Abbreviation fully expanded

<abbr> (abbreviation) contains the abbreviated part, i.e. the letters actually on the stone/page.

expan (expanded word) contains the full expansion of an abbreviated word or term.

ex (editorial expansion) contains the letters added by the editor when expanding an abbreviation.

In the simplest case of an abbreviation where a few letters are on the stone/support, and the editor expands this word by the addition of letters after the abbreviation—with parentheses in Leiden, for example "Aug(ustus)"—the TEI markup recommended by EpiDoc identifies three components of this word:

  1. the fully expanded word: the whole word, including both abbreviation and expansion, that appears in an editorial edition (in our example "Augustus"). This is tagged using the expan element.
  2. the abbreviation: the characters that are actually on the stone, or would appear in a diplomatic transcription (in our example "Aug" or "AVG"). These are tagged using the <abbr> element.
  3. the editorial expansion: the characters added by the editor in expanding the abbreviation (in our example, "ustus"). These are tagged with ex.

In full, our example would therefore be tagged so that each of the three elements contains the letters corresponding to the component in question, with no letters needing to be repeated:

<expan>
 <abbr>Aug</abbr>
 <ex>ustus</ex>
</expan>

Resolved and unresolved abbreviations are treated slightly differently, in that there is no expanded word nor editorial expansion, so only the <abbr> (whose semantics is therefore unchanged) need be tagged. See Abbreviation not expanded at all for more detail on marking up abbreviations for which the expansion is unknown.

Panciera 1991 X.1; SEG; Sosin 2011; : a(bc)

<expan>
 <abbr>a</abbr>
 <ex>bc</ex>
</expan>

Transformation using the example EpiDoc P5 stylesheets:

  • Default (Panciera) style: a(bc)

()
<expan>Καρ<ex>ανίδι</ex>
</expan>

Transformation using the example EpiDoc P5 stylesheets:

  • Duke Databank style: Καρ(ανίδι)

(Source)
Note: Note that for historical reasons the Duke Databank does not use the <abbr> element for expanded abbreviations.
<expan>
 <abbr>Iul</abbr>
 <ex>ius</ex>
</expan>

Transformation using the example EpiDoc P5 stylesheets:

  • Default (Panciera) style: Iul(ius)

<expan>
 <abbr>trib</abbr>
 <ex>unicia</ex>
</expan>

Transformation using the example EpiDoc P5 stylesheets:

  • London style: trib(unicia)

(IRT: 24)

A suspended rather than abbreviated word is tagged in precisely the same way, with the undertstanding (both semantically and by processing such as rendering XSLT and indexing tools) that multiple <abbr> elements within a single expan are considered to compose a single abbreviation between them. For example, in the case of "QQ" expanding to "q(uin)q(uennalis)", we might encode as follows:

<expan>
 <abbr>q</abbr>
 <ex>uin</ex>
 <abbr>q</abbr>
 <ex>uennalis</ex>
</expan>

If the abbreviated text tacitly implies a supplementary word or words which are understood but omitted from the original support (e.g., tribu, filius, etc.), the abbreviation may be encoded in two different ways depending on the rendering desired. In the first, all omitted text is enclosed within ex. In the second, only the characters which form part of the word actually abbreviated are enclosed within ex, while the word(s) omitted but understood are enclosed within a supplied element, whose reason attribute should be given a value of "subaudible". See examples below.


<!-- RIB 293 --><expan>
 <abbr>Poll</abbr>
 <ex>ia</ex>
</expan>
<supplied reason="subaudible">tribu</supplied>

Transformation using the example EpiDoc P5 stylesheets:

  • London style: Poll(ia) (tribu)

()

<!-- RIB 293 --><expan>
 <abbr>Poll</abbr>
 <ex>ia tribu</ex>
</expan>

Transformation using the example EpiDoc P5 stylesheets:

  • London style: Poll(ia tribu)

()

Responsibility for this section

  1. Charlotte Tupman, author
  2. Scott Vanderbilt, author
  3. Gabriel Bodard, author

EpiDoc version: 8.20

Date: 2014-12-04