Documentation

Studies.Kennedy2007

Kennedy (2007): Vagueness and Grammar #

Kennedy separates relative gradable adjectives (tall, long), whose positive form compares with a contextual standard, from absolute ones, whose standard is an endpoint of the scale: the minimum (wet, bent) or the maximum (full, dry). The positive form is ⟦pos⟧ = λg λx. g(x) ≥ s(g) (27), with the standard s fixed by the scale rather than by a comparison class, and Interpretive Economy (66) derives it from the scale structure. This file states the three positive forms and checks the diagnostics of Section 3.2 that separate them.

Main results #

References #

The positive form (§2.3, §3.1) #

def Kennedy2007.MinStandardPos {Entity : Type u_1} {D : Type u_2} [LinearOrder D] [OrderBot D] (μ : Entity → D) (x : Entity) :

A minimum-standard adjective (wet) holds of x when its degree is above the scale minimum.

Equations
Instances For
    def Kennedy2007.MaxStandardPos {Entity : Type u_1} {D : Type u_2} [LinearOrder D] [OrderTop D] (μ : Entity → D) (x : Entity) :

    A maximum-standard adjective (dry) holds of x when its degree is the scale maximum.

    Equations
    Instances For
      def Kennedy2007.RelativePos {Entity : Type u_1} {D : Type u_2} [LinearOrder D] (μ : Entity → D) (θ : D) (x : Entity) :

      A relative adjective (long) holds of x when its degree exceeds a contextual threshold.

      Equations
      Instances For

        Definite descriptions (§3.2, (53)–(54)) #

        theorem Kennedy2007.exists_relativePos_of_lt {Entity : Type u_1} {D : Type u_2} [LinearOrder D] (μ : Entity → D) {a b : Entity} (h : μ b < μ a) :
        ∃ (θ : D), RelativePos μ θ a ∧ ¬RelativePos μ θ b

        The long one succeeds because two objects of different length can always be told apart by a contextual standard.

        theorem Kennedy2007.maxStandardPos_and_not_iff {Entity : Type u_1} {D : Type u_2} [LinearOrder D] [OrderTop D] (μ : Entity → D) (a b : Entity) :
        MaxStandardPos μ a ∧ ¬MaxStandardPos μ b ↔ μ a = ⊤ ∧ μ b ≠ ⊤

        In the full one (54), a maximum-standard adjective tells two objects apart only when one is at the maximum and the other is not, so two partially full jars leave nothing to pick out.

        The Sorites second premise (§3.2, (57)–(58)) #

        theorem Kennedy2007.not_maxStandardPos_of_lt_top {Entity : Type u_1} {D : Type u_2} [LinearOrder D] [OrderTop D] {μ : Entity → D} {x : Entity} (h : μ x < ⊤) :

        A theater with one fewer occupied seat than a full one is not full (57), since any degree below the maximum fails a maximum standard.

        theorem Kennedy2007.not_minStandardPos_iff {Entity : Type u_1} {D : Type u_2} [LinearOrder D] [OrderBot D] (μ : Entity → D) (x : Entity) :
        ¬MinStandardPos μ x ↔ μ x = ⊥

        A rod with no bend is not bent (58), since exactly the minimum fails a minimum standard.

        theorem Kennedy2007.isCompl_minStandardPos_maxStandardPos_dual {Entity : Type u_1} {D : Type u_2} [LinearOrder D] [OrderBot D] (μ : Entity → D) :
        IsCompl {x : Entity | MinStandardPos μ x} {x : Entity | MaxStandardPos (⇑OrderDual.toDual ∘ μ) x}

        The door is not open entails the door is closed (47). A minimum standard and the maximum standard of the antonym, which measures on the dual scale, split the domain between them, since the minimum of the one is the maximum of the other.

        Comparatives (§3.2, (49)–(52)) #

        theorem Kennedy2007.minStandardPos_of_comparative {Entity : Type u_1} {D : Type u_2} [LinearOrder D] [OrderBot D] {μ : Entity → D} {a b : Entity} (h : μ b < μ a) :

        The floor is wetter than the countertop entails the floor is wet (49).

        theorem Kennedy2007.not_maxStandardPos_of_comparative {Entity : Type u_1} {D : Type u_2} [LinearOrder D] [OrderTop D] {μ : Entity → D} {a b : Entity} (h : μ b < μ a) :

        The floor is drier than the countertop entails the countertop is not dry (50).

        theorem Kennedy2007.relativePos_undetermined_of_comparative {Entity : Type u_1} {D : Type u_2} [LinearOrder D] (μ : Entity → D) {a b : Entity} (h : μ b < μ a) :
        (∃ (θ : D), RelativePos μ θ a) ∧ ∃ (θ : D), ¬RelativePos μ θ a

        Rod A is longer than rod B entails neither that A is long nor that it is not (51).

        Degree modifiers (§3.2, (61)) #

        A degree modifier is a maximizer (completely, fully) or a minimizer (slightly, partially).

        Instances For
          @[instance_reducible]
          Equations

          A modifier is licensed on a scale that has the endpoint it picks out.

          Equations
          Instances For

            table61 is table (61) as printed, indexed by the member of the pair.

            Equations
            Instances For
              theorem Kennedy2007.table61_iff_licenses (p : Polarity) (b : Degree.Boundedness) (m : DegreeModifier) :
              table61 p b m = true ↔ Licenses m (p • b)

              Every cell of (61) is the endpoint structure of the adjective's own scale.

              Interpretive Economy (§4.2–§4.3) #

              An open scale offers no endpoint, so its standard is contextual and the adjective needs a comparison class.

              A totally closed scale is interpretively variable, admitting both endpoint standards ((67)–(68), opaque/transparent, open/exposed); the maximum is only the default.