Loading field map
What is a text? Is it the words an author intended, the marks on a manuscript page, the collaborative product of a printing house, or something that only exists in the act of reading? Textual scholarship is the field that asks these questions not as abstract philosophy but as practical problems: every time an editor chooses one variant over another, decides which version to use as a base, or presents a text in a particular format, they are making a claim about what a text is and how it should be treated. The history of the field is a history of competing answers to that pressure, each framework arising from the limits of its predecessors and each leaving its mark on how we edit, read, and understand written works.
The first systematic framework for establishing reliable texts emerged from the study of ancient and medieval manuscripts. Stemmatics, developed most influentially by Karl Lachmann in the mid-nineteenth century, treated textual transmission as a genealogical tree. The scholar would collate all surviving manuscripts of a work, identify shared errors that revealed common ancestry, and reconstruct a stemma—a branching diagram showing how copies descended from lost ancestors. The goal was to recover the archetype, the earliest recoverable form of the text, by eliminating later corruptions. In practice, a stemmatic editor worked like a detective: variants were evidence of descent, and the stemma told the editor which readings were oldest and therefore most authoritative.
Stemmatics assumed that textual transmission was largely linear and that contamination—scribes copying from more than one exemplar—was rare enough to be manageable. By the early twentieth century, scholars working with vernacular traditions, especially medieval romance and early modern drama, found that contamination was the rule rather than the exception. A stemma could not represent a manuscript that drew on two different branches of the tradition. The framework's authority narrowed as its core assumption about linear descent lost credibility, but its basic logic—that textual relationships can be modeled as descent from common ancestors—survives today in computational phylogenetics and in the editorial practice of any scholar who still uses shared error as evidence of relationship.
New Bibliography emerged in the early twentieth century as a response to the limitations of stemmatics when applied to printed books. Where stemmatics had been designed for manuscript traditions with many copies and no authorial holograph, New Bibliography confronted the opposite situation: printed editions, often with authorial manuscripts or corrected proofs surviving, but with multiple editions produced during the author's lifetime. The framework's central figures—A. W. Pollard, W. W. Greg, and Fredson Bowers—shifted the editor's goal from recovering an archetype to recovering the author's final intentions.
Greg's 1949 essay "The Rationale of Copy-Text" became the framework's methodological cornerstone. Greg distinguished between substantives (the words themselves) and accidentals (spelling, punctuation, formatting). He argued that the editor should choose as copy-text the edition closest to the author's manuscript in accidentals—usually the first edition—but could emend its substantives using later editions if the author had revised them. This gave editors a principled way to combine the authority of the earliest witness with the author's later revisions. New Bibliography treated the printed book as a potentially corrupt but recoverable vehicle for authorial meaning, and the editor's task was to strip away the interventions of compositors, publishers, and censors to restore what the author intended.
By the 1970s, New Bibliography's dominance was under pressure. Critics pointed out that "final intentions" was a fiction for many works—authors revised erratically, left multiple versions, or collaborated with others in ways that made a single intention impossible to isolate. The framework had narrowed textual scholarship to a single authorial model, and that model was beginning to crack.
The 1980s saw the simultaneous emergence of two frameworks that rejected New Bibliography's core commitments, but from opposite directions. Genetic Criticism and Social Textual Criticism both argued that the single-authorial-intention edition was an artificial construct, but they disagreed fundamentally about what should replace it.
Genetic Criticism, developed primarily in France by scholars such as Louis Hay and Almuth Grésillon, turned the editor's attention away from the finished work and toward the process of composition. The object of study became the avant-texte—the full corpus of drafts, notes, sketches, marginalia, and revisions that precede the published text. Where New Bibliography had treated variants as errors to be eliminated, genetic criticism treated them as evidence of creative thinking. A genetic editor does not produce a single corrected text; instead, they document the layers of writing, showing how the work came into being.
In practice, a genetic critic works with facsimiles, transcriptions, and genetic dossiers that map the sequence of compositional stages. The framework assumes that the text is not a fixed object but a process, and that meaning resides in the movement from one version to another. This directly challenged New Bibliography's assumption that the author's final intention was the only legitimate editorial goal. For genetic criticism, the drafts are not inferior to the published text—they are the text in its fullest sense.
Social Textual Criticism, articulated most forcefully by D. F. McKenzie and Jerome McGann, argued that texts are never the product of a single author but are always socially produced. A published book involves authors, editors, compositors, publishers, booksellers, and readers, each of whom shapes the text. McGann's 1983 book A Critique of Modern Textual Criticism directly attacked New Bibliography's author-centered model, arguing that the editor's task was not to strip away social interventions but to document them.
