Seven ways to break suddenly
René Thom’s claim, developed between 1968 and 1972, was that sudden, qualitative jumps in a system’s behaviour, rather than smooth gradual change, can be sorted into a small, fixed list of geometric patterns, regardless of what kind of system is actually producing them. The mathematics behind this comes from studying degenerate critical points of a potential function, the places where not just the slope but several further derivatives vanish at once, which is where a system’s stable state can suddenly disappear or split. Thom proved that when a system depends on at most two variables that can actually change and up to four adjustable parameters, only seven distinct geometric forms of this kind of discontinuity are possible, ranging from a simple fold and a slightly more elaborate cusp through the swallowtail and butterfly forms to three so-called umbilic patterns.
From topology to a classification
The classification itself rests on Thom’s earlier background in topology, including singularity theory and the study of stable mappings, applied here to potential functions with a bounded number of variables and parameters. Proving that exactly seven qualitatively distinct local shapes exhaust the possibilities under those limits was a piece of pure mathematics, later given an independent structural grounding when Vladimir Arnold connected the same seven forms to the classification of simple Lie groups through what is now called the ADE system. The theory’s public life, though, was built largely by Christopher Zeeman, who through the 1970s took Thom’s geometric templates and applied them well beyond the physics problems they were first designed for, to a wide range of systems where a sudden shift in behaviour seemed to call for the same kind of explanation.
Where the geometry actually applies
The pure mathematics has never been challenged and remains standard: the proof that seven elementary catastrophes exhaust the possible forms under Thom’s constraints is accepted without dispute, and Arnold’s later link to Lie group classification only reinforced how solidly it sits within topology. Where a genuine, measurable potential function exists, the applications have held up well too. The geometry of the fold and cusp describes the edges of rainbows and other optical caustics, including the bending of light around massive objects, closely enough that these remain standard illustrations of the theory at work. Cases in structural fracture mechanics and in electron transfer reactions in chemistry are cited in the same way, as situations where an underlying potential can actually be written down and its predicted discontinuity checked against what is observed.
The stressed dog problem
What did not hold up was the wave of applications built without that measurable potential function behind them. A 1977 critique published in Nature accused a substantial share of the theory’s applications outside physics and mathematics, particularly in biology and the social sciences, of resting on incorrect reasoning, far-fetched assumptions and claims well beyond what the underlying mathematics could support, singling out models such as one purporting to describe a stressed dog’s behaviour as the kind of exercise that borrowed the geometry’s vocabulary without any of its rigour. That criticism landed, and the theory’s popularity in those looser applied fields declined sharply once the gap between a genuine potential function and a suggestive diagram had been so publicly pointed out.
A discipline correcting itself
Where the mathematics is used carefully, it continues to do real work: the same fold and cusp geometry that describes optical caustics and fracture behaviour has also been applied, more informally, to phenomena such as cloud condensation dynamics and even real estate price modelling, wherever a sudden qualitative shift needs a geometric account rather than a purely descriptive one. Beyond its direct applications, the episode has become a useful case study in its own right, showing how a genuinely rigorous piece of topology can be stretched into domains it was never built to describe once it becomes fashionable, and how the correction, insisting that a catastrophe-theoretic claim requires an actual, measurable potential function rather than a resemblance in shape, amounts to a discipline policing its own overreach.
Is it worth your time
This is worth understanding both for the mathematics, which is elegant and genuinely settled, and for the cautionary arc that followed it, which says something useful about how quantitative ideas travel between fields. Readers curious about how a small number of geometric shapes can classify every possible way a stable system can suddenly break will find real substance here, along with clear examples, in optics and fracture mechanics, of the theory doing exactly what it claims. It is a weaker guide for anyone hoping the same geometry explains behaviour or social phenomena, since that is precisely the extension the mathematics does not support, and precisely the one that damaged the theory’s reputation once critics looked closely at what such applications actually assumed.