Predicting L'Aquila Earthquake: is it right to blame the scientists?

Summary

---

Podcast Summary: More or Less

Episode: Predicting L'Aquila Earthquake: is it right to blame the scientists?
Date: October 29, 2012
Host: BBC Radio 4 – Jan Wong
Main Guest Experts: Prof. Ian Main (University of Edinburgh), Prof. David Spiegelhalter (Cambridge University)

---

Overview of the Episode

This episode examines the controversial prosecution of Italian scientists following the 2009 L’Aquila earthquake. Six scientists and one ex-government official were sentenced to jail for manslaughter after allegedly providing “falsely reassuring” statements before a deadly earthquake, sparking outrage in the global scientific community. The episode delves into what science can realistically say about earthquake prediction and the consequences of miscommunicating scientific uncertainty.

---

Key Discussion Points & Insights

1. The L’Aquila Earthquake & Scientists on Trial

• The 2009 earthquake in L’Aquila, Italy, led to 309 fatalities and immense public anger ([00:34]-[00:56]).
• Prosecuted scientists were accused not of failing to predict the quake, but of making statements interpreted as minimizing the risk ([01:16]-[03:16]).

2. Can Earthquakes Be Predicted?

• Host directly asks: Is it possible to predict earthquakes?
  • Prof. Ian Main: “The short answer is no.” ([01:21]-[01:23])
• Earthquake risk can be forecasted statistically, but not predicted precisely for a given time and place ([01:33]).
• After an earthquake swarm, probability of a major quake increases, but remains small—often up to 1% ([02:33]-[02:59]).
  • Prof. Main: “So the probability could be on the order of up to, say, 1% that a small Earthquake or series of small earthquakes might trigger a bigger one.” ([02:40])

3. Communicating Risk and Public Response

• Prior to the earthquake, residents would customarily sleep outside after tremors. This changed after scientists’ comments, resulting in some staying indoors ([03:16]).
• The issue was not failed prediction, but communication: scientists conveyed “unlikely” as “nothing to worry about”—a crucial distinction in risk messaging ([03:16]-[04:19]).
  • Key Quote (host): “This case isn't about the scientists’ ability to predict earthquakes. It's about their statements communicating the risk of an earthquake.” ([03:16])

4. Scientific Responsibility & Legal Consequences

• Worldwide scientific outrage at sentencing; concern over chilling effect on expert advice ([04:19]).
  • Prof. David Spiegelhalter:
    • “It seems to me this is a triple tragedy. ... But now there's been this additional tragedy of these seismologists being prosecuted for large fines and ludicrous six year sentences.” ([04:19])
    • “But there's the chance of a third tragedy which is just the withdrawal of the scientific community from engagement.” ([04:27])
• Scientists may avoid public advisory roles for fear of legal risk; possibility of “defensive science” ([05:39]-[06:05]).
  • Indemnity protections exist for government advisors and weather forecasters, but not universally for scientists ([05:03], [05:39]).
  • Spiegelhalter: “It would be terrible if we started practising defensive science... But of course, if people are worried, that's what will happen.” ([05:52])

5. The Broader Issue: Understanding Risk

• Host’s summary: Sometimes, misunderstanding numbers, risks, and uncertainty can be a matter of life or death ([06:05]).
• Urges need for clear public risk communication and protection for scientific advisors ([06:05]-[06:21]).

---

Notable Quotes & Memorable Moments

• Prof. Ian Main (01:21):
  “The short answer is no.”
• Prof. Ian Main (02:40):
  “So the probability could be on the order of up to, say, 1% that a small Earthquake or series of small earthquakes might trigger a bigger one.”
• Host Jan Wong (03:16):
  “This case isn't about the scientists’ ability to predict earthquakes. It's about their statements communicating the risk of an earthquake.”
• Prof. David Spiegelhalter (04:19):
  “It seems to me this is a triple tragedy... But now there's been this additional tragedy of these seismologists being prosecuted for large fines and ludicrous six year sentences.”
• Prof. Spiegelhalter (05:52):
  “It would be terrible if we started practising defensive science. And the only statements we made were sort of bland things that never actually changed, drew one conclusion or another.”

