Abstract: This paper introduces the Energetic Budget of Speech Systems (EBSS), a predictive, computational model that frames historical language change as a thermodynamic optimization problem constrained by human metabolic power. Traditional historical linguistics relies on qualitative taxonomics to describe structural shifts; by contrast, EBSS formalizes these mutations as system-wide re-optimizations balancing biomechanical vocal tract power, neurological processing strain and network error-correction loops. To evaluate the model, the paper presents an algorithmic simulation comparing two contrasting 500-year historical horizons: the high-velocity, open system of Middle English undergoing massive demographic admixture, and a low-velocity control group using the geographically insulated, endogamous system of Old-to-Modern Georgian. The simulation demonstrates that under high contact and low endogamy, a linguistic network experiences an exponential morphosyntactic software collapse, forcing syntactic fixity to prevent communicative failure while keeping total power below the critical 20-Watt metabolic ceiling. Conversely, the insulated control model confirms structural retention, proving that linguistic preservation is as thermodynamically predictable as simplification. These results demonstrate that language evolution can be modeled and verified as an emergent biological optimization event driven by generational processing bottlenecks. Keywords: Computational Linguistics, Evolutionary Biology, Information Theory, Language Evolution, Thermodynamic Optimization, Agent-Based Modeling, Neurobiology of Language.
Dan Vasiliu (Wed,) studied this question.