This chapter is concerned with the ‘official’ rules for inferring schizophrenia, as set out in the 3rd and 4th editions of the DSM. Although it is DSM-IV which is currently in use, I shall also examine DSM-III in some detail.1 There are several reasons for this. First, as I noted at the end of the previous chapter, DSM-III was the landmark edition of the DSM, the first to include correspondence rules (diagnostic criteria) for its concepts. Subsequent editions have not made fundamental changes to the actual criteria for inferring ‘schizophrenia’, although the justifications given for producing particular criteria have changed. It is therefore important to examine DSM-III to evaluate the process by which the initial criteria were produced. Second, although the process of producing diagnostic criteria is presented as one of continual progress, in practice it is one of trying to repair the shortcomings of previous versions of the DSM while avoiding a fundamental examination of the validity of its diagnostic concepts. It is therefore difficult to appreciate the approach to producing schizophrenia’s correspondence rules which was taken in DSM-IV, the reasons why it was taken and the problems encountered, without knowing something about the approach adopted in DSM-III and its problems. Finally, it is not unlikely that the approach or justificatory framework adopted by DSM-III will resurface at some future time when its problems have been obscured or forgotten, so that it might be as well to keep them in mind.