Documentation

Linglib.Fragments.Mayan.Chol.Agreement

Chol Agreement and Case Fragment @cite{coon-2013} @cite{imanishi-2020} #

Agreement morphology and case assignment for Chol (Cholan, Mayan), a low absolutive language with aspect-based split alignment. Per @cite{vazquez-alvarez-2011} §1.9.4, Chol is "an ergative language" in which "the ergative pattern is split in all non-perfective aspects, resulting in nominative-accusative alignment" — Set A indicates both transitive and intransitive subjects in non-perfective.

Descriptive vs analytical framing of the non-perfective pattern #

The descriptive grammar (@cite{vazquez-alvarez-2011}) characterizes the non-perfective alignment as nominative-accusative: Set A as a nominative-like marker grouping S with A, Set B as accusative on O.

The formal-syntactic analyses cited here label the same surface pattern differently: @cite{coon-2013} argues the non-perfective construction embeds a nominalized clause (with the aspectual predicate choñkol as matrix), so Set A on the embedded subject is genitive-from-D, NOT nominative — the morphological identity of Set A with the possessive marker (cf. tyi j-kajpelo 'in my coffee field' vs tyi j-k'el-e-ø 'I saw him', both with Set A j- per @cite{vazquez-alvarez-2011} p. 76) is taken as evidence. @cite{imanishi-2020} parameterizes Mayan split-ergative alignment via two parameters: (i) the Restriction on Nominalization (RON) on the nominalizing head n, and (ii) the Mayan Absolutive Parameter (high vs low ABS, after @cite{coon-mateo-pedro-preminger-2014}). For Chol, n does not impose RON, so S/A is the highest DP in the nominalized clause and receives genitive from D — matching Coon's analytical view. The substrate's extendedErgative.assignCase returns .gen (Coon's analytical view); a descriptive-grammar implementation would return .nom. The label "extended ergative" is Coon's coinage, generalizing one subtype of @cite{dixon-1994}'s split-ergative-on-TAM-lines pattern.

Morpheme order and word-class status #

Per @cite{vazquez-alvarez-2011} §3.4: in Chol "the aspect markers are auxiliaries" — tyi (perfective) and mi (imperfective) are aspectual auxiliaries (independent words preceding the verbal complex), not particles, not clitics, not verbal prefixes. Set A ergative/genitive markers are prefixes on the verbal complex (per @cite{vazquez-alvarez-2011} §4.1.1; some prior literature treats them as proclitics — @cite{martinez-cruz-2007}, @cite{arcos-lopez-2009}). Set B absolutive markers are suffixes (per @cite{kaufman-norman-1984} p. 90, originally cliticized pronouns).

Schematic order: [Aux] [ERG-modifier*-ROOT-DERIV-STATUS-ABS], with the bracketed verbal complex as a single phonological unit. Contrasts with Kaqchikel's [Aux] [ABS-ERG-Stem] (high-ABS).

The Two Agreement Paradigms (Set A / Set B) #

Case Licensing (analytical, per @cite{coon-2013}) #

Accusative Side (Non-Perfective) #

In non-perfective aspect, the aspectual predicate choñkol embeds a nominalized clause. The RON does NOT hold: the external argument may be generated inside the nominalized clause. Result (Coon analysis): S/A = GEN (from D), O = ABS (from Voice).

What this fragment doesn't model #

Per @cite{vazquez-alvarez-2011} §1.9.4, Chol exhibits all four Dixon alignment types: ergative-absolutive, nominative-accusative, Split-S (some intransitives obligatorily Sa = Set A on light verb cha'l; others obligatorily So = Set B), and Fluid-S (verbs like wäy 'sleep' that take either Set A or Set B). The current ArgPosition.intranS collapses Sa/So/fluid-S into a single intransitive subject category — sufficient for the perfective↔non-perfective split formalization but undermodels the agentive split. Future refinement: split into intranSAgentive / intranSPatientive / intranSFluid.

source
@[reducible, inline]
abbrev Fragments.Mayan.Chol.ArgPosition :
Type

Argument positions in a Chol clause. Aliased to the canonical Features.Prominence.ArgumentRole (S/A/P/R/T) so cross-Mayan and cross-framework code shares one inventory. Use the canonical constructor names .A / .P / .S directly.

Equations
Instances For
    source
    @[reducible, inline]
    abbrev Fragments.Mayan.Chol.ArgPosition.case :
    ArgPosition → Core.Case

    Perfective case assignment for Chol. Definitionally equal to Fragments.Mayan.ergCaseChol, which derives from Alignment.ergative.assignCase in Theories/Syntax/Case/Alignment.lean — the foundation makes the pattern (A → ERG, S/P → ABS) explicit; the per-language wrapper preserves dot-notation position.case for consumers, uniform with the other Mayan fragments.

    Equations
    Instances For
      source
      @[reducible, inline]
      abbrev Fragments.Mayan.Chol.ArgPosition.accCase :
      ArgPosition → Core.Case

      Non-perfective case assignment for Chol. Definitionally equal to Fragments.Mayan.accCaseChol, derived from Alignment.extendedErgative.assignCase. The "extended ergative" pattern (S/A → GEN, P → ABS) is shared with Q'anjob'al — both fragments call into the same caseExtErg substrate.

