Anneaux <i>p</i>-adiquement clos et anneaux de fonctions définissables (Q3984428)

Let \(p\) be a prime number and let \(\mathbb{Q}_ p\) denote the field of \(p\)- adic numbers. Let \(T_ p\) denote the theory of Henselian valuation rings (commutative with 1) which are not fields, whose residue field is elementarily equivalent to \(\mathbb{Q}_ p\) and whose value group is divisible. In this well written paper, the author finds a nice axiomatization of \(T_ p\) (and also of related theories). Then he gives a direct proof (based on the axiomatization) of the fact that the field of fractions of any model of \(T_ p\) is elementarily equivalent to \(\mathbb{Q}_ p\). He also shows that a ring \(A\) is a model of \(T_ p\) iff there are a valuation field \(K\equiv\mathbb{Q}_ p\) and a convex subgroup \(H\leq\text{val}(K)\) such that \(A\) is the valuation ring induced by \(H\), i.e. \(A=\{x\in K: \exists u\in K^*,v(x)\geq v(a)\}\). In the last part of the paper the author considers the ring of definable and continuous functions on a \(p\)-adic curve. Let \(V\subseteq\mathbb{Q}^ n_ p\) be a curve. Let \[ V(\mathbb{Q}_ p)=\{x\in\mathbb{Q}^ n_ p: f(x)=0\text{ for all }f\in I(V)\triangleleft K[X_ 1,\dots,X_ n]\}. \] Let \({\mathcal C}(V(\mathbb{Q}_ p))\) be the ring of definable, continuous \(\mathbb{Q}_ p\)-valued functions on \(V(\mathbb{Q}_ p)\). Finally let \(p\triangleleft{\mathcal C}(V(\mathbb{Q}_ p))\) be a prime ideal which is not maximal. By using the above result and inspired by the work of Dickmann in the real case, the author shows that \({\mathcal C}(V(\mathbb{Q}_ p))/p\) is a model of \(T_ p\). An interesting intermediate result is the following: \({\mathcal C}(V(\mathbb{Q}_ p))/p\simeq{\mathcal C}(\mathbb{Q}_ p)/q\) for some prime ideal \(q\) of \({\mathcal C}(\mathbb{Q}_ p)\) and that \(p\) is the kernel of some evaluation map at a non-standard point of \(V\). The following results also follow from the author's axiomatization by applying known variants of Ax-Kochen-Ershov principle: 1. \(T_ p=\hbox{Th}(\mathbb{Q}_ p[[t^ \mathbb{Q}]])\), where \(\mathbb{Q}_ p[[t^ \mathbb{Q}]]\) denotes the generalized formal power series, and therefore the theory \(T_ p\) is complete. 2. \(T_ p\) is decidable. 3. \(T_ p\) has (primitive recursive) elimination of quantifiers in the language of rings together with a predicate for the set of \(n\)-th powers for each positive integer \(n\). 4. \(T_ p\) is model-complete the language of rings together with the divisibility predicate: \(x\mid y\leftrightarrow\exists z(xz=y)\).