TAUS Review#4 - The Innovation Issue - July 2015

Page 18

The Language Perspective by Nicholas Ostler

New Lamps for Old Offering

to make sense of someone else’s language

is an old profession.

A word for interpreter, still current in Turkish (tercüman = archaic English dragoman), is first attested in Assyrian of 1800 BC (targumannum); while its synonym in Sumerian emebal goes back a further 500 years. (The first known Assyrian interpreter worked with Hurrian, in what is now south-eastern Turkey, while the Sumerian one dealt in the language of Meluhha, to the south, possibly the Indus Valley1. So to innovate in this business will be a testing ambition. But success is most likely to be achieved if the focus is not on means, but on ends. Machine translation, as developed and applied, has mainly been for decoding use: that is to say, the translator is trying to gain knowledge from a foreign source. This is one reason why poor-quality text is sometimes acceptable as an output: at least it will be a much less foreign text than the input, and possibly accessible to the user’s common sense. Encoding uses – which attempt to make the knowledge in a familiar text accessible to foreigners – are much less tractable: why should the target audience, who did not choose to have the text translated, bother with an imperfect product? As a result, the audience best served is the one that uses the regnant lingua franca, in our own era overwhelmingly English, though the same argument would also favour translation into other big-beast language communities, who can plausibly see the world accessible in their language as world enough. Above all, the availability of such translation spares them the massive effort of conscious language learning.

Machine translation, as developed and applied, has mainly been for decoding use.

1 I.J. Gelb, The word for Dragoman in the Ancient New East, Glotta 2:1 (1968)

18

Language learning is another field fit for technical support (from conventional grammars, dictionaries and tutors through to CALL). But the effort is only repaid if the student foresees massive use of the new skill: this means that once again the regnant lingua franca will be the preferred goal, providing lots of hitherto inaccesible texts (and people!). Speakers of minority languages will aim to widen their scope by joining the second-language learners of this (already vast) language community. So both these inter-lingual technologies – translation and learning – tend to favour the regnant lingua franca as their target language. The incidence and level of costs is different but in both cases, the implicit aim is to make the “foreign language problem” go away: wider communication is to be achieved, largely by adding to the effective reach of the language that is already dominant. To the languages that have speakers and writers worth attention, more

The audience best served is the one that uses the regnant lingua franca.


Issuu converts static files into: digital portfolios, online yearbooks, online catalogs, digital photo albums and more. Sign up and create your flipbook.