Attitude predicates: the classification #
This file is the root of the attitude API: the semantic classification a clause-embedding
predicate's lexical entry records, from which its combinatorial properties are derived. A
doxastic predicate is veridical or not, the [karttunen-1971] lineage of factive and non-factive
complement-taking verbs (Doxastic.Veridicality). A preferential predicate has an evaluative
valence, positive for hope and negative for fear (Preferential.Valence), and a
compositional strategy (Preferential.Strategy): the degree comparison of [villalta-2008],
whose question use is the existential over answers, or a relation to the question itself,
anxious uncertainty for worry ([anand-hacquard-2013]) or anticipation of resolution for
Mandarin qidai and relevance for care ([elliott-etal-2017]), which holds of no particular
answer. Attitude composes the two dimensions, and its projections are what verb entries and
the doxastic and preferential semantics of Doxastic.lean and Preference.lean read.
Implementation notes #
The binary cut between veridical and non-veridical is the classical default; [giannakidou-1998]'s three-way veridical, nonveridical, and antiveridical taxonomy and finer attitude typologies ([anand-hacquard-2013]) cut the space differently. The strategies are the ones whose clausal distributivity [qing-uegaki-2025] contrast; relevance-based is not a published label. Speech-act predicates are outside the classification.
References #
- [karttunen-1971]
- [villalta-2008]
- [anand-hacquard-2013]
- [elliott-etal-2017]
- [giannakidou-1998]
- [qing-uegaki-2025]
- [hintikka-1962]
Veridicality of a doxastic predicate: know and discover entail their complement, believe and think do not.
- veridical : Veridicality
- nonVeridical : Veridicality
Instances For
Equations
- Doxastic.instDecidableEqVeridicality 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
- Doxastic.instReprVeridicality = { reprPrec := Doxastic.instReprVeridicality.repr }
Evaluative valence of a preferential predicate: positive for hope and wish, negative for fear and worry.
Instances For
Equations
- Preferential.instDecidableEqValence x✝ y✝ = if h : x✝.ctorIdx = y✝.ctorIdx then isTrue ⋯ else isFalse ⋯
Equations
- Preferential.instReprValence = { reprPrec := Preferential.instReprValence.repr }
Equations
- One or more equations did not get rendered due to their size.
Instances For
The compositional strategy of a preferential predicate, from which its clausal
distributivity is derived: a degree comparison's question use is the existential over answers
(Preferential.mkDegreeComparison_isDistributive), while a relation to the question holds of
no particular answer (Preferential.PreferentialPredicate.not_isDistributive_of_forall_not).
- degreeComparison
(valence : Valence)
: Strategy
Degree comparison ([villalta-2008]): ⟦x V p⟧ = μ(x, p) > θ.
- uncertaintyBased : Strategy
Anxious uncertainty about the question (worry).
- relevanceBased
(valence : Valence)
: Strategy
Anticipation of the question's resolution (Mandarin qidai) or its relevance (care).
Instances For
Equations
- Preferential.instDecidableEqStrategy.decEq (Preferential.Strategy.degreeComparison a) (Preferential.Strategy.degreeComparison b) = if h : a = b then h ▸ isTrue ⋯ else isFalse ⋯
- Preferential.instDecidableEqStrategy.decEq (Preferential.Strategy.degreeComparison valence) Preferential.Strategy.uncertaintyBased = isFalse ⋯
- Preferential.instDecidableEqStrategy.decEq (Preferential.Strategy.degreeComparison valence) (Preferential.Strategy.relevanceBased valence_1) = isFalse ⋯
- Preferential.instDecidableEqStrategy.decEq Preferential.Strategy.uncertaintyBased (Preferential.Strategy.degreeComparison valence) = isFalse ⋯
- Preferential.instDecidableEqStrategy.decEq Preferential.Strategy.uncertaintyBased Preferential.Strategy.uncertaintyBased = isTrue ⋯
- Preferential.instDecidableEqStrategy.decEq Preferential.Strategy.uncertaintyBased (Preferential.Strategy.relevanceBased valence) = isFalse ⋯
- Preferential.instDecidableEqStrategy.decEq (Preferential.Strategy.relevanceBased valence) (Preferential.Strategy.degreeComparison valence_1) = isFalse ⋯
- Preferential.instDecidableEqStrategy.decEq (Preferential.Strategy.relevanceBased valence) Preferential.Strategy.uncertaintyBased = isFalse ⋯
- Preferential.instDecidableEqStrategy.decEq (Preferential.Strategy.relevanceBased a) (Preferential.Strategy.relevanceBased b) = if h : a = b then h ▸ isTrue ⋯ else isFalse ⋯
Instances For
Equations
- Preferential.instReprStrategy = { reprPrec := Preferential.instReprStrategy.repr }
Equations
- One or more equations did not get rendered due to their size.
Instances For
The valence of a strategy; worry is negative.
Equations
Instances For
The semantic classification of an attitude predicate: doxastic, with an accessibility semantics ([hintikka-1962]) and a veridicality, or preferential, with a degree semantics and a strategy.
- doxastic (veridicality : Doxastic.Veridicality) : Attitude
- preferential (strategy : Preferential.Strategy) : Attitude
Instances For
Equations
- instDecidableEqAttitude.decEq (Attitude.doxastic a) (Attitude.doxastic b) = if h : a = b then h ▸ isTrue ⋯ else isFalse ⋯
- instDecidableEqAttitude.decEq (Attitude.doxastic veridicality) (Attitude.preferential strategy) = isFalse ⋯
- instDecidableEqAttitude.decEq (Attitude.preferential strategy) (Attitude.doxastic veridicality) = isFalse ⋯
- instDecidableEqAttitude.decEq (Attitude.preferential a) (Attitude.preferential b) = if h : a = b then h ▸ isTrue ⋯ else isFalse ⋯
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
Equations
- instReprAttitude = { reprPrec := instReprAttitude.repr }
The veridicality of a predicate; preferential predicates are non-veridical.
Equations
Instances For
Whether the predicate is doxastic, as a Boolean for lexicon filters.
Equations
- (Attitude.doxastic a).isDoxastic = true
- (Attitude.preferential a).isDoxastic = false
Instances For
Whether the predicate is preferential, as a Boolean for lexicon filters.
Equations
- (Attitude.doxastic a).isPreferential = false
- (Attitude.preferential a).isPreferential = true
Instances For
The strategy of a preferential predicate.
Equations
- (Attitude.doxastic a).strategy? = none
- (Attitude.preferential a).strategy? = some a
Instances For
The valence of a preferential predicate.
Equations
- (Attitude.doxastic a).valence = none
- (Attitude.preferential a).valence = some a.valence