that of doctors and their patients. Likewise, the empathic dimension, considered to characterize the relationship in both cases, is not the same. In this article examples drawn from various case studies are used in a cross-cutting analysis of the nature and role of empathy in the contexts of medical and ethnological relationships. The part that empathy plays in the production of knowledge, in both types of relationship, is evaluated. What are the implications of empathy as regards the medical doctor's work and that of the ethnologist, and what is its heuristic value? In other words, what link is there between empathy and the knowledge production process? I show that even though empathy may be useful in both types of situation, ethnologists can derive a benefit from it only if, unlike doctors, they are able to control it and to distance themselves from it.
At the heart of the methodological and epistemological reflection involved in anthropological practice lies the following question: To what extent is empathy coextensive to the anthropological approach ? In other words, is the practice of the profession contingent on empathy, generally conceived of as the ability to put oneself in the place of the Other ? This is also a key question in the more particular context of medical anthropology, in so far as it constructs objects and conducts field research in which the anthropologist has a particular relationship with individuals, their ailments and their bodies an area traditionally occupied by medicine.
Just as anthropologists' relationship with their informants cannot be likened to that of doctors with their patients, so the empathic dimension characteristic of both types of relationship differs in each case . The aim of this paper is to examine the connection between empathy and knowledge and, more precisely, the specific role that empathy is likely to play, in anthropology, in the process of knowledge production. Are empathy and knowledge characteristic of a relationship of possibility or one of necessity ? What is the added value of the empathic dimension and of the phenomena traditionally associated with it in anthropological research (experience, emotion, understanding through emotion, etc.)?
Based on examples drawn from various field studies, I undertake a cross-cutting analysis of the nature and role of empathy in the contexts of the doctor-patient and the observer-informant relationships, and evaluate the part that empathy plays in producing a form of knowledge towards which both types of relationship tend. In concrete terms, I examine the way and the conditions in which the empathic relationship can be established, with a view to furthering insight into the knowledge-creation project of anthropology. This perspective on these different types of empathy and the examination of these diverse situations help me to answer the following questions: What are the implications of empathy on the anthropologist's work and what is its heuristic value? In other words, what impact does empathy have on the production of knowledge and in what conditions does the former facilitate the latter?
Empathy in the doctor-patient relationship
The first question we need to consider is whether it is necessary for doctors to have an empathic relationship with their patients in order to understand and treat them. This question has no simple answer because it immediately raises many others. First, do doctors need to understand their patients in order to treat them? What does 'understand' actually mean in this case? What are the implications, for the doctor, of having an empathic relationship? Does it mean that they have to identify with their patients? And in order to do so, do they need to have shared their experience? Do they need to have (or to have had) the same disease in order to be considered as 'a good doctor'? Most doctors would say 'of course not'. Their diagnosis and treatment do not require a sharing of the patient's suffering or the doctor's ability to put him- or herself in the patient's position. But is this how patients and some doctors actually see things? On this subject it is fruitful to examine empathy from the point of view of both the patient and the healer (a doctor or not) in different cultural contexts, before proposing an ethnological reading of what these perceptions reveal.
In the Bisa society, a West African lineal society, the therapeutic function is fulfilled by traditional healers. Two types of healer exist. The first type, in the majority, acquire their knowledge from their ancestors (agnatic or uterine lineage), from their father in the case of men and from their mother in that of women. The second type, the minority, become healers of a particular illness because they themselves had it in the past. This process of acquisition of therapeutic knowledge is an exception to the prevailing principle of genealogical transmission. It has the effect of training healers with a single speciality, able to heal an illness in which they have experience. The justification is then that it is necessary to have known the disease to be able to recognize and understand the patient. Due to the requirement of having been in the patient's position in order to be able to heal him or her, this is a perfect example of the idea of knowledge through experience and empathy. Even if their legitimacy is not superior to that of the healers from whom they inherited their function, it is at least equal as regards the institution. Here experience replaces learning through transmission and gives the healer the ability to understand the patient and the power to heal him or her (Fainzang 1986).
