CDL Encoding Guidelines for the TEI Header
DRAFT
Jared
Campbell
Brian
Tingle
Version 1.3
2006
University of California
The encoding guidelines provided here are unedited working drafts
produced by CDL's Structured Text Working Group. They should not be treated as final
documents. Updated guidelines will be available in November 2004.
Introduction
The following is a general guideline for creating the <teiHeader> for
documents to be submitted to CDL. Each document header should contain
<fileDesc>, <encodingDesc>, and a <profileDesc>.
The following codes are used to indicate whether an element is required or not:
R The element is "REQUIRED"
M The element is "MANDATORY WHEN THE INFORMATION IS AVAILABLE"
O The element is "OPTIONAL"
Encoding Guidelines
R <teiHeader>
Use the <teiHeader> type=" " attribute to encode the CDL TEI text type
guidelines that are to be used. Type values include CDL-TEI:MS for manuscripts, CDL-TEI:OH
oralhist for oral history transcriptions, and CDL-TEI:BK for other texts.
]]>
R <fileDesc>
The <teiHeader> should provide a relatively complete bibliographic record
for both the electronic version and the original source document. The data that describes
the electronic text should be encoded in the <fileDesc> section, while data
that pertains to the source text should be encoded within the <sourceDesc> (a
subsection of <fileDesc>) section. It should be noted that data
(such as <author>, <editor>, etc.) will often be identical in both
sections. Exceptions include:
The <fileDesc><titleStmt><title> will
include a mandatory "electronic version" subtitle.
The <fileDesc><respStmt> will include the name of
document encoders and individuals responsible for proofing or making general corrections
to the encoded file.
The <extent> in <fileDesc> will encode (in Kb. or Mb.)
the size of the electronic file. The <extent> in the
<sourceDesc> will record the number of pages, illustrations (if present),
and dimensions (see the MARC 260 field).
The <publicationStmt> in <fileDesc> should be used to
encode the institution responsible for the creation the electronic text. The
<publicationStmt> in <sourceDesc> should record the
publication information for the original source text.
<date> in <fileDesc> refers to the date which the
electronic version was encoded, while the <date> in
<sourceDesc> refers to publication or creation date of the original source
text.
<idno> in <fileDesc> should refer to numbers associated
with the encoded text, while those in <sourceDesc> should refer to the
source text (i.e. call no., manuscript no., etc.)
R <titleStmt>
The first element within a <titleStmt> must be <title>.
Enter here the main title, or a supplied title in square brackets [ ], followed by the
subtitle(s). Separate all subtitles by a space, colon, space (_:_). Titles should
end with the subtitle "electronic version".
A collection of poems : chiefly manuscript, and from living authors :
electronic version
[Letter to Chol Soo Lee] : electronic version]]>
M <author>
Record here the name, in the catalog form, of the author(s), personal or corporate, of
the work that has been encoded. Ideally, names entered here should be selected from a
national name authority such as the LC Name Authority File. If a personal name
authority record cannot be found, it should be entered as prescribed in ACCR2 2nd Ed.
Chapter 22 "Headings for Persons (usually Lastname, Firstname form, with the Middle (Initial or Name) and birth/death dates (YYYY-YYYY)
added when required for further clarification). If a corporate name heading cannot be found in the authority files, consult ACCR2
2nd ed. Chapter 24 (Headings for Corporate Bodies).
Texts that emanate from multiple authors should include multiple <author> elements. This rule
does not apply to anthologies or collections of works formed around a paticular literary
theme or form.
[Letter to Chol Soo Lee]
Lee, K. W. (Kyung Won)]]>
** Optional Use of the <name> element **
Encoders who wish to record names in both catalog form and the direct form (Firstname
Lastname) can use the <name> element within <author>
element. The catalog form of the name should be encoded using the <name>'s
reg=" " attribute. The direct form will be encoded within the <name> element.
[Letter to Chol Soo Lee]
K. W. Lee
]]>
M <editor>
Record here the name of the editor(s) of the work. Consult national name authority
files or, if no authority record exists, ACCR2 2nd ed. to construct a heading.
** Optional Use of the <name> element **
Encoders who wish to record names in both catalog form and the direct form (Firstname
Lastname) can use the <name> element within <editor>
element. The catalog form of the name should be encoded using the <name>'s
reg=" " attribute. The direct form will be encoded within the <name> element.
O <funder>, <sponsor>, <principle>
Should encoders feel the need to use these elements they should follow the same
guidelines set out in the <author> and <editor> elements
(including the optional use of the <name> element) as they pertain to the
form and source of the names used.
R <respStmt>
Group here other individuals or organizations that have contributed to intellectual or artistic
content of the work.
