Question 2.13 — (T.
Statement
(T. Stanford) Does Mn have an interesting complementary space in Vn? Consider, for example, the space Nn⊂ Vn of invariants v with the property that v(L) = 0 for any string link L such that π1(B3− L) is free. Is Nn nontrivial? Do Nn and Mn together span Vn? Here is some background and motivation. When considering finite-type invariants of string links, th e first ones that come to mind are the Milnor invariants. These were defined by Milno r [283] in 1954 as numbers associated to links. They are not quite invariant s of links, in the usual sense, because of some indeterminacy. They are, howev er, well-defined as invariants of string links, and this point of view was take n by Habegger and Lin [163]. After Vassiliev’s work appeared, Bar-Natan [ 26] and Lin [261] showed (independently) that the Milnor invariants are finit e-type invariants. Habegger and Masbaum [164] showed that on the chord diagram l evel, the Milnor invariants (including products of Milnor invariant s) are exactly the ones that vanish on Jacobi diagrams that contain internal loops, and also that the Milnor invariants are the only rational-valued finite-type invariants of string links which are also concordance invariants. String links may have local knots in the strands, and such kno ts are not detected by Milnor invariants. If a string link L has local knots, then π1(B3− L) is not free. Hence the question as to whether finite-type invariant s can show that the complement of a string link is not free. (M. Polyak) Let us review the constructions of Milnor µ -invariant in [89]. For a n-component link L = L1∪···∪Ln, regard the homotopy class of Ln as in π1 ( S3− (L1∪···∪Ln−1) ), and write it in terms of meridians m1,···, mn−1 of L1,···, Ln−1. Consider its Magnus expansion putting mi = 1 + Xi for non- commutative variables Xi. Then, Milnor’s µ -invariant µ i1···ik,n(L) is defined to be the coefficient of X i1···X ik in the expansion, which is an invariant under the assumption that the lower µ -invariants vanish. For example, µ 1,2 is equal to the linking number lk( L1, L2) of L1 and L2. Further, if µ i,j(L) = 0 for any i, j, then µ 12,3(L) = lk( L12, L3), where L12 denotes the link which is the intersection of Seifert surfaces of L1 and L2. In general, under the vanishing assumption of the lower µ -invariants, µ 12···n−1,n(L) = lk( L12···n−1, Ln) where L12···k (for k = 2, 3,···, n− 1) denotes the link which is the intersection of Seifert surfaces of L12···k−1 and Lk.
Record
- Source
- Added
Comments
Nothing has been published against this problem yet, and nothing has been said about it.
Sign in with an institutional address to take part in the discussion. Reading every thread stays open to everyone.
Sign inSolve with an agent
Open the statement in a chat, with the problem and the ground rules already written into the prompt.