Lifting Milnor Invariants for 3-Component Links
Christopher W. Davis, JungHwan Park · 2026
Read Lifting Milnor Invariants for 3-Component Links by Christopher W. Davis, JungHwan Park (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-2605-23086.txt.
Introduction
Davis and Park define integer valued invariants gamma^k for three component links. They prove they hold under concordance and more generally weak cobordism. The invariants lift certain Milnor invariants of three component links. The proof introduces h(L), a three component version of the Kojima-Yamasaki eta invariant, which gives the gamma values. As uses they obtain weak cobordism classes when the chosen component has trivial Alexander polynomial. They also describe knots bounding smoothly placed disks in 4D balls whose outsides have fundamental group Z.
Catalog record
| CREATOR | Christopher W. Davis, JungHwan Park |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 95204 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.23086
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Lifting Milnor Invariants for 3-Component Links Christopher W. Davis, JungHwan Park Abstract . We define a sequence of integer-valued invariants γ k ( L ) \gamma^{k}(L) for a 3 3 -component link L L . We prove that the resulting γ \gamma -invariants are invariant under concordance, and more generally under weak cobordism, and that they lift certain Milnor invariants of 3-component links. To establish this, we introduce an invariant h ( L ) h(L) , a 3 3 -component analogue of the Kojima–Yamasaki η \eta -invariant, and show that it recovers the γ \gamma -invariants. As applications, we obtain a weak-cobordism classification when the distinguished component has trivial Alexander polynomial and characterize knots that bound continuously embedded disks in B 4 B^{4} whose complements have fundamental group ℤ \mathbb{Z} . 1991 Mathematics Subject Classification 57K10 1. Introduction In 1954, John Milnor introduced the \bar{\mu} -invariants, higher-order linking numbers associated to a link L=(L_{1},\ldots,L_{m})\subset S^{3} . They are defined inductively by comparing the lower central series of the link group \pi_{1}(S^{3}\smallsetminus L) with that of the unlink [ 34 ] . Since then, they have been studied extensively. For instance, building on Stallings’ work [ 41 ] , Casson showed that Milnor invariants are invariants of topological locally flat link concordance [ 3 ] . They have also been reinterpreted in a variety of ways; see, e.g., [ 21 , 43 , 40 , 37 , 12 , 27 , 26 , 14 ] , and have since been developed in several different directions; see, e.g., [ 38 , 30 , 5 , 13 , 15 , 6 , 19 , 39 , 42 , 31 , 4 , 16 ] . These invariants generalize the classical linking number. Indeed, the length-two Milnor invariants \bar{\mu}_{L}(ij) coincide with the pairwise linking numbers of the components L_{i} and L_{j} of L . Given an m -component link L and a multi-index I=(i_{1},i_{2},\dots,i_{n}) with each i_{k}\in\{1,\ldots,m\} , the length- n Milnor invariant \bar{\mu}_{L}(I) is defined modulo the greatest common divisor of the lower-length invariants \bar{\mu}_{L}(I^{\prime}) over all proper subindices I^{\prime}\subset I . For example, if L is a two-component link with linking number \bar{\mu}_{L}(12)=1 , then every higher-order Milnor invariant of L is completely indeterminate. In 1985, Tim Cochran [ 10 ] discovered a beautiful method for lifting Milnor invariants of 2 -component links with vanishing linking number, introducing a sequence of integer-valued invariants \bigl(\beta^{1}(L),\beta^{2}(L),\dots\bigr)\in\mathbb{Z}^{\infty} , called the \beta -invariants . These give lifts of certain Milnor invariants in the sense that \beta^{k}(L)\equiv\bar{\mu}_{L}\bigl(1^{2k}22\bigr) modulo the indeterminacy of Milnor invariants. (Here 1^{2k}=11\cdots 1 is the multi-index of length 2k whose every entry is 1 .) In this article, we extend Cochran’s idea to 3 -component links and produce a sequence of invariants \bigl(\gamma^{1}(L),\gamma^{2}(L),\dots\bigr)\in\mathbb{Z}^{\infty} , which we call the \gamma -invariants . These lift Milnor invariants in the sense that \gamma^{k}(L)\equiv\bar{\mu}_{L}\bigl(1^{k}23\bigr), again modulo the indeterminacy of Milnor invariants. To state the main theorem precisely, we first introduce some terminology and notation. Unless otherwise specified, all links under consideration are ordered, oriented 3 -component links L=(L_{1},L_{2},L_{3})\subset S^{3} . We refer to the first component L_{1} as the distinguished component, and we further assume that \operatorname{lk}(L_{1},L_{2})=\operatorname{lk}(L_{1},L_{3})=0 . We define \gamma^{0}(L) to be the linking number of L_{2} and