Skip to end of metadata
Go to start of metadata

Pending Issues

  • Angela tasks
    • Verify bibliographic anatomy of the notes; make sure they confirm to MLA; update these guidelines and examples to reflect MLA style
    • consider also using the OCM– Outline of Cultural Materials from Yale – for topical (text and phrase-level) indexing
      • Already in XML form so we can covert into a plugin easily, I think. Need to look into licensing issues.
    • consider using FAST headings via's OCLC's Classify web service for topical headings
    • investigate GSAFD for genre terms: http://alcme.oclc.org/gsafd/OAIHandler?verb=ListIdentifiers&metadataPrefix=z39_19 (another OCLC web service)
    • need to update the <relation> description; mutual, active, passive need to be explained correctly
    • review pending encoding issues on VWWP Encoding Problems page; codify solutions in encoding guidelines; archive problems resolved and/or codified
    • begin brainstorming prosopography strategy with Rebecca; document on wiki somewhere
  • Michelle tasks
    • Notes
      • Done. Angela and Rebecca checking for MLA Style compliance.
    • New Schema (notes, reflect relation update and annotation/notes typology)
    • Update XSL that generates the TEI shell
    • Every div NEEDS a head
      • <head type=”supplied”>
    • Update references to critical intro and bios in here and in the VWWP header template; add new instructoins (place holder in the guidelines in gree)
    • Add to date description in guidelines and QC checks:  <date type="unknown">[n.d.]</date>
  • Angela and Michelle tasks
    • Finish first pass of editorial review
    • After Rebecca encodes the annotations, review files again (reference QC documentation)

Quick Links

Encoding Contextual Information

Reference

About TEI and XML

TEI P5 Chapters of Importance

About Victorian Women Writers Project


Getting Started

The books are encoded in TEI P5, with special attention to structural tagging (sections, paragraphs, page breaks, chapter headings, line groups, etc.) and other stylistic features such as textual emphasis (bold, italics, etc.) within the texts. For this project we will be exploring a limited range of semantic encoding at the division (e.g., poetry) and phase-level (e.g., personal names).

A shell XML/TEI document has been generated after concatenating the OCR text files. The header and body of text requires metadata and mark-up following the TEI P5 Guidelines and the encoding guidelines contained herein.

You will use the latest version of the Oxygen editor for encoding (download from IUWare. To learn more about the editor, reference the user manual. The DLP also has basic documentation about editor preference settings and short cuts.

Encoding Workflow

Encoding Setup and Validation

Before you begin encoding, you will need to setup your encoding environment and understand validation principles. This information is maintained in another page since you will only need to reference it as-needed.


Encoding Guidelines

Global Guidelines

Below are encoding scenarios that can be found across all texts.

What to Encode (and not to encode)

  • Intellectual Content: Only encode the printed text. Any handwritten notes or library/owner stamps should be ignored.
  • Running Headers / Running Titles: Do not encode running headers like chapter headings. Running headers may include such information as the book title, chapter title, author's name, etc. This text should be deleted from the XML/TEI document.
  • Page Numbers: Do not encode the page numbers that may appear at the top or bottom of each page. Delete the page number from the place it appears on the page and instead, follow the directions below for capturing page number information.

Running Headers

In general, we will not not encode running headers like chapter headings. Running headers may include such information as the book title, chapter title, author's name, etc. This text should be deleted from the XML/TEI document. However, running headers that are unique or meaningful (i.e., emphasize the storyline in some way; clearly the author had a hand in determining these headers, not the publisher) should be encoded. One indication of significant running headers are those that change from page to page.

Page numbers should be deleted from running headers and instead added to the page break tag.

Use the appropriate value for place as provided automatically by the TEI:

  • top and bottom being the more common values

Character Encoding

  • All characters, including "diacritics," should be entered directly (e.g., é) rather than using entity references, with the exception of the three predefined XML entities listed below:

    & = &amp;
    < = &lt;
    > = &gt;
    
  • If these entities are not used in place of the desired character, the TEI/XML document will be invalid.

