Documentation

Linglib.Studies.QingEtAl2025

Qing, Özyıldız, Roelofsen, Romero & Uegaki 2025: question-taking preferentials #

[QOR+25] classify non-veridical preferential predicates by two factors — clausal distributivity and evaluative valence — and show that only the distributive positive class (hope-type) is anti-rogative: non-distributive predicates (worry, Mandarin qidai, Japanese tanosimi) and distributive negative ones (fear, Japanese osore, Turkish kork-) take questions canonically, because the [US19] triviality needs both distributivity and the positive-valence Threshold Significance Presupposition (Studies/UegakiSudo2019.lean). PredicateClass and classify render the classification (their Table 1); the observations record the paper's cross-linguistic acceptability judgments over English, Mandarin, Japanese, and Turkish. Apparent hope + question cases are analyzed as non-canonical adjunction-style composition (their §4), with the highlighting analysis considered and dispreferred.

The classification (Table 1) #

The three classes of non-veridical preferential predicates: the two distributivity-by-valence cells that take questions, and the anti-rogative distributive positive class.

  • nonDistributive : PredicateClass

    Non-distributive (worry, qidai, tanosimi): the question semantics outruns the existential over answers.

  • distributiveNegative : PredicateClass

    Distributive with negative valence (fear, osore, kork-): no Threshold Significance Presupposition.

  • distributivePositive : PredicateClass

    Distributive with positive valence (hope, wish, expect): anti-rogative via the [US19] triviality.

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

      Languages represented in the data

      Instances For
        @[instance_reducible]
        Equations
        def QingEtAl2025.instReprLanguage.repr :
        LanguageStd.Format
        Equations
        • One or more equations did not get rendered due to their size.
        Instances For

          An empirical observation: predicate name, language, and acceptability.

          The semantic properties are stored in the corresponding Fragment entry. Here we just record the empirical acceptability judgments.

          • form : String

            Predicate form

          • language : Language

            Language

          • gloss : String

            English gloss (for non-English)

          • takesPolQ : Bool

            Empirical: Does it take polar questions?

          • takesWhQ : Bool

            Empirical: Does it take wh-questions?

          • notes : String

            Additional notes

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

                                                  Verifying predictions against observations #

                                                  Each predicate's class follows from its distributivity and valence via classify — distributivity proved from the semantics (Preferential.mkDegreeComparison_isDistributive, Preferential.worry_not_distributive) — and the class predicts question-embedding, checked against the observations:

                                                  Cross-linguistic verification #

                                                  LanguagePredicateClassPredictedObserved✓/✗
                                                  Englishhope3✗ questions
                                                  Englishfear2✓ questions
                                                  Englishworry1✓ questions
                                                  Mandarinqidai1✓ questions
                                                  Mandarinxiwang3✗ questions
                                                  Japanesetanosimi1✓ questions
                                                  Japanesekitai3✗ questions
                                                  Turkishkork-2✓ questions
                                                  Turkishum-3✗ questions

                                                  Key examples #

                                                  Hope cannot embed questions in English (*John hopes whether Mary will come; *John hopes who will come): hope is distributive and positive, so with the answers drawn from the comparison class the assertion is settled by the Threshold Significance Presupposition — the [US19] triviality, Studies/UegakiSudo2019.lean.

                                                  Mandarin qidai is positive yet embeds questions (Zhangsan qidai shei hui lai, "Zhangsan looks forward to who will come"): its question semantics carries an anticipation-of-resolution condition, so it is not distributive and the triviality derivation does not go through (§3.1).

                                                  Turkish kork- "fear" embeds questions with a symmetric interpretation — "John fears whether his neighbor will be home" is felicitous whether he fears the neighbor's presence or absence (their GoodFriend and NoiseHater contexts, §3.2) — because negative predicates do not trigger the Threshold Significance Presupposition.