« BSG is Detached Detail | Main | College Prestige Lies »

March 21, 2009

Comments

I hate to harp on the same theme on and on, but didn't you just tell me yesterday that you considered a microbiologist summoning the full knowledge of microbiology to determine that an antibiotic would work against a specific bacterium zero evidence (or no more evidence than the equivalent statement by a homeopathist) because it doesn't count unless there's a randomized clinical trial? Even though microbiologists have frequently proven to be right in the past?

To me, it seems like the same conflict between demanding high-status sophisticated tools that let you signal rigor and sophistication, and accepting messy hypotheses that are more likely to be true.

Yvian, in both cases I am persuaded enough by aggregate evidence to bring my specific beliefs in line with it. We have strong reasons to suspect signaling explains a lot of human behavior, so I expect to find signaling in most specific situations. Our best aggregate evidence is that medicine has near zero marginal health value, so I must conclude there are strong biases in docs usually glowing analyzes of individual treatments.

Robin,

reading the blog over the last two months or so leads me to believe that you will go with the "signaling (affiliation with) high status" explanation almost every time it is possible, even if there are other (more plausible) hypotheses on the table. In this case it seems to me that the more obvious explanation is that Arnold (correctly, in my view) prefers falsifiable hypotheses because unfalsifiable ones are scientifically useless (and he thinks that the counter-signaling concept makes signaling explanations unfalsifiable).

(In case you don't know it, you'll probably enjoy Erving Goffman's Stigma, which is largely about signaling nonassociation with low-status groups.)

This seems a clear choice between the status and truth; you can either study the clean hypotheses that let you show your rigor, mastery of sophisticated tools, and affiliate with high status "scientists", or you can study the messier hypotheses that you think are more likely to be true. I choose truth.

I fail to see this choice as quite so clear. Surely there is a cutoff where the expected gains from studying less-likely-but-more-tractable hypotheses outweigh those of their counterparts. Studying a particular hypothesis does not commit your beliefs to it.

I don't see the connection with status. If someone looks for their keys under the lamplight because that's where the light is, we don't assume they're seeking status.

So he'd prefer to focus on something that might be so, but not as likely, and that can be shown to be wrong so he can move on, rather than on something that's more likely but non-falsifiable? He can come back to the signalling theory after he's proven the other stuff wrong. Take the low hanging fruit?

Ditto Phil.

I have previously asked for empirical evidence in favor of the pervasiveness of signaling on this blog without receiving any. Could you give me an idea of what "broader evidence" you are referring to in this post?

I think your discussion of medical effectiveness has shown the sort of disparity that can exist between theory and reality.

Hi Robin,

Obviously what I'm trying to do in Be the Solution is to help do-gooders make the cognitive leap so that they understand that free market economists are not really bad people who need to be punished for not punishing free riders. Once they realize that there are actually strong altruistic grounds for advocating much of free market economics, perhaps they will praise economists instead of punishing them.

Historically, of course, there are two interesting precedents of such cognitive leaps achieving some popular status:

1. The 19th century classical liberal activist movement, with people such as Cobden and Bright leading political campaigns and Harriet Martineau writing popular books; for a brief period free market thought held the moral high ground, with popularizers celebrating those who advocated for free trade, etc.

2. The "freedom of speech" movement, especially from J.S. Mill onwards, and into the height of the ACLU principled commitments to free speech in the 50s, 60s, and 70s where, at least among the educated classes, free speech very much held the moral high ground.

In the case of free speech, a case may be made that the instinctive propensity was to punish both those who made socially unacceptable statements (be they statements that violated the sexual mores of the time or racist statements that violate our own sensibilities today) as well as a propensity to punish those who fail to condemn those who make harmful statements. (Google "fail to condemn," "fails to condemn," "failure to condemn," "failing to condemn," etc. for a sampling of the issues on which people exhibit these behaviors.)

But because classical liberals had the moral high ground in the late 18th and early 19th century, and because freedom of speech liberals had the moral high ground among the educated classes for much of a century (and outside of some university contexts they still do, but see FIRE, for problems on campus), there is reason to believe that it is at least possible for a cognitive elite to signal to others that the arguments on behalf of freedom are deserving of moral praise despite the fact that free systems may allow individual wrong-doers to go unpunished.

