The Pragmatic Translator and AI

The Pragmatic Translator and AI

Translation tantamount to a third eye and is an unequivalent to a medium between two languages, thus the Human Translator isn’t just a middleman that can easily be replaced with AI.

After internet-based machine translation came to surface in the early 2000s, many organizations became dependent upon AI. However, it is safe to say that this cost-effective solution couldn’t stand against the Human Translator for he is required to be highly pragmatic, especially in literary translation; a quality that Machine Translation couldn’t match till the present day.

 In the meantime, Arabic Translation significantly increased to cover new aspects of World Literature; masterpieces that remained foreign for far too long to the Arabic Language reader. Thanks to translation being a form of art in addition to it being a profession, a number of world-class Arab translators such as Sami Droubi provided great quality translation to Fyodor Dostoyevsky’s work from the Russian language, Saleh  Almani  who translated ‘One Hundred Years of Solitude’ written by G.G Marquez, and Ahmed Ali who translated many great works by different authors such as Paul Auster, Graham Swify and many others, bringing to the Arabic Language reader more than 36 new literary work.

 Such literary translation could never be done by machine translation for several reasons; no machine can study the author’s ideology in addition to her or his autobiography and previously translated work, let alone being able to see through the eyes of the author as if the translator is the one who wrote the original text and to pay great attention to details; mastering literary criticism and analyzing each and every hidden code mentioned in the literary work. This being said, “the translation world is not by any chance chaotic and isn’t subject to altitude but a strictly organized system where it is a must to respect the order of this complicated process.” Says Waleed Al-Farsheshee, a well known Arab translator and journalist, when asked about translation as a pragmatic process.

In the counterargument, it is AI that could decode complicated ancient language and translated it to us. Back in 1886. Sir Arthur Jhon Evans, an English archaeologist and pioneer in the study of Aegean Civilization in the Bronze Age had discovered a set of rather strange figures engraved in stones dated back to the year 1400B.c, he and a group of scientists had tried to decode the figures for years but for no avail.

 However, recently, linguistic researches provided huge databases and techniques with which AI could learn from and build upon; this allowed AI to give us a brand new way of language-thinking; opening the doors for a group of scientists from Massachusetts Institute of Technology and Google AI the programming of software that could decode the engraved figures mentioned previously.

 However, the method used didn’t match the standard AI translation. In this method, words became symmetrically connected and the key method laid in identifying such connections in a language which required a huge database of paragraphs where AI searches through and analyzes the number of word-by-word repetition. Such analysis creates what is called A “Simple Mathematical Formula” where a sentence is considered a consecutive scheme easing the process of determining the linguistic map of a language when compared to another.

Finally, understanding language and translating accordingly result in understanding history, explaining the present and predicting the near future where accuracy might stand side-by-side with style as the principal criterion for both the pragmatic translator and AI.









6. Waleed Al-Farsheshee’s interviews with Al-Arab Newspaper