Semi-topological Galois cohomology and Weierstrass realizability

This paper establishes a cohomology theory for the absolute semi-topological Galois group associated with Weierstrass polynomials, utilizing spectral sequences to derive obstruction theories and vanishing results that ultimately prove the $\pi_1$-detectable Weierstrass realizability conjecture for abelian varieties, smooth complex projective curves, and ruled surfaces.

The Gist

Imagine you are a detective trying to solve a mystery about the shape of a hidden world. In mathematics, this "world" is a geometric space (like a sphere, a donut, or a complex surface), and the "mystery" is understanding how things wrap around it.

This paper, written by Jyh-Haur Teh, introduces a new detective tool called Semi-Topological Galois Theory. It's a way of looking at shapes that sits right in the middle between two other ways of looking at them: pure geometry and pure algebra.

Here is the breakdown of the paper's ideas using simple analogies.

1. The Two Worlds: The "Rigid" and the "Flexible"

To understand this paper, imagine two ways to study a shape:

• The Classical Way (Topology): Imagine a rubber sheet. You can stretch it, twist it, and poke holes in it, but you can't tear it. Mathematicians use this to study "loops" (paths that start and end at the same spot). This is very flexible.
• The Algebraic Way (Galois Theory): Imagine a lock and key system. You have a polynomial equation (like $z^2 - 2 = 0$). To solve it, you might need to "unlock" a new world where the solution exists. This is very rigid and precise.

The Problem: Sometimes, the flexible rubber sheet world and the rigid lock-and-key world don't agree on what is possible.

2. The New Tool: "Weierstrass Polynomials" as Keys

The author introduces a specific type of mathematical key called a Weierstrass polynomial. Think of this as a special recipe for a function that changes as you move around your shape.

• The Analogy: Imagine you are walking around a city (your shape). At every street corner, you have a magic machine that spits out a polynomial equation.
• The Goal: You want to find a "Splitting Cover." This is like finding a special map of the city where, if you look at the machine at every corner, the equation instantly breaks down into simple, distinct answers (roots).
• The Discovery: The paper shows that for any such machine, there is a minimal map (a covering space) where this happens. This map is the "Splitting Cover."

3. The "Absolute Semi-Topological Galois Group"

Now, imagine you collect all possible minimal maps for all possible magic machines in the city.

• The Group: You take all these maps and look at how they relate to each other. The collection of all these relationships forms a giant, infinite structure called the Absolute Semi-Topological Galois Group ($\Pi_{ST}$).
• What it does: This group acts like a "super-filter." It tells you exactly which loops in your city can be "unlocked" by these polynomial machines.
  • If a loop is in this group, it means the polynomial machine can "see" it and break it down.
  • If a loop is not in this group, the machine is blind to it.

4. The "Cohomology" (The Detective's Notebook)

The paper builds a "notebook" (cohomology theory) to record what this group knows.

• The Comparison: The author compares this new notebook to the old, standard notebook (Singular Cohomology).
• The Map: There is a "translation map" ($\Phi$) that tries to translate notes from the new group into the old language.
• The Big Question: Does this translation cover everything? In other words, can every geometric feature we see in the standard notebook be explained by our polynomial machines?

5. The Main Discoveries (The "Aha!" Moments)

The author tests this new tool on different shapes and finds some surprising results:

• The "Free" Shape (Loose Strings): If your shape is like a tangled ball of string with no loops (a free group), the new tool is perfect. It sees everything. The new group is identical to the old one.
• The "Finite" Shape (Rigid Blocks): If your shape has a finite number of loops (like a projective plane), the new tool sees nothing. The group becomes trivial (empty). This means polynomial machines are completely blind to these shapes.
• The "Donut" (Torus): This is the star of the show. For a donut (or a multi-holed donut), the new tool works perfectly. The author proves that every geometric feature (specifically, divisor classes, which are like counting points or lines on the surface) that can be detected by loops can be "realized" by a polynomial machine.
  • Analogy: Imagine you have a complex pattern on a donut. The paper proves you can always find a polynomial recipe that "unlocks" and explains that pattern.

6. The "Realizability Conjecture"

The paper proposes a bold guess (conjecture):

"If a geometric feature on a shape can be detected by looking at its loops, then there is a polynomial machine that can explain it."

