We entirely agree with Dr Brown and Professor Peet (March 2002 JRSM¹) that the backward walking by their dystonic patient who subsequently developed psychosis was unrelated to his psychosis per se. Walking backwards in preference to forwards can be a feature of dystonia, whether primary (their patient initially had torticollis and later truncal torsion) or secondary (he was treated with neuroleptics, which may have caused superadded tardive dystonia). It is one of the many manifestations of ‘task-specificity’ that in the past caused dystonia and other movement disorders to be mislabelled psychogenic. Unfortunately this can still happen. The list includes not only patients with dystonia but also patients with Parkinson's disease who may be able to run up and down stairs but not walk on the flat, and patients with orthostatic tremor who are unable to stand still but have to keep moving like the white rabbit. The precise reasons for these disparities are not understood, but they are very strong pointers to organic extrapyramidal disease.