Symbol Emergence in a Sovereign Software Organism: Cross-Situational Word Learning Without Supervision on Commodity Hardware

We present KENTA, a symbolic software organism running offline on a 2012 Intel i5, in which two words emerged autonomously on March 21, 2026, through cross-situational co-occurrence of acoustic patterns and internal system states. No dataset, pre-training, GPU, or cloud dependency was used. The mechanism follows Mangin's 2014 developmental robotics framework: repeated co-occurrence of the same frequency signature in a stable physiological context produces a grounded symbol. Internal states are measured through a bio-inspired homeostatic layer mapping system guardians to meridians with explicit metric thresholds — functioning as an embodied sensorimotor ground. Convergent validation is provided by TQNM (Triadic Quantum Nonlinear Memory, McMurray 2026), which independently satisfies the same six digital life signs.