Translation Arrays: Version Variation Visualization (Phase 2)

Shakespeare is celebrated as "the world's playwright" at the 2012 Cultural Olympiad. His works have been and still are being translated into scores of languages. In many languages, they have been translated over and over again, scores of times. Translations vary in kind, from philologically precise study aids to free 'adaptations' or 'versions'. All are interpretations, displaying variation in language, concepts and values. Together, they are a very rich mine of information about world cultural variation and change, past and present.

Translations, being comparable, lend themselves to computational analysis. Digital media make it possible to bring together very large numbers of translations and explore how and where they differ. Data visualization tools make it possible to survey them and navigate easily through them, even without knowing the languages involved. Statistical analyses of variation among translations can be mapped onto Shakespeare's English play texts, so we can see where his work prompts more and less variation among translations, which passages are likely to be cut, or which translations expand upon which character parts.

We call this innovative concept a 'Translation Array', by analogy with an astronomical 'telescope array'. A translation array is a sophisticated tool for creating as yet unseen images of world culture. The concept applies to any multiply translated 'world cultural text' - in literature, religion, or philosophy. Implementing it requires solving some very interesting problems of multiple text alignment, fine-grained algorithmic analysis of variation, visualisation of the results, and interactive graphical interface design.

Our proposal now is to build a first prototype translation array, using a digital corpus of over 50 German translations and adaptations of Othello, dating from 1766 to 2006, and a selection in other languages. This corpus and some experimental visualizations were created in Phase 1 of our project at Swansea University in 2011. Othello, as a 'multicultural' play, makes a good basis for this experiment.

Our work is only conceivable in digital media, but our concept has no parallel in Digital Humanities as yet. It requires multi-disciplinary collaboration. Our team combines expertise in literature, languages and translation, text analysis, computational linguistics, data visualization, software development, interface design, and digital translation project management. Our larger aim is to make exploring highly complex, richly significant, multilingual text data practically feasible, instructive, enjoyable, and sociable. We use open source software to maximise social computational potential. We envisage online array users, in due course, not only exploring and analysing material we assemble, using built-in visualization tools, but sharing new data and interpretations, and developing their own analytic tools.

We foresee many applications of translation arrays, and of the software developed to build them: in research, education, creative industries, language industries, and cultural diplomacy.

However, our 'Phase 2' proposal here just focuses on building a first prototype for a single scene from Othello (1.3), in order to demonstrate the feasibility of the array concept, and in particular its potential value for monolingual (Anglophone) users, with a view to attracting further funding. We hope to launch a fully functioning array for Othello in multiple languages by April 2016, Shakespeare's 400th anniversary.

A successful prototype will confirm the plausibility of anticipated longer-term impacts. RDF work will develop the skillsets of team members, enhancing their experience, capacity for further work on this and related projects, and employability.

The longer-term project promises major impacts. If - as we propose - the Translation Array model is 'rolled out' to other content, impacts are potentially very far-reaching indeed: e.g., a Translation Array Bible. We envisage future collaboration with international cultural diplomacy organisations (British Council and national analogues; UNESCO).

A. Generic beneficiaries:

1) Language industries
Who:
Translation/localisation software and services companies, employees, freelancers
How:
A Translation Array constitutes a novel kind of translation memory with multiple target texts. It has applications in the refinement of statistical machine translation tools, such as Google Translate. Software enabling comparison of multiple translations has quality control applications. Translation arrays can be useful in advanced translator training - extension of both linguistic and cultural awareness.

2) Arts practices
Who:
Translators. Dramaturges, directors, actors. Artists/writers working in the areas of cultural memory, diversity, translation.
How:
Our work provides arts practitioners with very fertile resources, deepening and widening their awareness of alternative interpretations, and of the interpretability of those interpretations.

B. Shakespeare-specific beneficiaries:

3) UK cultural diplomacy
Who:
Shakespeare-specific UK organisations with international programmes of action (Shakespeare Birthplace Trust, International Shakespeare Association, Globe Theatre, Royal Shakespeare Company). Other UK-representative transnational organisations such as the British Council which have a stake in Shakespeare's global image.
How:
'Shakespeare' is global cultural currency, a key representative UK brand, and an important economic earning factor through exports and inward tourism (eg World Shakespeare Festival as flagship programme at Cultural Olympiad 2012). Representation of Shakespeare in a Translation Array framework manifests UK acknowledgement of the fact that his work circulates globally, now, in other languages as much as in English; and that (contrary to received ideas) the Anglophone world is curious about cross-language Shakespeare reception, and interested in learning both about and from 'others'.

4) Education
Who:
Secondary/tertiary teachers/students of both Shakespeare and languages/translation. Educational activities of cultural organisations promoting knowledge and understanding of Shakespeare's work (UK/worldwide). Translation arrays particularly suit multilingual classrooms and internationally networked classrooms. The national/regional Shakespeare associations worldwide will be approached in the next stage of our work as potential partners. They can assist greatly in prototype evaluation, data collection, and ongoing promotion and development of array resources.
How:
Shakespeare is studied and read, in English and in translations, by half the world's schoolchildren (www.worldshakespearefestival.org.uk/education). Modern English 'cribs' (in print and online) are widely used, though much criticized.
A Translation Array provides an alternative, inherently plural set of 'cribs', with authentic content, manifesting and making explorable the multiple interpretability of the original, and enabling collaboration. It promotes awareness of language diversity, cultural diversity, and historical cultural change; language learning; translation skills.