Glossary of terms used in the TEI XML critical-edition guides

From Wiki
Jump to navigation Jump to search

This glossary explains the editorial and technical vocabulary used throughout the collection Building critical editions from TEI XML with ConTeXt.

The collection brings together concepts from several fields:

  • textual criticism and scholarly editing;
  • XML;
  • the Text Encoding Initiative (TEI);
  • Lua and data processing;
  • ConTeXt and typesetting.

The same word may sometimes be used differently in different disciplines. The definitions below therefore explain how each term is used specifically within this collection.

This page is not intended to replace:

  • a general introduction to textual criticism;
  • the TEI Guidelines;
  • an XML reference manual;
  • a Lua programming manual;
  • the ConTeXt documentation.

Its purpose is narrower: to provide the vocabulary needed to understand and use the six guides.

Contents

How to use this glossary

The glossary can be used in three ways.

By subject

Readers who are new to a particular part of the workflow can begin with the thematic lists below.

Alphabetically

The main body of the glossary is organised alphabetically.

From encoding to output

Many entries describe different stages of the same process: 

editorial observation
        ↓
structured editorial concept
        ↓
TEI element or attribute
        ↓
Lua or ConTeXt processing
        ↓
printed or digital output

For example: 

a manuscript contains the reading “soul”
        ↓
witness B supports a variant reading
        ↓
<rdg wit="#B">soul</rdg>
        ↓
the witness reference is resolved
        ↓
soul B

The vocabulary by subject

Textual criticism and scholarly editing

XML

TEI

Lua and data processing

ConTeXt and typesetting

A quick correspondence table

Editorial concept TEI representation Processing task Possible printed form
Apparatus entry <app> Collect the lemma and its readings One entry in the critical apparatus
Lemma <lem> Identify the edited reading mind]
Variant reading <rdg> Collect an alternative textual form soul
Witness support @wit Resolve references to witness declarations B
Witness declaration <witness> with xml:id Retrieve metadata and display information A siglum such as A or B
Group of related readings <rdgGrp> Preserve or transform the grouping Several readings presented together

The four principal layers

Layer Main question Typical vocabulary
Editorial layer What textual evidence exists, and how should it be interpreted? Witness, lemma, reading, conjecture, omission
Encoding layer How should that information be represented explicitly? Element, attribute, <app>, @wit,

xml:id

Processing layer How should the encoded information be selected, checked, or transformed? Lua table, filtering, sorting, reference resolution
Typesetting layer How should the result be presented to the reader? Apparatus paragraph, lineation, layout, rendering

The same object can therefore be described at several levels: 

editorial level:     witness B supports the reading “soul”
encoding level:      <rdg wit="#B">soul</rdg>
processing level:    resolve #B and retrieve its siglum
typesetting level:   soul B

Alphabetical glossary

A

Apparatus

An apparatus is an organised body of information accompanying an edited text.

In this collection, the term usually refers to a textual apparatus recording differences between witnesses. A scholarly edition may also contain other apparatuses, such as:

  • an apparatus of sources;
  • an apparatus of parallels;
  • explanatory notes;
  • translation notes;
  • linguistic annotations.

The word does not by itself specify where the information appears on the page.

See also:

Apparatus entry

An apparatus entry is one structured record of textual variation within a critical apparatus.

It normally relates:

  • a passage in the edited text;
  • a lemma;
  • one or more alternative readings;
  • the witnesses supporting those readings;
  • sometimes an editorial explanation.

In TEI, an apparatus entry is commonly represented by <app>: 

<app>
  <lem wit="#A">mind</lem>
  <rdg wit="#B">soul</rdg>
</app>

A possible printed form is: 

mind] soul B

See also:

Apparatus layer

An apparatus layer is one distinct class of apparatus information presented alongside the edited text.

An edition might contain separate layers for:

  • textual variants;
  • editorial interventions;
  • sources and parallels;
  • translation notes;
  • linguistic commentary.

In ConTeXt, separate note series or apparatus mechanisms can be used to give each layer its own formatting and placement.

See also:

Apparatus paragraph

An apparatus paragraph is a paragraph containing one or more compact apparatus entries.

Instead of setting every entry as an independent footnote, a typesetting system may collect several entries into a continuous paragraph: 

12 mind] soul B   15 seeks] desires C   18 unity] harmony D

Its appearance is a matter of typesetting. The TEI data need not be stored as a preformatted paragraph.

Attribute

An attribute adds information to an XML element.