In the very different cultural context of former drinkers in Western societies, the subjects consider that only another alcoholic can really understand an alcoholic. At Vie libre one has to have personally experienced the ‘alcoholic's illness’ to understand it and to be able to treat others (Fainzang, 1996). In this respect, the healer's role is more readily recognized in peers (other former 'alcoholic patients') than in medical doctors. Although this phenomenon is characteristic of most organizations of this nature, there is divergence in the definition of the nature and subject of the experience. For instance, Alcoholics Anonymous and Vie libre have distinctly different conceptions of the idea of sharing an experience and the possibility of identification. The former believe that only alcoholics can understand, whereas at Vie libre the alcoholics' spouses, who have experienced the disease by suffering through life with an alcoholic, are called on to share the healing experience and are considered capable of sharing their suffering. (Alcoholics' spouses often claim to be former alcoholics or former patients when they visit alcoholics in hospital to try to persuade them to join the movement. The idea is that the treatment will be more effective if there is greater understanding.) A second difference concerns the group's relationship with the medical profession. Alcoholics Anonymous has a clear-cut position in this respect: they believe that there is no point in consulting doctors since they are not themselves alcoholics and are therefore unable to understand a heavy drinker. The Vie libre movement has a slightly different approach: it recommends cooperation with the medical profession even if it believes that true healing of an alcoholic is possible only among former alcoholics. Cooperation with doctors is considered necessary in so far as certain patients need detoxification treatment, even if it is insufficient. The real healing is achieved by being with those who are (or can imagine themselves) in the alcoholic's position. Most doctors agree; they consider these organizations to be useful and beneficial, and readily refer their patients to them, especially to Vie libre, as a complement to the detoxification cure. Their motto is: 'We treat, you heal' as Ambroise Paré put it. This tendency to see certain individuals' ability to understand and to feel what someone else is experiencing as a sign of a superior competence in the field of therapeutic care, is nevertheless problematical. Certain members of the movement who feel that they have been vested with the role of a 'healer' sometimes recommend therapeutic solutions to other alcoholics, when in fact this would normally be the doctor's role. For instance, they recommend taking esperal  or antidepressants that they themselves took in a situation they see as similar. Although some may qualify this as abusive, it is interesting to note that it is the competence endowed by empathy that prevails and prompts alcoholic peers to recommend treatment, regardless of medical advice. The link is nevertheless clearly expressed here between experience, empathy and knowledge. Like the African healers considered being able to cure a disease that they themselves have had, intimate knowledge of the disease and the possibility of identifying with the patient is believed to endow a person with superior competencies for overcoming the disease. It is as if, in the medical relationship, the healers' empathy, their understanding of their patients and their identification with them, through their personal experience of the disease, were an alternative to learning and equipped them with competence that was not identical but equal.
A third situation examined here is that of doctors, observed during their consultations (in French hospitals), and their relationship with their patients around certain pathologies. Various types of empathy can emerge in the practice of their profession, two of which seem to be dominant. The first type of empathy is induced by a specific illness that the doctor may or may not have had, for which he or she developed a particular interest, and that may even prompt him or her to choose the speciality corresponding to it. This is the case, for example, of a lung specialist who is very understanding with his asthmatic patients because he also suffers from asthma. It is very similar to the case of African healers considered above; only the learning process is different. While observing consultations we noted that doctors sometimes reveal this information to patients who seem reluctant to take their advice and the drugs prescribed. In this way the patient is informed that the doctor is familiar with the problem. The patient's attitude then changes, as if intimate knowledge of the illness and the resulting empathy reinforced the doctor's understanding and therefore his or her competence.
The other type of empathy is induced by the doctor's ability or tendency to identify with a patient because he or she sees him- or herself in that specific patient. Either the patient is of the same age or gender as the doctor, or the patient is a child or an aged person in whom the doctor sees his or her own child or parents. Empathy can also be experienced with someone close to the patient such as a family member. In all cases it stems from a tendency to see oneself through another person. Yet the question of the outcome of the relationship of proximity generated by empathy remains unanswered. Does it produce better understanding of the illness and a greater ability of cure it?
While medical empathy helps to establish a relationship of trust, there is no doubt that it can also be productive. For instance, it may allow the doctor to collect information required for diagnostic purposes and perhaps also to obtain stricter observance by the patient, for we know that a patient's observance is improved when there is a relationship of trust with his or her doctor (Haynes, Taylor & Sackett, 1979). Questioning the role of empathy in medical practice, Spiro & alii (1993) wonder whether it is dangerous for a patient if a doctor becomes too empathic. The authors unanimously answer in the negative. Empathy is fundamentally good in so far as it provides the doctor with information on the patient and can produce something good in the relationship, if only because it influences the patient's feelings towards the doctor. Is this trust contingent on empathy? It does not seem that the possibility of being seen as a good doctor (and actually being one) necessarily requires empathy. Although empathy can be useful and may even compensate for or enhance competence, it is not essential. It may have a technical role or function but it does not enter into the definition of healing, neither in clinical methods nor in therapeutic care. Medical empathy facilitates access to knowledge that doctors need to obtain on their patients, but it is not a condition on which the practice of their profession depends. Conversely, doctors' proximity with their patients (e.g. family proximity) that certain doctors refuse, does not preclude the ability to analyse a case (at least in somatic medicine). Certain doctors' choice not to treat their own family members relates either to the emotional charge of this type of situation, to the difficulty of taking decisions (e.g. an operation or chemotherapy), to the psychological or moral pressure that they are likely to experience, or, finally, to family and friends' judgement of their acts. Although empathy is not a prerequisite, it is not either an obstacle to knowledge .
