The role of empathy in psychoanalytic psychotherapy: A historical exploration

Abstract Empathy is one of the most consistent outcome predictors in contemporary psychotherapy research. The function of empathy is particularly important for the development of a positive therapeutic relationship: patients report positive therapeutic experiences when they feel understood, safe, and able to disclose personal information to their therapists. Despite its clear significance in the consulting room and psychotherapy research, there is no single, consensual definition of empathy. This can be accounted by the complex and multi-faceted nature of empathy, as well as the ambiguous and conflicting literature surrounding it. This paper provides a historical exploration of empathy and its impact on the therapeutic relationship across the most influential psychoanalytic psychotherapies: classic psychoanalysis, person-centered therapy and self-psychology. By comparing the three clinical schools of thought, the paper identifies significant differences in the function of transference and therapist’s role. Then, drawing on the different clinical uses of empathy, the paper argues that the earlier uses of empathy (most notably through Jaspers’ and Freud’s writings) are limited to its epistemological (intellectual or cognitive) features, whilst person-centered and self-psychology therapies capitalise on its affective qualities. Finally, the paper provides a rationale for further study of the overarching features of empathy in contemporary psychotherapy research.

Empathy is one of the most consistent outcome predictors in contemporary psychotherapy research. The function of empathy is particularly important for the development of a positive therapeutic relationship: patients report positive therapeutic experiences when they feel understood, safe, and able to disclose personal information to their therapists. Despite its clear significance in the consulting room and psychotherapy research, there is no single, consensual definition of empathy. This can be accounted by the complex and multi-faceted nature of empathy, as well as the ambiguous and conflicting literature surrounding it. This paper provides a historical exploration of empathy and its impact on the therapeutic relationship across the most influential psychoanalytic psychotherapies. In doing so, the paper distinguishes between epistemological (intellectual or cognitive) and affective uses of empathy in clinical treatment. Finally, the paper provides a rationale for further study of the overarching features of empathy in contemporary psychotherapy research. study of the overarching features of empathy in contemporary psychotherapy research.

Issues surrounding empathy in clinical literature
Empathy is one of the most consistent outcome predictors in contemporary psychotherapy research. The therapist's ability to identify and relate to patient's experiences is considered to be the key element of successful psychotherapy process (Elliott et al., 2018). It is also particularly important for the development of a positive therapeutic relationship: patients report positive therapeutic experiences when they feel understood, safe, and able to disclose personal information to therapists (Greenberg et al., 2001).
Despite its clear significance in the consulting room and psychotherapy research, there is no single, consensual definition of empathy in clinical practice. On one hand, empathy is frequently defined as an interpersonal trait that is somehow intrinsic or "given" to individuals, and not necessarily related to therapist's expertise or training (Greenberg et al., 2001;Hill et al., 2017). At the same time, empathy is also conceptualised as a capacity closely related to psychotherapy and counselling training. As a result, such interpersonal actions as understanding, sensitivity, awareness, relation, identification, sharedness, etc., have gradually become incorporated as clinical tools, techniques, and elements of psychotherapy training (see for example, Egan, 2002).
In order to shed light on some of the ambiguity surrounding the contemporary definition of empathy, this paper will present a historical analysis of empathy as a clinical tool in psychoanalytic psychotherapies. The first clear definition of empathy as a clinical tool that is part oftherapist's capacity was first formalised by the founder of humanistic psychology, Carl Rogers (1980): "[It is] the therapist's sensitive ability and willingness to understand the client's thoughts, feelings and struggles from the client's point of view […] to see completely through the client's eyes, to adopt his frame of reference" (p. 85). Similarly, psychoanalyst and founder of self-psychology, Heinz Kohut (1984), defined empathy as "the capacity to think and feel oneself into the inner life of another person" (p. 82).
Though both definitions appear to be strikingly similar at first glance, historical and clinical analysis shows that both authors had significant differences in their views on empathy, as well as its function and utility in psychotherapy (Kahn & Rachman, 2000). For Rogers, the role of empathy in clinical practice is not necessarily different from ordinary human relationships; acceptance, understanding, and genuine attention stand as the basic tenets of his person-centered therapy. For this reason, Rogers' ideas about empathy continue to permeate today's self-help groups, educational bodies, and politics (Caspary, 1991). For Kohut, however, empathy is an operational element that has a specific function in clinical practice: he defined it as an instrument by which therapists listen and attune to patients but also engage with the acquired material in a way that is beyond "regular" understanding (e.g., through clinical interpretation). Kohut wrote that merely adopting empathy without paying substantial attention to other clinical techniques results in patients' "temporary improvements" (Kohut, 1973(Kohut, / 1978. Whilst the two authors developed their theories at the same time and the same place (University of Chicago, 1945Chicago, -1957, the undeniable parallels between Rogers and Kohut are also shrouded by historical controversies, particularly the division between the departments of psychiatry and psychology (Kahn & Rachman, 2000).
But the history of empathy in clinical practice does not begin with Rogers or Kohut. Indeed, psychoanalysis has had an enormous interest in empathy since its inception. It is Freud, after all, who developed and pushed what one of his collaborator Josef Breuer's patients, Anna O., called "the talking cure" (Breuer & Freud, 1895. The first clinical application of psychology, psychoanalysis, to Freud was very much a scientific deal (although of course there are ongoing debates on the exact model of science that Freud had in mind for psychoanalysis) (Freud, 1914(Freud, , 1920. However, Freud was also very clear that psychoanalysis is a relational activity: it is "the talking cure", after all, not a "cure by interpretation". These relational dynamics are most evident from transference: a process by which patients transfer some of their past interpersonal experiences (usually with parental figures) onto therapists. These transferred feelings can range from extremely affectionate to hostile and ambiguous. In turn, therapists can access patients' life histories by re-experiencing their prototypic past relationships through patients' acting out (Levy & Scala, 2012). Thus, the process of transference is far more than the summation of verbal exchanges and narrative stories told by patients to their therapists-it is a relational way of showing different interpersonal reactions, emotions, and (unmet) desires that were first initiated in early childhood, and subsequently manifesting in later life. What happens on the receiving end of these transference experiences, however, depends on the particular therapeutic school of thought. Should we prioritise interpretation over understanding or seek to strike a balance between the two if possible?
