ITALIAN PROBABILITY

As a mathematician, I have a pretty good understanding of probability. In particular, I understand that probability is theoretical and relates to the future. It does not say what will happen, it only says what may happen. Even if the chances of getting six sixes in a row are very small, we can't say for sure in advance that it won't happen; it might, though it is unikely.

So this week's decision by an Italian judge to sentence six scientists and an official to six years in prison is extremely worrying. The case arose out of the aftermath of the April 2009 earthquake in L'Aquila, in which 309 people were killed. The seven were all members of the National Commission for the Forecast and Prevention of Major Risks, whose job is to assess the probability of such things as earthquakes. According to the court, they had issued wrongly reassuring statements, which had led (amongst other things) to people staying in the town and being killed, instead of moving to a place of safety. The defendants were convicted of manslaughter and banned from ever holding public office again.

The problem is that earthquakes are inherently unstable, and the world simply does not have the means of predicting when and where they will occur. Like volcanoes, there can be rumbling for a long period of time, yet nothing major happens; conversely all hell can break loose without warning. Those convicted included some of Italy's most internationally respected seismologists and geologists, who undoubtedly gave their considered view. I don't see that they could have done any more.

This being Italy, no sentence is certain until there have been appeals, and it is possible that the verdict will be overturned. More than 5,000 scientists signed an open letter to the Italian President, supporting the defendants. The probability is that the Italian authorities will yet see sense. But as I say, probability does not say what will happen, merely what may happen.

Walter Blotscher