Sir: I realise that classifying antipsychotic drugs into typical and atypical is simplistic — but, understandably, everybody does it. Few, however, define it. In a previous issue of the Bulletin, Paton et al do by stipulating the drugs thought to be atypical for their work but they do not explain why these and not others were considered (Paton et al, Psychiatric Bulletin, May 2002, 26, 172-174). As far as I understood, ‘atypicality’ was something to do with catalepsy in rats (Reference KerwinKerwin, 1994) or speed of dissociation from the dopamine d2 receptor (Reference Kapur and SeemanKapur & Seeman, 2001), or both. In the same issue of the Bulletin, Taylor et al neither define atypicality, nor list the drugs under consideration (Taylor et al, pp. 170-172). To further confuse matters, they describe a study supplementing clozapine with sulpiride as evaluating the effects of the combination of atypical and typical drugs. Turning to the Maudsley Guidelines (Reference Taylor, McConnell and McconnellTaylor et al 2001) for help I found none. Atypical antipsychotics are recommended for use for everyone with psychosis, yet a defined list is not provided. Using terms like new and old generation drugs is no better. It seems to be avoiding the key issue, which is the neuropharmacology/neurophysiology, not the age or cost of the compound.

This is a genuine plea to authors; if the classification of typical—atypical is being used, please list what is being considered as atypical, and why are some drugs being considered and not others.