The changing shape of clinical practice: Driven by science or by pragmatics?

Few would dispute that psychoanalytic practice has changed considerably over the last 50 years. The differences are at least three-fold: (a) a bimodal distribution of treatment lengths (some very long and some very short); (b) patients with personality pathology in addition to axis I symptoms; and (c) technical changes with a relational flavour, displaying influences from developmental theories including attachment theory. Are any of these changes associated with scientific research and, if so, how? A second question concerns the research base of psychoanalysis. If we take the discoveries of the last 25 years seriously, what kind of therapy might we recommend to our patients? Would it look anything like standard practice? I will attempt to pose some serious questions about the status of clinical practice, asking why it appears to be protected from scientific advances yet readily responsive to pragmatic considerations such as third party payments. To make the argument easier to absorb I will try and tell it as number of short heroic tales, each with its own beginning, middle and end, complete with the moral lessons to be learned.