Uniform definability of the Weierstrass \(\wp\) functions and generalized tori of dimension one (Q1777340): Difference between revisions

The authors' main theme is the generalization, to any algebraically closed field, of the classical result about the equivalence of complex tori, \(\mathbb{C}/\Lambda\), and non-singular cubic curves via the Weierstrass \(\wp\) functions. When the field is non-Archimedean, the quotient space construction does not work, and they invent the construction `gluing along a homeomorphism'. They carefully analyze the equivalence of tori, and of tori and curbs. A corollary to this analysis is: Let \(R\) be a non-Archimedean real closed field and \(K= R(\sqrt{-1})\), then there are some \(K\)-tori that are not \(K\)-biholomorphic to any \(K\)-curbs, and not algebraic even in any \({\mathcal R}\), o-minimal expansion of \(R\). A second theme of the authors is definability. If \({\mathcal R}\) is elementarily equivalent to \(\mathbb{R}_{\text{an.exp}}\), then any smooth cubic \(K\)-curb is definably \(K\)-biholomorphic to a \(K\)-torus. A third theme is the exponential function. For instance, the authors show: If the function \(\wp(\tau,z)\), \(-{1\over 2}\leq \text{Re}(\tau)<{1\over 2}\) and \(|\tau|\geq 1\), is definable in an o-minimal expansion of \(\mathbb{R}_{\text{an}}\), then so is the real exponential function.