Features.LicensingContext #
@cite{ladusaw-1979} @cite{kadmon-landman-1993}
The 22-case enum of licensing contexts for polarity-sensitive items. Theory-neutral data substrate: every framework that talks about polarity licensing (Ladusaw monotonicity, K&L domain widening, Israel scalar model, Chierchia exhaustification, Giannakidou nonveridicality) needs to talk about these context labels.
Provenance #
Extracted from Theories/Semantics/Polarity/Licensing.lean after
audit consensus that LicensingContext is data, not theory: the
22 cases are observational labels for syntactic environments, while
only the signatures assigned to them in contextProperties
(in the Theories file) carry theoretical commitment. Co-locating the
enum here makes it Fragment-importable substrate without dragging in
the Ladusaw/K&L theoretical apparatus.
The earlier 2-way split (Theories/Polarity/Licensing.lean for the
enum + theory together; Typology/PolarityItem.lean for entry +
Israel) forced Typology/PolarityItem.lean to import from
Theories/, the only such cross-layer import in linglib. This file
breaks that inversion: Typology/ and Theories/Semantics/Polarity/
both import from Features/, no peer-layer crossings.
Framework commitment #
The 22-case enum is theory-laden in its naming (constructor names
like adversative, doubtVerb, denyVerb, comparativeS reflect
specific traditions' classification of contexts), but the cases
themselves enumerate empirically-attested licensing environments any
framework needs to talk about. The DE/anti-additive/anti-morphic
labelling of these contexts is theory-laden and lives in
Theories/Semantics/Polarity/Licensing.lean::contextProperties
(Ladusaw/Zwarts canonical) — not here.
UNVERIFIED: The 22-case carve-up is English-anchored and may need
cross-linguistic restructuring (e.g., factor by veridicality + monotonicity
per Giannakidou 1998, rather than by surface construction); see the
Theories/Semantics/Polarity/Licensing.lean "Out of scope" section
for the documented gap.
Contexts that can license polarity-sensitive items.
Characterized by their logical properties:
- DE (Downward Entailing): reverses entailment direction
- Anti-additive: DE + distributes over disjunction
- Anti-morphic: anti-additive + distributes over conjunction (= negation)
Per-context theoretical classifications (DE strength, K&L mechanism,
Strawson-DE flagging) live in
Theories/Semantics/Polarity/Licensing.lean::contextProperties.
- negation : LicensingContext
- nobody : LicensingContext
- few : LicensingContext
- atMost : LicensingContext
- conditionalAntecedent : LicensingContext
- beforeClause : LicensingContext
- withoutClause : LicensingContext
- onlyFocus : LicensingContext
- question : LicensingContext
- comparativeNP : LicensingContext
- comparativeS : LicensingContext
- superlative : LicensingContext
- tooTo : LicensingContext
- modalPossibility : LicensingContext
- modalNecessity : LicensingContext
- imperative : LicensingContext
- generic : LicensingContext
- adversative : LicensingContext
- sinceTemporal : LicensingContext
- freeRelative : LicensingContext
- universalRestrictor : LicensingContext
- doubtVerb : LicensingContext
- denyVerb : LicensingContext
Instances For
Equations
- Features.instDecidableEqLicensingContext x✝ y✝ = if h : x✝.ctorIdx = y✝.ctorIdx then isTrue ⋯ else isFalse ⋯
Equations
- One or more equations did not get rendered due to their size.
Instances For
Equations
- Features.instReprLicensingContext = { reprPrec := Features.instReprLicensingContext.repr }