Piper et al1 report convincing evidence that a high proportion of transference interpretations (PTIs) in timelimited psychodynamic psychotherapy is correlated with a poor therapeutic alliance and poor outcome for some patients. Almost 10 years ago, in a similar clinical context, Buckley et al2 reported an identical finding with respect to interpretation and outcome. Some 10 years prior to that, Postner et al3 reported a similar finding in a different therapeutic context, using different outcome criteria. They found that a high proportion of interpretations as early as the second session was positively correlated with a high rate of early dropout in conjoint family therapy. The first and third groups of authors recognized that their evidence did not permit them to decide whether the patients' behavior determined the nature of the therapists' interventions, or the reverse.
By examining the relationship between PTI and the therapeutic relationship at the end