Fallibilism vs. Cynicism

Critical rationalism is a fallibilist epistemology. Man can err, and by an effort, he can learn from his mistakes. Ultimately, there’s no “criterion to distinguish with certainty between true statements and false statements. And this is the deepest reason for the fallibility of man”, argues Karl Popper, founder of critical rationalism. But he also says “very many of the statements which we hold for truth are true.”

In short, you may state the truth, but there’s no guarantee that it’s the truth. Tomorrow, you may find that you were wrong after all. That’s always possible. Though it may seem counterintuitive at first, that’s actually good news: it means there’s always room for improvement somewhere.

Note the words I am using. Man can err. He may state the truth. He may find that he was wrong. Over the years, it slowly dawned on me that physicist David Deutsch, arguably the most famous proponent of critical rationalism of our time, and his fans, the ‘crit rats’, are not really fallibilists. They’re what I’ve come to call cynics. They don’t think our ideas can be wrong – they think our ideas are necessarily wrong, no matter how hard we try to correct our errors.

I should clarify from the outset that I considered myself a crit rat once; I used to think virtually anything Deutsch said was correct. But over the years, I’ve developed some disagreements. These are friendly criticisms; I have known many crit rats for years, have met them in person, and consider them friends. As critical rationalism itself says, we should expect errors even in our best theories, so disagreements are bound to come up eventually. That said, I don’t claim to represent the One True Version of critical rationalism. Ultimately, my goal is to advance our understanding of rationality and to live a happier, more rational life.

In an interview with the Institute of Art and Ideas, Deutsch says “[a]ll human knowledge is false.” (!) He’s also written that “we expect, with Popper, that all our best theories of fundamental physics are going to be superseded eventually …” (emphasis mine). Brett Hall, arguably the most influential advocate of Deutsch’s ideas, once wrote to me that “we cannot ‘speak the truth’.” Other fans of Deutsch have picked up this idea as well. For example, at the time of writing, this Twitter user’s bio just says “Wrong.”, as if he never expects to say anything true. In The Beginning of Infinity (BoI) chapter 18, Deutsch writes:

[T]he desirable future is one where we progress from misconception to ever better (less mistaken) misconception. [T]he nature of science would be better understood if we called theories ‘misconceptions’ from the outset, instead of only after we have discovered their successors. … Einstein’s Misconception of Gravity was an improvement on Newton’s Misconception, which was an improvement on Kepler’s. The neo-Darwinian Misconception of Evolution is an improvement on Darwin’s Misconception, and his on Lamarck’s. If people thought of it like that, perhaps no one would need to be reminded that science claims neither infallibility nor finality.

There’s that same mistake again: calling all theories misconceptions as if we could know for sure that they contain mistakes. That is itself a claim to infallibility and finality. (Thanks to Zelalem Mekonnen for pointing out this example.)

The idea that we are always wrong is not Popper’s view; it’s not critical rationalism; it’s not fallibilism. To be sure, the names of the people involved are not important – I’d be saying the same thing if Popper got it wrong and Deutsch got it right. I mention names to give credit where credit is due, and to emphasize that crit rats do not represent Popper’s view, even though they seem to think they do. They call themselves critical rationalists, after all.

Claiming to have a guarantee that some knowledge is true, and claiming to have a guarantee that it’s false, are two sides of the same coin. Fallibilism says there are no guarantees either way. We may err, but we may state the truth, if only by accident. Popper quotes Xenophanes (brackets mine):

And even if by chance [man] were to utter
The perfect truth, he would himself not know it;
For all is but a woven web of guesses.

Both Popper and Xenophanes knew that it’s possible to speak “[t]he perfect truth” – we just can’t know that it’s the truth either way. Again, there are no guarantees. But consider what it would mean if all human knowledge were false. It would mean that, no matter how hard we tried, we could never, ever find any truth – which is why I call this stance cynicism. We would always be wrong. We would always be overwhelmed with error to the point of utter paralysis. Consider what it would mean in terms of honesty alone. We’d always be forced to lie! Liars would have the perfect excuse: ‘What did you expect? It’s impossible for me to state the truth anyway!’

The idea that all ideas are false rules itself out. But to my surprise, this simple argument does not convince cynics – after all, they think truth is unattainable anyway. If all ideas are false, then falsehood cannot be a reason to reject an idea. There’d be no ideas left to adopt! But if all ideas are false, how can we form a rational preference for one idea over another? Crit rat Liberty Fitz-Claridge recently suggested to me that all we can hope to achieve is a bit more truth, a little less falsehood – be “less mistaken”, as Deutsch calls it. But aren’t the true parts of an idea wholly true? ‘Less’ falsehood by what measure, by what standard? How much ‘less false’ or ‘more true’ does an idea get exactly when we improve it? Isn’t truth an absolute, leaving no room for deviation? Isn’t a ‘little false’ still false? Doesn’t the notion of ‘less false’ accommodate error rather than correct it? There’s not enough rigor here; it’s vibes.

