The Dadaism of "All Theories Are False"
dadaism: a particular absurd, nonsensical, deliberately anti-rational phrase, gesture, or work in the style or spirit of Dada.
How many times have scientists — natural and social — told me with casual confidence that “All theories are false”? I’ve long since lost count. My quixotic goal of the day is to write a piece to ensure that I never hear this absurd dogma again.
There is one legitimate sense in which a highly accurate theory might be false. Namely: If the theory claims to be true to an infinite number of decimal places. Suppose the true distance from Carow Hall to Kabob Corner is 1.8464864864463557967783 miles. Then in a trivial sense, the claims that the distance is 2 miles, 1.8 miles, 1.85 miles, and so on are all false. But each of these claims is literally true if you add “to the nearest mile,” “to the nearest tenth of a mile,” or “to the nearest hundredth of a mile.” And in ordinary English, such additions are implied. Near-automatic error only arises in the odd scenario where the speaker states “The distance is exactly 2.0000000000000000 miles.”
It’s tempting to dismiss “All theories are false” for ignoring probability. But the fundamental issue is the bizarre assumption that all assertions are meant to hold for an infinite number of decimal places. If I say, “The theory is 93% likely to be true,” you can interpret me as saying, “The theory is 93.00000000000000% likely to be true.” But unless I explicitly embrace this absurdly precise claim, it is absurd to so interpret me.
To make “All theories are false” true, however, it isn’t enough to find odd cases where the speaker claims accuracy to an infinite number of decimal places. “All” means “all,” and some theories definitely don’t claim to be accurate to an infinite number of decimal places. Many theories are even weaker, like “All humans are less than 20 feet tall.”
“All theories are false” is closely related to the Popperian dogma that observation can “disconfirm” a theory but never “confirm” a theory. This claim is correct for statements of the form “All X’s are Y” (unless it is possible to actually look at 100% of all X’s). If your theory states that “All swans are white,” the only way to fully confirm it is to find 100% of all swans and show that 100% are white. But the reverse holds for statements of the form “Some X’s are Y” (again, unless it is possible to actually look at 100% of all X’s). If your theory states that “Some swans are white,” you can instantly confirm it by finding a single white swan. The only way to disconfirm this claim, in contrast, would be to find 100% of all swans and show that not a single one was white.
What motivates people to declare, “All theories are false”? The most charitable story is that the speaker is warning us against dogmatically treating our Current Best Guess as the Final Answer for All Time. Einstein really did discover small errors in Newtonian physics (and large errors under rare conditions). If this is your point, however, the truism, “There’s almost always room for improvement” is much more protective than the dadaism, “All theories are false.”
Charity aside, “All theories are false” is usually invoked to promote radical skepticism or total nihilism. Yes, this promotion is almost always lazy and casual. But it gives aid and comfort to a vocal fringe of true believers in radical skepticism and total nihilism. Ayn Rand overrated the influence of these views, but they are poison nonetheless. Seriously, why search for answers if you can’t know anything, or there’s nothing to know?
There are two morals to today’s quixotic quest.
First: “All theories are false” is false. Some theories are true. Indeed, we’re awash in true theories. Many of these are only true to a few decimal places, but that’s fine. Some theories are true only under normal conditions, but that, too, is fine. Most listeners take such qualifications for granted — and any remaining ambiguity is easy enough to fix with explicit caveats.
Second: Scientists are bad at philosophy of science. Despite their sky-high IQs, despite their demonstrable ability to solve textbook problems, and even despite their original scientific research, scientists flounder at the meta level. How is this possible? Because even smart people have low Transfer of Learning. If you’re a scientist who wants to do good philosophy, you can’t just repeat snarky slogans. You’ve got to think long and hard about an unfamiliar subject. Curious? Start with Mike Huemer’s Knowledge, Reality, and Value — and before long, your philosophy of science will at least be competent.
I don’t think I’ve ever heard that phrase used except when followed by “but some are useful.” Which seems like it’s getting at the same thing you’re saying here, no?
You mis-understand what a theory is. It is NOT a prediction. It is an attempt to describe reality.
Newtonian gravity makes accurate predictions in almost all scenarios, but it is a completely false description of reality. There is NO force of gravity (as per our current best understanding) like the theory states. General relativity makes even better predictions, and is currently our best theory of gravity, but again we know it to be completely false as it assumes a classical world when the world is quantum in reality (again as per our current best understanding).
While the increases of predictive ability from Newtonian gravity -> General Relativity -> Some future Unified theory of gravity are only the smallest of decimals, it does not stop the first two theories from being completely false descriptions of reality (and likely the 3rd one once it is discovered).
Of course "All theories are false" is false, but it is not a contradiction to admit so. The statement conforms with our best understanding of how we create knowledge. It is your philosophy of Science that is sorely lacking.