Strutture morfoprosodiche del latino tardo nell’Appendix Probi

The work consists in the analysis of the linguistic content of two fragments of the Appendix Probi included in the manuscript Neapolitanus 1 from Bobbio (nrr. 3 and 4 according to Barwick’s numeration). The study starts from the new edition of the fragments, edited by Marina Passalacqua who has rectified Keil's edition in more than one point. The two short texts have never been subject to an accurate linguistic interpretation (Barnett 2006 is mainly a philological work), despite the fact that they are contiguous to the very famous antibarbarus (nr. 5 Barwick). According to the recent study forwarded by Mancini (2007), it is possible to date the works to the fifth century d.C. The fragment nr. 3 puts various «generis masculini» nouns together. Fragment nr. 4 lists some entries (mainly proper nouns of Greek origin) for which the paroxytone stress («cum accentu producto») or the proparoxytone one («cum accentu correpto») are preferred . One can infer the linguistic value of these two fragments for the reconstruction of the spoken Latin by applying a kind of reading ‘e contrario’, based on a comparison with what the late Roman grammarians included in their Artes. The lexemes of the two fragments only acquire true meaning when interpreted as correct forms, in strong opposition with stigmatized forms. The latter can also be found in late Roman grammars. In fact, the substantives found in the fragment nr. 3 use varied gender attribution in spoken Latin; likewise in fragment nr. 4, we can find some very important, and until now unknown, evidence regarding the assignment of the word-stress in Greek loanwords (Heliodorus Sibylla, tyrannus, etc.) and in some entries with the group 'muta cum liquida' (delubrum, baratrum) typical of late spoken Latin of the fifth century d.C.