On the history of the isomorphism problem of dynamical systems with special regard to von Neumann's contribution (Q666274): Difference between revisions

Let \((X, {\mathcal A}, \mu)\) be a measure space (i.\,e., \(X\) a set, \({\mathcal A}\) a \(\sigma\)-algebra of subsets of \(X\), and \(\mu\) a measure on \({\mathcal A}\)). Call a transformation \((X, {\mathcal A}, \mu, T)\) measure preserving (or, for short, an ``m.p.t.'') if, for all \(A\in{\mathcal A}\), \(\mu(T^{-1}(A))=\mu(A)\). Call two m.p.t.s\ \((X, {\mathcal A}, \mu, T)\) and \((Y, {\mathcal B}, \nu, S)\) isomorphic if there is a bijection \(\pi\): \(X\rightarrow Y\) that preserves \(\sigma\)-algebra, measure, and commutes, i.\,e., \(\pi T=S\pi\) and \(\pi^{-1}S =T\pi^{-1}\). (As usual, all these definitions hold modulo sets with measure zero, i.\,e., ``a.\,e.'') Finally, call a transformation \(T\) ergodic if only trivial sets are invariant under it, i.\,e., whenever \(T^{-1}(A)=A\) entails \(\mu(A)=1\) or \(\mu(A)=0\). A natural question to ask, then, is whether there are certain features whose presence would guarantee two m.p.t.s\ to be isomorphic. Considering that every m.p.t.\ induces a unitary operator (the so-called Koopman operator) on the associated Hilbert space \(L^2(X, {\mathcal A}, \mu)\) with a spectrum of eigenfunctions, John von Neumann had famously proved the first such isomorphism result [\textit{J. von Neumann}, Ann.\ Math.\ (2) 33, 587--642 (1932; Zbl 0005.12203)], namely, that m.p.t.s\ which are ergodic and have a discrete (pure point) spectrum are isomorphic iff their Koopman operators are unitarily equivalent. In the article under review, the two authors publish and translate for the first time a letter John von Neumann wrote to Stanislaw Ulam in 1941. The letter contains a proof sketch for a negative result, namely, that his theorem doesn't hold for ergodic m.p.t.s\ with mixed spectrum. The article proceeds as follows: Section 1 (pp.~71--73) provides a general introduction, Section 2 (pp.~73--76) introduces some technical language, Section 3 (pp.~76--79) reproduces the text of von Neumann's letter, Section 4 (pp.~79--84) renders the proof in sufficient detail, Section 5 (pp.~84--86) tracks scholarly references to the result, while Section 6 (pp.~86--90) summarizes subsequent developments in the field; the final Section 7 (pp.~90f.) discusses the repercussions the letter has for the history of the subject matter. I found the article well-written, with only a few and very minor typos, and think the authors achieve their goal of clarifying ``both systematically and in historical context \dots von Neumann's contribution to discovering the relationship between spatial isomorphism and spectral isomorphism'' (p.~71). While the letter itself was previously unknown, its result [\textit{P.~R.~Halmos}, Bull.\ Am.\ Math.\ Soc.\ 55, 1015--1034 (1949; Zbl 0036.35501), \S\,11] as well as the general historical picture were known. Many will appreciate, however, to have all pertinent historical information gathered in a single place, especially so since no other article on the history of this topic known to the reviewer paints the picture the way the authors do. No attempt has been made to round out the historical account by completing it with information from other, more personal sources -- for example, from biographies (like [\textit{P.~R.~Halmos}, I want to be a mathematician. An automathography. New York etc.: Springer-Verlag (1985; Zbl 0569.01020)] or [\textit{S.~M.~Ulam}, Adventures of a mathematician. New York: Charles Scribner's Sons (1976; Zbl 0352.01009)]) or from correspondence (just one letter is mentioned from [\textit{M.~Rédei} (ed.), John von Neumann: Selected Letters. Edited by Miklós Rédei. History of Mathematics (Providence) 27. Providence, RI: AMS (2005; Zbl 1092.01018)]). Since ergodic theory is fascinating and the protagonist famous, I'd consider this article to be a most welcome companion piece to another letter by von Neumann on ergodic theory recently published [\textit{J. D. Zund}, Hist.\ Math.\ 29, No. 2, 138--156 (2002; Zbl 1003.01011)].