DURING THE GLORY YEARS OF THE LATE 1960s AND EARLY 1970s, when reimbursement for insured outpatient psychotherapy was automatic, unchallenged and comprehensive, it's fair to say that diagnosis was of only fleeting concern to most therapists. Finding the right coded diagnosis in the Diagnostic and Statistical Manual of Mental Disorders-II (DSM), at that time a meager bookette of 134 pages, published in 1968 by the American Psychiatric Association (APA), was a mildly annoying formality required for insurance reimbursement, the key that opened the vault.

Probably every therapist in private practice had a favorite diagnosis; "I diagnosed almost everybody with 'adjustment reaction' because I didn't want to hurt them," says Dana Ackley, a clinical psychologist in Roanoke, Virginia. "I could hardly ever remember what I diagnosed for my clients it didn't really matter but I could tell you chapter and verse about the details of their lives."

Formal diagnosis, specifically the entire DSM enterprise, was regarded as a practice from the Pleistocene era of inappropriate medicalization, fossilized in a manual that nobody took very seriously. Even after the publication of the vastly expanded, explicitly medical DSM-III in 1980, and the revised edition (DSM-III-K)that appeared 7 years later (together adding 110 categories to DSM-//), diagnosis remained a marginal matter of…