Documentation

Linglib.Phenomena.Modality.Studies.Matthewson2013

@cite{matthewson-2013} — Gitksan Modals #

Lisa Matthewson. "Gitksan Modals." International Journal of American Linguistics 79(3): 349–394. DOI: 10.1086/670751.

Primary source for the Gitksan modal analysis. Three core contributions are formalized here against the existing infrastructure:

  1. Mixed-system thesis (Fig. 1): Gitksan encodes modal STRENGTH in the circumstantial domain (daakhlxw vs. sgi) but not the epistemic domain (imaa/gat are variable-force).

  2. No inherent future orientation (§3.3, §5.3): Gitksan modals are not lexically future-oriented. Future orientation comes from dim, obligatory with circumstantial modals and optional (only for future orientation) with epistemics. This contradicts @cite{condoravdi-2002}'s English analysis, where prospectivity is baked into may. Structurally: Gitksan imaa would be modeled with Condoravdi2002.mayCore (point evaluation), English might with Condoravdi2002.may (forward expansion). The relationship between them is Condoravdi2002.may_of_mayCore_dynamic. We do not introduce alias defs for the Gitksan/English projection here — that is a downstream choice that should land in a typed compositional dim operator (planned, see ProspectiveMarkerPolicy discussion in the integration audit).

  3. No actuality entailments for da'akhlxw (§4.1, fn 32): @cite{hacquard-2006} predicts AEs for the perfective + root-modal configuration. da'akhlxw's obligatory co-occurrence with dim blocks that configuration empirically. The explanation is given in @cite{matthewson-2012}.

Supporting comparisons: Peterson 2010's variable-force analysis of imaa contrasts with @cite{deal-2011}'s strengthened-possibility analysis of Nez Perce o'qa (§3.1, ex. 30 negation diagnostic). The diagnostic content (which scope ¬ takes relative to the modal) is not yet formalized here — currently only the labels.

The modal inventory is in Fragments/Gitksan/Modals.lean. The handbook chapter @cite{matthewson-2016} (Studies/Matthewson2016.lean) restates the survey-level claims; this file holds the primary-source theorems the chapter cites.

@[simp]

@cite{matthewson-2013} Fig. 1, §3.1: imaa is variable-force (Peterson 2010 analysis).

The mixed-system signature: circumstantial modals contrast in force, epistemic modals do not. The asymmetric encoding pattern is the paper's central typological observation (Fig. 1).

The flavor-keyed dim asymmetry from §3-4 lives in Fragments/Gitksan/Modals.lean (requiresDim_imaa_*, requiresDim_gat_*, requiresDim_circumstantial, dim_flavor_asymmetry). The deeper structural claim — that Gitksan modals project to Condoravdi2002.mayCore rather than Condoravdi2002.may — is currently expressed in the module docstring above; making it a typed compositional theorem requires promoting dim to a Theories-level operator.

@cite{hacquard-2006} predicts AEs in the configuration belowAsp + perfective. @cite{matthewson-2013} reports da'akhlxw lacks AEs. Per @cite{matthewson-2012}: da'akhlxw obligatorily co-occurs with prospective dim, blocking the perfective configuration empirically.

The §4.1 fn 32 explanation, schematically: da'akhlxw's obligatory dim co-occurrence (via requiresDim_circumstantial in Fragments/Gitksan/Modals.lean) means the perfective configuration that drives Hacquard's AE prediction is empirically inaccessible for this modal. The full structural realization — dim as a typed combinator that blocks the perfective configuration — requires the planned dim-as-operator refactor; currently this is asserted via the requiresDim policy, not derived.

@cite{matthewson-2013} §3.1 follows Peterson 2010 in analyzing imaa as variable-force. @cite{deal-2011} analyzes Nez Perce o'qa as strengthened possibility. The two analyses agree both modals admit necessity readings but disagree on the mechanism. The downward- entailing diagnostic (paper ex. 30) is consistent with Peterson's analysis for imaa: negated imaa yields "possibly not", i.e., the modal scopes above negation.

@cite{matthewson-2013} Fig. 4 (p. 369) cross-tabulates temporal perspective (past/present) with temporal orientation (past/present/ future) for the two epistemic modals. The two axes are the canonical Core.Modality.TemporalPerspective and Core.Modality.TemporalOrientation opened above.

A Figure 4 cell: a temporal perspective × orientation pair, with the paper's example number for grounding. The dim-requirement is NOT stored — it is derived from the orientation via the flavor-keyed requiresDim policy.

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

      Whether dim is required at this Fig. 4 cell, derived from the flavor-keyed policy on imaa (epistemic → required iff future).

      Equations
      Instances For

        The six cells of Figure 4 for imaa, with example numbers verified against the actual figure on p. 369. The figure also shows gat entries in the past-temporal-perspective row (47, 47, dim gat 48); those are not encoded here — this list is imaa-specific. The future-orientation cells (44, 42) are notated "ima('a) dim" in the figure, encoding the obligatory co-occurrence with prospective dim (which Fig4Cell.dimRequired recovers from requiresDim).

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