The unstable complex in Bruhat-Tits buildings for arithmetic groups over function fields

This paper establishes that for a principal congruence subgroup $\Gamma \subset GL_r(K)$ over a function field $K$, the $\Gamma$-unstable region of the Bruhat-Tits building for $GL_r(K_\infty)$ is homotopy equivalent to the spherical Tits building for $GL_r(K)$, extending Grayson's generalization of Serre's earlier result for $GL_2$.

The Gist

Imagine you are standing in a vast, infinite forest made of mathematical trees. This isn't a forest of oak or pine, but a Bruhat-Tits building. It's a giant, multi-dimensional map used by mathematicians to understand the hidden structures of numbers and shapes, specifically in a world where the "numbers" come from functions on a curve (like a circle or a twisted loop) rather than the usual integers.

In this forest, there are two main types of "terrain":

1. The Stable Ground: Areas where the rules are calm, predictable, and nothing moves.
2. The Unstable Ground: Areas where things are shaky, chaotic, and constantly shifting.

The Problem: The Shaky Forest

The authors of this paper, Gebhard Bockle and Sriram Chinthalagiri Venkata, are studying what happens when a specific group of "explorers" (called an arithmetic group, denoted as $\Gamma$) walks through this forest.

When these explorers walk around, they sometimes get stuck in the Unstable Ground. In the old days (when the forest was 2-dimensional, like a simple tree), mathematician Jean Serre discovered something amazing: The entire "shaky" part of the forest, no matter how messy it looked, could be squished down or "homotopically equivalent" to a simple, clean boundary line (called the Tits building).

Think of it like this: Imagine a crumpled piece of paper (the unstable forest). Serre found that if you carefully smooth it out, it turns into a perfect circle (the boundary). The messy details don't matter; the shape is fundamentally the same.

The Challenge: A Forest with More Dimensions

The problem is that this paper looks at forests with more than 2 dimensions (rank $r > 2$).

• In 2D, the unstable forest is a bunch of disconnected trees.
• In 3D or higher, the unstable forest becomes one giant, connected, tangled web. It's no longer a bunch of separate trees; it's a single, complex knot.

Because it's a knot, you can't just say "it's a bunch of trees." You need a new way to prove that this giant knot is actually equivalent to the clean boundary circle.

The Solution: The "Grayson" Map

The authors use a method developed by a mathematician named Grayson (who was inspired by another genius, Quillen). Here is the analogy of their method:

Imagine the giant tangled knot (the unstable complex) is a messy ball of yarn. The clean boundary (the Tits building) is a perfect sphere.

1. The Strategy: Instead of trying to untangle the whole ball at once, they look at specific "threads" in the yarn.
2. The Sub-complexes: They identify small, manageable chunks of the yarn (sub-complexes) that correspond to specific points on the perfect sphere.
3. The Magic Trick: They prove that each of these small chunks is contractible. In math-speak, this means if you grab that chunk of yarn, you can shrink it down to a single point without tearing it.
4. The Result: If every small chunk of your messy ball can be shrunk to a point, and those chunks cover the whole ball in a specific way, then the entire messy ball is topologically the same as the sphere.

The "Principal Congruence" Twist

The authors add a special condition: they only let the explorers walk if they are wearing very specific "uniforms" (called principal congruence subgroups). This is like saying, "We are only looking at the forest when the explorers are wearing red hats."

By restricting the explorers this way, they ensure the "shaky" areas behave nicely. They prove that even in these high-dimensional, tangled forests, if you look at the areas where the "red-hat explorers" get stuck, that entire area is mathematically equivalent to the clean boundary sphere.

Why Does This Matter? (The "Steinberg" Treasure)

Why do we care about squishing a messy knot into a sphere?

In the world of these numbers, there is a very special, valuable object called the Steinberg Module. You can think of this as a "treasure chest" of information about the group of explorers.

• The paper shows that the "messy knot" (the unstable complex) actually is the key to unlocking this treasure chest.
• They prove that the "homology" (the holes and loops) of the messy knot is exactly the same as the "homology" of the treasure chest.
• Furthermore, they show that as you change the "uniforms" (the ideals $I$), the way you unlock the treasure remains consistent. It's like having a master key that works for every version of the lock.