Unsurprisingly, the acknowledgment of the vehicular therapeutic role of transference has brought a variety of empathy-related issues to the table. This is particularly evident in therapists' countertransference feelings, which, just like transference, can range from anxiety, anger, and disappointment to affection, lust, and attachment (May, 1986;Winnicott, 1949). But as Celenza (2010) notes, "when the affective violence leans erotic, countertransference becomes once again a taboo" (p. 175). Patients are meant to-or so the theory teaches us-to have feelings for their therapists because that is precisely the function of repetition compulsion, a compulsion to repeat or re-enact conflictual childhood themes in transference (Freud, 1922). This standard is not reciprocated when it comes to therapists' feelings. Freud was very clear about this: the therapist evokes transference love in order to "cure the neurosis. For [the therapist], it is an unavoidable consequence of a medical situation, like the exposure of a patient's body" (Freud, 1915, p. 179). And because transference is likened to a medical situation, the therapist's role, according to Freud, should be similar to that of a scientist or a physician: "The struggle between physician and patient, between intellect and the forces of instinct, between recognition and the striving for discharge, is fought out almost entirely over the transference-manifestations" (Freud, 1912, p. 322). As such, the classic Freudian view urges that all therapists' feelings must be considered strictly as a part of the clinical process; they are not to trespass beyond the boundaries of scientific thinking.
But of course, the "scientific" or "medical" story of transference is not as linear as it may appear from the above citations. One only needs to take a quick look at the clinical literature that discusses empathy, therapeutic relationships and challenges evoked by countertransference processes to see that therapists, of all times and of all clinical perspectives, struggle to maintain an empathic or relational attitude toward their patients on one hand, and keeping up a more detached scientific stance on the other (Benjamin, 1994;Coen, 1994;Gorkin, 1987;Gabbard, 1994Gabbard, , 1996Rabin, 2003;Sharma & Fowler, 2016). Different stances on empathy have also contributed to the emergence of different kinds of psychoanalytic and psychotherapy schools; Kohut's self-psychology and Roger's person-centered therapies are, in a few significant ways, very different from the classic Freudian psychoanalysis precisely because of this reason. This paper will discuss the historical significance of empathy, and the divergent views employed toward it across different forms of psychoanalytic psychotherapies. In doing so, the paper seeks to demonstrate how the different views on empathy shaped our understanding of the role of the therapist, the therapeutic relationship, and the transference process. The paper will first outline empathy's clinical roots through the philosopher and psychiatrist Karl Jaspers' work and proceeds with its development in three major psychoanalytic schools of thought: Freud's psychoanalysis, Rogers' person-centered therapy, and Kohut's self-psychology.
Finally, the paper will provide a brief rationale for further study of empathy in contemporary psychotherapy research. Whilst there are significant variations between different clinical schools of thought and their definitions of empathy, all forms of psychotherapy are committed to understanding the experiential worlds of other people. As such, whatever may be the differences in the formal applications of empathy, complex interpersonal influences, non-verbal exchanges, and therapistpatient relationships will always play a crucial role in psychotherapy processes and outcomes (Dekeyser & Elliott, 2009). For this reason, it is important to have research that addresses both the multi-faceted nature and the historical controversies of empathy in clinical literature.

Main questions
(1) What is the historical role of empathy in psychotherapy?
(2) What is the clinical function of empathy in different forms of psychoanalytic psychotherapies?
(3) What are some of the issues caused by the ambiguous/conflicting definition of empathy in clinical practice?
3. In pursuit of clinical objectivity: empathy in the works of Jaspers and Freud Karl Jaspers, a psychiatrist-turned-philosopher, was one of the first to stress the psychological importance of subjective mental experiences (Jaspers, 1912). According to Jaspers, mental experiences and symptoms fall into two general classes: objective and subjective. The first category consists of symptoms and experiences that are accessible to us through visual perception, specific clinical tests, and patients' verbal reports. For example, we can observe patients' spontaneous movements, actions, appearances; we can (to varying degrees) test patients' intelligence, memory, and perception; and lastly, we can make rational observations from patients' narratives, reports, and dialogic engagement. Thus, Jaspers' first category subsumes what is now known as clinical "signs" (Oulis, 2014): clinical features of mental disorders that are either directly observable or can be observed through standardized techniques.
Jaspers' second category of subjective mental symptoms and experiences cannot be accessed in the same rational manner as the first one. According to Jaspers (1912), this is because "subjective symptoms cannot be perceived by the sense-organs but have to be grasped by transferring oneself, so to say, into the other individual's psyche, that is, by empathy" (p. 314). This is a crucial point in Jaspers' work: he acknowledges that clinical practice cannot be driven solely by intellectual effort, and that it needs to entail empathy (Einfühlung) and empathic understanding. In other words, it is not enough that we have objective measurements of mental experiences; for Jaspers, a genuine understanding of the patient requires that we have a phenomenological grasp of the patient's events "as from within" (Jaspers, 1912, p. 326), which means that there is something it is like to experience the world from each patient's perspective.
Jaspers' conception of empathy is relevant to our discussion in several ways. Firstly, by differentiating empathic understanding from objective (or scientific) understanding, Jaspers developed a dynamic view of mental disorders. Since each mental experience is personal, ever-changing, and dependent on a variety of contextual factors (temporal, social, cultural, etc.), clinical diagnostics can only be guiding and never fully representative of the patient's experiences. It is for this reason that Jaspers, like Freud, conducted lengthy patient case study notes: he was most interested in what his patients had to say about themselves, relying primarily on patients' self-reports. Therefore, Jaspers' approach prioritised the task of listening. However, Jaspers' listening is quickly followed by analysing patients' expressions, behaviours, and gestures, questioning patients' selfreported experiences, and studying analogies and metaphors in patients' speech.
The subjective patient reports, according to Jaspers, must be translated or reduced to reality; their subjective experiences must become a form of data accessible for the therapist. In turn, the therapist can engage with this data through the process of "actualization": drawing from their own cognitive and emotional resources and experiences, therapists can begin to understand the situation of another individual (Oulis, 2014). This is a crucial element of Jaspers' conception of empathy: he does treat it as an exclusively relational or affective understanding of how an individual feels. Instead, Jaspers' empathy entails awareness of one's own situation and experience by maintaining a hermeneutic distance from the other individual's experiences. This is because Jaspers believes that empathic understanding of the patient's experiences, however successful it may be, will always remain "second-hand"; we can only experience what it is like to feel like a particular patient indirectly because we rely entirely on the patient's self-reports. This is further articulated by Gatta (2014): The psychiatrist cannot possibly presume to experience what the patient is experiencing first-hand. By emphasizing the second-hand nature of the psychiatrist's experience […] Jaspers highlights something fundamental to his phenomenological approach: reliance primarily on the patient's own account and presentation, not on the psychiatrist's own response to those experiences. (p. 1005) Herein, certain issues arise in Jaspers' conception of empathy. These issues, as I will argue further, re-emerge (albeit in different ways) in psychoanalytic accounts dealing with empathy and therapeutic relationships. Although Jaspers stresses the importance of a "second-hand" approach to patients' irreducible subjective experiences, his account of empathy-and particularly, empathy as an epistemological tool in the clinical setting-appears to be ambiguous and conflictual. Jaspers makes numerous claims that therapists' empathic access to patients' mental experiences is a direct process, analogous to direct observation in the natural sciences (Oulis, 2014). This argument relies on what Jaspers calls the "immediate grasp of expressive phenomena" (Jaspers, 1919, p. 326); the idea that empathic understanding brings with itself a kind of immediacy of intelligibility. For Jaspers, this means that when the therapist empathically relates to and makes a meaningful connection with their patient, they become struck with this knowledge; it becomes immediately evident, as it sometimes is when an important observational discovery is made in the natural sciences.
