Levels of aligned translations

Luka Špoljarić and Petra Šoštarić
University of Zagreb
Faculty of Humanities and Social Sciences, Department of Classical Philology
GHLH: University of Zagreb Greek-Croatian and Latin-Croatian parallel digital corpus


Digital editions and corpora of aligned translations in which source language is Greek or Latin present theoretical and practical, as well as scholarly and educational challenges. The groundwork for aligned corpora has been laid down by the Alpheios Project and the Perseus Digital Library; there is much activity in the field of “translation memories” and aligned corpora of modern languages (mainly of administrative documents, with notable exceptions of projects such as MULTEXT). On the other hand, translations play significant role in education: many universities offer courses on “classics in translation”, relying heavily on relationship of the original and its modern language version. Similar aligned translations are nowadays often provided even by scholarly works (such as e. g. Baldi 2002, aimed both at students and at general linguists who do not necessarily have much competence in classical languages). Finally, a special set of questions and requirements may be introduced by aligning Greek or Latin with languages other than English.

How best to exploit opportunities provided at the crossroads of these trends, achievements, and demands? How to approach translations from a philological standpoint?

We will present results of research done on a 1000-sentence corpus of Greek and Latin aligned with Croatian. This corpus, called GHLH (for “grčko-hrvatski / latinsko-hrvatski”, Greek-Croatian / Latin-Croatian), is being realised as XML database, with relationships between source and target language encoded following the Alpheios scheme for aligned text. The encoding is done by university students (of Latin and Greek, native Croatian speakers). the collection, freely available under a Creative Commons license, will be accessible as a BaseX database, with a set of XQueries designed to stimulate research and interpretation of relationships between originals and translations. Texts and translations included in the GHLH corpus are special in two ways. First, our corpus is not limited to ancient text: we are including also samples of Latin texts written by Croatian authors, from the Middle Ages to the modernity. Second, the translations we have chosen are not only close and “literal”, i. e. striving for a 1:1 correspondence between source and target. We want to explore ways and procedures necessary for encoding relationship between a text and translation which is free, “culturally equivalent”, close to paraphrase. Our working hypotheses are that different target languages require different aligning practices (Croatian requires a different approach to aligning than English), and that different “levels of translation” again require different aligning practices. Our work intends to stimulate discussion about these levels.

GHLH Boards in Perseids


  • ἔς τε δὴ ὦν τὰς ἄλλας ἔπεμπε συμμαχίας καὶ δὴ καὶ ἐς Λακεδαίμονα. (Herodotus, 1, 82)
  • Slao ih je i drugim saveznicima , a i u Lakedemon . (Dubravko Škiljan, 2007)
  • So he sent to the Lacedaemonians as well as to the rest of the allies. (A. D. Godley, 1920)

  • poeticis magis decora fabulis quam incorruptis rerum gestarum monumentis traduntur (Liv. pr.)
  • pripovijeda se više kićeno u pričama pjesničkim negoli po neiskvarenim spomenicima povjesničkim (Koloman Rac, 1894)
  • are more fitted to adorn the creations of the poet than the authentic records of the historian (Rev. Canon Roberts, 1912)
  • are rather adorned with poetic legends than based upon trustworthy historical proofs (Benjamin Oliver Foster, Ph.D., 1919)

z/ghlh-paper.txt · Last modified: 2014/11/29 20:26 by njovanovic
Except where otherwise noted, content on this wiki is licensed under the following license: CC Attribution-Share Alike 3.0 Unported
Recent changes RSS feed Donate Powered by PHP Valid XHTML 1.0 Valid CSS Run by Debian Driven by DokuWiki