In our view, the conflict theory of imprinting explains the evolution of parental allele-specific gene expression patterns in the somatic tissues of mammals and angiosperms. Not surprisingly, given its importance in mammalian development and pathology, the evolution of imprinting continues to attract considerable interest from theoretical and experimental biologists. However, we contend that much of the ensuing debate is of poor quality. We discuss several problems with the manner in which workers in the field engage in this debate and we argue for a more formal approach to the discussion of theories of the evolution of imprinting.