What does it look like to read the texts we now call the gospels like first- and second-century readers? There is no evidence of anyone regarding the gospel as a book published by an author until the end of the second century. So, put differently, what does it mean to read the gospels "before the book"? For centuries, the ways people discuss the gospels have been shaped by later ideas that have more to do with the printing press and modern notions of the author than ancient writing and reading practices. In Gospels before the Book, Matthew D. C. Larsen challenges several subtle yet problematic assumptions about authors, books, and publication at work in early Christian studies. He then explores a host of under-appreciated elements of ancient textual culture such as unfinished texts, accidental publication, post-publication revision, and the existence of multiple authorized versions of the same work. Turning to the gospels, he argues that the earliest readers and users of the text we now call the Gospel according to Mark treated it not as a book published by an author, but as an unfinished, open, and fluid collection of notes (hypomnmata). In such a scenario, the Gospel according to Matthew would not be regarded as a separate book published by a different author, but as a continuation of the same unfinished gospel tradition. Similarly it is not the case that, of the five different endings in the textual tradition we now call the Gospel according to Mark, one is "right" and the others are "wrong." Rather each represents its own effort to fill a perceived deficiency in the gospel. Larsen offers a new methodological framework for future scholarship on early Christian gospels.