Lucas Smalldon, a crit rat with whom I’ve had several fruitful and stimulating discussions about epistemology, once told me that, “although problems are soluble, truth is unattainable.” But if truth is unattainable, how are problems soluble? I think he got this idea from the cynics – although he disagreed that he was a cynic, I believe cynicism is what ultimately led him to reject Alfred Tarski’s concept of correspondence as the aim of science. According to Tarski, truth means correspondence to the facts. It’s a commonsense, realist stance. But remember, cynics don’t believe truth can be achieved, meaning correspondence can’t be achieved. Smalldon concluded that instead, “the concepts of truth and falsity play no role in epistemology.” (!) But correspondence is our window on reality – so its rejection is rank cynicism.

I believe that, in reality, we can not only learn from our mistakes, but we often do state the truth – meaning 100% the truth, with zero errors or pending criticisms. Again, we cannot know for sure when or whether we have spoken the truth, and we should remain open to the possibility that we haven’t. We should look for error in any of our theories. But we can correct an error, be done with it, take steps to prevent its repetition, and move on to other errors. That’s progress.

Dirk Meulenbelt recently suggested a new argument against cynicism, invoking the multiverse and the laws of physics. If it were true that we could never utter true statements, even by chance, then there wouldn’t be a single universe where anyone does so. It would mean there’s some mechanism baked into the very laws of physics preventing people from making true statements. But it would be very strange indeed if the laws of physics had such a mechanism. I’m not a physicist, but as far as I know, none of our physical explanations refer to such a mechanism. Therefore, Meulenbelt concludes, cynicism is false. I agree.

Some cynics argue that there’s an inherent ambiguity in natural language, and that everything we say must therefore be erroneous. I don’t know if I agree that natural language is always ambiguous, but even if so, I don’t see how that implies error. We can make ambiguous but true statements. ‘I’m currently located in a hemisphere’ is ambiguous as to which hemisphere, but it’s still true. We could be silly and ask, on which planet? This one. Earth. We all know what we’re talking about.

In practice, cynicism has not done much damage – yet. Deutsch’s epistemology is still new and not as widespread as I expect it to be in the future. But, contrary to its stated aims, the cynical part of his epistemology undermines its content as a whole. Taken to its logical conclusion, cynicism is not compatible with realism or progress. To be sure, none of today’s crit rats consciously advocate anti-realism or stasis. They have good intentions; they want progress.

In this sense, cynicism is in a similar historical position as empiricism used to be. As Deutsch writes in BoI chapter 12:

[E]mpiricism initially played a positive role in the history of ideas by providing a defence against traditional authorities and dogma, and by attributing a central role – albeit the wrong one – to experiment in science. At first, the fact that empiricism is an impossible account of how science works did almost no harm, because no one took it literally. Whatever scientists may have said about where their discoveries came from, they eagerly addressed interesting problems, conjectured good explanations, tested them, and only lastly claimed to have induced the explanations from experiment. The bottom line was that they succeeded: they made progress. Nothing prevented that harmless (self-) deception, and nothing was inferred from it.

However, the role of empiricism turned bad as people took it literally:

Gradually, though, empiricism did begin to be taken literally, and so began to have increasingly harmful effects. For instance, the doctrine of positivism, developed during the nineteenth century, tried to eliminate from scientific theories everything that had not been ‘derived from observation’. Now, since nothing is ever derived from observation, what the positivists tried to eliminate depended entirely on their own whims and intuitions.

Deutsch goes on to say that, as a positivist, Ernst Mach rejected realism. Luckily, Einstein didn’t. But relativity might have suffered the same fate as quantum physics if he had.

So you can see how dangerous empiricism is when taken literally. Now, if you had pressed them, the original empiricists, the harmless kind, might have rejected anti-realism, and they would have struggled to explain how exactly we derive theories from the senses. But that did not stop later, literal empiricists from rejecting realism and advocating sense derivation of theories anyway.

Though we cannot predict the growth of knowledge – this is another core tenet of fallibilism – we can speculate that cynicism might develop similarly. Though a perverted view of fallibilism, it still contains a grain of truth in that utopia is impossible and error is commonplace. Cynicism may not do much damage as long as people don’t take it literally. Currently, few people do, and if you pressed them on it, they might even explicitly contradict it. But once people do take it literally, at scale, then they will reject progress and realism, and they will ‘achieve’ unhappiness and destruction as a result. We are starting to see the effects of cynicism already in, again, some people’s rejection of correspondence, and in the idea that all we can hope to achieve is ‘less falsehood’ – without any rigorous measure of just how much less. Just whims and intuitions.

Meulenbelt and I tried to trace this error historically. When exactly did people part with fallibilism and turn it into cynicism? Popper once wrote, “in our infinite ignorance we are all equal.” Deutsch quotes that sentence in BoI chapter 18. It’s possible that he misinterpreted “infinite ignorance” to mean “[a]ll human knowledge is false.” In that interview by the Institute of Art and Ideas, which I mentioned earlier, Deutsch’s full quote goes (starting at 2:50), “All human knowledge is false. As Popper said, we’re all alike in our infinite ignorance.” But I think Popper was simply saying that there are infinitely many things we don’t yet know, not that the things we do know are all false.