Documentation

Linglib.Phenomena.Agreement.Studies.Deal2024

Interaction, Satisfaction, and the PCC @cite{deal-2024} #

@cite{deal-2024} unifies the Person Case Constraint (PCC) typology under a single Agree operation with two independently parameterized conditions:

The probe encounters DO first ("DO preference"). If DO satisfies the probe, it halts before reaching IO — creating a PCC violation when IO requires licensing. If DO does not satisfy, the probe copies DO's features and continues to IO. Dynamic interaction narrows the probe's subsequent INT condition based on what was copied from DO.

PCC Typology #

Two parameters — satisfaction feature and dynamic interaction configuration — derive five attested PCC varieties:

SATDynINTPCC typeLicit
[PART]noneStrong3
[SPKR]noneMe-first6
none[PART]↑Weak7
[SPKR][PART]↑Strictly descending5
nonenoneNo PCC9

Bridge Results #

This study file connects @cite{deal-2024}'s framework to both:

  1. @cite{pancheva-zubizarreta-2018} (PConstraint.lean): exact match for Strong, Weak, and Me-first (all 9 cells); 7/9 match for Strictly Descending vs Ultra-strong (discrepancies on reflexive SAP combinations).
  2. @cite{bejar-rezac-2009} (CyclicAgree.lean): probe satisfaction in Deal's sense corresponds to residue depletion in cyclic Agree — a DP satisfies SAT:[PART] iff it depletes the partial probe's residue.

Person features in @cite{deal-2024}'s geometry.

[φ] → [PART] → [SPKR]
             → [ADDR]

Maps onto DecomposedPerson: [PART] = hasParticipant, [SPKR] = hasAuthor, [ADDR] = (person == .second).

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

      Dynamic interaction configurations.

      After copying features from DO, the probe's subsequent INT condition may narrow. The notation indicates which features trigger narrowing.

      • none_: no dynamic narrowing
      • part: [PART]↑ — if DO bears [PART], INT narrows to [PART]
      • spkr: [SPKR]↑ — if DO bears [SPKR], INT narrows to [SPKR]
      • partAndSpkr: both [PART]↑ and [SPKR]↑ — [SPKR]↑ takes priority
      Instances For
        Equations
        • One or more equations did not get rendered due to their size.
        Instances For

          A grammar in @cite{deal-2024}'s framework.

          Two parameters determine the PCC type:

          • satisfaction: which feature, if present on DO, halts the probe. none means the probe is never satisfied by DO alone.
          • dynInteraction: which features trigger dynamic narrowing of the probe's INT condition after interacting with DO.
          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

                  Is a clitic combination ⟨IO, DO⟩ licit under a Deal grammar?

                  The probe encounters DO first ("DO preference"):

                  1. If DO bears the SAT feature → probe satisfied → halts before IO → IO not licensed → illicit.
                  2. Otherwise, the probe copies features from DO. Dynamic interaction may narrow the probe's subsequent INT condition.
                  3. If narrowing occurred, IO must bear the narrowed feature to be visible to the probe. If IO is invisible → illicit.
                  4. If no narrowing, IO is always visible → licit.
                  Equations
                  • One or more equations did not get rendered due to their size.
                  Instances For

                    Strong PCC: SAT:[PART], no dynamic interaction. DO bearing [PART] satisfies the probe → only 3P DOs survive.

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

                      Me-first PCC: SAT:[SPKR], no dynamic interaction. Only 1P DOs satisfy the probe → only 1P DOs are banned.

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

                        Weak PCC: no satisfaction, [PART]↑ dynamic interaction. The probe is never satisfied by DO, but copying [PART] from a SAP DO narrows the probe so it can only see [PART]-bearing IOs.

                        Equations
                        Instances For

                          Strictly descending PCC: SAT:[SPKR], [PART]↑ dynamic interaction. 1P DOs satisfy (banned); SAP DOs narrow the probe to require [PART] on IO.

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

                            No PCC: no satisfaction, no dynamic interaction. The probe never halts at DO and never narrows.

                            Equations
                            Instances For

                              ⟨1,2⟩ is licit: 2P DO doesn't bear [SPKR] (no SAT); dynamic narrowing requires IO to bear [PART]; 1P bears [PART].

                              ⟨2,2⟩ is licit: 2P DO lacks [SPKR]; narrowing requires [PART] on IO; 2P IO bears [PART].

                              ⟨3,2⟩ is illicit: 2P DO lacks [SPKR] but bears [PART]; narrowing requires [PART] on IO; 3P IO lacks [PART].

                              Count licit combinations (out of 9 = 3×3).

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

                                Strong entails Me-first: anything licit under Strong is licit under Me-first. (Strong has SAT:[PART] ⊇ SAT:[SPKR].)

                                Strong entails Weak: anything licit under Strong is licit under Weak.

