What makes a signal meaningful? We show: meaning arises in the gap between structure and randomness — in the open interval C ∈ (0,1) of the MDL-ratio, where a receiver can partially but never fully compress a signal. We develop a formally grounded theory based on four ontological axioms, seven MDL axioms, and eight theorems (five core + three extensions on thermodynamics, learnability, and universal communication). The 1/f spectrum (pink noise) is the only scale-invariant carrier that simultaneously maintains a position in the interior of this gap across all time scales. Four independent measures confirm this empirically. Section 10 (Universal Communication Architecture): The framework is generalised to all conforming receivers — biological, artificial, or hypothetical — including AI-to-AI communication, SETI, and human-machine interaction. v17 additions: Generative encoder (messages → 1/f signal, C_spec = 0.749, BER = 0 up to SNR 3 dB); extended CTW analysis; Banach fixed point refined; Landauer coupling corrected. Independent confirmations: Chong & Feng (2024, Chaos, AIP): 1/f self-organisation in deep neural networks. Zhang et al. (2025, arXiv): MDL minimisation → explainable representations ("To understand is to compress"). This package contains: main paper (DE + EN), v17 supplement, Entropy Dynamics companion, narrative book excerpt, and all simulation code (reproducible, fixed seeds). Axiomatica Universalis — Beyer 2026f