Traumdeutung: grec ὄναρ, ὄνειρος, arm. anowrǰ, etc.

International audienceThe paper discusses recent attempts at etymologizing Gk. ὄναρ, Arm. anowrǰ, etc. ‘dream’, and refutes the hypothesis of any connection of its basis with the root *h2enh1- ‘to breathe’. After reviewing the typology of designations of the dream in Indo-European languages, one explores the possibility of deriving the basis of the cognate nouns of Greek and Armenian from the notion of ‘message’. It is proposed to connect the Latin noun ōmen, as derived from a root *h2eh3- (> Hitt. ḫā-) meaning ‘to believe, to trust’. The point of departure would be a delocatival derivative based on an original neuter *-men-stem *h2éh3-m(e)n-, which had been restructured, according to a well-known process, as *-(e)n- stem, hence an adverbial locative *h2h3én-er ‘in dream’ ; the latter was the source of the back-formed neuter stem *(h2)h3ón-r ‘dream’ that yielded the Greek form. Accordingly, the basis *Hanōr of Arm. anowrǰ reflects the amphikinetic stem *h2éh3nōr, leveled as *h2h3nōr (> *h2nōr, with Lindeman variant), which was internally derived from the neuter stem. These two allomorphs can account for all attested forms. Finally it is claimed that the amphikinetic stem did not have collective function: it was specifically animate in the first place. This substantive referred to the dream as divine agent or as personified force, delivering a message which is supposed to be believed