Ganea decompositions of classifying spaces

This paper investigates homotopy decompositions of classifying spaces of compact connected Lie groups via a relative fiber-cofiber construction, establishing rational sharpness and Cohen-Macaulay properties for the resulting spaces under specific cohomological conditions while providing explicit cohomological and K-theoretic computations for examples involving maximal tori and commuting elements.

The Gist

Here is an explanation of the paper "Ganea Decompositions of Classifying Spaces" using simple language and creative analogies.

The Big Picture: Taking Apart a Complex Machine

Imagine you have a very complex, mysterious machine (let's call it $BG$). This machine represents a "Lie Group," which is a mathematical object describing continuous symmetries (like rotating a sphere or shifting shapes). The machine is so complicated that looking at it all at once is overwhelming.

The goal of this paper is to figure out how to take this machine apart into smaller, simpler, and more understandable pieces, and then show how those pieces fit back together to recreate the original machine.

In mathematics, this process is called a homotopy decomposition. Think of it like taking a complex Lego castle apart brick by brick, studying the different types of bricks, and then proving that if you glue them back together in a specific order, you get the exact same castle.

The Main Characters

1. The Machine ($BG$): The complex object we are studying.
2. The Fibers ($F$ and $F'$): These are the "ingredients" or "building blocks" we use to build our decomposition.
   • One block is usually a Flag Manifold (a fancy geometric shape related to the machine's internal structure).
   • The other block is a Sphere (a simple, round shape).
3. The Join Operation ($*$): This is the paper's main tool. Imagine you have two shapes. The "Join" operation is like stretching a rubber band between every point on Shape A and every point on Shape B, creating a new, larger, more complex shape. It's like connecting two islands with a massive bridge network.

The Story: The "Ganea Tower"

The authors use a method invented by a mathematician named Ganea. Imagine you are trying to understand a tall, dark tower (the machine $BG$). You can't see the top, so you start building a scaffold from the bottom up.

1. Step 0: You start with a simple base (the Flag Manifold).
2. Step 1: You take that base and "join" it with a Sphere. This creates a slightly more complex structure ($X_1$).
3. Step 2: You take the result of Step 1 and join it with another Sphere. This creates an even more complex structure ($X_2$).
4. The Tower: You keep doing this forever. You get a tower of shapes: $X_0 \to X_1 \to X_2 \to \dots$

The Magic Result: As you keep adding these "sphere layers" (going up the tower), the shapes get closer and closer to looking exactly like the original complex machine ($BG$). If you look at the "limit" of this infinite tower, you have perfectly reconstructed the machine.

The "Quasi-Invariants": The Secret Code

Why do we care about this tower? Because the authors discovered that the mathematical "DNA" (cohomology rings) of these tower steps looks like a special code known as Quasi-Invariants.

• The Analogy: Imagine a secret language used by a group of spies (the Weyl Group).
  • Standard Invariants: These are messages that remain exactly the same no matter how the spies rearrange the letters. (Very rigid).
  • Quasi-Invariants: These are messages that almost stay the same. If you rearrange the letters, the message changes, but only by a tiny, specific amount (like a whisper). It's a "fuzzy" version of the secret code.

The paper proves that the shapes in our Ganea Tower ($X_m$) are the physical, geometric homes of these "fuzzy" secret codes. As you go up the tower (increasing $m$), the "fuzziness" decreases, and the code becomes stricter, eventually becoming the perfect, rigid secret code of the original machine.

The "Join" Recipe

The authors realized that to make this work for any Lie Group (not just the simple ones), they needed a specific recipe:

• Ingredient A: A complex shape with no "odd" holes (rational cohomology vanishing in odd degrees).
• Ingredient B: A simple sphere with a transitive action (a sphere that the group spins around perfectly).

When you mix these two ingredients using the "Join" operation, you get a tower that behaves perfectly. The math shows that the resulting shapes are "formal" (their shape is determined entirely by their algebraic DNA) and "Cohen-Macaulay" (a fancy way of saying they are structurally sound and well-behaved).

The Appendix: The "Infinity" Toolbox

The end of the paper (the Appendix) is a bit different. It's like the authors stepping back to look at the tools they used. They realized that the "Join" operation they were using is actually a special case of a much broader, abstract concept in Infinity Category Theory (a high-level branch of math that treats shapes and spaces as if they are made of infinite layers of connections).

They proved a generalized version of Ganea's old theorem using this modern "Infinity" toolbox. It's like taking a specific screwdriver and proving it works because of the fundamental physics of how metal and torque interact. This ensures their results are rock-solid and apply to almost any mathematical universe you can imagine.