Where genetic criticism looked inward at the author's creative process, social textual criticism looked outward at the institutional and material forces that produce texts. A social textual editor produces an edition that records not just variant readings but the full history of the text's production and reception: who set the type, who paid for the printing, who censored what, and how readers responded. The framework assumes that meaning is distributed across a network of agents, and that the editor's job is to make that network visible rather than to erase it in favor of a single authorial voice.
Genetic Criticism and Social Textual Criticism share a rejection of New Bibliography's single-authorial-intention model, but they diverge sharply in their positive commitments. Genetic criticism treats the text as a temporal process unfolding in the author's mind and hand; social textual criticism treats it as a spatial network of social relations. A genetic editor might produce a digital edition showing every stage of a poem's composition; a social textual editor might produce an edition that includes the publisher's ledger, the compositor's bills, and the reviews alongside the text itself. The two frameworks have influenced each other—some genetic critics now attend to the social conditions of composition, and some social textual critics attend to the process of revision—but they remain distinct in their primary focus and in what they consider the most important evidence.
Digital Textual Scholarship emerged in the 1990s as computing became powerful enough to handle large textual corpora. At first, it seemed like a purely instrumental development: computers could collate texts faster, store more variants, and produce editions that were searchable and hyperlinked. The Text Encoding Initiative (TEI), founded in 1987, developed a standard markup language for encoding texts in digital form, and by the 1990s scholars were producing electronic editions that could include facsimiles, transcriptions, and apparatus in ways impossible in print.
But digital textual scholarship quickly revealed itself as a framework with its own interpretive commitments. Encoding a text requires decisions about what counts as a variant, what counts as a version, and what counts as a text. TEI markup, for example, forces the editor to decide whether a crossed-out word is a deletion to be recorded or a stage in composition to be preserved. The digital edition does not simply present the text; it models the text, and the model reflects the editor's theoretical commitments. A digital edition built on genetic principles looks very different from one built on social textual principles, and both differ from a digital edition that aims to present a single reading text.
Digital Textual Scholarship has largely been absorbed as infrastructure by both genetic and social textual criticism. Most genetic editions today are digital, and most social textual editions use digital tools to manage their complex apparatus. But the framework also retains its own distinctive contribution: it has made textual scholarship more transparent by forcing editors to make their encoding decisions explicit, and it has made textual scholarship more collaborative by enabling multiple scholars to work on the same corpus. The tension within digital textual scholarship is whether the digital edition should aim to represent the text as faithfully as possible or to enable as many different readings as possible—a tension that echoes the older debate between recovering authorial intention and documenting social production.
Material Philology, also emerging in the 1990s, took a different path. Developed by scholars such as Stephen G. Nichols and the contributors to the journal Speculum's 1990 special issue on "The New Philology," Material Philology argued that the manuscript itself—the physical object—is the text, and that any edition that abstracts the words from the manuscript's material form is a distortion. Where stemmatics had treated manuscripts as witnesses to a lost original, Material Philology treated each manuscript as a unique artifact with its own authority.
In practice, a material philologist does not produce a critical edition. Instead, they study the manuscript as a physical object: its parchment, its ink, its layout, its marginal annotations, its binding, its ownership marks. The framework assumes that meaning is inseparable from material form—that the layout of a page, the quality of the illumination, and the wear of use are all part of what the text means. This puts Material Philology in direct tension with every editorial framework that produces a critical edition, because the edition necessarily abstracts the text from its material context. For the material philologist, the edition is not a solution but a problem.
Material Philology revives elements of pre-stemmata manuscript study—the careful description of individual codices—but with a new theoretical sophistication drawn from book history, art history, and anthropology. It coexists uneasily with Digital Textual Scholarship, which often treats the manuscript as a source of data to be extracted, and with Social Textual Criticism, which attends to material conditions but still produces editions. The living disagreement is whether the edition is a necessary tool for making texts accessible or an inherently distortive practice that should be abandoned in favor of direct engagement with the material artifact.
Four frameworks remain active today: Genetic Criticism, Social Textual Criticism, Digital Textual Scholarship, and Material Philology. Stemmatics and New Bibliography survive in transformed form—stemmatics as computational phylogenetics, New Bibliography as a set of editorial principles still used in some scholarly editions—but they no longer define the field's leading edge.
The leading frameworks share a fundamental agreement: editing is always interpretive. No edition is neutral; every editorial choice reflects a theory of what the text is and how it should be treated. This consensus is the legacy of the 1980s watershed, which permanently broke the assumption that editing could be a purely technical or objective procedure.
Where the frameworks disagree is on where meaning resides. For genetic criticism, meaning is in the process of composition. For social textual criticism, meaning is in the network of production and reception. For material philology, meaning is in the physical artifact. Digital textual scholarship, as infrastructure, does not take a side in this debate but forces every editor to make their commitments explicit. The unresolved question—whether editions are necessary or inherently distortive—continues to drive the field forward, and the answer depends on which framework's model of the text you adopt.