For help with how to type diacritics, reference the:

  • "Character Map" utility (Windows, in the Start menu, Programs=>Accessories=>System Tools=>Character Map)
    • Click on the character you want, and then click on "Select" (the character should appear in the "Characters to copy" box) and then click "Copy". You can then paste the character into your XML file in Oxygen.
  • OR the "Character Palette" (Macintosh, in System Preferences, under International, select "Character Palette" and "Show input in menu bar" for easy access to the palette)

Em-Dash and End-of-Line Dashes

Many authors of this period are fond of using em dashes--like this--in order to add grammatical emphasis or a parenthetical comment. In poetry, em dashes may appear in the middle or at the end of a line. These parenthetical pauses shouldn't be confused with naturally hyphenated words or words that run off the page and are finished on the next line. Em dashes serve a unique grammatical purpose and are to be reproduced faithfully.

OCR usually interprets em dashes as a single dash. To correct this, replace the single dash with the code, --- as in the example below.

For end-of-line dashes used as a printing convention (word wrapping due to preset margins), delete the dash or hyphen so the word appears as one (e.g., laugh-ing should appear as laughing, without the dash).

Text Formatting

Typographically distinct text should be encoded with hi (e.g. <hi rend="bold"> unless it is contained by some other take to which the rendering value can be applied (e.g., <head rend="i">}}Love me or leave me.</head>). By default, the {{<title level="m"> is rendered in italics.

  • i for italics
  • b for bold
  • super for superscript (e.g., a footnote number)
  • sub for subscript
  • uc when a word is all uppercase capitals and the font is larger than the surrounding text
  • sc when a word is all small but the font size is the same as the surrounding text
  • blockquote for content that needs to be blockquoted
  • center for content centered on the page
  • left for content to the left of the page
  • right for content to the right of the page

Note: the <emph> element will NOT be used. Always use the <hi rend="i"> to mark up any phrase-level text that is italicized, but is not contained in some other tag (e.g. <head>).

Indentation as Commonly Used in Poetry


In poetry, the following values can be used with the rend attribute of the <l> element to indicate the indentation of lines of poetry (see also Tags Used Especially for Poetry):

  • ti-1 (text indent level 1)
  • ti-2 (text indent level 2)
  • ti-3 (text indent level 3)
  • ...
  • ti-10 (text indent level 10)

Lists


For lists, the following values can be used with the rend attribute of the <list> element to indicate styling:

  • Bulleted lists (type="bulleted" rend="disc")
    • disc (filled circle)
    • circle
    • square
  • Ordered lists (type="ordered" rend="decimal")
    • decimal (1, 2, 3, 4, 5, etc.)
    • lower-latin (a, b, c, d, e, etc.)
    • lower-roman (i, ii, iii, iv, v, etc.)
    • upper-latin (A, B, C, D, E, etc.)
    • upper-roman (I, II, III, IV, V, etc.)

List of type="simple" do not require a rend attribute.

Tables


Tables currently have two rendition values:

  • rend="border" for tabular data displayed with borders
  • rend="noborder" for tabular data displayed without borders

An example of the latter can be found in the title page of Notes upon some of Shakespeare's plays (see TEI snippet below):

Not all possible rendition values for tables have been identified so if you encounter a table with a particular border that should be displayed differently, please indicate this on the VWWP Encoding Problems page.

These rend values may be combined, separated by white space, e.g., <hi rend="u b">this text will be bold and underlined</hi>, except when marking up poetry lines (e.g. <l n="3" rend="ti-1 b"). Instead use the <hi> element after the <l> element (e.g. <l n="3" rend="ti-1"><hi rend="b">The content of the line</hi></l>).

Page Breaks

Page breaks can come at any point in the text. You mark page breaks with a self-closing <pb/>. Page breaks are marked at the beginning of the content for that page.