---

Timestamps for Important Segments

• 00:34 – Introduction: Sentencing of Italian earthquake scientists
• 01:21 – Can earthquakes be predicted?
• 02:08 – Swarms and probability increases
• 03:16 – The crux: Communication, not prediction
• 04:19 – International scientific reaction with Prof. Spiegelhalter
• 05:29 – Risks in other professions (weather forecasting)
• 06:05 – Lessons on risk, statistics, and communication

---

Tone and Style

The episode maintains a clear, rational, and slightly wry tone, presenting complex statistical nuance in plain language. Guest experts speak directly and candidly, emphasizing the limitations of science and the societal importance of honest risk communication.

---

Conclusion

This episode vividly illustrates the challenges scientists face when communicating complex risks to the public. While science cannot offer certainty, misunderstanding probabilities can profoundly affect public safety and societal trust in experts. The host and guests call for better public understanding of statistical uncertainty and for protections enabling scientists to advise frankly without fear of prosecution.

---

Transcript

[A]

Thank you for downloading from the BBC. For details of our complete range of podcasts and our terms of use, go to bbcworldserveys.com podcasts

[B]

hello and welcome to More or Less on the BBC World Service, your weekly ration of numerical nutrition. I'm Jan Wong. This week we'll be reporting on a rogue Indian statistic which has fueled a full scale political row.

[C]

While in his address to the Punjab university, Rahul said seven out of 10 youth in Punjab were drug addicts.

[B]

But first, this week, six Italian scientists and one ex government official were sentenced to jail for manslaughter. Why? Because they made overly reassuring statements before this happened.

[D]

A big earthquake in Italy. There's widespread damage to a medieval town where many are dead.

[B]

The earthquake, which hit the Italian town of Lacchila in March 2009, killed 309 people. Perhaps we shouldn't be surprised that amid the grief and anger which followed, some in Italy wanted to apportion blame. But is it right to blame the scientists? Well, at its heart, this is a story about numbers and probability. So let's start with an obvious Is it actually possible to predict earthquakes?

[E]

The short answer is no.

[B]

That's Ian Main, professor of seismology and rock physics from the University of Edinburgh, Scotland, and also a member of a committee on earthquake forecasting following the tragedy in La Quila.

[E]

In another sense, we can forecast that they are almost certain to happen with a very low probability, maybe once in a hundred, once in a thousand years, depending on the size of the earthquake in a given zone. And that's a hazard that's there all of the time. A fault can go at any time, but we know it's going to happen.

[B]

And in the case of this earthquake in Italy, it was preceded by a whole series of smaller tremors. I assume that that means you can change the estimate of probability that a major earthquake will happen after that. It changes the equation somewhat.

[E]

Yes. One of the things that's fairly consistent about earthquake population dynamics is that they often occur in clusters, and we can understand the physics of that in terms of stress reliability redistribution caused by small earthquakes or bigger ones, and the fact that that then makes earthquakes more likely to occur.

[B]

So how much more likely is an earthquake to occur if there's been an earthquake swarm, as they're called beforehand.

[E]

So it can be significantly above the background rate. So instead of once in 100 years, it might be once in a few months or something like that. So the probability could be on the order of up to, say, 1% that a small Earthquake or series of small earthquakes might trigger a bigger one that would cause significant damage.

[B]

So the probability of an earthquake happening after a series of these events could go up by more than a hundredfold, but it still goes up to a very small amount, 1% or so.

[E]

It's a significant probability gain, but it's still low absolute probability. There's on the order of or less than a 1 in 100 chance of a swarm ending in a larger earthquake or a smaller earthquake triggering a bigger one. It's very rare, but it does occur.

[B]

So 99 times out of 100, a series of tremors won't mean there's about to be a larger earthquake. But this time there was. Now in the past, in La Killa, after numerous tremors, some residents would leave their houses and sleep in their cars. However, on this occasion, comments by the authorities prompted some of those people to change their usual response and stay in their homes. This case isn't about the scientists ability to predict earthquakes. It's about their statements communicating the risk of an earthquake. The Italian seismologists thought an earthquake was unlikely, but in a press conference the message seemed to be that that meant there was nothing to worry about at all. This is the falsely reassuring statement which formed part of the case against them. The scientists do have the right to appeal, but their jail sentences have been viewed with surprise and in some cases outra by statisticians and scientists worldwide like David Spiegelhalter, professor of the Public Understanding of Risk at Cambridge University in the uk.