However, elsewhere in his work, Jaspers makes quite opposite claims: that relying on such an immediacy is dangerous, and that, as a result, we can fall short of genuine understanding altogether. In order to support this claim, Jaspers describes empathic understanding as an "exploration by direct questioning of the patients" that requires "our [therapists'] guidance" (Jaspers, 1912, p. 320). In this part of the argument, he notes that subjective experiences also "include those mental processes which we have to infer from fragments of […] previous kinds of data" (Jaspers, 1912, p. 314). These two definitions of empathy appear to be in great conflict in Jasper's work. It is clear from the second definition of empathy as a form of indirect understanding that Jaspers became increasingly concerned with prejudice and preconceptions that accompany all scientific work, and perhaps most of all, clinical work that relies on empathy.
The latter is one of many Jaspers' contentions with Freud's work: he was sceptical of Freud's clinical interpretations of patients' unconscious material, directly critiquing Freud for producing fictionalism in his case studies (Monti, 2013). For Jaspers, the conditions necessary for psychoanalytic treatment-namely, the "asymmetrical" communication between the expert psychoanalyst and the instinct driven patient-are detrimental to the establishment of a genuine empathic connection with the patient. Furthermore, the "asymmetrical" therapeutic relationship poses issues to the generation of knowledge that is factually representative of patients' experiences.
The latter criticism partly derives from Jaspers' own interest in the clinical practices of institutional psychiatry, and especially in the psychiatric treatment of psychotic and schizophrenic patients. It is in this context of clinical treatment that our ability to understand and relate to patients is shown to be painstakingly limited. Since empathic understanding is based on therapists' successful "actualization" of patients' subjective experiences as they are expressed by the patients themselves (Jaspers, 1919), it can only reveal something about each patient's conscious mental life. As such, Jaspers' interest in schizophrenia is based on its outside-of-consciousness nature: because the schizophrenic patient considers their mental life from an interventionist point of view rather than relying on their own agency, establishing an empathic connection becomes nearly impossible. Jaspers called this "ununderstability" (Jaspers, 1912): one cannot empathically understand or relate to an individual who is severely limited in their expression of consciousness. Ununderstability itself acts as an indicator of patient's impaired consciousness and delusional thinking; however, to make any firm clinical assessments about their unconscious mental life is, according to Jaspers, to thread on dangerous ground of fictionalism, prejudice, and pseudotherapy.
It is not hard to see why psychoanalysis, and particularly Freudian psychoanalysis, became Jaspers' arch enemy. Despite the fact that, at one-point, Freud conducted much of his clinical and theoretical work in one of the largest European hospitals at the time, Salpêtrière, he had almost no interest in institutional psychiatry or the clinical (mal)practices that took place there. This is exemplified by the following passage from Freud's case of Mrs Emmy von N. in "Studies on Hysteria" (Breuer & Freud, 1895: [The patient] brought out new fears about asylums-that people in them were treated with douches of ice-cold water on the head and put into an apparatus which turned them round and round till they were quiet […] I appealed to [the patient's] good sense and told her she really ought to believe me more than the silly girl from whom she heard the gruesome stories about the way in which asylums are run. (p. 61) This passage serves as a good example for considering Jaspers' criticism of the "detrimental" psychoanalytic clinical setting, in which the medical authority ascribed to the psychoanalyst is so high that it prevents any empathic engagement with patients' actual-consciously expressedexperiences. The patient is clearly signalling through her verbal reports that she is horrified by institutional asylums. Freud noted elsewhere that the patient even "clenched her hands in horror; she saw all this before her eyes" (Breuer & Freud, 1895, indicating a physical sign of resistance to the very idea of being admitted to an asylum. Freud reassures the patient as a clinical authority, an expert who knows the inside-out of the psychiatric practice; however, he does not seek to make a meaningful connection with the patient's emotions, to re-experience them himself (even if it is a "second-hand" experience). At this point, Jaspers criticises the excessive importance given by Freud to unconscious and infantile sexual life so much so that his criticism "reaches the point of wishing the extinction of psychoanalysis as an obstacle to human freedom" (Monti, 2013, p. 29). For Jaspers, Freud's psychoanalysis has failed on both ends: it failed to engage with the patients empathically, and it failed to produce objective and factual clinical knowledge about patients' psychic life. That is to say, Freud was neither subjective enough in the way he engaged with his patients relationally (thus undermining the phenomenological importance of empathy), nor was he objective enough to produce clinical data that contains "the unprejudiced direct grasp of the mental as it is" (Jaspers, 1912, p. 318).
Jaspers' criticism of Freud is interesting, seeing that he himself struggled with a similar dialectic of objectivity and subjectivity in clinical practice, which subsequently resulted in his shift toward philosophy. The ambiguity in Jaspers' definition of empathy stems from the very same clinical struggle that, as I will argue further, Freud experienced with transference: on one hand, there is a deep acknowledgment of patients' subjective experiences and their engagement with the therapeutic relationship, and on the other, there is a clear scientific pursuit for clinical objectivity. As such, even though Jaspers considered empathy to be the key piece in clinical treatment of patients, he nevertheless maintained that there are severe limitations to 1) patients' ability to express their mental experiences, and 2) therapists' capacity to empathically engage with the said experiences without projecting their own social stereotypes and prejudices. This is why Jaspers' empathic understanding becomes increasingly categorically limited: as we saw with schizophrenia, Jaspers acknowledges that only some mental experiences are fully understandable (similar to therapists' own experiences), but most will be only partly understandable (dissimilar in their intensity to therapists' experiences) or completely ununderstandable. Therefore, Jaspers' conception of empathy remains largely as an epistemological (rather than affective) tool: it does not provide an immediate or direct understanding of patients' experiences (despite Jaspers claiming this in some areas of "General Psychopathology" (Jaspers, 1963(Jaspers, /1997, and it still relies on additional scientific observation and therapists' capacity to appropriately "actualize" patient experiences (Oulis, 2014). It is for these reasons that Jaspers famously claimed in his book "Psychology of Worldviews" (Jaspers, 1919) that "even the most elevated psychological understanding is not a loving understanding" (p. 127).