It is written inside the element’s start tag: 

<rdg wit="#B" type="orthographic">soule</rdg>

Here:

  • wit identifies the supporting witness;
  • type classifies the reading.

An attribute has a name and a value: 

name:   wit
value:  #B

See also:

C

Child element

A child element is an element directly contained inside another element.

In this example: 

<app>
  <lem>mind</lem>
  <rdg>soul</rdg>
</app>

both <lem> and <rdg> are children of <app>.

The relationship can be represented as: 

app
├── lem
└── rdg

See also:

Conjecture

A conjecture is a reading proposed by an editor rather than directly attested by the known witnesses.

A conjecture may be introduced when the transmitted readings appear corrupt, incomplete, or otherwise unsatisfactory.

The encoding should distinguish a conjecture from an attested witness reading. The exact TEI representation depends on the project’s encoding policy.

See also:

ConTeXt

ConTeXt is a programmable typesetting system built on TeX and LuaMetaTeX.

In this collection, ConTeXt is used to:

  • load and inspect XML documents;
  • associate processing instructions with XML elements;
  • format the edited text;
  • generate apparatus entries;
  • control lineation and references;
  • place apparatus material on the page;
  • produce PDF output.

ConTeXt is responsible primarily for presentation, but it can also participate in selecting and transforming XML data.

See also:

Correction

A correction is a change made to an earlier textual form.

A manuscript may contain:

  • an original reading;
  • a deletion;
  • an addition;
  • a corrected reading;
  • several successive states.

A critical edition may need to distinguish: 

original hand
        ↓
correction
        ↓
later correction

Corrections are therefore not always equivalent to simple variant readings.

See also:

Critical apparatus

A critical apparatus records textual variation and editorial decisions in relation to an edited text.

A typical entry may identify:

  • a location in the text;
  • a lemma;
  • one or more variant readings;
  • the witnesses supporting each reading;
  • editorial additions or conjectures.

A critical apparatus may be printed at the bottom of the page, but it is not merely a footnote system. Its entries represent structured relationships between textual forms.

See also:

Critical edition

A critical edition presents a text established through the comparison and interpretation of textual evidence.

It normally includes:

  • an edited text;
  • an account of the witnesses or sources;
  • editorial principles;
  • a critical apparatus;
  • sometimes commentary, translation, or additional apparatus layers.

A critical edition is both: 

a scholarly argument about the text
                  +
a designed publication for readers

D

Data record

A data record is a structured collection of related values.

For example, an apparatus entry may be represented internally as a Lua table: 

{
  lemma    = "mind",
  readings = {
    { text = "soul", witness = "B" }
  }
}

The record keeps the editorial components separate from their printed presentation.

See also:

Diplomatic transcription

A diplomatic transcription attempts to reproduce the features of a particular source as closely as the editorial method permits.

It may preserve:

  • original spelling;
  • punctuation;
  • abbreviations;
  • deletions and additions;
  • line and page divisions;
  • corrections;
  • scribal features.

The same TEI data may contribute both to a diplomatic transcription and to a normalised reading text, depending on the encoding and processing rules.

See also:

Document tree

A document tree is the hierarchical structure formed by the elements of an XML document.

For example: 

TEI
├── teiHeader
└── text
    └── body
        └── p
            └── app
                ├── lem
                └── rdg

The tree representation makes parent, child, and sibling relationships visible.

See also:

E

Edited text

The edited text is the text presented by the editor as the principal reading text of the edition.

It may reproduce one witness closely or may combine evidence from several witnesses.

The edited text should not be confused with:

  • the complete textual tradition;
  • one individual witness;
  • the critical apparatus;
  • the TEI source file.

See also:

Element

An element is one of the basic structural units of an XML document.

An element normally consists of:

  • a start tag;
  • content;
  • an end tag.

For example: 

<rdg>soul</rdg>

Here:

  • <rdg> is the start tag;
  • soul is the textual content;
  • </rdg> is the end tag.

Elements may also contain other elements: 

<app>
  <lem>mind</lem>
  <rdg>soul</rdg>
</app>

See also:

Encoding

Encoding is the act of representing textual or editorial information through an explicit system of markup.

In this collection, encoding means more than inserting visual formatting commands. It means identifying the scholarly structure: 

printed notation:      mind] soul B

encoded structure:     lemma = mind
                       reading = soul
                       witness = B

TEI XML provides a shared vocabulary for this representation.

See also:

F

Filtering

Filtering means selecting records that satisfy particular conditions and excluding others.