      Equations
      Instances For
        source
        def Fragments.Mayan.Chol.absPosition :
        ABSPosition

        Chol's absolutive morphemes appear in low position (at the end of the verb stem, post-stem). Observable from morpheme order: ASP-ERG-ROOT-(DERIV)-SUFFIX-ABS.

        Equations
        Instances For
          source
          def Fragments.Mayan.Chol.extractionProfile :
          Typology.ExtractionProfile

          Chol's extraction profile: no special morphology for any extraction. Unlike Q'anjob'al, Chol requires no Agent Focus morphology for A-extraction (the diagnostic for "lacking syntactic ergativity" in the @cite{coon-mateo-pedro-preminger-2014} sense). The substantive claim is extractionProfile.strategy = .none; per-position extractability follows trivially (every argument extracts freely), so no per-position table is stipulated — the contrast with Q'anjob'al lives at the strategy field.

          The resulting ambiguity (when both arguments are 3rd person) is a natural consequence of the absence of AF marking: Maxki₁ tyi y-il-ä (___₁) jiñi wiñik (___₁)? 'Who saw the man?' / 'Who did the man see?'

          Equations
          • One or more equations did not get rendered due to their size.
          Instances For
            source
            def Fragments.Mayan.Chol.absObjectInNonFinite :
            Bool

            In Chol non-finite embedded clauses (aspectless), absolutive objects ARE available. This follows from Chol being LOW-ABS: v⁰ assigns case to the object without needing Infl⁰.

            Mejl [i-k'el-oñ] 'She can see me.' (ABS object ✓) Choñkol [k-mek'-ety] 'I am hugging you.' (ABS object ✓)

            Equations
            Instances For
              source
              def Fragments.Mayan.Chol.absIntranSInNonFinite :
              Bool

              Absolutive intransitive subjects are NOT available in Chol non-finite clauses: they must be marked with the ergative/possessive prefix instead.

              Choñkol [k-ts'äm-el] 'I am bathing.' (ERG prefix, not ABS) *Choñkol [ts'äm-i-yoñ] intended: 'I am bathing.' (ABS ✗)

              Equations
              Instances For
                source
                def Fragments.Mayan.Chol.setAExponentPreC :
                PersonNumber → String

                Set A (ergative/possessive/genitive) markers: pre-consonantal allomorphs (@cite{vazquez-alvarez-2011} Table 10, p. 83). Note the morphophonemic rule k- → j- /_k (1sg before a /k/ root, e.g., tyi j-kajpelo 'in my coffee field' p. 76). Plural forms use the inclusive clitic =la for 1pl (the unmarked plural form per VA §4.2); the exclusive paradigm with =l(oj)oñ is a per-language refinement not exposed by the pan-Mayan PersonNumber substrate.

                Equations
                Instances For
                  source
                  def Fragments.Mayan.Chol.setAExponentPreV :
                  PersonNumber → String

                  Set A markers: pre-vocalic allomorphs (@cite{vazquez-alvarez-2011} Table 10, p. 83). 1sg k- is identical pre-C and pre-V; 2sg surfaces as aw-, 3sg as (i)y- (some speakers omit the initial vowel — @cite{vazquez-alvarez-2011} examples (12)-(13) p. 76-77).

                  Equations
                  Instances For
                    source
                    @[reducible, inline]
                    abbrev Fragments.Mayan.Chol.setAExponent :
                    PersonNumber → String

                    Canonical Set A exponent table for cross-Mayan typology. The pre-consonantal allomorph is the citation form (matching Q'anjob'al's convention); per-context realization uses setAExponentPreV before vowel-initial roots.

                    Equations
                    Instances For
                      source
                      def Fragments.Mayan.Chol.setBExponent :
                      PersonNumber → String

                      Set B (absolutive) markers: suffixes (@cite{vazquez-alvarez-2011} Table 10, p. 83). 3rd person is null (-∅). Plurals use the inclusive =la clitic for 1pl per the convention above.

                      Equations
                      Instances For
                        source
                        theorem Fragments.Mayan.Chol.p3sg_abs_null :
                        setBExponent PersonNumber.p3sg = "-∅"

                        3rd person absolutive is null — invariant across the standard Mayan branches (Cholan, Q'anjob'alan, Tseltalan, K'ichean) per @cite{kaufman-norman-1984} Table 8 reconstruction. Not universally pan-Mayan: Mam's default Set B tz'= surfaces in the 3sg slot (@cite{scott-2023}), and MayanLang.isStandard excludes Mam from the relevant cross-Mayan theorem (mayan_p3sg_abs_null).

                        source
                        theorem Fragments.Mayan.Chol.p3sg_erg_allomorphy :
                        setAExponentPreC PersonNumber.p3sg = "i-" ∧ setAExponentPreV PersonNumber.p3sg = "(i)y-"

                        3rd person Set A allomorphy: pre-consonantal i- vs pre-vocalic (i)y-. Distinct from Q'anjob'al's s- vs y- (the proto-Mayan *s- ~ *y- allomorphy was leveled to *r → y in proto-Tseltalan, and Chol inherited the leveled form per @cite{kaufman-norman-1984} p. 91).