A measurable cardinal carries at least one normal measure — an ultrafilter with completeness and normality properties that enable the ultrapower construction. How many normal measures can a measurable cardinal carry? The Kimchi-Magidor theorem showed that for strongly compact cardinals, the number can be prescribed arbitrarily. But what about measurable cardinals that sit above supercompact cardinals, where inner model theory breaks down?
Note (2026-11-10): I checked the abstract of Apter, Kaplan and Poveda. It says the classical Kimchi-Magidor result is that the first n measurable cardinals can be strongly compact, that this can be combined with an arbitrary prescribed pattern of normal measure counts, and that the first measurable above a supercompact cardinal, and the first measurable limit of supercompacts, can carry any given number of normal measures, using consequences of the Ultrapower Axiom. The essay describes Kimchi-Magidor as prescribing the number of measures on strongly compact cardinals and says any measurable above a supercompact can carry any number; the abstract states the second for the first such measurable cardinal, and the conclusion that the count is independent of the cardinal's strength is the essay's own.
The count can still be prescribed (arXiv:2603.11710). Using the Ultrapower Axiom — an axiom about the linear ordering of ultrapowers — arbitrary patterns of normal measure counts can be imposed on measurable cardinals above supercompact cardinals and on measurable limits of supercompact cardinals. The technique avoids core model methods entirely.
The structural insight: the number of normal measures is not determined by the cardinal's position in the large cardinal hierarchy. A measurable cardinal above a supercompact cardinal can carry exactly one measure, or uncountably many, or any number in between — the count is independent of the cardinal's strength. The measures are not forced by the cardinal's properties; they're compatible with them. The Ultrapower Axiom provides enough structure to control the count from above, replacing the core model theory that would provide control from below. Different tools, same precision.