🔢 mathematics

Alexander Torsion and Floer Comparisons in the BV-BFV Formalism

This paper investigates the relationship between relative Abelian BF torsion and knot Floer homology within the BV-BFV formalism, establishing connections to the Alexander polynomial and analyzing specific algebraic obstructions, though it concludes that a full geometric equivalence between the two frameworks remains unproven.

Original authors: Nima Moshayedi

Published 2026-10-23

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Nima Moshayedi

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

In the vast landscape of modern mathematics, there is a persistent desire to connect two very different ways of looking at the same shapes. On one side, there are knots, the tangled loops of string that mathematicians have studied for centuries, trying to understand their hidden complexity. On the other side, there are fields of physics, specifically theories that describe how particles and forces behave in space and time. For a long time, these two worlds have spoken different languages. One uses tools from geometry and topology to count and classify knots, while the other uses equations of motion and quantum mechanics to describe physical systems. The challenge has been to find a bridge between them, a way to show that the mathematical patterns found in a knotted string are the same as the patterns found in a physical field theory. This is not just an academic exercise; it is an attempt to see if the deep structures of the universe can be described by a single, unified set of rules, whether we are looking at a piece of string or a quantum particle.

A recent paper by Nima Moshayedi takes a significant step toward building this bridge, but it also draws a very clear line in the sand about what is possible and what is not. The author focuses on a specific type of knot, the kind that lives in a three-dimensional sphere, and compares two powerful methods used to study them. The first method comes from a field called Floer homology, which is a sophisticated way of counting the ways a knot can be twisted and turned, producing a rich, multi-layered map of information. The second method comes from a theoretical physics framework known as the BV–BFV formalism, which is used to describe how physical systems behave at their boundaries. The goal was to see if the simple, numerical output of the physics method could perfectly reconstruct the complex, multi-layered map of the knot method.

The researchers found that the physics method does indeed capture a crucial piece of the knot's identity. By carefully setting up the boundaries of their physical model—specifically by choosing a particular type of boundary condition that mimics a loop around the knot—they were able to extract a mathematical object known as the Alexander polynomial. This polynomial is a famous invariant in knot theory, a single algebraic expression that changes depending on the knot's shape. The paper demonstrates that the physics model produces this exact polynomial, confirming that the two approaches agree on this fundamental level. Furthermore, the study shows that if you evaluate this polynomial at specific, carefully chosen points related to prime numbers, you can actually recover the entire structure of the knot's complexity, provided the knot is not too complicated. This means that for a certain class of knots, a single numerical value from the physics model is enough to rebuild the full mathematical picture.

However, the paper also delivers a sobering conclusion about the limits of this connection. While the physics model can reproduce the overall "shadow" or summary of the knot's complexity, it cannot recreate the full, detailed map on its own. The researchers proved that the simple, scalar values produced by the physics theory are not enough to distinguish between certain knots that are actually different. There are knots that look identical to the physics model but are fundamentally distinct in the more detailed mathematical view. The paper explicitly rules out the idea that a simple, one-dimensional version of the physical theory can fully replace the complex, multi-layered Floer homology. It shows that the physics model, in its current form, loses too much information to be a complete substitute.

To understand why this happens, the author delves into the algebraic machinery behind the scenes. The paper constructs a detailed algebraic model of the boundary where the knot lives, which acts like a dictionary for translating between the two languages. This dictionary reveals that the physical theory, when reduced to its simplest form, collapses into a single point, losing the intricate internal structure that the knot theory preserves. The researchers show that this loss of structure is not a minor glitch but a fundamental obstruction. Even when they try to fix the model by adding more layers of complexity or changing the way they calculate, the core difference remains. The physical theory, as it stands, cannot generate the specific algebraic relationships needed to match the knot theory perfectly.

The study does offer a way forward, however, by showing that if one is willing to keep the full, complex algebraic structure of the boundary rather than simplifying it, a perfect match becomes possible. By resolving the algebraic equations into a more detailed form, the researchers created a version of the model that retains all the necessary information. This version preserves the ability to glue different pieces of the knot together correctly, a property that the simplified version lost. They also showed that this complex algebraic structure can be realized as a classical physical system in one dimension, providing a concrete, tangible example of how the mathematics works. This suggests that the bridge between the two fields is not broken, but it requires a much more sophisticated and detailed construction than a simple, direct comparison would allow.

Ultimately, this work clarifies the relationship between knot theory and quantum field theory by showing exactly where they align and where they diverge. It confirms that the physics model can successfully predict the most basic numerical invariants of a knot, and it provides a method to recover the full complexity for a specific class of knots. But it also proves that a direct attempt to equate the two will fail, because the physical model, in its simplest form, is too blunt an instrument to capture the full nuance of the knot. The paper leaves the reader with a clear understanding that while the two fields are deeply connected, bridging them requires respecting the full complexity of both, rather than trying to force one into the shape of the other. The path forward lies in building more elaborate models that can carry the necessary weight of information, ensuring that the rich tapestry of knot theory is not lost in the translation to physics.

Technical Summary: Alexander Torsion and Floer Comparisons in the BV–BFV Formalism

Problem Statement
The paper investigates the relationship between two distinct mathematical frameworks for studying knots in $S^3$ : the cellular Abelian BF theory (a topological field theory formalism) and knot Floer homology (a categorified invariant). Specifically, it addresses whether the Alexander polynomial, recoverable from the Reidemeister torsion of the knot exterior, can be identified with the decategorified Euler characteristic of knot Floer homology, and whether a deeper, categorified equivalence exists between the underlying chain complexes or boundary algebras of these theories. The central question is whether the "scalar" data of BF torsion is sufficient to reconstruct Floer invariants, and if not, what algebraic or geometric obstructions prevent such a reconstruction.