Empathy in the anthropological field
Generally considered as intrinsic to ethnological field work, empathy is a complex attitude or disposition. It is not always clear whether it is related to the observer or to the object studied. How is this empathy constructed? Is it a necessary or useful tool? Is it deliberate or inevitable? Is it controlled?
In this section I examine the conditions in which empathy can or must exist. I question the connection between the existence of empathy and the type of research (the object, the group, the research question), with a view to establishing what this empathy contributes to the knowledge project. For this purpose I draw on some personal experiences in the field and, above all, on a study on African polygamy in immigrant families. In this research project I investigated the way in which women deal with their sexuality and their reproductive capacity in the context of this type of marriage in France, and the future of polygamy in situations of immigration. The method employed was maximum immersion in the field, which involved spending long days with these women, in their homes, helping them to prepare meals, feeding their babies when they were busy, or going shopping with them. In this way I was able to experience albeit only partly the woman's condition in this type of household.
The first problem was that of the criteria of empathy. It was very difficult to identify (among those who enabled me to feel what they were feeling, or to 'put myself in their shoes') the decisive criteria that allow empathy. Even if my empathic relationship with African women was facilitated by the fact that I am also a woman (which is what some of them thought since a person's definition in their society is based primarily on gender), it was nevertheless probably limited, as was my ability to really understand them because I am neither African nor married in a polygamous system (which is what others thought). Hence, there are not simple criteria on which identification can be based, but a set of entangled positions that combine or cancel one another out, making the definition of an optimum position for the realization of empathy an eminently complex affair. We thus see to what extent the construction of empathy is relative, as is that which serves to consolidate it. Is it a question of sensitivity to the difficult material and social conditions in which immigrants live? Is it the possibility of sharing feelings experienced by women married to the same man? Does the social condition of a woman suffice for this empathy to exist or is it necessary to add the experience of particular emotions?
On this point it is relevant to mention what Max Weber (1971) says in the first pages of his book Economy and Society on the methodological foundations of sociology, where he examines the modes of understanding behaviours and social activity, showing that it can be either rational or empathic (that is, emotional). Yet he shows the limits of understanding in the latter case since the phenomena can be radically different from our own perceptions (or what he calls our 'evaluations'). He takes the example of certain religious or mystical phenomena that may be difficult to understand for someone who is inaccessible to this type of experience. According to Weber, when it comes to understanding through empathy, we understand behaviours, reactions or feelings (jealousy, anger, fear, etc.) only to the extent that we are ourselves accessible to those emotions and that they constitute significant entities for us. In a sense, rational understanding is seen to offset the lack of understanding through empathy, which is considered to be productive from the outset.
However, being able to feel what individuals feel, to put oneself in their place, can also be problematical. In this respect I wish to revert briefly to the case of the African women mentioned above, and to report some ethnographic findings. Despite the cramped conditions in which African immigrant families that are often large live in France, separation between the sexes persists. The female sphere remains associated with the private domain and the male sphere with the public domain. Consider the case of a family living in Paris in two rooms, each situated on a different floor of an old building in the 20th arrondissement. The smallest room serves as a bedroom for the children and those wives who are not assigned to sexual services. The largest room (13m²) is where all the members of the family meet at mealtimes, especially when visitors are present, and in which the domestic space is restructured on the public/private model. This family comprises a man, his three wives and their twelve children. The furniture in the room consists of three beds arranged in a U-shape (the base of the U against the back wall). The women are seated towards the inside (at the base of the U, against the wall at the end of the room) and the men towards the outside of the U, closer to everything connected with the outside world: the entrance door, the TV set near the door and the telephone, placed on the TV. The public/private opposition is coupled with an outside/inside opposition, also found in the respective positions of the man and the woman when they are in bed. In relation to the opening of the room, the woman always lies behind the man. This occupation of space is significant: women are placed towards the inside not only because they are responsible for domestic work but also because their access to the outside world is controlled and mediated by men. The place assigned to them in the room reflects and symbolizes the position granted to them as regards both the outside world and public life. The husband occupies a strategic intermediate position, between the woman and the door; he is the gateway to the outside world or to society (cf. Fainzang & Journet, 1988).
Reference to this situation has the particular advantage of showing how empathy can be constructed and revealed through use of the space in which the ethnological relationship is situated, where 'putting oneself in the place of' is fully meaningful, both literally and metaphorically. In fact, my presence was incorporated into this process on the basis of the place that I occupied in their view. Invested with the status of a foreigner, a 'visitor', I was initially invited to sit at one of the ends of the U formed by the three beds (that serve both as bed and as seats). In other words, I was seated among the men because I was associated with the outside world. But as the bonds became stronger I lost the status of a representative of the institution (I was neither a social worker likely to provide them with decent accommodation nor an official checking whether they had a valid residence permit). Consequently, I also lost the status of a foreigner. I simply became a woman, a friend of the family who was given a place at the base of the U (either on the bed at the back, or on the most distant part of the lateral beds), that is, among the women. I was not the one who, through my ethnological inquiry, wanted to put myself in the women's place to understand them, but I was effectively put in the women's place. Almost against my will my informants put me in an absolutely empathic position, not only socially but also spatially, that is, both in the context of social relations and in the domestic space.
