The question
The record reads a system’s architecture from its coupling. Mutual coupling makes peers, one-way drive makes a hierarchy (CLAIM-172, CLAIM-167). Often only an outcome is observed: units in phase, or a shared quantity falling. The question is what must be measured to tell common drive from coupling. A second part asks how to tell one route to an outcome from another.
Why the record needs it
A timing test that sees only phase alignment would pass units that are synchronized by a common cause and never coupled. CLAIM-172’s criterion requires that “each side’s timing shifts with the other’s”. That calls for an intervention on one member, which the record has not specified. QUESTION-033 wants drive apportioned among members. An outcome reached by several routes cannot apportion drive until the members are measured separately. These are direct cases of multi-unit biological systems, the kind QUESTION-035 must handle. They are not analogies.
What the record holds
- THEORY-129: in coral spawning, colonies spawn together because each responds to “the natural night-light regime”. Displacing it “delays or abolishes” release. It “shows synchrony by a shared cue, not coupling between colonies”. Limit: one species, small groups, and coupling is untested, not excluded.
- THEORY-130: coral bleaching is “reached by more than one route”. The same fall in symbiont density came from fewer symbionts per host cell in one species and from the loss of host cells in another. “Only measuring host and symbiont together showed it.” Limit: reviews and one octocoral study. It says the exchange is regulated from more than one side, and “nothing about agency follows”.
- THEORY-275 (LIT-947, NOTE-788): between two symmetrically coupled colloids, directional information measures follow the path of a recorded drive, not the symmetry of the coupling. Under strong drive of one unit a reciprocal coupling reads as one-way flow. It answered the question in part, negatively: an observational test of direction can misreport the coupling. That part is now the child QUESTION-263 (Answered, 2026-10-23). Limit: the measures follow a recorded drive on one unit. No shared unrecorded drive is tested, which is this question’s case.
- THEORY-276 (LIT-960, NOTE-775): in a network of identical units, the orbits of the graph’s symmetry group are the synchronization clusters, and one cluster can lose synchrony while the others keep it. Limit: the paper never mentions external drive. Reading such a cluster pattern as a coupling signature that a shared drive does not predict is the record’s inference, valid only for near-identical units on a known graph.
- LIT-933 (Nowak et al., NOTE-768) is an instance of the conflation this question names. It counts synchrony “through their mutual influence or through their common response to an external signal” as the same functional unit.
What would answer it
An intervention design for the record’s candidate systems that separates the possibilities. Displace the shared drive for one member only and see whether the others follow. Perturb one member and measure the others. With that, a statement of which of QUESTION-035’s tests can and cannot see the difference.
Note of 2026-10-23: a human dyad with both routes manipulated
Bigand et al. (2024, LIT-1424, read in NOTE-1093), cited at A5 of the exchange of 2026-10-23, “Framework Overview Request”, ran in pairs of dancers the separate interventions this question asks for. They removed the shared drive (the same music, or music shifted 8.5%) and the coupling channel (a curtain) independently. Each raised synchrony in different movement components, and the authors found “no evidence of an interaction between musical input and visual contact, and this was so for all PMs”. Timing could not tell the two apart: “when (unlike where) participants move does not reflect whether synchrony is established through music or partner vision.” Filed as CASE-149 and THEORY-891 (Proposed), which answers this question in part.
What stays open: the perturb-one-member intervention, so direction and peers against hierarchy; an unrecorded shared drive, since here the drive was the experimenters’ known music; and the question’s second part, telling one internal route from another. The absence of an interaction rests on a null test, not an equivalence test.
Note of 2026-10-26: shared disturbances against organization in a simulated firm
A second setting. The Northstar simulation reported at T20 of the exchange “Experimental Infrastructure Review” (CASE-166; reported as “PerspectiveLab v2.11” and not reproduced, since dmarx/PerspectiveLab holds no v2.10 or v2.11) fitted overlapping activity modes to employee trajectories. The reply said they “could reflect shared environmental disturbances rather than genuine higher-level interpretive organization”. T22 named the confound again: “correlated activity may be caused by common external events. Employees responding to the same incident don’t necessarily constitute one emergent perspective.”
A check made when CASE-166 was filed gives this question’s shared drive in that setting: twelve employees with no communication, each seeing two of three shared disturbances, give three clear correlation modes, each spanning all four departments. T22 proposed interventions of the kind this question asks for: team rewiring, incident-command handoff, departure of participants and a change of authorization pathway, each scored against what the communication and authorization graph already predicts. CASE-101’s first population is the confound’s constructed form. None of these interventions has been run. QUESTION-276 asks the question for an organization’s inferred perspectives. The question is unchanged.
Note of 2026-10-27 (evening): the readings of the U15–U23 works
Stewart, Golubitsky and Pivato (NOTE-1217) treat drive and coupling on one footing. A drive is just another cell, and balance counts inputs whoever sends them. In Pecora and Carroll’s master–slave network (Fig. 19, p. 635), cells that are only driven are robustly synchronous with cells that also drive back. In Example 5.1 (p. 626), cells with no arrows among them are robustly synchronous. So a pattern forced by the architecture means “equal input”, not “coupled to each other”. A cluster is evidence of coupling only relative to a graph that includes every source of input. That is the record’s reading of the paper, which never discusses drive.
Note of 2026-10-31: shared cue or spillover, in two readings of affect
Two readings filed today disagree on this question’s distinction, for affect in groups, and neither measures it.
- Spillover, which is coupling. Bencherki, Brummans and Vézy describe one person’s anger as augmenting others’ ability to act, “as her anger spills over into her inert colleagues and turns them into active strikers” (LIT-1902, p. 1354), and agency as one process taking up “the excess action of another” (p. 1351). That is influence passed between members.
- Shared cue, differential response, which is common drive. Massumi (LIT-1905, pp. 6, 9; THEORY-1286): affect “is not a content that can be transported from one event to another”; bodies cued by the same event, an alarm, “will have been attuned – differentially – to the same interruptive commotion”, “responding differently together”. Uniform response, which looks like contagion, needs “other factors” that “pre-channel tendencies” (p. 6). That is this question’s shared external drive, with the members’ differences kept.
Massumi’s alarm is THEORY-129’s coral, for people. The record’s test is the one CASE-149 ran in dancers: remove the shared cue and the channel between members independently. Neither reading proposes it. A collective emotion read off synchrony alone (CLAIM-355’s “integrated regulatory pattern”) would not tell the two apart.
The question is unchanged and stays Open.