We note that the Darwinian models of psychosis reviewed by Kelleher et al ^{Reference Kelleher, Jenner and Cannon1} in their editorial were all variants of the ‘costly by-product’ evolutionary model whereby an adaptive neurobiological system that enhances fitness in the vast majority of the population generates the risk of error in a small minority, resulting in psychosis (including schizophrenia). Burns ^{Reference Burns2} identified the frontotemporal and frontoparietal cortical connections of the social brain, whereas Crow ^{Reference Crow3} proposed that the dysregulation occurs in the language centres.

We wish to propose a different and entirely environmental Darwinian formulation for the non-affective psychoses based on an ‘environmental mismatch’ model. We have explained elsewhere ^{Reference Abed and Abbas4} that, although we agree with Burns’ proposal regarding locating the dysregulation and dysconnectivity within the social brain, we contend that the aetiology of the dysregulation relates to the effects of the novel post-Neolithic social environment. Although the susceptibility to non-affective psychosis, including schizophrenia, is likely to be ancient, the schizophrenic and the non-affective psychosis phenotype did not manifest itself until very recently in our species’ history. In other words, the risk of these disorders lay dormant and did not become evident until the post-Neolithic period.

Hence, we have proposed a reformulation of the social brain theory of schizophrenia and contend that schizophrenia (and the non-affective psychoses) are novel human phenomena that arose following the establishment of large permanent human settlements that accompanied the advent of agriculture and the abandonment of the hunter–gatherer way of life. We have contended that the blurring of the demarcation between in-group and out-group membership and living in close proximity to strangers is a stressor that can lead to perturbation in the development of the social brain in vulnerable individuals, resulting in the syndrome of schizophrenia. Hence, according to our formulation, schizophrenia is the result of a mismatch between the post-Neolithic human social environment and the design of the social brain. We highlight the importance of the distinction between in-group and out-group membership that lies at the heart of intergroup conflict, violence and xenophobia. Our hypothesis (the out-group intolerance hypothesis) provides an explanation for the disparities in the prevalence of schizophrenia across the world and for the higher risk of this condition among immigrants and city dwellers. We propose that our hypothesis can account for a range of disparate epidemiological and other findings regarding schizophrenia that have thus far defied explanation by other theories, including the Darwinian by-product formulations reviewed by Kelleher et al.