Remarks on Relational Nouns and Relational Categories
Kow Kuroda, Keiko Nakamoto, Hitoshi Isahara · 2006
It is sometimes suggested that a certain kind of common nouns have their own “arguments,” at least at some conceptual level. For example, in recent research [8, 9] (cf. [2]), Gentner and her colleagues claim that nouns such as victim, barrier, gift have their own arguments. Such nouns are called “relational nouns,” distinguished from “object/entity nouns” that do not have their own arguments. A useful analogy “relational nouns: entity nouns:: verbs: nouns” [9, p. 154] is proposed to test for argumenthood. The distinction is based on their hypothetical distinction between “relational categories” and “object/entity categories.” A similar distinction was independently proposed in [13] between “role(-denoting) names” and “object(denoting) names,” which echoes Gentner’s distinction between relational categories and object categories. In Kuroda-Isahara’s theory, role-denoting nouns refer to either “situational roles” (specific to a “situation”) or “functions” (specific to “components” of an object), while object-denoting nouns refer to objects as “entities.” From a different perspective, Nishiyama proposed a relevant distinction between “saturated” and “unsaturated” nouns [18] and “referential” and “nonreferential” nouns [19], both of which seem to be somehow related to Gentner’s and Kuroda-Isahara’s distinctions.