Methodology
The author employs a combination of algebraic topology, homological algebra, and the BV–BFV (Batalin–Vilkovisky/Batalin–Fradkin–Vilkovisky) formalism.

Torsion and Boundary Conditions: The study utilizes relative cellular BF theory with a specific "meridional annulus" boundary condition. This setup allows for the direct computation of the Alexander polynomial as a determinant-line invariant (torsion) without the $(t-1)$ factor typically present in absolute torsion.
Arithmetic Reconstruction: The paper analyzes the reconstruction of Alexander polynomial coefficients from a single evaluation of the torsion at a primitive root of unity (cyclotomic evaluation). It utilizes Galois trace formulas and genus bounds to establish conditions under which the entire polynomial is uniquely determined.
Gluing and Decategorification: The paper compares the gluing operations of both theories. On the BF side, it uses linear homotopy pullbacks of cellular cochain complexes. On the Floer side, it uses the box tensor product of bordered Floer modules. It examines whether these operations commute with decategorification (taking Euler characteristics).
Algebraic Obstructions: To test for a full equivalence, the author constructs explicit algebraic obstructions. This involves analyzing the "degree-zero torus algebra" (the bordered Floer algebra for the torus boundary) and comparing it to the quantum BFV state space and observable algebras derived from the BF theory. Tools include derived tensor products, Hochschild homology, and dg resolutions.

Key Contributions and Results

Meridional Relative Torsion and Alexander Polynomial:
The paper establishes that the twisted cochain complex of a knot exterior $X_K$ relative to a meridional annulus $R_-$ is acyclic for non-resonant holonomy. The torsion of this relative complex, when normalized symmetrically, is exactly the Alexander polynomial $\Delta_K(t)$ . This provides a field-theoretic realization of the Alexander polynomial that matches the Alexander-graded Euler characteristic of knot Floer homology.
Cyclotomic Reconstruction and Sharpness:
A significant result is the proof that the entire Alexander polynomial (and thus all Alexander-graded Floer Euler characteristics) can be uniquely reconstructed from a single exact algebraic value of the torsion at a primitive $p$ -th root of unity $\zeta$ , provided the prime $p$ satisfies $p \ge 2g(K) + 1$ (where $g(K)$ is the genus). The paper proves this bound is sharp by constructing fibered knots where a smaller prime fails to distinguish the Alexander polynomial from that of the unknot.
Gluing Compatibility:
The paper demonstrates that the gluing formulas for both theories are compatible at the level of decategorified invariants. The relative BF torsion of a connected sum is the product of the individual torsions, matching the convolution of Floer Euler characteristics. However, this compatibility does not extend to a quasi-isomorphism of the underlying chain complexes.
Obstructions to Equivalence:
The paper identifies several fundamental obstructions to a direct equivalence between the BF and Floer theories:
- Acyclicity Obstruction: The standard twisted relative BF complex is acyclic (cohomology vanishes), whereas the associated graded knot Floer complex has non-trivial homology. Thus, they cannot be quasi-isomorphic in the ordinary derived category.
- Scalar Reduction Obstruction: The paper proves that reducing the bordered Floer algebra to scalar fields (via augmentation or semisimple quotient) destroys the "pairing defect" necessary to reproduce the gluing of modules. Specifically, the derived pairing matrix of the torus algebra cannot be factored into independent scalar tensor products.
- Hochschild Obstruction: In a specific non-resonant quantum BFV sector (characteristic zero, $\hbar$ inverted), the quantum state space and observable algebra contract to a vacuum line. The Hochschild homology of this vacuum algebra ( $HH_0 \cong K$ ) differs from that of the degree-zero torus algebra ( $HH_0 \cong K^2$ ), obstructing a Morita equivalence.
L-Space Knots:
For the specific class of positive L-space knots, the paper shows that the arithmetic reconstruction of the Alexander polynomial, combined with the known "staircase" classification of their Floer complexes, allows for the reconstruction of the full filtered chain homotopy type of the knot Floer complex from a single torsion value.
Algebraic Resolution:
The author constructs a six-generator dg resolution of the torus algebra that preserves the derived category of perfect modules and the derived pairings. This resolution admits a classical one-dimensional BV–BFV realization. However, the paper clarifies that this is an algebraic construction starting from the torus algebra, not a derivation of the algebra from the three-dimensional BF theory.

Significance and Claims
The paper modestly claims to establish a precise "decategorified" comparison between the two theories while rigorously delineating the limits of extending this to a full geometric equivalence.

It confirms that the Alexander polynomial serves as a common "shadow" for both the relative BF torsion and the Floer Euler characteristic.
It demonstrates that while arithmetic data (torsion values) can reconstruct Floer invariants for specific classes of knots (L-space knots) or at the level of Euler characteristics, it is insufficient to recover the full categorified structure for general knots.
It explicitly states that the results do not establish a full geometric BV–BFV/Floer equivalence. The paper argues that a full equivalence would require an independent construction of a boundary category with non-trivial off-diagonal morphisms (modeling Reeb chords) and a sewing map that respects the specific gradings and peripheral parametrizations of the manifolds.
The work serves as a rigorous boundary analysis, showing that simple scalar reductions or standard quantization of the linear BF theory are insufficient to reproduce the rich algebraic structure of bordered Floer theory.

More like this