The Conway polynomials and Self Delta-equivalence of pretzel links
Yasutaka NAKANISHI, Tetsuo SHIBUYA, Tatsuya Tsukamoto · 2026
Read The Conway polynomials and Self Delta-equivalence of pretzel links by Yasutaka NAKANISHI, Tetsuo SHIBUYA, Tatsuya Tsukamoto (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-2604-03698.txt.
Introduction
Yasutaka Nakanishi, Tetsuo Shibuya, and Tatsuya Tsukamoto studied self delta-equivalence of pretzel links, a way of judging when links can be turned into each other by a local move. For two-component links, they worked out the full classification in terms of Conway polynomials, computing the values outright. For links with three or more components, they gave a necessary and sufficient condition for self delta-equivalence.
Catalog record
| CREATOR | Yasutaka NAKANISHI, Tetsuo SHIBUYA, Tatsuya Tsukamoto |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 59808 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.03698
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
The Conway polynomials and Self Delta-equivalence of pretzel links Yasutaka NAKANISHI, Tetsuo SHIBUYA, Tatsuya Tsukamoto Abstract . In this paper, we study the self Δ \Delta -equivalence of pretzel links. If the number of components is 2 2 , then we know the complete invariants in terms of the Conway polynomial for classification. We calculate the values. For pretzel links with μ \mu ( ≥ 3 ) (\geq 3) components, we give a necessary and sufficient condition to be self Δ \Delta -equivalent. 1. Introduction Throughout this paper, a link means a set of tame, ordered, oriented, and simple closed curves in the oriented 3 -space \mathbb{R}^{3} . If the number of components of a link \kappa is \mu , we say that \kappa is a \mu - component link . A 1 -component link is sometimes called a knot, simply. For two links \kappa and \lambda , if there is an ambient isotopy which maps \kappa to \lambda preserving the order and orientation, we say that \kappa is equivalent to \lambda , denoted by \kappa\approx\lambda . For two link diagrams K and L which differ only in one place as illustrated in Figure 1 , a local move between K and L is called a \Delta -move , denoted by K\stackrel{{\scriptstyle\Delta}}{{\leftrightarrow}}L . Then, for two links \kappa and \lambda with diagrams K and L , \kappa and \lambda are said to be transformed into each other by a \Delta -move . Furthermore, for two links \kappa and \lambda , if \kappa can be transformed into \lambda by a finite sequence of \Delta -moves, \kappa and \lambda are said to be \Delta -equivalent , denoted by \kappa\stackrel{{\scriptstyle\Delta}}{{\sim}}\lambda . In the case that all three arcs illustrated in Figure 1 are contained in the same component, the above move is called a self \Delta -move , denoted by K\stackrel{{\scriptstyle{\rm s}\Delta}}{{\leftrightarrow}}L . Then, for two links \kappa and \lambda with diagrams K and L , \kappa and \lambda are said to be transformed into each other by a self \Delta -move [ 13 ] . Furthermore, for two links \kappa and \lambda , if \kappa can be transformed into \lambda by a finite sequence of self \Delta -moves, \kappa and \lambda are said to be self \Delta -equivalent (or \Delta link homotopic [ 8 , 9 ] ), denoted by \kappa\stackrel{{\scriptstyle{\rm s}\Delta}}{{\sim}}\lambda . Figure 1. \Delta -move It is known the following result by Matveev [ 7 ] and by Murakami and the first author [ 6 ] . Proposition 1.1 . A pair of knots ( or links ) are \Delta -equivalent if and only if they have the same number of components and the same linking numbers between the corresponding components. Claim 1.2 . Let two links \kappa and \lambda be self \Delta -equivalent. Then we have : (1) Each corresponding pair of sublinks of \kappa and \lambda are self \Delta -equivalent. (2) Each corresponding pair of 2 -component sublinks of \kappa and \lambda have the same linking numbers. The first author and Y.Ohyama studied the self \Delta -equivalence of links in terms of the coefficients of the Conway polynomial ( [ 8 ] , [ 9 ] ). The Conway polynomial \nabla_{\kappa}(z)=a_{0}(\kappa)+a_{1}(\kappa)z+ a_{2}(\kappa)z^{2}+\cdots+a_{m}(\kappa)z^{m} is an invariant of a link \kappa which is a polynomial in a variable z with integral coefficients a_{0}(\kappa),a_{1}(\kappa),a_{2}(\kappa),\cdots,a_{m}(\kappa) . It is known that if \kappa is a \mu -component link (\mu\geq 2) then a_{0}(\kappa)=\cdots=a_{\mu-2}(\kappa)=0 , and if i and \mu have the same parity then a_{i}(\kappa)=0 (cf. [ 2 ] ). Proposition 1.3 . ( [ 8 , Theorem 2] ) If two \mu -component links \kappa=\kappa_{1}\cup\cdots\cup\kappa_{\mu} and \lambda=\lambda_{1}\cup\cdots\cup\lambda_{\mu} are self \Delta -equivalent, then we have (1) a_{\mu-1}(\kappa)=a_{\mu-1}(\lambda) and (2) a_{\mu+1}(\kappa)-a_{\mu-1}(\kappa)\times\left(\overset{\mu}{\underset{i=1}{\sum}}a_{2}(\kappa_{i})\right)=a_{\mu+1}(\lambda)-a_{\mu-1}(\lambda)\times\left(\overset{\mu}{\underset{i=1}{\sum}}a_{2}(\lambda_{i})\right) . The above is a necessary condition, but not a sufficient condition in general. For instance, consider the following two 3 -component links; the 3 -component trivial link, and the split sum of a 2 -component Hopf link and the unknot. They have the same Conway polynomial \nabla(z)=0 . However they are not self \Delta -equivalent, since they have different linking numbers of components and a (self) \Delta -move does not change the linking numbers of components. But for 2 -component links, the above is also a sufficient condition as follows. Proposition 1.4 . ( [ 9 , Theorem 3] ) Two 2 -component links \kappa=\kappa_{1}\cup\kappa_{2} and \lambda=\lambda_{1}\cup\lambda_{2} are self \Delta -equivalent if and only if (1) a_{1}(\kappa)=a_{1}(\lambda) and (2) a_{3}(\kappa)-a_{1}(\kappa)\times\left(a_{2}(\kappa_{1})+a_{2}(\kappa_{2})\right)=a_{3}(\lambda)-a_{1}(\lambda)\times\left(a_{2}(\lambda_{1})+a_{2}(\lambda_{2})\right). In this paper, we study a necessary and sufficient condition for pretzel links to be self \Delta -equivalent. For a sequence \bm{v}=(k_{1},k_{2},\ldots,k_{u}) of non-zero integers, the unoriented link diagram shown in Figure 2 is denoted by P(k_{1},k_{2},\ldots,k_{u}) or P_{\bm{v}} , which is obtained as follows: First take u pairs of parallel strands, introducing k_{i} half-twists on the i -th pair, with the convention k_{i}>0 for right-handed half-twists and k_{i}<0 for left-handed half-twists. Then, connect each pair of parallel strands with k_{i} and with k_{i+1} by a pair of a top bridge t_{i} and a bottom bridge b_{i} , where i=1 , \dots , u and k_{u+1}=k_{1} . We call \bm{v} a pretzel sequence . Figure 2. a pretzel link diagram P(k_{1},k_{2},\ldots,k_{u}) Claim 1.5 . For a pretzel link \kappa with a diagram P(k_{1},k_{2},\ldots,k_{u}) , the following can be seen. (1) \kappa is a knot if and only if all the parameters k_{i} are odd and u is odd, or exactly one parameter is even. (2) \kappa is a 2 -component link if and only if all the parameters k_{i} are odd and u is even, or exactly two parameters are even. (3) \kappa is a \mu -component link with \mu\geq 3 if and only if exactly \mu parameters are even. If P(k_{1},k_{2},\ldots,k_{u}) is oriented, then there are four types of orientations on each pair of parallel strands as illustrated in the left four of Figure 3 . An enhanced pretzel sequence \widehat{\bm{v}}=(k_{1}\epsilon_{1}, k_{2}\epsilon_{2}, \ldots,k_{u}\epsilon_{u}) is a pretzel sequence with an information such that \epsilon_{i} is \mathrm{s} (resp. \mathrm{r} ) if the i -th pair of parallel strands with k_{i} has anti-parallel (resp. parallel) orientation ( i=1 , 2 , \dots , u ). We also define 0\mathrm{s} , 0\mathrm{r} , \infty\mathrm{s} , and \infty\mathrm{r} as illustrated in the right eight of Figure 3 , which we use in the following sections. Then our diagram is denoted by P(k_{1}\epsilon_{1},k_{2}\epsilon_{2},\ldots,k_{u}\epsilon_{u}) or by P_{\widehat{\bm{v}}} . Note that as we can see in Figure 3 , there are two possibilities to decide the orientation of parallel strands with k_{i}\epsilon_{i} . Also note that if we decide an orientation of a certain top bridge t_{i} , say t_{1} , then the orientations of the diagram is decided by \widehat{\bm{v}} . Here we define the orientation of t_{i} and b_{i} as positive (resp. negative ) if the orientation is from the strands with k_{i} (resp. k_{i+1} ) to the strands with k_{i+1} (resp. k_{i} ) (i=1 , \dots , u) . Then we define an oriented pretzel link diagram P(k_{1}\epsilon_{1},k_{2}\epsilon_{2},\ldots,k_{u}\epsilon_{u}) , if one exists, as the oriented diagram such that the orientation of t_{1} is positive. Proposition 1.6 . There exists an oriented pretzel link diagram P_{\widehat{\bm{v}}}=P(k_{1}\epsilon_{1},k_{2}\epsilon_{2}, \ldots, k_{u}\epsilon_{u}) if and only if the following conditions hold: (1) \epsilon_{i}