PREVIOUS   NEXT   CONTENTS   HOME 

4.1 Multilingual resources: how far can we go?

Translation of basic resources, such as web sites themselves, can be achieved through a time-consuming and complex series of translations. It is quite possible to create whole web sites in a number of languages. Good examples of these are the Pathways to Cultural Landscapes and Planarch project web sites. The Pathways project allows the user to read the site in ten languages accessed at the beginning of the web site by clicking a flag. Planarch utilises a drop-down list to allow the user to select one of four languages. In both cases the web site content has been translated, a process that requires some care to ensure that meaning in certain contexts is not lost. The Planarch project used a specialist organisation to achieve this, reflecting the need to ensure that translation is made accurately and meaningfully.

The ARENA project focused its multilingual work on the interfaces to those shared resources developed through the project. These are:

In both cases the resources are entered at a point allowing the user to select one of six languages. Text describing the archives and those short terms which allow cross-searching are translated to create multilingual interfaces. It was interesting to see how the process of translation had to be addressed both prior to the design of the interface, to facilitate its actual construction, and afterwards when it became clear exactly what each button or search box would do. The important factor for meaningful translation is context.

In the case of both ARENA archives and the ARENA portal the actual data are only available in the native language of the appropriate partner. The translation processes involved in preparing data for resources of many hundreds or even hundreds of thousands of records in multiple languages are beyond current technology and too time-consuming to be done by hand. In the medium term this situation will remain, although automatic translation such as that offered by AltaVista Babel Fish may develop to overcome the problems of meaning in many different contexts.

In working on the translation of resources the ARENA partners were confronted with the question, 'how far can we go'? A dichotomy was identified among the partners while discussing this issue. Providing a resource in a native language is of course important, and the project invested much time and effort into making this a reality. But in addition to this, many of the ARENA partners also elected to translate archive resources into English to encourage international readership. In the end, of course, both issues have merit and indeed were achieved. The role of English is discussed by Oberländer-Târnoveanu elsewhere in this issue of Internet Archaeology.

The Planarch project was a partnership of organisations with heritage management and sites and monuments data responsibilities. It set about the concept of multilingual information sharing in a different way. ARENA concentrated on the technical demonstration of interoperable searching of large data-sets held independently at the partner host organisations, while Planarch set about creating a smaller data-set that contained translated elements. Individual partners selected important sites and monuments from their local resources and made them available in a single database. The descriptive elements of the database were then translated, leaving the summary data and search terms in English. The problems that would have been faced in doing this for a full sites and monument database made it out of the question. The result was a cleaned and translated data set of some 50 records from each of the seven partners. In constrast, the ARENA portal provides access to millions of records from six partners without the same depth of translation.

In both cases a decision had to be made regarding how far the data could be translated. This issue will continue to be confronted by trans-national projects in Europe until automated translation becomes available or a single common language is agreed.


 PREVIOUS   NEXT   CONTENTS   HOME 

© Internet Archaeology/Author(s)
University of York legal statements | Terms and Conditions | File last updated: Tue Sep 27 2005