Text Encoding Fundamentals: Element list

Elements for basic TEI documents

This is more of a brief reference sheet than an exhaustive list of TEI elements: it is intended to provide you with a way to look up the most commonly used elements, grouped together for the exercises in which we’ll be encountering them. For detailed information about the contents and semantics of these elements (and for other more arcane elements), have a look at the TEI Guidelines.

Element groups

structure
TEI, back, body, front, group, teiHeader, and text
general purpose block-level
ab, argument, div, head, item, label, list, p, quote, and said
general purpose phrase-level
bibl, date, distinct, emph, foreign, hi, mentioned, name, q, quote, rs, said, seg, soCalled, and term
poetry
l, lg, and rhyme
drama
castGroup, castItem, castList, role, roleDesc, sp, speaker, and stage
diary entries, letters, etc.
closer, dateline, opener, postscript, salute, signed, and trailer
alternative transcriptions
abbr, choice, corr, expan, orig, reg, and sic
manuscripts and physicality of documents
add, addSpan, cb, del, delSpan, handShift, lb, milestone, and pb
editorial annotation
app, damage, gap, lem, rdg, restore, subst, supplied, and unclear
hypertextual
anchor, note, ptr, and ref

Elements (in alphabetical order)

TEI
The outermost (or root) element for any TEI P5 conformant document. It groups together the TEI header and the document text. It must have the TEI namespace specifed, and should have an xml:lang attribute, i.e. TEI xmlns="http://www.tei-c.org/ns/1.0" xml:lang="en".
ab
An anonymous block, that is, a paragraph-like chunk that does not carry the semantic weight of a paragraph. Use type and maybe subtype to categorize.
abbr
An abbreviation; may be used alone or, when inside choice, in combination with expan which holds an expanded reading.
add
A handwritten addition. The hand attribute indicates the handwriting in which the addition is made. This attribute contains an identifier which points to a hand element in the profileDesc of the TEI header; this hand element contains an extended description of the handwriting, ink, and other details.
addSpan
An empty element which marks the starting point for a handwritten addition that either is too long to be encoded with add, or overlaps an element boundary. Its spanTo attribute points to an anchor element which marks the endpoint of the added material. The hand attribute indicates the handwriting in which the addition is made (see above for details).
anchor
An anchor point, usually used as a place for some other element (such as a note) to point to, using the anchor’s xml:id attribute.
app
Contains one entry in a critical apparatus, with an optional lemma and at least one reading.
argument
A short summary or description of the contents of the following section. Contains one or more p or lg elements.
back
Contains the back matter of the document, if any: indices, appendices, epilogues, colophons, errata lists, etc. May be subdivided into div elements if necessary.
bibl
Used to encode bibliographical references, either in a list (using listBibl) or in running prose.
body
Contains the main body of the document, not including front matter and back matter. The body element typically includes one or more div elements. It may start with a head. (Think about where the head belongs—is it the heading for the body, or the heading for the first division?)
castGroup
A grouping of related items in a cast list, containing one or more castItem elements and an optional head and trailer.
castItem
An item in a cast list, containing a role and an optional roleDesc.
castList
A cast list in a dramatic text, listing the roles in the drama. It consists of one or more castItem or castGroup elements.
cb
An empty element which marks the break between one column and the next. Equivalent to milestone unit="column".
choice
Groups together two or more alternate encodings of a phrase-level passage, using the elements listed below.
closer
Very similar to opener, but located at the end of the div instead of at the beginning.
corr
A corrected reading of a typographical error or oddity in the original; may be used alone or, when inside choice, in combination with sic, which holds the original reading.
damage
A damaged portion of the original text; the type attribute allows you to classify the damage, and the extent attribute allows you to indicate the extent of the damage.
date
Used to encode dates. The when attribute can be used to encode a regularized form of the date (e.g. <date when="2001">The first year of the new century</date> or <date when="2005-05-29">Sun, 29 May 05</date>).
dateline
Used within opener and closer to encode the date and place of writing. Contains words and phrase-level encoding.
del
A deletion. The hand attribute indicates the handwriting in which the addition is made (see above for details).
delSpan
An empty element which marks the starting point for a deletion that is either too long to be encoded with del or that overlaps an element boundary. Its spanTo attribute points to an anchor element which marks the endpoint of the deleted material. The hand attribute indicates the handwriting in which the deletion is made (see above for details).
distinct
Used for linguistically distinct words (e.g. dialect words, regionally accented words).
div
A division of a text: for instance, an act, a chapter, a section, a poem, a letter… Use the type attribute to indicate what kind of division.
emph
Used to encode emphasized words or phrases.
expan
The expanded reading of an abbreviation; typically used inside choice, in combination with abbr which holds the corresponding abbreviated reading. Rarely used alone.
foreign
Used for foreign-language words when no other element (e.g. quote) is already present.
front
Contains the front matter of the document, if any: title pages, tables of contents, introductory essays, and so forth. The front element contains an optional titlePage and may be subdivided into div elements.
gap
A gap in the original text (either from damage, deletion, excerption, or some other cause). The desc child element provides a description of what is missing, and the reason attribute provides the reason for the omission.
group
This element is used to represent documents which contain more than one independent text. It appears instead of body in the overall TEI document structure, and groups together multiple text elements, with an optional front and back.
handShift
An empty element which marks the boundary point at which a change of handwriting takes place. Its new attribute indicates the handwriting that begins at the point being marked. The new attribute functions just like the hand attribute, in pointing to a hand element in the TEI header, which provides detailed information on the handwriting in question.
head
The heading of a division: contains words and phrase-level encoding. head may appear at the start of div, but also at the start of body, front, back, list, and lg.
