Gospels before the Book is a book about unfinishedness and unfinalizability. Our approach to the early Christian gospels particularly suffers from an undertheorized appreciation of unfinishedness. Thus, my book offers a new narrative for the emergence and reception of early Christian gospels in general and the textual tradition we now call the Gospel according to Mark in particular. Before previewing the methodologies and conclusions, one might ask: what are the stakes of the project?
Pick up almost any monograph, commentary, or article on the Gospel according to Mark (or, for that matter, the Gospel according to Matthew, Thomas, and so forth), and you will find such concepts as book and author at work within the scholarly discourse. A standardized set of historical questions are brought to bear on the text: Who wrote it? When did the person write it? From where did the person write? Who is the intended audience? In seeking answers to such questions, scholars try to locate a specific time and place of origin where an author worked on and finished a book. From such an origin, then, historians try to configure a scenario in which the meaning of the text can be attained in its proper first century context.
All these inquiries have a priori assumptions that are rather modern and “bookish.” They presuppose an author producing some sort of finished book at a particular time and for a specific situation. It is not uncommon to see scholars even use the word publication, often in scare quotes, to talk about a moment when a gospel was first put into circulation.
There is a subtle yet serious problem here. There is no evidence of someone regarding the gospel as a discrete, stable, finished book with an attributed author until the end of the second century CE, and a gospel qua discrete authored book does not really become a dominant discourse for talking about “the gospel(s)” until the third century CE. In other words, although a fair bit of the billowing and pluriform “gospel” tradition became textualized in the first and second century CE, there is no evidence of the idea of gospel as a finished book published by an author until much later.
The earliest evidence for the latter comes from the Christian apologist Irenaeus of Lyon around 180–190 CE. In Against Heresies 3.1.1, Irenaeus defends his gospels by describing them as published books, created in specific times and places by known authors, traceable back to stable and distinct origin points. While his approach may seem intuitive to modern readers, his comments stand in stark contrast to other discourses of gospel textualization and authorship. For others in the second century like Celsus or Justin Martyr, and in works like the Didache and 2 Clement, the gospel is a textualized tradition, but the configurations and metaphors for making sense of that tradition are far too unbounded, uncontained, and open to be a good categorical match with concepts like book, author, and publication. In the first century and throughout most of the second century, the gospel, though textualized, nevertheless remains contingent, malleable, and subject to change—more rhizomatic than arboressent.
For instance, Celsus, the second century Christian opponent, based on a comment recorded in Origen, Cels. 2.27, seems to have viewed the gospel as a textual tradition that, while it did have an initial textual shape, was in a constant state of flux and reworking; an open textual tradition and not a stable and discrete book (or books) attributed to an author. Similarly, when Theophilus of Antioch cites gospel tradition in Autol. 3.13–14, he does not cite an author-figure or (to use the language of Eva Mroczek) employ the metaphor of the book. Rather he imagines the gospel tradition as a textual constellation issuing forth directly from “the gospel voice” (hē… euaggelios phōnē). In that context, the gospel voice functions analogously to such author-figures as Isaiah, Jeremiah, Hosea, Joel, and Zechariah (3.12). While the gospel tradition cited in 3.13–14 bear strong family resemblances to bits of the textual traditions we now call the Gospels according to Matthew and Luke (and perhaps Q), the gospel is not envisioned here as a book published by an author.Rather the gospel in Autol. 3.13–14 is personified and somehow self-authorizing. For many throughout the first few centuries of the Common Era, textualized gospel tradition constituted a diverse, living, mushrooming textual constellation (or constellations) that has been altered and remodeled time and time again.
Yet many modern critical scholars of first century CE gospel texts participate, unwittingly perhaps, in later Christian discourses about the textuality and authorship of the gospels, presupposing and prioritizing a definitive moment when a stable book and a unique author simultaneously came into existence. The final version of the text and the author’s intention are conjoined, locked in time for later readers to decode. The earliest discourses about gospel textuality and authorship, as we see above, do not share such assumptions. At the foundation, then, of the historical-critical questions mentioned above lies a misleading anachronism regarding books, authors, and publication. Ignoring, or at least unaware of, the disjointed discourses about gospel textuality and authorship within the first centuries of the Common Era, modern historians of ancient Christianity speak about first century gospel texts in ways unknown in the first and second century discourses about the gospel.
Ironically, then, while modern critical scholars deny the historical reliability of ancient Christian apologists’ constructions about the historical origins of the gospels, many have been blind to the same apologists’ more significant ideological invention: the very idea that there are separate, finished, and fully authored books called the Gospels according to [Name], as opposed to a more fluid constellation of written residue. What if we take Celsus, Justin Martyr, and others seriously, not equating the process of textualization of the gospel with the creation of gospel “books” nor the producers of textualized gospel tradition with “authors”? What if gospel texts were not (at least not initially) stable, finished, discrete books? How would our constructed knowledge about “the gospels” shift? What would it look like to construct a narrative of gospel production without recourse to ideas like books, author figures, or publication? What would it mean to think about the textual tradition we now call the Gospel according to Mark before the book, before authors, and before publication?
