Almost \(\omega \)-categorical weakly o-minimal theories of convexity rank 1 (Q2221962)

This paper studies almost \(\omega\)-categorical weakly o-minimal theories of convexity rank \(1\). Recall that a weakly o-minimal structure is a linearly ordered structure \(M \equiv \langle M, <, \dots \rangle\) in which every definable subset of \(M\) is a finite union of convex sets. Weakly o-minimal theories are classified by the convexity rank: this is a means of measuring how far such structures are from o-minimal ones, since an o-minimal theory has convexity rank one. In this respect, weakly o-minimal theories of convexity rank 1 are closest to o-minimal theories; note that, however, some weakly o-minimal theories of convexity rank one are not quite o-minimal. A countable complete theory \(T\) is said to be almost \(\omega\)-categorical whenever there are finitely many types \(q(x_1, \cdots, x_n)\) such that \\ \(q(x_1, \cdots, x_n) \supseteq \bigcup_{i=1}^n p_i(x_i)\) for any set of \(n\) \(1\)-types \(p_1(x_1), \cdots, p_n(x_n)\). The concept of almost \(\omega\)-categorical theory is closely related to that of Ehrenfeucht theory; a theory is said to be Ehrenfeucht whenever the number of its (pairwise nonisomorphic) countable models is finite, and \(>1\). Considering almost \(\omega\)-categorical weakly o-minimal theories of convexity rank 1, the authors prove the following results: \begin{enumerate} \item any such theory is binary (all types are determined by types in at most two free variables) (Theorem (3.6)), and, \item orthogonality of any family of pairwise weakly orthogonal non-algebraic types (Theorem (2.15)). \end{enumerate} The proof makes extensive use of results around the notion of \(p\)-stable convex-to-the right (or convex-to-the-left) formula, where \(p\) is any non-algebraic \(1\)-type. More precisely, in Proposition (2.7) it is proved that in an almost \(\omega\)-categorical weakly o-minimal theory \(T\) there are no \(p\)-stable convex to the right (or convex to the left) formulas, for \(p\) any non-algebraic type \(\in S_1(\emptyset)\). An important Corollary (2.8) is that the set of realizations of a non-algebraic type \(p \in S_1(\emptyset)\) is dense. A further key element in the proof is Lemma 2.10 (originally Lemma 3.7 in [the first author and \textit{S. V. Sudoplatov}, Ann. Pure Appl. Logic 168, No. 1, 129--149 (2017; Zbl 1454.03046)]), expressing a weak form of binarity.