Page breaks will have 2 attributes:

  1. number (n="")
  2. xml:id (xml:id="")

Paragraphs and Line Breaks

  • <p>: Used to surround paragraphs.
    • Do not close a paragraph at the end of a page unless the paragraph actually ends there. Instead, nest the <pb/> tag between the <p> and </p> tags at the appropriate location in the text.
  • <lb/>: A line break is used when it is important to represent a break in the textual flow.
    • Generally used in headings, on the title page, lists, etc.
    • Do not use in standard prose paragraphs.

Number Attributes in Page Breaks

  • The actual page number must be entered for the "n' attribute.
    • If a page number is printed on the page, enter it as it appears (e.g. n="32")
    • If a number does not appear on the page, but is part of the numbering sequence (e.g. chapter title page), enter the page number in brackets (e.g. n="[45]")
    • If a page does not have a number, enter empty brackets for the 'n' attribute (e.g. n="[]")

The page break tag can come at nearly any point in the text, no matter where one is in the encoding. For example, if the page break comes before the end of a paragraph, encode the <pb/> inside the paragraph tags:

In the following, more complicated example:

  • The first page is the frontispiece, which has no page number and is not part of a numbering sequence.
  • The title page has the number i, but that number is not printed on the page.
  • The table of contents begins on page ii (no number printed on the page) and continues on page iii (the number is printed on the page).
  • The first chapter begins on page 1, however the number is not printed on the page.
  • The chapter continues on pages 2 and 3, and both of those numbers are printed on the page.
  • Notice that the pb tag for page 4 comes between the closing div tag for chapter 1 and the opening div tag for chapter 2. Chapter 1 ended on page 3, so the closing div tag must come before the pb tag for page 4. Since chapter 2 begins on page 4, the opening div tag for chapter 2 must come after the pb tag for page 4.
  • Notice also that no pb tags appear between the front and body tags (nor should a pb tag appear between the body and back tags).

xml:id attribute

The second attribute on the page break element will be the xml:id. The xml:id will be used to coordinate the transcribed text with the scanned images of the original print text. It will also be used to coordinate the table of contents entries with the correct section of the text (see linking section below). The xml:id will be generated automatically from the unique identifier for the electronic text and a three digit sequence. For instance, if the identifier for the text is VAB1983, and the page number is 3, then the xml:id might look like: VAB1983-003. As the encoder, you:

  • will not alter that number
  • will want to check that the page break xml:id numbers correspond to the correct scanned image.

Linking the page breaks and table of contents

In order to link the table of contents to the correct section of the electronic text a <ref> element will be placed around the part of text that you want to link with another section of the text. A target attribute is used to reference the xml:id of the page number where the corresponding content begins. The pound sign (#) signals that the link is being made within the text rather than outside it.

In this example, the Chapter 1 label is linked to the page number where chapter one begins by using the xml:id of that page:

There may be cases, such as in poetry, where two poems may occur on the same page. In such a case, place the same xml:id as the value in both the target attributes.


Lists

Lists are ordered, itemized information that are marked by the using the <list> and <item> elements. Lists can have headings, come within paragraphs and divisions, or even within another list (see examples in TEI P5 guidelines).

With the use of attributes, one can indicate the number of items in a list and whether a list is bulleted, numbered, or otherwise marked. The most common type of list will be "simple", which merely consists of a list of items, as seen in the first example below. It is also important to indicate the styling with a rend attribute in the <list> tag:

  • Bulleted lists (type="bulleted")
    • disc (filled circle)
    • circle
    • square
  • Ordered lists (type="ordered")
    • decimal (1, 2, 3, 4, 5, etc.)
    • lower-latin (a, b, c, d, e, etc.)
    • lower-roman (i, ii, iii, iv, v, etc.)
    • upper-latin (A, B, C, D, E, etc.)
    • upper-roman (I, II, III, IV, V, etc.)

List of type="simple" do not require a rend attribute.

