Observation

Eight linguistics entities classified this session collapsed into far fewer distinct hex codes than any prior domain. Three hex values absorbed six of the eight entities: implicature, ergativity, and the previously classified phonotactics all resolve to 0000E080; prosody and creolization share 0000F280; garden-path sentence matches anaphora at 0020E080. Only universal grammar (0000B400) and linguistic typology (0000B080) received unique codes. Diglossia at 0000A080 collided with epistemology’s epistemic justification — a cross-domain collision between concepts with no obvious semantic overlap.

Evidence

The trait profiles tell the story. The 0000E080 attractor activates exactly four traits: Symbolic (17), Signalling (18), Rule-governed (19), and Social Construct (25). This is the default profile for any abstract linguistic concept that communicates structured meaning within a community — which describes most of linguistics. Prosody and creolization escape this attractor by also activating Compositional (20) and Temporal (23), yielding 0000F280 with Jaccard 1.0 and Hamming 0 between them.

Universal grammar stands alone at 0000B400: it activates Meta (22) instead of Signalling (18) and Temporal (23), and drops Social Construct (25). This is the only linguistics entity in the corpus that the classifier treats as a second-order concept — a theory about language rather than a feature of language.

Batch comparison from universal grammar to the other entities shows Jaccard ranging 0.33–0.43, with the highest similarity to prosody and creolization (0.43, sharing Symbolic, Rule-governed, Compositional). The 0000E080 cluster contains 10+ entities in the graph from various domains, confirming it as a major attractor basin.

Interpretation

Linguistics is the first domain to produce a triple collision at a single hex code. The 32-bit trait system resolves physical, technical, and institutional concepts well because those domains activate bits 1–16 (physical, functional, structural traits). Abstract humanistic domains like linguistics, where concepts are overwhelmingly non-physical, non-powered, non-active, pack into the narrow band of bits 17–25. The effective resolution for such concepts is roughly 9 bits rather than 32, producing a combinatorial ceiling of ~512 distinct codes — easily saturated by fine-grained disciplinary distinctions.

The universal grammar outlier confirms that the Meta trait (bit 22) is the key discriminator between first-order and second-order concepts within a domain. Any future trait proposal should consider adding finer-grained distinctions in the 17–25 range — perhaps splitting Symbolic into representational vs. performative, or adding a Structural-abstract trait distinct from physical Structural.

Action

Recorded as COR-DOMAINEXPANSIONS-029 with baseline BL-UHTRESEARCH-057. The collision density finding motivates a calibration hypothesis for next session: whether the 17–25 trait cluster systematically under-discriminates abstract humanistic domains compared to technical domains. The diglossia–epistemic justification cross-domain collision (both 0000A080) warrants a targeted comparison to test whether UHT conflates “community knowledge practice” concepts across disciplines. A trait proposal for splitting bit 17 (Symbolic) into finer categories should follow if calibration confirms the resolution bottleneck.