For example, a project may filter apparatus data in order to:

  • include only substantive readings;
  • exclude orthographic variants;
  • show only selected witnesses;
  • produce a shorter reading edition;
  • generate a report of omissions.

Filtering may be performed with ConTeXt XML selection mechanisms, Lua, or a combination of both.

See also:

Function

A function is a named or anonymous block of Lua code that performs a task.

For example, a function might:

  • remove the initial # from a witness reference;
  • look up a witness declaration;
  • sort sigla;
  • format an apparatus record;
  • report missing identifiers.

A schematic Lua function may look like: 

local function normalise_reference(reference)
    return reference:gsub("^#", "")
end

The guides introduce functions only when they solve a concrete processing problem.

H

Hierarchy

A hierarchy is an organisation in which structures are nested inside other structures.

XML is hierarchical: 

document
└── text
    └── paragraph
        └── apparatus entry
            ├── lemma
            └── reading

This differs from a flat list of unrelated character strings.

Hierarchy is one of the main reasons XML is suitable for representing complex documents.

See also:

I

Identifier

An identifier is a value used to distinguish one object from others.

In TEI XML, a witness may receive an identifier through xml:id: 

<witness xml:id="B">Witness B</witness>

Another element can then refer to that witness: 

<rdg wit="#B">soul</rdg>

The identifier is intended for machine-readable linking. It is not necessarily identical to the siglum printed in the apparatus.

See also:

Inline apparatus

An inline apparatus is an apparatus whose TEI apparatus entries occur directly at the relevant locations in the encoded text.

For example: 

<p>
  The
  <app>
    <lem wit="#A">mind</lem>
    <rdg wit="#B">soul</rdg>
  </app>
  seeks unity.
</p>

The apparatus entry replaces or contains the relevant textual passage within the document flow.

Advantages may include:

  • a direct connection between text and variation;
  • simple local processing;
  • easy access to the lemma and readings together.

The suitability of this method depends on the structure of the edition.

See also:

L

Layout

Layout is the spatial organisation of material on the page.

It includes:

  • page dimensions;
  • margins;
  • text width;
  • columns;
  • headers and footers;
  • note areas;
  • apparatus placement;
  • spacing between textual layers.

Layout belongs primarily to the typesetting layer rather than to the TEI encoding layer.

Lemma

A lemma is the passage in the edited text to which an apparatus entry refers.

In a printed apparatus: 

mind] soul B

mind is the lemma.

In TEI, it may be represented by <lem>: 

<lem wit="#A">mind</lem>

The lemma may be:

  • a word;
  • part of a word;
  • a phrase;
  • a line;
  • a longer passage.

See also:

Lemma separator

A lemma separator is the sign that separates the lemma from the readings in a printed apparatus.

A common example is the closing square bracket: 

mind] soul B

The separator is a typographical convention. It is not part of the lemma itself and does not need to be stored as editorial data in TEI.

Other traditions may use different punctuation or formatting.

Lineation

Lineation is the numbering or identification of lines in an edited text.

Line numbers may be used to:

  • identify the location of apparatus entries;
  • support scholarly citation;
  • connect text and commentary;
  • compare parallel versions.

Lineation may follow:

  • the lines of a source;
  • the lines of the printed edition;
  • poetic verses;
  • project-specific units.

ConTeXt is responsible for rendering line numbers, while TEI may record or support the textual divisions to which they refer.

Location-referenced apparatus

A location-referenced apparatus stores apparatus entries separately from the main text and identifies their location through references.

The entry may refer to:

  • a line number;
  • a paragraph;
  • an identifier;
  • a passage boundary;
  • another encoded location.

This approach separates the textual flow from the apparatus records, but requires reliable reference resolution.

See also:

Lua

Lua is a lightweight programming language integrated into LuaMetaTeX.

In this collection, Lua may be used to:

  • inspect XML nodes;
  • retrieve attributes;
  • build data records;
  • resolve witness references;
  • sort readings and sigla;
  • validate project-specific rules;
  • transform TEI structures;
  • prepare information for ConTeXt.

Lua is a processing tool. It does not replace TEI encoding or ConTeXt typesetting.

See also:

Lua table

A Lua table is Lua’s principal data structure.

Tables can represent:

  • ordered lists;
  • associative records;
  • nested structures;
  • witness declarations;
  • apparatus entries.

For example: 

local witness = {
    id          = "B",
    siglum      = "B",
    type        = "manuscript",
    description = "A later manuscript",
}

A table can preserve the distinction between the different components of a record.

