Documentation

Linglib.Studies.Sauerland2004

[Sau04] — Scalar Implicatures in Complex Sentences #

[Sau04]

Sauerland, U. (2004). Scalar implicatures in complex sentences. Linguistics and Philosophy 27(3): 367–391.

The paper's derivation for disjunction, run through the consistency-gated algorithm in Pragmatics/NeoGricean/Basic.lean (SecondaryLicensed, implementing the paper's (42)/(43), verified p. 383): asserting A or B against the alternatives {A, B, A∧B} yields the primary implicatures ¬KA, ¬KB, ¬K(A∧B). Of the candidate secondary implicatures, K¬(A∧B) is consistent with the commitments and licensed (conj_secondary_licensed), while K¬A is blocked (disjunct_secondary_blocked): K¬A together with K(A∨B) forces KB, contradicting the primary ¬KB. (The paper frames the same block dually: K¬A contradicts the possibility implicature PA entailed by the assertion plus ¬KB.) This asymmetry — "not both" arises but "not A" does not — is the paper's signature prediction for disjunction, unavailable to accounts that negate all stronger alternatives indiscriminately.

The four-world model DisjWorld distinguishes worlds by which disjuncts hold; the assertion A or B excludes only neither.

Worlds distinguished by which of the two disjuncts hold.

Instances For
    @[implicit_reducible]
    Equations
    Equations
    • One or more equations did not get rendered due to their size.
    Instances For
      @[implicit_reducible]
      Equations
      • One or more equations did not get rendered due to their size.

      The first disjunct A.

      Equations
      Instances For

        The second disjunct B.

        Equations
        Instances For

          The assertion A or B.

          Equations
          Instances For

            The conjunctive alternative A and B.

            Equations
            Instances For

              The scalar alternatives to A or B: each disjunct and the conjunction.

              Equations
              Instances For

                The licensed secondary implicature: K¬(A∧B) is consistent with the assertion and all primary implicatures — witnessed by the state considering exactly onlyA and onlyB possible. This is the "not both" inference of A or B.

                The blocked secondary implicature: K¬A is inconsistent with the commitments. From K(A∨B) and K¬A every possible world satisfies B, so KB holds — contradicting the primary implicature ¬KB. The disjuncts therefore yield only ignorance inferences, never "not A".

                By the A↔B symmetry of the model, K¬B is blocked identically.

                The strengthened reading of A or B the algorithm predicts: assertion plus the licensed "not both", realizable at exactly the one-disjunct worlds.