The author proves this is TRUE for:

1. Abelian Varieties: Complex, multi-dimensional donuts.
2. Smooth Curves: Shapes like a donut with holes (genus $\ge 1$).
3. Ruled Surfaces: Shapes made by dragging a line along a curve.

7. Why Does This Matter? (The "So What?")

This isn't just about abstract math; it connects to Projective Monodromy.

• The Analogy: Imagine you are a pilot flying a plane (a projective representation). You know your path, but your compass is slightly off (it's "projective," not "linear").
• The Result: The paper shows that if you fly through a "Splitting Cover" (a special polynomial route), you can fix your compass. You can turn your "off" compass into a perfect "on" compass.
• The Condition: You can only fix the compass if the "error" (the Schur multiplier) is something that the polynomial machines can detect.

Summary

Jyh-Haur Teh has built a bridge between the flexible world of shapes and the rigid world of equations.

• He created a new "lens" (the Semi-Topological Galois Group) to look at shapes through the eyes of polynomial equations.
• He proved that for many important shapes (like donuts and complex surfaces), this lens is powerful enough to explain every geometric feature that loops can detect.
• He showed that this lens can also fix "broken" mathematical compasses (linearizing monodromy), but only if the breakage is "polynomial-friendly."

In short: Polynomials are surprisingly powerful detectives. They can solve almost every mystery that loops can find, provided the shape isn't too rigid or too simple.

Technical Summary

Here is a detailed technical summary of the paper "Semi-topological Galois cohomology and Weierstrass realizability" by Jyh-Haur Teh.

1. Problem Statement and Motivation

The paper addresses the intersection of algebraic geometry, topology, and Galois theory. The central problem is understanding the extent to which geometric classes (specifically divisor classes on projective manifolds) can be realized via polynomial data, specifically Weierstrass polynomials.

• Context: Classical covering space theory classifies finite coverings of a space $X$ via the profinite completion of the fundamental group, $\widehat{\pi_1(X, x)}$. However, not all finite coverings arise from the splitting of polynomials with continuous coefficients.
• The Gap: There is a discrepancy between the "classical" topological regime (all finite covers) and the "semi-topological" regime (covers arising specifically from the splitting of Weierstrass polynomials).
• The Goal: To develop a cohomology theory that captures the "semi-topological" fundamental group and to determine which cohomology classes (specifically those related to divisors) are "Weierstrass realizable." This leads to the formulation of the $\pi_1$-detectable Weierstrass realizability conjecture.

2. Methodology and Theoretical Framework

The author constructs a new cohomology theory based on the inverse limit of deck groups of splitting coverings.

A. Semi-Topological Galois Theory

• Weierstrass Polynomials: Defined as monic polynomials $f \in C(X)[z]$ where the specialization at every point $x \in X$ has distinct roots.
• Splitting Coverings: For such a polynomial, there exists a minimal finite covering $E_f \to X$ where $f$ splits completely into linear factors. This is the topological analogue of a splitting field.
• The Group $\Pi_{ST}(X, x)$: The absolute semi-topological Galois group is defined as the inverse limit of the deck groups of all such splitting coverings:
  $$\Pi_{ST}(X, x) := \varprojlim_{(f, E_f)} \text{Deck}(E_f/X)$$
  There is a canonical surjection $\rho: \widehat{\pi_1(X, x)} \twoheadrightarrow \Pi_{ST}(X, x)$. The kernel $K_{ST}(X, x)$ measures the "non-realizable" topological covers.

B. Cohomology Theory

• Definition: For a discrete $\Pi_{ST}(X)$-module $A$, the semi-topological Galois cohomology is defined as the continuous group cohomology:
  $$H^n_{ST}(X, A) := H^n_{cont}(\Pi_{ST}(X), A)$$
• Comparison Maps: The paper constructs a canonical chain of maps relating this new theory to classical singular cohomology:
  $$\Phi^n_X: H^n_{ST}(X, A) \xrightarrow{\rho^*} H^n_{cont}(\widehat{\pi_1}, A) \xrightarrow{\iota^*} H^n(\pi_1, A) \xrightarrow{c^*} H^n_{sing}(X, A)$$
  A class $\alpha \in H^n_{sing}(X, A)$ is Weierstrass realizable if it lies in the image of $\Phi^n_X$.

