Ergativity in Proto-Indo-European

Most Indo-European languages display the nominative-accusative grammatical pattern. It has been suggested by some linguists, however, that PIE had passed through a stage which exhibited an ergative case pattern in a certain type of NPs. Based on the accumulation of knowledge about ergative languages in the world so far, this paper examines the ergative theories of PIE and show that, in general, they are typologically untenable. One possibility which is typologically tenable is that in pre-historical times animate pronouns and nouns exhibited the ergative pattern, but neuter or inanimate nouns had a single non-distinct form for the three functions (transitive and intransitive subject and transitive object), which is attested in IE.