Malle’s Conjecture for Galois octic fields over ℚ {\mathbb{Q}}

Arul Shankar, Ila Varma · 2025

Read Malle’s Conjecture for Galois octic fields over ℚ {\mathbb{Q}} by Arul Shankar, Ila Varma (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-23690.txt.

Introduction

Arul Shankar and Ila Varma count a special family of number fields in a 2025 study. They look at degree-eight fields whose symmetry group matches that of a square, and they count those fields by the size of a standard measure called the discriminant.

Their count confirms the strong form of Malle's conjecture for this family, and it pins down the exact constant in front of the growth rate. The result also shows that Galois octic fields with this non-abelian group do not form a positive share of all such fields. The lead constant splits into local factors, matching a guiding principle, even though that principle fails for the related quartic case. The work is a first count of its kind for this group type.

Catalog record

CREATORArul Shankar, Ila Varma
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE118 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-23690

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Malle’s Conjecture for Galois octic fields over ℚ {\mathbb{Q}} Arul Shankar, Ila Varma Abstract We compute the asymptotic number of octic number fields whose Galois groups over ℚ {\mathbb{Q}} are isomorphic to D 4 D_{4} , the symmetries of a square, when ordering such fields by their absolute discriminants. In particular, we verify the strong form of Malle’s conjecture [ 20 ] for such octic D 4 D_{4} -fields and obtain the constant of proportionality. Our result answers the question of whether a positive proportion of Galois octic extensions of ℚ {\mathbb{Q}} have non-abelian Galois group in the negative. We further demonstrate that the constant of proportionality satisfies the Malle–Bhargava principle [ 20 , 5 ] of being a product of local masses, despite the fact that this principle does not hold for discriminants of quartic D 4 D_{4} -fields (see [ 8 , 3 ] ). This is the first instance of asymptotics being recovered for a non-concentrated family (in the sense of [ 2 ] ) of number fields of Galois group neither abelian nor symmetric. Previously, this was only known for abelian fields [ 25 ] , degree- n n S n S_{n} -fields for n = 3 , 4 , 5 n=3,4,5 (in [ 10 , 4 , 6 ] ), and degree- 6 6 S 3 S_{3} -fields [ 7 ] . 1 Introduction The purpose of this article is to prove the strong form of Malle’s conjecture [ 20 ] on the count of octic number fields having Galois group D_{4} and bounded discriminants, by determining the asymptotic growth of this count. Let N_{8}(G,X) denote the number of (isomorphism classes of) normal octic extensions over {\mathbb{Q}} with Galois group isomorphic to G . Work of Wright [ 25 ] on counting abelian fields that predated and inspired Malle’s formulation of his conjectures demonstrated \displaystyle N_{8}(C_{8},X) \displaystyle\sim c_{8}X^{1/4}, \displaystyle N_{8}(C_{2}\times C_{4},X) \displaystyle\sim c_{2,4}X^{1/4}\log^{2}X,\mbox{ and } \displaystyle N_{8}(C_{2}\times C_{2}\times C_{2},X) \displaystyle\sim c_{2,2,2}X^{1/4}\log^{6}X, where C_{n} denote the cyclic group of order n , and c_{8} , c_{2,4} , and c_{2,2,2} are explicit constants. For the quaternion group Q_{8} , Kluners [ 12 ] determined that dX^{1/4}\leq N_{8}(Q_{8},X)\leq d^{\prime}X^{1/4} for some positive real constants d and d^{\prime} , and recent work of Koymans–Pagano [ 15 ] determines exact asymptotics for the count. The remaining and most elusive Galois octic case occurs when G=D_{4} , the dihedral group of order 8, which is the subject of our first main result. If M denotes a number field of degree 8 that is normal over {\mathbb{Q}} with Galois group D_{4} , we refer to M as an octic D_{4} -field . In this article, we prove the following: Theorem 1 . Let N_{8}^{(\mathrm{r})}(D_{4},X) ( respectively, N_{8}^{(\mathrm{c})}(D_{4},X)) denote the number of isomorphism classes of totally real ( respectively, complex ) octic D_{4} -fields with absolute discriminant bounded by X . Then \displaystyle N_{8}^{(\mathrm{r})}(D_{4},X) \displaystyle\sim\frac{56+3\sqrt{2}}{256}\cdot\prod_{p\geq 3}\Bigl(1+\frac{3}{p}+\frac{1}{p^{3/2}}\Bigr)\Bigl(1-\frac{1}{p}\Bigr)^{3}\cdot X^{1/4}\log^{2}X; \displaystyle N_{8}^{(\mathrm{c})}(D_{4},X) \displaystyle\sim\frac{56+3\sqrt{2}}{128}\cdot\prod_{p\geq 3}\Bigl(1+\frac{3}{p}+\frac{1}{p^{3/2}}\Bigr)\Bigl(1-\frac{1}{p}\Bigr)^{3}\cdot