Since I first started interpreting in my earliest youth, people would ask me whether I would soon be replaced by a machine or computer program. And yes, computer-assisted translation and glossaries are vitally important. I use them regularly. But I don’t expect to be completely replaced during my professional lifetime. And I have a theory on it, based mostly on my ignorance, and my imagination, two things that computers are not conscious of having, by the way.
My theory is that to the degree that what we are asking a computer to do requires cultural context and filtering of meaning, focusing on areas of knowledge, and discarding what doesn’t feel right or seem right, the human mind will still be needed. Can a machine hold glossary terms to a larger degree than a flawed and aging human? No doubt. And the machine will not forget items. Which I reliably will, based on infrequency of use.
To the degree that massive amounts of linguistic data are input, a program such as Google Translate can get you a pretty good facsimile of whatever you search for the major language pairs. The way they do this is through what is now called corpus linguistics (body of language) where data is culled and processed from real-world documents, then analyzed in a myriad of ways, including glossary and translation formation. But the human mind must still look at it and decide if it fits the context, makes sense, and seems right. And sometimes override the offered translation.
Case in point where we needed both human and computer assistance, and then this human to override the computer. I was assisting a new patient in filling out her medical history. She was telling me that she had the little bell removed, you know, that thingie that hangs down in your throat, not your tonsils, but the thingie right in the middle, the little bell. I told her I knew exactly what she was talking about but for some reason, I could not think of the actual word in either language and I had to think about it for a moment. It is not a very talked-about body part, as body parts go.
She told me not to worry about it because she uses Google translate all the time and she can find it in a second. She then shows me her phone and says, “Here it is. I had my ‘Tinker Bell’ removed!”
I let her know that Tinker Bell in a fairy-like character in a famous story over a hundred years old called Peter Pan about a boy who never grows up and it is now a movie as well. Tinker Bell is so named because she is a tinker, a person who repairs old pots and pans, in the fairy world, and when she talks it sounds like ringing bells so only magic people can understand her. She can also give people the ability to fly. And she is so tiny that she can only hold one emotion at a time, which I believe may have been the author’s reflection upon the smallness of his own neglectful mother’s personality. But Tinker Bell is decidedly not a body part.
The patient had seen the children’s movie and we started giggling together and all of a sudden from the back of my mind, I called out, “uvula!” And uvula is the actual formal name of this body part in both languages. Then we started joking about what the doctor would think if we had walked into the pre-surgery appointment and told her that “somebody took my Tinker Bell!” The doctor would have questioned either my abilities or the patient’s connection to our shared reality.
If the word had not come to me spontaneously, I could have done a deeper internet search on the word the patient was using, and likely found uvula. I could have looked up throat anatomy in the target language, and found the body part that way. Computer research capabilities are almost endless. I could also, if pushed to the wall, have interpreted for the doctor that the patient has had “the little bell” that dangles in her throat removed, and the doctor would likely have looked at me quizzically and offered “uvula” to us. But I feel confident in claiming with certitude that the doctor would never have suggested Tinker Bell, as Google Translate did.
So yes, as Wikipedia likes to term it, disambiguation is a valuable part of our language processing. Being able to access glossaries such as Linguee that are based on terms translated in context culled from millions of existing published translations is hugely important. But so is knowing the context and even the cultural background of the speech communities in which we interpret. Because with all the data in the world, our human judgment, and our human ability to create meaning and to communicate it, have not yet been replaced. And this makes me happy.