Why Mathematics Even Works

Many objects that mathematics uses to solve a problem do not belong to the setting in which the problem is posed: imaginary numbers are not real, infinitesimals are not real numbers, roots of an irreducible polynomial need not lie in the base field, and a cohomology class is not a function on vertices. Yet the calculations end with answers in the original setting. We make this pattern precise. A lower problem fixes what its own candidates can try and which answers it accepts; a promoted layer supplies a hidden structure with no lower counterpart and a selected expression that returns an accepted answer under explicit checks. We bundle these data into a Strict Audited Utility certificate and prove that it licenses the returned answer but not the hidden structure as a lower object, and that concluding usefulness also requires the judging instrument’s permission. For root problems, non-descent is forced: if a lower operation F has no k-th root among the lower operations, and a promoted operation G, observed through a surjective map q, satisfies q ∘ Gkm = F ∘ q, then G has no lower counterpart through q; examples show that the hypotheses are needed. In an accepted computation modelled as a finite dependency graph, hidden content on which the answer depends must cross back to the lower layer at a checked boundary, and the rest can be replayed from lower data under explicit conditions; but an audit can make a hidden step essential even though the step descends, so essential use alone does not force non-descent. Complex numbers, dual numbers, Galois roots, and graph cohomology are worked out exactly, including the four-phase model of a hidden square root of reversal, polynomial first jets, the field ℚ(√2) with trace 0, norm −2 and orbit polynomial t2 − 2, and the triangle datum (1, 0, 0); they carry four distinct labels in a declared grammar of return constructors, and a shared shape transfers obligations, not evidence. The headline results are verified in Lean 4; source, Lean code, and reproducibility scripts are at https://github.com/ioannist/six-birds-math-usefulness. Version 2 replaces the general trace-necessity, normal-form, canonical-minimality, and six-role statements of version 1 by the results above, records the stronger statements as open problems with counterexamples to their unrestricted forms, and corrects several definitions; Appendix F of the paper lists the changes.

Authors

Institutions

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-10-06
DOI
https://doi.org/10.5281/zenodo.20712760
Primary Topic
History and Theory of Mathematics
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
preprint

Why Mathematics Even Works

Ioannis Tsiokos
Zenodo (CERN European Organization for Nuclear Research)
History and Theory of Mathematics
preprint

Why Mathematics Even Works

Ioannis Tsiokos
preprint en

Abstract

Many objects that mathematics uses to solve a problem do not belong to the setting in which the problem is posed: imaginary numbers are not real, infinitesimals are not real numbers, roots of an irreducible polynomial need not lie in the base field, and a cohomology class is not a function on vertices. Yet the calculations end with answers in the original setting. We make this pattern precise. A lower problem fixes what its own candidates can try and which answers it accepts; a promoted layer supplies a hidden structure with no lower counterpart and a selected expression that returns an accepted answer under explicit checks. We bundle these data into a Strict Audited Utility certificate and prove that it licenses the returned answer but not the hidden structure as a lower object, and that concluding usefulness also requires the judging instrument’s permission. For root problems, non-descent is forced: if a lower operation F has no k-th root among the lower operations, and a promoted operation G, observed through a surjective map q, satisfies q ∘ Gkm = F ∘ q, then G has no lower counterpart through q; examples show that the hypotheses are needed. In an accepted computation modelled as a finite dependency graph, hidden content on which the answer depends must cross back to the lower layer at a checked boundary, and the rest can be replayed from lower data under explicit conditions; but an audit can make a hidden step essential even though the step descends, so essential use alone does not force non-descent. Complex numbers, dual numbers, Galois roots, and graph cohomology are worked out exactly, including the four-phase model of a hidden square root of reversal, polynomial first jets, the field ℚ(√2) with trace 0, norm −2 and orbit polynomial t2 − 2, and the triangle datum (1, 0, 0); they carry four distinct labels in a declared grammar of return constructors, and a shared shape transfers obligations, not evidence. The headline results are verified in Lean 4; source, Lean code, and reproducibility scripts are at https://github.com/ioannist/six-birds-math-usefulness. Version 2 replaces the general trace-necessity, normal-form, canonical-minimality, and six-role statements of version 1 by the results above, records the stronger statements as open problems with counterexamples to their unrestricted forms, and corrects several definitions; Appendix F of the paper lists the changes.

Zenodo (CERN European Organization for Nuclear Research)
Intelligent Automation (United States) (US)
History and Theory of Mathematics
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.