Co-authored with Sunyoung Jung and Achinthya Sonil.
The natural sciences wear their hierarchy on their sleeve. Biology reduces to chemistry, chemistry to the motion of atoms, and atoms to quantum physics. Modern science has, in a real sense, a bottom.
Mathematics is stranger. Calculus, geometry, topology, number theory: they feel less like branches of a single tree and more like neighbouring countries, linked by style and temperament rather than by any single underlying principle. In the late nineteenth century a small group of mathematicians set out to change that. They wanted for mathematics what quantum physics has since given the natural sciences, namely a single source from which the rest could in principle be derived. That ambition is what we now call the foundations of mathematics, and it is mostly a question about whether a given system of axioms is both complete and consistent.
Frege's logicism
Gottlob Frege thought he had found the answer. His position, known as logicism, was that mathematics could be reduced to logic. Arithmetic, and everything built on it, would turn out to be logical consequence wearing numerical clothes.
Against Aristotle, Frege argued that numbers are not properties of objects but properties of concepts. A concept is any idea you can pick out: the colour red, the shape of a square, the property of being a prime. Each concept has an extension, which is the set of all things falling under it. The extension of red contains every red thing. The extension of square circle is empty because nothing falls under it.
From there Frege made the bold move. Numbers themselves are extensions of concepts. The number four is the extension of the concept "sets containing exactly four things." Counting becomes a logical operation on extensions.
To make this work Frege needed one axiom: every well-formed concept has a corresponding extension. He called it the General Comprehension Principle. With it, logicism had its foundation. Or so it appeared.
The letter
In June 1901, Bertrand Russell sent Frege a letter. Frege's work was weeks from the printing press. Russell's letter contained a single question:
Consider the set of all sets that are not members of themselves. Is that set a member of itself?
Formally, let
and ask whether .
If , then by the defining property of , . If , then satisfies the defining property, so .
Either answer contradicts itself. The set Russell described is well-defined in Frege's system. It corresponds to a perfectly reasonable concept, "set that does not contain itself", and yet it cannot exist. The General Comprehension Principle produces a contradiction out of its own most natural use.
Frege's reaction was not stoical. He suffered a breakdown severe enough to put him in hospital, and he eventually abandoned much of what he had written on the foundations of mathematics.
Two ways to see it
Russell's paradox is sharper than it first looks. Two old analogies help bring it into focus.
The liar's note. Imagine a piece of paper with two statements:
- Side A: The statement on the other side is false.
- Side B: The statement on the other side is true.
If A is true, then B tells us A is false. If A is false, then B is true, which makes A true. Neither assignment of truth values is stable. The two statements refer to one another in a way that forbids consistency.
The barber. A barber in a small village cuts the hair of all and only those men who do not cut their own hair. Does the barber cut his own hair? If he does, then by his rule he must not. If he does not, then by his rule he must. The barber cannot exist, not because of plumbing but because the definition is self-contradictory.
Russell's paradox is the set-theoretic version of the same move. Most sets are not members of themselves. The set of all flowers is not a flower. The set of all fish is not a fish. But some sets are members of themselves: the set of all things that are not dogs is itself not a dog, so by its own rule it belongs inside itself. Once you allow unrestricted set formation, the set of sets that do not contain themselves is perfectly grammatical and perfectly contradictory at the same time.
Why it mattered
Frege's project depended on the General Comprehension Principle as an axiom. Axioms are the starting point of a formal system. If an axiom produces a contradiction, anything can be derived from it and the whole system collapses. That is why Russell's letter was not a technical nuisance but an existential one. The paradox did not show that some particular theorem of Frege's was wrong. It showed that the bedrock he had chosen could not support any theorems at all.
What replaced it
The foundations of mathematics as they are practised today rest on Zermelo–Fraenkel set theory, ZF, or ZFC once you add the axiom of choice. ZF keeps much of Frege's spirit: sets, predicate logic, rigorous deduction. What it gives up is the General Comprehension Principle in its unrestricted form. You cannot form the set of all satisfying a property out of nothing. You can only form the set of all in some existing set that satisfy :
This is the axiom of separation (or specification), and it blocks Russell's construction cleanly. You cannot write because there is no ambient set from which to separate.
Set theory is still a live field. Questions about large cardinals, the continuum hypothesis, and the independence of axioms are all open. What Russell's paradox settled permanently is that you cannot build a foundation by allowing every concept to correspond to a set. Some restriction is not optional. It is the price of consistency.
References
- Zalta, E. N. (2022). Gottlob Frege. Stanford Encyclopedia of Philosophy.
- Zalta, E. N. (2023). Frege's Theorem and Foundations for Arithmetic. Stanford Encyclopedia of Philosophy.
- Irvine, A. D. (2020). Russell's Paradox. Stanford Encyclopedia of Philosophy.