Hungarian possession profile #
[Sta09b] [Nic86] [Hei97] [KVF98] [Rou01]
Per-language possession data for Hungarian (Uralic; ISO hun), per the
project's "per-language data flows through Fragments" rule, as bare
field-by-field defs. Substrate types (PredicativeStrategy,
AdnominalMarking, …) live in Linglib/Features/Possession.lean.
Cross-linguistic theorems consuming these values live in
Studies/NicholsBickel2013.lean.
Examples: nekem van konyvem, Janos kalap-ja. Dative possessor + van 'exists' + head-marked possessum; possessive suffixes obligatory on relational nouns.
The adnominalStrategy := .headMarking choice is consistent with both
standard reference grammars: [KVF98] §1.10
analyzes possession via the possessive suffix on the possessum
(Péter vers-e 'Peter's poem' = "Peter poem-POSS"), with the optional
dative possessor (Péter-nek a vers-e "Peter-DAT the poem-POSS")
when extracted for specificity reasons (per Szabolcsi 1986/1992, 1994,
as cited there). Both KVF and [Rou01] treat the
possessor-marker as dative, not genitive — Hungarian has no
morphological GEN, and Fragments/Hungarian/Case.lean reflects this
by omitting .gen from its caseInventory.