Lists of Definitions and Terms, Glossaries

Lists are not always confined to a group of items. Lists are also used to mark up glossaries. Glossaries are marked up using the <list type="gloss"> and <label> tags. The <item> tag is then used to denote the definition, as show in the example below.

Figures

Figures and illustrations that appear in the text are represented with the <figure> tag. The actual graphic will not be included in the mark up, ONLY the caption or other accompanying text. The caption or accompanying text is marked up using a <p> element.

Example:

Frontispieces are frequently found in front matter and are described in more detail in the front matter section of these guidelines.

Tables

A few texts may include tables. Tables are marked with the <table> tag.

Each row in the table is marked with a <row> tag. The <row> element contains a role attribute to determine whether the information in the table is a heading or data. Within the <row> tag, there are <cell> tags. These tags indicate the specific units within the table. There should be as many cells as there are columns in the table. The rows are in order, but cells are used to indicate columns, rather than a separate <col> tag. The text from the table is placed within the <cell>.

Tables can come inside of paragraphs, lists and many other elements.

Tables should have a rendition (@rend) style designated; either rend="border" or rend="noborder." Learn more about table renditions in the Text Formatting/Rendition section of the Guidelines.

Headings/Titles

Headings include the titles of lists, chapters, sections, etc. in a work. Most commonly, you will use them for chapters, lists and sections. Headings are marked up using the <head> element.

Closers

Closers are most commonly seen at the ends of letters, but they also appear at the end of prefaces, pamphlets and poems. They include the closing salutation, such as "sincerely," the author's name, and sometimes information about the date of publication or the place. Closers are grouped inside the <closer>, which can contain various other elements.

Corrections

Always correct errors or typos found in the printed texts (these errors are distinctly different from OCR errors):

  • the <sic> element indicates the printed error
  • the <corr> element indicates your correction
  • the <choice> element allows both the apparent error and its editorial correction to be recorded, as in the following examples:

Notes

Use the note element to encode the text of a margin note, footnote, endnote, or other note found in the source document. Typically these notes are provided by the:

Notes by the author do not require a type value of "author." By default, all notes that are not typed are considered notes provided by the author. Please do type notes from the editor and publisher as:

  • <note type="editor">
  • <note type="publisher">
  • Use the <ref>element to surround the number, symbol, etc used to link the point of reference to the note.
    • Use the "target" attribute to point to the actual note
    • Use the "rend" attribute if applicable (e.g., to denote superscript; see allowable values above)
  • Use the <note>element around the content of the note.
    • Use the "xml:id" attribute to facilitate linking from the "target"
    • Use the "place" attribute to indicate where the note should appear. Sample values include:
      • top
      • bottom

      • end
      • side
      • inline
  • If portions of the note run on to other pages, place all of the text of the note at the point where the note begins.

Note ID/target schemes should be defined in numerical order as notes (numbered or unnumbered) appear in the text. The id scheme is as follows:

  • note_001, note_002, etc. Each note should be uniquely identified.
    • Always use leading zeros to indicate values up to 999.

Notes that Span Pages

When you encounter a note that spans pages, for readability, collapse the note into one page, the first page it occurs.  

To signify the page break, which is important for citing and other reasons, we will impose the following editorial markup:
 

 

  • type="editorial" (always true)
  • resp=encoder/editor ID as stated in the statement of responsibility section of the TEI Header
  • rend="i" (always true)
  • notation about pagination should be structured as follows:
    • [con't on p. # (reference printed page number or supply the page number if it is not printed)]
For example, text VAB7401.xml contains the following collapsed note:

 

 

Citations

For notes in the form of a citation even a loose citation, tag the note using a <bibl> tag (as opposed to <biblStruct> or <biblFull>. For example:

Use <biblScope> to indicate volumes, issues and pagination by selecting the appropriate "type" value (as shown in the example above). Other values are also possible:

  • pp = pagination
  • chap = chapter title
  • issue
  • ll = line numbers
  • part
  • vol

When using the <title> element in the text be sure to use the level attribute when ever possible. The TEI provides values such as "m" for monographs, "a" for analytic (which would be used for a journal article, for instance), etc.

