I read with interest the title of the editorial by Harland et al (Reference Harland, Morgan and Hutchinson2004), which promised a fascinating synthesis of phenomenology, anthropology and the psychology of the self to formulate a new model for the aetiology of psychosis. Sadly, this was not achieved by the authors and I was left wondering how this had been lost on the Editor.

My response to the authors' striking claim that they propose a new model (even if only ‘an outline sketch of a potential model’) is: what model? Unfortunately, careful reading of the paper in search of this model resulted in little that is new and certainly nothing approaching a testable hypothesis.

Is it a new model to state that the environment affects the configuration of the self and that this can somehow lead to psychosis? The authors cite the example of migration and its association with increased risk of psychosis to illustrate their point, but it does not seem that they have anything new to say about this fascinating area.

The authors call for greater rigour in future conceptual models that integrate the biological and the social in the aetiology of psychosis. It is, therefore, doubly surprising that they advocate the integration of hermeneutically oriented social sciences into such future models. Given that hermeneutics is disconnected from the rigours of the laws of causation that govern the material world, one wonders how this would reduce the ‘vagueness’ that the authors warn us against. It is also of interest that the two other concepts central to the authors' model are the ‘self’ - a concept that has numerous competing definitions (one article cited 21 competing concepts of the self; see Reference Zahavi, Kircher and DavidZahavi, 2003) - and ‘social capital’, which has no operational definition (see Reference McKenzie, Whitley and WeichMcKenzie et al, 2002). It seems to me that the authors should have followed their own admonition against vagueness or else produced their own clear definitions of these concepts.

Also, the authors commit an elementary error by confusing the concept of ‘biological’ with ‘genetic’ or ‘genomic’ in their critique of current theories on schizophrenia, citing the work of Eisenberg (Reference Eisenberg2004). When we consider the role of a given environmental factor in shaping a particular trait, we are most certainly dealing with a biological process. Can we discuss the effect of sunlight on tanning of the skin without considering melanocytes and melanin (see Reference Gaulin and McBurneyGaulin & McBurney, 2001)? Similarly, if the human brain/mind has the propensity, under certain environmental conditions and given a particular genetic make-up and early-life experience, to produce the clinical picture we call psychosis, this cannot be understood outside of biology. Phenotypes, we should remember, are not simply the obligate expression of genes but the complex outcome of the interaction of the genome with the environment. In other words, the identification of an environmental risk factor for a particular disorder is not the end of the story. To achieve a real understanding of how the phenotypic trait was shaped, we still need to understand the intra-organismic process that led to the said trait.