4. "The talking cure": transference, therapeutic relationship, and the role of the therapist Earlier in this paper, I have referred to Freud's approach to transference as a "consequence of a medical situation" (Freud, 1915, p. 179), in which the therapist functions as a physician. Freud chose the medical analogy here because the role of the psychoanalytic therapist requires a similar detachment from the patient in order to properly engage with the unconscious transferencemanifestations. This detachment is necessary for two reasons.
Firstly, the therapist cannot rely on the patient's direct responses and engagement with the therapeutic process precisely because the patient is repeating or re-enacting early childhood themes in the transference relationship (Freud, 1922). This re-enactment, according to Freud, is unconscious; the libidinal patterns formed in early childhood have been "held up in development" and subsequently "withheld from the conscious personality and from reality" (Freud, 1923, p. 313). This is particularly clear with patients who experienced parental deprivation or neglect; the unmet libidinal needs once experienced with parental figures are likely to re-emerge in subsequent interpersonal relationships, including therapeutic relationships. Here, Freud (1912) helpfully relies on Jung's example of the father-imago: the patient may connect the therapist to their father (and the interpersonal similarities may be realistic in some sense), however, this connection is not based on conscious ideas but by those that are "under suppression". Therefore, the success of psychoanalytic treatment is largely based on the therapist's ability to identify the patient's unconscious influences, anxieties, and defences, and to demonstrate these unconscious processes to the patient (to make them conscious). In this sense, classic Freudian psychoanalysis privileges interpretation over relation to the patient; the therapist's interpretation remains (largely) as an object outside of the patient's subjectivity (Gabbard, 1997).
The second reason for the therapist's detachment has to do with maintaining a clinical (nonpersonal) relationship with the patient. Freud is very clear that, in the cases of positive transference relationships in which patients demonstrate affectionate attitudes toward their therapists, transference love and love outside treatment are two different experiences and that they should not be conflated. In "Observations on Transference-Love: Further Recommendations on the Technique of Psychoanalysis" (Freud, 1915), Freud states that transference love: … has its distressing and comical aspects, as well as its serious ones. It is also determined by so many and such complicated factors, it is so unavoidable and so difficult to clear up, that a discussion of it to meet a vital need of analytic technique has long been overdue. But since we who laugh at other people's failings are not always free from them ourselves, we have not so far been precisely in a hurry to fulfil this task. We are constantly coming up against the obligation to professional discretion-a discretion which cannot be dispensed with in real life, but which is of no service in our science. (p. 159) In this passage, Freud implies that there is a real, pressing limitation of a professional discretion to patients who may not wish to be identified by their neurotic symptoms in public discourse. However, Freud also notes that therapists themselves "are not always free" from the same kind of "failings" their patients experience. This seems to be the real crux of the issue; that even psychoanalytic experts positioned on the higher ladder of what Jaspers called the "asymmetrical" clinical arrangement are not infallible and may conflate transference love with real love.
The danger of treating transference love as real is partly caused by the very technique of psychoanalytic treatment. As Celenza (2010) argues, clinicians must be able to understand and feel patients, especially when it comes to their most sensual and vulnerable experiences: "Our patients often need to bring their sexuality into the foreground and not just dance with us but tango as well; we need to be pushed and pulled; held in open embrace with a seemingly dissociated body" (p. 180). Freud acknowledged this as well by stating that it is absolutely crucial for therapists to maintain patients' transference-manifestations: "The patient's need and longing should be allowed to persist in her, in order that they may serve as forces impelling her to do work and to make changes" (Freud, 1915, pp. 164, 165). This is perhaps one of the first moments where Einfühlung (or as it is often (mis) translated, "sympathetic understanding") emerged in Freud's work: the therapist must establish an empathic attachment with the patient in order to be connected with one of the patient's prototype figures (Freud, 1913). Without this empathic stance and transference connection, psychoanalytic work cannot begin or progress.
However, this clinical principle and method come with a hefty technical price. Because love and sexuality are often the central themes of clinical treatment, they challenge any possible therapeutic detachment. The reception of complex patient feelings-and in particular, affectionate or sexual feelings-can lead to a violation of professional and personal boundaries. A further example from Celenza's (2010) article demonstrates the difficulty of balancing an empathic, relational, playful attitude toward a patient with a detached, expert-like clinical mode. She and her patient "had a special way of playing-'deep play' that had sexual overtones […]" (p. 181), which gradually led to an immense discomfort on her part as a therapist: "The discomfort was more in the way of a nagging guilt-was I enjoying him too much? Did our play cross over to flirtation in a way that seductively offered something I could not and would not deliver? […] Was I behaving with his best interests in mind?" (p. 181). Celenza notes that all these elements can be part of an empathic attachment with the patient; however, the therapist must be able to think about these exchanges from a detached, objective point of view. This is precisely the kind of detached balance Freud advocates in "Observations on Transference-Love" (Freud, 1915) and "Constructions in Analysis" (Freud, 1937). When it comes to empathy and relational dynamics with patients, Freud suggested extreme caution: to neither gratify transference love, nor to suppress it. As discussed previously, Freud saw the necessity of maintaining transference as a methodological tool that allows the therapist to re-experience patients' prototypical relationship patterns.
Differently from Jaspers' "second-hand" empathy, however, Freud's transference has to do with re-experiencing mental symptoms that are unconscious to the patients. As such, psychoanalysis relies largely on therapists' approach to these unconscious experiences rather than patients' expressions of them. And it is precisely because the therapist plays such an active role in reconstructing patients' unconscious material that they must pay a hefty technical price: they must not respond to transference in a way that would violate the boundary between a clinical relationship and a personal relationship. If this boundary is violated and the patient's erotic transference (or therapist's erotic countertransference) is gratified, then the conditions for psychoanalytic treatment no longer exist. These conditions, as Pinsky (2014) reminds us, are fundamental for the production of a cure (interpretation): "The analyst's interpretation is another form of 'No, we may not,' the same proscription that ignites desire for both people" (p. 462). The patient's idea of a boundary violation is productive for maintaining the transference relationship; however, it must continue existing as an idea, a phantasy, without ever transgressing the psychoanalytic arrangement.