Encoder Supplied Content: Annotations, Citations, etc.

Annotations

English graduate students compile annotations as part of their scholarly encoding process. These annotations are distinct from "notes" provided by the author, editor or publisher.

Notes provided by scholars will have the following attributes:

  • type (emphasize supplied note as opposed to printed notes part of the text)
    • scholarly
  • subtype (nature of the note)
    • gloss
    • explanatory
    • translation
    • citation
  • resp (the encoder/scholar)

Annotations that contain in-line citations regardless of their subtype, should use <bibl> to identify the citation (even if just a title or author).

Placement of Annotations and ID schemes

The scholar-supplied notes shoul be place in the <back> section of the TEI document with the following attributes:

  • type="notes"
  • subtype="scholarly"

For example:

As stated in the Notes section of this document, the Note ID/target schemes, for both notes found in the printed texts or supplied by scholars, should be defined in numerical order as notes appear in the text (whether numbered or unnumbered). The id scheme is as follows:

  • note_001, note_002, etc. Each note should be uniquely identified.
    • Always use leading zeros to indicate values up to 999.

Encoding the Referent

  • ptr v. ref

Types of Annotations

Four types of annotations have been defined:

Gloss Annotation

A gloss annotation is usually a brief notation of a word or phrase in text often in the form of a definition.

Explanatory Annotation

An explanatory annotation is an extended explanation intended to provide additional context of the word, phrase or passage annoated.

Translation Annotation

A translation annotation is often just that a translation or explanation of a word or phrase in a language other than the dominant one of the text. Make sure you indicate the language if known.

Citation Annotation

A citation annotation is typically a complete bibliographic citation following the MLA Style that is otherwise lacking
in the source text.

THIS ESPECIALLY NEEDS TO BE CHECKED AND UPDATED TO REFLECT MLA STYLE

Clarifying or Enhancing Citations

Often, along with providing a citation annotation (as defined above), students will conduct additional bibliographic research to determine attribution or source for quotes or epigraphs cited in the source text. These citations should be tagged much like a citation annotation as described above.

Typically the additional bibliographic information will be added to an epigraph or the <cit>/<quote>/<q> tags.

For example:

Note: The use of <ptr> or <ref> once it is defined.


Encoding Structure

This is a typical encoding structure for the monographs:

  • teiHeader
  • text
    • front
      • advertisement
      • title page
      • preface
    • body
      • chapter or poem or act, etc
    • back
      • appendix
      • advertisement

Please note that NOT every monograph contains this structure. Some may not have back matter and the internal structure within the body of the book will change depending on its genre:

  • Prose usually contains chapters, but not always
  • Verse may be sub-divided by parts, but not always and will contain a series of line groups
  • Dramas may be divided by sections (e.g., acts), and will contain various speech conventions

TEI Header

Reference: <teiHeader>

Every XML file contains a basic TEI Header with boilerplate information already completed and bibliographic metadata extracted from the MARC record. You will complete/update the information represented by a dollar-sign variable (e.g., $Encoder's First and Last Name). The following TEI Header information will need to be completed (see example header below):

  • In the <fileDesc> <titleStmt>: <title type="filing">?Title without articles</title>
  • In the <fileDesc> <respStmt>: <name>$Encoder's First and Last Name</name>
  • In the <fileDesc> <publicationStmt>: <date>$Year of Encoding</date> (4 digit year)
  • ADD INFO ABOUT RELATED MATERIALS
  • In the <fileDesc> <profileDesc> <particDesc: in support of prosopography
  • In the <fileDesc> <profileDesc><textClass><keywords scheme="#mla">: <term>:$genre heading</term> (see explanation below)
  • In the <fileDesc> <profileDesc><textClass><keywords scheme="#victbib">: <term>:$subject heading</term> (see explanation below)

