Oh My God! People Don't Behave Rationally!

I just stumbled across an article from May about the rationality of getting painful experience over with and, conversely, postponing pleasurable experiences. Not surprisingly, humans as a rule fail to behave according to economic models of rationality:

The research, being published today in the journal Science, is "terrific, " said a leading expert on brain imaging, Dr. Read Montague, a professor of neuroscience at Baylor College of Medicine who was not involved in the study. It demonstrates that the brain "assigns a cost to waiting for something bad, so that the bad thing is worse when it's delayed farther in the future," Dr. Montague said.

"Hence," he said, "the 'let's get it over with' bit when we're at the doctor's office waiting to get a shot."

The research also sheds light on economic behavior, said George Loewenstein, a behavioral economist at Carnegie Mellon University. According to standard economic models of human behavior, choosing more pain in the short run is irrational, Dr. Loewenstein said: if you know something bad is going to happen, you should postpone it as long as possible, and if something good is going to happen, you should want it right away.

In real life, people often do the exact opposite, he said. They delay gratification to savor a sweet sense of anticipation, and accelerate punishment just to get it over with. The new study sheds light, he said, on how the act of waiting can be used to describe economic behavior more accurately.

I've been having these arguments recently about the idealization (indeed, elevation to a transcendental level) of reason and rational behavior. If we're going to learn to deal with humans, we need to learn how to think about irrational behavior. I had thought that logical positivism was as dead as the idea of homo oeconomicus, but both are apparently alive and well. Anecdotally speaking, it sure looks to me like -- you know, based on the empirical evidence -- many scientists and philosophers of science have an easier time recognizing the limits of empirical evidence -- and inductive inference therefrom -- than do their popularizing propagandists.

My argument has been that adopting the inductive principle is, strictly speaking, neither scientific (where that means empirically grounded) nor logically sound (where that means consistent with widely recognized logical principles, like the invalidity of circular arguments). It simply cannot be empirically verified that the inductive principle is the case; the principle itself can only be inductively inferred from a body of evidence. But the use of induction to confirm induction begs the question -- it's a textbook example of assuming the very thing you are trying to prove. But adoption of the inductive principle is of course perfectly reasonable/rational (in the broader sense of "reasonable" or "rational"), as long as we recognize that it is not and cannot be absolute.

Two consequences follow from this observation. First, science understood as an empirical system is substantially grounded in a principle unsusceptible of being proven scientifically. In that respect, then, the inductive principle is simply a specific case of the general rule, noted aleady by Aristotle and proven mathematically by Gödel (the incompleteness theorems), that formal systems cannot be both strictly complete and strictly consistent. So we are obligated by a commitment to empiricism to recognize empiricism's limitations, as we are obligated by a commitment to logic to recognize that a radical empiricism violates all sorts of logical principles. Even the logical positivists eventually pulled the plug on their own ideology when they recognized that it was precisely illogical; indeed, even worse from their perspective, it was ideological, i.e., metaphysical!

Second, reasonableness (or rationality) is not reducible to rationality understood as empirical science -- put another way, rationality is not reducible either to empiricism or to obedience to logical laws. Empiricism has a certain reasonableness to it, but to espouse a radical empiricism (that is, reductionist empiricism along the lines of the logical positivists) is inconsistent with both logic and science. But if a certain kind of common-sense empiricism is reasonable, as I suppose it to be, then what is reasonable about it is precisely a reasonable, commmon-sense form of it (more is not always better, and more empiricism is not necessarily more correct). Radical positions require radical rigor in their conformity to principles, and it is in that context that radical-reductive empiricism's violation of its own core principle is a problem. If we hold to a more reasonable standard of reasonability, however (along the lines, say, of the phrase "reasonable doubt"), what we can ultimately say is that it is perfectly reasonable to espouse a reasonable (ie, not absolute) empiricism; or, to come at it from the other direction, radical-reductive empiricism is precisely irrational because it is unreasonable, in common sense terms; irrational, in logical terms; and unscientific, in empirical terms.

