The compound as a phonological domain in Indo–European

Abstract

Archaic Indo–European compounds show some marked reductions of their individual parts. Phonetic rules can be formulated for the following phenomena: (1) Accented zero–grade in the first member; (2) thematic –o– > –i– in both members, in part with loss of consonantal material, as (2.1) –ro– > –i– (via asyllabic –ry–), which explains much of the ‘Caland system’, and (2.2) ‘root nouns’ replacing agent–nouns with suffixal –t–, and gerundives in –yo– instead of –tyo– (vs. simplex –two–), depending on root structure. The alternations seen in Lat. tellūs : meditullium, NE axle : Lat. axis, the lack of –m– in pronominal feminine forms like Gen.sg. *tesyah2s, and the interchange of ro– and u–stem adjectives are explained as side–effects of the rules which are throughout phonetic, not functional.