LuaMetaTeX

LuaMetaTeX is the engine used by current ConTeXt distributions.

It combines TeX-based typesetting with an embedded Lua environment.

This integration allows ConTeXt documents to move between: 

XML structures
      ↓
Lua data
      ↓
TeX and ConTeXt typesetting

M

Markup

Markup consists of signs added to content in order to identify its structure or function.

In XML: 

<rdg wit="#B">soul</rdg>

the markup includes:

  • <rdg>;
  • wit="#B";
  • </rdg>.

The word soul is the textual content.

Markup is not necessarily visible in the final edition.

Namespace

An XML namespace identifies the vocabulary to which element names belong.

A TEI document commonly declares the TEI namespace on its root element: 

<TEI xmlns="http://www.tei-c.org/ns/1.0">
  ...
</TEI>

The namespace distinguishes TEI elements from elements belonging to other XML vocabularies.

Namespaces affect how XML processors, including ConTeXt, identify and select elements.

Negative apparatus

A negative apparatus normally omits explicit support for the lemma and records only the witnesses containing alternative readings.

For example: 

mind] soul B; understanding C

The apparatus implies that the remaining relevant witnesses support mind, according to the conventions of the edition.

A negative apparatus is compact, but its interpretation depends on knowledge of the witness base and editorial conventions.

See also:

Node

A node is one component of an XML tree.

Depending on the processing model, nodes may include:

  • elements;
  • text;
  • attributes;
  • comments;
  • processing instructions.

In the guides, the word often refers more narrowly to an XML element being inspected or processed.

Note series

A note series is a separately defined class of notes in ConTeXt.

Different series may be used for:

  • ordinary footnotes;
  • textual variants;
  • source notes;
  • translation notes;
  • commentary.

Each series may have its own numbering, formatting, and placement.

A critical apparatus can sometimes use note mechanisms, but its underlying data model is more structured than that of an ordinary note.

O

Omission

An omission occurs when a witness lacks material present in another witness or in the edited text.

For example: 

A  The mind seeks unity.
B  The mind seeks.

Witness B omits unity.

An omission should be represented as a textual relation, not merely as a blank character string. The precise TEI encoding depends on the project’s apparatus model.

See also:

Output

Output is the result produced from the encoded and processed source.

Possible outputs include:

  • a printed critical edition;
  • a PDF;
  • an HTML edition;
  • a reading text;
  • a diplomatic transcription;
  • a witness report;
  • a list of variants;
  • a validation report.

The same TEI source may support several outputs. 

one TEI source
├── reading edition
├── critical PDF
├── HTML edition
└── editorial report

P

Parallel text

A parallel text presents two or more textual streams in coordinated positions.

Examples include:

  • original text and translation;
  • two versions of a work;
  • two witnesses;
  • text and commentary;
  • diplomatic and normalised transcriptions.

ConTeXt can control the visual alignment, while TEI records the structures and relationships being presented.

Parent element

A parent element directly contains one or more child elements.

In this example: 

<app>
  <lem>mind</lem>
  <rdg>soul</rdg>
</app>

<app> is the parent of both <lem> and <rdg>.

See also:

Parser

An XML parser reads an XML document and interprets its markup structure.

A parser can determine:

  • whether the document is well formed;
  • where elements begin and end;
  • which attributes belong to which elements;
  • how the document tree is organised.

A parser does not by itself understand the scholarly meaning of a lemma or a witness. That meaning comes from the TEI vocabulary and the project’s encoding rules.

Positive apparatus

A positive apparatus explicitly records the witnesses supporting the lemma as well as those supporting the variant readings.

For example: 

mind A D] soul B; understanding C

Here, witnesses A and D are explicitly associated with the lemma.

A positive apparatus contains more explicit information than a negative apparatus but usually occupies more space.

See also:

R

Reading

A reading is a textual form attested, reconstructed, or proposed at a particular location.

A reading may be:

  • the form adopted in the edited text;
  • an alternative form found in a witness;
  • an omission;
  • an addition;
  • a conjecture;
  • a corrected state.

In TEI, a variant reading is commonly represented with <rdg>.

See also:

Reading text

A reading text is a version of the edited text intended primarily for continuous reading.

It may:

  • omit the critical apparatus;
  • simplify editorial signs;
  • normalise some textual features;
  • include only selected annotations.

A reading text can be generated from the same structured source as a fuller critical edition.

Reference

A reference connects one part of an XML document with another object or location.

