Ribbon-clasp surface-links and normal forms of immersed surface-links (Q2405072)

There is now a substantial literature on methods to construct and study embedded surface links, and the paper under review represents another step towards adapting this theory to immersed surface links. Here, an immersed surface link is a closed, oriented surface generically immersed in \(\mathbb{R}^4\), meaning that self-intersections arise precisely as finitely many transverse double points; two immersed surface links are considered equivalent if they are ambient isotopic. The first main result, Theorem 1.4, states that an immersed surface link is equivalent to one placed in a certain normal form in \(\mathbb{R}^3 \times [-4,4]\), meaning that for some \(k,k'\geq 1\), the intersection of each connected component with the hyperplane \(\mathbb{R}^3 \times \{t\}\) for \(t=-3,-2,1,0,1,2, 3\) is, respectively: a \(k\)-component unlink (just above the minima), a \(k\)-component split union of an unlink and some Hopf links (one for each positive double point), a knot, a link, a knot, a \(k'\)-component split union of an unlink and some Hopf links (one for each negative double point), and a \(k'\)-component unlink (just below the maxima); for other values of \(t\in (-3,3)\) adjacent level sections are connected by isotopy. The result for when the surface link is embedded (and the specifications at the \(t=-2\) and \(t=2\) levels are not necessary) was proven in [\textit{A. Kawauchi} et al. Math. Semin. Notes, Kobe Univ. 10, 75--126 (1982; Zbl 0506.57014)], and it is worth noting that normal forms in this embedded case have found numerous non-trivial applications (see, for example, [\textit{J. S. Carter} and \textit{M. Saito}, Knotted surfaces and their diagrams. Providence, RI: American Mathematical Society (1998; Zbl 0904.57010)]). The new ingredient in the present paper is to replace a simple crossing change sequence in a motion picture description (a.k.a. a link transformation sequence) of an immersed surface link by a pair of band surgeries followed by the unlinking and capping off of a Hopf link. For a ribbon surface link (i.e., the result of performing 1-handle surgeries on a trivial link of 2-spheres), \textit{A. Kawauchi} et al. also showed in [Math. Semin. Notes, Kobe Univ. 11, 31--69 (1983; Zbl 0568.57004)] that the normal form may be chosen symmetrical with respect to the hyperplane \(\mathbb{R}^3\times \{0\}\). The second main result of the paper under review (Theorems 1.3 and 1.5) extends this to the normal form of immersed surface links that are obtained from ribbon surface links by performing some finger moves. The proof is a consequence of the work of [loc. cit.] and the observation that finger moves and 1-handle surgeries are supported near arcs.