“One machine can do the work of 50 ordinary men. No machine can do the work of one extraordinary man.” (Elbert Hubbard, American writer, 1856 to 1915)
The robots are coming! The robots are coming!
Or so the media keep telling us. In jobs ranging from assembly line worker to lawyer, robots are starting to outperform humans and will soon replace us. And yes, translators are among those on the robot hit list. But does that mean we’ll all be out begging for bread in a few years?
I’m not so sure.
The leisurely evolution of machine translation
Back in the 1950s, some of the first computer scientists looked at the code-breaking successes of the Second World War and thought that those techniques could be applied to translation. Languages are just another type of code, they reasoned. These scientists figured they could find the Holy Grail (the Babel fish or the universal translation machine) within a few years, and no one would ever need an interpreter or translator again.
70 years later, they’re still working on it.
But that doesn’t mean nothing has changed in the world of translation because of their work. Indeed, almost everything has changed. Computers and then the Internet have revolutionized the field, increasing both the amount a human can translate and the quality of the translations.
A before and after view
Let’s do a quick comparison of translation in a government context before and after the computer revolution. The first column describes the step in the translation process. The middle column shows how things used to work, based on what I’ve heard from my … how shall I put this delicately? … more experienced colleagues. The last column is current practice.
Step in the process | Stone Age | Silicon Age |
---|---|---|
Delivery of document to the translation division | The client sends a typed paper document by internal mail to the translation division. | The client uploads a digital document to the translation division’s server or submits it by email. |
Storage and consultation of previous versions of the same document | The translator opens his or her filing cabinet to find similar texts or previous versions of the same document and props them up on a document holder for consultation. | A computer program searches the document text and compares it to a database of previous translations. Exact matches can be automatically inserted into the translation. The translator can consult the database and see the texts aligned side-by-side. |
Use of dictionaries | The translator looks up words in paper dictionaries. | The translator types words into digital dictionaries. |
Use of terminology banks | The translator consults a terminology bank by thumbing through a card catalogue. | The translator searches for terms in an online terminology database. |
Consultation of reference materials | The translator visits a reference library to consult a limited selection of authoritative sources and comparable texts. | The translator uses the Internet to access an almost infinite range of information and comparable texts. |
Revision | The translator types out a draft translation on a typewriter and revises it on paper. | The translator types out a translation in a word processor and revises it onscreen. |
Preparation of the final version | The marked-up paper copy is given to a typist, who types up the final version. | The final version is saved digitally. |
Delivery of translation to client | The finished translation is returned to the client by internal mail. | The client is notified that the translation is ready and can download the file from the translation division’s server or receive it by email. |
The advantages for translators today
This list illustrates only a few of the many computerized tools that translators use in their work. I like to think of translation and technology interacting on a continuum from 0 (pen and paper) to 10 (fully automatic machine translation). The greatest achievement of the technological progress of recent decades may simply be the number of positions along that continuum that translators can take. With so many options, we can use the technology that works best for us or for a given assignment.
So, while the pursuit of the machine translation Holy Grail gets all the headlines, today’s technology is most often put in service of a human translator, who can make use of the proven advances to produce a better product in less time.
In short, the robots are coming, but quite slowly, and our experience tells us they’re friendly.
What about you? What do you think is the greatest benefit (or the greatest drawback) of language technology in your life? Share your opinion in a comment.