I’ve tried to raise my kids in an environment of learning foreign languages, since I’ve seen that (especially Spanish) as something that could be extremely helpful to them in the 21st century job market. But then it occurred to me, as software like Google Translate seems to get better and better (though still far from perfect), that at some point we’re going to have the Star Trek style Universal Translator, where people can talk to each other and have it automatically translated, UN style, into their headphones (or earplugs, or whatever). At that point, I suppose actually knowing someone’s language will become a lot less important.
Are we still many decades away from this reality? Or will it make my middle schooler’s foreign language ability redundant and thus obsolete by the time he’s out of grad school?
Never, I hope, as (I believe) learning a foreign language broadens the mind and induces a more cosmopolitan and diverse perspective. Even if it became functionally useless I’d want people to still do it.
There really isn’t a better way to understand a different culture like learning a language. Even if they never use the skill, it’ll add depth to learning about any culture not their own.
Before technology translated everything for us, the whole world was heading towards a common language. Now, it’s the other way around. Although most people who don’t speak English will still want to learn it as the language of business, they are unlikely to forsake their traditional tongue now. So I think the need to learn languages for practical purposes is actually increasing in a few ways. As mentioned above, it also is good exercise for the brain.
Tough to say for sure how far away we are from really good machine translation. It wouldn’t surprise me at all if it’s decades.
Even if we had it tomorrow, there are benefits to knowing other languages beyond being able to communicate with people who don’t speak your native tongue. Speaking another language lets you experience culture much differently than speaking through a machine. The guy with the machine is going to be on the outside looking in.
I learned English as a child by immersion, and outside a home where only Spanish was spoken. I still think in Spanish, and use both languages interchangeably. As I recall the road to learning English was not an easy one, and I remember pretending I could understand words that were frighteningly difficult for me at the time. But the effort was worth it, and mastering a different language, as a child, added to my self esteem, and I think helped me along with higher education.
Machine translation, for some things, has gotten better. For others, it’s still hit and miss. Just this morning, I used machine translation to check a tax document. I’d say that about 40% of the resulting translation was very good, about 20% good. The entire document, though, required human perusal and correction.
What I never appreciated as a child being force fed French and Spanish is how much it teaches about the usage and construction of my own language. It would have been very helpful for me to have understood it was about communication and how to use English effectively.
Even if one were to have an ap that could translate on the fly tomorrow, what do you do when the batteries die out? Limiting all interactions to less than a handful of hours or withing range of an electrical socket would kind of suck, no?
I agree. That’s something that a translator just can’t do. There is so much nuance in language and while translators have the uses, we’re nowhere near the point that they can replace learning the language.
I don’t think it will ever be pointless, but it might become less vital in the future for pure communication. I agree with others that say it will always be a good way to learn another culture.
Based on my own translation app (on my iPhone and iPad), we aren’t there yet, but it’s getting closer and closer. When I used to travel outside of the US I’d often bone up on the language of where ever I was going, so that I’d at least have the basics. If I was going to be in country for a while I’d actually get the Rosetta Stone course for the language. These days, while I don’t travel outside of the US nearly as much, I just get an app on my iPhone, once that takes my spoke English into whatever language they speak in the country I’m in. It works pretty well, even though sometimes what I say (or vice versa) get’s translated in hilarious ways. Thus far I haven’t had anyone offended or angry that I use a phone to translate for me…in fact, I generally have gotten a pretty positive response.
What I think this will do, eventually, is allow us all to communicate more freely with a larger percentage of folks (many of who just aren’t going to invest the time in learning a language other than their own), but the cool part, IMHO, is that people will hold onto their own languages and dialects. And it will never be ‘pointless’ to learn some of them.
One interesting thought about universal translators – they ENABLE language learning. Why? Because you can switch your translator on “Japanese” and suddenly you have full immersion learning (well, for the verbal component at least).
Also, if there’s a universal translator you still need, y’know, a programmer and translator and whatnot, so SOMEBODY at least has to learn foreign languages (or at the very least linguistics).