This immersion and the resulting empathy entail difficulties related to the fact that the observer is unwittingly involved, by the subjects themselves, in their relationships to the point of finding him- or herself in one camp or another in the event of conflict. This inevitably raises the question of whether one can be in an empathic relationship simultaneously with two individuals in conflict. In the case described here, considering the rivalry between co-spouses inherent in the polygamous situation, it was impossible for me, in the family's eyes, to be the confidant of all concerned. The fact of developing a relationship with a woman and being considered by the family as 'her friend' meant that I implicitly supported her in her rivalry with her co-spouse. It followed that I was unable to form a close relationship with the latter. If one confided in me regarding her problems of sterility that she imputed (like all reproduction-related diseases) to persecution by the co-spouse, it stood to reason that the other spouse would not also confide in me. In short, due to the empathy I felt for the first spouse, I was considered unable to put myself in the place of the second one. I consequently had to undertake a painstaking process of reconstruction, consisting in putting together the pieces and compensating for my shortcomings regarding the experience of the second wife in families where I had privileged relations with the first, by striving to build up privileged relations with the second or third wife in other homes. Although empathy was achieved through a long period of frequentation of these women, I experienced the need to withdraw from these situations in order to avoid confining myself to the women's point of view. Not only did I not want to limit myself to understanding the viewpoint of the first wife and not the others (whose social position, via their rank in the order of marriages, is attended by a different way of conceiving of their relations with others, with their husband and with the community at large); I also wanted to avoid limiting my understanding to women only. I wished to understand the social pressure that forces men to take several wives and, within the social relations implicit in this institution, the distress it sometimes causes them.
For the most objective analysis possible of the situation under study, the ability to identify with some, in the process of familiarity of the inquiry, should not stop us from understanding and possibly of also developing empathy for the others, those considered by the former to be their rivals. In rigorous ethnological practice researchers should be able to feel empathy for different individuals in different situations. In this way they are in a position to analyse the power struggles and behavioural logics of the subjects under study. Empathy, inherent in the object and the approach, is therefore something ethnologists have to be able to withdraw from at some point in time. This is necessary in order to be able to 'put themselves in the place of' the different informants involved in a play of social relations in which none have exactly the same place, and to be able to stand back from their representations, for analytical purposes.
In this respect, in the research mentioned above on Alcoholics Anonymous, I found myself, from the outset, in the difficult position in which the people studied asked me if I had an alcoholic in my family or if I was concerned in one way or another by the problem, in order to be able to understand them. In all honesty I answered 'no' since I have never experienced alcoholism myself or the suffering induced by the fact of living with an alcoholic. In terms of the categories defined by the organization itself, which distinguishes between 'former drinkers', 'voluntary abstainers' and 'sympathizers', I did not claim to be anything more than a 'sympathizer'. The term itself put me in a relationship of sympathy and not empathy (a distinction that is worth highlighting and to which I will revert). Yet the members of the branch that I frequented the most regularly asked me to become a 'pink card', that is, to have the same status as themselves, the same place, that of a 'voluntary abstainer'. In short, I found myself snapped up by an object and by people whose pain I had come to understand, intellectually and emotionally, by spending a lot of time with them. But I refused this status, not only because I had no desire to become abstinent but also because it seemed to me imperative to allow myself the means to achieve empathic detachment with a view to coming as close as possible to the truth of the object. In this respect it seems that empathy is a feature not of the ethnologist but of ethnology, in so far as it is not a personal predisposition of the observer but a dimension of his or her approach. Empathy is so closely connected to the ethnologist's approach that it can be achieved without necessarily being sought. It is strongly correlated with the object in so far as it is more easily felt when that object involves emotions (and suffering and disease are strong triggers of emotions), precisely because there is a possibility of identification. I therefore had to try rather to leave the naturally empathic position in which the research put me, and seek the weaknesses in the subjects' behaviours with a view to identifying apparent inconsistencies in some of the observed practices. By putting oneself too much in the place of the subjects studied, and especially remaining there, one runs the risk of entering into their representations to the point of no longer seeing, for instance, that some of their behaviour is paradoxical a question that ethnology is supposed to investigate. Therefore, I once again felt the need to detach myself from this empathy, although for different reasons to those noted above. Whereas my field research on polygamy raised the question of whom the ethnologist could feel empathy for, considering the multiplicity of places that he or she can occupy in the field, in the case of alcoholics I was confronted with the necessity to find weaknesses, to act as if I was unable to understand them from an emotional point of view, precisely in order to be able to understand them intellectually. Empathy therefore appeared to be a positive stage but one that should last only for a while. Whereas it can be useful to feel empathy, one also has to move beyond it. Hence, empathy can be demanded provided that it is also mastered, allowing one to stand back from it or to deconstruct it.