One of the reasons I'm frustrated with academia is that they ought to be the cognitive elite that does this signaling, much as they have done with freedom of speech. But, outside of econ departments, they mostly continue to punish free market economists rather than praise them. But there are some issues on which the case for economic freedom is overwhelming: For instance, Hernando de Soto, the World Bank's Doing Business Index, and the Fraser Economic Freedom Rankings, all show that rich nations are economically free and poor nations are not. And at an intuitive level, the obstacles to creating legal businesses in the developing world are simply ridiculous. Free market economists are pretty much the only people advocating for greater economic freedom for the world's poor. And I know of no one who defends the massive obstacles to legal business creation that characterize most developing world nations. But the extraordinary harm due to these obstacles were only visible to those economists who were thinking about the benefits of economic freedom for releasing entrepreneurial wealth creation. If poverty is bad, and if free market economists have been leaders in identifying the most effective means of eliminating poverty, then free market economists, insofar as they promote policies that eliminate global poverty, are good. See my "Milton Friedman, A Modern Galileo," for one articulation of this line of thought.

More details to be added here, but this is certainly part of the path I'm taking so that do gooders can make the cognitive leap needed to understand why economists should not be punished. In the book I also provide the history of property rights solutions to tragedy of the commons problems, and the fact that free market economists have been advocating such solutions for decades, whereas only recently have mainstream environmentalists begun to recognize how powerful and versatile a solution they are. Again the message is that free market economists are not the enemies of the good that you do gooders thought they were.

All, yes on the margin one can prefer to study simpler hypotheses where one can more easily make progress, but that effect seems too weak here to justify ignoring signaling. If you are pretty sure your keys aren't at the lamppost, since you've already looked there several times, and there is enough moonlight to let you look down the street where you think you left them, you really should go down the street.

Michael, getting academics to agree with economists about what acts are cooperative seems much harder than getting them to see that we are paying costs to promote the acts we agree with them are non-cooperative.

@Michael Strong

How nice to see you here at OB, Michael. It's so great to see a fellow Johnnie make good. I have been so impressed with your career over the years - all of our fellow Johnnies are so proud of what you've been able to do.

And let me take this time to once again thank you for letting live in on your couch for 3 weeks all those years ago. I do wish you all the very best. "Hey, hey, arete, ego kale hedone!"

Robin:

Shouldn't you be ashamed of yourself?

You criticize Arnold because he says you do not have enough evidence for your claims about signaling, and because he says that often signaling claims don't seem to be falsifiable. But you offer no evidence for your signaling claims, or indication of how they might be falsifiable.

Are you aware that your claims would be more convincing if you provided some evidence that your claims about signaling are true, and some evidence to show that signaling claims are falsifiable?

Instead, you attack Arnold, and make fun of him for 'tripping over his positivism, 'head in the sand.' Then you accuse him of holding his beliefs because he is seeking status, and say that he is choosing status over truth.

You have often made ad hominem arguements before in your blog and in person, belittling and insulting your opponents, and you must be aware that such arguments have no logical force.

Shouldn't you be ashamed of yourself?


Just read the forward of "The Bell Curve Wars" by Steven Fraser, (Gould's 1995 "Curveball" article.) In refuting Charles Murray's stastical argument for a biological basis for IQ differences across (racial/ethnic) groups, Gould sites a study of two cohorts of children of Black GI's born to German women after WWII and reared in Germany, who had no discernable differences in IQ from other German children. Well, how sexist is that? Who says IQ is transmitted by males? To make Gould's argument, IQ would have to be a function of Y chromosomes.

@Thug Shrink

You are presenting a false dilemma. I do not believe you have exausted all possible alternative theories.

I'm inclined to agree with Bruce, only I haven't seen Robin make personal attacks like this one before. To the contrary, he's impressed me with his level of detached-ness. (And it's technically not the ad hominem fallacy, because the subtext is that Arnold is bad because of the things he says, not the other way around.)

I'm quite willing to consider the importance of signaling, but I think Robin should be open about the lack of empirical evidence or falsifiability if it so happens that there is no empirical evidence / no falsifiability.

The comments to this entry are closed.

Less Wrong (sister site)

May 2009

Sun Mon Tue Wed Thu Fri Sat
          1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30
31