The paper argues that human speech evolved as an elaboration of ancient non‑verbal "honest signals"—mimicry, activity, influence, and consistency—that coordinated early hominin groups. It defines these four signal components, explains their neurobiological bases (mirror neurons, autonomic nervous system), and describes how they can be measured using custom sociometric badges that capture audio and coarse body‑language features. The author presents several empirical studies: (1) business‑plan pitch experiments where signal patterns predicted plan selection; (2) speed‑dating, job interview, and salary‑negotiation outcomes predicted with ~80 % accuracy using only signal data; (3) a Mission Survival task where audio‑visual signal features classified social and task roles with accuracies comparable to human raters; (4) turn‑taking prediction using activity and influence models; (5) a large‑scale group‑performance study showing that aggregated signal features (high activity, frequent back‑channeling, balanced turn‑taking) correlate strongly (r≈0.7) with collective intelligence scores across diverse tasks. The findings support the hypothesis that semantic role structures and group problem‑solving processes run in parallel with unconscious signaling mechanisms. The paper concludes that signaling constitutes a "para‑semantic" channel and can be harnessed in real time via the "meeting mediator" system—badge plus mobile visualization—to improve group collaboration, increase activity and back‑channeling, and boost performance.
Topics
Evolutionary Foundations of Honest Signals and Speech
Explores how pre‑linguistic non‑verbal signals (dominance, interest, warning) provided a coordination framework that later gave rise to human speech, positioning speech as an elaboration of these ancient mechanisms.
Four Core Honest Signals: Mimicry, Activity, Influence, Consistency
Defines the four signal components studied: mimicry (reflexive copying of facial/gestural cues), activity (energy/excitement level), influence (extent a speaker shapes another's speaking pattern), and consistency (fluidity of speech/movement perceived as expertise).
Neurobiological Underpinnings of Honest Signals
Links each signal to brain structures: mirror neurons for mimicry, autonomic nervous system for activity, and discusses how these ancient systems relate to mental health (e.g., depression) and to clinical applications.
Measuring Unconscious Signals with Sociometric Badges
Describes the custom electronic badges that capture audio amplitude, pitch, pauses, and coarse body motion; outlines algorithms for utterance parsing, back‑channel detection, and statistical influence modeling.
Influence of Signals on Decision Outcomes
Presents experimental evidence that signal patterns alone predict business‑plan selection, speed‑dating matches, job interview success, and salary negotiations with ~80 % accuracy, highlighting the predictive power of non‑verbal cues.
Parallelism of Signaling and Semantic Role Classification
Details the Mission Survival task study where audio and visual signal features classified social (protagonist, supporter, neutral) and task roles (information giver, seeker, etc.) with accuracies comparable to human raters, supporting the hypothesis of parallel signaling and semantic structures.
Influence Modeling for Turn‑Taking and Role Prediction
Explains the Hidden Markov Model‑based influence model that captures conditional dependencies among speakers to predict turn‑taking order and role assignments, achieving human‑level performance.
Signaling Correlates of Collective Intelligence and Group Performance
Analyzes data from 51 groups across four tasks, showing strong correlations (r≈0.7) between aggregated signal features (high activity, frequent back‑channeling, balanced turn‑taking) and group performance scores, indicating a "para‑semantic" performance predictor.