Reflections on Feynman per J. Case

My next serious read (what's yours?) is Competition, a (mostly nontechnical) plea by prominent mathematician James Case for a post-equilibrium economics. A browse in the bookstore reveals that he has little respect for the equilibrium models favored by quantitative economists. I think he is starting from results in game theory. More on this anon; I haven't read it yet.

For the present, I just wanted to discuss something from his epilogue. An important message for "believers" (as I (not to mention James!) am described elsewhere) and "skeptics" alike:

Seldom has anyone explained what science is - and is not - as simply and well as Richard Feynman in his 1974 commencement address to the students at Caltech. Science, he said on that occasion, is nothing more than a method developed over the years for separating ideas that work from ideas that don't. Anyone who observes the same natural phenomena day after day, such as the ebb and flow of the tides or the barking of dogs in a village street, will begin to develop ideas about them. Try it and see. There's nothing scientific about having ideas. Everyone does that. Science, said Feynman, begins when somebody figures out a way to test an idea to see if it works or not.

...

Feynman devoted a substantial portion of his 1974 commencement address to the subject of scientific integrity. Scientists, he said, have a responsibility to other scientists - and perhaps to the public as well - not to fool themselves. "after you've not fooled yourself" he assured his listeners, "it's easy not to fool other scientists." But not fooling yourself is far from easy because, liking your own ideas, "you are the easiest one to fool". Scientists have been learning for generations - indeed are still learning - ways of avoiding self-deception. One such way, he hastened to add is to divulge every reason you can think of why your conclusions are only tentative and may yet be proven wrong.

...

The perpetrators of pathological science are guilty not of fraud but of self-deception. Enamored of their own ideas, and fully expecting their experiments to confirm their theories, they find confirmation where none exists and - entirely too often - rush into print with results that are easily disproved. Such behavior is irresponsible, because it creates unnecessary work for others. Yet those who engage in it are seldom accused of dishonesty.

...

There is a scene in Bertolt Brecht's play Galileo in which the master and his assistants are preparing to test the Copernican notion that the earth revolves around the sun. Galileo explains to the others that as a matter of discipline, their purpose must be to prove the earth stationary. Only if the ascertainable facts render that position untenable will they allow themselves to find in Copernicus' favor. In fact, says Galileo, "if we find anything which would suit us, that thing will we eye with particular distrust."

(Emphasis added; resemblance to any real "AGW skeptic" living or dead except those who happen to be economists is coincidental; Case is talking about the failure of mainstream economics to attend to this ideal.)

Let me add my own taxonomy here, back to the AGW issue and related themes. There are "skeptics" and "believers" and there are also investigators and pseudo-investigators.

Investigators cultivate certain habits of mind that enable the advancement of science, while pseudo-investigators cultivate those habits of mind which advance a particular agenda even as they attempt to make use of the justified credibility of the real investigators.

Unfortunately, and increasingly, the habits of mind of the investigator are somewhat unfamiliar to a public that when it rises above distraction and confusion retains a very utilitarian frame of mind. The network of trust among serious investigators no longer extends to the general public, which is prone to various distractions, some well-intended and some amazingly malign.

To be sure every auto mechanic and every plumber, every engineer and every MD, not to mention many other professionals of modest or exalted reputation, is an investigator in a very real sense. And one thing that infuriates me as well as some serious skeptics is the arrogant refusal of science to learn from more utilitarian professions some of the commercially successful techniques for refining a solution to a particular problem.

However, it doesn't follow that the expert in some domain understands and has thought about the methods by which the expertise was brought into the world. And here we see a certain hubris appearing again and again: "I don't understand it so it must be mumbo-jumbo". Not to say that every scientist really thinks about epistemology or needs to. But what scientists are looking for is deeper than a diagnosis or a repair strategy, and is based on a more diffuse platform. Furthermore, some things are hard to understand, because not many people understand them, and they may not be experts in explaining it, and there may not be enough demand for the knowledge to support people expending too much effort on the explanation, and it may take years of study to see the picture emerge.

The failure to provide a "complete explanation" is interpreted as caginess, but with twenty dedicated skeptics for each even marginally first rank climate scientist who has many other responsibilities, there simply isn't enough response to go around. Gatekeeping is the inevitable result. (Sometimes excessive gatekeeping happens for this and other reasons, but under the present constellations of forces and social groups some gatekeeping is inevitable.)

I've been exposed to a couple of great climatologists.

(Update: Specifically Ray Pierrehumbert and Francis Bretherton, who bear no responsibility for my beliefs, but whom I have been privileged to interact with at length, and who jointly hold primary responsibility for my deep respect for climatology as a respectable branch of physical science. And I cannot imagine myself or anyone else explaining much of what they say in a typical single hour's peer group conversation to an above-average electrical engineer (say, a graduate of Northwestern University's tech institute from the 1970s like myself) in less than three months of one-on-one full time exposition. )

There is more there than curve fitting or squashing facts into a preordained pattern. You can tell a person who has a healthy skepticism about his or her own ideas once you listen to them expound for a few dozen hours. That doesn't mean you can explain it.

On the other hand an AGW "skeptic" can be relied upon to celebrate certain results and mock others not depending on the quality of their evidence or their reasoning, but on whether or not the evidence is convenient for their beliefs. This is, in other words, pseudo-investigation, or in John McCarthy's formulation "lawyers' science".

On the other other hand, it is rare for a theoretical science of modest means and accomplishments (such as climate science) to abruptly become not only an applied science but a subject of controversy. And therein lies the reason we are still arguing over the parts of the picture that are not in great doubt among the community of actual investigators.

I don't think there is any poll, per Bi's recent suggestion, that can capture this. The only thing I can imagine improving matters is a re-established network of trust, a thing which has been deliberately undermined by the obvious malign social forces which seem to exult in promoting fear, division, hostility and suspicion whether it is warranted or not.