For example: 

<rdg wit="#B">soul</rdg>

The value #B refers to the element whose identifier is B: 

<witness xml:id="B">...</witness>

The initial # signals a reference to an identifier within the same document.

See also:

Reference resolution

Reference resolution is the process of following a reference and retrieving the object it identifies.

For example: 

@wit = "#B"
        ↓
remove or interpret the initial #
        ↓
find xml:id="B"
        ↓
retrieve witness metadata
        ↓
print the siglum B

Reference resolution may be performed with XML selection mechanisms, Lua, or both.

Rendering

Rendering is the process of converting structured information into a particular visible or audible form.

In this collection, rendering normally means producing typographical output.

The same encoded reading may be rendered as: 

soul B

or: 

Witness B reads “soul”.

or omitted entirely from a reading edition.

Rendering belongs primarily to ConTeXt and the processing rules, not to the TEI data alone.

Root element

The root element is the single outermost element of an XML document.

In a TEI document, the root element is normally <TEI>: 

<TEI xmlns="http://www.tei-c.org/ns/1.0">
  <teiHeader>...</teiHeader>
  <text>...</text>
</TEI>

Every other element in the document is contained directly or indirectly inside the root element.

S

Schema

A schema describes the structures permitted in a class of XML documents.

A schema may define:

  • which elements are allowed;
  • which elements may contain other elements;
  • which attributes are available;
  • which values are permitted;
  • which structures are required.

A TEI schema can therefore be used to check whether a document follows a particular TEI model.

A valid document may still contain scholarly mistakes. Schema validation does not replace editorial review.

Selection

Selection is the process of locating particular XML elements or data records.

Examples include selecting:

  • all <app> elements;
  • all readings supported by witness B;
  • all omissions;
  • all entries of a particular type;
  • all witness declarations.

Selection identifies the relevant material. Filtering may then include or exclude records according to additional conditions.

Siglum

A siglum is the short label used to identify a witness in scholarly writing.

Typical sigla include: 

A
B
P46
M
α
Vat.

A siglum is intended for human readers.

It should be distinguished from an XML identifier:

Function Example
XML identifier xml:id="ms-paris-123"
Printed siglum P

A project may use the same value for both, but the two functions remain conceptually distinct.

See also:

Sorting

Sorting places records in a defined order.

Witnesses may be sorted:

  • alphabetically;
  • by siglum;
  • by date;
  • by manuscript family;
  • according to a project-specific editorial order.

Readings may also be sorted by type, support, or another rule.

The required order should not be inferred accidentally from the order in which elements happen to occur in the XML source.

Source

A source is material from which textual or editorial information is derived.

Sources may include:

  • manuscripts;
  • printed editions;
  • inscriptions;
  • papyri;
  • typescripts;
  • quotations;
  • translations;
  • earlier scholarly editions.

In textual criticism, a source is not always identical with a witness. The exact distinction depends on the editorial model.

Stand-off apparatus

A stand-off apparatus stores apparatus information separately from the main textual sequence.

The apparatus entries refer back to passages or locations in the text through identifiers or pointers.

A simplified model is: 

text
├── passage xml:id="p1"
└── passage xml:id="p2"

apparatus
├── entry referring to #p1
└── entry referring to #p2

This can keep the reading text less interrupted, but requires careful reference management.

See also:

Structured data

Structured data are data whose components and relationships are identified explicitly.

A printed apparatus string: 

mind] soul B

can be analysed as: 

entry
├── lemma: mind
└── reading
    ├── text: soul
    └── witness: B

Structured data allow those components to be selected, checked, transformed, and rendered independently.

See also:

T

TEI

TEI stands for the Text Encoding Initiative.

The TEI provides guidelines and an XML vocabulary for representing texts and textual scholarship.

TEI includes mechanisms for encoding:

  • document structure;
  • manuscripts and sources;
  • names and places;
  • editorial interventions;
  • quotations and references;
  • textual variation;
  • critical apparatuses.

TEI does not determine the typography of the final edition. It records structures and relationships that can be rendered in different ways.

See also:

TEI document

A TEI document is an XML document using the TEI vocabulary and namespace.

A minimal structural model is: 

TEI
├── teiHeader
└── text
    └── body

A TEI document for a critical edition may also contain:

  • witness declarations;
  • apparatus entries;
  • lemmas;
  • readings;
  • editorial metadata;
  • references between the text and the apparatus.

See also:

TEI Guidelines

The TEI Guidelines are the documentation defining and explaining the TEI encoding model.

