Kolakowski’s Law Located
In my previous post, I lamented not being able to find the source of The Law of the Infinite Cornucopia, also known as Kolakowski’s Law. Internet searches turned up a bunch of similarly worded references on blogs, all with no citations to original works.
A reader has pointed out that, using Google Books, a reference to the law can be found in a few philosophy books in which the original work by Kolakowski is cited. Google Books is not yet a part of my search habits, and in the past Google Books results have turned up in regular Internet searches, so I didn’t think to use it. My mistake.
It still strikes me as odd, though, that such an interesting argument from such a famous philosopher, with implications for political theory as well as the philosophy of religion and other fields, has not been cited to its original source far more often than it has.
The book in which it appeared is a 1982 work entitled Religion. Sadly, the section of this book containing an exegesis of the infinite cornucopia does not appear on Google Books. One of the secondary texts that mentions the law, however, cites the following quote from Kolakowski:
“The law of the infinite cornucopia…applies not only to philosophy but to all general theories in the human and social sciences: it states that there is never a shortage of arguments to support any doctrine you want to believe in for whatever reasons. These arguments, however, are not entirely barren. They have helped in elucidating the stats questiones and in explaining why these questions matter.”
Interesting: the law originated in a discussion by Kolakowski not of Marxism but of religion. According to Google Books, the law is also cited by the perennial sophist Cornel West and by the anti-Cold Warrior novelist John le Carre.
My next task is to get a copy of that book by Kolakowski and further explore the law’s implications.
Thank you, reader.
The Law reminded me of something. Then it came to me: The Logic of Scientific Discovery.
You can always find evidence to support any hypothesis that you happen to find attractive. IF you really want to know the truth, then you must ask yourself, what evidence would convince me that my hypothesis is wrong? Then you must look for such evidence.
But that’s only if you really want to know the truth.
I think you’re on to something, although I believe Popper actually stated it in terms of experiments: you can perform infinite experiments to verify an observation, but you need experiments intended to falsify to test it.
Definitely. Popper and Kolakowski, at least in this sense, go hand in hand. Not just with falsification, too. In The Open Society and Its Enemies, Popper blends his views of the philosophy of science with his political theory.
I’m interested. I don’t care that much one way or another, but what sort of experiment would falsify macro-evolution? (And not just create another mutation of the theory?)
I think it’s not just a law but a theorem. Let L be a logical system defined by axioms, and p be the proposition you want to prove. Extend L making your axiom system (L\union p). I think you can then prove by induction on the length of an argument that there are countably many arguments in the system that leap to concluding \entails p.
I wonder what Wittgenstein’s take on this would be, as he was both a brilliant mathematician and brilliant philosopher of language. Kolakowski’s law seems to concern both.
Let’s see if I understand: you are saying that there are countably infinite ways of “proving” something by petitio principii?
Yup. Isn’t that what constructing arguments to support a hypothesis (as opposed to falsify) would be?
Thank you. Then you were correct: the relationship to Popper is somewhat remote.
In one case we are dealing with an infinite number of “proofs” of the hypothesis, in the other with an infinite number of experimental results “supporting” the hypothesis.
Also, in one case the hypothesis must be compatible with only a finite set of axioms, in the other with all the potentially infinite experimental results.
Are you familiar with WorldCat? Many libraries provide a portal to WorldCat, enabling you to place an interlibrary loan request yourself, rather than going through a librarian.
Robert, I see from your bio that you are writing about the history of molecular biology. The law of the infinite cornucopia applies to molecular biology too. Most studies suffer from the moderator vs mediator issue. That is, many factors (eg, genes) affect a phenotype and therefore can be reported as ‘significant.’ But the challenge is to identify the set of factors (mediators) that are modulated by all the other factors (moderators). The problem is partly due to the misuse of statistics. The effect is that many of the results reported in molecular biology studies are unreliable.
Thanks for bringing this up. You’re absolutely right. Though I’m a mere historian and not a scientist, statistician, mathematician, logician, or philosopher of science, in my role I have tried to learn as much about these fields as possible. My experience has been that molecular biologists, as brilliant as they often are, are woefully deficient in the philosophical aspects of their work. I don’t blame them: the amount of information generated on a daily basis by even a few labs is too much to digest, and bench scientists who spend hours with microassays don’t think to stay up to date on the philosophy of science. There was a great story about this a few months back in I forget where, but it focused on how much peer-reviewed research in molecular biology has turned out to be non-reproducible. Nevertheless, amazing progress is made due, at least in part, to the enormous hyper-specialization in the field.
In my arrogant opinion, it is pointless, even counter productive, to be up to date on the philosophy of science: Aristotle, Hume, and Popper are all what a practicing scientist really needs to know.
Having said that, recent developments on the rationale of Occam’s Razor seem to be very interesting, not to say potentially revolutionary. But they come from Comp.Sci. departments, not Philosophy departments. I am thinking of Markus Hutter’s work in particular.
> You can always find evidence to support any hypothesis that you happen to find attractive.
I guess that all depends on how you define evidence. I can think of a few current hypotheses that don’t seem to be able to find any evidence. Many Worlds, for instance.