What to do with complexity ?
Résumé
Complexity is everywhere. It is one of these contemporary words that has colonised both scientific and lay discourses. In the field of major accident or safety research, complexity has played a central role in structuring ideas about disasters as socio technical phenomena. The use of complexity as a key concept has been extensive in the field of safety science for the past 30 years. With early emphasis for example on technology (Perrow, 1984), on human-machine interactions (Rasmussen, Lind, 1981) or on organisations (La Porte, 1975), complexity has been shaped in relations to these different contexts of use. As a consequence, one could say that complexity is part of the genesis of the field. It helped to build a rhetoric linking together surprises, features of high risk socio - technological systems (aviation, nuclear industry, etc) and various category of operating personnel (e.g. pilots, process operators, engineers, managers, inspectorate) studied by different scientific disciplines (e.g. cognitive engineering, ergonomics, sociology, management, etc). But the use of complexity has evolved throughout the years, revealing sometimes diverging or competing meanings, some criticising for example the limits of a 'technologically' centred view of complexity in favour instead of a more 'socially' oriented one (Vaughan, 2005 ) . Similarly, if one looks beyond the field of safety, the genesis of complexity is older than otherwise presented in recent accounts (e.g. Mitchell, 2009) and reveal many different nuances and orientations. Let us think for example of Weaver's 1948 paper on 'organised complexity' (Weaver, 1948). Thus, some authors have written about a history of complexity ideas (e.g. Le Moigne, 2003) linking its more recent developments of the twenty first century to the mid twentieth century with cybernetics, second cybernetics (self organisation), artificial intelligence, system theory. There are, therefore, several 'waves' of complexity discourses. But today, the list of writings on the topics is endless, and anyone investigating this area is impressed by the number of books, articles and papers dealing with the subject from so many different angles. In this respect, Thrift sees complexity as 'one of the first fully mediatised scientific theories' (Thrift, 1999, 61). The aim of this presentation, as part of the symposium on complexity, is to offer a historical perspective on the notion of complexity both inside and outside the field of safety and the relationship between the two.