Multi-language texts


Where isolated foreign words appear in a text otherwise in a single language, or when a text has significant parts in two or more languages (bilinguals, glosses, etc.).

1. Foreign words

The <foreign> element identifies a word or phrase as belonging to a language other than that of the surrounding text. It is usually useful to apply the attribute xml:lang with a language code (see Languages and Scripts) to identify the language of the embedded word.

For the lost line 6 Aurigemma, loc. cit., suggests <foreign xml:lang="la">matris castrorum</foreign>

Transformation using the example EpiDoc P5 stylesheets:

  • Default (Panciera) style: For the lost line 6 Aurigemma, loc. cit., suggests matris castrorum
  • London style: For the lost line 6 Aurigemma, loc. cit., suggests matris castrorum
(IRT: 21)

2. Meaningful divisions

Where significantly sized and meaningfully divided passages of the text and in different languages, it is advisable to use <div type="textpart"> with an xml:lang on each, perhaps rather than declaring a single default language on the <div type="edition">.

<div type="editionxml:space="preserve">                 <div type="textpartxml:lang="lan="a">                     <ab><lb/>Caecilius Diodorus <gap reason="lost"/>                         <lb/>Caesaris delubrum a<gap reason="lost"/></ab>                 </div>                 <div type="textpartxml:lang="grcn="b">                     <ab><lb/>Καικίλιος Διόδωρος ἅμα <gap reason="lost"/>                        <lb/>ἐκ τῶν ἰδίων εὔξατο θ<gap reason="lost"/></ab>                 </div>             </div>
(IRT: 481)

3. Arbitrary blocks

Where significant, multiline blocks of text are in each of two or more languages, but these blocks are not considered otherwise meaningful or part of the citation scheme of the document, it is enough to tag each block with an <ab>, to which the xml:lang may be added.

<div type="editionxml:space="preserve">                 <ab xml:lang="la"><lb/><gap reason="lost"/>dium murum supra<gap reason="lost"/>                    <lb/><gap reason="lost"/>tribunicia potestate<gap reason="lost"/></ab>                 <ab xml:lang="grc"><lb/><gap reason="lost"/>υς μόνος τὸν ναὸν <gap reason="lost"/>                    <lb/><gap reason="lost"/>δεων αὐτοκράτορ<gap reason="lost"/></ab>             </div>
(IRT: 481)

4. Glosses and dictionaries

A text that is made up of words or phrases in one language, and glosses or translations in one or more others, may be tagged with a series of <term> and <gloss> elements, inline, and each bearing the xml:lang specifying the language of the short phrase. Alternatively, if the specific and rich semantics of <term> and <gloss> are to be avoided, <seg> may be used to mark arbitrary spans of text as being in one language or another.

 <term xml:id="seq1xml:lang="san-Brah">mahībhujām·</term>
 <gloss target="#seq1xml:lang="pyx">tg'am·ḥ d'iṁ tiṁ pmir·ḥ CV naḥ</gloss>
 <term xml:id="seq2xml:lang="san-Brah">°unnata</term>
 <gloss target="#seq2xml:lang="pyx">kd'ir·ṁ tra v'a kv'iṁ</gloss>
 <term xml:id="seq3xml:lang="san-Brah">porusa</term>
 <gloss target="#seq3xml:lang="pyx">°o saṁḥ pir·ṁ tg'a</gloss>

In this example, from the Corpus of Pyu Inscriptions project, xml:id and corresp attributes are used to link the Pyu <gloss>es with the Sanskrit <term>s of which each is a translation or equivalent.

Responsibility for this section

  1. Charlotte Tupman, author
  2. Gabriel Bodard, author
  3. Arlo Griffiths, contributor
  4. Marc Miyaki, contributor

EpiDoc version: 9.0

Date: 2017-12-12