To the Editor: A perusal of any issue of the Journal can only lead to the conclusion that psychiatry and neuroscience are now well integrated. A recent overview by Joseph B. Martin, M.D., Ph.D. (1), and an editorial by Stuart C. Yudofsky, M.D., and Robert E. Hales, M.D., serve to remind us that this integration has not gone far enough, and, more important, clinical psychiatry and neurology have not breached the barriers between the two disciplines. In Australia, as in many other countries around the world, the teaching of the two disciplines remains frozen in tradition, with only weak attempts at integration. This is partly due to the comfort offered by continuity. It is also because clinicians are pragmatic by nature and will change their teaching and management practices only if they are convinced that real difference to the patient is in the offing.

Most clinicians would accept that there have been remarkable changes in neuroscientific understanding in recent decades. The change in psychiatric practice has, however, been brought about more by developments in diagnostic practice and psychopharmacology and a greater empiricism in clinical care. Specialist neurological training is still not necessary for good psychiatric practice. A considerable proportion of neurological practice does not require psychiatric knowledge. While integrated teaching of the neurosciences is laudable at the undergraduate level, specialists continue to usefully train in one discipline or the other. Given the increasing complexity of diagnostic and treatment practices, there appears to be no alternative but to continue with such a division so as not to dilute expertise.

There is, of course, a middle path, which Drs. Yudofsky and Hales suggested: that of clinical neuropsychiatry. This emerging discipline defines itself as "the application of neuroscientific principles to psychiatric practice," thereby claiming all of psychiatry (2). In reality, its practice relates to the disorders that require comfortable expertise in both psychiatry and neurology. The basic training can be in either discipline, with advanced training in neuropsychiatry itself. The field defines itself by what it does—treats disorders such as dementia, epilepsy, traumatic brain injury, substance-related neuropsychiatric problems, movement disorders, secondary psychoses, etc. It obviates the need to indulge in boundary disputes. It recognizes the need for both psychiatry and neurology and the place for a hybrid discipline for neurological diseases that have psychiatric manifestations. The integration of neurology and psychiatry is thereby seeing the emergence of a new discipline rather than the disappearance of old ones. The leaders of this discipline must ensure that its training is robust and that its boundaries remain permeable in both directions.