Summary

In a nutshell:
The authors built a Lego tower to take apart complex symmetry machines.

1. They started with a base and kept adding sphere layers (the Join operation).
2. They proved that as the tower grows, it perfectly reconstructs the original machine.
3. They discovered that the mathematical "blueprints" of these tower steps correspond to Quasi-Invariants—a fascinating, slightly fuzzy version of symmetry codes.
4. They used modern Infinity Math to prove that this method works universally, not just for specific cases.

This work bridges the gap between abstract algebra (polynomials and symmetries) and topology (shapes and spaces), showing us how to build complex geometric worlds out of simple, repetitive steps.

Technical Summary

Here is a detailed technical summary of the paper "Ganea Decompositions of Classifying Spaces" by Yuri Berest, Yun Liu, and Ajay C. Ramadoss.

1. Problem Statement and Motivation

The paper addresses the problem of constructing homotopy decompositions of the classifying space $BG$ of a compact connected Lie group $G$. Specifically, the authors seek to realize the algebraic structures known as quasi-invariants of Weyl groups in a topological setting.

• Algebraic Context: In the 1950s, Borel and Chevalley established that the rational cohomology $H^*(BG, \mathbb{Q})$ is isomorphic to the ring of Weyl group invariants $H^*(BT, \mathbb{Q})^W$. Later, the theory of quasi-invariants $Q_m(W)$ was developed (by Finkelberg, Varchenko, Etingof, Ginzburg, etc.). These are subalgebras of the polynomial ring $k[V]$ defined by congruence conditions modulo powers of reflection hyperplanes. They form a filtration between the full polynomial ring and the invariant ring, possessing remarkable properties: they are free modules over the invariant ring and are Cohen-Macaulay.
• Topological Gap: In a previous work [BR], the authors successfully constructed a topological filtration of spaces $X_m(G, T)$ for $G=SU(2)$ (rank 1) using the classical fiber-cofiber construction (Ganea construction). These spaces had cohomology rings isomorphic to the quasi-invariants $Q_m(W)$.
• The Challenge: For higher-rank Lie groups ($rk(G) > 1$), applying the classical fiber-cofiber construction to the fundamental Borel fibration $G/T \to BT \to BG$ fails. The resulting spaces do not have Cohen-Macaulay cohomology rings, and thus cannot represent the quasi-invariants. The paper aims to generalize the construction to arbitrary compact connected Lie groups and identify the necessary topological conditions to recover the "quasi-invariant" properties.

2. Methodology

The authors employ a combination of homotopy theory, rational homotopy theory, and $\infty$-category theory.

A. The Relative Ganea (Join) Construction

Instead of the classical fiber-cofiber iteration, the authors utilize the join of fibrations.

1. Setup: Given two Borel fibrations over $BG$:
   • $F \to X = EG \times_G F \to BG$
   • $F' \to X' = EG \times_G F' \to BG$
2. Iteration: They define a tower of spaces $X_m(F, F')$ inductively. The fiber $F_m$ is the join of the previous fiber with $F'$ (specifically, $F_{m} = F_{m-1} * F'$), and $X_m$ is the homotopy quotient $EG \times_G F_m$.
   • This creates a mapping telescope (tower) $X_0 \to X_1 \to \dots \to BG$ which converges to $BG$.
3. Key Innovation: The authors replace the standard cofiber step with a relative join over the base $BG$. This allows them to control the cohomological properties of the resulting spaces by choosing specific pairs of fibers $(F, F')$.

B. Rational Homotopy and Spectral Sequences

• The authors analyze the rational cohomology $H^*(X_m, \mathbb{Q})$ using the Leray-Serre spectral sequence.
• They prove that under specific conditions on $F$ and $F'$, the spectral sequence collapses, making $H^*(X_m, \mathbb{Q})$ a free module over $H^*(BG, \mathbb{Q})$.
• They utilize the Eilenberg-Moore spectral sequence to compute the cohomology of the homotopy pushouts involved in the join construction.

C. $\infty$-Categorical Framework (Appendix)

A significant portion of the paper (Appendix A) re-examines the Ganea construction in the language of $\infty$-categories (specifically Quillen model categories and $\infty$-topoi).

• They define fiber and cofiber functors as adjoint functors on the $\infty$-category of morphisms.
• They construct a canonical Ganea tower functor $\Xi$ taking values in Mather cubes (cubical diagrams satisfying specific pushout/pullback axioms).
• They prove a generalized Ganea Theorem for $\infty$-topoi, establishing the convergence of the tower to the base space under hypercompleteness conditions. This provides a rigorous foundation for the topological constructions used in the main text.

