Jan 2, 2004

MORE CRICHTON OUT LOUD

by Cliopatria

Ralph Luker sent me the following link to another speech of Michael Crichton. It was the Caltech Michelin Lecture, given January 17, 2003

Unlike the previous speech that I railed about a couple of weeks ago, Crichton does give some interesting facts on science and public policy, particularly concerning second-hand smoke.

However in other ways, it is as dishonest as that first speech. The worst statement was this one:

"There is no such thing as consensus science. If it's consensus, it isn't science. If it's science, it isn't consensus. Period."

To put it simply, he is on, if not over, the ragged edge of lying. But he is very subtle at it. He's doing it this way: he's conflating the idea"a consensus of scientists in a field indicates that the best science concludes that a=b" with the idea"a consensus of scientists proves that a=b."

The latter is indeed false; the former is not. A consensus generally does indicate the best science. But he wants the reader or listener to confuse the second statement with the first and then agree with him that citing a consensus indicates a weakness in an argument.

Why? Apparently because he wants to discredit the use of current environmental science in public policy. He creates a situation in which anyone who cites a scientific consensus as evidence in support of an environmental policy is discredited while anyone who cites a lone figure in opposition to the same policy is treated with respect.

The really sad thing is that at the end of this speech he makes an interesting suggestion about how to do better science on public policy issues. It is one worth considering.

But given the context it is hard not to wonder if this is another slight-of-hand like the one I dissected above.

comments powered by Disqus

More Comments:

Oscar Chamberlain - 1/6/2004

Fine. When you discuss the question of identifying the most qualified scientists on a given question, you are demonstrating part of what I would like to see. I don't say that it is a neat process; I do think it is doable.

On your last point: you and Crichton apparently agree that scientists have been hopelessly corrupted. Yet you both give examples of good science or scientists. This suggests that it is not hopelessly corrupted at all.

Orson - 1/6/2004

"As an example, if a large majority of climatolgists agree with the research this has concluded that global warming is a real phenomena, occurring in part because of human action, and having major unfortunate consequences, that consensus can and should be useful in shaping policy."

But the standing implication is false. Actual, practicing climatologists are a famoulsy small and very cautious group. The "scientists" who clamor about the "consensus" on anthopogenic climate warming are itself a creature of public policy funding (currently arround $2 billion annually in the US alone), whose backgrounds and expertise are generally in other fields. This situation led the famous hurricane predictor, Professor William Gray of Colorado State University, to decry a world where all sorts of fanciful global warming research gets funding while his emminently practical and more successful research does not!

QED - Crichton.

But I would go further than Crichton and assert the relvance of the corruption of science through government- as demonstrated early in the nnuclear age and its "Atoms fo Peace" mantra half a century ago -simply continues today. Only under more "fashionable" agencies - almost all "environmental, " and universally and unjustly bearing equally erroneously sainted white hats.

Oscar Chamberlain - 1/5/2004

I think part of the problem here is that Crichton was not simply talking about how to do good science. He was talking about how science should be used in public policy.

Non-scientists who are politicians, regulators, citizens, whatever, need a functional way to find out what good science is, whether this involves global warming, earthquake impacts, or something else.

As Crichton himself points out, there is a problem here because a great deal of science is funded by--and therefore potentially influenced by--people and groups with political interests.

In short, you can find someone with the right abbreviations at the end of his/her name who supports or opposes almost any conclusion.

It is at this point that consensus among scientists becomes useful. As an example, if a large majority of climatolgists agree with the research this has concluded that global warming is a real phenomena, occurring in part because of human action, and having major unfortunate consequences, that consensus can and should be useful in shaping policy.

Does this "consensus" guarantee that they are right? No. In any field, there might be a new Galileo out there right now whose work will lead to the overturning the existing consensus. Even without a revolution, any consensus will evolve.

But if one wants relevant good science to help in determining public policy, then one must look at the science of the present. And, on the whole, it is better to have government officials look for the exisitng consensus because non-scientists are competent to do that.

And because that approach is more likely to lead to good information and advice than they would get by looking for the new Galileo.

C.R.W. - 1/5/2004

Oscar,

Without having read his speech in its entirety, I believe the first statement can be logically permissible. Science is a process of producing new knowledge. Whatever knowledge (or, if you like, "consensus,") already exists, does not constitute science. Etymology aside, science in and of itself is merely a process to be conducted - entailing the design of an experimental set of controls and gathering the resulting data. Analyzing and discussing the results and how they fit into the current framework of knowledge is another crucial process that must be conducted before consensus even means anything.

The fact that all four of these activities can be repeated, reapproached and reinterpreted means that the static connotation implicit in a term such as "consensus" is decidedly less meaningful when it comes to science. Until such time as everything in the universe is deterministically known to an absolute degree of certainty, anything may be up for debate - at any time. In other words, once we have absolute consensus, the need for conducting science will no longer exist.

Meanwhile, for the rest of us, a p-value of about 5% is usually considered sufficient. ;-)

(That and subjectively satisfactory definitions and premises.)

Jonathan Dresner - 1/3/2004

His handling of SETI and the Drake equation reveals his deceptive intent. (Full disclosure: I've got SETI screen savers on my computers) He describes it as "literally meaningless." It's a theorem, and no, it can't be thoroughly tested with current technology, but that doesn't mean that the SETI project is entirely meretricious. Strict adherence to a particular set of numbers as inputs might fit his definition of the SETI project as "a religion": in other words, to preclude the possibility that the Drake equation has a positive result is a faith in and of itself. Contrast his handling of SETI with the discussion of consensus in science and he's a push-me-pull-you with a big speaking fee.