Re: A Critique of Social Science (part I)

Hank, sorry it's taken so long to respond on this. You've got a lot of interesting stuff there. I agree with the general assertion that there is some barrier of irreducible complexity beyond which the scientific method cannot yield verifiable results, and that as a general matter the social sciences fall on the wrong side of that barrier. I like your thought that deductive arguments should be given greater consideration in social sciences, and that it is possible and desirable to more frequently argue from first principles.

I was just talking with Mr. Veritas about this, and we formulated a number of criticisms. In some instances your theory seems underinclusive (i.e. much of this criticism could apply to elements of the hard sciences), in others, overinclusive (experimentation in social sciences can uncover something of truth). Implicit in that discussion is exactly what composes this barrier of complexity, and where it precisely lies. I think it's also an interesting question to what extent that barrier is malleable based on technological capability and keen insight. I am also not sure that we know that there are significant quantum effects that govern the operation of the brain which would place human behavior conclusively on the far side of the complexity barrier. V raises the question of whether Chaos Theory can shine any light on complex human interactions despite the inability to model a single human. Also, I'm not certain that the application of research in the social sciences is as separate from deductive reasoning as you suggest. I'm going to address this last point now, with the added hope of fleshing out your theory through some discussion of application. I will try to take up some of the other topics in the coming week or two as time allows, I know V is hoping to post some on this, and we're hoping to drag Ryan in as well, particularly on the fact questions relating to the brain and chaos theory.

First, I disagree with your characterization of Rabin's paper as purely a quixotic exercise attacking a long-defunct economic model. The rational, utility-maximizing actor is a staple of Western economic theory (although apparently not the Austrian theory). I would very much appreciate it if you could elucidate the basic deductive argument for Mises' irrational capitalism. I understand that as an advocate of the Austrian School, you reject Rabin's position for the reasons you stated, but as your audience is not necessarily familiar with (or in agreement with) the Austrian economic theory, we would be well served to hear where the foundation lies, if not on rational actors. The balance of the criticism of Rabin's positions seems largely reliant upon this.

Moreover it doesn't strike me that Homo Economicus is a concession to modeling. As far as I can tell, he rather resembles a deductive device. As an example here is a premise from James Mill's deductive argument for democracy (my paraphrase): Human beings apply accessible government power (rationally) to serve their own pleasure and avoid pain. The premise of Homo Economicus (again, my paraphrase): Human beings act (rationally) in markets to serve their own pleasure and avoid pain. These sound very similar to me. I know the former is an element of deduction, why not the latter? On what distinction do you separate Homo Economicus from a premise of deductive argument and name it a predictive modeling approximation? To the extent that Rabin is criticizing a deductive premise, is he not applying research in the role that you suggest for the social sciences?

I also don't see the inherent conflict between relevant data and calculable data. It is indisputable that deductive arguments call up many controversies of fact (see any legal decision). It is also true that some issues of fact are more critical in a given argument than others, and that some issues of fact yield more easily to research than others, and finally that the ones more easily researched are not necessarily going to be the ones that are most critical. However, I see no problem with researchers, faced with this dilemma, plucking the low-hanging fruit. If some premise of the argument is easily assessed through research, why not engage in that research even if it is not the most critical premise? Some premises will necessarily remain within the realm of rhetorical debate indefinitely (or nearly so), but those for which useful insight can be gained through research should be examined. There is the possibility that undue focus and disproportionate resources will be applied to those areas which do yield insight through research and experimentation, and I agree that this could be a problem. However, this is a problem of resource allocation, and I don't think taints the results of the research.

I'll hold off until hearing more of the Austrian School before really engaging this issue, but my feeling is that the rational maximizer is a fundamental underlying premise of nearly all Enlightenment social, political, and economic thought -- fundamental to the point of frequently going unstated. And thus data that suggests limits on the rational maximizing function carry wide-ranging ramifications. I would also suggest, that behavioral criticism aside, the rational maximizer is a generally defensible, useful, and important premise.

p.s. Mr. V highly recommends the book Consilience by Edward O. Wilson on this topic. Hopefully he'll say more on that. If you get a chance you might check it out and provide some feedback.