All documents should have at least one <respStmt> for the individual
responsible for the encoding of the electronic text. Each <respStmt> should
contain a least one pair of <resp> and <name> elements. The
<resp> element should contain a statement that indicates the nature of the
contribution. Each <name> element should contain the name as it appears in
on the work. Within <respStmt>, repeat <resp> and
<name> for each contributor that is to be included in the description.
** Use of the <name>'s reg=" " attribute. **
Names that are recorded solely in the <respStmt> (those that have not
been already entered in catalog form in the <titleStmt>) that encoders
wish to have appear as access points in the Dublin Core record (as dc:contributor) should use the
<name>'s reg=" " attribute to encode the catalog form of the name. Do not
use the reg=" " for names that have been recorded in the <titleStmt>
(author, editor, funder, sponsor, principle), since they will be entered as dc:creator (<author>) or dc:contributor (<editor>, <funder>, <sponsor>, and <principle>) in the Dublin Core record (see Appendix B for a Tei Header to Dublin Core Crosswalk).
. . .
Encoded by
Leigh Rios
Proofing and Corrections to markup by
Charlotte Payne
]]>
R <extent>
Record here the approximate size of the TEI XML file to nearest 10 Kb for files under
100 Kb, to the nearest 50 Kb for files over 100 Kb, and to the nearest 100 Kb for files
over 1 Mb. Always precede the entry with "ca."
If the encoded text includes images, follow the file size statement with a space,
semicolon, space (_:_) and an approximate number of images associated with the file ( ; ca. 1200
images).
ca. 30 Kb
ca. 150 Kb
ca. 1.3 Mb
ca. 2.2 MB ; ca. 1100 images]]>
R <publicationStmt>
Record here the publication information for the electronic version of the document.
Within the <publicationStmt> use <publisher> to encode the
name of the person or organization responsible for to production of the electronic text,
<pubplace> for the City and State where the publisher is located,
<date> for year of the file's creation, and <idno> for any
identifying numbers related to the digital text.
** Note On <date> Normalization **
Normalize dates using the value=" " attribute. Use ISO 8601 : 2000 to formulate the
attribute value. The following format is recommended YYYY-MM-DD (using 4 digits for
years and hyphens between the two digit month and date elements). For example March 6,
1975 would be normalized as value="1975-03-06", April 2004 as value="2004-04-00", and
1942 as value="1942-00-00".
R <idno>
An <idno> element that records the CDL System ARK key is REQUIRED for
all CDL TEI texts. This <idno> should be qualified with the type="ARK"
attribute. Local <idno>'s should use the type="LOCAL" attribute.
University of California, Davis
Davis, Calif.
2004
ark:/13030/kt8n39p8kw
81
]]>
O <seriesStmt>
Record here information describing the bibliographic series that the electronic text
is a part. At a minimum the <seriesStmt> should include
<title>, and <respStmt>. Include an <idno>, using the type="VOL" attribute, if the
text is part of a numbered series.
If available, encode the series' homepage using the <xref> element within the <title>. Guidlines for <xref>/<xptr> linking can be found in Section 2.14 of the CDL TEI Base Encoding Guidelines.
<xref href="http://digital.lib.ucdavis.edu/projects/bwrp/index.html"
type="url" rend="new">British Women Romantic Poets</xref>
General Editor
Nancy Kushigian
Managing Editor
Charlotte Payne
81
]]>
R <sourceDesc>
The <sourceDesc> contains bibliographical information pertaining to the
original source of the encoded text. Within <sourceDesc>, record bibliographic
information in the in the <bibl> or <bibFull> subelement.
R <bibl> or <biblFull>
The choice of bibliographic citation subelement to be used is up to encoders. The <bibl> allows for loosely structured bibliographical data. This option may be more useful for projects dealing with source texts that have not been formally published (such as manuscript materials). The <biblFull> provides for a fully-structured bibliographic citation, in which all components of the TEI <fileDesc> are present. The choice of which type of bibliographical citation is to be used, is up to encoders, project managers, etc.
or ]]>
Guidelines for the use of subelements within <biblFull>
Elements in this section follow the same guidelines as the <fileDesc>
unless otherwise noted.
R <titleStmt>
Must include minimally a <title> child element.
R <title>
Within the <title> subelement, transcribe the title as it appears in the
source document, using "_:_" for each subtitle. Supplied titles should be placed within
[ ]'s. Do not include the "electronic version" subtitle.
M <author> and <editor>
Enter names, personal or corporate, in CATALOG FORM.
Texts that emanate from multiple authors should be encoded using multiple
<author> or <elements>.
Optionally: Encoders may use <name> as a child element of
<author> or <editor> and its reg=" " attribute to encode both the catalog and direct
forms of author names. *NOTE: This optional encoding has no effect on the Dublin Core
access points.
