Adventures in Ethics and Science

One hundred years ago today, 18 April 1906, a major earthquake (estimated to be 7.9 on the richter scale) nearly destroyed the city of San Francisco and did some serious damage to other communities in the area. Here in the Bay Area, there are various commemorations of the event taking place, and the local papers have all hit the vaults to dig up accounts of the quake, and of the fires that followed. (See, for example, the Chronicle’s “Great Quake” page. Of course, the U.S. Geological Survey has a page with great quake links, too.)

So in a very obvious way, you could say there’s lots of earthquake “awareness” in these parts. But the commemoration (some even call it a “celebration”, although no one is actually celebrating the destruction and deaths caused by the quake) has an uneasy feel to it.

We know a lot more about earthquakes than they did in 1906 — and here, by “we” I mean both scientists and lay people. (You can look at this very nice GSA Today article about how our understanding has improved. Or, check out this set of ten short articles from USGS targeted at students, teachers, and parents.) Plate tectonics, man; the plates are in motion, there’s friction, stress builds up and needs to be released. You can work it out a little at a time (perhaps with frequent, smallish earthquakes). Or, after a period of relative quiescence something big can happen.

Those of us who live in the Bay Area know that there are faultlines running through our whole lives — home, school, work, recreational areas, the roads and transit routes that connect them all (’cause everyone commutes here). We know that stress is building up along each of the major faults, and that eventually something will have to give. Maybe a few times a year we feel a rumble and we go online to see if it was a quake or just a big truck driving by. Some of us find the little temblors reassuring — there goes a little of the stress.

But, they could just be the prelude to The Big One.

And, the geologists tell us, the probability of a major quake (magnitude 7 or more) in the next few decades is better than 50%. It’s not a matter of if, but when.

Yet hardly any of us have real earthquake kits, or plans for dealing with the aftermath of a major quake.

It’s not because we think the geoscientists can prevent a quake. Nor, I think, is it because we have total faith in all the seismic retrofitting that has been undertaken since the Loma Prieta quake (a 6.9) in 1989.

I think it may come down to the probabilistic nature of our knowledge here.

Geologists aren’t in a position to say, “We’re expecting a quake along the Hayward fault, with this approximate epicenter, in the 6.8 to 7.3 range, in May of 2006; take appropriate precautions.” A month is a ridiculously small window of time, geologically speaking. Even if it weren’t, how high a probability of seismic activity would there have to be before you could announce that something is going to happen? Greater than 0.5, surely, and probably less than 1.0 … but where’s the cut-off? And, given that improbable stuff actually happens (and probable stuff sometimes fails to happen), how do you avoid becoming the earthquake predicting agency that cried wolf? (“Well, they said all hell was going to break loose on that fault, and we battened down, and not a darned thing happened. Alarmist scientists!”)

David Hume’s problem of induction just doesn’t bother most folks. The ground stood still yesterday. The ground stood still today. No good reason to plan on the ground not standing still tomorrow.

Maybe the risk of earthquakes would feel more real to us, and have more of an effect on our planning, if we could feel the seismic storm gathering like a thunderstorm or a blizzard. Right after the Loma Prieta quake, and then after the largish aftershock we had about 6 months later, people talked about “earthquake weather”, but I’ve felt that same “unseasonable” weather dozens of times in the 16 years since with no shaking associated. Unless you’ve got instruments in the ground monitoring the stress along the faults, you’ve got nothing to go on. From the individual’s point of view, it might as well be 1906. The Big One is going to surprise you.

And maybe this is just how things are. Maybe we expected science to turn the whole world into something completely predictable, whose trajectory we can compute accurately to many decimal places given sufficiently precise information about where things are right now. But maybe the best science can do, at least in certain realms, is tell us what forces are at work, and what outcomes are most probable, and we’ll just have to wait and see which of the possibilities become actualities. Maybe we’ll be in a position to understand quite a lot, but not always in a position to predict precisely what’s going to happen at any given moment.

Probably, embracing the limits of our science shouldn’t put us off from assembling emergency supplies. Really, I should do that soon. But what are the odds that the next Big One would strike at such a perfect interval from the last Big One?


  1. #1 Kevin Vranes
    April 18, 2006

    Good stuff. Some thoughts:

    First, this article is illustrative of where you’re going here.

    In doing my current quake research I came across this Peter May paper (he’s a U of Washington science policy prof). He essentially says that people in high-risk quake areas are well aware of the risks yet ignore them. Lately I’ve been arguing that the same might be true of climate change and that if it is, it changes the policy outlook.

    Quite a few politicians spoke at the conference today and I was impressed that almost all of them, but especially Mayor Newsom and the Governator, mentioned the lack of personal preparedness. They all repeated the stats that only 25% of Bay Area households have emergency plans and even fewer SF households (8%?) do. Newsom is pushing hard

    I’m staying with friends who live on what I think is a very vulnerable slope of the Oakland Hills. They don’t have quake insurance so I last night told them the probabilities and asked if it changed their outlook on insurance. The probabilities are 62% chance of a Mw 6.7 or greater quake in the next 30 yrs. I expected them to say that it didn’t sound like a big deal, but they said the opposite — that now that I put the risk in real terms with a real timeline, it sounded more ominous. Maybe it is a matter of public education.