As such, Freud's understanding of transference promotes a detachment from the patient in two ways: 1) the therapist must remain substantially detached from the patient in order to be able to observe (and interpret) their unconscious mental experiences, and 2) the therapist must maintain a clinical (non-personal) transference relationship with the patient in order to preserve the psychoanalytic arrangement. In this way, Freud's definition of transference is paradoxical and ambiguous in a similar way to Jasper's empathy: there is an attempt to understand, feel, and even maintain patients' subjective experiences, but also to represent them in an objective and clinically accurate way (which is where the therapist's empathic involvement becomes limited).
The subjectivity/objectivity paradox in Freud's and Jaspers' work can be described in the words of psychoanalyst Max Hernandez: "Love is the motor of the analytic cure as well as the main obstacle to it" (Hernandez, 1993, p. 98). Therapists strive to understand and empathically relate to their patients; however, as we now know from empirical research on empathy, our human capacity to empathically or intuitively understand other human beings is fairly low at its best (Stueber, 2006). In the clinical realm, this is also accompanied by prejudices and preconceived theoretical assumptions, further limiting our genuine capacity to understand and experience the world from patients' perspectives (Oulis, 2014).
One may argue, however, that Freud goes one step further than Jaspers. Because Freud's focus is on the unconscious rather than conscious mental experiences, the patient appears to be heavily dependent on the psychoanalyst and their interpretive work. Freud controversially stated that even if an interpretation is wrong, there will still be "no change in the patient; but if it is right or gives an approximation to the truth, he reacts to it with an unmistakable aggravation of his symptoms and his general conduct" (Freud, 1937, p. 265). For this reason, Freud, unlike Jaspers, does not prioritise relying on patients' verbal reports and expressions: since patients are often unconsciously reenacting their early experiences, most of their direct responses to treatment are forms of resistance. This resistance is not caused by patients' inability to express their consciousness or their delusional thinking; rather, it is a common ego defence against a subconsciously perceived threat. The psyche resists access to the unconscious material, collaboration in the therapeutic relationship, and it even resists the obtainment of a therapeutic relief. As a result, the psychoanalytic method largely depends on indirect observation of reactions, aggravations, and anxieties that may be easy to observe, or so repressed that they may emerge only through free associations, dreams, or parapraxes. From this point of view, Freud treats all mental experiences as ununderstandable: even the most plain, direct responses are to be treated with careful consideration and additional observation. This is also why it may seem that the methodological significance of transference overrides its relational meaning. After all, Freud states that we must maintain patients' painful and traumatic experiences until we find an appropriate cure: "Cruel though it may sound, we must see to it that the patient's suffering, to a degree that is in some way or other effective, does not come to an end prematurely" (Freud, 1919, p. 162-163). In other words, the clinical quest of addressing patients' unconscious afflictions overrides the suffering patients may experience in the process of clinical treatment. Herein, empathy seems to be scarce in Freud's priority of interpretation versus relation.
But to argue that empathy is insignificant in Freud's psychoanalysis would be a grave misunderstanding. Freud placed an enormous significance on Einfühlung in his work; however, as Pigman (1995) points out, this term has been often mistranslated to "sympathy" or "sympathetic understanding", both of which are semantically different terms from empathy (Gerdes & Segal, 2011). Freud was clear that psychoanalysis should not promote sympathising with patients, especially in the context of feeling pity or sorrow. Instead, he frequently wrote about being "guided by an intuitive perception [intuitive Einfühlung]" (Freud, 1925, p. 237). This kind of intuitive knowledge shows the therapist when to keep silent and await further therapeutic developments, and when to disrupt the silence in order to reduce patients' suffering and uncertainty (Ferenczi, 1928).
Furthermore, Freud considered empathy to be the key element in understanding patients' subjective experiences. He recognised that a purely intellectual understanding of patients' pathological ailments is not enough; that, in order to be a good clinician, one must also be able to take up a mental attitude similar to the patient's (Freud, 1912). In doing so, Freud considered the limitations around empathising with specific patients. For example, he wrote about his own inability to understand children and members of primitive tribes: "It is not easy to feel one's way into primitive modes of thinking. We misunderstand primitive men just as easily as we do children, and we are always apt to interpret their action and feelings according to our own mental constellations" (Freud, 1912(Freud, -1913. Therefore, just like Jaspers, Freud was concerned about theoretical and personal projections that can take place when we fall short of genuine empathic understanding. Freud saw empathy as simultaneously the drive and the challenge of psychoanalytic practice. On one hand, without empathy, observation of patients' subjective and unconscious experiences would be impossible; on the other, patients' subjectivities may still appear "foreign to our ego" (Freud, 1912, p. 110). This adds yet another layer to Jaspers' concerns over the limitations of understanding other people's consciousness: the unconscious ununderstanding. For Freud, this capacity (or lack thereof) to understand other individuals extends beyond consciousness, by which he means that, instead of empathising with our patients, we may be unconsciously projecting our own feelings onto them all along. Freud seeks to respond to this problem by outlining the technical steps necessary for maintaining transference relationships and avoiding boundary transgressions. In doing so, however, Freud almost single-handedly focuses on the intellectual features of empathy whilst neglecting the affective ones (Pigman, 1995). Thus, much like Jaspers, Freud maintained the use of empathy on a strictly epistemological level: because of its suspiciously affective nature (Freud, 1927), empathy must be "kept in check" by adopting a detached therapeutic stance toward transference relationships otherwise filled with the most sensual and fundamental of human emotions.

Empathy as an affective tool in clinical practice
The empathy-versus-clinical objectivity dialectic contributed to a variety of different perspectives in psychoanalytic theory. As could be seen from Jaspers' and Freud's accounts of empathy, their active pursuit for clinical objectivity led to concerns of arbitrary, subjective, and prejudiced forms of treatment resulting from empathy's affective features. However, the very notion of clinical objectivity quickly fell under scrutiny amongst many post-modern thinkers. With the advent of relativism, perspectivism, and intersubjectivity, the ideas of a fundamental clinical truth and a "blankscreen analyst" (Gabbard, 1997, p. 15) ultimately fell out of favour. Instead, a variety of different scenarios have been put forward about the roles of empathy, therapeutic relationship, and transference. For instance, relativist authors proposed that there is a "crisis of representation" (Marcus & Fischer, 1999), by which they argue that interpretations of each patient's experiences will always remain flawed due to our inability to perceive the world from another's point of view. A relativist view is pessimistic in our capacity to escape our own social, personal, and cultural filters and thus represent the world "as it is" (Lightburn & Sessions, 2005).
