Flowing from intersection product to cup product

Greg Friedman, Anibal M. Medina-Mardones, Dev Sinha

This paper, "Flowing from intersection product to cup product," provides a long-sought-after bridge between two fundamental ways of multiplying "shapes" in topology: the geometric intersection of manifolds and the algebraic cup product of cochains.

TL;DR

The authors use a "logistic" vector field flow to reconcile geometric intersection theory with combinatorial cubical cohomology. They prove that by flowing submanifolds along this field, their physical intersections eventually align perfectly with the standard algebraic cup product at the granular cochain level.

What Problem Does This Paper Address?

In algebraic topology, there are two primary ways to think about the "product" of two cohomology classes (essentially, ways to multiply topological data):

  1. Geometric Intersection: If you have two submanifolds, you can look at where they intersect. This is intuitive but mathematically "messy"—it’s only partially defined (the manifolds must be in "general position" or transverse) and depends on smooth geometry.
  2. The Cup Product: This is a purely algebraic operation used in discrete settings (like a grid of cubes or triangles). It is always defined, but it loses the immediate geometric intuition of "overlapping shapes" and is famously non-commutative at the cochain level.

While mathematicians have known these two approaches are equivalent at a high level (homology), they lacked an explicit, step-by-step comparison at the "cochain level"—the most detailed level of the theory. This paper fills that gap by showing how to "flow" the geometric version until it snaps into place with the algebraic one.

Methods

The authors employ several key mathematical structures:

  • Cubulation: They represent a manifold $M$ as a collection of $n$-dimensional cubes (a cubical complex).
  • Geometric Cochains: They use manifolds with corners to represent cohomology classes. This "analog" representation allows for a more flexible geometric treatment than standard singular simplices.
  • The Logistic Vector Field: This is the paper's primary innovation. On a standard unit interval $[0, 1]$, the logistic field is defined by the equation $x(1-x)$. This field has "zeros" at the boundaries (0 and 1). When extended to an $n$-dimensional cube, it creates a flow that pushes points toward the "terminal" faces of the cube.
  • The Comparison Map ($\mathcal{I}$): A map that converts a geometric manifold into a discrete cubical cochain by counting how many times the manifold intersects the various faces of the cubulation.

The core of the method involves taking two manifolds, $W$ and $V$, and flowing $W$ forward in time ($f_t$) and $V$ backward in time ($f_{-t}$) along this logistic field.

Datasets

Not discussed in the paper. This is a work of theoretical mathematics; the results are proven via rigorous logical deduction and differential geometry rather than through empirical data or benchmarks.

Key Insights and Ideas

  • The "Analog-to-Digital" Bridge: The logistic flow acts as a synchronizer. It "squashes" the tangent spaces of the geometric manifolds in a way that forces their intersections to occur only at the specific locations where the cubical cup product expects them.
  • Reciprocity and the Serre Diagonal: The cup product in cubical topology relies on something called the Serre diagonal. The authors discovered that the logistic flow naturally respects "reciprocal" faces—pairs of faces where one is "initial" and the other is "terminal." This mirrors the algebraic requirements of the cup product perfectly.
  • Global Smoothness: Unlike previous attempts to relate these fields which used non-continuous vector fields, the logistic flow is globally smooth. This makes it a much more robust tool for connecting physical models in geometry to discrete models in topology.

Key Results

  • The Flow Comparison Theorem (Theorem 1.1/5.1): This is the main result. It states that for a sufficiently large time $t$, the intersection of the flowed manifolds $f_t(W)$ and $f_{-t}(V)$ is exactly equal to the cubical cup product of their associated cochains: $$\mathcal{I}(f_t(W) \cap f_{-t}(V)) = \mathcal{I}(W) \smile \mathcal{I}(V)$$
  • Quasi-isomorphism: The authors prove that their intersection map $\mathcal{I}$ is a quasi-isomorphism, meaning it preserves the underlying topological structure while moving between the geometric and cubical worlds.
  • Explicit Cochain Comparison: This is the first result to provide an explicit, cochain-level formula relating these two fundamental products.

Limitations and Future Work

  • Cubical Limitation: The current proof is tailored to cubical complexes. While the authors suggest a version for simplicial complexes (triangles/tetrahedra) is possible, it is not explicitly developed here.
  • Higher Structures: The authors aim to extend this "flow" idea to higher-order operations, such as "cup-$i$" products and $E_\infty$ structures. These are essential for capturing more subtle topological invariants that the standard cup product misses.
  • Applications: The authors suggest this framework could be used to define new knot invariants or to better understand configuration spaces in physics and robotics.

Summary updated

Abstract

We use a vector field flow defined through a cubulation of a closed manifold to reconcile the partially defined commutative product on geometric cochains with the standard cup product on cubical cochains, which is fully defined and commutative only up to coherent homotopies. The interplay between intersection and cup product dates back to the beginnings of homology theory, but, to our knowledge, this result is the first to give an explicit cochain level comparison between these approaches.
BibTeX
@article{friedman2024flowing,
  author = {Greg Friedman and Anibal M. Medina-Mardones and Dev Sinha},
  title = {{Flowing from intersection product to cup product}},
  journal = {Journal of Topology and Analysis},
  year = {2024},
  doi = {10.1142/s1793525324500377},
}
Figure