The Reconstruction of the Proto-Semitic Genitive Ending and a Suggestion on its Origin

The Proto-Semitic genitive ending on triptotic nouns is commonly reconstructed as *-im (unbound state)/*-i (bound state). In Akkadian, however, this case ending is long -ī- before pronominal suffixes. Since the length of this vowel is unexplained, I argue that it is original and that the Akkadian bound state ending -i should also be reconstructed as long *-ī, explaining its retention in word-final position. This form seems more original than Proto-West-Semitic *-i. Hence, the Proto-Semitic bound state genitive ending should also be reconstructed as *-ī. Through internal reconstruction supported by the parallel of kinship terms like *ʔab-um ‘father’, I arrive at a pre-Proto-Semitic reconstruction of the genitive ending as *-ī-m (unbound), *-ī (bound). This paper then explores a hypothetical scenario where the genitive ending *-ī is derived from the adjectivizing ‘nisbe’ suffix through reanalysis of adjectival constructions like *bayt-u śarr-ī ‘the/a royal house’ as construct chains with meanings like ‘the/a king’s house’; with the addition of mimation and the resultant vowel shortening, this yielded the Proto-Semitic construction with a genitive, *bayt-u śarr-im. The genitive case failed to develop with diptotic nouns because they did not take mimation and in the dual and plural because the nisbe adjective was derived from the uninflected (singular) noun stem; hence, these categories all retain the more original contrast between the nominative and and an undifferentiated oblique case.