X^{1/4}\log^{2}X. The above result concludes the strong form of Malle’s conjecture for Galois octic fields. While we were working on Theorem 1 , Koymans–Pagano [ 15 ] proved using different techniques that N_{8}(D_{4},X)\ll cX^{1/4}\log^{4}(X) for some (unknown) constant c>0 . Previous to either of these works, the best known result was verifying the weak form of Malle’s Conjecture (see [ 19 ] ), which was accomplished by Kluners–Malle [ 14 ] in 2004. In particular, they proved that X^{1/4}\ll N_{8}(D_{4},X)\ll X^{1/4+\varepsilon}. In conjunction with [ 25 , 12 ] , both Theorem 1 and [ 15 ] imply that 100% of Galois octic fields over {\mathbb{Q}} (when ordering such fields by their discriminants) have Galois group isomorphic to C_{2}\times C_{2}\times C_{2} . Note that it does seem possible that by keeping track of local conditions during their argument, Koymans–Pagano’s methods could achieve the sharp upper and lower bounds we prove here to obtain Malle’s conjecture! However the authors of [ 15 ] remark that their techniques do not give an explicit handle for computing the constant of proportionality given in Theorem 1 . These constants in Theorem 1 satisfy the Malle-Bhargava principle for the density of discriminants of number fields of a fixed Galois group, as predicted by recent work of Loughran–Santans [ 16 ] . This is especially noteworthy given that the constant of proportionality for the number of quartic D_{4} -fields, i.e., non-Galois quartic number fields with degree 8 normal closure over {\mathbb{Q}} , ordered by discriminant was determined by Cohen-Diaz y Diaz-Olivier [ 8 ] ) and it does not satisfy the Malle-Bhargava principle (see §5.3 of [ 3 ] ). If a group G has a transitive embedding into S_{n} , what we mean by an n -ic G -number field is a degree n extension M of {\mathbb{Q}} such that the action of the Galois group of the normal closure \widetilde{M} of M over {\mathbb{Q}} on the embeddings of M into \overline{{\mathbb{Q}}} is isomorphic to G . The Malle-Bhargava principle relies on the expectation that the number of (isomorphism classes of) global n -ic G -number fields of discriminant D is proportional to the product indexed by v of the “weighted” number of local G -compatible extensions of {\mathbb{Q}}_{v} that are discriminant-compatible with D , where v ranges over all places \nu({\mathbb{Q}}) of {\mathbb{Q}} . Using this expectation, Bhargava [ 5 ] conjectures the precise constant in the asymptotics of degree- n S_{n} fields. The Malle-Bhargava principle has been verified in only a handful of cases. It is true for abelian G with prime exponent due to work of Maki [ 17 , 18 ] and Wright [ 25 ] . It has been verified for degree n S_{n} -fields when n=3 by Davenport-Heilbronn [ 10 ] , and n=4,5 by Bhargava [ 4 , 6 ] , as well as for sextic S_{3} -fields by Bhargava-Wood [ 7 ] . However, the Malle-Bhargava principle does not hold in general, failing for example in the cases of cyclic fields of degree p^{2} for any prime p ; it also does not hold for quartic D_{4} -fields (proven by Cohen-Diaz y Diaz-Olivier [ 8 ] ). A more robust form of the Malle–Bhargava principle involves studying subfamilies of the original family constructed by imposing finitely many splitting conditions (see [ 22 ] ). To illustrate this, we obtain more general asymptotics for any family of octic D_{4} -fields satisfying any finite set of local conditions. For a place v of {\mathbb{Q}} , we define a D_{4} -type as a conjugacy class of continuous group homomorphisms \rho_{v}:{\rm Gal}(\overline{{\mathbb{Q}}}_{v}/{\mathbb{Q}}_{v})\rightarrow D_{4} . An octic étale algebra M_{v} over {\mathbb{Q}}_{v} is of D_{4} -type if there exists a continuous group homomorphism \rho_{v}:{\rm Gal}(\overline{{\mathbb{Q}}}_{v}/{\mathbb{Q}}_{v})\rightarrow D_{4} such that M_{v}\cong{\rm Ind}_{{\rm im}(\rho_{v})}^{D_{4}}\overline{{\mathbb{Q}}}_{v}^{\ker(\rho_{v})}, i.e., M_{v} is isomorphic to \#{\rm coker}(\rho_{v}) copies of the fixed field of \ker(\rho_{v}) in \overline{{\mathbb{Q}}}_{v} . In this case, we further say that M_{v} is of D_{4} -type [\rho_{v}] . For a finite place v of {\mathbb{Q}} , a local specification \Sigma_{v} is a set consisting of pairs (M_{v},[\rho_{v}]) , where M_{v} is (an isomorphism class of) an octic étale algebra of {\mathbb{Q}}_{v} of D_{4} -type [\rho_{v}] . For a collection \Sigma=(\Sigma_{v})_{v} of local specifications, we let N_{8}(\Sigma,X) denote the number of

Read the complete text