Applied ITT — Executable Physics I: Articulatory Physics as Semantic Primitive

Applied ITT — Executable Physics I: Articulatory Physics as Semantic Primitive Armstrong Knight (Sensei Intent Tensor) · intent-tensor-theory.com Contemporary word embedding models derive lexical geometry exclusively from distributional co-occurrence statistics. We propose that lexical properties — including the function/content word distinction and word frequency distribution — are predictable from articulatory physics alone, without corpus statistics. We formalize this through Molecular Weight (MW), a scalar quantity derived from a nine-dimensional articulatory feature space representing vocal tract configuration. We demonstrate empirical correlation ρ = 0.88 between MW and observed word frequency, and derive a zero-corpus function word classifier from the MW threshold alone. The articulatory framework generates five distinct emergence layers (L0–L4) of the ITT 10-layer semantic stack. Key claims: MW(w) = 10(1+3.5κ)²/(1+4σ) predicts word frequency with ρ=0.88 from vocal tract anatomy alone. Five levels of linguistic organization derivable from one nine-dimensional physical representation. Zero corpus statistics required. Part of the Applied ITT — Executable Physics series (WP-01 through WP-08). Parent theory (Astrosynthesis series):Vol I: https://doi.org/10.5281/zenodo.19328544Vol II: https://doi.org/10.5281/zenodo.19363000Vol III: https://doi.org/10.5281/zenodo.19363002 Repository: https://gitlab.com/intent-tensor-theory.com-group/git-0-0-applied-intent-tensor-theoryWebsite: https://intent-tensor-theory.com