I wrote the previous post about a week ago, but didn’t feel like publishing it before anyone commented on the previous four published in rapid succession. It was only while telling H.A about Ashwin Parameswaran that I was inspired by the latter to re-check chapter nine of “Seeing Like a State”. That chapter is a celebration of “metis” or local, practical, or personal knowledge over “techne”, what we would consider more scientific knowledge. It included not only the folk-wisdom of agriculturalists, but the medical knowledge of doctors. I was particularly struck by the passage in which a doctor claims he can just by looking identify to a great degree of accuracy whether an infant was seriously ill and needed medical attention, but couldn’t fully explain what visual cue alerted his judgment. It brought back to me memories of Robyn Dawes’ “House of Cards” in which he gives numerous examples of psychiatrists claims to experience derived insight & intuition completely unsupported by performance data. In another of his books “Rational Choice in an Uncertain World” he explains (with some of the figures providing examples of Goofus behavior) how our minds are prone to that kind of error due to our inability to understand how randomness works and our susceptibility to confirmation bias. Dawes’ experience was with the notably flaky field of mental health, surely physical doctors are much better. Robin Hanson has been on a one-man crusade to take them down a notch and insist on the use randomized studies on health outcomes rather than relying on their expertise. Has anyone tried reconciling the clash of metis and techne? Malcolm Gladwell may have fumbled in that direction with “Blink”, but I don’t feel like reading it.
November 28, 2010
Robyn Dawes & Robin Hanson as antidotes to James Scott?
Posted by teageegeepea under Uncategorized[17] Comments
Advertisements
November 28, 2010 at 10:23 pm
Good post. Don’t have much to add right now which is a compliment.
December 12, 2010 at 3:43 pm
Does James Scott provide an antidote to Robin Hanson? That is the kind of antidote we seriously need. Prof. Hanson feels compelled as a blogger to dream up and post some new daydream of random grandiosity every day. He used to invoke supposed Bayesian reasoning but these days he merely labels the novelty a “forager” idea — despite not knowing much of anything about what foragers actually believed, or about much of anything else about any other real society for that matter. He then flames anybody who disagrees with his idiocy-du-jour as a “farmer”, as if that is somehow more backwards than foraging and as if believing in ideas that weren’t just dreamed up yesterday is hopelessly backwards.
Hyperrationality almost always leads to absurd results, but Hanson has moved goalposts of the dream-it-all-up-from-scratch worldview down field to blatantly preposterous ignorance on display. Forget you might have ever known anything about anything, much less that other people in the world know things, and just dream up ideas from scratch and insist that people take them seriously. (e.g. “You haven’t even read the paper” in response to criticism of a blog post). Despite said ideas being utterly untested insist that they are true and that people who disagree are backwards. When invoking Bayes leads to readers exposing the garbage assumptions that lie at the bottom of hypperationalist theories, switch to an absurd attempt to invoke traditional authority by imputing said novelties to illiterate “foragers” who conveniently never left records. Sit back and watch the gullible suck it all up.
December 12, 2010 at 4:22 pm
You’re criticizing Prof. Hanson on the grounds of irrationality but you’re calling yourself antihyperationalist. I don’t get the disconnect, and I think you’re naming yourself (and your opponent) unhelpfully.
December 12, 2010 at 5:09 pm
One could say that hypperationality is the irrational overuse of rationality. But to say that would still be to privilege rational process over accumulated knowledge and observation as the _sine qua non_ of productive thought and communications. Such privileging is the basic flaw that leads to hyperrationality.
What I mean by hyperrationality is the making a fetish out of rational processes (e.g. the Bayesian cult) at the expense of examining the assumptions fed into said processes (which it turns out, once you have wasted your time discovering them, are almost always very flawed), at the expense of empirical observation, and especially at the expense of the vast knowledge that has already been accumulated by billions of brains as smart and observant as ours that have already lived and experienced the world. A hyperrationalist thinks that methodologies as seemingly rigorous as those of theoretical physics are the most productive ways to understand the relationships of people. A hyperrationalist is a smart-sounding know-nothing.
Hanson illustrates many of the problems with hyperrationality by taking it to such absurd lengths. Yudkowsky is another example. Bloggers who present statistics nearly in isolation from other kinds of knowledge and suggest those statistics are far more meaningful than they actually are are a less extreme but more common example.
December 13, 2010 at 6:47 am
I think you’re on a better nomenclature track with “fetishizing” than with “hyper”.
