I’m a big fan of Karl Popper myself, right down to Popper’s distinctive, controversial reading of Plato. Briefly, Popper argues in The Open Society and Its Enemies that Socrates was in many ways a forerunner of classical liberalism, and that Plato twisted his master’s teachings to favor pro‐Spartan authoritarianism. This solves what would otherwise be a major problem in philosophy — that Plato often seems to talk out both sides of his mouth on questions of individual freedom and dignity.
In Popper’s view, the texts that seem to favor the equality and universal dignity of mankind, like the Meno, are representative of what Socrates really thought, while the texts that favor natural inequality, like the Republic , are the results of Plato’s tinkering. It’s possible that the truth is a lot messier than this, of course, and that Plato and Socrates weren’t such polar opposites, but I do find it an appealing thesis.
And what of Popper himself? He was no authoritarian, to be sure. But is Popper, or his politics, liberal or conservative? (Yes, yes, I know. There are other choices. But some people can only count to two, and it may be interesting to follow them for a while.)
One could certainly make the case for placing Popper with the conservatives. Critical rationalism sounds a lot like Michael Oakeshott’s conservatism (though, admittedly a lot less like some others). Here’s Lester:
Put simply and starkly, critical rationalism is the view that absolutely all alleged knowledge is ultimately only fallible theory: mere guesses that we can test but which never become more probable by passing those tests. No truth is ever established to any degree at all…
But Popper noticed a crucial asymmetry: the falsification of universal theories is logically possible; we need just one counter instance. ‘All swans are white’ cannot be verified by any finite number of positive instances of white swans. It can be falsified by one instance of a non‐white swan; as this ‘well‐supported’ theory eventually was falsified by the discovery of black swans in Australia. So, methodologically, we can make a virtue of producing bold universal conjectures that we do not pretend are ultimately supported by evidence and then test these conjectures as severely as we can: both by observation and criticism. We can happily admit that, in principle, we might be mistaken about any theory regardless of the amount of testing. This became known as ‘falsificationism’.
Which I think is a fair summary of Popper’s view of knowledge. Compare this to Oakeshott, who in this passage is criticizing the Cartesian tradition in philosophy:
[T]he lesson his successors believed themselves to have learned from Descartes was the sovereignty of technique… a beneficient and infallible technique replaced a beneficient and infallible God; and where Providence was not available to correct the mistakes of men it was all the more necessary to prevent such mistakes… (Michael Oakeshott, “Rationalism in Politics”)
Popper recommends to us a technique, yes, but its results are never infallible. Empirical science falsifies, but it does not and can not prove. Even the most apparently certain empirically grounded theories might always be knocked down by a startling new piece of evidence. (Sometimes the evidence would have to be mind‐bogglingly startling, but that’s okay. Minds can and do boggle.)
This modesty is not a defect of science. On the contrary, it’s science’s greatest strength. Overbearing claims of certainty are for geometers or theologians, but here on earth, I believe that we can do no better than falsification. And, whenever geometers or theologians make falsifiable claims, they too have to contend with the possibility that their claims will, in fact, be falsified — regardless of how little used to it they may be on their home turf, and regardless of how annoying they may find it.
Popper’s approach to knowledge counsels caution. It doesn’t say “don’t even bother to build your knowledge” — which, in politics, wouldn’t be liberal or conservative, but rather totalitarian. What it does say is “build slowly.” Try things. Err. And when you err, try other things. Question claims of universal truth. Learn to be suspicious of them. And this sounds a lot like Oakeshott.
Now, as they say, for the other side of the story.
A curious thing often happens when governments begin to play by these rules, and when they respect the human limitations to knowledge: Human knowledge increases. Really, really fast. In the private sector, new inventions, new discoveries, and new cultural forms proliferate — possibly, I’d say, to the dismay of someone with Oakeshott’s conservative disposition toward life in general. A restrained government typically doesn’t result in a restrained society. Quite the opposite, in fact.
That’s why, as a libertarian, I think the best approach to politics is one that happily combines a free market, a limited government, and an open, tolerant, liberal culture. As Virginia Postrel wrote in The Future and Its Enemies — a book I never get tired of praising –
[T]rial-and-error evolution is very different from the concept of progress popular earlier in this century… We make progress not toward a particular, certain, and uniform destination but toward many different, personally determined, and incremental goals. In a global sense, “progress” is the product of those parallel individual searches: the extension of knowledge and the gradual improvement of people’s lives — an increase in comfort, in life options, in the opportunity for “diversified,” worthwhile experience.
That’s a very liberal‐sounding world, in a sense. But to get there, we need to respect the limits to knowledge. Nature, to be commanded, must be obeyed.
Could falsification itself be falsified? I have to disagree with Lester here, in that I do not think the question is properly formed. Falsification does not rest on any empirical justifications. It is a method for producing claims, not a claim in itself. We should more properly ask whether the method of falsification is invalid — that is, whether falsification is ever capable of starting with true premises and yielding untrue conclusions. Falsification, however, is ultimately just the use of modus tollens on empirical data and the universal claims that we make about it. Modus tollens can’t be denied without contradiction, which means that falsification is a valid method as long as universal truth claims continue to imply particular results — in other words, we’re stuck with falsification whenever P implies Q.