Learning a language does not just gives you a new means of communicating, but also a new way of understanding the world. A good explanation of this is in theGeography of Thought. There is ample evidence that language can fundamentally change how we perceive and interact with the world. In order to really fully understand another culture, you probably need to be familiar with their language.
Learning another language, of course, also leads to a new awareness of your own language, both in mechanics and usage.
Finally, there will always be some languages that are not going to make the translation scheme. There are hundreds of languages right now that don’t even have their own dictionary. These languages tend to be fairly difficult languages to learn. it’s a lot easier to learn your third or fourth language than your second one. So if people stop learning Spanish and French, they probably going to stop learning Quechua and Hausa. This could have huge effects- essentially, we could see the development of language haves and have-nots. One portion of people will be able to interact internationally while their own languages are probably evolving more and more apart (making learning new languages even harder), meanwhile, the have-nots are left unable to communicate with much of anybody. The minority languages where people have access to education will probably die off, while the speakers of minority languages in other areas will probably become completely marginalized, at a huge disadvantage of communicating with the rest of the world.
I’m sure you know this, and were just using the UN reference to explain your idea, but the ‘automatic translation’ provided at the UN is provided by simultaneous interpreters working in teams of 2 or 3, who trade off every 15 minutes or so due to the high level of concentration required.
And as hard as it is to simultaneously interpret, listening to simultaneous interpretation ain’t no picnic either! It definitely gets the message across, and if the interpreters are given the material to be covered in advance, it will sound pretty smooth. But it’s not automatic, and it’s not instant. It’s anywhere from half a sentence to 2-3 sentences behind the speaker.
It is gratifying to be able to hold a conversation with someone with whom you don’t share a common language–and the smoother it is, the more fun it is–for both the speaker AND the interpreters!
It doesn’t always get better, let’s not overlook the fact that some popular translation services such as Google’s base their translations on bilingual text found online. Since more and more of the bilingual text on websites comes from automatic translation, the greater proportion of garbage being put in will lead to a greater proportion of garbage coming out.
I agree that learning foreign languages has its own inherent cultural and cognitive benefits. Still (and maybe this reflects the economic anxiety in the zeitgeist right now) I will be a little frustrated if my kids are all fluent in Spanish in a 21st century world where the Latino population has skyrocketed, but they don’t have a significant leg up on other jobseekers because everyone’s smart phone can easily and accurately translate for them.
I guess it’s a bit like how I felt when spellcheckers were invented. I was always a perfect speller growing up, and I thought that would always give me an advantage over others–so I found it a bit annoying when this skill became significantly less advantageous.
I imagine that people who are truly bilingual will still have an advantage in understanding contextual and cultural subtleties, untranslatable idioms, etc. But I can imagine a future in which someone with what would now be considered a very solid second language ability would still not be able to communicate as well without a translation app as they could with one. In that world, the economic advantage of years of study of that second language is pretty well negated, even if there are other advantages. And can we be sure those years of study and thousands of hours of practice might not have been better spent studying/practicing something else, like art history or tennis or organic gardening?
That depends, this is getting into magical future land territory though, be warned. I think that if machine learning and linguistics both ever become advanced enough, it may be possible to “train” the translation machine. This means that rather than have a team of scientists program the individual rules for each language, they could have a set of algorithms that, as exposed to a new language (utilizing vision, referential intention, inference etc) can piece together what it means, and then translate it into (and from) other languages it already knows. This would allow linguistic diversity to thrive and be cataloged at the same time.
Of course, we’re really far away from any technology like that, but as we make advances in machine learning and linguistics I think it’s certainly possible, some day. Maybe not for 100+ years, but possible.
We’ll develop translators based on word definition long before that. If a language is worth keeping, all it takes is someone who speaks it to spend the time with the linquist/programmer.
It definitely gets the message across, and if the interpreters are given the material to be covered in advance, it will sound pretty smooth. But it’s not automatic, and it’s not instant. It’s anywhere from half a sentence to 2-3 sentences behind the speaker.