 

Noting Pseudonyms

If you enounter a pseudonym whether or not it is reflected in the bibliographic information already provided in the Header, update the Header as follows:

Adding the Filing Title

  1. For those titles beginning with an article:
    1. Add a second title tag with a filing type only in the fileDesc/titleStmt:
    2. Remove the leading articles (English and non-English)

 

Linking to Related Materials: Intros and Bios

  • NEED TO ADD INFO

Selecting Subject and Genre Headings

You will need to select subject and genre headings for the text you are encoding. The headings come from two sources:

  • Victorian Studies Bibliography taxonomy, which we will use to identify the prominent subject(s).
  • MLA thesaurus and the Oxford Dictionary of Literary Terms, which we will use to identify the genre (e.g., pamphlet, sensational novel, etc.). Please select the most specific term(s).

Adding Subject Terms

To select the subject terms:

  1. Place the cursor between the opening and closing <keywords scheme="#victbib">
  2. Right-click
  3. Select from the menu: Plugins => Terminology Service Plugin
    1. A window will pop-up with two tabs: Victbib and MLA
  4. Select the "Victbib" tab
  5. Select "Add Terms" button, which will launch a search window
    1. If you know the term, search
    2. If you don't know the term, browse the list of subject headings
  6. Find the relevant term
  7. Click "Add Term"
  8. When you are done adding the term, select "Update Selection" ("update selection" saves the term you selected)
  9. If more terms apply, repeat the previous two steps as necessary
  10. When you are happy with all the terms selected, click on "Update Document" (your cursor will jump down to the bottom of the text, sadly)
  11. Scroll back up to the <keywords> section of the TEI Header to confirm terms were added

Adding Genre Terms

Repeat these steps for the genre terms, except:

  1. Place the cursor on the first placeholder <term> in <fileDesc> <profileDesc><textClass><keywords scheme="#mla">
  2. Select the "MLA" tab (note: MLA genre terms include descriptions to help you identify the proper genre)
  3. Follow the steps as outline above

Prosopographical Information

Note!

Capturing prosopographical information is experimental in nature. We hope to learn from you as you record this information about what works and what doesn't. We will also perform analysis of this work after the texts have been encoded to determine how to best proceed. But in the meantime, have fun with it! Recording information about personages has lots of great potential.

Broadly speaking, prosopography is the study of individuals within a specific context and their relationships to one another. The TEI P5 Guidelines now dedicate a section to the compilation of prosopographical information. While it seems more geared toward demographic information, it is also applicable to fictional and non-fictional works.

The prosopography section, included in the <particDesc> section of the TEI Header, is comprised of one main section, <listPerson>, which is further subdivided by many components, the main ones being:

  • <person>
  • relationGrp>

(Reference the TEI Header template above to see the whole of how the prosopographical data is contructed. Below, examples are included, but in chunks to illustrate the topic at hand).

<listPerson>

This is the container tag that will contain your list of people! It is important to assign it a type attribution with one of the following values:

  • fictional
  • historical
  • biblical
  • mythical

If you encounter another type of person, report this issue in the VWWP Encoding Log page.

<person>

Many characteristics can be recorded about a particular personage. Below are some examples, but feel free to reference the section on Biographical and Prosopographical Data in the TEI P5 Guidelines. At a minimum, you should have, if relevant to your particular work, a person's name. None of the other characteristics listed below or otherwise documented in the P5 Guidelines are required.

All persons required an xml:id, the value for which you will create based on some combination of fore- and surnames. Remember that xml:id values must be unique so if you have two characters, Joe and Jane Garcia, you can't asdcribe jgarcia as the xml:id for both. Instead one would have to be joegarcia and the other janegarcia.