The evidence, I argue, suggests a first- or second-century reader of the textual traditions we now call the Gospel according to Matthew and Gospel according to Mark would not have thought of them as two separate books by two different authors. Rather they would have regarded them as the same open-ended, unfinished, and living work: the gospel—textualized. Thus, the validity and utility of source-, redaction-, and textual-criticism as traditionally practiced are called into question. For example, what does it mean to talk about the “Synoptic Problem” without recourse to ideas like books, authors, and textual finality?
In order to reimagine the narrative of gospel textualization and proliferation in ways native to the first and second century discourses of textual fluidity and growth, one must explore some often underappreciated writing practices and conceptions of authorship throughout the ancient Mediterranean world. Here, I return to the complex constellation of ideas which I signaled above: textual unfinishedness, unauthored texts, “publication,” textual revision, and a variety of diverse uses and functions of different kinds of texts.
In Gospels before the Book I attend to ancient writing and reading practices around such concepts among different writers and textual objects from Jewish, Greek, and Roman contexts across the Mediterranean basin. I then turn to the earliest readers and users of the textual tradition we now call the Gospel according to Mark. I show how they regarded and used it not as a finished and stable book published by an author, but as an unfinished and fluid constellation of notes (hypomnēmata/commentarii).
This has implications for how the textual traditions we call the Gospels according to Luke and Matthew relate to the Gospel according to Mark, as well as how we make sense of the pluriform endings in the textual tradition of the Gospel according to Mark. My approach works to break down binaries (e.g. “right” vs. “wrong” endings or the discreteness of book A vs. book B) and invites readers to think beyond such binaries in a more textured way about the open-ended process of gospel writing. In the end, I question the validity of drawing of hard divisions between a text, the tradition that accretes around it, and its modern scholarship.
My approach situates the gospels in general, and the textual tradition we have come to know as the Gospel according to Mark in particular, in a moment in the first few centuries CE when gospel is still primarily a speech genre. It is something one speaks. This gospel, even when already textualized, is also still oral, still pliable, still open. The textualization of the gospel is an ongoing process, an energy. The reader becomes paradoxically the unfinalizable ending and the “perpetual present” of the Gospel according to Mark. Whether or not the textual tradition we now call the Gospel according to Mark was an intentionally unfinished work or not, ancient and late antique readers and users often regarded it as one, and modern readers may treat it as such, if they wish.
 Ad. haer. 3.1.1 is discussed in detail in chapter five of Gospels before the Book.
 See, for example, Origen, Cels. 2.27; Justin Martyr, Dial. 101.3; 106.4 et al; Did.8.2 et al; 2 Clem. 8.5.
 Gilles Deleuze and Félix Guatarri, A Thousand Plateaus: Capitalism and Schizophrenia 2, trans. Brian Massumi (Minneapolis: University of Minnesota, 1987), 15: “We're tired of trees. We should stop believing in trees, roots, and radicles. They've made us suffer too much. All of arborescent culture is founded on them, from biology to linguistics. Nothing is beautiful or loving or political aside from underground stems and aerial root, adventitious growths and rhizomes.”
 Matt 5:28, 32, 44, 46; Luke 6:28; 32–33; 16:18.
 On “perpetual present” of writerly texts, see Barthes, S/Z, 5. See also Michael J. Thate, Remembering Things Past?: Albert Schweitzer, the Anxiety of Influence, and the Untidy Jesus of Markan Memory (Tübingen: Mohr Siebeck, 2013), 270: “With these approaches in mind, something startling happens for the Reader: they become the star of the Markan Script, and themselves become the missing ending / new beginning of Mark’s Gospel.”
 Bakhtin, Problems of Dostoevsky’s Poetics (Minneapolis: University of Minnesota Press, 1984), 166: “Nothing conclusive has yet taken place in the world, the ultimate word of the world and about the world has not yet been spoken, the world is open and free, everything is still in the future and will always be in the future.”
Matthew Larsen is a Cotsen Fellow in the Princeton Society of Fellows and a Lecturer in the Department of Religion at Princeton University. He is a cultural historian whose work focuses on the lived experiences of Jewish and Christian communities in antiquity and late antiquity. Matthew earned a Ph.D. in religious studies from Yale University in 2017. The contents of this article are adapted from his book, Gospel before the Book (Oxford University Press, 2018). His work has also been published in the Journal for the Study of Judaism, the Journal for the Study of the New Testament, the Journal of Early Christian Studies (forthcoming), and various other venues. He is currently working on a project on early Christians and incarceration.