Acton, Eliza, 1799-1859]]>
M <respStmt>
Transcribe responsibility statements as they appear in the source text. Statements
that describe the nature of the intellectual or artistic contribution should be encoded in
<resp>. Names of people or organizations should be placed in
<name>. Repeat <resp> and <name> for each contributor.
Optionally: Encoders may use <name>'s reg=" " attribute to encode both the catalog and direct
forms of names. *NOTE: This optional encoding has no effect on access points in the Dublin Core record.
By
Mrs. Addison
]]>
M <editionStmt>
Using the <edition> subelement, transcribe the edition statement if one
appears in the source document.
3rd ed.
]]>
R <extent>
Minimally, enter the number of pages of the source document. If a page approximation
if being encoded use "ca." Ideally, encode extent in AACR2 form if available.
50 p.
ca. 600 p.
100 p. : ill ; 32 cm]]>
R <publicationStmt>
Encode the name of the publisher <publisher>, place of publication
<pubPlace>, date of publication <date>, and any identification
numbers (such as a call number) <idno>.
Materials that have not been previously published (letters, reports, memos, etc.) may
not contain publisher or place of publication information. For this reason
<date> is the only required element.
Supplied information should be enclosed in [ ]'s. Date approximations should be
encoded with a ca. before the date. Dates that are uncertain should be followed by a
question mark (?).
** Note On <date> Normalization **
Normalize dates using the value=" " attribute. Use ISO 8601 : 2000 to formulate the
attribute value. The following format is recommended YYYY-MM-DD (using 4 digits for
years and hyphens between the two digit month and date elements). For example March 6,
1975 would be normalized as value="1975-03-06", April 2004 as value="2004-04-00", and
1942 as value="1942-00-00".
M <seriesStmt>
Encode information about the bibliographical series that source text may be part of.
Use same <seriesStmt> guidelines as described in the
<fileDesc>. For the series <title>, consult national authority
files and use the established form of the series title if available.
Guidelines for the use of subelements within <bibl>
Encoders using <bibl> have much more flexibility in the how they can construct citations of source texts.
For example, within <bibl>, encoder may transcribe a simple text citation
Blain, Clements and Grundy: Feminist Companion to Literature in
English (Yale, 1990)
]]>
Encoders may also use many of the elements available in the <fileDesc>:
TITLE
FIRSTNAME LASTNAME
Revisions by:
FIRSTNAME LASTNAME
# of pages, leaves, etc.
DATE OF CREATION
LOCALLY DEFINED ID#
]]>
O <encodingDesc>
Institutions should consult specific local guidelines and CDL TEIFORM Guidlines for the
encoding of <projectDesc>, <editorialDecl>,
<refsDecl>, <samplingDecl>, and <tagsDecl>
O <classDecl>
If keywords from a controlled vocabulary are to be applied to the description the
source, they should be declared in the <classDecl> section. For each subject or
name thesaurus that is going to be referenced use the element <taxonomy>.
Within each <taxonomy> enter a <bibl> element. Finally, embed
<title> within the <bibl> tags. Encode the title of the
thesaurus in the <title> element.
Use the <taxonomy> attribute id="" to encode an identifier that will allow subject
headings (in <textClass>) to be linked to their source thesauri. Repeat
<taxonomy>/<bibl>/<title> elements for each
controlled vocabulary to be used.
Library of Congress Subject Headings
Library of Congress Name Authority File
]]>
Encoders who wish to encode local or project defined <taxonomy>'s can do
so using the <category> and <catDesc> elements. Like other
taxonomy encodings, an id=" " attribute must be encoded so that terms in
<textClass> can be linked to their source vocabulary. Use
<catDesc> to record a description or title of the thesauri or categorical system.
Original UC Berkeley Dynaweb collection structure. July
2002
Library of Congress Name Authority File
]]>
O <profileDesc>
O <langUsage>
Encode in <profileDesc> one <langUsage>. Embed in each
<langUsage> a <language> element for each language that is used.
** Note on <langUsage> **
The <langUsage> element is required when the lang=" " attribute has
been used within the <foreign> that requires special fonts, left-right handling, or other special treatment. Encode the name of the Unicode chart (http://www.unicode.org/charts/) needed to render the character data.
(Range: 0370-03FF)
(Range: 0400-04FF)
]]>
O <textClass>
Use <textClass> to wrap <keywords> subelement headings
that come from the controlled vocabularies referenced in the
<classDecl>. The <keywords> element is always required as the
root element for grouping a list of headings from a particular controlled vocabulary. Use
the <keywords> attribute scheme="" to identify the thesaurus that is to be
used for the list. The value for the scheme="" attribute should be taken from one of the
<taxonomy> attribute id=" "'s that were declared in the
<classDecl>. Within <keywords> use <list> and
its subelement <item> to form a list of headings.
Separate subdivisions with "--" (do not insert spaces between headings and subdivisions " -- ").
