I admire someone like R. Walter Heinrichs who is not afraid to wrestle with a moving mountain. A neuropsychologist at the University of Toronto, Heinrichs has attempted to evaluate all studies on schizophrenia published since 1980 and to weigh the quality of evidence on various theories of this disease. To his credit, he does so by using rules of evidence, meta-analyses, calculations of effect size, and more than 1,000 references. In Search of Madness: Schizophrenia and Neuroscience is well organized and well written.

Heinrichs concludes that measures of cognitive dysfunction—for example, verbal memory—have the most evidence in their favor of being abnormal in schizophrenia. Second place goes to measures of neurophysiology—for example, P50-evoked potential. Third, and a distant last, come measures of neurobiology—for example, excess presynaptic glutamate. Heinrichs then ties his conclusions to the diathesis-stress model of schizophrenia in general and to neurodevelopmental models in particular. It is a thought-provoking and interesting exercise.

Despite its strengths, the book also has many shortcomings, and the reader gets the impression that the mountain may have won. One such shortcoming is that the author, although critical of virtually every other aspect of schizophrenia research, never questions the assumption that genes are an essential part of the etiology of schizophrenia. We all accept the fact that genes play some predisposing role, but if genes are critical in etiology, why didn't the incidence of schizophrenia decrease over the past two centuries, when most of the affected individuals were locked up and their rate of reproduction was only one-fourth that of other people? Why is the incidence of schizophrenia not higher in populations with a higher rate of first-cousin marriages?

Omissions are also a problem. Neurological soft signs are not discussed in this book, despite the numerous studies since 1980. How do neurological soft signs differ from neurophysiological measures, such as P50-evoked potentials, other than by the fact that the former are traditionally measured by neurologists and the latter by psychologists? Omissions are also evident in the discussion of the season-of-birth effect. Heinrichs found just 17 studies, yet a 1997 review lists more than 250. By arbitrarily restricting himself to studies from 1980 onward, Heinrichs has failed to include a majority of the largest season-of-birth studies.

The author can also be criticized for assuming that schizophrenia affects only one part of the brain or that it will ever be measured in all cases by a single marker of brain function, such as P50. The results of recent studies strongly suggest that the disease affects multiple structures at the base of the brain and, like multiple sclerosis, may affect somewhat different structures in different patients.

However, perhaps the greatest shortcoming of In Search of Madness is its omission of a majority of recent neurobiological studies. Most of the book appears to have been written in the mid-1990s, yet the most interesting neurobiological studies have been published since that time. Among the author's 1,018 references, only 18 percent are from 1997 or after. Heinrichs includes a few references from 2000, which suggests that he had the option of including these more recent data. Given that the most recent studies are now available online, there was no reason not to include them. Furthermore, because we have been measuring verbal memory and P50 for 35 years but presynaptic glutamate for only five, the comparisons hardly seem fair.

Notwithstanding such criticisms, I came away with great respect for the author's efforts and enthusiastically recommend In Search of Madness to schizophrenia researchers and especially neuropsychologists. Until the rest of us are willing to wrestle with this moving mountain of data, Heinrichs' achievements will stand as the mark to beat.

Dr. Torrey is executive director of the Stanley Medical Research Institute and professor of psychiatry at the Uniformed Services University of the Health Sciences in Bethesda, Maryland.