Appearance
Choose a documentary edition when your goal is to represent a single witness faithfully — every reading, deletion, and layout feature as it stands — and a critical edition when your goal is to reconstruct an authoritative text from multiple witnesses, recording their disagreements in an apparatus. The deciding question is whether the document or the work is your object of study. In digital form you are not strictly locked in: TEI lets you encode witnesses documentarily and still generate a critical view, but you must pick a primary intent before you design your schema.
What does each edition type actually deliver?
A documentary edition is a transcription of one physical witness, presented as it is — diplomatic in its fidelity, including scribal errors and false starts. A critical edition presents a text that may match no single surviving witness, because the editor has weighed the variants and chosen what they believe the author wrote, banishing the alternatives to a footnote apparatus. The documentary editor says "this is what the page shows"; the critical editor says "this is what the text should read, and here is my evidence."
How do I decide between them?
Work through three questions, in order:
- Is there one authoritative witness, or many that disagree? One witness pushes you toward documentary.
- Are your users evidential (archivists, palaeographers) or interpretive (literary scholars)? Evidential users want the witness.
- Do you have the comparative scholarship to justify editorial choices? Without it, a critical edition is hard to defend.
| Factor | Documentary | Critical |
|---|---|---|
| Witnesses | One, central | Many, collated |
| Editorial intervention | Minimal, transparent | Substantial, justified |
| Apparatus | Optional notes | Required variant apparatus |
| Effort driver | Layout/diplomatic fidelity | Collation across witnesses |
| Primary audience | Archival, evidential | Interpretive, literary |
What does a documentary encoding look like?
Documentary work prioritises the document's surface. TEI's <sourceDoc> model lets you anchor text to physical zones rather than a logical structure:
xml
<surface ulx="0" uly="0" lrx="2480" lry="3508">
<zone xml:id="z1" ulx="120" uly="200" lrx="2300" lry="480">
<line>Deare Sister, I haue receaued your <del>letre</del> letter</line>
</zone>
</surface>Notice that the <del> stays in the text — the deletion is evidence, not noise to be cleaned. That is the documentary instinct.
What does a critical encoding look like?
Critical work prioritises the variation. The same passage, collated against three witnesses, becomes an <app>:
xml
<app>
<lem wit="#A #B">received</lem>
<rdg wit="#C" type="orthographic">receaued</rdg>
</app>Here the editor has chosen "received" as the lemma and demoted the witness spelling to a reading. The document's exact surface is gone from the main text; it survives in the apparatus.
Can I have both without doubling the work?
Yes — encode each witness once, documentarily, and let collation software build the critical layer. Tools like CollateX take several plain transcriptions and emit a variant graph you can serialise as TEI <app> markup:
bash
collatex --format tei --output app.xml witnessA.txt witnessB.txt witnessC.txtThis is the digital advantage: your documentary transcriptions are the raw data, and the critical apparatus is a generated product you can regenerate whenever a witness is corrected.
When does the choice bite an archivist specifically?
If your institution holds the single surviving autograph of a letter, a documentary edition preserves its evidentiary integrity — users can cite it as the witness. A critical edition would conflate it with later printed copies and weaken that evidential chain. Conversely, for a heavily copied text where no manuscript is authoritative, a documentary edition of an arbitrary witness misleads readers into treating one copy as canonical. Match the edition to the evidentiary status of your holdings.
Key Takeaways
- Documentary editions represent one witness faithfully; critical editions reconstruct a text from many.
- The deciding question is whether the document or the work is your object of study.
- Documentary work uses
<sourceDoc>and zones; critical work uses thetextcritapparatus. - Documentary fidelity keeps errors and deletions visible as evidence.
- Digital editions can be hybrid: encode witnesses documentarily, generate a critical view.
- CollateX can build a TEI apparatus from several plain transcriptions automatically.
- Archivists usually prefer documentary editions to preserve evidentiary value.
Frequently Asked Questions
What is the core difference between a documentary and a critical edition?
A documentary edition reproduces one witness exactly as it stands, including its errors. A critical edition reconstructs an editor's best text from many witnesses, recording variants in an apparatus. One is faithful to a document; the other is faithful to a hypothesised work.
Which edition type is faster to produce?
Documentary, usually. You transcribe a single witness without collation, so you skip the most labour-intensive critical step. But high-fidelity documentary work — capturing every deletion, layout zone, and abbreviation — can rival a critical edition in effort.
Can a digital edition be both documentary and critical?
Yes, and digital media make hybrids practical. You can encode each witness documentarily in TEI, then generate a critical view with apparatus from those transcriptions. The data supports multiple reading interfaces from one source.
What TEI elements signal each approach?
Documentary editions lean on <sourceDoc>, <surface>, and <zone> for layout fidelity. Critical editions use the textcrit module — <app>, <lem>, <rdg> — to record variation across witnesses declared in <listWit>.
When should an archivist prefer a documentary edition?
When the document itself is the object of study — a single authoritative letter, charter, or autograph manuscript — and when you want to preserve evidentiary value without imposing editorial reconstruction. Archival users usually want the witness, not a conflation.
Does choosing critical mean I ignore the physical document?
No. A good critical edition still links its reconstructed text to facsimiles via IIIF and records each witness's physical description. The reconstruction sits on top of, not instead of, documentary evidence.