Wrong Question

Walter E. Williams asks "Can we trust experts?" That question is not the right question to ask in the current circumstances. It's premature, because two other questions need to be answered first.

1) Is expertise in this area possible?

2) If it is, how would we know who the experts are?

Both of these are old problems. The first one is Socrates' problem. The second one is the one that concerned Plato in his later political writings, starting with The Republic but also The Laws.

There are fields in which expertise is demonstrably possible. In ancient Greece, this kind of expertise was called techne -- the root of our word for 'technology.' It is a kind of craft knowledge, an ability to attain particular purposes. What Socrates wanted to know is whether those who claimed knowledge of things like the virtues could demonstrate their knowledge in the same way as shoemakers or shipwrights. It turns out that there isn't a clean break:
Still, by including commanding knowledge, the Visitor has left a middle ground between the purely theoretical and the practical. Certainly architecture is not practical since it does not directly produce anything, in the way carpentry does. However, it does give commands, whose effects are practical; thus, it is not for knowledge only, in the way in which calculation is for knowledge only. Insofar as architecture is an analogue for the political craft, the Visitor seems to be exploiting this middle ground[.]
Economics and politics would both like to think of themselves as possessing a kind of 'commanding knowledge,' akin to architecture. It is not clear that they do have this capacity, however, and we should insist that economists and politicians demonstrate a capacity for such knowledge before ceding to them any powers. We should certainly be ready to snatch away power from any of these so-called experts who prove to be incapable.

Even if that can be done, it still leaves the second problem. Let's say that economic expertise is possible, and that some small number of people really have it. How would I know who they are, if I don't have it myself? I can't judge directly because I lack the knowledge that I would need in order to recognize that they possessed it. I could trust the opinions of others who had the knowledge, but I have the same problem with identifying them. You are likely to end up exactly where we are, i.e., with a large class of mutually-reinforcing "experts" who affirm each other's claims to knowledge, but who really just agree with each other. If they're wrong, they are likely all wrong. Anyone they point to as another of their class is likely to be wrong too.

You might be able to judge from pragmatic experience, but that would require putting people in charge without being sure of their expertise (and, indeed, in spite of being told by all the experts that they were mad, fools, and the like). If they managed to produce good outcomes reliably over time, you could judge that they knew what they were doing.

I think we may be in the middle of such an experiment right now. It's a dangerous thing to do, but it may be the only way to discover expertise in the face of an established class of mutually-reinforcing pretenders to knowledge.


Assistant Village Idiot said...

We are experimenting with whether children need two parents. We shall see.

David Foster said...

I just put up a related post: On Trusting Experts--and Which Experts to Trust:


Grim said...

That’s an interesting post. The discussion in your comments section also has some insights.

J Melcher said...

This reminds me of James Randi and the psychics. Can a person do what he (or she) says he can do? If so, how? Can the method be taught?

Often, "how" was (Randi revealed) essentially "cheating". A prognosticating psychic would for instance predict multiple outcomes to multiple people and eventually a few people would be left who had only seen the outcomes that had turned out to be correct. Or the clairvoyant psychic would "peek" around the blindfold or open the box, when it was thought nobody else was obverving. Cheats.

One interesting case was a guy who claimed to be able to read the pattern of grooves in an LP phonograph record and so identify the piece of music recorded. Randi showed up with unlabled records -- and the guy COULD identify classical music, symphonies and such, but not Alice Cooper or other modern pop music. "Not much structure" the guy said of the grooves in such records... To me, that particular experiment revealed less about the guy making exceptional claims than the fundamental experimental honesty of James Randi. Why, yes, sometimes it does turn out that the oddball crank claim is true.

The point of it all though is that ordinary expertise, making an UN-exceptional claim -- is pretty safe. You do, or do not, have a broken tie rod, sayeth the auto mechanic. You do, or do not, owe taxes on withdrawals from that particular investment, sayeth the CPA.

What trips us up are the exceptional claims for which there is only subtle evidence.

douglas said...

I'm not sure even the CPA you can readily say is expert in the material way- as systems like tax codes get more and more complex, it's impossible to master the entirety of it, and you've really been reduced to a guide into the unknown, but you only know what you know- and there's still a fair amount of unknown, including whether or not this particular course is the best possible course or not- that's become almost impossible to determine anymore. I speak to this from my experience with building codes and their expansion.

David Foster said...

The rigidity problem faced by the Kaiser in the WWI deployment decision rigidity, (at least, as interpreted to him by Moltke) was interestingly mirrored in the French response to the Rhineland crisis of 1936. According to Andre Beaufre, who was then a young officer on the French general staff, there was only then *one* mobilization plan, and it called for calling up millions of men, requisitioning large numbers of private vehicles, etc. When asked by the political leadership whether it would be possible to conduct a smaller-scale mobilization...which was all that would have been required to repel the German forces at the time...general Gamelin's response was that NO, the necessary planning could not be done on short notice.

The politicians were unwilling to disrupt the entire country to the degree required by the existing mobilization plan, so nothing was done at all.