hi
Used to encode words or phrases which are highlighted for reasons which the encoder either does not know or chooses not to analyse.
item
An item in a list: contains an optional label followed by words and phrase-level encoding, or a series of paragraphs.
l
A single verse line: contains words and phrase-level elements. May have a met attribute to formally specify the metrical pattern.
label
The label of an item (e.g. a letter, number, or word indicating its order or other facts about it): contains words and phrase-level encoding. Note that label can also be the first element inside a paragraph.
lb
An empty element which marks a typographical line break. Equivalent to milestone unit="line".
lem
A lemma; e.g., the reading from the base text.
lg
A group of verse lines: contains one or more l elements. May have a rhyme attribute to formally specify the rhyme scheme, e.g. lg rhyme="ABAB".
list
A list: contains a series of item elements.
mentioned
Used for words which are mentioned but not used (for instance, for spelling or definition purposes).
milestone
An empty element which marks a boundary point in the text according to some standard reference system, such as signatures, scrolls, leaves. Use the unit attribute to indicate the reference system whose units are being marked at this point.
name
Used to encode all kinds of names, i.e. proper nouns and noun-phrases. If you want to distinguish between different kinds of names, you can use the type attribute (e.g. name type="person"). TEI also includes specific elements for different kinds of names (e.g. persName) for projects that need more detailed encoding. The rs element is a more generic version of name, which may be used to encode common nouns and noun phrases.
note
A note (a footnote, endnote, marginal note, or inline note). Link the note to the point where it’s anchored using xml:id and target. note contains most anything, including words and phrase-level encoding, or one or more p elements.
opener
This element may appear at the start of a div, text, front, or back, and it groups together the elements that appear at the start of a letter or similar document: the date and place of writing (using dateLine, and the salutation to the person being addressed (using salute).
orig
An unmodernized reading in the original; may be used alone or, when inside choice, in combination with reg, which holds a regularized reading.
p
A prose paragraph: contains words and phrase-level encoding.
pb
An empty element which marks the break between one page and another. By convention, information stored in the attributes of pb refer to the page that follows the break. Equivalent to milestone unit="page".
ptr
Indicates a reference to some other XML element (either in the current document or some other accessible document) by pointing to it with a URI on the target attribute. Must not have content. E.g., ptr target="#art08_sec08"/.
postscript
Used to encode a postscript, e.g. of a letter.
q
Used to encode passages surrounded by quotation marks, when you don’t want to bother with a more precise element like said. Roughly the same as hi rend="surrounded-with-quotation-marks".
quote
Used to encode quotations from other sources; contains words and phrase-level encoding.
rdg
A single reading, e.g. from a particular witness.
ref
Indicates a reference to some other XML element (either in the current document or some other accessible document) by pointing to it with a URI on the target attribute. May (and probably should) have content. E.g., <ref target="#art08_sec08">the <soCalled>IP</soCalled> clause</ref>.
reg
A modernization of a reading in the original; may be used alone or, when inside choice, in combination with orig, which holds the corresponding unmodernized reading.
restore
Indicates restoration of text to an earlier state by cancellation of a marking or instruction; in particular, useful to indicate that a deletion was restored, e.g. by the notation stet.
rhyme
May be optionally used to indicate the portion of the metrical line that rhymes, and with its label attribute which part of the rhyme scheme is in play.
role
The name of a role in a cast list.
roleDesc
The description of a role in a cast list.
rs
Used to encode all kinds of references to people, places, and things; i.e., nouns and noun phrases. If you want to distinguish between different categories of entity being referred to, you can use the type attribute (e.g. rs type="person"). The name element is a more specialized version of rs, reserved for proper nouns and noun-phrases.
said
Passages spoken aloud or thought, e.g. by a character in a novel.
seg
General-purpose prhase-level segment: use type and maybe subtype to categorize.
salute
Used within opener and closer to encode the salutation to the person being addressed (e.g. Dear Sir, or I remain faithfully yours…). Contains words and phrase-level encoding.
sic
A typographical error or oddity in the original; may be used alone or, when inside choice, in combination with corr, which holds a corrected reading.
signed
Used within closer to encode the signature or name of the person writing. Contains words and phrase-level encoding.
soCalled
Used to encode (or express) authorial distance; e.g., phrases that were or should be in scare quotes.
sp
A dramatic speech; usually begins with a speaker element, followed by a p or lg.
speaker
A speaker identification printed in the text.
stage
A stage direction. The type attribute may be used to identify the kind of stage direction; suggested values include:
  • business
  • costume
  • delivery
  • entrance
  • exit
  • location
  • narrative
  • novelistic
subst
Groups together an add and a del so that the addition is understood as being a substitution for the deletion.
supplied
Indicates that a given word or passage cannot be read in the original and is being supplied (either through editorial judgment or from some other textual source).
teiHeader
The wrapper for all of the document’s metadata. The elements that go inside the TEI header are too numerous to list usefully here; see the templates for details.
term
Used to encode specialized terminology; often associated with a gloss.
text
The wrapper element which contains all of the document’s content. The text element is most often used for a single work (i.e. a single published document, or a single aesthetic unit such as a play or a work of fiction). Terms like single work and aesthetic unit need to be defined by the individual project. A text element contains an optional front, a mandatory body, and an optional back.
trailer
This element is used for things that come at the very end of the document or section, such as The End.
unclear
Indicates that a given word or passage is unclear, but not entirely illegible (expresses uncertainty rather than absolute lack of information); multiple alternative readings may be grouped in a choice element.