In contrast, the perspectivist view argues that the idea of a clinical reality is a useful one (Aron, 1996;Orange, 1995). Although constructivist authors acknowledge the inherent dangers of relying on patients' and therapists' subjectivities (and the social projections involved in them), they nevertheless maintain that subjective experiences are crucial for the construction of "clinical truth" (Stern, 1992). That is to say, "clinical truth" is not perceived as something "out there", an object that can be slowly accumulated through observation; rather, it became seen as a construction of our shared perspectives and experiences-our joint subjectivities. This paper will not go into detail about the modern/post-modern dialectic of objectivity and subjectivity in psychoanalytic treatment. However, being aware of this shift in the discourse of clinical objectivity is useful when thinking about the emergence of other kinds of psychoanalytic psychotherapies that utilise empathy for its affective rather than intellectual features. Whilst there is a great deal of conceptual diversity in post-modern perspectives critiquing clinical objectivity, they do seem to agree on one fundamental thing: clinical treatment is "inescapably subjective" (Renik, 1993, p. 560). This means that the therapeutic relationship cannot be detached or neutral; the therapist's subjectivity will inevitably enter clinical observations and interpretive material. And as most of these perspectives contend, this inevitable subjectivity need not be unproductive or incompatible with the clinical task of producing a cure. However, it means that a certain reconceptualisation of the classic psychoanalytic cure was required to reflect this subjectivity. This subsequently led to a series of developments to Freud's original psychoanalytic arrangement: the transference struggle between the physician's intellect and the patient's forces of instinct became a struggle of extending and fulfilling the patient's self-identity.
It is because of this shift that the definition of empathy moved toward a different direction in psychotherapy, which this paper will call affective. Though the term affective empathy is typically used in neuroscience to define the capacity to infer and respond to another individual's feelings (Healey & Grossman, 2018), it may also serve purpose in our discussion of empathy in clinical contexts. In neuroscience literature, affective empathy is clearly distinguished from cognitive empathy (Healey & Grossman, 2018). Similarly to Jaspers' and Freud's early definitions of empathy as an intellectual and epistemological tool, cognitive empathy is defined as the ability to model others' emotional states. The difference between affective and cognitive empathy is a difference between feeling what another individual feels and understanding what another individual feels, respectively. Therefore, affective empathy is based on not only the recognition of another individual's emotional state but also the adoption of this emotional state as a response.
Recognising and adopting patients' emotional experiences were central to both Rogers' and Kohut's definitions of empathy. As explained earlier in the paper, they both define empathy as the therapist's capacity to adopt patients' experiences; to see the world through another's eyes (Kohut, 1984;Rogers, 1980). Rogers sought to put forward a more rigorous definition of empathy in "A Theory of Therapy, Personality and Interpersonal Relationships as Developed in the client-centered Framework" (Rogers, 1959): The state of empathy, or being empathic, is to perceive the internal frame of reference of another with accuracy and with the emotional components and meanings which pertain thereto as if one were the person. Thus, it means to sense the hurt or the pleasure of another as he senses it and to perceive the causes thereof as he perceived them, but without ever losing the recognition that it is as if I were hurt or pleased and so forth. If this 'as if' quality is lost, then the state is one of identification (p. 210-211).
Later in his writings, Rogers clarified that empathy is not a state, but a process by which one enters another individual's private perceptual world (Rogers, 1975). To capture this process, Rogers refers to a variety of felt emotions that therapists experience when they empathise with their patients: he uses the words "sensitivity", "fear", "rage", "tenderness", "confusion" (Rogers, 1975, p. 4). These emotions, according to Rogers, are not to be transmitted through interpretations or other clinical techniques immediately; this would potentially jeopardise the therapist's affective experience of the patient's inner world. Prejudice or preconceived theoretical assumptions about patient's feelings would lead to an incongruent empathic response: it would become inappropriate for the mood and content of patient's statements (Truax & Carkhuff, 1967).
Even from a brief theoretical review, it is clear that Rogers has a very different conception of empathy from Freud. Firstly, empathy is not ambiguous in Rogers' work and his understanding of psychotherapy: "The ideal therapist is first of all empathic" (Rogers, 1975, p. 5). Although Rogers, like Freud, expresses concern over becoming too empathic in one's relation to the patient (Rogers responds to this by differentiating empathy from identification), he nevertheless maintains that empathy is not merely an epistemological (or intellectual) tool. Therefore, one must not be "detached" (in ways of therapeutic abstinence or neutrality). On the contrary, Rogers writes about engaging with each patient via unconditional understanding: "There have been speculations to the contrary, that an appealing or seductive client might be responsible for drawing understanding from the therapist. The evidence does not support this" (Rogers, 1975, p. 5). As such, an empathic therapist-one who is capable of listening and understanding patients whilst maintaining an awareness of their own point of view and experiences-is equated by Rogers to a responsible therapist.
From this comes the second difference: there is far less concern about empathy's affective features in Rogers' work. His qualitative language seeks to demonstrate the emotions and feelings as they are experienced by both therapists and patients in psychotherapy. Freud's descriptions of empathy-even when he describes it as a prerequisite for interpretation (Freud, 1913)-remain at a largely cognitive level of thoughts and beliefs.
This also pertains to differences in clinical technique. The task of listening is, of course, central to all schools of psychotherapy. The difference in Rogers' person-centred therapy and classic Freudian psychoanalysis, however, lies in the prioritisation of listening over interpreting patients' experiences. For Freud, interpretation is the key tool in psychoanalytic treatment, and as such, listening to patients' life experiences remains to be a prerequisite for putting forward the interpretive material (Freud, 1937). In this sense, the clinical technique of listening is of limited value to the overall process of treatment: as could be seen from Freud's discussions of transference, the therapist should be cautious about engaging with the patient's direct responses because they are (unconsciously) enacting transference manifestations. For Rogers, the task of listening is clearly more important than the task of interpreting. In fact, Rogers (1942) appreciation of patients' subjective experiences led him to advocate a mode of listening that, to a certain degree, excludes interpretation: The more accurate the interpretation, the more likely it is to encounter defensive resistance. The counsellor and his interpretations become something to be feared. To resist this temptation to interpret too quickly, to recognize that insight is an experience which is achieved, not an experience which can be imposed, is an important step in progress (p. 195-196).
Whilst most therapists consider empathic listening as an important element of their clinical technique, Rogers (1975) asserts that this is often not carried out in practice: "Though therapists regarded empathic listening as the most important element in their ideal, in their actual practice they often fall far short of this" (p. 6). For this reason, he argued that therapists are unable to assess their own degree of empathy in therapeutic relationships and thought that patients should be able to let therapists know-directly-whether or not they are being understood adequately. For this reason, Rogers did not regard empathy as something that is innate or "given"; he thought that everyone can learn how to be empathic toward other individuals from other empathic persons. For example, one can learn empathy from their supervisors, teachers, parents, and other individuals who are able to genuinely listen and provide sensitive understanding. The realm of psychotherapy just happens to be one instance in which empathy can be used for building positive relationships with other individuals.
