Integration in algebraically closed valued fields (Q638502): Difference between revisions

\textit{E.~Hrushovski} and \textit{D.~Kazhdan} [``Integration in valued fields'', in: V. Ginzburg (ed.), Algebraic geometry and number theory. Basel: Birkhäuser. Progress in Mathematics 253, 261--405 (2006; Zbl 1136.03025)] introduced a new theory of motivic integration for algebraically closed valued fields in a quite general context. \textit{Y.~Yin} [``Special transformations in algebraically closed valued fields'', Ann. Pure Appl. Logic 161, No.~12, 1541--1564 (2010; Zbl 1226.03044)] started to rewrite that theory in a more specific context, which allows some simplifications and which therefore is more accessible. The present article is the second part of that rewriting. More precisely, Hrushovski-Kazhdan work in an arbitrary ``\(V\)-minimal'' theory \(T\), a quite general expansion of the theory of algebraically closed valued fields of equi-characteristic zero, \(\text{ACVF}(0,0)\); Yin only considers expansions \(T\) of \(\text{ACVF}(0,0)\) by constants for a \((\text{VF},\Gamma)\)-generated substructure; the main advantage of this more specific context is the presence of syntax. Recall that the heart of motivic integration in the sense of Hrushovski-Kazhdan consists of the ``motivic measure'' which one associates to any definable set in the valued field sort \(\text{VF}\) of the expansion \(T\) of \(\text{ACVF}(0,0)\). This motivic measure itself is essentially a definable set in the \(\text{RV}\)-sort (where \(\text{RV}(K) = K^\times/(1 + \mathcal{M}_K)\) is the ``leading term structure'' of a valued field \(K\)). More precisely, motivic measure is an isomorphism \(\int_+: \mathbb{K}_+\mu\text{VF}_* \longrightarrow \mathbb{K}_+\mu\text{RV}[*]/\mu I_{\text{sp}}\) (Theorem~7.10) between Grothendieck semirings, where \(\mu\text{VF}_*\) is a category of definable sets in the valued field with volume form, \(\mu\text{RV}[*]\) is a similar category of definable sets in the \(\text{RV}\)-sort, and \(\mu I_{\text{sp}}\) is a suitable semigroup congruence relation. To obtain \(\int_+\), one first constructs a natural surjective homomorphism in the other direction: \(\mathbb{L}: \mathbb{K}_+\mu\text{RV}[*] \longrightarrow \mathbb{K}_+\mu\text{VF}_*\); this has been done in the first article of Yin [loc. cit.]. In the present article, the equivalence relation \(\mu I_{\text{sp}}\) on \(\mathbb{K}_+\mu\text{RV}[*]\) is defined (it is generated by a single, very natural equivalence); the main technical work then consists in proving that \(\mu I_{\text{sp}}\) is exactly the kernel of \(\mathbb{L}\). After that, \(\int_+\) can simply be defined as the inverse of \(\mathbb{L}\).