Blind Identification of Social-System Morphology: A Structural Validity Test of the Complex Adaptive Model of Societies

Abstract

The Complex Adaptive Model of Societies (CAMS) posits that human collectives can be represented as eight-node institutional networks whose dynamics encode recoverable information about social-system morphology. This paper reports a blind identification experiment testing whether anonymised CAMS panels contain sufficient society-specific structural information to permit recovery of national identity from trajectory shape alone. Three blind sets comprising twenty-four societies (labelled only as Society_A through Society_H) were analysed by a protocol-bound analyst denied all labels, dates, and geopolitical context. The analyst recovered exact identity in 13 of 24 cases (54.2%; P ≈ 1.2 × 10⁻⁶ against eight-candidate forced-choice null) and functional family in all 24. All eleven misses were within-family; none was random. The error structure, initially hypothesised to reflect a latent morphospace in the panels, was tested and refuted: errors tracked the analyst's historiographic narrative templates rather than feature-space proximity. The panels nonetheless volunteered genuine structural adjacencies (Australia–Norway, UK–USA, Thailand–Hong Kong, Russia–China) in a label-free embedding. Pre-registered nulls behaved as nulls (feedback loops, coupling–viability divergence, most synchrony), one pre-registered positive fired on exactly one society (Sweden's armed-neutral compensation signature), and the protocol's discipline clauses (magnitudes quarantine, carry-forward gating, null reporting) operated as designed. The results establish CAMS panels as carrying recoverable social-system morphology while specifying the replication programme required to demonstrate scorer-independent measurement.

Keywords