Simple Lie Groups of type A n A_{n} as Galois groups over ๐ \mathbf{Q}

Stepan Nesterov ยท 2026

Read Simple Lie Groups of type A n A_{n} as Galois groups over ๐ \mathbf{Q} by Stepan Nesterov (2026) 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-2604-27402.txt.

Introduction

Nesterov turns his earlier work on prime-modular monodromy for cyclic covers of the projective line into a new payoff. Whole families of PSL and PSU groups appear as Galois groups over the rationals. For the first time, one fully explicit infinite family lets the field degree over the prime field grow without bound while the group stays distinct from its PGL and PGU counterparts. The build thus parts these Galois groups from their projective linear and unitary kin.

Catalog record

CREATORStepan Nesterov
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE33200
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.27402

Browse the full library ยท About the Tower of Babel

Full text

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

Simple Lie Groups of type A n A_{n} as Galois groups over ๐ \mathbf{Q} Stepan Nesterov Abstract In this paper, we utilize our previous results on mod p p monodromy of cyclic coverings of the projective line to realize a large series of groups of the form PSL โก ( n , q ) \PSL(n,q) and PSU โก ( n , q ) \PSU(n,q) as Galois groups over ๐ \mathbf{Q} . We achieve for the first time a fully explicit infinite series of such groups where simultaneously the field can have arbitrarily large degree over the prime field and the group does not coincide with PGL โก ( n , q ) \PGL(n,q) or PGU โก ( n , q ) \PGU(n,q) , respectively. 1 Introduction The Inverse Galois Problem is to find, for a given finite group G , a Galois extension of \mathbf{Q} with Galois group G . We use [ 6 ] as a general reference regarding the Inverse Galois Problem. In view of the classification of finite simple groups, it is natural to attempt to solve the Inverse Galois Problem for the simple groups frirst. Despite a lot of progress, for many of these groups the Inverse Galois problem remains open. To solve the Inverse Galois Problem for the alternating groups A_{n} is not that difficult, and was already done by Hilbert. One can even give explicit examples: indeed, for any n , the splitting field of a polynomial x^{n}-x-1 has Galois group S_{n} ; the polynomial 1+x+\frac{x^{2}}{2}+\ldots\frac{x^{n}}{n!} has Galois group S_{n} if n is not a multiple of 4 , and A_{n} otherwise. See [ 9 ] for a large family of explicit polynomials with Galois group A_{n} . One of the most productive methods to solve the Inverse Galois Problem is the rigidity method, originally presented in [ 10 ] . In addition to recovering the result for S_{n} or A_{n} , it allows us to solve the Inverse Galois Problem for a lot of simple groups of Lie type over prime fields \mathbf{F}_{p} , and all the sporadic groups with the exception of M_{23} and M_{24} . See [ 6 , Chapter II] for an extensive list of groups where the rigidity method applies. This leaves us with a large class of simple groups of Lie type over finite fields \mathbf{F}_{q} , where q is not a prime, but a prime power. In [ 6 , Chapter III.10] a generalisation of the rigidity method is discussed, which allows us to cover many simple groups associated to symplectic and orthogonal groups over finite fields. An alternative approach is to prove that Galois representations attached to automorphic forms often have full image [ 11 ] [ 12 ] . However these results are ineffective in the sense that it you do not know which field \mathbf{F}_{p^{k}} you get as a field of coefficients, only that k can be arbitrality large. In this paper we give, for the first time, a realisation of an infinite set of simple groups of the form \PSL(n,q) and \PSU(n,q) as Galois groups over \mathbf{Q} , where simultaneously n can be arbitrarily large, \mathbf{F}_{q} can have arbitraily large degree over the prime field, and the condition on (n,q) to arise as a Galois