TY - JOUR T1 - Diagnosis and Wittgenstein's theories of language JF - British Journal of General Practice JO - Br J Gen Pract SP - 480 LP - 481 VL - 54 IS - 503 AU - Kevin Barraclough Y1 - 2004/06/01 UR - http://bjgp.org/content/54/503/480.abstract N2 - This short discussion paper will examine the applicability of Wittgenstein's two theories of language to understanding the nature of diagnoses.It can be argued that ‘the diagnosis’ is the elemental concept of clinical medicine.1 Without it little analysis is possible, and such analysis as is possible slides around uncertainly among arguments so metaphysical as to be meaningless. Diagnoses are the hooks on which we hang all of medicine.As a newly qualified doctor I believed not only in the utility of diagnoses but also in their objective reality. The patients on the ward with a prolactinoma, heart failure or clinical depression had those conditions. Their illnesses were as real as their pyjamas. Indeed, as a student of medicine, those patients were defined by their diagnosis.As I progressed in medicine my view of the nature of ‘the diagnosis’ changed.The first realisation on the ward round was that the definitive diagnosis was made by the doctor at the apex of the pyramid of authority. The consultant neurologist's diagnosis of ‘ice-pick headache syndrome’ was the gold standard. There was nothing to measure it against. This was diagnosis defined by authoritative opinion.The next realisation was that the biological variation of disease and people means that diagnostic labels sometimes have limited meaning. What exactly do we mean when we make the diagnosis of ‘heart failure’ in an 84-year-old with multiple comorbidities, each of which affects her biochemistry, organ function, response to treatment and prognosis?2 The question becomes particularly pertinent when the label carries with it … ER -