It is also worth noting that science as an empirical endeavor is in no way undermined by this admission. Only if science vs. religion is an all-or-nothing, winner-take-all struggle between two absolute worldviews is such a thing even conceivable, and even then it would not follow. But let's acknowledge that while science and religion overlap in some important areas, that overlap is not definitive of either religion or science, so that one is not reducible to the other. Even where they do overlap they need not always conflict, even if specific religions will almost certainly conflict with specific scientific positions. The fact that many "scientific" and religious people think the war is precisely all or nothing does not make it so in principle, and only makes it so in practice to the extent that everyone else plays along.

In general, the religion textbook comparison of "science" and/or "humanism" to what most of us would call religion strikes me as specious, at best, and an insidious form of obfuscation, at worst. Is humanism a worldview? Sure, but that doesn't make it a religion except in a view of religion that is drained of any real meaning. That said, a lot of intelligent people seem to take science precisely religiously (just like many intelligent people believe in Yahweh, Jesus, and Brahman). The inductive principle is for such people the transcendent guarantor of truth, itself beyond disproof (because it proves itself), immune to contradiction or question, a transcendent logos that organizes the universe while remaining aloof from the very rules it imposes upon that universe. Does the Inductive Principle exist somewhere "out there"? No, of course not. But its more, um, committed (not to say "fundamentalist") adherents will brook no question as to its authority, will hardly contemplate any limit to its infinite perfection, completeness, finality, unity, etc. (add other properties of God here), and spend more time thinking about how wrong any acknowledgement of the contradictions of radical-reductive empiricism must obviously be.

I am fairly well convinced that for many such persons, the concern at bottom is to protect science from anti-empiricism, and particularly from the encroachments of explanation-oriented religious thought into scientific territory. Surely everyone who knows me knows I share this concern. But I don't think the way to fight fire is with fire -- either when it comes to fighting terrorism or to fighting imperialist religious thought. Let science be honest science instead of turning it into a religion in order to fight religion(s). To turn it into a religion, to commit to the clash of civilizations version of a war between radical-reductive empiricism and repressive-fundamentalist religion, is to play into the hands of religious absolutists, just like becoming terrorists to fight terrorism is to compromise the very values we claim to be fighting for. Perhaps there would be a lot less resistance to evolution if people didn't think just that it contradicts a literalist reading of the Bible, but that admitting it to be true would mean that they'd have to be atheists like Richard Dawkins. But it doesn't mean that, any more than the shift from a geocentric to a heliocentric model of the world meant that.

For the record, I am committed to empiricism as a scientific principle and as a principle of common sense decision-making. I consider myself an atheist, not terribly far in this respect from Richard Dawkins, who rates himself a "6" on a 7-point scale of disbelief in religion. I would also rate myself a 6, although my logic would be a little bit different from Dawkins's. I am grateful, however, that physics has gone beyond naive realism (so that we know that the chair I'm sitting on is actually moving at the sub-atomic level, despite all appearances to the contrary, and that it is mostly empty space, despite the fact that I can sit on it). I am glad that no one told Einstein that he was being impractical when he made hay about the fact that the universe behaves differently at different speeds, especially at speeds human beings never actually attain. Conversely, I am convinced that we "know" things in other ways than a pure or simple empiricism (however theoretical that empiricism might be), that it is possible and even necessary for us to explore those other ways of knowing and the "objects" of those other ways of knowing. I do not think this position amounts to a validation of, or apologia for, traditional religions of any stripe. It's a recognition that we need to think about what it means to construct the values that interact with our empirically known realities. It's a question about how we "know" that, for example, rape is wrong. Like knowing that my hands are typing right now, I cannot possibly be mistaken about it. But how do I know it and what exactly is it that I think I know? I can't literally look at anything for empirical verification, even in principle, that makes it plain that I'm right. So what is it that I know when I know that and how do I know it? And what does it mean if I say slavery is wrong, when hundreds or thousands of years ago, it was so plain to everyone that it was right? Did the empirical evidence change? Did we get new "value telescopes" that allowed us to see empirical evidence we couldn't see before? Of course not.

So in the end, I say: Empiricists, be true to yourselves and your principles, and you cannot be imperialist empiricists. At this rate, however, I fear I will soon hear some of you claiming that "extremism in the defense of empiricism is no vice."