Kohut shared many of Rogers' concerns about the lack of genuine empathic understanding and listening in clinical practice. Whilst initially an ardent Freudian (Strozier, 2001), Kohut gradually moved away from many traditional psychoanalytic concepts, including Freud's structural theory of the id, ego, and superego. This has much to do with Kohut's increasing appreciation of patients' selfhood: he emphasized the importance of early caregiving experiences to the development of a cohesive sense of self (Marmarosh & Mann, 2014). Kohut referred to these developmental needs as selfobject needs that make up the self-structure of cohesion and steadiness (Kohut, 1984). He specified three categories of selfobject needs: mirroring, idealization, and twinship. Mirroring involves facilitating self-esteem, ambitions, and assertiveness; idealisation fosters a sense of ideals and internal values that are introjected from an idealised image; twinship encourages one's sense of acknowledgment, feelings of belongingness and connection to other individuals.
According to Kohut, the development of a cohesive and stable sense of self depends on how successfully mirroring, idealization, and twinship needs are met. Although each of the three selfobject needs involves a range of interpersonal actions, they all depend on a sufficient exposure to empathy. If exposure to empathy is insufficient or lacking in early development (e.g., through parental neglect), then the selfobject needs cannot be met. For Kohut, this is the root of all psychopathology: "By failing to provide appropriate empathic feedback during critical times in a child's development, the child does not develop the ability to regulate self-esteem, and so the adult vacillates between an irrational overestimation of the self and feelings of inferiority" (McLean, 2007, p. 41).
The theoretical focus on selfhood and empathy had a significant impact on Kohut's clinical work, particularly with narcissistic patients. Kohut observed that narcissistic patients' inability to develop transference relationships is often caused by parental deprivation and lack of empathy in the first years of development. As a result, narcissistic individuals are incapable of recognising emotions and responding to them congruently; they cannot regulate their self-esteem without other people's validation; and they struggle to internalise values and ideals due to unavailable parental objects.
Kohut insisted that classic psychoanalytic approach with its heavy focus on interpretation will not be effective for narcissistic patients with the aforementioned problems. He cautioned that interpretations or direct therapeutic interventions will be perceived as insults, criticism, and rejections (Kohut, 1984). Thus, treating patients with unmet selfobject needs required a new approach that would involve a different kind of transference relationship, and a different conception of empathy from the ones in classic psychoanalysis.
Kohut developed the self-psychology approach to provide both. Self-psychology defines empathy as the central executor of the human psyche: individuals with healthy empathic abilities are capable of generating meaningful and congruent responses toward other people and events (Kohut, 1981). In clinical setting, empathic understanding allows for an experience-near observation of the patient's feelings and perceptions as they are, without theoretical preconceptions or impositions (Kohut, 1968). This means that the therapeutic relationship in Kohut's self-psychology treats the therapist and the patient on equal footing: patient's direct responses to the treatment are considered to be significant, and at times, more significant than the therapist's views (Kahn & Rachman, 2000).
Kohut was clear, however, that empathy in itself is not an action, nor a form of identification with the patient. Instead, it is seen as a long-term process of attunement to each particular patient's self-expressions, emotions, and gestures (Goldberg, 1980). In this sense, self-psychology functions as a continuous feedback loop in which the therapist provides a reparatory function that is necessary for the completion of patient's interrupted or lacking developmental process. In order to achieve this, the therapist must be able to "think and feel oneself into the inner life of another person" (Kohut, 1984, p. 82). Here, Kohut begins to drastically distance himself from the classic Freudian model: the therapist's task is to, first and foremost, define and fulfil the patient's developmental trauma and unmet selfobject needs. For this reason, Kohut (1984), like Rogers, also emphasised the task of listening, and especially listening without producing interpretations: If there is one lesson that I have learned during my life as an analyst, it is the lesson that what my patients tell me is likely to be true-that many times I believed that I was right and my patients were wrong, it turned out, though often only after a prolonged search, that my rightness was superficial whereas their rightness was profound (p. 93-94).
It is clear that Kohut sought to reconceptualise the classic view of transference relationship. From a self-psychology perspective, transference is understood not as an expression of unconscious drives, but as a patient's need to complete the interrupted developmental process (Marmarosh & Mann, 2014). Therefore, the therapist's role in self-psychology is not equated to that of a clinical expert or physician dealing with a medical situation. Instead, the therapist is seen as a figure that acts out an interpersonal function by meeting the patient's self-object needs. And because these needs can only be met through patient's sufficient exposure to empathy, Kohut argues that the therapist cannot be "detached": they must reflect patient's feelings and thoughts, as well as provide for their lacking structural parts of self. For instance, if the patient expresses unmet mirroring needs, then the therapist should encourage their sense of validation and assertiveness through transference. If, on the other hand, the patient tends to idealise the therapist by admiring their authority, knowledge, or interpersonal skills, then such an idealised transference should be accepted as a patient's attempt to rebuild their psychic structure (McLean, 2007). The hallmark of successful self-psychology, then, is to gradually establish a stable foundation that will enable the patients' self-structure through empathic and compassionate interpersonal experiences that they have likely never experienced before.
As noted earlier in the paper, Kohut and Rogers had significant theoretical differences in approaching empathy. Rogers' view of empathy was very broad and largely non-clinical: all one needs is unconditional positive regard (what he called non-judgmental acceptance) and empathic understanding in order to develop a healthy self-esteem. Kohut, apparently, was sceptical of Rogers' techniques by claiming that therapists who only use empathic listening are like repairmen trying fix clocks without knowing anything about them; the analogy goes to show that such empathic listeners only repeat patients' problems without actually "working through" them (Kohut, 1973(Kohut, /1978. It is interesting to observe that Rogers frequently cited and commented on Kohut's work in agreement (Rogers, 1986), but also noted that there were significant moments of disagreement. For example, Rogers was particularly critical of Kohut's use of empathy for "data collection" in pursuit of a dynamic interpretation. That said, it is unclear whether these disagreements were caused by genuine differences in clinical technique or historical controversies. Some of these disagreements can be attributed to then-existing divisions between psychology and psychiatry that could be observed in both Universities and clinical practice (Kahn & Rachman, 2000).
