Why Mathematics Works
Dissolving Wigner’s 64-Year-Old Puzzle
In 1960, the physicist Eugene Wigner published an essay with one of the great titles in the history of ideas: “The Unreasonable Effectiveness of Mathematics in the Natural Sciences.” His observation was deceptively simple. Mathematics — a discipline pursued largely for its own internal beauty — keeps turning out to be exactly what physics needs. And nobody can explain why.
Wigner called it “a wonderful gift which we neither understand nor deserve.” Sixty-four years later, we still don’t understand it. Or so the standard story goes.
The standard story is wrong. Not because the puzzle isn’t real, but because it rests on a hidden assumption that everyone shares and no one examines. Pull that assumption out and the puzzle dissolves.
The Puzzle in Three Parts
Wigner’s puzzle isn’t really one puzzle. It’s three, braided together.
The puzzle of applicability. Why does abstract mathematics — a discipline about structures, relations, and proofs, with no essential reference to the physical — apply to the natural world at all? The integers count physical objects. Differential equations govern the evolution of physical systems. Group-theoretic symmetries classify fundamental particles. That there should be any systematic correspondence between these two domains is, on its face, surprising.
The puzzle of indispensability. Mathematics isn’t merely a convenient shorthand for physics. General relativity doesn’t just use Riemannian geometry as an expository aid; it is, at the level of its fundamental commitments, a theory about the geometric structure of spacetime. Quantum mechanics doesn’t borrow from Hilbert space theory; it identifies the state of a physical system with a vector in Hilbert space. The entanglement between mathematical structure and physical theory runs all the way down.
The puzzle of serendipity. This is the really uncanny one. Again and again, mathematical structures developed for purely internal reasons — with no physical application in view — later turn out to be precisely the right tools for new physics. Riemann developed his generalized geometry in 1854 as a contribution to pure mathematics. Six decades later, Einstein found it provided the exact framework needed for general relativity. Galois developed group theory to solve algebraic problems about polynomial equations. A century later, it became the organizing principle of quantum mechanics and the Standard Model. The theory of fiber bundles was a creature of pure topology before anyone recognized it as the natural language of gauge theories.
These aren’t isolated episodes. They constitute a pattern, and the pattern demands explanation.
The Usual Suspects
Philosophers have offered a range of responses over the decades. Each one handles part of the puzzle. None handles all of it.
Platonism says the physical world instantiates mathematical structure. Fine — but why does it instantiate the particular structures that mathematicians find beautiful and worth studying? And Platonism says nothing about whether the laws are necessary or contingent. A Platonist who thinks the laws could have been otherwise faces the full force of Wigner’s puzzle.
Nominalism denies abstract mathematical objects and reconstructs mathematical truth in terms of conventions or fictions. This makes the puzzle worse, not better. If mathematical truths are merely features of our linguistic practices, why should those practices track the fundamental structure of reality with such eerie precision?
Structural realism collapses the gap between mathematical and physical structure by identifying the latter as the proper object of scientific knowledge. This is the most promising standard response — it handles applicability and indispensability well — but it stalls on serendipity and is typically silent on the modal status of physical structure. Is the structure of the world necessary or contingent? The structural realist usually doesn’t say.
The deflationary response says there’s no real puzzle: mathematics works because it was developed to work, honed by centuries of empirical feedback. This handles elementary mathematics (counting, measuring) but falls apart when you notice that the mathematics most central to modern physics — non-Euclidean geometry, group theory, fiber bundles, noncommutative algebra — was developed in response to internal mathematical pressures, not empirical needs.
Tegmark’s mathematical universe hypothesis goes nuclear: every consistent mathematical structure is physically instantiated in some universe. Problem solved — or rather, problem dissolved by brute ontological force. This works, technically, but at the cost of postulating an incomprehensible infinity of physically real universes for which we have no evidence. It explains the effectiveness of mathematics the way “everything happens somewhere” explains any particular occurrence: by draining the explanandum of all content.
The Hidden Assumption
Here’s what everyone is missing — or, more precisely, assuming without noticing.
Wigner’s puzzle has the force it does because of a background commitment that is almost universally shared but rarely made explicit: the assumption that the fundamental laws of nature are contingent. That the laws could have been otherwise. That there are metaphysically possible worlds with different fundamental physics.
Grant that assumption and the puzzle is devastating. If the laws are contingent, then the space of metaphysically possible physical worlds is vast — worlds with different force laws, different symmetry groups, different dimensionality, perhaps worlds with no mathematically tractable structure at all. Against this backdrop, the fact that our world’s laws happen to be deeply mathematizable — and happen to be mathematizable by the very structures that pure mathematicians find independently compelling — looks like an extraordinary cosmic coincidence. It’s the contingency of the laws that makes the coincidence brute.
The explanatory pressure here is real enough to have been recruited as an argument for theism. William Lane Craig, among others, has argued that the unreasonable effectiveness of mathematics points toward a divine mind: a God who selected, from the space of possible law-packages, the one with deep mathematical structure. That's not a frivolous move. If the laws really are contingent, then something has to explain why the actual laws are so thoroughly mathematizable, and a selecting intelligence is one candidate. But notice how the argument depends on there being a space of alternatives from which a selection is made. The contingentist assumption isn't incidental to the theistic argument; it's load-bearing. Which means that if the assumption is wrong — if the laws couldn't have been otherwise — the explanatory vacuum that the theistic argument is designed to fill doesn't open up in the first place. The puzzle doesn't need an answer. It needs its generating assumption identified and questioned.
