Positive knots are rationally anisotropic

Let $K$ be a knot, let $\overline{X}$ be the infinite cyclic cover of its exterior, and consider the associated Seifert-form data over $\mathbb{Q}$. A knot is $\mathbb{Q}$-anisotropic when it has no nontrivial isotropic subspace for this rational form. Rational anisotropy conjecture for positive knots. Every positive knot is $\mathbb{Q}$-anisotropic. This property would, together with the Kervaire–Gilmer result, remove the signature hypothesis used in the source to deduce isomorphism of rational Alexander modules for concordant positive knots. The conjecture is presented as an unresolved condition in the paper.