The Conway knot has infinite concordance order
Chiara Donatone, Marc Kegel, Lukas Lewark, Paula Truöl · 2026
Read The Conway knot has infinite concordance order by Chiara Donatone, Marc Kegel, Lukas Lewark, Paula Truöl (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2606-30422.txt.
Introduction
Chiara Donatone, Marc Kegel, Lukas Lewark, and Paula Truöl settle a famous knot's status in the concordance group. Knots are concordant when their joined difference bounds a smooth disk. The concordance order counts how many copies must join to become slice. Infinite order means no number of copies ever works. The Conway knot was already known not to be slice.
Piccirillo's celebrated proof showed the Conway knot is not slice. Every additive invariant vanishes on it, blocking the usual route onward. Infinite order needs a different technique entirely. The authors combine three tools to supply it. The Rasmussen invariant provides the key measurement. Satellite operations and null-homologous twists move it in controlled ways. The twist inequality lets the invariant survive those moves. Together they force the concordance order to be infinite. The proof also showcases the method for other knots.
Catalog record
| CREATOR | Chiara Donatone, Marc Kegel, Lukas Lewark, Paula Truöl |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 33 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2606.30422
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
The Conway knot has infinite concordance order Chiara Donatone, Marc Kegel, Lukas Lewark, Paula Truöl Abstract . We examine how the Rasmussen invariant, satellite operations, and null-homologous twists can be used to establish infinite order of knots in the smooth concordance group. As an application, we show that the Conway knot has infinite concordance order. 1. Introduction A knot in the 3 -sphere S^{3} is slice if it bounds a smoothly and properly embedded disk in B^{4} , the 4 -ball bounded by S^{3} . Two knots K and J are concordant if the connected sum K\#{-J} is slice. Here -J is the inverse of J , which is the image of J under an orientation-reversing diffeomorphism of S^{3} with the opposite orientation on the knot. Knots up to concordance form an abelian group, the concordance group \mathcal{C} . The group operation is induced by connected sum, slice knots represent the identity element, and the additive inverse of a concordance class [K] is [-K] . While the study of the group \mathcal{C} has advanced in recent years, particularly since the advent of Heegaard Floer and Khovanov homology theories, its structure—let alone its isomorphism type—remains mysterious. In particular, the structure of the torsion subgroup of \mathcal{C} is an open problem (see e. g. [ 4 , Problem 1.38] ). The concordance order of a knot K is its order in \mathcal{C} , i. e. the least positive integer n for which the connected sum K^{\#n} of n copies of K is slice, or infinity if no such n exists. Theorem 1 . The Conway knot has infinite concordance order. It had previously been known that the Conway knot C is not slice, as demonstrated in the celebrated work of Piccirillo [ 39 ] . Piccirillo’s proof uses a clever construction involving finding a knot K^{\prime} with the same 0 -trace as C . According to the trace embedding lemma (see Section 2 ), this gives K^{\prime} the same sliceness status as C . However, computing Rasmussen’s s invariant [ 42 ] for K^{\prime} shows that K^{\prime} is not slice. One difficulty in determining the non-sliceness of the Conway knot was that all known sliceness invariants vanished for this knot; in particular, all known additive invariants, which could otherwise have been used to prove Theorem 1 , vanish on C . For this reason, we use an additional technique called twisting . Definition 2 . For knots J,K\subset S^{3} , we say that K is obtained from J by a right-handed null-homologous twist , or simply, a twist , if there is an unknot \alpha\subset S^{3}\setminus J with \lk(\alpha,J)=0 such that (-1) -surgery on S^{3} along \alpha turns J into K . We say that a real-valued knot invariant y satisfies the twist inequality if y(K)\leq y(J) whenever K is obtained from J by a twist. For example, a positive-to-negative crossing change is a