Japan's part-natural, part-human disaster is an extraordinary event. As well as dealing with the consequences of an earthquake and tsunami, rescuers are having to evacuate thousands of people from the danger zone around Fukushima. In addition, the country is blighted by blackouts from the shutting of 10 or more nuclear plants. It is a textbook case of how technology can increase our vulnerability through unintended side-effects.
Yet there had been early warnings from analysts. In 2006, the Japanese professor Katsuhiko Ishibashi resigned from a nuclear power advisory panel, saying that the policy of building in earthquake zones could lead to catastrophe, and that design standards for proofing them against damage were too lax. Further back, the seminal study of accidents in complex technologies was Charles Perrow's Normal Accidents, published in 1984.
Perrow, a Yale professor, analysed accidents in chemical plants, air traffic control, shipping and dams, as well as his main focus: the 1979 accident at the Three Mile Island nuclear plant in Pennsylvania. Things can go wrong with design, equipment, procedures, operators, supplies and the environment. Occasionally two or more will have problems simultaneously; in a complex technology such as a nuclear plant, the potential for this is ever-present. Perrow took five pages to sketch what went wrong in the first 13 seconds of the incident. He concluded that in complex systems, "no matter how effective conventional safety devices are, there is a form of accident that is inevitable" – hence "normal accidents".
...
Unfortunately, such events are often made worse by the way the nuclear industry and governments handle the early stages of disasters, as they reassure us that all is fine. Some statements are well intentioned. But as things get worse, people wonder why early reassurances were issued when it is apparent that there was no basis for them. It is simply too early to say what precisely went wrong at Fukushima, and it has been surprising to see commentators speak with such speed and certainty. Most people accept that they will only ever have a rough understanding of the facts. But they instinctively ask if they can trust those in charge and wonder why governments support particular technologies so strongly.
Industry and governments need to be more straightforward with the public. The pretence of knowledge is deeply unscientific; a more humble approach where officials are frank about the unknowns would paradoxically engender greater trust. Likewise, nuclear's opponents need to adopt a measured approach. We need a fuller democratic debate about the choices we are making. Catastrophic potential needs to be a central criterion in decisions about technology. Advice from experts is useful, but the most significant questions are ethical in character.