  2. #2 Janet D. Stemwedel
    April 18, 2006

    We did actually have emergency supplies assembled for Y2K (which … yeah, not so much). And we have our Costco stockpile in the garage (with juice pouches and Fresca in place of the bottled water), a hand-crank radio, and such. Probably we should progress past the half-assed “we won’t starve” supplies to a dedicated “emergency” stockpile, but then we’d need remember to roll over the supplies every so often.

    More importantly, we ought to have an assembly plan (given that the four Free-Riders currently are in four different places during the school/workday) and more robust agreements with neighbors about who will hang out where under various emergency circumstances. If the cell phone networks (and internets) go down, there will be an awful lot of making things up as we go.

    The wide range of outcomes even from a big quake make it seem that much harder to plan rationally. The Loma Prieta was my first quake; a bookshelf fell and a cantaloupe rolled off the top of the fridge. And, a water main broke on campus, so I had to bike to a friend’s house to shower. And, there was a big old mess at the chem stockroom, but I wasn’t teaching lab that day; I can’t even remember is we missed a class meeting. Of course, other people lost their homes (or worse) in the same quake. Which outcome should you expect the next time around? (Does prudent planning mean you give up driving on bridges altogether?)

  3. #3 boojieboy
    April 19, 2006

    I harp on this point with my students all the time. Science cannot give us perfectly certain knowledge. It can only give us imperfect knowledge, but with the imperfection decreasing over time, until (maybe) the imperfections become so slight that it resembles certain knowledge.

    Earthquake prediction. Weather prediction. Epidemiology. Predictions of what an individual might do in the future. All suffer the same problem: lots of causal models and mechanisms understood, but very few of those models allow fine-grained predictions. Problem is that people want certainty, not “calculated uncertainty” and unfortunately too many trained professionals are willing to feed them predictions framed in language that gives the illusion of certainty to feed that need.

    Witness the nightly weather reports, with their 5 day forecasts. One meteorologist in my area gives 7 day forecasts. What a joke.

    Witness the psychologists who are all too willing to go on Oprah and expound on individuals with various problems, or to declare after the fact why exactly some kid got a gun and shot up his school, or what sort of person is currently running around killing people in some area (remember the psychological profiling of the DC sniper?)

    In terms of earthquake prediction, this happily seems to be one realm in which the working scientists remain firmly in control of the language in which predictions are expressed. Predictably, many in the public are unhappy with this kind of probabilistic prediction, and so many choose to ignore them. What can you do?

    Now we’re faced with another such situation: the predictions that some are trying to convey about the possibility of an upcoming pandemic. Same problem. Here, even the working scientists disagree so much that people choose to believe whoever’s prediction fits their preexisting biases. Most don’t want to think about the possibility that 100M people might die worldwide in a flu pandemic. So they choose to believe the few experts out there who’s message seems to be it won’t happen, some even declaring it CAN’T happen. See? Language that conveys certainty, where no such certainty exists.

  4. #4 Uncle Fishy
    April 19, 2006

    We actually ate the earthquake survival kit on Sunday. (Country Ham may actually not be the best kit as it needs cooking. I’m going with a whole prosciutto next time around.) But other than the gallons and gallons of water, the over stocked pantry and the gobs and gobs of camping gear what’s one to do? We’ve identified a meeting place (your house. Did you know that?) As renters, insurance isn’t really an issue and should the house collapse anyhow there goes our earthquake survival ham, err I mean kit.

  5. #5 David Harmon
    April 19, 2006

    People acclimatize to their local risks, especially the most routine. (Traffic deaths, anyone?) And they’ll happily ignore uncertain hazards in the face of present rewards. Remember, the people of Pompeii knew perfectly well that Vesuvius was an active volcano. So did the folks growing grapes on its fertile slopes.

  6. #6 dan c. williamson
    April 21, 2006

    I’d like to add that there’s a specific economic aspect to all of this. Not so many people even have garages to stuff survival gear into because they don’t have the house attached to that garage. This doesn’t account for the fact that many people live on the edge of economic survival to begin with. What would happen to all of the homeless and marginally homeless in a burg like San Jose, let alone San Francisco? One shudders to think. There are going to be different needs for people at different economic strata as well as different living arrangements.
    This adds up to a complex of problems that could reduce to more “community awareness” as well as city awareness of the same problems. Since when was this remotely possible, esp. city/county governments? I’m not convinced they even have really substantial plans for ANY kind of disaster, including earthquakes; perhaps I’m being overly cynical. I know there’s some stuff out there, but where’s the coordination and the public awareness of same? For things like bio-toxins, and other pieces of nastiness? It always seems to be the bureaucrats answer–we’ll send it to committee. Academics like us are only to aware of the Kafka like quality of that “syndrome”!! Or should I say symptom?