They describe:

  • TEI elements and attributes;
  • recommended structures;
  • alternative encoding methods;
  • manuscript description;
  • critical apparatuses;
  • editorial interventions;
  • many other forms of textual representation.

The six guides use only a limited part of the TEI Guidelines.

A project should document any choices it makes between alternative TEI methods.

TEI header

The TEI header contains metadata about the TEI document and the work it represents.

It is encoded with <teiHeader>.

It may contain information about:

  • the title and responsibility;
  • publication;
  • the source;
  • encoding practices;
  • revisions;
  • witnesses and manuscripts;
  • project-specific editorial policies.

The TEI header is not merely a decorative title page. It documents the identity, provenance, and editorial framework of the encoded text.

See also:

Textual content

Textual content is the character data contained within an XML element.

In: 

<rdg wit="#B">soul</rdg>

the textual content is: 

soul

The tags and attributes are markup, not textual content.

An element may contain both textual content and child elements.

Textual variant

A textual variant is a difference between two or more textual forms.

Variants may involve:

  • spelling;
  • individual words;
  • word order;
  • omission;
  • addition;
  • correction;
  • punctuation;
  • longer passages.

Not every difference must necessarily be printed in the final apparatus. The edition’s policy determines which variants are recorded, selected, or suppressed.

Transformation

A transformation converts structured data from one form into another.

Examples include:

  • converting TEI apparatus entries into Lua tables;
  • replacing XML identifiers with display sigla;
  • producing a negative apparatus from positive data;
  • generating prose descriptions of variants;
  • producing ConTeXt commands from TEI records.

A transformation should preserve the relevant scholarly meaning even when the visible form changes.

Transposition

A transposition occurs when the same textual material appears in a different order.

For example: 

A  reason and freedom
B  freedom and reason

A transposition is more complex than a simple substitution because the same material may remain present but be reordered.

Its encoding requires a project-specific method capable of preserving the relationship between the reordered passages.

Typesetting

Typesetting is the process of composing text and related material into a designed visual form.

It includes:

  • font selection;
  • spacing;
  • line breaking;
  • page breaking;
  • headings;
  • notes;
  • apparatus paragraphs;
  • line numbers;
  • page layout.

In this collection, ConTeXt performs the typesetting of information encoded in TEI and prepared through XML setups or Lua.

V

Validation

Validation checks whether an XML document conforms to a defined schema or set of rules.

Schema validation may detect:

  • missing required elements;
  • forbidden structures;
  • invalid attribute values;
  • elements used in the wrong context.

Project-specific validation may also detect:

  • references to undeclared witnesses;
  • duplicate sigla;
  • unsupported reading types;
  • missing lemmas;
  • inconsistent witness groups.

Validation improves consistency, but it does not prove that the editorial interpretation is correct.

Validation rule

A validation rule is a condition that data must satisfy.

Examples include: 

Every @wit reference must resolve to a declared witness.

Every apparatus entry must contain at least one lemma or reading.

Every printed siglum must be defined.

No two witnesses may share the same xml:id.

Some rules can be expressed in an XML schema. Others are easier to check with Lua or project-specific tools.

W

Well-formed XML

A document is well formed when it follows the fundamental syntax rules of XML.

Among other requirements:

  • there must be exactly one root element;
  • every start tag must have a matching end tag;
  • elements must be properly nested;
  • attribute values must be quoted;
  • reserved characters must be escaped where required.

This is well formed: 

<app>
  <lem>mind</lem>
  <rdg>soul</rdg>
</app>

This is not well formed: 

<app>
  <lem>mind
  <rdg>soul</app>

A well-formed XML document is not necessarily valid TEI.

Witness

A witness is a source that preserves or transmits a form of the text.

Depending on the edition, witnesses may include:

  • manuscripts;
  • printed editions;
  • papyri;
  • inscriptions;
  • typescripts;
  • quotations;
  • translations;
  • transcriptions.

A witness is usually identified by a siglum and declared in the TEI document.

For example: 

<witness xml:id="B">
  <abbr>B</abbr>
  <desc>A fifteenth-century manuscript.</desc>
</witness>

See also:

Witness agreement

Witness agreement occurs when two or more witnesses support the same reading.

For example: 

A D  mind
B C  soul

In TEI, several witnesses may be cited in one @wit value: 

<rdg wit="#B #C">soul</rdg>

Witness agreement may be significant for textual interpretation, but the TEI encoding records the relation without deciding its scholarly importance.

Witness family

A witness family is a group of witnesses thought to share textual or historical characteristics.

