Documentation

Linglib.Studies.Bondarenko2022

Bondarenko 2022: Anatomy of an Attitude #

[Bon22]

Embedded CPs receive two denotations, tracking a left-peripheral syntactic distinction (§1.1.1, ex. 1): a Cont-CP denotes a predicate on content individuals via CONT ([Kra06]; [Mou09]; [Mou15]; [Ell20a]) and projects a ContP; a Sit-CP denotes a predicate on minimal exemplifying situations ([Kra89]) and lacks one. ContP is overt in Korean (-ta) and Buryat ( 'say'), null in English and Russian.

Formalized here: equality semantics for displacement (eq. 7) against subset (eq. 10) and existential (eq. 15), whence the impossibility of true CP conjunction ([BB21]); Sit-CP transparency vs Cont-CP opacity (§1.1.1.1 ex. 5-6); and the transparent syntax-semantics mapping thesis (§1.1.2, ch. 4) — bare CPs compose only as modifiers, nominalized CPs only as arguments — later attacked by [Ang26] (comparison in Studies/Angelopoulos2026.lean, per the chronology rule).

Main declarations #

Implementation notes #

Bare §/ex./eq. locators refer to the dissertation and were verified against its text. Sections track different chapters: nominal sorts through the BE2026 bridge follow the ch. 1 summary, the type-theoretic sections ch. 4 (§§4.2-4.5), the co-occurrence and head-denotation sections ch. 2 (§2.2.3, §2.3). The ch. 1 summary repeats later material: Table 1.1 = Table 4.1, and ch. 1 ex. 1 = §4.2 ex. 1-2.

Out of scope: ch. 3 clausal coordination (only the conjunction-impossibility theorem lands here), ch. 5 polarity subjunctives (Russian NPI data, L-analyticity), and the morphological exposition of ContP, which lives in Fragments/Buryat/Complementizers.lean and Fragments/Korean/Complementizers.lean.

Typed paradigm sentences (§2.2.3 ex. 105–112, 120–122; §4.3.1 ex. 30–32) live in Bondarenko2022.Examples, generated from Data/Examples/Bondarenko2022.json.

Nominal sorts #

The dissertation's typology of nouns combining with embedded CPs (§2.2). Studies-local: the substrate (ContentIndividual, SituationIndividual) lives in Semantics/Attitudes/ClauseDenotation/.

  • content : NominalSort

    Content nouns (§2.2): idea, rumor, hypothesis, claim, fact, lie.

  • situation : NominalSort

    Situation nouns (§2.2): situation, event, case, circumstance, state of affairs.

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

      Equality vs. subset vs. existential semantics #

      §1.1.1.2. The substrate exposes equality as compC and subset as ContentIndividual.entails; only the existential candidate is new here.

      Existential semantics for a Cont-CP (eq. 15): CONT(x) ∩ p ≠ ∅. Rejected by the thesis in favor of equality (compC, eq. 7); subset is eq. 10.

      Equations
      Instances For

        Equality is stronger than subset (§1.1.1.2); re-export of ContentIndividual.eq_implies_entails.

        Equality implies existential for nonempty content; the empty-content witness in ContentIndividual shows the proviso is real.

        CP conjunction impossibility #

        Equality semantics makes CONT a function, so true CP conjunction is impossible ([BB21], §1.1.1.2).

        Equality forces functionality: two contents assigned to the same individual coincide.

        theorem Bondarenko2022.subset_does_not_force_functionality :
        ¬∀ (xc : Semantics.Attitudes.ContentIndividual Bool) (p1 p2 : BoolProp), xc.entails p1xc.entails p2p1 = p2

        Subset does not force functionality: a non-vacuous content individual entails every superset of its content. Witness over Bool.

        Sit-CP transparency vs. Cont-CP opacity #

        The substitution test (§2.2.3 ex. 120-122; summarized at ch. 1 ex. 5-6): coextensional DPs substitute salva veritate under a Sit-NP but not under a Cont-NP — Cont-CPs are a referentially opaque domain ([Bar81], [Hig83]). Everywhere-coextensional transparency is funext-trivial; the substantive notion is transparency at the evaluation world.

        def Bondarenko2022.ReferentiallyTransparentAt {S : Type u_1} (clause : (SProp)SProp) (s : S) :

        Weak transparency at s: substituting predicates that agree at s preserves clause truth.

