Thursday, November 26, 2009

Keeping your cool with global warming

Well, we just had to write, even over this holiday break (in the US), rather than just leave any hungry or un-turkeyed readers with just our Oulipo challenge to keep them occupied.

The story about the hacked emails from the University of East Anglia's program on climate change is making the rounds. A hacker discovered correspondence, including from faculty here at our own Penn State group on climate change, that appeared to indicate cover-up and cooking of the data about whether global warming is a true, anthropogenic (human-produced) fact.

As expected, the anti group, who want to keep on truckin' (with the exhaust emissions that go with it), are reveling in what they claim is a "Gotcha!" moment, showing the faking of evidence, with the earth-huggers responding in outrage that their innocent statements have been taken out of context.

Neither party seems to be against science in this case, so we're not into religion vs science conflicts here. But who's right? Clearly each side is coloring the evidence to support its pre-existing perspective. Here is yet another instance where tribalism, bullying, and ideology reflect the realities of the real world of human beings. So we offer another installment in the "What is 'evidence-based?' " series of posts.

We are scientists ourselves, so we know how language used as shorthand can be twisted, and how data can also be ambiguous. One of the hacked emails referred to a 'trick' referred to in a Nature article, which the smoke-spewers say shows that data were fudged if not faked. But 'trick' clearly meant analytic method of blending different kinds of data into a single unified analysis, done completely in the open, and not fraud in any sense. To get a continuous climate record from 1400AD to now, the authors had to blend indirect data such as from tree-ring patterns with direct instrumental measurements which were not available in the past. Was this legitimate? Yes, it seems to have been an entirely routine and appropriate.

There was other intercepted email correspondence about 'hiding' some tree-ring data set because of a change in responsiveness to climate. So says the news story. The Penn State correspondent is quoted as replying that nothing bad was done and the dirt-digging was just a way to undermine the 'strong consensus' that global warming is real.

But that is no kind of defense! Science is not a democracy and a 'consensus' is in itself no indicator of truth (or else we'd still be practicing Galenic four-humors medicine, and who knows what else). Comparably illegitimate would be the argument that we should excuse hiding data because, overall, the 'consensus' must be right. But the underpublication of negative results is well known and is a problem of that sort.

We don't know the details in this particular instance, but science always works from a theory or hypothesis. We are only human, and we like our own ideas, so we try to defend them against the evidence. It is routine to find ways to fit new data into our preferred theory, even when the data on the surface seem to conflict. There is a subjective element to this, in that we try to show how our theory really is true. Eventually, if enough 'bad' data are found, some other theory is needed, as the pretzel of contorted explanations is replaced by something more plausible. But no scientist gives up easily!

Scientists certainly color evidence and the way they present it. Negative results are under-reported, if reported at all, and weak positive evidence is highly touted. Of course this will happen, when careers, prestige, money, research resources, and publications are at stake! It's not just science -- after all, shell-games are how bankers picked your pockets and got away with it, no?

In the same way, those opposed to a given view will stress the problematic data and minimize the oomph of positive findings. We write about this all the time. It's a perfectly natural part of our culture. If you think it is not part of genetics and evolutionary biology and biomedical research, you are being naive (and have negligently not been reading our recent posts!)

In a way it's a kind of built-in dishonesty in what is supposed to be a bluntly honest area of life. Part of the problem here, as in other aspects of human social life is that most of us who pass judgment don't have the facts at hand, but have to base our judgment on what we read (even in science articles, since we, after all, didn't do the actual work ourselves). Yet, it's right and important to question and criticize the system, to keep it as straight as possible.

But criticism of a theory, as in the gloating embers of the coal-huggers over these hacked emails, can also be pure opportunism of those with something to gain that has zero to do with the real issues--such as profiting by clear-cutting a forest whether or not it damages our climate. In this case, from what we know, the emails were the usual kind of informal chat, perhaps even about coloring the data to present the best case of a theory in which the correspondents believed. So we think there was no fraud here, and no scandal.

But, of course, we're tree-huggers!


Anonymous said...
This comment has been removed by a blog administrator.
Ken Weiss said...
This comment has been removed by a blog administrator.
Anonymous said...

"But, of course, we're tree-huggers!"

I'll applause the honesty here... Now, if only the honest ones in the Crugate-scandal owns up.

Ken Weiss said...

No one is so objective as not to have a point of view. Everyone in science should be as aware as possible of his/her own predispositions, biases, and so on. The facts themselves always have subjective elements -- even if they are just about what things got studied and in what way.

It might be easier and more productive to argue about the facts if one didn't have to do it through a fog of rhetoric that hides or colors these underlying agendas.

But in any human society that's asking a lot.