The question of the bases on which empathy can be realized is crucial when one is working on different types of groups. This was the case, for example, of research that I undertook on the use of medication and medical prescriptions, depending on individuals' religious cultural origin (Catholic, Protestant, Jewish or Muslim) (Fainzang, 2001). Here again the question of criteria arises, and these criteria are in a sense always uncertain and sometimes unpredictable. Does the methodological choice of feeling close to people and putting oneself in their place enable one to put oneself in everyone's place? Can the ethnologist put him- or herself in the place of people who are culturally close or, by contrast, different from him or her? What does culturally close mean in this case? Is it a common cultural or religious origin of the ethnologist and his or her informants? Or is it rather the fact of having in common, as in my case, no religious belief at all? The difficulty becomes that of the possibility and viability of empathy in a study that is supposed to be comparative from the outset. If empathy cannot be achieved with everyone, perhaps it would be better to have it with no one (if can be controlled), for nothing enables us a priori to know in which case and with whom, in a heterogeneous population, empathy will have the greatest chances of being felt.
Empathy and experience
Some people would say that it is impossible to put oneself completely in someone else's place, precisely because one is not in that place. There cannot be complete identification since this move is imagined and not lived. Even when it has an emotional dimension, empathy is always an intellectual construction. And even when it is rooted in experience, it is a theoretical construction. An ethnologist who has experienced neither alcoholism nor the suffering of living with an alcoholic spouse, can still imagine and reconstruct that experience. Empathy is then an intellectual construction.
In Dilthey's definition of empathy as the situation in which “the soul follows the accustomed paths in which it enjoyed and suffered, desired and acted in similar situation” (Dilthey, 1926: 159), we see that for him empathy can be experienced only by a person who has lived an identical experience to that of the subject. In other words, one has to have had the same experience to be able to develop an empathic position. Yet, as seen above, it seems that the conditions of the possibility of empathy are as much in an intellectual or emotional position as in the studied subject’s former experience of the situation. Hence, in my opinion, Martinelli (2000) is right when, in contrast with the idea defended by Dilthey that empathy feeds partly on experience and memory, he argues that one cannot always find these resources within oneself. According to him, in certain cases the only possibility is a theoretical construction, for instance the experience or hunger or a tramp's life.
The very notion of 'empathic understanding' as forged by the social sciences, with its association of experience and imagination, following Weber and Dilthey, accounts for this intellectual construction. As Gauss (2003) writes: “In the social sciences, […] empathic understanding refers to our deliberate attempts to identify ourselves with another, accounting for his actions by our own immediate experience of our motivations and attitudes in similar circumstances as we remember or imagine them” (my underlining) . Thus, it is no longer as necessary to have lived the same experience as to be able to imagine what we would feel if we were to live it. The Encyclopaedia Britannica (1999 edition) defines empathy as: “the ability to imagine oneself in another’s place and understand the other’s feelings, desires, ideas and actions” (Vaknin 1999). The experience is therefore not the one that we need to have had prior to the empathic feeling, but the one that actually produces that feeling. This is what emerges from the definition of Ch. Morris (1996: 442, quoted by Vaknin), for whom: “closely related to the ability to read other people’s emotions is empathy [...] empathy depends not only on one’s ability to identify someone else’s emotions but also on one’s capacity to put oneself in the other person’s place and to experience an appropriate emotional response”.
The conceptualization proposed by Allison Barnes and Paul Thagard for the terms empathy and sympathy (Barnes & Thagard, 1997) enables us to specify the status of the experience in these two situations. For them, empathy “is a more extensive concept than sympathy; it signifies the ability to comprehend another’s state without actually experiencing that state”. On this point they cite Lauren Wispe (1991: 80): “In empathy one acts “as if” one were the other person… the object of empathy is understanding. The object of sympathy is the other person’s well-being. In sum, empathy is a way of knowing; sympathy is a way of relating” . Barnes & Thagard comment on this as follows: “sympathy is performed with altruistic ends, but empathy may or may not be motivated by good intentions”. This distinction is subject to debate because there is no guarantee that the doctor needs to feel sympathy to treat and thus to do good to a patient. What is assumed to characterize sympathy as opposed to empathy is no longer relevant when the concern to do good enters into the definition of a professional activity. The fact that doctors are not required to feel sympathy is attested de facto by many medical professionals. As Martinelli (2000) says, empathy and antipathy can go together.
