The concordance genus of knots (Q1880616): Difference between revisions

A knot consists of a smooth oriented pair \((S,K)\) where \(S\) is diffeomorphic to the 3-sphere, \(S^3\), and \(K\) is diffeomorphic to the 1-sphere, \(S^1\). Knots are equivalent if they are oriented diffeomorphic. With this, such notations as connected sum are defined as they usually are for oriented manifolds and pairs. We write \(-K\) to denote the knot \((-S^3,-K)\), the mirror image of \(K\) with orientation reversed. A knot \((S,K)\) is called slice if there is a proper 4-manifold pair \((B,D)\) with \(B\) diffeomorphic to the 4-ball, \(B^4\), \(D\) diffeomorphic to the 2-disk, \(B^2\), and with \(\partial(B,D)=(S,K)\). The knots \(K_1\) and \(K_2\) are called concordant if the connected sum \(K_1\sharp -K_2\) is slice. Concordance is an equivalence relation and the set of equivalence classes forms a group under the operation induced by connected sum. This group is called the concordance group, denoted by \(\mathcal C\). An alternative approach to concordance is the following. The knots \((S_1,K_1)\) and \((S_2,K_2)\) are called concordant if there is a properly embedded annulus, \(A\), in \(S^3\times [0, 1]\) with \(\partial(S^3\times [0, 1],A)= (S_1,K_1)\cup-(S_2,K_2)\). In knot concordance, three genera arise naturally, \(g(K),\,g_4(K)\) and \(g_c(K)\): these are the classical genus, the 4-ball genus, and the concordance genus, defined to be the minimum genus among all knots concordant to \(K\). Clearly \(0\leq g_4(K)\leq g_c(K)\leq g(K)\). \textit{A. Casson} and \textit{Y. Nakanishi} [Math. Semin. Notes, Kobe Univ. 9, 99--108 (1981; Zbl 0481.57002)] gave examples to show that \(g_4(K)\) need not be equal to \(g_c(K)\), independently. Associated to a knot \(K\) and a choice of Seifert surface, \(F\), there is a Seifert form \(V_K\): this is an integral matrix satisfying \(\det(V_K-V_K^t)=\pm 1\), where \(V_K^t\) is the transpose. There is a Witt group of such Seifert forms, denoted \(\mathcal G_{-}\), defined by \textit{J. Levin} [Comment. Math. Helv. 44, 229--244 (1969; Zbl 0176.22101)]. Levin showed that the map \(K\rightarrow V_K\) induces a homomorphism \(\psi: \mathcal C\rightarrow \mathcal G_{-}\). This map has kernel denoted \(\mathcal A\), the concordance group of algebraically slice knots. In this paper, the following theorem is proved: For every \(N\) there exists a knot \(K\in \mathcal A\) with \(g_4(K)=1\) and \(g_c(K)=g(K)=N\). The argument begins by reviewing and extending Casson's and Nakanishi's results, and several concrete examples are included. Further there is a table of \(g_c\) for all prime knots with 10 crossings.