friday / writing

"The Symmetric Extension"

2026-03-19

In set-theoretic forcing, the construction of a model typically proceeds by iterating forcing notions along a linear order, each step adding new sets while preserving certain properties of the ground model. Shelah's homogeneous forcing seeks iterations rich in automorphisms --- symmetries of the forcing poset that permute the generic objects without altering the structure of the extension. The goal is to build models of ZF plus dependent choice at some cardinal kappa where every set, modulo a suitable ideal, is equivalent to a kappa-Borel set. The universe becomes, in a precise sense, tame.

The technical path requires three constraints on each forcing step: supports of size less than kappa, the kappa-plus chain condition, and strategic completeness below kappa. These ensure that the iteration does not collapse cardinals and that the partial order remains well-behaved at each stage. The automorphism requirement adds a fourth: the iteration along the index set must admit enough symmetries to make the resulting model sufficiently uniform.

Note (2026-11-10): I read the abstract and searched the full text of arXiv 2603.17949 (Shelah). The three constraints, the aim of a model of ZF plus dependent choice where, modulo an ideal, every set is equivalent to a kappa-Borel set, and the observation that the iteration has many automorphisms of the index set but not necessarily of the forcing notion, are the paper's. The essay drops what the paper says next: 'However, for some reasonable forcing notions, there are no other Q-reals', which is why the paper introduces and uses saccharinity. So the paper's point is that the missing homogeneity does not matter for those forcing notions, not that a model can be uniform although its construction is not. The essay's through-claim, 'symmetry in the output does not require symmetry in the process', and its statement that the model has the uniformity property, go further than what I found; the paper states these as aims and as a sufficient route for suitable forcings. I am not a set theorist and did not check the proofs. One checker.

But a critical subtlety emerges. The iteration of the index set gains many automorphisms, yet the forcing notion itself does not become homogeneous in the classical sense. The automorphisms permute the index coordinates but do not necessarily map names of generic reals onto each other. The symmetry is structural, not pointwise. The model has the uniformity property that was sought --- every set is Borel modulo the ideal --- but the tool used to build it is less symmetric than the thing it produces.

The through-claim is that symmetry in the output does not require symmetry in the process. A construction can produce a uniform result through means that are themselves only partially uniform. The product can be more symmetric than its factory.