Radon measure definitions vary but converge on locally compact spaces

A long-running debate among mathematicians over the precise definition of a Radon measure has resurfaced on Math StackExchange, where users are once again grappling with a concept that carries surprisingly little consensus across textbooks.

The original poster observed that several non-equivalent definitions for Radon measure circulate in the literature, yet all of them collapse into a single working definition when restricted to locally compact Hausdorff spaces. That apparent paradox, the user noted, likely explains why the term survives with such divergent formulations rather than being pinned down to a canonical statement.

The question quickly drew responses pointing to the practical reality that textbooks in measure theory, functional analysis, and probability each adopt slightly different conventions depending on whether the underlying space is assumed to be Hausdorff, locally compact, or fully general. Some authors require a Radon measure to be a positive Borel measure that is finite on compact sets and inner regular with respect to compact sets. Others extend the definition to arbitrary Hausdorff spaces by demanding inner regularity with respect to all finite Borel measures, while still others impose outer regularity on open sets.

Commenters on the thread pointed to Folland’s Real Analysis and Rudin’s Real and Complex Analysis as two of the most frequently cited references, noting that each frames the concept differently and that neither definition strictly contains the other outside the locally compact setting. The discussion also touched on Schwartz’s treatment of Radon measures on general topological spaces, which remains a standard reference for practitioners working beyond the locally compact framework.

The practical takeaway for working analysts, several respondents argued, is that the choice of definition rarely matters in the settings where Radon measures are most commonly applied, including integration on manifolds, representation theorems for linear functionals, and duality in measure theory. In those contexts, the locally compact Hausdorff assumption does the heavy lifting of reconciling the competing formulations.

The thread has continued to attract answers from users seeking clarity ahead of graduate exams and research work, suggesting that the ambiguity remains a genuine source of friction for those moving between courses and textbooks that adopt conflicting conventions.

For anyone working with Radon measures today, the consensus emerging from the discussion is straightforward: identify which definition your source uses, verify whether your underlying space is locally compact Hausdorff, and proceed accordingly. Outside that setting, the distinctions between formulations become mathematically real rather than merely cosmetic.

Found a mistake? See our corrections policy. Have a tip? Contact the editor.