The research cycle: Correspondences and Mathematical Physics

One equation, three readings

The same equation can be read through its solutions, through its symmetries, or through the functions recording their traces. How do you pass from one reading to another without losing what gives them meaning? That is the question that opens volume four. And with it, what the collection calls the research cycle. Articulated Course in Mathematics. Reader’s guide, episode eight: the research cycle.

Two volumes after Graduate Mathematics

Volumes one to three form a continuous progression, from secondary school to the master’s level. Volumes four and five open the research cycle. Correspondences runs from class field theory to the Langlands program. Mathematical Physics, from symmetries to strings. Both assume Graduate Mathematics, but can be read independently of one another. A bridge, at the start of each volume, sets out the change of perspective.

Volume IV: Correspondences

Volume four follows the question from number fields to the geometric correspondence, and a path through its 2024 proof. Thirty-seven chapters, two hundred and thirty-eight articles, eight parts. The first five build the arithmetic program: reciprocity, modular forms, arithmetic geometry, Shimura varieties, motives. The last three change the objects of the correspondence: they prepare the geometric statement, follow the course of its proof, then open toward physics. The book addresses readers who want to understand how the domains connect, then return to the sources for the complete constructions and proofs.

What the book keeps distinct

Distinguishing these routes is part of the book’s purpose. The unramified de Rham geometric theorem does not resolve the whole arithmetic program. Its physical interpretation does not by itself prove a duality of gauge theories. Hypotheses, normalisations and the nature of the categories are retained in the statements. And a difficult result is identified as used without proof when an article explains its role rather than reproducing its demonstration. A dictionary becomes useful only when it preserves what sentences allow us to do. Matching names does not construct a correspondence.

Volume V: Mathematical Physics

Volume five studies how mathematical structures become physical descriptions. A symmetry specifies possible transformations; it does not choose a dynamics by itself. An action yields equations; it does not yet determine their quantisation. An effective theory produces predictions at a specified precision; it does not claim to describe every scale. Forty-four chapters, four hundred and six articles. Mechanics, representations and relativity prepare quantum states. Field geometry leads to quantisation and gauge theories. Then dualities, strings, holographic dictionaries. A descending route finally recovers certain physical regimes through calculated limits.

The article and the research question

In both volumes, the article remains the unit of reading: a question, a construction, a worked example, and its interpretation. Further readings lead to the sources. And each chapter closes on a research question. They are not all open problems, and the book says so. Their role is to give the reader an operation to pursue: extend a calculation, test a distinction, situate a conjecture.

Two volumes that answer each other

Volume four leads to geometric Langlands. Volume five meets its physical reading through gauge dualities and branes, without confusing it with the proof. The two volumes speak to each other, and either can be approached using the tools of volume three.

To remember

To remember. Two volumes, one research cycle, one shared base: Graduate Mathematics. They keep their distinctions: what is proved, what is used without proof, what remains a question. Each chapter ends on an operation to pursue. That was the last episode of the reader’s guide. The link is in the description.