From Eusebius of Caesarea, who first compiled a collection of martyr narratives around 300, to Thierry Ruinart, whose Acta primorum martyrum sincera et selecta was published in 1689, the selection and study of early hagiographic narratives has been founded on an assumption that there existed documents written at the time of martyrdom, or very close to it. As a result, a search for authenticity has been and continues to be central, even in the context of today's secular scholarship. But, as Eric Rebillard contends, the alternative approach, to set aside entirely the question of the historical reliability of martyr narratives, is not satisfactory either. Instead, he argues that martyr narratives should be consider as fluid "living texts," written anonymously and received by audiences not as precise historical reports but as versions of the story. In other words, the form these texts took, between fact and fiction, made it possible for audiences to readily accept the historicity of the martyr while at the same time not expect to hear or read a truthful account. In The Early Martyr Narratives, Rebillard considers only accounts of Christian martyrs supposed to have been executed before 260, and only those whose existence is attested in sources that can be dated to before 300. The resulting small corpus contains no texts in the form of legal protocols, traditionally viewed as the earliest, most official and authentic records, nor does it include any that can be dated to a period during which persecution of Christians is known to have taken place. Rather than deduce from this that they are forgeries written for the sake of polemic or apologetic, Rebillard demonstrates how the literariness of the narratives creates a fictional complicity that challenges and complicates any claims of these narratives to be truthful.