Hindi Case Inventory @cite{blake-1994}

Hindi has a split-ergative case system: ergative -ne marks the transitive agent in perfective aspect only.

Hindi postpositions mark 7 case functions:

• NOM (unmarked), ERG (-ne, perfective A only)
• ACC / DAT (-ko, syncretic), GEN (-ka / -ke / -ki)
• LOC (-mem), ABL/INST (-se, syncretic)

The ACC/DAT syncretism (-ko) and ABL/INST syncretism (-se) are cross-linguistically common patterns.

Split-Ergative Connection

This fragment connects to the hindiSplit already defined in Core.Case.SplitConditions, which formalizes the perfective to ergative conditioning.

def Fragments.Hindi.Case.caseInventory : Finset Core.Case

Hindi case inventory. ACC/DAT share -ko; ABL/INST share -se. Both syncretic pairs are included as distinct Core.Case values since they occupy different positions on Blake's hierarchy.

Equations

• Fragments.Hindi.Case.caseInventory = {UD.Case.nom, UD.Case.erg, UD.Case.acc, UD.Case.dat, UD.Case.gen, UD.Case.loc, UD.Case.abl, UD.Case.inst}

theorem Fragments.Hindi.Case.acc_dat_syncretic_marker : True

ACC/DAT syncretism (-ko marks both).

theorem Fragments.Hindi.Case.abl_inst_same_tier : Core.Case.hierarchyRank UD.Case.abl = Core.Case.hierarchyRank UD.Case.inst

ABL/INST syncretism (-se marks both). Same-tier adjacency.

theorem Fragments.Hindi.Case.hindi_perfective_is_ergative : Core.hindiSplit.alignment Core.Aspect.perfective = Core.AlignmentFamily.ergative

The split-ergative system defined in SplitConditions.lean.

theorem Fragments.Hindi.Case.hindi_imperfective_is_accusative : Core.hindiSplit.alignment Core.Aspect.imperfective = Core.AlignmentFamily.accusative