A Story About the Past, Present & Future of Translation
Hi there! If you’re reading this, you’re probably interested in getting your message across in another language and are looking for the best way to do so.
You may have asked a polyglot or used translation companies before. Or you may, like me, be a frequent user of Google Translate.
But have you wondered how translation got to where it is today? What is a translation memory and why should you care? Why is translation so expensive when services like Google Translate are free? To answer these questions, and to explain why we created Honyaku Cloud, we may want to travel back in time.
55 BCE in Rome.
In the Western world, Saint Jerome is the patron saint of translators, known for his rendering of the scripture from Greek into Latin—what would later be known as the Vulgate—in 405 CE into a language the masses could understand. He advocated for sense-for-sense (communicative) translation as opposed to word-for-word (semantic)—or direct translation—cementing his position as a true saint of the ancient translation world.
Fast Forward to Medieval Europe.
Fast forward to 1439, when the Bible revolutionized the world of translation again. Gutenberg’s printing press democratized literacy and ushered in the era of mass communication. It is estimated that a single Renaissance printing press could produce 3,600 pages per workday, compared to only a few by hand-copying. Not only did the printing press make translation and copying faster, but it made it easier and cheaper, too.
16th Century Explosion of Literacy
With more and more literate populations in the 16th century across Europe and in England (Say hello to the Great Vowel Shift), there was a growing need for shared orthographic standards of pronunciation, spelling, and punctuation. One significant milestone was the publication of Samuel Johnson's Dictionary of the English Language in 1755. This would be followed less than a century later by American lexicographer Noah Webster, who published Webster’s American Dictionary of English Language in 1828, complete with alternate American spellings and the etymology for each word.
So how does this all relate back to translation? Well, translation requires consistency. A shared lexicon is needed not only to understand what you want to say but also to accurately convey your message to the intended audience, what people in the translation industry call a "termbase."
150 Years After Webster
Nearly 150 years after Webster, in 1976, a 19-year-old student from Japan was toying with computers to reinvent dictionary technology. With the help of professors at the University of California, Berkeley, now Softbank CEO and billionaire Masayoshi Son invented an electronic pocket translator that he sold to Sharp Corporation for $1.7 million, one of the first successful attempts of computer-assisted translation and one of the first successful electronic termbases.
Computing in the Late 20th Century Aids Productivity
With the rise of globalism and multinationals at the end of the 20th century, the need for translation expanded exponentially while the productivity of translators remained unchanged. This created an expensive and time-consuming bottleneck for companies and foreshadowed the need for better tools than Son’s electronic translator. The result: computer-assisted translation (CAT) tools and machine translation technology. These technologies have aided the productivity and consistency of individual translators since the 1990s, which take morpheme-level termbases and sentence-level translation memories and pair them with machine translation technologies.
Then Came the Internet.
Since former Google CEO Eric Schmidt coined the term “cloud computing” in 2006, technology has advanced rapidly, bringing us to where we are today. In terms of translation, cloud technology has made CAT tools lighter, faster, and more collaborative. Translation agencies are now able to crowdsource translations from people around the globe with the click of a button.
Machine translation Has Also Evolved.
The spread of the Internet made automated translation tools readily available. The phrase-based machine translation (PBMT) model calculates the similarity between phrases in the source and target languages according to a statistical model and rearranges them into the appropriate word order of the target language, much like a translation memory might. However, the substandard quality of PBMT translations remained largely unchanged. However, with recent developments to machine translation technology, such as neural machine translation (NMT), the quality of these translations is improving through use of neural networks designed to imitate the neurons of the human brain. Even so, algorithms are what ultimately dictate quality, which means these machines remain dogged by problems such as cultural context and precision, and you see more and more of these mistranslations in use today.
This was why we first developed Honyaku Cloud: to integrate computing and cloud technologies with specialist translators to deliver quality translation.
Curation + Computing
Honyaku Cloud combines recent developments in computing technologies with human curation and professional expertise. We automate what we can, which makes us faster, more agile, more organized, and more affordable than your usual translation agency. But we rely on people as much as technology.
We use CAT tools and refer to machine translations, but we also see translation as inherently creative and collaborative. We use cloud technology so multiple translators can work on a single translation simultaneously. We collaborate with clients in real-time, too, addressing feedback instantly to speed up the translation process. We work exclusively with particular clients to better understand and represent their brand, even going beyond translation and data management to help with visual and verbal identity.
Because we are an agile team with low overhead, we can afford to charge less, which means clients can afford to use our services without fretting over whether or not to hire someone new in-house. And best of all, we organize all of the data and give everything we translate on Honyaku Cloud back to the client in the form of easy-to-use and completely searchable bilingual databases. All data ends up where it rightfully belongs: in the hands of the client. No strings attached. Ever.