A Constructional Analysis of Old English Nominal Compounds

Analysis of Old English Compounds. English Language and Linguistics 24.3, 1-45. This study provides a constructional analysis of Old English (OE) nominal (N) compounds (8,034 in total) based on the theoretical framework of Construction Morphology (Booij 2010c). A constructional analysis gives the solution to overcome the limitation of a traditional descriptive analysis, which makes a dichotomous classification, ‘Endocentric’ and ‘Exocentric’. This study mainly shows two things. First, OE N compounds are seen as constructions with form and meaning, and hence the formal and semantic properties of both regular and idiosyncratic N compounds can be simply and straightforwardly specified by abstract schemas (templates). Second, with the notion of ‘Default Inheritance’, this study demonstrates that both schemas for regular and idiosyncratic N compounds inherit the default properties from the most abstract schema for OE N compounds ([[X] k [N] i ] Nj ↔ [SEM i with relation R to SEM k ] j ) at the top of the hierarchy: a default override of idiosyncratic compounds is allowed. As a result, a classification of OE N compounds can be made in a unified and simple way with an inheritance tree.