On the isotopy classes of embeddings of surfaces in 5-manifolds

This paper establishes that two homotopic smooth embeddings of a closed surface into a closed oriented 5-manifold are isotopic if they share a common algebraic dual 3-sphere or if the ambient manifold is simply connected, by introducing a new invariant defined via the manifold's homotopy groups that classifies such isotopy classes.

The Gist

Imagine you are an artist working in a very strange, five-dimensional art gallery (a 5-manifold). Your job is to place a flat, closed piece of fabric (a surface, like a donut or a sphere) into this gallery.

You have two pieces of fabric, let's call them Fabric A and Fabric B. You know they are "homotopic," which means you can wiggle Fabric A around in the air until it looks exactly like Fabric B without tearing or cutting it. In the world of math, this is like saying they are the same shape in a loose, flexible sense.

The Big Question:
If you can wiggle Fabric A into Fabric B, can you do it without the fabric ever passing through itself? In other words, can you slide them past each other smoothly (an "isotopy") without any knots or self-intersections?

In lower dimensions (like 2D or 3D), the answer is often "yes, if they look the same, you can slide them." But in higher dimensions, things get tricky. Sometimes, two fabrics can look identical and be transformable into one another, yet they are "knotted" in a way that prevents them from sliding past each other smoothly.

The Paper's Discovery

This paper by Ruoyu Qiao solves a specific puzzle: When can we be sure that two surfaces in a 5D gallery are actually the same "slide" (isotopic), even if we only know they are the same "shape" (homotopic)?

The author builds a special mathematical "detector" (an invariant) to check this.

The "Self-Intersection Detector"

Imagine you try to wiggle Fabric A into Fabric B. As you move it, the fabric might accidentally bump into itself.

• The Detector: The author creates a scorecard. Every time the fabric bumps into itself during the transformation, the detector records a "point."
• The Score: It doesn't just count the points; it records where the bump happened relative to the loops and tunnels in the 5D gallery. It's like a GPS log of every collision.
• The Rule: If you can wiggle the fabric into the new shape without any collisions, the score is Zero. If the score is Zero, it means the transformation was a smooth slide (an isotopy). If the score is not Zero, the fabric is "stuck" in a different class.

The "Magic Keys" (When the Answer is Always "Yes")

The paper finds two special situations where the detector always reads Zero, meaning you can always slide the fabric smoothly.

1. The "Ghost Sphere" (Algebraic Dual):
   Imagine the gallery has a hidden, invisible 3D sphere floating around that acts like a "ghost" or a "key." If your fabric has a "partner" sphere that intersects it exactly once (like a key fitting into a lock), the geometry of the 5D space forces the fabric to be unknottable. It's like having a magic wand that untangles any knot instantly.

   • Analogy: If you are trying to untie a knot in a string, but you have a second string that crosses it exactly once, the laws of physics in this 5D world say the knot must be untieable.

2. The "Empty Room" (Simply Connected):
   If the 5D gallery has no holes, loops, or tunnels (mathematically, it's "simply connected"), then there are no places for the fabric to get tangled around.

   • Analogy: If you are trying to move a rug in a room with no pillars or furniture, you can slide it anywhere without it getting caught.

The "Infinite Knots" (When the Answer is "No")

The paper also shows that if the gallery is full of complex loops and tunnels (specifically, if the fundamental group is complicated), you can have infinitely many different ways to place the fabric that look the same but cannot be slid into each other.

• Analogy: Imagine a room filled with an infinite number of pillars. You can wrap your fabric around these pillars in infinitely many different patterns. Even if you can wiggle the fabric to look like another pattern, you might need to cut it to get from one pattern to the other because the pillars are in the way.

Summary in Plain English

• The Problem: In 5D space, can you always slide one surface into another if they look the same?
• The Solution: Not always. But we can build a "collision counter" to tell the difference.
• The Good News: If the space is simple (no holes) or if the surface has a special "partner" sphere, the answer is YES, you can always slide them.
• The Bad News: If the space is complex, you might have infinite versions of the same surface that are stuck in different "knots" and can never be slid into each other.

This work generalizes previous discoveries about spheres in 4D and 5D, providing a complete rulebook for understanding how surfaces behave in these high-dimensional worlds.

Technical Summary

Here is a detailed technical summary of the paper "On the Isotopy Classes of Embeddings of Surfaces in 5-Manifolds" by Ruoyu Qiao.

1. Problem Statement

The paper addresses the classical topological problem of distinguishing between homotopy and isotopy for smooth embeddings of closed oriented surfaces ($\Sigma$) into closed oriented 5-manifolds ($N$).

Specifically, given two homotopic embeddings $f, f' : \Sigma \hookrightarrow N$, the author seeks to determine the conditions under which $f$ and $f'$ are isotopic (can be deformed into one another through a family of embeddings). While the relationship is well-understood for spheres in 4-manifolds (Gabai's Light Bulb Theorem) and 2-spheres in 5-manifolds (Kosanovi´c–Schneiderman–Teichner), this paper extends the classification to surfaces of arbitrary genus in 5-manifolds, particularly in cases where the fundamental group $\pi_1(N)$ is non-trivial and no geometric dual sphere exists.