                                theorem Phenomena.Agreement.Studies.Deal2024.sd_matches_ultra_non_reflexive :
                                isLicit strictlyDescending Core.Prominence.PersonLevel.first Core.Prominence.PersonLevel.second = Minimalism.PConstraint.pccLicit Minimalism.PConstraint.ultraStrongGrammar Core.Prominence.PersonLevel.first Core.Prominence.PersonLevel.second isLicit strictlyDescending Core.Prominence.PersonLevel.first Core.Prominence.PersonLevel.third = Minimalism.PConstraint.pccLicit Minimalism.PConstraint.ultraStrongGrammar Core.Prominence.PersonLevel.first Core.Prominence.PersonLevel.third isLicit strictlyDescending Core.Prominence.PersonLevel.second Core.Prominence.PersonLevel.first = Minimalism.PConstraint.pccLicit Minimalism.PConstraint.ultraStrongGrammar Core.Prominence.PersonLevel.second Core.Prominence.PersonLevel.first isLicit strictlyDescending Core.Prominence.PersonLevel.second Core.Prominence.PersonLevel.third = Minimalism.PConstraint.pccLicit Minimalism.PConstraint.ultraStrongGrammar Core.Prominence.PersonLevel.second Core.Prominence.PersonLevel.third isLicit strictlyDescending Core.Prominence.PersonLevel.third Core.Prominence.PersonLevel.first = Minimalism.PConstraint.pccLicit Minimalism.PConstraint.ultraStrongGrammar Core.Prominence.PersonLevel.third Core.Prominence.PersonLevel.first isLicit strictlyDescending Core.Prominence.PersonLevel.third Core.Prominence.PersonLevel.second = Minimalism.PConstraint.pccLicit Minimalism.PConstraint.ultraStrongGrammar Core.Prominence.PersonLevel.third Core.Prominence.PersonLevel.second isLicit strictlyDescending Core.Prominence.PersonLevel.third Core.Prominence.PersonLevel.third = Minimalism.PConstraint.pccLicit Minimalism.PConstraint.ultraStrongGrammar Core.Prominence.PersonLevel.third Core.Prominence.PersonLevel.third

                                Strictly descending agrees with ultra-strong on 7 of 9 cells. All non-reflexive-SAP combinations match.

                                Strictly descending diverges from ultra-strong on ⟨2,2⟩: Deal allows it (2P lacks [SPKR]; narrowing requires [PART]; 2P bears [PART]), P&Z bans it (P-Uniqueness violated, 2P lacks [+author] so no P-Primacy rescue).

                                Probe satisfaction in @cite{deal-2024}'s sense (SAT:[PART]) corresponds to residue depletion in @cite{bejar-rezac-2009}'s cyclic Agree.

                                A DP fully checks the partial probe [uπ, uParticipant] iff it bears [PART]. In Deal's terms, such a DP satisfies a SAT:[PART] probe. In B&R's terms, it depletes the active residue to ∅.

                                This bridges the two frameworks: "the probe is satisfied" (Deal) ↔ "no active residue remains" (B&R).

                                SAT:[PART] yields the Strong PCC regardless of dynamic interaction.

                                This is the key insight of Table (53): when DO bears [PART], the probe is satisfied before any dynamic narrowing can take effect. Since [SPKR] and [ADDR] geometrically entail [PART], any DP that would trigger dynamic interaction via [SPKR]↑ or [PART]↑ also satisfies SAT:[PART]. Therefore, dynamic interaction is irrelevant when SAT = [PART].

                                You-first PCC (predicted): SAT:[ADDR], no dynamic interaction. 2P DOs satisfy the probe → ⟨IO, 2P DO⟩ is illicit. Predicted by @cite{deal-2024} §6.1 but not yet robustly attested.

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

                                  A-descending PCC (predicted): SAT:[ADDR], [PART]↑ dynamic interaction. IO must outrank DO on the hierarchy 2 > 1 > 3. @cite{deal-2024} §6.1 notes hints of this pattern in Catalan and Italian speaker variation.

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

                                    Reverse PCC: the probe encounters IO before DO (IO preference).

                                    @cite{deal-2024} §6.2: in structures where IO moves to a position between v and DO, the probe encounters IO first. This reverses the PCC restriction — now IO features can satisfy/narrow the probe, and DO is the argument that may fail to be licensed.

                                    Attested in Shapsug Adyghe, varieties of Swiss German, Czech, and Slovenian.

                                    Equations
                                    Instances For

                                      Reverse strictly descending PCC (Shapsug Adyghe): DO must outrank IO on the hierarchy 1 > 2 > 3 — the mirror image of forward SD.

                                      Equations
                                      Instances For

                                        Strong PCC (2a): DO must be 3P. Any IO is licit with a 3P DO; any SAP DO is illicit regardless of IO.

                                        Weak PCC (2b): if IO is 3P, DO must be 3P. Equivalently: the only illicit cells are ⟨3P IO, SAP DO⟩.

                                        Me-first PCC (2c): if 1P is present, it must be IO. Equivalently: DO cannot be 1P.

                                        Strictly descending PCC (2d): IO must outrank DO on 1 > 2 > 3. For the 6 non-diagonal cells, this is exactly the pattern. (Diagonal cells: ⟨3,3⟩ and ⟨2,2⟩ are licit; ⟨1,1⟩ is illicit because SAT:[SPKR] fires.)

                                        Strong entails strictly descending: anything licit under Strong is licit under SD. (Strong bans all SAP DOs; SD bans only 1P DOs and ⟨3P IO, SAP DO⟩.)

                                        dpBears is grounded in the shared feature geometry decomposePerson from PersonGeometry.lean. This makes the connection structural — Deal's person features are not independently stipulated but derived from the same privative geometry used by @cite{pancheva-zubizarreta-2018}'s satisfiesProminence and @cite{bejar-rezac-2009}'s personSpec.