Compositional CDRT Fragment #
@cite{muskens-1996}
§III.4 and §IV: Lexical translations (T₀) and generalized coordination for a fragment of English in Compositional DRT.
Semantic Types (Table 2, p. 164) #
| Muskens Type | Lean Type | Description |
|---|---|---|
et | E → Prop | Static predicate |
e(et) | E → E → Prop | Static relation |
s(st) | DRS S | Dynamic proposition |
[π] | Dref S E → DRS S | Dynamic predicate |
[[π]] | DynPred S E → DRS S | Dynamic quantifier |
Composition #
Meanings compose via function application (T₂) and sequencing (T₃).
These are native Lean operations, so the composition rules T₁–T₅
need no special formalization — they are just function application,
dseq, and λ-abstraction.
Generalized Coordination (§IV) #
and = sequencing applied pointwise; or = DRS disjunction applied
pointwise. The same schema works at every syntactic category (S, VP, NP).
Dynamic one-place predicate: type [π] in @cite{muskens-1996}.
Equations
Instances For
Dynamic generalized quantifier: type [[π]] in @cite{muskens-1996}.
Equations
Instances For
Common noun: farmer ↝ λv[|farmer v].
Type: [π] (dynamic one-place predicate).
Equations
Instances For
Intransitive verb: stink ↝ λv[|stinks v].
Type: [π] (same as common noun).
Equations
Instances For
Transitive verb: love ↝ λQλv(Q(λv'[|v loves v'])).
Type: [[π]] → [π]. Takes an NP (object) and produces a VP.
Equations
- Semantics.Dynamic.CDRT.Fragment.tv R Q u = Q fun (v : Semantics.Dynamic.Core.DynamicTy2.Dref S E) => [Semantics.Dynamic.Core.DynamicTy2.atom2 R u v]
Instances For
Indefinite determiner: aⁿ ↝ λP'λP([uₙ]; P'(uₙ); P(uₙ)).
Type: [π] → [[π]]. Takes a noun, produces a quantifier.
Introduces discourse referent u.
Instances For
Universal determiner: everyⁿ ↝ λP'λP(([uₙ]; P'(uₙ)) ⇒ P(uₙ)).
Type: [π] → [[π]]. Dynamic implication gives universal force.
Equations
- Semantics.Dynamic.CDRT.Fragment.detEvery u noun vp = [Semantics.Dynamic.Core.DynProp.dimpl ([u]ᵈʳᵉᶠ ⨟ noun u) (vp u)]
Instances For
Negative determiner: noⁿ ↝ λP'λP[|not([uₙ]; P'(uₙ); P(uₙ))].
Type: [π] → [[π]].
Instances For
Proper name NP: Maryⁿ ↝ λP.P(Mary).
Type: [[π]]. Applies the VP directly to the name's dref.
Equations
- Semantics.Dynamic.CDRT.Fragment.properNP name P = P name
Instances For
Pronoun NP: heₙ ↝ λP.P(uₙ).
Type: [[π]]. Picks up the dref from the antecedent.
Equations
- Semantics.Dynamic.CDRT.Fragment.pro u P = P u
Instances For
Conditional: if ↝ λpq[|p ⇒ q].
Type: DRS → DRS → DRS.
Equations
Instances For
Auxiliary negation: doesn't ↝ λPλQ[|not Q(P)] (T₀, p. 165).
Type: [π] → [[π]] → DRS. Takes VP (P) then subject NP (Q),
matching @cite{muskens-1996}'s argument order.
Equations
Instances For
and = generalized sequencing; or = generalized DRS disjunction.
The same schema works at every syntactic category by applying the
Boolean operation pointwise to the innermost DRS layer.
Sentence-level and: K₁ and K₂ = K₁; K₂.
Instances For
Sentence-level or: K₁ or K₂ = [K₁ or K₂] (disjunction test).
Equations
Instances For
VP-level and: λv(P₁(v); P₂(v)).
Equations
- Semantics.Dynamic.CDRT.Fragment.andVP P₁ P₂ u = P₁ u ⨟ P₂ u
Instances For
VP-level or: λv[P₁(v) or P₂(v)].
Equations
- Semantics.Dynamic.CDRT.Fragment.orVP P₁ P₂ u = [Semantics.Dynamic.Core.DynProp.ddisj (P₁ u) (P₂ u)]
Instances For
NP-level and: λP(Q₁(P); Q₂(P)).
Equations
- Semantics.Dynamic.CDRT.Fragment.andNP Q₁ Q₂ P = Q₁ P ⨟ Q₂ P
Instances For
NP-level or: λP[Q₁(P) or Q₂(P)].
Equations
- Semantics.Dynamic.CDRT.Fragment.orNP Q₁ Q₂ P = [Semantics.Dynamic.Core.DynProp.ddisj (Q₁ P) (Q₂ P)]
Instances For
Example derivation: "A¹ man adores a² woman. She₂ abhors him₁."
Translation (p. 167, derivation tree (39)):
[u₁]; [man u₁]; [u₂]; [woman u₂]; [u₁ adores u₂]; [u₂ abhors u₁]
Truth conditions (formula (24)):
∃x₁ x₂ (man(x₁) ∧ woman(x₂) ∧ adores(x₁,x₂) ∧ abhors(x₂,x₁))
Equations
- One or more equations did not get rendered due to their size.
Instances For
Example: "Every¹ farmer who owns a² donkey beats it₂."
The donkey sentence, with universal force from detEvery
and anaphoric it₂ picking up the dref from the indefinite.
Translation: ([u₁]; [farmer u₁]; [u₂]; [donkey u₂]; [u₁ owns u₂]) ⇒ [u₁ beats u₂]
Equations
- One or more equations did not get rendered due to their size.
Instances For
Example: "A² cat catches a¹ fish and eats it₁." (§IV, p. 179, tree (56))
VP coordination with cross-conjunct anaphora: it₁ in "eats it₁"
picks up the dref introduced by "a¹ fish" in "catches a¹ fish".
This works because andVP sequences the VP meanings, so the dref
introduced by the first conjunct is accessible in the second.
Equations
- One or more equations did not get rendered due to their size.
Instances For
VP and is definitionally sequencing at the dref argument.
VP or is definitionally disjunction at the dref argument.
Sentence and is sequencing.