The distinction between these terms relates partly to the debate relative to the status of experience in the empathic phenomenon. In many studies that set out to define the nature of the difference between the two concepts, the emphasis is on the act of understanding underlying the feeling of empathy. Thus, for Switankowsky (2000), sympathy is “feeling another individual’s predicament […] leading to a type of emotional identification between two individuals. Empathy consists of understanding another person’s situation”. Yet understanding does not necessarily mean knowledge. As Wispe (1991: 89) points out: “it is a very long journey from understanding to knowing and understanding, while a component of knowledge cannot be construed as knowledge”.
Empathy and knowledge
In medicine, as in anthropology, empathy may be seen as a tool or a strategy. In fact it is certainly more than that: doctors' own relationship with death, for instance, certainly impacts on their empathic relationship with patients with serious diseases. While the empathic relationship is believed to be useful in establishing a relationship of proximity between the players in many interactions, including between doctors and patients, it is often seen as indispensable in the relationship between observer and observed in anthropology. Most social scientists accept this idea, following Weber and Dilthey who, as we have seen, considered empathy as the very foundation of the method peculiar to the social sciences (on this point see also Bakker, 1999). Empathic understanding is a prerequisite to sociological understanding.
Apart from the consensus around this issue in social science, it seems relevant to wonder whether the usefulness of empathy implies its necessity. Or, more precisely, does the empathic relationship have to remain a constant for knowledge to be possible? Although there is no doubt that empathy is a specific characteristic of the anthropological approach, it is also true that the extended periods of field research required in ethnology favour empathy. But even if regular frequentation of the people studied facilitates the establishment of an empathic relationship, the knowledge resulting from the former is not necessarily contingent on the latter. This is probably where the polysemy of the word 'understand' (understand the Other) applies most. One can know and frequent someone enough to understand their acts (to explain their motives and recognize their meaning) without understanding emotionally and without ever being able to put oneself in that person's place. Consider the following case. During research on the models of interpretation of disease undertaken in Paris and surrounding areas, I studied a family in which the father had kidney failure and was dialysed at home. This family consisted of this man (a tax collector), his wife (a nurse), and their two children, a girl and a boy. The father shared his bed with his daughter and the mother shared hers with her son, and both ardently defended far right-wing ideas. While the fact of frequenting these people enabled me to understand (and to give meaning to) the way in which they perceived and dealt with disease, I do not think that it ever enabled me to 'put myself in their place' (regarding either their domestic choices or their political opinions). In this case, empathy was not necessary for the analysis. And despite its utility in understanding the phenomena and subjects studied, it must be followed by an objective view. Putting oneself in someone else's place means trying to have the same point of view on things. If I sit on the chair on which X is seated it will enable me to see what and how he or she sees, but I would have to change chairs to see what his or her place is in the overall game of social relations. It is therefore necessary to leave empathy to attain knowledge.
The possibility that we have of understanding what a person feels in a given situation, that prompts them to act in a particular way, tells us nothing about our perception of that person and our feelings for him or her . The same applies to doctors. Even if a family has a distinct dislike for a doctor, if their child is running a high fever their antipathy will not prevent the doctor from feeling empathy and putting himself in the parents' place, that is, understanding their concern because it reminds him of his own experience (although this will not make him any more capable of dealing with the fever). The ability to understand patients or informants, on an emotional level, can be facilitated if the experience they are having resonates with that of the doctor or ethnologist. On the other hand, unlike the doctor, the ethnologist necessarily has to deploy a huge effort to stand back from his or her own perception of the phenomenon under analysis, in order to avoid the risk of transferring his or her own system of representations onto it.
This is the risk that Gay Becker (1999) took, for instance, when she chose to study the way in which individuals experience the appearance of a disease or a sudden disruption in their lives. A noteworthy feature of her approach is the fact that her personal experience is at the centre of her intellectual interest in the subject. This study of disturbed, even broken lives results, in a sense, from a reflexive approach. Gay Becker tried to understand how people experience a disruption, in order to deal with the one that she had experienced in her own personal life. She explains to the reader that she spent her childhood in a marginal community, her family being members of what she refers to as 'an atypical protestant sect', the Christian Scientists, in which her grandmother was a healer. As a child she never saw doctors nor participated in activities at school, and it was in this context that she developed an allergic pathology and asthma. She believes that what she was experiencing was a situation of 'incorporated distress'. Her condition and her family's failure to provide her with the possibility of healing caused her to turn to another type of medicine, bio-medicine. This constituted a major break with family traditions. Her experience enabled her to be readily received by the anthropological community that tends, she says, to celebrate marginality, and prompted her to choose as a research subject biomedicine, a field that was, in a sense, foreign to her. She explains that she considered herself in a suitable position to take it as a subject of anthropological investigation and that, by contrast, she would have had difficulty studying popular healing systems that were familiar to her precisely because she felt she was too close to be able to study them 'without passion'. I can only approve of a choice made out of a concern to preserve the precious distance from the object of anthropological study; on the other hand, I cannot entirely subscribe to her reflexive analysis when she affirms that, not having been 'socialized' in the context of biomedicine, at least before adulthood, she did not consider it with the same emotional charge as Christian Scientism. We can legitimately reply that the fact of belonging to a marginal group like the Christian Scientists is hardly likely to make her view biomedicine in a neutral way, without passion! Finally, of the different events likely to disrupt a person's existence, Gay Becker focused in particular on the subject of 'sterility', also inspired by her own personal experience for several years. Even if the subjectivity of her approach (attested by her interest in problems that marked her personal experience) does not reduce the acuity of her analysis, I wonder to what extent the empathy that she felt for her asthmatic or sterile informants was productive, and whether she did not superimpose her own representations on those of the individuals studied.
