This article starts with the observation that the Hittite 3sg.pret.act. form šipantaš, šipandaš (OH/MS) ‘(s)he libated’ can hardly be analysed as consisting of a tarna-class inflected stem šipant... Show moreThis article starts with the observation that the Hittite 3sg.pret.act. form šipantaš, šipandaš (OH/MS) ‘(s)he libated’ can hardly be analysed as consisting of a tarna-class inflected stem šipant/da- + the 3sg.pret.act. ending -š, since the OH/MH verbal paradigm of ‘to libate’ contains no other tarna-class inflected forms. It is therefore argued that šipantaš, šipandaš should be analysed as consisting of the consonantal verbal stem šipant- + -š, which implies that the a in šipantaš, šipandaš is an empty vowel. In order to explain the spelling -ntaš, -ntaš vs. the spelling -nza, which is commonly used to note down the sequence /-nts/ < PIE *-nts, it is argued that -ntaš, -ndaš denotes /-ntːs/, the regular outcome of a PIE sequence *-nds. Show less