2. Methodology

The core methodology involves constructing a new algebraic invariant that classifies isotopy classes within a fixed homotopy class. The approach relies on the following steps:

A. Intersection Number of Homotopies

The author defines an intersection number invariant, $\mu$, for homotopies between embeddings.

• Context: Unlike Wall's intersection number (which applies to regular homotopy of immersions), this invariant is defined for homotopies between embeddings.
• Construction: For a generic homotopy $H: \Sigma \times I \to N \times I$, the invariant counts self-intersection points of the trace of the homotopy.
• Algebraic Structure: The value of the invariant lies in a quotient of the free $\mathbb{Z}$-module generated by the set of double cosets $G \backslash \pi_1(N) / G$, where $G = f_*(\pi_1(\Sigma))$.
• Relation to Isotopy: A homotopy $H$ is homotopic (relative to the boundary) to an isotopy if and only if its intersection number $\mu([H])$ vanishes in the target group.

B. Handling Self-Homotopies (The Affine Action)

A major technical challenge is that the intersection number depends on the choice of homotopy between $f$ and $f'$. If $H$ is a homotopy from $f$ to $f'$, and $G$ is a self-homotopy of $f$ (a loop in the space of embeddings), then $G \cdot H$ is another homotopy from $f$ to $f'$.

• The author analyzes the group of self-homotopies $H^f_0$ and decomposes it based on the CW-decomposition of the surface $\Sigma$ (0-skeleton, 1-skeleton, 2-skeleton).
• This decomposition reveals that the group of self-homotopies acts on the space of homotopies and the target invariant group.
• Key Insight: The action of self-homotopies on the invariant is affine, not linear. Specifically, the action involves conjugation by elements of $\pi_1(N)$ and a translation term derived from the intersection number of the self-homotopy itself.

C. The Invariant Map

By quotienting out the action of self-homotopies, the author constructs a well-defined map:
$$f_q: p^{-1}([f]) \to A_f$$
where:

• $p^{-1}([f])$ is the set of isotopy classes of embeddings homotopic to $f$.
• $A_f$ is a specific quotient group derived from the fundamental group of $N$ and the homotopy groups of $N$ (specifically $\pi_2(N)$ and $\pi_3(N)$).

3. Key Contributions

1. Generalization of KST23: The paper generalizes the classification results of Kosanovi´c, Schneiderman, and Teichner (who treated 2-spheres) to surfaces of any genus $g \geq 0$.
2. New Invariant Construction: It introduces a complete invariant for isotopy classes of surfaces in 5-manifolds within a homotopy class. This invariant is defined purely in terms of the homotopy groups of the ambient manifold ($\pi_1, \pi_2, \pi_3$) and the image of the fundamental group of the surface.
3. Resolution of the "Dual Sphere" Condition: The paper clarifies the role of algebraic dual spheres. It proves that if an algebraic dual sphere exists, the invariant group $A_f$ becomes trivial, forcing all homotopic embeddings to be isotopic.
4. Counter-Examples: The paper establishes conditions under which there are infinitely many non-isotopic embeddings within a single homotopy class, specifically when $\pi_2(N)$ and $\pi_3(N)$ are trivial but the double coset space $G \backslash \pi_1(N) / G$ has infinitely many conjugacy classes.

4. Main Results

Theorem 1.1 (Classification Theorem)

There exists a bijection between the set of isotopy classes of embeddings homotopic to $f$ and a specific algebraic quotient group $A_f$:
$$p^{-1}([f]) \cong A_f$$
The map is induced by the intersection number of homotopies.

Corollary 1.3 (Triviality Conditions)

If any of the following conditions hold, then any embedding $f'$ homotopic to $f$ is isotopic to $f$:

1. $N$ is simply connected ($\pi_1(N) = 0$).
2. $f$ admits an algebraic dual sphere (an element in $\pi_3(N)$ with algebraic intersection number 1 with $f$).
3. The induced homomorphism $f_*: \pi_1(\Sigma) \to \pi_1(N)$ is surjective.

Corollary 1.4 (Non-Uniqueness)

If $\pi_2(N) = 0$ and $\pi_3(N) = 0$, and the set of double cosets $G \backslash \pi_1(N) / G$ contains infinitely many conjugacy classes under the action of $G$, then there exist infinitely many embeddings homotopic to $f$ that are pairwise non-isotopic.

5. Significance

• Theoretical Advancement: This work bridges the gap between 4-dimensional and 5-dimensional embedding theory. It demonstrates that while the "Light Bulb Theorem" logic (dual spheres implying isotopy) holds in 5-manifolds for surfaces, the absence of such duals leads to a rich structure of isotopy classes governed by the fundamental group.
• Independence of Geometric Duals: Unlike previous results that relied heavily on the existence of geometric dual spheres to perform isotopies, this paper provides a classification that works even in the absence of such spheres, using algebraic obstructions instead.
• Algebraic Topology Application: The paper provides a concrete application of homotopy groups ($\pi_1, \pi_2, \pi_3$) and group actions (conjugation, double cosets) to solve a geometric topology problem, offering a template for analyzing embeddings in higher dimensions.

In summary, Qiao establishes that the isotopy classification of surfaces in 5-manifolds is entirely determined by an algebraic invariant derived from the fundamental group and higher homotopy groups of the ambient space, generalizing previous sphere-specific results and providing a complete dichotomy between uniqueness and infinite multiplicity of isotopy classes.