The psychometric structure of personality has been a topic of enduring interest for some decades (John, Naumann, & Soto, 2008). As a result of this research, it is now uncontroversial to assert that a small number of latent factors – often five (Costa & McCrae, 1992a) or, less frequently, three (Eysenck & Eysenck, 1975), or six (Lee & Ashton, 2004) – account for the bulk of reliable variance in a wide spectrum of traits and behaviors. However, alongside this descriptive research, a parallel debate has focussed on whether these domains reflect unitary underlying biological systems (McCrae & Costa, 1999), or are instead better understood as convenient heuristics, valid only at the phenotypic level (Jang et al., 2002, Paunonen and Jackson, 1996, Saucier and Goldberg, 1996). Note, this latter perspective does not necessarily suggest that heritable effects on personality are absent; rather, it posits that heritable effects do not form a unitary underlying genetic architecture. While much has been published on this topic (e.g. Costa & McCrae, 1998), tests of the unitary basis of personality domains have largely taken place at the phenotypic level. This taxonomic approach has helped to advance the field by providing a common language for the structure
