Truong, Q. H. X.; Truong, X. K.

The origin of life remains fragmented across disciplinary boundaries, with competing models emphasising genetic replication (RNA-world) or metabolic networks (metabolism-first) without a unifying physical principle explaining why either should emerge from non-living matter. We propose a theoretical framework combining Entropy-Oriented Mechanics (EOM) and Information Force Fields (IFF) that connects Big Bang nucleosynthesis to biological emergence without invoking biological selection a priori. EOM is formalised through the local entropy production rate Sigma(x), defined rigorously via probability flux in a driven Markov chain. IFF introduces the information quasi-potential Phi_I(x) = -ln p*(x), tied to the non-equilibrium stationary distribution of the driven system. We show that the gradients of Sigma and Phi_I are generically linearly independent off equilibrium (Theorem 1), so that the two gradient fields act independently in the combined Langevin dynamics. England's dissipative-adaptation framework develops a single-field description centred on dissipation; the present two-field formulation additionally yields a non-monotonic relationship between entropy flux and polymer yield (scaling Phi* proportional to Delta-Phi_I / sqrt(2D), R^2 = 0.885 on the shock-synthesis data of Blank et al. 2001), which a single-field framework does not address. Within EOM-IFF, we identify a cascade of information attractors (defined via Kramers-Eyring escape times) spanning nuclear (C-12, N-14), chemical (amino acids, nucleobases), and polymeric scales. Life is defined formally as a dynamical regime satisfying three simultaneous conditions on the effective quasi-potential landscape, independent of molecular substrate. Carbon-based life is viewed here as a statistically favoured outcome of entropy-driven quasi-potential deepening in carbon-nitrogen chemistry, rather than a singular improbable event.