This paper examines the reasons for a lack of full critical engagement with textual process (the draft materials of the literary work) in Anglo-American scholarship. It engages with French theoretical influences of the late twentieth century, arguing on the one hand that "theories of the text" have had the effect of discouraging systematic study of textual origins but that, on the other, French genetic criticism offers a way forward. The paper gives an overview of key accounts and the relationship between them, as well as looking at the usefulness of French genetic criticism as a methodology. The second half of the paper seeks to develop an alternative position for Anglo-American genetic scholarship which might acknowledge the dangers of an overly expressive or "originary" model but retain space for some degree of agency. Drawing on Barthes' concept of the reader as a "spilt subject" (The Pleasure of the Text, 47) it articulates a readerly and writerly contradiction in relation to draft materials. The Freudian concept of "disavowal" is brought into play as a way of understanding a doubled relationship towards the materials of textual process. Both writer and reader are "split subjects" knowing the autonomy of language, yet responding to the making of the literary work in authorial ways by means of the materiality of the manuscript object.