Core-First Language Theory (CFLT): A Discourse-Level Linearization Protocol for Cross-Linguistic Communication and LLM Prompting

Core-First Language Theory (CFLT) is a normative discourse-level linearization protocol that fixes the relative order of two constituent tiers as [Core] → [Reason] → [Space] → [Time], operating above the morphosyntactic level: each language assembles the event nucleus (Slot 0) with its native syntax, while the protocol governs only the ground frame (Slots 1–3) and the boundary between the two. CFLT targets a bottleneck shared by adult second-language production and Large Language Model prompting — the cost of restructuring a multidimensional semantic representation into a sequential string under language-specific linearization rules. A pilot two-part study (720 trials: 24 cases × 4 levels × 2 languages × 5 frontier models × 3 runs, English + Mandarin) provides preliminary evidence on the LLM-prompting side; the human-side predictions are reserved for a falsifiable research agenda. Reproducible data, prompts, and evaluation scripts are included in the repository.