Dr Pearce continues an argument he has been making consistently and persuasively for decades,1 2 that diseases should not be confused with diagnoses and syndromes. In particular, he argues that when diagnoses and syndromes are offered, they should not imply an unproven cause when none is known (eg, chronic whiplash syndrome, Gulf War Syndrome). He explores what people mean by the word disease and comes to the unfortunate conclusion that no one is all that sure. Ultimately, we want to classify the problems that we see according to aetiology (like tuberculous meningitis) but most of the time we have to settle for pattern recognition corresponding to pathology (eg, Parkinson's disease, multiple sclerosis). In many conditions in neurology we have to rely on pattern recognition alone (eg, migraine, epilepsy) without any pathology or even imaging abnormalities.

Dr Pearce brings some philosophical heavyweights to the argument but the final definitions of ‘disease’ are so broad that it is hard to see many people using or understanding them. If one does define disease according to Scadding or even the Oxford English Dictionary, any group of patients with a common set of characteristics, such as …