I · Analytic foundations and direct methods
Chapters 1–4 build the common trunk: complex singularities, monodromy and connection matrices, the Heun laboratories, Wronskians, recurrences, and verified numerical computation.
A linear ODE can be solved locally without revealing how its solutions fit together globally. This book begins where local existence theory leaves off: with the connection matrices, monodromy, Stokes data, and boundary functions that turn local solutions into spectra and physical response.
For canonically normalized fundamental solutions and in two regions, the organizing equation is
The matrix is simple to write and often difficult to compute. Its entries can be obtained directly from Wronskians, analytic continuation, recurrences, or numerical matching. In suitable families, selected parts of the same global data can instead be encoded by isomonodromic tau functions, conformal blocks, quantum periods, Seiberg–Witten/Nekrasov–Shatashvili quantities, spectral determinants, ODE/IM functional relations, or TBA equations.
The book develops that problem for second-order complex linear ODEs and their locally equivalent traceless rank-two systems. Globally, three constructions are kept separate: a single-valued trace-zero gauge of a given system uses a horizontal determinant trivialization; an lift of its projectivization is a lifting or twisting problem; and a scalar oper realization uses a theta characteristic. Higher-order equations enter only where they are intrinsic to ODE/IM.
Write a fundamental matrix with columns formed from two independent scalar solutions,
A connection matrix is meaningful only together with its normalization ledger. If the two bases are changed by constant invertible matrices,
then the same continuation problem is represented by
Thus a bare gamma-function formula, continued fraction, tau function, or period is not yet a connection formula. One must also specify the equation, bases, leading normalizations, branches, continuation path, resonant exceptions, and—at an irregular singularity—the sectors and lateral summation prescription.
The book is organized around global connection data. Solid arrows are direct ODE constructions or checks; dashed arrows require a model-specific dictionary and stated analytic hypotheses.
Three layers recur throughout the book:
| Layer | Typical objects | Question being answered |
|---|---|---|
| Local | Frobenius bases, formal normal forms, canonical sectorial solutions | What are the independent solutions near this point or in this sector? |
| Global | Connection matrices, monodromy and Stokes matrices, Wronskians, boundary functions | How do normalized solutions compare after continuation? |
| Spectral or physical | Eigenvalues, resonances, quasinormal modes, Green functions, correlators | Which global coefficient vanishes, has a pole, or gives a response ratio? |
The middle layer is indispensable. A boundary condition becomes a spectral condition only after it has been translated into a statement about a normalized connection coefficient or boundary Wronskian.
No single equation naturally illustrates every method. The book therefore uses two families in parallel.
| Laboratory | Progression | What it teaches |
|---|---|---|
| Connection problems | Hypergeometric HeunG confluent Heun equations | Frobenius bases, Wronskians, recurrences, monodromy, isomonodromy, conformal blocks, AGT, black-hole boundary data |
| Quantum curves | Airy and Weber Mathieu and modified Mathieu anharmonic and Razavy systems | Turning points, exact WKB, resurgence, quantum periods, Seiberg–Witten/NS theory, ODE/IM, TBA, tunnelling spectra |
Each recurring example receives a persistent data sheet: equation and normal form, singularities and ranks, local and formal exponents, preferred bases, branch cuts and cycles, boundary conditions, known global data, and independent numerical checks.
I · Analytic foundations and direct methods
Chapters 1–4 build the common trunk: complex singularities, monodromy and connection matrices, the Heun laboratories, Wronskians, recurrences, and verified numerical computation.
II · Isomonodromy and conformal blocks
Chapters 5–7 develop monodromy-preserving deformation, Painlevé systems, BPZ equations, and the distinct roles of analytic blocks and classical large- blocks.
III · Exact WKB and resurgence
Chapters 8–9 pass from formal WKB geometry to Borel summation, Stokes graphs, Voros symbols, wall crossing, transseries, and chamber-correct quantization.
IV · Seiberg–Witten theory and AGT
Chapters 10–11 establish the cycle, normalization, mass, momentum, and accessory-parameter dictionaries needed to compare ODE, CFT, and gauge descriptions.
V · ODE/IM and TBA
Chapters 12–13 derive spectral-determinant functional relations and show when analyticity data turn them into nonlinear integral equations.
VI · Integrating applications
Chapters 14–15 compare the methods on quantum-mechanical potentials, black-hole perturbations, quasinormal modes, and holographic response.
Chapters 1–4 are the common prerequisite trunk. Later branches can be read selectively, but the AGT dictionary assumes both the CFT and Seiberg–Witten/NS branches, while the TBA chapter assumes exact WKB and the ODE/IM functional-relation language.
The correspondences in this book are complementary tools, not interchangeable names for one universal object.
| Method | Primary output | What it does not provide automatically |
|---|---|---|
| Wronskian matching | Exact ratios of connection coefficients or a boundary spectral condition | A stable computation when the chosen bases cannot be evaluated in a common domain |
| Recurrences and continued fractions | Minimal-solution conditions, connection amplitudes, spectra | Convergence and conditioning without asymptotic analysis |
| Isomonodromy | Monodromy-preserving flow, tau function, accessory data | A deformation family for an arbitrary isolated ODE |
| Analytic blocks | Fourier/structure-constant expansions of isomonodromic tau functions | The classical large- BPZ connection problem |
| Classical large- blocks | Accessory parameters and, with extra data, connection coefficients | Fusion, braiding, normalization, and analytic continuation factors by themselves |
| Formal WKB | Asymptotic solutions and formal quantum periods | A unique exact value for a divergent series |
| Exact WKB and resurgence | Directional Borel sums, Stokes jumps, chamber-dependent quantization | A chamber-independent formula without summability and boundary hypotheses |
| Seiberg–Witten/NS theory | Quantum periods and twisted-superpotential data | A model-independent nonperturbative completion |
| ODE/IM | Spectral determinants and functional relations for suitable ODE families | A construction for every linear ODE |
| TBA or Riemann–Hilbert integral equations | Nonlinear integral equations and high-precision resummed data | Equivalence between ODE/IM TBA and exact-WKB/GMN equations in general |
This distinction is a practical safeguard. For example, a zero of an isomonodromic tau function becomes a quantization condition only after the relevant monodromy and boundary constraints are imposed. Likewise, a classical conformal block principally fixes accessory data; a full connection coefficient also needs basis normalization and degenerate fusion or braiding data.
The book labels advanced statements by status:
Substantial formulas include a normalization ledger. Computational claims aim for two independent checks chosen from an exactly solvable limit, a Wronskian identity, direct integration, a recurrence, an independent period calculation, spectral comparison, or a controlled confluence or symmetry limit.
The main audience is graduate students and researchers in mathematical physics, special functions, spectral theory, gravity, quantum field theory, and integrable systems. The common trunk assumes:
The just-in-time toolkit supplies the required fundamental groups, local systems, Riemann surfaces, absolute and relative homology, intersection pairings, and minimal operator theory. CFT, supersymmetric gauge theory, and integrability are introduced only to the depth needed for the ODE problem.