In the beginning, when the Internet moved at the slow pace of 56k modems, and computers were still not very powerful with only Pentium III’s onboard, the dream of having an automatic translator seemed a science fiction dream, only possible in movies and books. As a matter of fact, the author Douglas Adams invented a universal translator for his 1979 book “A Hitchhiker’s Guide to the Galaxy.” In his book, space travelers needed only to place a tiny fish in their ear to hear any language in the known universe translated directly into their brain. The little fish was called a Babelfish.
Fast forward to the early Nineteen-nineties, and the website Altavista became one of the very first websites to offer a free online multi-language automatic translator. It was called “Babelfish,” and it’s still available for use online. In this website you can enter a word or a phrase and ask the website to translate it into your language of choice, from a list of available languages. For the most part, it works pretty well, but sometimes there are many errors because of the context or because of how idiomatic your phrase might be. However, it still took a lot of time to turn on your computer, open a browser, go to the website and then type all the information you wanted translated. Wouldn’t it be wonderful if you could have a real-time automatic translator? Well, that still seemed a science fiction dream…
Except not. Fast forward again to the last years of the first decade of the Twenty-First Century, when, with much fanfare, Google announced that shortly they would be able to add voice recognition to their own version of their free online automatic translator “Google Translate.” For a few years Google had been developing their online automatic translator, and it grew from being a website similar to Altavista’s Babelfish to being a plug-in for any web browser (a plug-in is an auxiliary program that runs additional functionality to a web browser; for example, the search bar on the tool bar of the browser is a plug-in). Google Translate was even able to translate whole web pages from one language to another with a few simple clicks of the mouse. But soon enough the voice recognition programs were powerful enough to understand most anybody, and they were incorporated to the Translate application. The result was a website in which you pressed a button and recorded a phrase, and the program would automatically translate what was said and spoke the translated text back to you!
Fast forward again to a few years later. The proliferation of “smart phones” and ultra-fast wireless Internet connections made possible for anyone who owned one of these devices to connect remotely to the Internet. People were able to install “apps” into their phones, which allowed them to use their phone as a portable computer. One of the apps available is, of course, Google Translate. Now anyone can have an automatic translator in the palm of their hand. And that is not the only automatic translator available. There are others, with varying degrees of dependability and accuracy. Of several I have seen, the better ones seem to be Google Translate and iTranslate (except this last one offers only limited functionality for free, and full service for a fee.)
But how reliable or useful are these automatic translators in the medical field? Now, that’s the $64,000 question. As an interpreter, I can tell you that all those programs are not only impressive in their technological prowess, but also in their high level of accuracy. Long gone are the days when all automatic translators made laughable mistakes. The way they work is that, in several buildings choke-full of servers, at several locations in different countries, there are vast and extensive databases of terminology available for these programs to be able to translate not only automatically, but also in real-time. However, they do depend a lot on the skill of the user. For example, if you are not familiar with the more standard language of your native tongue and rely too much on idiomatic expressions, the error factor in the translation is very large. And sometimes you have to know more or less what the grammatical structure of the foreign language would allow before you submit your phrase to be translated. If you cannot formulate your phrases accordingly, you will get the wrong translation. A pretty straight forward example is the simple phrase, “when he ate last?” (when talking about a patient.) Both Google Translate and iTranslate make valiant efforts in translating this phrase, with varying results which could be usable, but will still be mostly incorrect. A human interpreter might never have difficulties with this phrase, because he understands the context of the phrase. But a machine, as powerful and efficient as it might be, still depends on the user to give it sufficient information for it to perform correctly. In this case, the user might be better served if he were to ask, “at what time did he eat the last meal?” This phrase seems cumbersome and stilted, but it would yield a better automatic translation into Spanish than the more idiomatic and common phrase.
I imagine that, at the pace in which technology advances, in a few years all these problems of context might be resolved and the accuracy of automatic translators might start to rival that of human interpreters.
But even then, there will be many other considerations, such as cultural brokerage or advocacy, which cannot be downloaded in an app.