The Big Picture

In simple terms, this paper says:

"Even when the mathematical forest gets incredibly complex and tangled in high dimensions, if you look at the specific places where our number-theory explorers get stuck, that messiness isn't random. It has a hidden, perfect shape underneath it. We found a way to prove that this messy shape is actually the same as a clean, simple sphere, and this connection helps us understand the deep 'treasure' of these number systems."

It's a bit like realizing that a chaotic, swirling storm cloud is actually just a perfect, rotating sphere in disguise, and knowing that helps meteorologists (or in this case, number theorists) predict the weather of the mathematical universe.

Technical Summary

Here is a detailed technical summary of the paper "The Unstable Complex in Bruhat-Tits Buildings for Arithmetic Groups over Function Fields" by Gebhard Böckle and Sriram Chinthalagiri Venkata.

1. Problem Statement and Context

The paper addresses the topological and homological properties of Bruhat-Tits buildings associated with the group $GL_r(K_\infty)$, where $K$ is a function field of positive characteristic and $K_\infty$ is the completion at a fixed place $\infty$.

• Background: For $r=2$, J.-P. Serre established that the $\Gamma$-unstable region of the Bruhat-Tits tree (the subcomplex consisting of simplices with non-trivial stabilizers under an arithmetic subgroup $\Gamma$) is homotopy equivalent to the spherical Tits building (the projective space $P(W)$). This result is foundational for understanding the cohomology of arithmetic groups and the Steinberg module.
• Generalization: D. Grayson extended Serre's result to higher ranks ($r > 2$) but utilized Harder-Narasimhan stability (interpreting vertices as vector bundles) to define the unstable region.
• The Gap: The authors aim to generalize Serre's specific result for congruence subgroups to higher ranks ($r \geq 2$) without relying on the Harder-Narasimhan filtration. Instead, they seek to define the unstable complex directly via the action of principal congruence subgroups $\Gamma_P(I)$ on the building. The challenge is that for $r > 2$, the unstable complex is connected (unlike the disjoint union of trees in the $r=2$ case), requiring a more sophisticated homotopy-theoretic approach to show equivalence to the Tits building.

2. Methodology

The authors employ a blend of combinatorial topology, group theory, and algebraic geometry, adapting the methods of Grayson and Quillen to an equivariant setting.

• Definitions and Setup:

  • Let $W$ be a $K$-vector space of dimension $r$. The Bruhat-Tits building $B_\bullet(W)$ is defined as the simplicial complex of homothety classes of $R$-lattices in $W$ (where $R$ is the discrete valuation ring at $\infty$).
  • The $\Gamma$-unstable complex $B_\bullet(W)_{\Gamma\text{-un}}$ is the subcomplex of simplices whose stabilizer in $\Gamma$ is non-trivial.
  • The Tits building $T_\bullet(W)$ is the spherical building of proper non-zero subspaces of $W$.

• Equivariant Homotopy Theory:

  • The core methodology relies on equivariant Whitehead's Theorem and a specific lemma (an equivariant version of Quillen's Theorem A, adapted from Grayson and W. T. T. [TW91]).
  • The strategy involves constructing a family of subcomplexes $\{B_\bullet(W)_\sigma\}_{\sigma \in T_\bullet(W)}$ indexed by the vertices (and simplices) of the Tits building.
  • Key Technical Step: They prove that for any simplex $\sigma$ in the Tits building, the corresponding subcomplex $B_\bullet(W)_\sigma$ (consisting of lattices stabilized by an element acting trivially on the subspace $\sigma$) is contractible.
  • They further show that the poset of indices $\sigma$ for which a given unstable simplex $\gamma$ belongs to $B_\bullet(W)_\sigma$ is itself contractible under the stabilizer action.

• Vector Bundle Interpretation:

  • To handle the arithmetic groups, the authors utilize the correspondence between $R$-lattices and rank $r$ vector bundles on the curve $C$ (with fixed behavior at $\infty$). This allows them to explicitly describe the stabilizers of simplices using cohomology of sheaves (specifically $H^0(C, E^\vee \otimes E(I))$).
  • They prove that for principal congruence subgroups $\Gamma_P(I)$, the stabilizers of simplices are finite $p$-groups, which is crucial for the homological arguments.

