REJ:2026.07.31.0001 soc.TH (Social Thermodynamics) Rejected
No Evidence of Person-to-Person Transmission in Cultural Cascades: The SEED-2 Randomised Sharing Trial, with a Common-Source Reanalysis of 4.1 Billion Adoption Events (1897–2025)
Fenella Aguirre, Kwabena Otoo, Ilkka Väisänen, S. Marilena Prokopiou, Hanh T. Duong, Rufus Ekwueme & The SEED-2 Investigators
Comments: 7 pages, 3 figures, 0 reproducible results. Rejected in 22 minutes on Jul 31, 2026.
Abstract: Cultural items are said to spread from person to person. The claim rests on curve-fitting. Adoption curves are S-shaped, contact-transmission models produce S-shaped curves, and reproduction numbers above unity are reported on that basis. A common-source model containing no contact term produces the same curves. We therefore randomised the act of telling someone something. In SEED-2, 21,088 sender–recipient dyads were allocated 1:1 to share or to withhold a designated item, and recipient adoption at 30 days was ascertained by blinded assay. Adoption was 11.4% in the share arm and 11.1% in the withhold arm (risk difference 0.30 percentage points, 95% CI -0.57 to 1.17 , p = 0.51 ). There was no dose–response across 1, 3 and 9 transmissions and no heterogeneity across nine item strata. The implied reproduction number for person-to-person transmission is R 0 = 0.043 (95% upper limit 0.17 ), against 2.31 (95% CI 2.18 – 2.44 ) from conventional observational fitting. In a parallel audit, 71.3% of adopters named a person as their source; 96.1% of those named had no contact capable of transmitting the item, and 62.8% adopted it after the respondent did. Registries should not record a named source as a transmission event.