        Equations
        Instances For
          theorem Bondarenko2022.sit_cp_transparentAt {S : Type u_1} (s : S) :
          ReferentiallyTransparentAt (fun (p : SProp) (s' : S) => p s') s

          A Sit-CP evaluates at the actual situation, hence is weak-transparent.

          compC is not weak-transparent (witness over Bool). This is the propositional-operator shadow of the §2.2.3 ex. 120 DP-level de dicto claim, whose binding-theoretic encoding ([Ell20a]'s CONT-as-content-restrictor) is not yet substrate.

          The transparent syntax-semantics mapping thesis #

          §1.1.2: bare CPs must compose as modifiers (PM via the situation argument), nominalized CPs as arguments (FA via the DP argument). [Ang26] §7.3 attacks this thesis.

          The four logically possible (clause shape, composition path) combinations.

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

              The (bare, argument) cell is empty — the cell [Ang26] attacks (see Angelopoulos2026).

              Composition paths (Table 1.1) #

              Table 1.1's two composition paths: PM with the verb's situation argument, or FA into a DP argument slot.

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

                  The (Sit-CP, viaSituation) cell is empty by semantic triviality (§1.1.2).

                  Bridge to BondarenkoElliott2026 #

                  Bondarenko & Elliott 2026's MSI/MSO/TECM apparatus presupposes the equality semantics CONT(x) = ⟦S⟧ defended here; the cross-file bridge awaits a stable BE2026 lemma (this file does not import BE2026).

                  Type-theoretic apparatus (§4.2) #

                  The §4.4 bans are type-theoretic, not structural: bare CPs are ⟨e,t⟩, nominalized CPs ⟨e⟩, Θ-heads require ⟨e⟩; N/D presence is the exponence of the type-shift, not its cause. In Bare Phrase Structure the X-bar argument/modifier distinction is gone, so the "modifier" claim is semantic.

                  The two semantic types the derivation uses: saturated ⟨e⟩ vs unsaturated ⟨e,t⟩.

                  Instances For
                    @[implicit_reducible]
                    Equations
                    def Bondarenko2022.instReprSemType.repr :
                    SemTypeStd.Format
                    Equations
                    • One or more equations did not get rendered due to their size.
                    Instances For

                      The bare-vs-nominalized cut (§4.2), named by semantic content per Bare Phrase Structure.

                      • predicateOfIndividuals : ClauseType

                        Nominalized CP: type ⟨e⟩. Buryat participial nominalization, with or without the say-root (§4.3.1 ex. 31–32); Korean -nun + kes clause.

                      • predicateOfSituations : ClauseType

                        Bare CP: type ⟨e,t⟩, predicate over (minimal) situations. Buryat gɘžɘ-clause; Korean -ko-clause; Russian bare čto-clause.

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

                          The Θ-heads of §4.4, all sharing one selectional restriction: an individual argument. Per-Θ presupposition flavor is deferred.

                          • causer : ThetaHead

                            Θ_Causer. §4.4.1.

                          • theme : ThetaHead

                            Θ_Theme (internal-argument introducer). §4.4.2.

                          • aboutAttitude : ThetaHead

                            Θ_About: introduces the res-argument the attitude is about ([Qui56]; [CvS82]); carries a pre-existence presupposition not formalised here. §4.4.3.

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

                              Every Θ-head requires an individual argument.

                              Equations
                              Instances For

                                A clause type saturates a Θ-head iff its semantic type matches the head's required type (§4.2).

                                Equations
                                Instances For
                                  @[implicit_reducible]
                                  Equations

                                  Causer, Theme, About (§4.4) #

                                  The three bans are one type-mismatch mechanism, not three stipulations.

                                  §4.4.2: bare CPs cannot be Theme-arguments; cf. the explain-class data ([Pie00], [HS13], [Ell16], [RU21]).

                                  The modifier path (§4.5) #

                                  Bare CPs compose with the verb's situation argument by Predicate Modification; composesViaPM is the qualitative projection of the Modifier.intersective substrate (Semantics/Modification/Basic.lean), whose Frame-typed instantiation is queued.

