Iowa Type Theory Commute

Iowa Type Theory Commute is a technical listening room for people who like computation before the slogans get simplified. Aaron Stump uses the commute format to work through type theory, computational logic, and pure lambda calculus in a direct, reflective register. The episodes are short in setup but dense in substance. A question such as “what is control flow analysis?” quickly becomes a discussion of lambda abstractions, program points, sound approximations, closure analysis, and the correctness property behind the analysis. The show is at its best when it pauses over distinctions that are easy to skip. In episodes on terminating computation, Stump considers the possibility of starting with an untyped language where every program terminates, then letting types describe other properties instead of carrying the whole burden of logical soundness. That leads into affine lambda calculus, structural recursion, closed construction, closed reduction, and the cost of restricting a language strongly enough to prevent divergence. The mood is curious rather than declarative. He is willing to say when an idea seems original, when it may be too restrictive, and when a paper finally states the property he had been looking for. Several episodes also return to the finite developments theorem in untyped lambda calculus. There, the show walks through marked redexes, residuals, beta reduction, multiplicity measures, and the elegance of using intersection types to get termination as a corollary. Names and papers matter because the arguments matter. Jens Paulsberg, Jean-Louis Cravin, Krivine, Royald de Vrier, and Victor Talon appear as reference points in a larger conversation about how proofs are shaped. The show does not aim for broad tech commentary. It rewards listeners who already care about formal systems, programming language theory, and the small proof moves that make a theorem suddenly feel clean.