Even with all these technical and personal differences in mind, it is evident that both Rogers and Kohut pioneered a completely different view of empathy. It was no longer perceived as a strictly epistemological or intellectual tool that brings with itself a bundle of problems that warrant a "detached" therapeutic stance, as could be seen with Freud's psychoanalysis and some of Jaspers' philosophical work. Instead, person-centered and self-psychology approaches utilised empathy for its affective features, causing a huge shift in the therapist's role. No longer an expert but a depersonified medium, the therapist promotes missing or impaired segments of the patient's selfhood.
The combined efforts between Kohut's self-psychology and Rogers' person-centered therapy, however, continue to entertain a similar dialectic of objectivity and subjectivity that could be perceived in Freud's and Jaspers' work. To this day, there is no consensus over the dialectic between empathy as a universally applicable tool (Rogerian approach) and empathy as a particular clinical technique in combination with interpretation (Kohutian approach). But perhaps this dialectic need not be resolved in its entirety. As a relational practice involving (at least) two active psyches and innumerable unique situations and contexts, psychotherapy ought to benefit from the continuous debate over various forms of empathy and their utility for each patient and situation (Bohart & Greenberg, 1997). Our approach to empathy in the psychotherapy context, however, would undoubtedly benefit from a clearer direction and specificity. The next section outlines some future research trajectories that could facilitate this.

Empathy in contemporary psychotherapy research
Despite the conflicting definitions and ambiguous uses of empathy in psychological and clinical literature, it remains to be one of the most important predictors in therapeutic outcome across every therapeutic modality (Constantino et al., 2008;Elliot et al., 2011;Watson et al., 2013). This raises two important points: 1) If empathy plays a central role in all therapeutic modalities (as present data indicate), then each therapeutic modality should present a working definition of empathy. Otherwise, conflicting or lacking definitions contribute to inaccurate and misleading psychotherapy process measurements; 2) If the use of empathy is prevalent in all therapeutic modalities, then we ought to assess its function and impact on different therapeutic relationships.
It is only with Rogers that empathy came to be seen as an important (and measurable) predictor for positive therapeutic outcome. This is because Rogers was able to put forward a clearer definition of empathy as a sensitive (affective) understanding of patients' experiences that can be achieved through listening, unconditional positive regard, and congruent empathic responses (Rogers, 1975). This allowed to incorporate empathy into research studying therapeutic relationships, patients' constructive learning, and therapeutic change. However, even in contemporary psychotherapy research, researchers often do not define empathy or they use it interchangeably with other terms, such as sympathy, relatedness, or therapeutic alliance (Bruce et al., 2010). This is partly because, historically, psychotherapy research mostly focused on therapeutic techniques that provide effective symptom alleviation (Roth & Fonagy, 2005). Empathic and relational qualities were not seen as therapeutic techniques, and thus they were largely omitted from psychotherapy process research.
Similarly, the specific function and impact of empathy on different therapeutic relationships often goes unrecorded. As this paper sought to demonstrate, empathy is a wide and overarching concept; it ranges from cognitive (ability to understand others' feelings) and affective (ability to feel others' feelings), with many different dimensions and models in between (see for example, Berrol, 2006;Stueber, 2012). Therefore, empathy in psychotherapy is not necessarily limited to Rogers' unconditional positive regard or Kohut's attunement. As Orange (1995) helpfully points out, empathy is "value-neutral" (p. 171): its function is to perceive another individual's internal experiences and to respond to them in a congruent way.
However, there may be clinical instances where the congruent response is not that of positive regard or need fulfilment. This is precisely why Freud advocated a "detached" therapeutic stance in his technical papers; since patients form transference patterns characteristic of their early interpersonal experiences long before entering clinical treatment, their psychic tendency to repeat these interpersonal experiences must be met cautiously (with a certain sense of detachment) by the therapist (Ferenczi, 1928).
Contemporary clinical cases concur that relational or "engaged" therapeutic responses can sometimes be ineffective when treating patients with personality disorders and defence mechanisms (Galloway & Brodsky, 2003). In cases of overinvolvement, therapeutic impasse or therapistpatient identification, the congruent response is often not one of affective empathy; in these situations, therapists are encouraged to adopt the therapeutic mode of distancing, which sets clearer boundaries between therapists and patients. As such, there is an overarching necessity to (re)consider what we mean by congruent empathic response (and, specifically, the wide array of possible congruent responses) in different clinical situations across all forms of psychotherapies.

Conclusion
Although there is a great deal of diversity, conflict, and controversy surrounding empathy, this concept re-emerges in psychotherapy literature more frequently than any other (Patterson, 1984). This is unsurprising, considering that the notion of empathy was at the forefront of Freud's ideas about psychoanalytic treatment, and as such, it was integral to the first clinical application of psychology.
Furthermore, both Rogers' person-centered therapy and Kohut's self-psychology were developed largely as responses to Freud's epistemological view of empathy. These approaches flipped the original psychoanalytic arrangement by prioritising listening over interpreting and equalising the roles of the therapist and the patient (Kahn & Rachman, 2000). As such, the concept of empathy has played an important historical part in shaping the understanding of the therapist's role, transference, and therapeutic relationship across some of the most influential psychoanalytic psychotherapies.
The task of contemporary psychotherapy researchers is to integrate the wide (and at times conflicting) psychoanalytic legacy on empathy. This requires a further study of empathy's multifaceted nature and its impact on different therapeutic relationships. More specifically, further research is necessary into what appears to be the overarching feature of empathy: congruent responses to others' emotional experiences. From an immediate point of view, a congruent empathic understanding is one that appreciates patient's circumstances and experiences whilst providing a helpful and productive response. But as previously noted, this response need not be limited to Rogers' unconditional positive regard or Kohut's attunement, the two positions that are almost exclusively associated with empathic understanding. Acongruent empathic response can also involve therapeutic "detachment", by which therapists distance themselves from patients, set clearer relationship boundaries, generate interpretations, collect additional data, etc.
In this way, both Jaspers' and Freud's views of empathy as an epistemological (intellectual or cognitive) tool can be integrated into what might be considered as empathic psychotherapy; after all, clinical situations and individual patients are complex and diverse, and it is therefore unsurprising that there is a need for both cognitive and affective uses of empathy. Guidance on how to identify patients' needs and provide appropriate empathic responses would assist therapists in making clinical decisions that could have an immense impact on therapeutic outcome. Whilst classic and contemporary clinicians often do not explicitly deal with issues relating to empathy in writing, these issues must be "worked through" in order to assist patients in practice.

Arts and Humanities Research Council (AHRC) and
Consortium for Humanities and the Arts South-East England (CHASE) Doctoral Training Partnership, Award Number [AH/L50 3861/1].