In this paper, I consider how human psychological variation should affect methodology in two-stage psychiatric research. I examine recent empirical evidence and develop what I call the Cross-Cultural Diversity picture of the human mind, according to which variation in the underlying causal structure of the human mind implies variation in mental illnesses. I then examine the implications of the Cross-Cultural Diversity view for a discipline that tries to separate the descriptive from the evaluative in taxonomization of mental disorders, by examining one methodological proposal given by philosopher Dominic Murphy in his book Psychiatry in the Scientific Image. I argue that, even in an idealized case, Murphy’s methodology is hindered by its reliance on a conception of ‘normal human nature,’ and thus does not adequately accommodate cognitive diversity. Next I sketch a promising way to revise Murphy’s proposed methodology, by examining Grant Ramsey’s recent work on human nature and his Life-history Trait Cluster view. I end with some notes on how these considerations are beginning to shape inquiry in the form of the Research Domain Criteria project.