In a recent paper, McGee, Del Vento, and Bavelas (2005) wrote about how therapists' questions constitute a form of intervention. By that, they mean that therapists' questions often carry with them a framework of presuppositions that constrain the client to answer in such a way as to ratify, and hence to affiliate with, the presuppositions informing the questions. Such affiliation involves the co-construction or sharing of the perspective of the therapist. But sometimes clients resist sharing therapists' perspectives, a situation known to even highly experienced therapists: “What therapist is not familiar with the experience of feeling his or her body tense as a client replies with ‘yes, but’ to everything that is discussed?” (Lipchik, 2002, p. 17).

How much more stressful, then, might it be for training therapists to manage such resistance? The impetus for the particular study on which this chapter is based was provided by a masters student in a university-based programme in couple and family therapy. As part of their internship, students in this programme work as individual therapists, and sometimes as co-therapists, providing counselling that is supervised by a programme faculty. This student was participating in a research project I was conducting using conversation analysis (CA) to study therapy interactions between training clinicians and their clients.