Documentation

Linglib.Studies.Kalin2018

Kalin (2018) — Licensing and Differential Object Marking #

[Kal18] [Mar91]

[Kal18] derives differential object marking (DOM) from nominal licensing rather than object visibility, raising, or differentiation. Two parameters interact: (i) which nominals require licensing (in Senaya, only specific ones), and (ii) where the licensers are — every clause has one obligatory primary licenser (always merged, licensing the closest nominal) plus secondary licensers that merge only as a last resort, when their absence would leave some needy nominal unlicensed. DOM is the visible signature of a secondary licenser activating.

The motivating data are from the Neo-Aramaic language Senaya, where DOM surfaces as differential verbal agreement (an L-suffix), not case — and [Kal18] argues case and agreement are two reflexes of one licensing process, so we model the agreement marking abstractly through the licensing substrate's outcome (Syntax/Case/Licensing.lean). The Senaya facts (paper examples around the object-agreement and aspect-split data): a specific object triggers agreement, a nonspecific one does not, and — the crux — in the perfective base the object position is unlicensed, so a specific object (which needs licensing) is banned there entirely.

That perfective ban is [Kal18]'s argument against a no-licensing view of case ([Mar91], [Pre14]): if nominals never needed abstract licensing, the ban would be unexplained. The flagship theorem below states this divergence formally, via the shared Assigner harness (Syntax/Case/Assigner.lean): on the perfective object, a Marantz-style total configurational account assigns a case, while Kalin licensing crashes — the two accounts disagree precisely where licensing is unavailable.

Senaya clause configurations #

The aspect split as licenser availability: the imperfective base offers a secondary licenser for a specific object (yielding DOM agreement); the perfective base offers none (so a specific object cannot be licensed). The assignedCase fields abstract the agreement marking — nom for the primary (subject) relation, acc for the object relation.

Imperfective base: a secondary licenser is available for a specific object.

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

    Perfective base: no secondary licenser — the object position is unlicensed ([Kal18]'s central Senaya claim).

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

      A transitive clause with a specific object: both nominals carry the licensing requirement.

      Equations
      Instances For

        A transitive clause with a nonspecific object: the object lacks the licensing requirement, so it is interpretable in situ.

        Equations
        Instances For

          The Senaya DOM pattern #

          Perfective: a specific object cannot be licensed (no secondary licenser available) and crashes — Senaya's ban on specific objects in the perfective base.

          A nonspecific object does not need licensing, so it is fine in the perfective (licensed trivially by the primary, no DOM marking).

          The flagship divergence: licensing vs. no-licensing #

          On the perfective object, the two accounts assign incompatible verdicts: a Marantz-style total configurational account gives it a structural accusative, while Kalin licensing crashes it (uncased, caseless). This is the witness behind the divergence.

          Licensing diverges from total configurational case assignment ([Kal18] vs [Mar91]). The two accounts disagree on the surface case of the perfective object: the dependent-case account assigns it accusative (case assignment is total — it never crashes), whereas Kalin licensing leaves it unlicensed. [Kal18]'s point: under a no-licensing view the perfective ban on specific objects is unexplained; under licensing it follows, because the perfective object position offers no licenser.