                                  PM-compatibility with a verbal situation argument: bare CPs qualify, nominalized CPs do not.

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

                                    Deriving the mapping table #

                                    The §1.1.2 table agrees cell-by-cell with the type-theoretic derivation: the four-cell pattern is derived, not stipulated.

                                    Cross-linguistic ContP exponence #

                                    Per-language Cont-exponence analyses (ch. 4), tied by law to the fragment inventories. Buryat carries the licenser-conditioned Comp allomorphy (§4.3.1 ex. 33); Korean's parallel rule (§4.3.2 ex. 46) is not yet law-checked, so its witness stays at ContAnalysis. [Cac25]'s Tigrinya extension lives in Studies/Cacchioli2025.lean per the chronology rule.

                                    A Cont-exponence analysis of one language's clause-typing inventory (ch. 4): the morpheme, if any, overtly exponing Cont (none = null exponence: English, Russian). A structure, not a class — rival frameworks construct rival witnesses.

                                    Instances For

                                      A full Cont/Comp head assignment: ContAnalysis plus the licenser-conditioned Comp allomorphy and its laws.

                                      Instances For

                                        Buryat (§4.3.1 ex. 33): expones Cont; the suffixes are Comp allomorphs — participial -Aːša next to nominal projections, converbial -žA next to verbs.

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

                                          Korean (§4.3.2 ex. 46; cf. [BAM18]): -ta expones Cont; adnominal -nun and adverbal -ko are the Comp allomorphs, parallel to Buryat ex. 33.

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

                                            In Buryat the Cont exponent is exactly the non-suffixal say-root — a Buryat-specific alignment, not a ContAnalysis law (Korean's -ta is itself a suffix).

                                            Each of hanaxa's two frames (§4.4.3) is realized by exactly one clause-typer — bare by converbial -žA, nominalized by participial -Aːša — and the say-root realizes neither.

                                            hanaxa think~remember (§4.4.3): the attitude flips with the frame — nonveridical/opaque on the bare CP, veridical on the nominalized frame. Previously unstateable: attitude was per-lexeme.

                                            Noun-predicate co-occurrence (§2.2.3) #

                                            Cont-NPs combine with truth-evaluable predicates, Sit-NPs do not (ex. 105-107); Sit-NPs combine with occurrence predicates, Cont-NPs do not (ex. 108-112). Both asymmetries project the NominalSort distinction: contents are truth-evaluated, situations occur.

                                            Combines with 'true' / 'false' / 'mistaken' (ex. 105-107).

                                            Equations
                                            Instances For

                                              Combines with 'occur' / 'happen' (Russian proizojti, slučit'sja; Korean ilena) — ex. 108-112. The footnote-28 hedge on alachay is aggregated with the main paradigm.

                                              Equations
                                              Instances For

                                                Cont and Comp head denotations (§2.3) #

                                                Ex. 150: ⟦Cont⟧ = λp.λx. CONT(x) = p — exactly the substrate compC. Ex. 151: ⟦Comp⟧ = λp.λx. x ⊑ s ∧ x ⊩ p — parthood plus exact exemplification. The dissertation's own abbreviation of ex. 151 drops the anchoring conjunct; its footnote 37 notes the anchoring matters in §5.2.3 (clauses as restrictors of existential quantifiers). compHead keeps it.

                                                Ex. 150 is the existing compC — the ch. 2 proposal introduces no new machinery. The equality shape is further defended in §2.4 (no-stacking, ex. 171; the fact definiteness, ex. 177-178; the complex-claim argument, ex. 183-184), unformalized.

                                                def Bondarenko2022.compHead {S : Type u_1} [Preorder S] (p : SProp) (x evalSit : S) :

                                                Ex. 151: x is part of the evaluation situation and exactly verifies p (verifier-membership, vs the part-existential inexactVer of Semantics/Truthmaker/Inexact.lean).

                                                Equations
                                                Instances For
                                                  theorem Bondarenko2022.compHead_implies_inexactVerifier {S : Type u_1} [Preorder S] (p : SProp) (x evalSit : S) (h : compHead p x evalSit) :
                                                  uevalSit, p u

                                                  Exact exemplification implies inexact: a verifier is a part of itself (cf. inexactVer_of_exact).