Classification of nominal compounds containing mimetics

Abstract

In Japanese, some nominal compounds have mimetic components (Nominal Compounds with Mimetics (NCMs)) (e.g., zaazaa-buri [mimetic(the sound of heavy rain)-a fall(from the sky)] ‘a downpour’). This paper examines how mimetics participate in word-formation of nominal compounds, applying Construction Morphology. Examination of representative NCMs indicates: (i) NCMs are mostly right-headed, although some are double-headed, and (ii) mimetics combine with the types of nouns that combine with non-mimetic components. Given this, the paper proposes NCMs are part of the inheritance hierarchy for nominal compounds; specifically, their top node diverges according to the head position, building on Booij (2010: 7). The hierarchy consists of different constructional schemas, such as <[x i-hada] nk ↔ [hada ‘skin’ with attribute SEMi]k>, wherein the variable x can be replaced by a mimetic, as in gasagasa-hada ‘rough skin’, or a non-mimetic, as in yawa-hada ‘soft skin’. The paper argues that mimetics are an integral part of nominal compound word formation, enriching lexical varieties of nominal compounds. The Construction Morphology representational system proves useful to indicate where NCMs appear in the word network.