C. Obstruction Theory

• The paper utilizes the Lyndon-Hochschild-Serre (LHS) spectral sequence associated with the exact sequence $1 \to K_{ST} \to \widehat{\pi_1} \to \Pi_{ST} \to 1$.
• This spectral sequence provides an obstruction theory for semi-topological embedding problems (lifting projective monodromy to linear monodromy via splitting covers).

3. Key Contributions and Structural Results

The paper establishes fundamental properties of $H^n_{ST}$ and computes them for various topological spaces:

1. ST-Fullness for Free Groups: If $\pi_1(X)$ is a free group, then every finite covering is dominated by a splitting covering. Consequently, $\Pi_{ST}(X) \cong \widehat{\pi_1(X)}$, and the comparison map is an isomorphism.
2. Vanishing for Torsion Groups: If $\pi_1(X)$ is a torsion group (e.g., finite fundamental groups like $\mathbb{Z}/2$ for $\mathbb{RP}^2$), then $\Pi_{ST}(X)$ is trivial.
   • Result: $H^n_{ST}(X, A) = 0$ for $n > 0$.
   • Implication: For spaces like $\mathbb{RP}^2$, the comparison map $\Phi^2_X$ is the zero map, meaning non-trivial singular cohomology classes are not Weierstrass realizable.
3. Braid-Free Spaces: The paper introduces "braid-free" groups (those with no non-trivial homomorphisms to braid groups $B_n$). For such spaces, all splitting coverings are trivial, leading to the same vanishing results as torsion groups.
4. Tori and Abelian Varieties:
   • For the torus $T^2$, the paper proves that the map $\Phi^2_{T^2}$ is surjective for all finite coefficients.
   • This result is extended to higher-dimensional tori and complex abelian varieties.
5. Linearization of Monodromy: The paper proves that a finite projective representation $\bar{\rho}: \Pi_{ST}(X) \to PGL_n(\mathbb{C})$ can be lifted to a linear representation $GL_n(\mathbb{C})$ after passing to a splitting cover if and only if its associated Schur-multiplier class in $H^2(G, \mu_m)$ is semi-topologically realizable.

4. Main Theorem: The $\pi_1$-Detectable Weierstrass Realizability Conjecture

The paper formulates and proves the following conjecture for specific classes of varieties:

Conjecture: For a smooth complex projective variety $X$, every divisor class (modulo $m$) that is detected by the fundamental group (i.e., lies in the image of $H^2(\pi_1(X), \mathbb{Z}/m) \to H^2_{sing}(X, \mathbb{Z}/m)$) is Weierstrass realizable.

Proofs of the Conjecture:
The author proves this conjecture holds for:

1. Abelian Varieties: By reducing to the torus case ($T^{2g}$) and using the surjectivity of $\Phi^2$ on tori.
2. Smooth Complex Projective Curves ($g \ge 1$): By constructing specific Weierstrass polynomials ($z^m - u_i$) whose splitting coverings generate the necessary cohomology classes. (Note: The conjecture fails for $\mathbb{P}^1$ as $\pi_1$ is trivial).
3. Ruled Surfaces over Positive-Genus Curves: The paper shows that for $X = \mathbb{P}(E) \to C$ (where $g(C) \ge 1$), the image of the comparison map coincides exactly with the $\pi_1$-detectable classes.

5. Significance and Impact

• New Invariant: The introduction of $\Pi_{ST}(X)$ and $H^n_{ST}(X, A)$ provides a new topological invariant that distinguishes between "algebraically constructible" covers and general topological covers.
• Bridge between Algebra and Topology: The work rigorously connects the algebraic notion of Weierstrass polynomials with the topological notion of covering spaces and cohomology.
• Obstruction Theory: It provides a concrete cohomological criterion (via the LHS spectral sequence) for when projective monodromy can be linearized, a problem relevant to the study of vector bundles and local systems.
• Resolution of Realizability: By proving the conjecture for abelian varieties and curves, the paper establishes that for these important geometric objects, the topological constraints imposed by the fundamental group are the only obstructions to realizing divisor classes via polynomial splitting.

In summary, Teh's paper develops a robust cohomological framework to quantify the "polynomial-ness" of topological coverings, proving that for many complex algebraic varieties, the fundamental group fully controls the realizability of divisor classes via Weierstrass polynomials.