In The Poverty of Historicism (1957), the great philosopher of science Karl Popper claims, “I have shown that, for strictly logical reasons, it is impossible for us to predict the future course of human history.” By showing this, Popper thought he had fatally wounded Marxism and other varieties of what he calls “historicism.” As Popper uses the term, it refers to the view that scientific laws allow us to predict the future. Marxists believe that the laws of economics establish that capitalism will face crises that will grow in severity. Eventually, capitalism will be overthrown and socialism will replace it. If Marx is correct, then a scientific law allows us to predict the future course of human history. This is exactly what Popper says is impossible. Is he right?
Some people might object that this isn’t a correct way to look at Marxism. According to this objection, Marx does not make unconditional predictions about the future course of history. He instead points to tendencies that aren’t inevitable. I’m going to ignore this objection here. If it’s right, then Popper is wrong to class Marxism as a variety of historicism. But I’m interested in his argument against historicism, which doesn’t refer to Marxist doctrines. Rather, it claims to refute any view that purports to predict the future course of human history. Once more, my question is whether Popper is right about historicism, as he characterizes this term. (The term usually means something else, but this is irrelevant to our discussion.)
Popper’s argument against historicism is this: (1) the course of human history is influenced by the growth of human knowledge; (2) we cannot predict (by rational methods) the future growth of human knowledge; (3) we cannot, therefore, predict the future course of human history.
The key step in this argument is second premise. Why does Popper think that we can’t predict (by rational methods) the future growth of human knowledge? The chief philosophical interest of the argument is his answer to this. He says that if we could predict future human knowledge then we would know this knowledge now. But in that case, it wouldn’t be future knowledge. The claim that we can now know that future knowledge leads to a contradiction. By the way, when Popper talks about “knowledge,” he means, roughly, what people accept as part of science, mathematics, and other disciplines. He thinks that epistemologists who look for knowledge that withstands skeptical doubt are pursuing a hopeless project.
Some people respond to Popper’s argument by trying to find “wiggle room” that allows them to modify the second premise. “Maybe we can’t know the details of future knowledge,” they say, “but can’t we sometimes get a general idea of what lies ahead?” As an illustration of what they mean, there were people at the beginning of the twentieth century who predicted the coming of television, even though they didn’t understand how to make television sets.
I’m not going to go down this path, because Popper’s ingenious argument can be refuted more directly. People in the future will, we can safely assume, retain at least a great deal of what we know at present. They will continue to know the mathematical and scientific laws we know today. So can’t we predict at least in part people’s future knowledge?
The response to this is obvious. “How stupid can you get? Popper isn’t talking about knowledge we have now that is retained in the future. He means knowledge that we don’t have now but people in the future will have. It is this sort of knowledge that he says it’s impossible for us to have now.”
This is indeed what Popper means, but this is exactly where the problem with his argument comes in. He has simply defined future knowledge as “knowledge we don’t have now but that people will have in the future.” It follows from the definition of “future knowledge” that we don’t know any of it now, but that doesn’t rule out predictions about what people will accept as knowledge in the future. It is just that successful predictions won’t be future knowledge, by Popper’s definition.
The fallacy in Popper’s argument is simple, once you think about it. It has fooled a lot of people, though, so it might be helpful to give an analogy where the same fallacy is more apparent. Suppose a professor gives his class a multiple-choice exam. He says, “You won’t know the right answers to some of the questions, because it’s impossible to know the right answers to questions you don’t know the right answer to.” The professor is correct that you don’t know the right answer to questions you don’t know the right answer to, but his “argument” doesn’t establish that there are such questions. In like fashion, Popper’s argument doesn’t establish that there are things people know in the future that we don’t know now.
Once more, though, an obvious objection arises. “Isn’t it the case, when we look back at the past, that there has been an enormous amount of knowledge that no one anticipated? Isn’t it a good conjecture that this will continue to be the case in the future? Who knows what people living a hundred years from now will discover?”
I would not deny that for a moment. I would bet that the course of history will be influenced by what we don’t know now. But Popper claims to establish this by a logical argument, and this he has failed to do.
There is another problem with Popper’s argument, to the extent to which he wants to use it to refute Marxism and other forms of historicism. Suppose, I think contrary to fact, that Popper’s argument is correct and also that Marxists who say that scientific laws predict the demise of capitalism are also right. These two assertions can be true at the same time. Popper’s argument, if correct, shows that we can’t know the future course of history, if by that we mean all the important developments in the future course of history. It surely doesn’t rule out knowing some things about the future, though, and maybe one of the things we know is that capitalism is doomed. There is an obvious Popperian rejoinder to this, but I will leave it to readers to figure out what it is and also why it fails.