Reconstruction method
This is a bibliographic and methodological reconstruction, not a reproduction of Singer’s papers, lectures, or archival materials. The corpus expands papers, named theorems, proof techniques, mathematical-physics translations, and institutional roles into 300 lecture-style cases. Strategy tags overlap; percentages do not sum to 100%.
Core thesis
Singer’s method repeatedly converts a difficult analytic or geometric situation into a stable invariant: holonomy from curvature, index from kernel-cokernel imbalance, K-class from symbol, local density from heat kernels, boundary correction from eta, and physics from Dirac-type operators.
Deep reading unit
Each case is read as a pipeline: object → operator/connection/spectrum → invariant → translation → proof mechanism → bridge theorem or field-building artifact.
Why this style
The page is designed as a research instrument: strategy cards expose method, formulas compress the move, ranking shows prevalence, corpus rows provide cases, and demonstrations show reusable workflows.
The Singer strategy engine
Overlapping prevalence ranking
Bars show count divided by 300 cases. Since a case carries multiple strategy tags, the displayed prevalence is a method-frequency map rather than a probability distribution.
Decision tree for reading Singer as method
300-case corpus
| # | Case | Family | Deep reading move | Strategies |
|---|