The different studies mentioned above were chosen to contribute to reflection on the role of empathy in the knowledge production process. I hoped in this way to establish whether empathy produces knowledge and whether knowledge production requires empathy or the sharing of an experience.
Although the aim of medical activity is generally to heal, whereas that of ethnological study is only to analyse, the question relative to the doctor’s and the ethnologist’s respective positions regarding the subjects treated or studied, and regarding the role of empathy in the knowledge process, enables us to distinguish the necessity of empathy in these two types of relationship. While the doctor's empathy can be seen as an asset (and not as a sufficient condition to constitute his or her competence, as opposed to the lay representations mentioned above, borrowed from different cultural contexts), the ethnologist's empathy is probably far more problematical. Although ethnologists have to observe and share their subject's experience very closely, and although this affords them a more immediate grasp of the subject, they must also be able to understand another individual, different to the first one, who does not experience things in the same way.
If we examine these diverse research situations we clearly see that understanding has to leave the emotional field if it is to be truly intellectual. The ethnologist's empathy can then contribute to the knowledge process provided that he or she is able, at some point, to move away from it. Perhaps the ethnologist should assume the loser's role in the game of musical chairs; he or she should be able to change places constantly, to sit on different chairs, except on his or her own chair that has been removed. And if he or she is left without a chair, it is precisely to avoid the place of an individual engaged in this play of social relations. This is what will finally enable him or her to be excluded from the game and therefore to analyse it better. To sum up, whereas medical empathy may appear as purely beneficial, the ethnologist's empathy is beneficial for a while only. Although it opens doors that possibly remain closed to other disciplines, it is something that has to be jettisoned at some point in the process.
All these examples highlight the ambiguity that can exist when empathy is conceived of by representing the Other (the one with whom the empathic relationship is realized) as a monolithic entity. In fact, prevailing discourse on the ethnologist's need to develop an empathic relationship with his or her informants contains the implicit idea that informants are an homogeneous whole, occupying a single place only, in which the ethnologist can put him- or herself to feel what this Other (conceived of in the singular and thus from a simplified point of view) might experience. But informants have different experiences and social positions, and are engaged in different power struggles, so that it is impossible to develop empathy with all of them. This is a reality that reflection on empathy generally seems to overlook.
This distancing is necessary if empathy is not to produce an over-simplified view of the object. When the ethnologist puts him- or herself in an informant's place, it enables him or her to see what the informant sees. However, it is then necessary to leave that place in order to see the informant and to see how others see him or her. A way of avoiding this situation of reductive empathy is perhaps dual or even multiple empathy, facilitated by an understanding of the different protagonists in a social relationship. For instance, anyone who wants to study the issue of information and the role of lies in the doctor-patient relationship (a subject on which I am currently working), has to try to put him- or herself 'in the place' of both doctors and patients. Feeling empathy for only one party in this relationship would involve the risk of obtaining biased knowledge on the issue of information.
In medicine the distancing is obtained immediately via objects measuring instruments, tensiometer, stethoscope, radiography, blood tests, clinical examination , all of which enable the subject to remain in the position of an object. In contrast, anthropologists have to make an effort to extract themselves from that place, to reverse the point of view. Empathy then naturally produces knowledge. While doctors do not necessarily need detachment since it will not change their analysis of the situation, for ethnologists the break with empathy is essential to obtain critical knowledge. After savouring its benefits in terms of data collection, they have to give up empathy before analysing that information.
Finally, a fundamental difference lies in the very nature of the relationship in which empathy is felt. The medical relationship is in fact dual: it involves the doctor and the patient, and the knowledge produced by the former is intended for the latter. In contrast, the ethnological relationship is ternary: it involves the ethnologist, the informants, and the public in general (readers, society) for which the knowledge is produced. In other words, in one instance empathy serves the patient; in the other, it serves the public. Not only is the purpose not the same; the actual use of empathy has a different function in the process of knowledge production. We see here the cognitive and therefore heuristic virtue of empathy, provided it can be controlled. Because it can be useful, even necessary, in the research process, but also because it has to be discarded in the knowledge project, empathy in ethnology is perhaps both the condition and the illusion of knowledge.