- Scottish Poetry--Women Authors--18th century
- Scottish Poetry--Women Authors--History and criticism
- Bannerman, Anne, 1765-1829
- Blamire, Susanna, 1747-1794
]]>
M <revisionDesc>
Record here any significant changes made to encoding or the content of the file.
M <change>
Record each revision using the <change> element. List all changes in
order from most recent to the earliest. Each <change> should include
<date>, <respStmt>, <resp>, and
<item> in that order. The <item> element should contain an
explanation of the revision that has been made to the text.
02/28/2004
Charlotte Payne
Project Manager
- Converted SGML to XML
06/14/1999
Charlotte Payne
Project Manager
- Revised teiHeader
]]>
Minimum and Full CDL TEI HEADERS
STWG Header: Minimum
TITLE OF TEXT : ELECTRONIC VERSION
Text encoder:
FIRSTNAME LASTNAME
in Kb. or Mb.
NAME OF PUBLISHER
PLACE OF PUBLICATION
DATE OF PUBLICATION
OAC ARK#
BIBLIOGRAPHIC CITATION
]]>
STWG Header: FULL
TITLE OF TEXT : ELECTRONIC VERSION
LASTNAME, FIRSTNAME, YYYY-YYYY
LASTNAME, FIRSTNAME, YYYY-YYYY
by
Firstname Lastname of AUTHOR
Text encoder:
FIRSTNAME LASTNAME
ELECTRONIC EDITION STMT
in Kb. or Mb.
NAME OF PUBLISHER
PLACE OF PUBLICATION
DATE OF PUBLICATION
OAC ARK#
Local ID#
<xref href="Series or Collection URL"
type="url" rend="new">SERIES TITLE</xref>
FIRSTNAME LASTNAME
TITLE
LASTNAME, FIRSTNAME, YYYY-YYYY
LASTNAME, FIRSTNAME, YYYY-YYYY
By
FIRSTNAME LASTNAME
ed.
p. : ill. ; cm
NAME OF PUBLISHER
PLACE OF PUBLICATION
DATE OF PUBLICATION
LOCALLY DEFINED ID#
SERIES TITLE
FIRSTNAME LASTNAME
NAME OF CONTROLLED VOCAB
NAME OF ANOTHER CONTROLLED VOCAB
NAME OF LOCAL OR PROJECT DEFINED CONTROLLED VOCAB
(Range: 0370-03FF)
'S">
- NAME OR SUBJECT HEADING
-
">
- NAME OR SUBJECT HEADING
-
Date of revision
PERSON RESPONSIBLE FOR REVISION
- EXPLANATION OF REVISION
]]>
TEI HEADER to DUBLIN CORE CROSSWALK
The following crosswalk describes the how data encoded in the TEI Header is used to create
Dublin Core records that will be used in the texts' Metadata Encoding and Transmission
Standard (METS) records. The Dublin Core record will be used within the METS descriptive
metadata section (<dmdSec>)
DC:ELEMENT
LOCATION OF DATA IN TEI HEADER
dc:title
tei.2/teiHeader/fileDesc/titleStmt/title
dc:creator
tei.2/teiHeader/fileDesc/titleStmt/author
dc:subject
tei.2/teiHeader/profileDesc/textClass/keywords/list/item
dc:description
tei.2/teiHeader/encodingDesc/projectDesc
tei.2/teiHeader/encodingDesc/editorialDesc
tei.2/teiHeader/encodingDesc/refsDecl
Others?
dc:publisher
tei.2/teiHeader/fileDesc/publicationStmt/publisher/publisher and pubplace
dc:contributor
tei.2/teiHeader/fileDesc/titleStmt/editor
tei.2/teiHeader/fileDesc/titleStmt/funder
tei.2/teiHeader/fileDesc/titleStmt/sponsor
tei.2/teiHeader/fileDesc/titleStmt/principle
tei.2/teiHeader/fileDesc/respStmt/name/@reg
(Names from the respStmt will only be pulled in the DC record when the
catalog form of the name has been entered in the @reg)
dc:date
tei.2/teiHeader/fileDesc/publicationStmt/date
dc:type
tei.2/teiHeader/@type (for CDL TEI Form)
dc:format
text/xml
dc:indentifier
tei.2/teiHeader/fileDesc/publicationStmt/idno/@type="ARK"
dc:source
tei.2/teiHeader/sourceDesc/bibful/publicationStmt/publisher, pubPlace, and date
tei.2/teiHeader/sourceDesc/bibl
dc:language
tei.2/teiHeader/profileDesc/langUsage/language
dc:relation
tei.2/teiHeader/fileDesc/seriesStmt/title OR title/xref/@href="url"
dc:coverage
N/A
dc:rights
tei.2/teiheader/fileDesc/publicationStmt/availability
Or does CDL boilerplate this?