So let's chew on it, then

(It is necessary to read my preceding post to understand this follow up.)

My title is derived from my previous post which raises an almost 'perfect' storm' of intractable problems.  First of all, should scientists be held legally responsible for forecasts they make in areas of science which are inherently 'chaotic', that is, areas in which exactitude is impossible and only probabilities can be offered?  A prime and infamous example was our very own Michael Fish from the Met Office who, back in the '80s, laughed off the silly notion of approaching storms and we all woke up the next morning with perhaps one the biggest weather disasters the country had ever experienced.  Mind you, Mr. Fish, rather foolishly, did not talk in terms of probabilities, he simply denied that it would happen. 

Returning to L' Aquila in Italy, the seismic scientists, who try to seek certainty in an equally 'chaotic' sytem, did not tell the people that there would be no earthquake, they limited their forecasts to the fact that there was a 2% chance of an earthquake.  At this point we move from science to language.  How many people truly understand the inherent meaning of 2%?  And how many can weigh that 2% against the catastrophic disaster if they 'bet' the wrong way?  I use the word 'bet' advisedly because there are far more punters than there are bookmakers and on the whole punters tend to end up poorer than bookmakers which indicates to me that very few people understand the mathematics of probability, or, 'odds', as the bookies call it.  Let me put it this way, if a seismologist tells you that the chances of an earthquake is only 2 in a hundred - but you will die if the 2% comes up, then that sets the problem in a new framework for a punter to decide.  But, I am tempted to ask, why 2%?  Why not 3%, or 1%, or even 5%?  Scientists like exactitude but is it responsible of them to provide an exact probability when in fact there can be no exactitude?

This is an important point because increasingly inexact science is being used to propel social-political aims.  I think back to the days when the anti-smoking lobby insisted that lung cancer was caused by smoking where-as even a quick glance at the charts indicated that age was obviously an important factor but somehow the slogan 'Die young and you won't get lung cancer' is hardly inspiring.   Also, sex rears its ugly head!  As you can see from this chart, smoking trends are down in men and so are the incidents of lung cancer.  But women, too, are smoking less but their cancer rates are increasing!

I have no wish to be drawn into that old debate, I use it merely as an illustration that scientific pronouncements of a statistical nature should come with a 'health warning', particularly if they are issued by scientists expert in one specialist subject but not necessarily equally expert in the arcane complexities of statistical mathematics.

But the overhanging question is, should scientists be held legally responsible for their pronouncements?  Should we, to put it crudely, have sued the arse of Michael Fish?  Here-in lies the dilemma facing the Italian judges asked to pronounce on the guilt or otherwise of the Italian seismologists whose pronouncements were wrong and 308 people died.

(Alas, I am off, reluctantly, to the dentist but I shall return - I hope! - because there are even juicier difficulties arising from this tragedy.)

Comments

[Peter MacFarlane]

What Mr. Fish actually said what that there wasn't a hurricane coming.

That statement is unexceptionable because hurricanes don't - cannot - happen far outside the tropics (and 55N is a long way out of the tropics).

Since he was a real meteorologist (not like the weather-girls of today who are chosen solely for their appearance), he meant what he said, and he was correct.

People chose to confuse the terms, that's all, and that is why he's been laughed at for it ever since.

Unfair really.

[Whyaxye]

Yes, Peter MacFarlane is correct. Fish gave a very clear warning that severe storms were on the way, but his response was to a question about rumours of a hurricane. At the time, the met office in general would not have been able to accurately gauge the severity of the storms, so to expect more from him would be wrong in several respects.

You pose an intriguing set of questions. My take on this issue is that we have a natural need to feel safe, and as large-brained speculative creatures we partly accomplish this by means of prediction. Hitherto, this was (arguably) met by the extremely broad brush-strokes of religion, and with increasing secularisation this burden is now likely to be placed onto scientists and other "experts". Of course, scientists are the first to acknowledge the limits of their predictions - hence the probability stats, which are currently creeping into weather forecasts, health warnings, and the like. But as these deal with the "big picture", and individual punters want to know about what will happen to them. Their needs for certainty are unassuaged, and they will just make up the detail to suit themselves.

Add lawyers and a "compo culture", and stand well back!

You might also want to consider the legal/ethical results of expert pronouncements in, say, economics. Apparently, some Health Authorities have been saddled with huge ongoing debts as a result of Gordon Brown's fondness for PFI. Now who could possibly have forseen that?

[dearieme]

"Should we, to put it crudely, have sued the arse of Michael Fish?"

But his arse was guiltless.

[Whyaxye]

dearieme

I thought it was a different typo, and Mr. Duff was suggesting sewing up poor Michael's orifice as a punishment

[David Duff]

DM, what would I do without you and your eagle eyes? I shall not correct it, some typos are too good to alter!

Gentlemen, from Wiki (and therefore with the usual 'health warning'!

"he said during a forecast: "Earlier on today, apparently, a woman rang the BBC and said she heard there was a hurricane on the way... well, if you're watching, don't worry, there isn't!". That evening, the worst storm to hit South East England since 1703 caused record damage and killed 18 people." [...]
In later years, Fish claimed that he had been referring to a hurricane in Florida,[2] USA in a link to a news story that preceded the weather bulletin. But he did not mention Florida in his forecast, which was made amid widespread worries about a coming storm: that morning, the Surrey Record had warned of "furious gales", so both his caller and his viewers likely believed he was referring to Britain. Fish did go on to warn of high winds for the UK, warning viewers to "batten down the hatches", although the storm that actually occurred was far stronger than he had predicted (albeit, technically not a hurricane)."

All of that (including my typo) takes me on to the next complexity in this confusing tale - language. (See above, er, when I've written it, that is!)