Bakker H., 1999, Wilhelm Dilthey: Classical Sociological Theorist, Quarterly Journal of Ideology, Vol 22 N°1 & 2: 43-82.
Barnes, A. & P. Thagard, 1997, Empathy and analogy. Dialogue: Canadian philosophical Review, 36: 705-720.
Becker G. 1999, Disrupted Lives. How people Create Meaning in a Chaotic World, Berkeley/ Los Angeles/ London, University of California Press.
Dilthey, W. 1959 . "The Understanding of Other Persons and Their Life-Expressions." Tr. J.J. Kuehl. Pp. 213-225 in Gardiner, Patrick (ed.) Theories of History: Readings in Classical and Contemporary Sources. New York. Free Press. [From Gesammelte Schriften Volume VII: 205-220.]
Fainzang S. & O. Journet, 1988, La femme de mon mari. Ethnologie du mariage polygamique en Afrique et en France, Paris, L'Harmattan.
Fainzang S., 1986, "L'intérieur des choses". Maladie, divination et reproduction sociale chez les Bisa du Burkina, Paris, L'Harmattan (Préface de M. Augé).
Fainzang S., 1996, Ethnologie des anciens alcooliques. La liberté ou la mort, Paris, Presses Universitaires de France (Collection: Ethnologies).
Fainzang S., 2001, Médicaments et société. Le patient, le médecin et l’ordonnance, Paris, PUF.
Gauss Ch. E., 2003, « Empathy », Dictionary of the History of Ideas, The Electronic Text Center, University of Virginia Library, Charlottesville, The Gale Group [http://etext.lib.virginia.edu/DicHist/dict.html].
Haynes R.B., Taylor D. & Sackett D.L. (eds.), 1979, Compliance in Health Care, Baltimore, John Hopkins University Press.
Morris Ch., 1996, Psychology. An Introduction (9th edition), Prentice Hall.
Martinelli, B., 2000, « L’empathie en ethnologie », introductory paper to the Idemec/MMSH Seminar on Empathy (Aix-en-Provence, April 7, 2000).
Snow N.E., 2000, Empathy, American Philosophical Quarterly, Vol. 37, n°1: 65-78.
Spiro H. & alii, 1993, Empathy and the Practice of Medicine, New Haven & London, Yale University Press.
Stein H.F., 1980, "Medical Anthropology and Western Medicine", The Journal of Psychological Anthropology 3/2: 185-205.
Switankowsky I., 2000, “Sympathy and empathy”, Philosophy today: 86-92.
Vaknin S., 1999, “On empathy” [http://samvak.tripod.com/empathy.html].
Weber M., 1971, Economie et société, Paris, Plon.
Wispe L., 1987, “History of the concept of empathy”, in: N. Eisenbers & J. Strayer eds., Empathy and its development, Cambridge, Cambridge University Press: 17-37.
Wispe L., 1991, The Psychology of Sympathy, New York: Plenum Press.
 The idea here is not to compare what happens between the anthropologist and his or her informant, on the one hand, with the identification that doctors are likely to have with their patients during psychotherapeutic or psycho-analytical treatment, on the other. (On this point see Stein (1980) who in this regard examines the concepts of transfer and counter-transfer and the conditions of their applicability.)
 Esperal is a substance (disulfiram) whose properties have the effect of making any subsequent ingestion of alcohol extremely unpleasant. Generally administered in the form of tablets taken daily, it produces what is known as the anti-abuse effect, a set of physical reactions such as congestion of the face, tachycardia, vomiting and low blood pressure.
 Empathy may nevertheless be an obstacle to the adoption of adequate behaviour from a therapeutic point of view. This is the conclusion of Snow (2000) who describes a case in which a doctor's excessive empathy with a patient is seen as preventing him from administering appropriate treatment in an emergency.
 Gauss notes that the development of the method of Verstehen and empathic understanding occurs in Wilhelm Dilthey, Ideen über einer beschreibende und zergliedernde Psychologie,Leipzig, 1894; and in Max Weber, Gesammelte Aufsätze zur Page 89, Volume 2, Wissenschaftslehre, Tübingen, 1920. See also Dilthey (1959).
 While it is true that empathy can hardly be conceived of without its connection to sympathy, this association should not allude to their synonymy nor lead us to conclude that it exists. We can put ourselves in someone's place and understand their thoughts and acts even if we do not like them; conversely, we can experience difficulty in identifying with someone for whom we feel sympathy. As Snow (2000) notes, 'there is a difference between feeling an emotion for someone, and feeling an emotion with someone' (my underlining). The condition of empathy is then the fact of moving closer to what the subject is experiencing or to his/her position in society, rather than to what he/she is.