special type of twist. (This is why twisting is also called a positive-to-negative generalised crossing change , or adding a generalised negative crossing [ 8 ] .) Alongside many other knot concordance invariants, the Rasmussen s invariant satisfies the twist inequality (see Section 2 ). Another key element in our proof of Theorem 1 is the behaviour of satellite operations under twists. Recall that a pattern P is a knot in the standard solid torus S^{1}\times D^{2} , and P(K) denotes the satellite knot with pattern P and companion K ; see, for example, [ 45 , Section 4D] . Theorem 3 . Let J,K\subset S^{3} be two knots that can be transformed into the unknot U via a finite sequence of concordances and twists. If a knot concordance invariant y exists that satisfies the twist inequality and a pattern P such that {y(P(K))\neq y(P(U))} , then K\#J is not slice. In particular, K has infinite concordance order. To prove Theorem 1 using Theorem 3 , we must first find a suitable pattern P . Building on Piccirillo’s work [ 38 , 39 ] , we construct for all knots K with unknotting number 1 and unknotting crossing c a pattern P_{K,c} such that the satellite knot P_{K,c}(J) has the same 0 -trace as K\#J for every knot J . By the trace embedding lemma, the sliceness statuses of K\#J and P_{K,c}(J) are identical. Experiments suggest that P_{K,c}(U)\neq K unless K is a twist knot or a twisted Whitehead double (see, for example, [ 18 , Section 5] ). We are particularly interested in the pattern P_{K,c} when it additionally satisfies s(P_{K,c}(U))\neq 0 . Proposition 4 . Let K be a knot with unknotting number 1 and unknotting crossing c . If s(P_{K,c}(U))\neq 0 , then K has infinite concordance order. According to KnotInfo [ 26 ] , the only prime knot with at most 11 crossings for which the concordance order is currently unknown is the Conway knot C=K11n34 . Theorem 1 thus concludes the computation of concordance orders for knots with at most 11 crossings. However, the following question remains open. Question 5 . Does the concordance class [C] of the Conway knot C=K11n34 generate a \mathbb{Z} -summand of \mathcal{C} ? The infinite concordance order of a knot is often demonstrated using knot concordance homomorphisms that do not vanish for the knot in question. In fact, the concordance class [K] of a given knot K generates a \mathbb{Z} -summand of \mathcal{C} if and only if there is a knot concordance homomorphism to \mathbb{Z} that takes the value \pm 1 on K . We do not know if the Conway knot admits such a homomorphism. While finishing this paper, the authors learned that Golla–Pinzón-Caicedo are preparing a paper that answers 5 positively, using different methods. Organisation We prove Theorem 3 in Section 2 . Section 3 is devoted to constructing and defining suitable patterns P_{K,c} , and proving 4 and Theorem 1 . Section 4 describes our experiments with knots of low crossing number. Acknowledgements We thank Peter Feller, Lisa Piccirillo, and Claudius Zibrowius for helpful remarks, Marco Golla for comments on a first draft, and Livio Ferretti for useful discussions on related topics. Part of this work was carried out when CD and MK visited LL at ETH Zürich. We thank ETH for the hospitality. CD and PT would like to thank the Max Planck Institute for Mathematics in Bonn for the hospitality during the initial stages of this project. MK is supported by a Ramón y Cajal grant (RYC2023-043251-I) and PID2024-157173NB-I00 funded by MCIN/AEI/10.13039/501100011033, by ESF+, and by FEDER, EU; and by a VII Plan Propio de Investigación y Transferencia (SOL2025-36103) of the University of Sevilla. PT acknowledges the support of the Swiss National Science Foundation Postdoc.Mobility fellowship 230329, and thanks the University of Glasgow for its hospitality and support. 2. Twists, monotonicity, and non-sliceness Unless stated otherwise, all manifolds are assumed to be oriented, compact, and smooth throughout. We consider manifolds up to orientation-preserving diffeomorphism and links up to isotopy. Let us begin with the definition of a relation on the concordance group \mathcal{C} . Similar relations have been considered e. g. in [ 6 , 7 , 8 ] . Definition 6 . For concordance classes of knots J,K\subset S^{3} , we write