TY - JOUR T1 - Writing for journals: a paradigm lost? JF - Journal of Epidemiology and Community Health JO - J Epidemiol Community Health SP - 642 LP - 643 DO - 10.1136/jech.54.9.642 VL - 54 IS - 9 AU - TIM ALBERT Y1 - 2000/09/01 UR - http://jech.bmj.com/content/54/9/642.abstract N2 - Writing for journals, like giving to the poor and taking exercise, is one of those institutions that seem beyond reproach. One of the encouraging characteristics of our civilisation (as the ideology has it) is this constant stream of high quality information about our bodies and our health, contributing piece by piece to an expanding knowledge base, and taking us on the long slow march to a brighter and healthier future.The evidence for these assumptions is remarkably thin, which is somewhat surprising in view of the fact that they are usually expounded by a profession that prides itself on its objectivity. The assumption that publication is the same as dissemination has never been satisfactorily tested; indeed one study found that those who claimed to have read journals recently had retained surprisingly little of the information contained in them.1 This does not surprise professional communicators2 because most scientific papers break the principles of effective style, famously outlined by George Orwell, such as short sentences, short and familiar words and active voice.3 They ignore the popular technique, well validated by the research and experiences of publishing houses, of starting with an explicit key message; instead they favour the tortuous four section structure (also known as the IMRAD—Introduction, Methods, Results and Discussion—structure) where a clear message will appear in the last sentence in just under 50% of articles.4 The text in … ER -