Attributes (in alphabetic order)

met
May be used to specify the metrical pattern a line (or line group).
n
Provides a label or identifier for this particular element, not necessarily unique.
next and prev
Allow what is logically a single text object (e.g. a quotation) to be encoded as a series of two or more discrete XML elements, as a work-around for overlap problems. These attributes represent the connections between these fragmentary elements, by pointing to a prior or subsequent element in the chain of fragments. They do so by referring to that element’s xml:id value. That is, if next is specified on a said element, then its value should be a hash mark (#) followed by the value of the xml:id of another said element, the one that is the next part of the spoken passage. For example, <said xml:id="s01" next="#s02">Hey</said>, he said, <said xml:id="s02" prev="#s01">What’s up?</said>.
rend
May be used to specify how the element looked in the source. E.g., head rend="align(center)" or l rend="indent(2)".
rhyme
May be used to specify the rhyme scheme of a line group.
target
Provides a URI (e.g. http://bauman.zapto.org/gallery/Niagara_Falls_2008-01/2008_01_07T16_35_39 or #sect08) that points to either another document or an element within an XML document (including the current one).
xml:id
Provides a unique identifier for this particular element, thus allowing other elements to point to it (using their target, next, prev, etc.).
xml:lang
Used to indicate the language of an element’s content. Its value conforms to BCP 47 (a standard system for defining language codes). For information on how BCP 47 codes are constructed, see the note in the data.language documentation. Some sample values for the xml:lang attribute are:
English en
French fr
German de
Italian it
Latin la
Arabic as spoken in Iraq ar-IQ
Chinese zh
simplified Chinese zh-Hans
Taiwanese zh-TW
If further explanation is required, a language element with an ident attribute of the same BCP 47 code can be specified in the TEI header.

Copyleft 2008 Syd Bauman and Julia Flanders; source available at http://www.wwp.brown.edu/encoding/current/handouts/elementList.tei.