[F]

It seems to me this is a triple tragedy. Obviously we shouldn't forget the terrible case of 300 people that died there. And that's the main thing. But now there's been this additional tragedy of these seismologists being prosecuted for large fines and ludicrous six year sentences. But there's the chance of a third tragedy which is just the withdrawal of the scientific community from engagement. I mean, who's going to be a seismologist in Italy now? I mean, Italy has got continuous earthquake problems. They've got one in Calabria at the moment. Now who is going to engage in that now in Italy? It really makes you wonder about the risks to scientists of engagement and being advisors.

[B]

What do you feel needs to happen now to encourage people, I guess, to give independent advice without getting into these problems?

[F]

Well, I think they have to be reassured that their work is going to be used appropriately and communicated properly. And I think that's also the scientists responsibility. I actually think there are legal aspects as well about getting indemnity to civil cases. I mean, government advisors are supposed to have indemnity. I've been on a committee once and I found I was the only person in the room who didn't have indemnity. And at that point, you start feeling slightly exposed given the increasingly litigious society. And that's an awful shame.

[B]

I presume there are other areas where professionals make forecasts and where they may or may not be met. And the consequences could also be devastating.

[F]

Well, I understand for weather forecasters who might fail to predict a hurricane or some severe weather event that might in fact lead to loss of life, my understanding is that they are indemnified against that. Nobody's going to be able to sue them for that statement. It would be terrible if we started practising defensive science. And the only statements we made were sort of bland things that never actually changed, drew one conclusion or another. But of course, if people are worried, that's what will happen.

[B]

David Spiegelhalter and one of the lessons of all this is that understanding numbers, risks and uncertainty can sometimes be a matter of life and death. Now, a little while ago, we received an email from listener Movin Miranda in Malaysia.

[F]

There is some furore in India, my home country, about the prince in Waiting Rahul Gandhi's comments on drug addiction levels in Punjab, India. I thought you might be interested.

[B]

So Raul Gandhi, the General Secretary of India's Congress Party, caused a political storm by claiming that 70% of young people in Punjab are drug addicts. Well, more or lesser's Charlotte Pritchard has been looking into this one. Charlotte, what have you got?

[D]

Well, our listener is right. Gandhi's comments have triggered a huge row in India where he's been accused by other politicians of being flat out wrong. And I have to say they have a point. The number has been quoted a lot, not just by Rahul Gandhi. It's even been referred to in the Indian High Court, which is why some politicians say they thought it was right. To get to the truth, I spoke to Dr. Sandhu from the Guru Nanak Dev University in Punjab because the number comes from his study. Dr. Sandhu says his research has been repeatedly misquoted, including by Rahul Gandhi. His study didn't look at a sample of young people to see how many of them were drug addicts. Instead, he looked at a sample of drug addicts to see how many of them were young people. A very different thing.

[B]

Okay, so that statistic is plain wrong. It's like saying 90% of murderers are men. So therefore 90% of men are murderers. That would make the world a much more dangerous place.

[D]

And that's why this statistic is so incorrect. However, there's a pretty big drug problem in Punjab. To find out why, I called Rajeev Shaw, who runs an organization called Sharan, which works with drug users across India. He says that there's a plentiful supply of drugs in Punjab because they're smuggled over the border and that traditionally there has been a high use of opium in the region.

[C]

There are a large number of people who are addicted to drugs and the numbers are increasing.

[D]

Why is the problem particularly acute in Punjab in comparison to the rest of India?

[C]

I think it's got more than one reason. It's very close to the borders. There's a lot of trade of heroin which comes into the state. Then, of course, also the average youth or the average family in Punjab is, I won't say cash rich, but they're better off than most people in other parts of the country. So they have easy access to money then. Traditionally, Punjab has had a sort of culture of opium use, so drugs have been sort of a part of the state for a long time.

[B]

Rajeev Shaw, thanks to him and of course to you, Charlotte. And that's all from us this week. Do please keep your questions and comments coming more or less at BBC Code. As you've just heard, we love to set out on statistical quests set for us by you, our listeners. Don't forget our website too. BBC.co.uk More or less. Where you'll find some of our best stories. More or less. We'll be back next week. Until then, goodbye.

[A]

There are dozens of different podcasts now available from the BBC, including news, documentaries, science, business, arts and sports. The details of them all go to bbcworldservice.com podcasts.