The field of definition of point sets in \(\mathbb P^1\) (Q375203)

The paper under review concerns \(n\)-point sets in the projective line and their fields of moduli. The author works with a perfect field \(K\) of characteristic different from two as base field. Let \(T\) be an \(n\)-point set in \(\mathbb{P}^1(\bar{K})\). The field of moduli of \(T\) is defined as the fixed field of the subgroup of the Galois group \(G_K\) consisting of elements mapping \(T\) to another \(n\)-point set that is projectively equivalent to \(T\). All fields of definition of \(T\) contain the field of moduli of \(T\), but the field of moduli of \(T\) need not be a field of definition. It is an interesting question to ask if they are equal. The main results of this paper answer this question when \(n \geq 5\). For \(n \geq 5\) odd, the field of moduli is a field of definition (Theorem 1). For \(n \geq 6\) even, under a mild hypothesis about the cohomology of \(K\), the field of moduli is not a field of definition; this is demonstrated by exhibiting an \(n\)-point set which is defined over a quadratic extension of \(K\) and has field of moduli \(K\), but is not defined over \(K\) (Theorem 2). The last result of the paper (Theorem 3) states a local-global principle for \(n\)-point sets in \(\mathbb{P}^1(\bar{\mathbb{Q}})\). For a given \(n \geq 5\), if an \(n\)-point set is defined over all primes, then it is defined over \(\mathbb{Q}\).