The pattern of the genetics of psychiatric disease is various, in detail and in kind; one cannot always expect to find simple, trustworthy explanations of what are complex relationships. There will be some diseases, mostly of the disruptive type, that will prove to be Mendelian defects; but it would be idle to expect that of most. Many disorders will be elucidated only when the nub of the problem is better defined. Mindless application of standard genetic techniques devised for quite different purposes is no substitute for articulate speculation built on sound fact and cogent testing. In this respect genetic evidence is no different from that of physiology, pathology and pharmacology, which are perhaps less unfamiliar to readers than genetics. Finally, no answer has been found unless it deals in those terms in which the question first arose. If the topic of the genetic analysis is some psychosis, the final predicate must be a statement about the psychosis, not about some arbitrary abstraction from it.