Conchordal: Emergent Harmony via Direct Cognitive Coupling in a Psychoacoustic Landscape

Abstract

This paper introduces Conchordal, a bio-acoustic instrument for generative composition whose sonic agents are governed by artificial life dynamics within a psychoacoustic fitness landscape. The system is built on Direct Cognitive Coupling (DCC), a design principle requiring that generative dynamics operate directly within a landscape derived from psychoacoustic observables and read from that landscape without symbolic harmonic rules. The environment integrates roughness and harmonicity into a continuous consonance field without presupposing discrete scales or explicit harmonic rules. Agents adjust pitch through local proposal-and-accept dynamics under a crowding penalty, regulate survival via consonance-dependent metabolism, and entrain temporally through Kuramoto-style phase coupling. Four experiments are reported: (1) consonance search produces structured polyphony with enriched consonant intervals; (2) consonance-dependent metabolism yields survival differentials that vanish when recharge is disabled; (3) a minimal hereditary adaptation assay shows that parent-guided respawn plus metabolic selection can accumulate more structured polyphony without adult hill-climbing; and (4) a shared oscillatory scaffold organizes rhythmic timing under external forcing. A supplementary mechanism check reports one possible composer-configurable bridge by which spectral state can modulate temporal coupling. These findings show that a psychoacoustically derived landscape serves as an effective artificial-life terrain, yielding self-organization, selection, synchronization, and lineage-level accumulation in a non-traditional computational medium. At the level of the model, the same landscape therefore functions both as ecological terrain and as an internal proxy for musical coherence.

Figures (5)

• Figure 1: Psychoacoustic landscape around 220 Hz. (A) Harmonicity $H_{01}$; (B) roughness $R_{01}$; (C) consonance field $C_{\text{field}}$; (D) consonance density $C_{\text{density}}$. Grey: 0.5-oct grid; dashed: integer ratios. Simple-ratio peaks in $H$ and troughs in $R$ create attractor structure in both views. Reduced density outside $\pm 1$ oct is a finite-order property ($N\!=\!16$).
• Figure 2: Consonance search. Top---(A) Mean LOO consonance $C_{\text{score}}$ over time (95% CI); vertical line = phase switch, shaded band = burn-in excluded from statistics. (B) Scene consonance $G(F)$ (Eq. \ref{['eq:consonance-core']}) after singleton subtraction (95% CI). (C) Representative local-search trajectories (cents from drone). (D) Final unique pitch bins (95% CI). Bottom---(E) Pairwise interval histogram (25 ct bins): local-search concentrates mass at consonant intervals; random-walk spreads broadly. (F) Entropy of the 5 ct interval distribution (95% CI).
• Figure 3: Consonance as selection pressure. (A, B) Kaplan--Meier survival by early consonance ($C_{\mathrm{firstK}}$) median split: with recharge, high-consonance agents survive longer; without recharge, the separation disappears. (C, D) Lifetime versus early consonance shows the same contrast. Curves pool events across 20 seeds for display; inference is seed-level ($n=20$).
• Figure 4: Hereditary adaptation through lineage-biased respawn and metabolic selection (20 matched seeds, 4 conditions). (A) Mean leave-one-out contextual $C_{\text{score}}$ over simulation steps (95% CI). (B) Pitch heatmap for heredity$+$selection over simulation steps. (C) Final leave-one-out contextual $C_{\text{score}}$ by condition (95% CI). (D) Final just-intonation proximity score by condition (95% CI). Teal denotes heredity and rose denotes matched random; dark colors indicate selection-on and light colors indicate no-selection.
• Figure 5: Temporal scaffold assay (20 seeds $\times$ 3 conditions). (A) Group mean PLV to the canonical 2 Hz (120 BPM) beat over time (95% CI). (B) Seed-averaged onset phase histogram relative to the canonical beat (bin probabilities normalized within seed, then averaged across seeds; shaded bands indicate 95% CI across seeds). (C) Seed-level onset vector strength: shared rhythm produces the strongest phase concentration, scrambled forcing is weaker, and off is lowest.