Unknotting genus one knots

Alexander Coward, Marc Lackenby

This paper provides a rare "finiteness" result in the complex world of knot theory. The authors prove that for any knot with genus one and unknotting number one, there is exactly one way to unknot it with a single crossing change—unless it is the figure-eight knot, which has exactly two.

TL;DR

While it is notoriously difficult to determine how many ways a knot can be "untied" with a single move, this paper proves that for the simplest class of knots (genus one), the answer is almost always "exactly one." The figure-eight knot is the lone exception, admitting two distinct unknotting procedures.

What Problem Does This Paper Address?

In knot theory, the unknotting number is the minimum number of times you must pass one strand of a knot through another (a "crossing change") to turn it into a simple, unknotted circle. While this sounds straightforward, calculating this number is a major challenge in topology.

Even when we know a knot can be unknotted in a single move (unknotting number one), a deeper question remains: How many different ways can you do it? Are there infinitely many pairs of strands you could swap to reach the unknot, or just a few? This paper tackles this question for genus one knots—knots that can be "skinned" by a surface with exactly one hole (like a torus with a boundary).

Methods

The authors combine classical geometric topology with combinatorial tools:

  • Sutured Manifold Theory: They use existing theorems (by Scharlemann and Thompson) to translate the physical act of a crossing change into a topological operation called "plumbing." Essentially, unknotting a knot is equivalent to decomposing its associated surface (the Seifert surface) into simpler pieces, one of which must be a "Hopf band."
  • Product Discs: They establish a one-to-one correspondence between these surface decompositions and specific geometric objects called "clean alternating directed product discs."
  • The Arc Complex: For "fibred" knots (like the trefoil and figure-eight), the problem becomes a study of how the knot's internal "twist" (monodromy) acts on the arcs of a punctured torus. The authors use the Farey graph—a beautiful geometric representation of rational numbers—to track these arcs and prove that only one or two can satisfy the unknotting conditions.

Datasets

As a work of pure mathematics, this paper does not use external datasets or benchmarks. It relies on logical proof and the classification of knots.

Key Insights and Ideas

  • The "Doubled Knot" Connection: The authors leverage the fact that every genus one knot with unknotting number one is a "doubled knot." This provides a specific geometric template to work from.
  • Equivalence is Key: A major part of the work is defining what it means for two unknotting moves to be "the same." They define equivalence based on ambient isotopy—if you can slide one crossing change site to another without cutting the knot, they are considered the same.
  • The Figure-Eight Exception: The most striking insight is why the figure-eight knot is different. Most knots are "non-fibred" and have very rigid structures that allow only one unknotting site. The figure-eight knot is "fibred" with a specific type of internal twist (pseudo-Anosov monodromy) that creates exactly two distinct, non-equivalent sites where a crossing change works.
  • The Trefoil’s Illusion: The trefoil knot (the simplest non-trivial knot) appears to have three different unknotting sites, but the authors show that because of the trefoil's symmetry, all three are actually equivalent.

Key Results

  • Theorem 1.1 (The Main Result): For a genus one, unknotting number one knot $K$:
  • If $K$ is not the figure-eight knot, there is precisely one crossing change that unknots it (up to equivalence).
  • If $K$ is the figure-eight knot, there are precisely two.
  • Theorem 1.2: The authors provide explicit diagrams (Figure 3 in the paper) showing exactly where these crossing circles are located for doubled knots.
  • Conjecture 1.3: Based on their findings, the authors propose that this finiteness isn't just a quirk of genus one knots, but a universal rule: For any knot $K$, there are only finitely many ways to unknot it in one move.

Limitations and Future Work

The primary limitation is that the proof is specific to genus one knots. While the authors suspect the result holds for higher genus knots (Conjecture 1.3), the complexity of the "arc complex" grows significantly as the genus increases.

The authors also note a reason for caution: in the related field of 3-manifold surgery, there are known examples (by Osoinach) where a single manifold can be represented as surgery on infinitely many different knots. This suggests that while knots themselves might behave "finitely," the spaces they live in can be much more deceptive. Future work would involve extending these "plumbing" and "arc complex" techniques to knots of genus two and beyond.

Summary updated

Abstract

For any knot with genus one and unknotting number one, other than the figure-eight knot, we prove that there is exactly one way to unknot it by means of a crossing change. In the case of the figure-eight knot, we prove that there are precisely two unknotting crossing changes. The proof uses sutured manifold theory and an analysis of the arc complex of the once-punctured torus.
BibTeX
@inproceedings{coward2011unknotting,
  author = {Alexander Coward and Marc Lackenby},
  title = {{Unknotting genus one knots}},
  booktitle = {Commentarii},
  year = {2011},
  doi = {10.4171/cmh/227},
}

Loading related content...

Figure