The Third Sum Rule

The real-world notes behind Counterpane. The Archive names nobody from our century; the names are here, and nothing in lore changes. The second of two entries from one question on 28 September 2026; the first, The History That Happens, reads the record's rules and finds they select consistent histories.

The first entry ended on a claim it did not argue: that the choice of interpretation matters beyond taste. The argument is a finding, and it is one the record had already made without the literature.

The lore page's title is Hartle's phrase. Generalized quantum mechanics was James Hartle's name, in lectures of the early 1990s, for the decoherent-histories framework with its Hilbert-space specifics loosened, so that ordinary quantum mechanics becomes one instance of a wider theory rather than the only possible one. Rafael Sorkin gave the loosening its sharpest form in 1994. Quantum mechanics, he showed, can be written as a measure on histories obeying a hierarchy of sum rules. Classical probability obeys the second: no interference between two alternatives. Quantum mechanics breaks the second and obeys the third: whatever interference three alternatives show is already accounted for by their pairs. A theory that broke the third would be no less lawful, only not ours. In 2010 Urbasi Sinha and colleagues tested the third rule with three slits and found it held to their precision. The reconstructions of quantum theory from axioms that followed, Lucien Hardy's in 2001 and Jonathan Barrett's framework of 2007, showed something else the record relies on: the rule for composing independent systems is a postulate of its own, and a theory may agree with quantum mechanics on every single system and differ on how systems combine.

That is Counterpane, exactly. An atom is an atom there, the second is the second, and the joint statistics of three or more independent sources are not ours. One precision, because a loosely true claim is a misunderstanding waiting to happen: Sorkin's third rule concerns three alternatives open to one system, and Counterpane's finding concerns three independent systems combined, which is the composition axis and not the interference one. The two are separate postulates in the reconstructions, and the record's neighbour varies the second while keeping the first. The page was written as a scoped hedge made literal, with none of these names in view; it describes a point in a theory space physicists have mapped, on the axis the reconstructions say is separable. And it is why the interpretation matters. An account that makes the universal wavefunction the furniture of the world has no room for a neighbour with a different kernel. An account that treats our quantum mechanics as one lawful measure among those a mathematical regime permits has exactly that room, and the record's Metawrights and Archewrights already stand where the measure is chosen. Consistent histories is not only the interpretation the Cascade's rules select; it is the one whose own founders built the door Counterpane walks through.

The M-theory side is geometry, borrowed as shape. The home floor of three extended dimensions, seven compact and one of time is the eleven of M-theory, established as such by Edward Witten in 1995, which reaches four dimensions by compactifying seven. In Petr Hořava and Witten's 1996 picture the eleventh dimension is an interval with a boundary at each end; matter lives on the boundaries and gravity alone crosses the bulk, which is what the record's Interval, and its rule that only gravity registers across a narrowed one, describe. The landscape of vacua that Raphael Bousso and Joseph Polchinski counted in 2000, and Leonard Susskind named in 2003, is Max Tegmark's Level II multiverse of the same year, different constants in different regions. The Ensemble is that, and not his Level III of Everett branches. None of this is confirmed physics, and lore asserts none of it; the record takes the geometry and marks it as ours.

That is the pattern the energy entry named: take real physics as far as it goes, say plainly where the deviation is, and put the deviation where the real physics is already thinnest. Here the physics runs further than expected. Whether quantum mechanics had to be the theory it is remains open in our world, and the programme that asks it is the same one the record's neighbour answers.

References: R. D. Sorkin, "Quantum mechanics as quantum measure theory", Modern Physics Letters A 9 (1994); U. Sinha, C. Couteau, T. Jennewein, R. Laflamme and G. Weihs, "Ruling out multi-order interference in quantum mechanics", Science 329 (2010); L. Hardy, "Quantum theory from five reasonable axioms" (2001); J. Barrett, "Information processing in generalized probabilistic theories", Physical Review A 75 (2007); E. Witten, "String theory dynamics in various dimensions", Nuclear Physics B 443 (1995); P. Hořava and E. Witten, "Heterotic and Type I string dynamics from eleven dimensions", Nuclear Physics B 460 (1996); R. Bousso and J. Polchinski, "Quantization of four-form fluxes and dynamical neutralization of the cosmological constant", Journal of High Energy Physics (2000); L. Susskind, "The anthropic landscape of string theory" (2003); M. Tegmark, "Parallel Universes", in Science and Ultimate Reality (2003).