3. Key Contributions

1. Direct Equivalence for Congruence Subgroups: The paper establishes a natural $\Gamma_P$-equivariant homotopy equivalence between the $\Gamma$-unstable complex $B_\bullet(W)_{\Gamma\text{-un}}$ and the spherical Tits building $T_\bullet(W)$ for principal congruence subgroups $\Gamma = \Gamma_P(I)$ in any rank $r \geq 2$.

   • Significance: Unlike Grayson's approach which used Harder-Narasimhan stability, this result is intrinsic to the action of congruence subgroups and recovers the "boundary" intuition of Serre's rank 2 case directly.

2. Resolution of the Steinberg Module:

   • The authors construct a chain complex of $\Gamma$-stable chains, $C_\bullet(B_\bullet(W))_{\Gamma\text{-st}}$, defined as the cokernel of the inclusion of the unstable complex into the full building.
   • They prove that the top homology of this complex, denoted $St_I(P)$, is isomorphic to the Steinberg module $St(W)$ (the top homology of the Tits building).
   • Crucially, they show that $St(W)$ is a finitely generated projective $\mathbb{Z}[\Gamma]$-module.

3. Compatibility of Isomorphisms:

   • The paper demonstrates that the isomorphisms between the homology of the stable complexes and the Steinberg module are compatible as the ideal $I$ varies (i.e., as one moves to smaller congruence subgroups). This provides a coherent system of resolutions for the Steinberg module across the tower of congruence subgroups.

4. Main Results

• Theorem 4.8 (Main Theorem): There exists a natural $\Gamma_P$-equivariant homotopy equivalence between the geometric realization of the unstable complex $|B_\bullet(W)_{\Gamma\text{-un}}|$ and the geometric realization of the Tits building $|T_\bullet(W)|$.

  • Note: The authors clarify that while the spaces are homotopy equivalent, the inverse map is not necessarily $\Gamma_P$-equivariant in the strict sense (unlike the $r=2$ case where the unstable complex is a disjoint union of trees), but the equivalence holds in the homotopy category of $G$-spaces.

• Theorem 5.8: For ideals $J \subset I$, there is a canonical $\Gamma_P$-equivariant homomorphism $i_{J,I}: St_J(P) \to St_I(P)$ that commutes with the isomorphisms to the Steinberg module $St(W)$. This establishes a directed system of resolutions.

• Corollary 5.7: The Steinberg module $St(W)$ is a finitely generated projective module over the group ring $\mathbb{Z}[\Gamma]$ for any principal congruence subgroup $\Gamma$.

5. Significance and Future Directions

• Theoretical Impact: This work bridges the gap between the combinatorial geometry of Bruhat-Tits buildings and the representation theory of arithmetic groups in higher ranks. It provides a robust framework for studying the Steinberg module without relying on the specific geometry of vector bundles (Harder-Narasimhan), making the results more accessible to purely group-theoretic or combinatorial approaches.
• Cohomology of Arithmetic Groups: By establishing that the Steinberg module is a projective $\mathbb{Z}[\Gamma]$-module, the paper provides a powerful tool for computing the cohomology of arithmetic groups over function fields.
• Harmonic Cohomchains: The authors indicate that these results are a stepping stone toward understanding the relationship between harmonic cochains on the Bruhat-Tits building and the Steinberg module in higher ranks. Specifically, they aim to generalize Teitelbaum's results (from rank 2) regarding the extension of harmonic cochains from the stable part to the full building.
• Modular Symbols: The work connects the geometric cycles of stable chambers to modular symbols (Ash-Rudolph), offering a potential geometric interpretation of these symbols in higher rank settings.

In summary, the paper successfully generalizes Serre's classical results on the unstable complex of trees to higher-rank buildings for congruence subgroups, providing a new, equivariant homotopy equivalence and a refined understanding of the Steinberg module's structure in the context of function fields.