Accident Prone: A History of Technology, Psychology, and Misfits of the Machine Age

Before reviewing John Burnham's thoughtful book, let us review the Times Higher Education books editor. That lady suggested to your reviewer, a former RAF and airline pilot, that this book, Accident Prone, addressed his precise area of expertise, while denying that she had confused him with the famous Second World War aviator Pilot Officer Prune.

Prune, the dim-witted creation of cartoonist Bill Hooper, was the RAF's wartime specialist in accidents. In other words, he specialised in having them. He had an unfailing ability to find the means to turn a serviceable aeroplane into scrap metal, a role model that RAF pilots were urged not to emulate. The point was that, whereas Prune always emerged from the wreckage to provide next month's lesson (Good Show!), any pilot daft enough to fly like him would not be there to read the advice. Sadly, neither Prune, nor even the ubiquitous Murphy and his Law, is cited in Burnham's otherwise comprehensive work (Bad Show!).

Accident Prone is not an analysis of the concept of accident proneness, but a scholarly history of the studies of the concept. It is well researched; the 200-plus pages of text are accompanied by nearly half as many pages of smaller-print notes. Burnham is a historian; his "biography of an idea" (of accident proneness) gives 1925-26 as the birth and christening dates, announced concurrently in seminal British and German papers; thereafter came decades of changes in character, culminating in the idea's current unfashionability.

Some may doubt the validity of the author's attribution of the birth of the concept to industrialisation. It implies that thitherto the accident prone had not been identified; this, however, is at odds with history's extensive parade of duffers. His next point is more persuasive. He quotes the British historian Roger Cooter, who "identified a defining 'moment' when injuries changed from private concerns to public concerns", and for a simple reason: the industrial advances that brought more people into more and longer contact with more and bigger machinery were accompanied by rising numbers of industrial accidents. For example, there were more than 2 million industrial accidents in Germany between 1885 and 1908. The economic and personal costs became a concern for insurers and philanthropists, and for the occasional caring mill owner.

Burnham suggests that accident prevention first centred on the human factor. He quotes the admirable H.B. Rockwell, who cited bad engineering and faulty management as human errors. This 1905 American accident expert was well ahead of his times and to some degree, I am sorry to say, of ours. Accident proneness dominated accident-prevention thinking for most of the 20th century, until the idea "faded away", with some even calling it a myth. "Accident proneness was effectively replaced by engineering - in the safety movement and as a general social strategy." Burnham does not say here that the main reason for this was the successful use of selection processes to weed out the accident prone.

Accident Prone does not purport to be an action guide; it offers only two answers to the practitioners' "so what?" test. First, beware of any ill-defined or general concept of accident proneness. Second, be aware that the curse of political correctness is a new and serious challenge to accident prevention. A voluble lobby argues that the right of individuals to be "different" or "other" is a human right that transcends all else. The label "accident prone" is no longer socially acceptable, and therefore it must go. But the inept will never be banished from the world. Keep a safe distance from them; you ignore any aspect of the safety equation at your peril. And please, someone, write a book on stupidity proneness.

The reality is that only a holistic approach can achieve the highest standards of system safety. It must consider the properties of the individual components of the system, hardware, software and humanware, their interfaces, their organisation and their behaviour both as parts and as a whole. There is lots to go wrong and it may. And if we choose to overlook the variability of human competence, it certainly will.

Accident Prone has no bibliography and the quality of the illustrations is deplorable (Bad Show!). Other than that, Burnham has written a good academic book for academics (Good Show!).