Categoricity and multidimensional diagrams

Categoricity and completeness of theories (03C35) Classification theory, stability, and related concepts in model theory (03C45) Abstract elementary classes and related topics (03C48) Properties of classes of models (03C52) Set-theoretic model theory (03C55) Other infinitary logic (03C75) Other combinatorial set theory (03E05) Large cardinals (03E55)

Review text

This paper develops the tool of multidimensional diagrams to give a partial answer to \textit{S. Shelah}'s categoricity conjecture [Isr. J. Math. 46, 212--273 (1983; Zbl 0552.03019)]:\N\NConjecture: Let \(\mathbf{K}\) be an abstract elementary class with Löwenheim-Skolem number \(\mathrm{LS}(\mathbf{K})\). If \(\mathbf{K}\) is categorical in some \(\lambda\geq\beth_{(2^{\mathrm{LS}(\mathbf{K})})^+}\), then it is categorical in all \(\lambda'\geq\lambda\).\N\NEarlier results have given a positive answer, assuming a mix of\N\begin{itemize}\N\item[--] non-ZFC axioms (weak general continuum hypothesis, large cardinals);\N\item[--] nice structure of \(\mathbf{K}\) (amalgamation, primes, tameness);\N\item[--] specific properties of \(\lambda\) (successor cardinal, of certain cofinality).\N\end{itemize}\N\NTheorem 14.5 provides a positive answer assuming a proper class of strongly compact cardinals. This improves on \textit{S. Shelah} and \textit{M. Makkai}'s result [Ann. Pure Appl. Logic 47, No. 1, 41--97 (1990; Zbl 0704.03015)] which focuses on the special case that \(\lambda\) is a successor and that \(\mathbf{K}=\mathrm{Mod}(\psi)\) where \(\psi\) is a \(L_{\kappa,\omega}\) sentence for some strongly compact \(\kappa\).\N\NCorollary 14.10 (together with a later improvement by \textit{S. Vasey} [Sel. Math., New Ser. 25, No. 5, Paper No. 65, 51 p. (2019; Zbl 1468.03042)]) also provides a positive answer but assumes the weak general continuum hypothesis and that \(\mathbf{K}\) has amalgamation.\N\NWhile other structural properties of \(\mathbf{K}\) such as primes and tameness are not explicitly assumed, they are derived from the original assumptions using multidimensional diagrams -- the \(n\)-dimensional analog to the 2-dimensional independence notion called \textit{nonforking}. Multidimensional diagrams allow a better transfer of \textit{extension} and \textit{uniqueness} properties across different cardinals. Ultimately, a nice enough independence relation (called \textit{excellent}) on multidimensional diagrams leads to the proof of the categoricity conjecture.\N\NSections 1--7 introduce the basic notions and background results, including nonforking. Sections 8--12 build an independence relation on multidimensional diagrams from nonforking and study its properties. Section 13 proves the existence of primes so as to invoke the categoricity transfer by \textit{S. Vasey} [Math. Log. Q. 64, No. 1--2, 25--36 (2018; Zbl 1521.03076)]. Section 14 contains the main theorems and their variants.\N\NSome minor typos in the paper: On page 2314 Remark 3.14(4), ``reasonable should have been ``resolvable. On page 2365 Definition 13.8(1), the embedding should send \(f(a)=b\).

Recommendations

Cites work

Cited in

(3)

This page was built for publication: Categoricity and multidimensional diagrams

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6566412)