L_{3} , that is, \gamma^{0}(L):=\operatorname{lk}(L_{2},L_{3})=\bar{\mu}_{L}(23). To define the higher invariants \gamma^{k}(L) , we require the notion of derivatives of links , a geometric construction introduced in [ 10 ] . Given a link L , the derivative link D(L) is defined by first finding Seifert surfaces G_{1} and G_{2} for L_{1} and L_{2} that intersect in a knot. The result of pushing this knot off of G_{1} in the positive normal direction is denoted L_{12} . Replacing L_{2} with L_{12} yields a new 3 -component link: D(L)=(L_{1},L_{12},L_{3}). More generally, we define the k -fold derivative by iterating this process, obtaining D^{k}(L)=\bigl(L_{1},L_{1^{k}2},L_{3}\bigr). We then define \gamma^{k}(L):=\bar{\mu}_{D^{k}(L)}(23)=\operatorname{lk}(L_{1^{k}2},L_{3}) and let \gamma(L)=\bigl(\gamma^{0}(L),\gamma^{1}(L),\dots\bigr)\in\mathbb{Z}^{\infty} be the sequence of these \gamma -invariants. We remark that it is important to fix G_{1} throughout this iterative process. Indeed, the precise value of the \gamma -invariants depends on this choice G_{1} , which introduces an indeterminacy. When we want to emphasize this dependence on Seifert surface we will write D(L,G_{1})=(L_{1},L_{12},L_{3},G_{1}) and \gamma^{k}(L,G_{1}) . While we discuss this indeterminacy in more detail in Section 4 , we briefly summarize it here. Let T\colon\mathbb{Z}^{\infty}\to\mathbb{Z}^{\infty} be the right-shift operator , T(a^{0},a^{1},\dots):=(0,a^{0},a^{1},\dots). The sequence \bigl(\gamma^{0}(L),\gamma^{1}(L),\dots\bigr) is an invariant of the concordance class of L modulo the action of T+\mathrm{Id} . Theorem 1.1 . Let L=(L_{1},L_{2},L_{3}) be a 3 -component link with a distinguished component. Then \gamma(L)=\bigl(\gamma^{0}(L),\gamma^{1}(L),\ldots\bigr)\in\mathbb{Z}^{\infty} is a link concordance invariant modulo the action of T+\mathrm{Id} . Moreover, we have \gamma^{k}(L)\equiv\bar{\mu}_{L}(1^{k}23)\pmod{\gcd\left\{\,\bar{\mu}_{L}(1^{i}23)\mid i<k\,\right\}}. We remark that, given two sequences of \gamma -invariants, it is straightforward to determine whether they represent the same element modulo the action of T+\mathrm{Id} ; see Corollary 4.6 . In fact, we will see that if \gamma^{k}(L) is the first nonvanishing \gamma -invariant of L , then \gamma^{k} is an invariant and \gamma^{k+1} is well defined modulo \gamma^{k} , as in the case of Milnor invariants. After making a normalization to account for this indeterminacy, \gamma^{n}\in\mathbb{Z} is well defined for all n>k+1 . Therefore, we do indeed obtain examples that are indistinguishable by Milnor invariants but can be distinguished by the \gamma -invariants; see Section 2.1 . Cochran [ 10 ] further related his \beta -invariants to a \mathbb{Q}(t) -valued invariant \eta(L) due to Kojima–Yamasaki [ 29 ] , where \mathbb{Q}(t) denotes the field of rational functions. This invariant can be interpreted as a self-linking invariant of the lift of L_{2} to \widetilde{E(L_{1})} , the infinite cyclic cover of the exterior of L_{1} . Our proof of the invariance of \gamma(L) proceeds by defining h(L)\in\mathbb{Q}(t) , an analogue of \eta for 3 -component links, checking the invariance of h(L) , and then establishing the connection between h and \gamma . More precisely, we define h(L) in terms of the linking of chosen lifts of L_{2} and L_{3} in \widetilde{E(L_{1})} . Since this quantity depends on the choice of lifts, it is defined only up to an indeterminacy. In Proposition 4.2 , we show that h(L) nevertheless determines a well-defined element of \nicefrac{{\mathbb{Q}(t)}}{{\doteq}} , where \doteq is the equivalence relation generated by p(t)\doteq tp(t). We then prove that the Taylor coefficients of h(L) at t=1 recover the \gamma -invariants of L . Theorem 1.2 . Let L=(L_{1},L_{2},L_{3}) be a 3 -component link with a distinguished component. Then h(L)\in\nicefrac{{\mathbb{Q}(t)}}{{\doteq}} is a link concordance invariant, and h(L)\doteq\sum_{k=0}^{\infty}\gamma^{k}(L)\,(t-1)^{k}. Remark 1.3 . The reader may notice that we have not specified whether we are working in the smooth or locally flat category. This is because the invariants considered here are in fact invariants of the weakest natural notion of concordance for tame links, namely I -equivalence. For convenience, we work primarily in the smooth category. In Remark 4.3 we explain how to remove the smoothness assumptions. To obtain the main results, we develop several structural and computational consequences of the \gamma -invariants. We place the construction in the framework of weak cobordism , a weaker equivalence relation than concordance in which the distinguished components cobound an annulus and the remaining components cobound disjoint