Freud coined the term 'countertransference' to designate the analyst's 'unconscious feelings' towards the patient. Although Freud only used the term very rarely, it became much more widely used in psychoanalytic theory after his death. In particular, analysts soon divided over the role allotted to countertransference in discussions of technique. On the one hand, many analysts argued that counter-transference manifestations were the result of incompletely analysed elements in the analyst, and that such manifestations should therefore be reduced to a minimum by a more complete training analysis. On the other hand, some analysts from the Kleinian school, beginning with Paula Heimann, argued that the analyst should be guided in his interpretations by his own countertransference reactions, taking his own feelings as an indicator of the patient's state of mind (Heimann, 1950). Whereas the former group regarded countertransference as an obstacle to analysis, the latter group regarded it as a useful tool.
In the 1950s, Lacan presents countertransference as a RESISTANCE, an obstacle which hinders the progress of psychoanalytic treatment. Like all resistances to treatment, countertransference is ultimately a resistance of the analyst. Thus Lacan defines countertransference as 'the sum of the prejudices, passions, perplexities, and even the insufficient information of the analyst at a certain moment of the dialectical process' of the treatment (Ec, 225).
Lacan refers to two of Freud's case studies to illustrate what he means. In 1951, he refers to the Dora case (Freud, 1905e), and argues that Freud's countertransference was rooted in his belief that heterosexuality is natural rather than normative, and in his identification with Herr K. Lacan argues that it was these two factors which caused Freud to handle the treatment badly and provoke the 'negative transference' which led to Dora breaking off the treatment (Lacan, 1951a).
In 1957 Lacan presents a similar analysis of Freud's treatment of the young homosexual woman (Freud, 1920a). He argues that when Freud interpreted the woman's dream as expressing a wish to deceive him, he was focusing on the imaginary dimension of the woman's transference rather than on the symbolic dimension (S4, 135). That is, Freud interpreted the dream as something directed at him personally, rather than as something directed at the Other.
Lacan argues that Freud did this because he found the woman attractive and because he identified with the woman's father (S4, 106-9). Once again, Freud's countertransference brought the treatment to a premature end, though this time it was Freud who decided to terminate it.
The preceding examples might seem to suggest that Lacan aligns himself with those analysts who argue that the training analysis should give the analyst the capacity to transcend all affective reactions to the patient. However, Lacan absolutely rejects this point of view, which he dismisses as a 'stoical ideal' (S8, 219). The training analysis does not put the analyst beyond passion, and to believe that it does would be to believe that all the passions stem from the unconscious, an idea which Lacan rejects. If anything, the better analysed the analyst is, the more likely he is to be frankly in love with, or be quite repulsed by, the analysand (S8, 220). If, then, the analyst does not act on the basis of these feelings, it is not because his training analysis has drained away his passions, but because it has given him a desire which is even stronger than those passions, a desire which Lacan calls the DESIRE OF THE ANALYST (S8 220-1).
Hence Lacan does not entirely reject Paula Heimann's position. He accepts that analysts have feelings towards their patients, and that sometimes the analyst can direct the treatment better by reflecting on these feelings. For example, if Freud had reflected a bit more on his feelings towards the young homosexual woman, he might have avoided interpreting her dream as a message addressed directly to him (S4, 108).
No one has ever said that the analyst should never have feelings towards his patient. But he must know not only not to give into them, to keep them in their place, but also how to make adequate use of them in his technique. (Sl, 32)
If countertransference is condemned by Lacan, then, it is because he defines it not in terms of affects felt by the analyst, but as the analyst's failure to use those affects appropriately. In the 1960s Lacan becomes very critical of the term countertransference. He argues that it connotes a symmetrical relationship between the analyst and the analysand, whereas the transference is anything but a symmetrical relationship. When speaking of the analyst's position it is both misleading and unnecessary to use the term countertransference; it is sufficient to speak of the different ways in which the analyst and analysand are implicated in the transference (S8, 233). 'The transference is a phenomenon in which subject and psycho-analyst are both included. To divide it in terms of transference and counter-transference . . . is never more than a way of avoiding the essence of the matter' (Sll, 231).