This study problematizes the notion of “a text” by examining the medieval Hebrew re-translations of the book of Tobit, particularly the two “versions” published by Moses Gaster in 1896. An experiment with different approaches, it combines insights from folklore and literary studies to address the question of when is a text as testified in a particular manuscript or printed book still a telling (a term to be preferred over “version”) of a known text, and when is it better to be considered as a new or different text? Noticing, with Stanley Fish, that a text does not exist outside of the “interpretative communities” can help to “rehabilitate” the often ignored later witnesses to Tobit. “Meanings” of a story are “actualized” as it is read (or told) and adjusted over time in different communities, and attested in different manuscripts and printed editions. The re-appreciation of the medieval Hebrew Tobit tradition proposed in this article thus contributes to a fuller understanding of the story of Tobit.