A family may represent:

  • common ancestry;
  • a recension;
  • a regional tradition;
  • a group of closely related manuscripts;
  • a project-specific classification.

Grouping witnesses can assist analysis and output, but the group should not replace the declaration of individual witnesses when individual support matters.

Witness list

A witness list is the part of a TEI document in which witnesses are declared and described.

It is commonly represented with <listWit> containing one or more <witness> elements: 

<listWit>
  <witness xml:id="A">
    <abbr>A</abbr>
    <desc>The principal manuscript.</desc>
  </witness>
  <witness xml:id="B">
    <abbr>B</abbr>
    <desc>A later manuscript.</desc>
  </witness>
</listWit>

Readings elsewhere in the document can refer to these declarations.

See also:

X

XML

XML stands for Extensible Markup Language.

XML provides a syntax for representing structured information through elements, attributes, textual content, and hierarchy.

XML does not itself define the scholarly meaning of:

  • a witness;
  • a lemma;
  • a reading;
  • a critical apparatus.

TEI supplies a vocabulary for expressing those concepts in XML.

The relationship can be summarised as: 

XML
└── general syntax for structured documents

TEI
└── scholarly vocabulary expressed in XML

XML declaration

The XML declaration may appear at the beginning of an XML document: 

<?xml version="1.0" encoding="UTF-8"?>

It identifies:

  • the XML version;
  • optionally, the character encoding;
  • optionally, whether the document depends on external declarations.

The declaration is not the root element and is not part of the TEI vocabulary.

XML setup

An XML setup is a ConTeXt mechanism for associating processing and typesetting instructions with XML elements.

A setup may determine what happens when ConTeXt encounters:

  • a paragraph;
  • a lemma;
  • a reading;
  • a witness declaration;
  • an apparatus entry.

Conceptually: 

TEI element
      ↓
ConTeXt XML setup
      ↓
typeset output

XML setups can call ConTeXt commands, retrieve attributes, select child elements, or invoke Lua functions.

XML tree

An XML tree is the hierarchical representation of an XML document.

The term is largely synonymous with document tree in these guides.

For example: 

TEI
├── teiHeader
└── text
    └── body
        └── p
            └── app
                ├── lem
                └── rdg

Lua and ConTeXt can inspect and process nodes within this tree.

xml:id

xml:id is the standard XML attribute used to assign a unique identifier to an element.

For example: 

<witness xml:id="A">
  <abbr>A</abbr>
</witness>

The identifier A allows other parts of the document to refer to this witness: 

<rdg wit="#A">mind</rdg>

The value stored in xml:id is not necessarily the same as the siglum printed in the critical apparatus.

See also:

TEI elements and attributes

This section gathers the principal TEI names used repeatedly in the guides.

<app>

The TEI element <app> represents an apparatus entry.

It commonly contains:

  • one <lem>;
  • one or more <rdg> elements;
  • sometimes groups, notes, or additional editorial structures.

Example: 

<app>
  <lem wit="#A">mind</lem>
  <rdg wit="#B">soul</rdg>
  <rdg wit="#C">understanding</rdg>
</app>

Conceptually: 

app
├── lemma
├── reading
└── reading

See also:

<lem>

The TEI element <lem> represents the lemma associated with an apparatus entry.

Example: 

<lem wit="#A">mind</lem>

It may identify:

  • the reading adopted in the edited text;
  • the witnesses supporting that reading;
  • additional editorial information.

The printed form need not reproduce the element directly. ConTeXt determines how the lemma appears.

<listWit>

The TEI element <listWit> contains a list of witness declarations.

Example: 

<listWit>
  <witness xml:id="A">...</witness>
  <witness xml:id="B">...</witness>
</listWit>

It provides a central location from which apparatus references can retrieve witness information.

<rdg>

The TEI element <rdg> represents a reading.

Example: 

<rdg wit="#B">soul</rdg>

It may record:

  • the textual form;
  • the supporting witnesses;
  • the type of variation;
  • responsibility or certainty;
  • other project-specific information.

See also:

<rdgGrp>

The TEI element <rdgGrp> groups readings that belong together for some editorial reason.

For example: 

<rdgGrp type="orthographic">
  <rdg wit="#B">soule</rdg>
  <rdg wit="#C">soul</rdg>
</rdgGrp>

The group may express:

  • related forms;
  • a shared classification;
  • a hierarchy of readings;
  • a project-specific analytical relation.

The exact meaning of the grouping should be documented by the edition.

<TEI>

The TEI element <TEI> is the root element of an individual TEI document.

