ABSTRACT
All those who live on or near an active volcano are making a volcanic prediction. They may not think about it in those terms, but they balance the benefits of staying near a volcano against the chances of losing their lives or jobs. They predict that they, and their property will survive. If, however, they live on materials erupted from an active volcano, they run the risk that a similar eruption will overwhelm them in the future. The longer their homes are spared, the more confident the people will become that they are safe-and the more likely it is, in fact, that they will be proved wrong, because active volcanoes that have been quiet for a long time have a way of breaking their silence by cataclysmic explosions with little warning. But it is hard to predict when an eruption will take place, to assess its likely power, and to calculate the risks facing the local population. In some countries the inhabitants scarcely have a choice. Many towns in Japan and Indonesia, for example, are too close to explosive volcanoes. It is hardly surprising, then, that these countries account for two-thirds of the deaths caused by volcanoes since 1600. Safe choices are not always made anyway. People often knowingly put themselves in danger, which they clearly perceive as small. They fall prey to the gambler’s fallacy: that an event is unlikely to repeat itself in the same place. Catania was rebuilt on the lavaflows from Etna that had destroyed most of the city in 1669. St Pierre was rebuilt over its own ruins soon after the eruption of Montagne Pelée in 1902. Although the risks were re-emphasized by the eruptions in 1929-31, St Pierre now has 7000 inhabitants.