group over \mathbf{Q} is complicated, but completely explicit. We achieve this by considering the action of the absolute Galois group G_{\mathbf{Q}} on the etale cohomology with finite coefficients of cyclic coverings of \mathbf{P}^{1} . The deck transformation of the curve coming from the covering map to \mathbf{P}^{1} is what endows the cohomology with the structure of an \mathbf{F}_{q} -vector space. The main result of this paper is: Theorem 1 . Let l be an odd prime, and let n be a natural number, not congruent to -2 modulo l . Let e=\mathrm{ord}\ p\mod l . Then: โ€ข If \frac{l-1}{e} is even, then \PSL((l-1)n-2,p^{e}) is a Galois group over \mathbf{Q} . โ€ข If \frac{l-1}{e} is odd and, in addition to all the previous assumptions, either e is a multiple of 4 or if 2^{r} is the largest power of 2 dividing p+1 , then \frac{l-1}{2}n\equiv 1\mod 2^{r} , then \PSU((l-1)n-2,p^{\frac{e}{2}}) is a Galois group over \mathbf{Q} . For various choices of l and p , we have the following corollaries: โ€ข For a prime p=3k+1 and n not divisible by 3 , \PSL(2n,p) is a Galois group over \mathbf{Q} . Under these congruence conditions, the corresponding groups \PGL(2n,p) are rigid, so under the additional assumption \mathrm{gcd}(n,p-1)\leq 2 , the simple groups \PSL(n,p-1) were previously obtained as Galois groups over \mathbf{Q} by the rigidity method [ 6 , Theorem III.10.3] . In the case n=2 , the Inverse Galois Problem is solved in [ 13 ] . โ€ข For a prime p=5k-1 , and n odd, but not divisible by 5 , \PSL(2n,p^{2}) is a Galois group over \mathbf{Q} . In particular, \PSL(2,p^{2}) is a Galois group over \mathbf{Q} for such p . This complements a result of Mestre [ 7 ] , where \PSL(2,p^{2}) is realised as a Galois group over \mathbf{Q} for p=5k\pm 2 . โ€ข For a prime p=5k+1 , and n odd but not divisible by 5 , \PSL(2n,p) is a Galois group over \mathbf{Q} . โ€ข If p and l are two odd primes, and p has multiplicative order \frac{l-1}{2} mod l , then \PSL(l-3,p^{\frac{l-1}{2}}) is a Galois group over \mathbf{Q} . So this method allows us to obtain groups \PSL(n,q) where q is an arbitrarily large power of a prime, but the rank of the group will also become arbitrarily large. โ€ข For a prime p=12k+7 and n not divisible by 3 , \PSU(2n,p) is a Galois group over \mathbf{Q} . Again, these congruence conditions also arise when applying the rigidity method. [ 6 , Theorem III.10.3] โ€ข For a prime p=28k+13 and n odd, but not congruent to -2 mod 7 , \PSU(6n-2,p) is a Galois group over \mathbf{Q} . โ€ข For a prime p=28k+5,28k+17 , and n odd, but not congruent to -2 mod 7 , \PSU(6n-2,p^{3}) is a Galois group over \mathbf{Q} . โ€ข If p and l are two odd primes, such that l=4m+1 , and p is a primitive root mod l , then \PSU(l-3,p^{\frac{l-1}{2}}) is a Galois group over \mathbf{Q} . 2 Proof Outline In a previous article, we studied the monodromy action on cohomology of cyclic coverings of \mathbf{P}^{1} and proved the following theorem: Theorem 2 . If n\geq l , then the image of the monodromy representation of the pure braid group on the cohomology of cyclic covering of \mathbf{P}^{1} branched at n points is either the group \SlU(n-1,q):=\{A\in\GL(n-1,q^{2}):A\cdot\mathrm{Fr}_{q}(A)^{t}=1,\det A\in\mu_{l}\} or \SlL(n-1,q):=\{A\in\GL(n-1,q):\det A\in\mu_{l}\} , depending on whether or not \frac{l-1}{\mathrm{ord(p\mod l)}} is odd or even. The naive model of a cyclic covering of \mathbf{P}^{1} with monodromy vector (k_{1},\ldots,k_{n}) branched at (\lambda_{1},\lambda_{2},\ldots,\lambda_{n-1},\infty) is the normalization of the projective closure of the affine curve y^{l}=(x-\lambda_{1})^{k_{1}}\ldots(x-\lambda_{n-1})^{k_{n-1}} , with the deck transformation (x,y)\mapsto(x,e^{\frac{2\pi i}{l}}y) . If we take \lambda_{i} to be rational and consider the curve given by the same equation, then the deck transformation is not defined over \mathbf{Q} , which forces the corresponding field extension to not be normal. Therefore, the first step in the proof is to

Read the complete text