Example: 

<TEI xmlns="http://www.tei-c.org/ns/1.0">
  <teiHeader>
    ...
  </teiHeader>
  <text>
    ...
  </text>
</TEI>

It normally contains:

  • one TEI header;
  • one encoded text.

<teiHeader>

The TEI element <teiHeader> contains the metadata and documentation associated with the encoded text.

It may record:

  • title and responsibility;
  • publication details;
  • source description;
  • encoding principles;
  • witness declarations;
  • revision history.

See also:

<witness>

The TEI element <witness> declares or describes a textual witness.

Example: 

<witness xml:id="A">
  <abbr>A</abbr>
  <desc>The principal manuscript.</desc>
</witness>

The element usually has an xml:id so that readings can refer to it.

@wit

The attribute @wit identifies the witness or witnesses supporting a lemma or reading.

Example with one witness: 

<rdg wit="#B">soul</rdg>

Example with several witnesses: 

<rdg wit="#B #C #D">soul</rdg>

The values are references, not necessarily printed sigla.

The processing workflow is: 

@wit="#B #C"
      ↓
split the references
      ↓
resolve B and C
      ↓
retrieve their display sigla
      ↓
print B C

Common distinctions

A note is not an apparatus entry

Ordinary note Critical apparatus entry
Usually contains a relatively independent comment Records a relationship between textual forms
Often attached to one point in the text May concern a word, phrase, line, or passage
May be stored as formatted prose Benefits from structured representation
Usually does not cite witness support systematically Commonly associates readings with witnesses

A siglum is not an XML identifier

Siglum XML identifier
Intended primarily for human readers Intended primarily for machine-readable linking
May contain typographical conventions Must follow XML identifier rules
May change according to the publication Should remain stable within the encoded document
Example: Par. Example: ms-paris-123

XML is not TEI

XML TEI
A general markup syntax A scholarly vocabulary and encoding model
Defines elements, attributes, and hierarchy Defines elements such as <app>,

<lem>, and <rdg>

Does not define textual criticism Provides recommendations for textual scholarship
Can express any suitable vocabulary Is one particular XML vocabulary

TEI is not the printed edition

TEI encoding Printed rendering
Records semantic and editorial structures Presents selected information to readers
May preserve more information than is printed May suppress or abbreviate information
Uses elements, attributes, identifiers, and references Uses typography, punctuation, spacing, and layout
Remains independent of one output design Is produced for a particular publication

Validation is not editorial proof

A document may be valid XML and valid TEI while still containing:

  • an incorrect witness attribution;
  • a mistaken lemma;
  • an incomplete collation;
  • an unjustified conjecture;
  • an inconsistent editorial decision.

Validation answers: 

Does the document follow the declared structural rules?

Editorial review answers: 

Does the document represent the textual evidence correctly?

Both are necessary, but they are not interchangeable.

The collection workflow in one view

TEXTUAL EVIDENCE
manuscripts, editions, inscriptions, other witnesses
                         ↓
EDITORIAL ANALYSIS
collation, lemmas, readings, omissions, conjectures
                         ↓
TEI XML ENCODING
elements, attributes, identifiers, references
                         ↓
VALIDATION
schema rules and project-specific consistency checks
                         ↓
PROCESSING
ConTeXt XML setups and Lua functions
                         ↓
SELECTION AND TRANSFORMATION
sorting, grouping, filtering, reference resolution
                         ↓
TYPESETTING
edited text, apparatus paragraphs, lineation, layout
                         ↓
OUTPUT
critical PDF, reading text, HTML edition, reports

Guides in which the terms are developed

Guide Main vocabulary
Understanding TEI documents for critical editions XML, element, attribute, hierarchy, root element, TEI document, TEI header
Declaring witnesses in TEI critical editions witness, siglum, identifier, xml:id,

<listWit>, <witness>

Encoding a basic critical apparatus in TEI apparatus entry, lemma, reading, <app>,

<lem>, <rdg>, @wit

Encoding complex textual variation in TEI omission, correction, conjecture, transposition, reading group,

positive and negative apparatus

Processing TEI critical apparatus data with Lua XML tree, node, selection, filtering, sorting, Lua table,

reference resolution, validation rule

Typesetting TEI critical editions with ConTeXt XML setup, rendering, lineation, apparatus paragraph, apparatus layer,

layout, output

Related pages

Retrieved from "https://wiki.contextgarden.net/index.php?title=Glossary_of_terms_used_in_the_TEI_XML_critical-edition_guides&oldid=42986"