Documentation

Linglib.Phenomena.Gradability.Studies.Rouillard2026

Cross-Domain Licensing Agreement Bridge #

@cite{champollion-2017} @cite{kennedy-2007} @cite{krifka-1989} @cite{krifka-1998} @cite{rouillard-2026} @cite{zwarts-2005} @cite{vendler-1957}

The showcase theorem: six independent classification systems from different linguistic subfields all agree on licensing predictions, because they all route through LicensingPipeline.isLicensedBoundedness.isLicensed.

Compositional Chain Architecture #

MereoTag (QUA/CUM) serves as the canonical bottleneck — all binary domain classifications route through it, and MereoTag.toBoundedness is the single point where the licensing prediction is made:

VendlerClass ─.telicity─→ Telicity ─.toMereoTag─→ MereoTag ─.toBoundedness─→ Boundedness
PathShape ─pathShapeToTelicity─→ Telicity ─.toMereoTag─→ MereoTag ─.toBoundedness─→ Boundedness
SituationBoundedness ─.toMereoTag─→ MereoTag ─.toBoundedness─→ Boundedness
BoundaryType ─.toBoundedness─→ Boundedness (direct — concepts are identical)
EpistemicTag ─.toBoundedness─→ Boundedness (direct — weak mereological link)

Each arrow encodes an independently motivated theoretical insight:

The Six Sources #

SourceTypeClosed variantModule
Scale boundednessBoundedness.closedCore/Scale
Mereological tagMereoTag.quaCore/Scale
TelicityTelicity.telicEvents/Aspect
Vendler classVendlerClass.accomplishmentEvents/Aspect
Interval boundaryBoundaryType.closedCore/Time
Situation boundednessSituationBoundedness.boundedCore/Time

All six have LicensingPipeline instances mapping their closed/open variants to Boundedness.closed/.open_, yielding identical licensing predictions.

The epistemic probability scale is closed [0,1] (@cite{kennedy-2007}), so epistemic adjectives like "certain" license degree modification ("completely certain") for the same structural reason that "full" does — closed-scale boundedness routes through Boundedness.closed → isLicensed = true.

This connects the seventh licensing source (epistemic modality) to the existing six, demonstrating that the LicensingPipeline unification extends beyond the adjectival and verbal domains into epistemic modal semantics.

Epistemic probability licensing agrees with Kennedy adjective licensing: both are Boundedness.closed → isLicensed = true. "completely certain" is licensed for the same reason as "completely full".