A large group of compound verbs in Ancient Greek such as στεφανη-πλοκέω, στρατ-ηγέω, and νομο-θετέω, can be seen, from a synchronic point of view, as noun-incorporated verbs. The morphological phenomenon of noun incorporation is well known mostly from polysynthetic languages, such as Native American languages. However, looking at this phenomenon typologically (based mostly on Mithun 1984, but also on other typological and theoretical work), one can argue that Ancient Greek also displays this morphological process.
These verbs have been traditionally analyzed as denominative verbs, derived from verbal governing nouns or adjectives. However, for many of these verbs – most of them hapax legomena – the purported noun from which they were allegedly derived is not attested in our corpora. The talk will strive to show that these verbs can be better analyzed and understood as noun incorporated verbs, all sharing the same basic morphology of a nominal component followed by a verbal component.
In many cases it is possible to outline direct evidence for a noun incorporation analysis of these verbs. Often we see in texts the independent noun (e.g., ἀγών), the independent verb (e.g., τίθημι), and then the noun-incorporated verb (e.g., ἀγωνο-θετέω as in D.9.32). The incorporation also occurs across two different texts, such as in comments made by the scholia (e.g., scholia ad Ar.Nu.997 & Ar.Th.400-1) or in ancient translations (e.g., in the Septuagint). In cases like this the verbal compounding reveals the synchronic status of this verbs as noun-incorporated and not merely as denominative. In addition, in a few cases (e.g., Hdt.4.194.1) these verbs do not function as ‘anaphoric islands’ (Postal 1969) and the incorporated noun is actually referred to by anaphoric pronouns later on in the text.
From a morpho-syntactic point of view, the incorporation of the semantic patient of a transitive verb causes the argument structure of the verb to become saturated, that is to say most noun-incorporated verbs are intransitive (e.g., στεφανη-πλοκέω). However, in some well-defined exceptions – attested also typologically – these verbs indeed do take a full-fledged noun as a direct object, such as in cases of lexicalization (e.g., χειρο-τονέω), ‘case manipulation’ (e.g., δειρο-τομέω, in which the possessor of the body part appears in the acc., while in the non-incorporated version it appears in the gen.), and classificatory incorporation (e.g., οἰκο-δομέω, in which the acc. complement can be either truly cognate with the base noun (e.g., οἰκία) or merely semantically cognate, i.e. a type of substantial building like νηόν). These limitations on the argument structure of the verb are best accounted for under a noun incorporation analysis.
Due to the prevailing historical approach to the Classical languages, this phenomenon has hardly been studied in Ancient Greek (though it was studied in Latin: Fruyt 1990 & Fugier 1994). The only significant exception is Pompei 2006, but this article does not fully explore the phenomenon and its morpho-syntactic ramifications. The suggested talk will adduce additional examples and arguments in order to show that one can actually add this phenomenon to the inventory of Ancient Greek morphological mechanisms.
Greek and Latin Linguistics