03_Samet_375_375_390 30/10/15 13:05 Pagina 386 386 NILI SAMET 386 Babylonian text but is absent from the older versions, and two more passages that were part of the Old Babylonian text, and are lacking in the later recension. We seem to be dealing here with indirect evidence for the existence of a peculiar version of Gilgamesh, not known to us in its original form, which included ancient components from the early second millennium BCE. Our assumption may find support in Stephanie Dalley’s recent study of the transmission of canonical cuneiform texts in the Se- leucid period as reflected in Berossos’ writings 34. As demonstrated by Dalley, the third century BCE saw the coexistence of different Gilgamesh narratives, and the one used by Berossos was not iden- tical with Sin-Leqe-Unninni’s so-called canonical SB version. Moreover, Dalley shows that the version reflected in the writings of Berossos included details that, while absent from the SB version, are known to us from other literary compositions, and may go back to as early as the 11th century BCE 35. On the methodological level, Dalley suggests abandoning the model of single authorship and canonical text, and adopting instead a more polyphonic conception of coexisting different versions, in both earlier and later stages of the epic’s life 36. In addition to Dalley’s insightful conclusions, it is noteworthy that ancient traditions, which deviate from allegedly canonical versions of famous literary works, are often likely to be found in the periphery of cuneiform culture rather than in its main- stream centers 37. Recent scholarship of another cuneo-biblical case study, which involves Berossos yet again, has interestingly presented similar conclusions. John Day has pointed to several traits of the flood story that are shared by the biblical source P and Berossos, while 34 S. DALLEY, “First Millennium BC Variation in Gilgamesh, Atrahasis, the Flood Story and the Epic of Creation: What was Available to Berossos?”, The World of Berossos. Proceedings of the 4th International Colloquium on the Ancient Near East between Classical and Ancient Oriental Traditions, Hatfield College, Durham, 7th – 9th July 2010 (eds. J HAUBOLD et al.) (Clas- sica et Orientalia 5; Wiesbaden 2013) 165-176. 35 S. DALLEY, “Variation”, 167. 36 A similar suggestion, though more cautiously formulated and restricted to oral tradition, is made by Andrew George in his edition of the epic. See GEORGE, Gilgamesh I, 56. 37 See e.g.: Y. COHEN, Wisdom of the Late Bronze Age (WAW 34; Atlanta, GA 2013) 124-127.