I thank Drs Feeney and McQueen for their thoughtful observations. Dr Feeney is of course correct in pointing out that post-termination boundary violations, much like boundary violations during treatment, occupy a broad range. Sexual activity is at one end of spectrum and attracts most of the negative consequences, both for the patient and the therapist. It also attracts the more punitive sanctions, in civil as well as criminal courts. The ‘milder’ forms of boundary violation can be seemingly innocuous, for example employing an ex-patient to do small jobs. It is debatable whether such actions can cause harm to the patient, but it can be argued with relative force that the new relationship (say of employer and employee) is based on something that misuses trust, or trust obtained in the course of a fiduciary relationship. The damage caused may not be immediately obvious but is there for all to see if one is so minded. Given that the therapist can access a large pool of people for establishing such a relationship, it remains open to interpretation why a patient is recruited. Some authors (Reference Gutheil and GabbardGutheil 1993) call these ‘minor’ transgressions, boundary crossing, rather than boundary violations. I personally do not believe that this artificial distinction actually adds anything to the discourse, a discourse based on the timeless notion of trust in the profession (Reference SarkarSarkar 2004). More than the actual act, it is the degree of exploitation, not as felt by the therapist necessarily but as perceived by others (including the patient), that essentially influences outcomes.

Dr McQueen very correctly brings the concept of attachment into the mix. Although schools of psychotherapy are divided on how much ‘transference’ is responsible for therapy, the concept of attachment is less contentious. Attachment is all pervasive, and healthy attachment is necessary not only in therapy but in general development. Research is emerging on the fact that early attachment behaviour is replicated in adult behaviour, including attachment behaviour as adults. This could be attachment to partners, children or therapists. Attachment behaviour is central to interpersonal relationships, of which the therapist–patient relationship is but one. Whether one calls it transference or attachment matters little as long as it is agreed that it is the dynamic of a therapeutic dyad that is perhaps curative or facilitatory in treatment. The scenarios I discussed in my original article (Reference SarkarSarkar 2004) have dysfunctional attachment (on the part of both the therapist and the patient) at their core, which may manifest as abuse of transference or merely abuse of one's position as a doctor.