Alternating groups and moduli space lifting invariants (Q607846)

From author's abstract: ``The genus of a curve discretely separates decidely different algebraic relations in two variables to focus us on the connected moduli space \(\mathcal M_g\). Yet, modern applications also require a data variable (function) on the curve. The resulting spaces are versions, depending on our needs for this data variable, of Hurwitz spaces. A Nielsen class ({\S}1.1) consists of \(r \geq 3\) conjugacy classes \(\mathbf C\) in the data variable monodromy \(G\). It generalizes the genus.'' ``Some Nielsen classes define connected spaces. To detect, however, the components of others requires further subtler invariants. We regard our main result as level 0 of Spin invariant information on moduli spaces.'' ``In the main result, \(G=A_n\) (the alternating group), \(r\) counts the data variable branch points and \(\mathbf C={\mathbf C}_{3^r}\) is \(r\) repetitions of the 3-cycle conjugacy class. This Nielsen class defines two spaces called absolute and inner: \(\mathcal H(A_n,\mathbf C_{3^r})^{\mathrm{abs}}\) of degree \(n\), genus \(g=r-(n-1)>0\) covers and \(\mathcal H(A_n,\mathbf C_{3^r})^{\mathrm{in}}\) parametrizing Galois closures of such covers.'' The main result shows that a \(\frac{1}{2}\)-canonical class (spin) invariant (noted by a \(\pm\) subscript) completely determines components of \(\mathcal H(A_n,{\mathbf C}_{3^r})^{\ast}\), (\(\ast=\)in or abs), namely, ``the parity of a spin invariant precisely identifies the two components of each space. The inner result is the deeper.'' ``We examine the effect of combining the main result, [\textit{M. Artebani} and \textit{G. P. Pirola}, ``Algebraic functions with even monodromy'', Proc. Am. Math. Soc. 133, No. 2, 331--341 (2005; Zbl 1058.14041)] and \(\frac{1}{2}\)-canonical classes on \(\mathcal M_g\). First: \S 5.2 considers an analog of a famous conjecture of Shafarevich: With \(H\) the composite group of all Galois extensions \(K/\mathbb{Q}\) with group some alternating group, does the canonical map \(G_{\mathbb{Q}}\to H\) have pro-free kernel. Second: Theorem 6.1.5 produces nonzero automorphic (\(\theta\)-null power) functions on the reduced Hurwitz spaces \(\mathcal H_+(A_n,\mathbf C_{3^r})^{\mathrm{abs},\mathrm{rd}}\) (resp.~\(\mathcal H_-(A_n,\mathbf C_{3^r})^{\mathrm{abs},\mathrm{rd}}\)) when \(r\) is even (resp.~odd), for either \(g=1\) or \(n\geq 12g+4\).''