Pancheva & Zubizarreta (2018): The Person Case Constraint #
@cite{pancheva-zubizarreta-2018}
The Person Case Constraint: The Syntactic Encoding of Perspective. Natural Language and Linguistic Theory 36: 1291–1337.
Summary #
The Person Case Constraint (PCC) restricts person combinations in ditransitive clitic clusters cross-linguistically. P&Z re-analyze the PCC as a syntax-semantics interface phenomenon rooted in the encoding of perspective. The core mechanism is a P-Constraint triggered by an interpretable person feature on Appl, which marks the indirect object as a point-of-view center.
Key Contributions Formalized #
Full PCC typology: five attested varieties (strong, ultra-strong, weak, super-strong, me-first) plus three predicted grammars, all derived from four binary parameters of the P-Constraint.
Logophoric grounding: the P-Prominence settings correspond to logophoric roles (Sells 1987): pivot, self, source.
Markedness predictions: grammar markedness follows from parameter departures from the default (strong PCC).
Impossible grammar predictions: me-first + *<3,3> restriction is ruled out by incompatible parameter settings.
Cross-linguistic data: French strong PCC, Catalan ultra-strong, Spanish weak, Kambera super-strong, Bulgarian me-first.
The Person Hierarchy is derived from suitability for point-of-view center, not stipulated as a primitive.
Equations
- Phenomena.Agreement.Studies.PanchevaZubizarreta2018.personHierarchyRank Core.Prominence.PersonLevel.first = 2
- Phenomena.Agreement.Studies.PanchevaZubizarreta2018.personHierarchyRank Core.Prominence.PersonLevel.second = 1
- Phenomena.Agreement.Studies.PanchevaZubizarreta2018.personHierarchyRank Core.Prominence.PersonLevel.third = 0
Instances For
Strong PCC (§4.1.1, (14a)): DO must be 3P.
Weak PCC (§4.1.3, (14b)): 1P/2P co-occurrence allowed.
Ultra-strong PCC (§4.1.2, (14d)): ⟨1,2⟩ OK but ⟨2,1⟩ banned.
Me-first PCC (§4.3, (14c)): only ⟨3,1⟩ and ⟨2,1⟩ banned.
Super-strong PCC (§4.2, (14e)): IO must be 1P/2P, DO must be 3P. ⟨3,3⟩ banned (IO not [+participant]).
CLR effects predicted in all [+proximate] grammars. Me-first grammars do NOT predict CLR in ⟨3,3⟩ (domain restriction exempts).
*<3,3> effects compatible with [+proximate] family.
*<3,3> incompatible with me-first (impossible parameter combination).
PG1: [+participant] + P-Primacy.
PG2: [+participant], no P-Uniqueness. IO must be SAP.
PG3: [+author], unrestricted domain. Only 1P IO.
"Direct" order: IO satisfies the prominence requirement.
Equations
- Phenomena.Agreement.Studies.PanchevaZubizarreta2018.isDirectOrder ioPerson prominence = Minimalism.PConstraint.satisfiesProminence prominence ioPerson
Instances For
French strong PCC (§4.1.1, (16)).
Catalan ultra-strong (§4.1.2, (20)).
Spanish weak PCC (§4.1.3, (23)–(24)).
Kambera super-strong (§4.2, (27)).
Bulgarian me-first (§4.3, (29)).
Me-first and weak have different licit sets.