December 13, 2010 at 12:11 am
Funny, since my impression is that Hanson inclines more toward “farmer” norms (one litmus test would be how one resolves the Repugnant Conclusion).
I agree that Hanson has gotten more speculative and less rigorous. It seems that once it became a solo blog it shifted its focus from Overcoming Bias to just being Robin Hanson’s blog.
I believe I’ve written earlier hear about my problems with rationalism vs empiricism (epistemologically) and pluralism (more normatively). But I’ll still defend statistics. As some wag put it, statistics are applied & refined common-sense (the same could be said of Bayesian probability).
Since this is rather general, I suggest you highlight the OB post is the most egregious example of the tendencies you described.
December 13, 2010 at 6:51 am
I don’t like calling statistics a derivative of common sense, because I think it both isn’t and it’s a major corrective against it.
I think some past statistics academic wrote a book called something like “Statistics as rhetoric” and I think that’s a better lense to view statistics. Everything still arises somewhat magically from the first principle intuitions in our brains and organizations, but statistics is kind of a persuasive argument on how to view the world based on its claims of improved prediction -at least that’s my understanding of the author’s work, mostly from the title and my own grasping understanding of elementary statistics.
December 13, 2010 at 7:05 am
I can’t find the book I’m looking for discussing statistics as rhetoric (I found a lot of false hits with almost the exact same title) but the closest similar thing I found was “economics as rhetoric” by Diedre McCluskey, who seems to me to be well-credentialed.
December 13, 2010 at 9:59 am
To add to this, I think (initial mover) rhetoric pulls audience from their “common sense”/1st line intuition to a different position. I think statistics serves that role.
It also serves in the role of subsequent/multi-mover rhetoric, which includes defenses against other rhetorical agents.
December 14, 2010 at 10:56 pm
Economics is not statistics. I use statistics just to keep track of how well our software is performing over time and in different configurations. This serves for purposes of discovery, not rhetoric.
The statement about statistics being common sense (modulo blah blah blah) has spread around enough it is tough to find the origin. Someone here says it was Laplace.
McCloskey was last discussed at this blog here. Since then Robin Hanson has criticized Greg Clark’s theory.
December 15, 2010 at 3:33 am
“Economics is not statistics. I use statistics just to keep track of how well our software is performing over time and in different configurations. This serves for purposes of discovery, not rhetoric.”
Rethink this and look for your first principle assumptions.
That “statistics … keeps track of how well our software is performing over time and in different configurations” seems to me to be rhetorical, and the first principles seem to me to remain (currently unescapably magical) intuition.
I encourage you to take a look at the Diedre McCluskey cite, I think it’s germane although it’s a bit buried in her literary quirks.
December 16, 2010 at 11:04 pm
A quick google search didn’t turn up “economics as rhetoric”. Did you mean to write “the rhetorics of economics”?
I’m confused by your statement about statistics. My code is used by others in my company. I compare how it performs to other statistics that have been gathered that I am trying to roughly match. If the performance is significantly worse it suggests my implementation is sub-optimal and I can expect my co-workers to come to me with complaints. Such statistics helped reveal a serious flaw in my code that was causing lots of data to be lost. One common statistic that I temporarily use when testing is the amount of time taken in certain sections of the code, which helps to reveal which portions I should look into speeding up.
December 17, 2010 at 12:41 am
I think we’re hitting a wall with this topic. I’ll try to jump back into it later when I have more time.
December 28, 2010 at 11:34 pm
[…] towards its genre. Because of my predisposition toward that narrative, I look out for contrary ones. I’m currently also enjoying “The Art of Not Being Governed”. I saw on the […]
January 26, 2011 at 11:49 pm
[…] I’ve been reading James Scott, this brings to mind again the dark side of tacit knowledge. One of Scott’s books, “Weapons of the Weak“, I think is largely about […]
October 21, 2014 at 10:53 am
“Has anyone tried reconciling the clash of metis and techne?”
Techne is useful where we’ve got really strong delusional intuitions, like in medicine, schooling and alms. I think that’s the short version of Hanson’s view.
October 21, 2014 at 11:59 am
My impression of Hanson is that delusional intuitions go far beyond that. And this case of medicine isn’t merely the folk beliefs of lay people (which Hanson seems more interested in, though he thinks they’re shared by doctors) but an unsupported claim of expertise.