In psychiatry one may generally distinguish two main approaches toward major psychiatric disorders: a categorical approach, which regards the concepts of "normality" and "mentally ill" as two clearly separated spheres; and a dimensional approach, which sees the two states as two poles of a continuum. It is clear to every psychiatrist that there are good and less good empirical reasons for favoring the one approach over the other. What is often less clear is the extent to which a decision for categorization or for dimensionality is also a decision in favor of a certain epistemological and/or ontological view of the nature of the relationship between disease and health, deviancy and normalcy, patient reality and doctor reality. In other words, the imperatives of the diagnostic enterprise ensure that philosophical considerations—usually unexamined—play a role in psychiatry from the very outset.