The case
A12 §7 stated the three arms:
- A, common input. “Individuals have little interaction, but all respond to the same repeating external events. The aggregate is highly predictable because the environment is predictable.”
- B, reproducing organization. “Individuals inherit norms from previous members, modify them through interaction, and transmit them onward.”
- C, institutionally maintained. Individuals turn over and stay coordinated “because an enduring external institution selects, trains, and governs them.”
The arms are built so that observable aggregate statistics and turnover rates match. A11 §9 had a third arm of its own instead of C: an original population that splits in two after a relocation (CASE-105).
What it would show
Where the criteria disagree. A12: a common-input population “might display high observational persistence but little endogenous historical organization”; an institutionally maintained one “might exhibit a highly effective macrostate while its boundary must include the external governing institution”; a reproducing one “might show strong genealogical continuity while lacking exact dynamical independence”. The criteria to compare are Krakauer et al.’s A*, A and nC (LIT-1137, unread), dynamical independence (THEORY-257), effective information (THEORY-222) and CLAIM-157’s priced block.
The record already holds the answer for arm A on one criterion, which the exchange did not cite: Bourrat (LIT-005, NOTE-027) argues that an environment that fully determines S makes A* positive for “a reconstructed entity”, and that only an interventional  scores it zero. So the study should add Â, with the caution NOTE-027 records: Bourrat’s claim that  equals Krakauer’s A is false in general. Arm C is CLAIM-194’s made-versus-imposed boundary in population form.
A negative result is informative too. A11 §9: if the measures cannot tell the arms apart, “temporal information propagation alone is insufficient to recover the kind of historical social individuality that fandom seems to exhibit.” Fictional by construction; nothing is run.
Note of 2026-10-23: a precedent for arm B
A5 of the exchange of 2026-10-23, “Framework Overview Request” cited Jacobs and Campbell (1961, LIT-1423, Unavailable) for this design’s transmission arm: a norm surviving the replacement of all who set it. A5: CASE-101’s contribution “should concern discriminating mechanisms of persistence, not demonstrating intergenerational transmission itself.” CASE-148, perspective transplantation, is the extension A4 proposed.
Note of 2026-10-29: Krakauer et al. read, and the split into S and E
LIT-1137 is now read (NOTE-1400), so “unread” above no longer holds. The reading confirms that arm A, common input, is where A* fails, as Bourrat argued and THEORY-424 holds. It adds a condition on the design. A conditions on the environment, so it scores the common-input arm near zero only if the common input is among the environment’s variables. All three of Krakauer et al.’s measures are relative to the split between system and environment, so the study should state which variables it puts in E for each arm. The proposal is otherwise unchanged.