The <person> tag can contain any number of predetermined characteristics. Below are just a few that are possible to capture. Again, consult the P5 Guidelines for additional options, if you are interested. Remember that none are required and each personage could have a different set of characteristics. In other words, they do not need to be uniform across personages.

relationGrp

Personages or characters can be explicitly related in the TEI Header. This ability is probably the least intuitive with respect to markup so hopefully my limited examples will capture the range of relationships you may encounter. Again, if you encounter a relationship not defined, report this in the VWWP Encoding Problems page.

Relationships may be either mutual or non-mutual. A mutual relationship is one in which the participants are equal, such as a sibling, co-worker, or spousal relationship.

A non-mutual relationship currently includes only two types of roles: active and passive. According to the TEI Guidelines 13.3.2.4, these designations suggest that the relationships are best described by using a transitive verb, requiring a subject and object. The subject is active and the object is passive. In a parent-child relation for example, the parent is active in a “parent” (read “is parent of) relationship and the child is passive. In a “child” (is child of) relationship the active and passive roles are reversed.

Possible relationship values for the attribute "name" include:

  • spouse
  • sibling
  • parentchild
  • fiancé(é)
  • friend
  • lover
  • apprentice
  • adversary
  • colleague
  • coworker
  • masterservant
  • others you encounter (the possibilities are endless)

Sections of the Text:

Front Matter

The front matter consists of any writing or images that comes before the beginning of the main text (e.g., table of contents, preface, frontispiece, etc.). Most texts in the collection should have front matter, but some texts may have more front matter materials than others. All front matter will be marked up using the <front> element. The most basic front matter will include only a title page. The following example show the markup for a title page.


Some other common features that occur in the front matter after the title page are dedications or epigraphs and table of contents. Both features use the <div> element to contain the information and are marked up as follows:


For more detailed information about what may appear in the front matter how to encode it see the Front Matter page.

Body

The body of the text consists of the chapters, poems, and other matter that make up the primary text. The body is encompassed by the <body> tag.

Divisions of the Text

For the Victorian Women Writer monographs, we use divisions (<div>) with type attributes to denote sections of the text. (See one of the specific genre types for more detail.)

A few exceptions that may occur in the body of any genre:

Embedded Text/Floating Text

Sometimes an embeddd text may appear in a work (common in prose): a letter, journal entry, poem, songs, etc. If this is the case, use the <floatingText> element to separate this embedded text from the rest of the work.

What to include inside <floatingText>:

  • <body>, <div> with a type attribute which describes the text that follows (e.g., letter, poem, song, etc.), any other tags needed to encode the embedded text.

For instance, below is an example of how to encode a letter with the <floatingText> element.

For more information on how to encode embedded texts and which division types you may encounter, please consult the prose guidelines.

A work that contains more than one book by an author

Although it is rare, there may be a case where a work has more than one book inside it. The most common occurrence of this is poetry work that includes several book titles by one author. When this occurs, replace the global <body> element with a <group>.

What comes inside the <group> tag:
  • <text>
  • <front> (optional)
  • <body>
  • any tags needed to encode the embedded text
  • <back> (optional)

The <back> element, as well as the <front> element, are optional when using <group>. Only use these elements if there is front matter and back matter that needs encoding.

Personal Names

In conjunction with the prosopographical information we are compiling in the TEI Header, personages will be tagged accordingly throughout the text. In most cases, this will occur in prose texts that contain dialogue. In those cases, the <said> tag will be used with a who attribute. However, names may appear across genres and in various contexts. In those cases, where names are mentioned, use <persName> with the ref attribute to points to the xml:id of the name established in the TEI Header as part of your prosopographical work. For example:

See a specific genre for detailed guidelines:

Back Matter

Oftentimes the books in this collection will not contain back matter. If the text does contain back matter, use the <back> element to encode it. Back matter content can be unpredicatble, but common content include indices, appendix, advertisments, etc. Below is an example of a short back matter that consists of only a note.

For more information how you to compose specific parts of the back matter see the Back Matter page.

  • No labels