What if the laws aren’t contingent?
The Necessitarian Dissolution
Necessitarianism about laws of nature is the thesis that the fundamental laws hold in all metaphysically possible worlds. There is no possible world in which masses fail to attract, in which the gauge symmetries of the Standard Model are replaced by different symmetries, or in which the Schrödinger equation takes a different functional form.
This is not a fringe position. It has a distinguished philosophical pedigree — from Parmenides through rationalist metaphysics to contemporary dispositional essentialism (Shoemaker, Ellis, Bird), Della Rocca's principle-of-sufficient-reason program, and the modal firewalls framework I’ve developed elsewhere.
It does something remarkable to Wigner’s puzzle.
If the laws are necessary, then both mathematical truths and physical laws belong to the same modal category: both hold in all metaphysically possible worlds. The correspondence between them is no longer a bridge across a modal gap — the necessary and the contingent — but an internal relation within the domain of necessary truth.
Now add one further observation. The laws aren’t just any necessary truths. They instantiate structurally deep, highly constrained patterns: symmetry, geometric organization, invariance, topological structure. Mathematics, meanwhile, is the systematic investigation of exactly such structures — and mathematicians, for their own internal reasons, are drawn toward structures that are deep, unifying, elegant, and richly interconnected. If the laws instantiate precisely such structurally privileged features, then repeated convergence between advanced pure mathematics and fundamental physics isn’t miraculous. It’s intelligible.
The puzzle of applicability dissolves because the gap between the abstract and the physical is, at the level of fundamental law, a gap between two departments of necessary truth, not a gap between the necessary and the contingent.
The puzzle of indispensability dissolves because mathematics is indispensable for formulating the laws precisely because the laws are, at the level of their structural content, mathematical — necessary truths about structures that are simultaneously mathematical and physical.
And the puzzle of serendipity — the hardest piece — dissolves because mathematicians exploring the space of necessary mathematical truths are exploring the same broad landscape that physics inhabits. They converge on the same structures not by accident but because both inquiries are constrained by the same structurally privileged features.
Consider the historical cases again under this lens. Riemann, in exploring the space of possible geometries, was necessarily exploring structures that include the geometry of spacetime — because that geometry is necessary. Group theory, as the mathematical science of symmetry, was always going to converge on the symmetry structure of the fundamental interactions — because that structure is necessary. Topologists investigating fiber bundles were investigating a space that necessarily includes the structures governing gauge theories. In each case, the necessitarian explanation has the same form: coincidence is replaced by constrained convergence.
Wigner’s Puzzle as Evidence
Here’s where it gets really interesting. The argument can be run in reverse.
If the necessitarian account of mathematical effectiveness is the best available — more unified, more parsimonious, and more explanatorily powerful than its rivals — then the observed effectiveness of mathematics is itself evidence for the necessity of laws.
Under contingentism, the systematic correspondence between internally motivated mathematics and fundamental physics is surprising. The contingentist must regard it as either a brute fact, an observer selection effect, or the product of some as-yet-unidentified mechanism. None of these is satisfactory.
Under necessitarianism, the correspondence is positively expected. The depth and systematicity of the alignment — the very features that make it seem “unreasonable” under contingentism — are significantly less surprising if the laws are necessary.
By standard abductive reasoning, the effectiveness of mathematics in the natural sciences is genuine, defeasible evidence for the necessity of laws. Wigner’s puzzle isn’t merely a problem that necessitarianism alleviates. Properly understood, it’s a datum that necessitarianism fits better than any competing view.
What Remains
Necessitarianism explains the core of Wigner’s puzzle — the modal mystery of why mathematics and physics correspond — but a pragmatic residue remains. The effectiveness of idealized models, the uncanny accuracy of approximate calculations, the fruitful application of continuous mathematics to a possibly discrete world — these pragmatic successes go beyond bare structural correspondence.
Necessitarianism reduces even this residue. If the laws are necessary, there’s a principled distinction between what an idealization must get right (necessary nomological structure) and what it may safely ignore (contingent particulars). Many of the most striking cases of pragmatic effectiveness — perturbation theory in QED, renormalization group methods, symmetry arguments — are cases where the success is underwritten by structural features of the laws that, under necessitarianism, are themselves necessary.
After the necessitarian explanation has done its work, what remains is not a fundamental mystery about the very possibility of mathematical physics, but a family of more specific, more tractable questions about modeling, approximation, and scientific discovery. That’s philosophical progress, even if it’s not a total dissolution.
The Upshot
Wigner’s puzzle has persisted for over six decades not because it’s genuinely insoluble but because the assumption that generates it — the contingency of fundamental law — is so deeply entrenched that almost no one questions it. The effectiveness of mathematics appears unreasonable precisely because we’re implicitly imagining that the laws could have been otherwise and measuring the improbability of the correspondence against that imagined space of alternatives.
Remove the alternatives and the apparent improbability vanishes.
The effectiveness of mathematics in the natural sciences is not a wonderful gift we neither understand nor deserve. It is a structural inevitability that follows from the necessity of the laws that mathematics describes. Wigner’s puzzle is not an embarrassment for necessitarianism. It is one of its strongest philosophical opportunities.
Modal Firewalls: Why Metaphysics Keeps Telling Explanation to Stop