3. Key Contributions and Results

A. Main Theorem (Theorem 1.4 / Theorem 3.1)

The authors establish sufficient conditions for the join construction to yield spaces with "quasi-invariant" properties:

• Assumptions:
  1. $F$ is a $\mathbb{Q}$-finite $G$-CW complex with rational cohomology vanishing in odd degrees (e.g., $F = G/T$ or $E_{com}G_1$).
  2. $F'$ has the $G$-homotopy type of an odd-dimensional sphere $S^{2n-1}$ with a transitive $G$-action (e.g., $G/H$ where $G/H \cong S^{2n-1}$).
• Conclusions:
  • The rational cohomology rings $Q_m(F, F') := H^*(X_m(F, F'), \mathbb{Q})$ are free modules over $H^*(BG, \mathbb{Q}) \cong \mathbb{Q}[V]^W$ of rank $\dim_{\mathbb{Q}} H^*(F, \mathbb{Q})$.
  • These rings are Cohen-Macaulay.
  • The maps in the tower induce an injective filtration:
    $$Q_0 \leftarrow Q_1 \leftarrow \dots \leftarrow Q_m \leftarrow \dots$$
    such that $\varprojlim Q_m \cong H^*(BG, \mathbb{Q})$.
  • The decomposition is sharp: the associated Bousfield-Kan spectral sequence degenerates.
  • The spaces $X_m$ are rationally formal and Cohen-Macaulay.

B. Explicit Presentations

The authors provide explicit algebraic presentations for the cohomology rings:
$$Q_m(F, F') \cong \{ f \in Q_0(F, F') \mid s_\alpha(f) \equiv f \pmod{\theta^m} \}$$
where $\theta$ is the Euler class of the spherical fibration $F' \to X' \to BG$, and $s_\alpha$ are reflection operators. This is a topological analog of the algebraic definition of quasi-invariants.

C. Examples

The paper constructs numerous examples satisfying the theorem's conditions:

1. Flag Manifolds: Taking $F = G/T$ (the flag manifold) and $F' = G/H$ where $G/H$ is a homogeneous sphere (e.g., $SU(n)/SU(n-1) \cong S^{2n-1}$, $Spin(7)/G_2 \cong S^7$).
2. Commutativity Classifying Spaces: Taking $F = E_{com}G_1$ (the identity component of the universal bundle for commuting elements). This yields new analogs of quasi-invariants related to the cohomology of $B_{com}G$.
3. Generalized Join: The authors extend the construction to filtered joins involving sequences of subgroups, allowing for more complex combinatorial structures (Section 5).

D. Equivariant K-Theory

The paper computes the equivariant K-theory $K_G^*(X_m)$ for these spaces. They show that $K_G^*(X_m)$ is isomorphic to a subalgebra of the representation ring $R(T)$ defined by congruence conditions modulo powers of an element $\Theta$ in $R(G)$, analogous to the cohomology results.

4. Significance

• Unification of Algebra and Topology: The paper successfully bridges the gap between the algebraic theory of quasi-invariants (originally arising in representation theory and integrable systems) and the homotopy theory of Lie groups. It proves that the "quasi-invariant" filtration is a genuine topological phenomenon for a wide class of Lie groups, not just $SU(2)$.
• Resolution of the Rank Problem: It solves the issue where the classical Ganea construction failed for higher-rank groups by introducing the relative join construction. This provides a robust method for generating Cohen-Macaulay spaces from classifying spaces.
• New Invariants: The construction yields new topological invariants for classifying spaces of commuting elements ($B_{com}G$), expanding the understanding of these recently studied spaces.
• Foundational $\infty$-Categorical Results: The Appendix provides a modern, rigorous treatment of the Ganea construction using $\infty$-categories. By proving the convergence of the Ganea tower in hypercomplete $\infty$-topoi, the authors generalize classical theorems (like the Blakers-Massey theorem) to a broader categorical context, which may have applications beyond Lie groups.
• Computational Power: The paper provides explicit formulas for cohomology rings and Hilbert series, offering concrete tools for further research in equivariant topology and representation theory.

In summary, this paper establishes a powerful framework for decomposing classifying spaces of Lie groups into towers of spaces whose cohomology rings mimic the structure of quasi-invariants, thereby deepening the connection between Lie theory, invariant theory, and algebraic topology.