Non-meridional epimorphisms of knot groups (Q276600)

For a knot \(K\) in \(S^3\), the knot group is \(G(K)=\pi_1 (S^3 -K)\). An important problem in the study of knots is to determine when there is an epimorphism between two knot groups. Such an epimorphism \(G(K)\to G(K')\) is \textit{meridional}, if a meridian of \(K\) is sent to a meridian of \(K'\). There are examples of non-meridional epimorphisms \(G(K)\to G(K')\), but for any such previously known examples, there also exists a meridional epimorphism \(G(K)\to G(K')\). This leads to the question of whether there is a meridional epimorphism, whenever there is an epimorphism. The main result of the paper is that the answer is in the negative: There are infinitely many distinct pairs of prime knots \((K,K')\) for which there is an epimorphism \(G(K)\to G(K')\), but there is no meridional epimorphism \(G(K)\to G(K')\). The authors start by constructing pairs of knot groups which admit non-meridional epimorphisms and present a geometric method (using the satellite construction) to produce from these examples infinitely many knot groups which admit non-meridional epimorphisms. By developing computational results of certain twisted Alexander polynomials they show the following: Let \(T\) be the trefoil knot or the figure-eight knot. Then there are infinitely many prime knots \(K_1 ,K_2 ,\dots \) such that there is an epimorphism \(G(K_n )\to G(T)\), but there is no meridional epimorphism \(G(K_n )\to G(T)\).