No more language barriers?

Ten years ago, I used Babelfish language translation once in a blue moon and often for entertainment. For the last six months or so I’ve been using Google’s translation tool on almost daily basis. (I read a French and a Japanese blog as well as an Italian magazine.)

The prospect of cheap portable translation tools that are contextually savvy is thrilling. What it will do to culture, I can’t say. But these comments following a New York Times story on Google’s translation breakthroughs may illuminate the next few years.


My friend RC replies:

Interestingly, I think the advent of these types of technology for translating foreign text will make it all the more impressive when people are able to speak and understand foreign languages without the assistance of programs or gadgets.

It’s probably making a fetish of something or other, but I do think that as we move into a world where people are able to use programs to translate texts (and inevitably speech), it’ll be only the most committed who will take the effort to try and master other languages — without assistance. The upside in the advent of these technologies is self-evident (i.e., access to information we wouldn’t otherwise have). The downside, I’d say, is that it may discourage people from trying to learn these foreign languages that they can have instantly translated. Learning other languages can, not to be too Whorfian about it, change the way you think, and the way you perceive the world.

In a way, these programs and gadgets, while ostensibly expanding the material we have access to, also limit and confine us in our native language, as everything is automatically converted into a language in which we are comfortable, and we no longer really need to venture out into a language in which we are not so comfortable, in which the concepts are completely different (and tied to the language, untranslatable), etc. The technology, I fear, even as it opens doors, will also, in a way, lock us in.

But the downsides I’m pointing to are plainly outweighed by the upsides. One does not need to learn Russian to appreciate Tolstoy (though it would probably heighten and enrich one’s appreciation), and we English speakers are happy to have access to translated Tolstoy.

And putting aside for the moment the spectrum of translatability (running from Car User’s Manual (easily translatable) to poetry (less so)).

And, I shouldn’t overlook the potential to help people learn languages that some of these tools will probably have (somewhat like reading Wittgenstein with one page in German and the facing page in English). Though I do think as these technologies develop, there won’t even be the opportunity to compare. E.g., if your default language is English on your browser, say, all sites, regardless of the original language they were written in, will immediately be presented to you in English, and you won’t have the opportunity to compare the English translation to the original text.

Perhaps with devices that will be able to translate speech, the listener will be able to compare the translation to what is being spoken in the foreign language (as a moviegoer might with subtitles). But I think it’s depressing to imagine people going to Italy, Japan, Peru or wherever and just toting along their Universal Translator app instead of trying to learn a new language.

See also calculators and math.

To which I responded: Thanks. I didn’t want to focus on the negative implications because they are many.

Language is a culture’s most deeply rooted reason. Languages underpin or re-enforce religion, morality, art, science, etc. They embody a people’s inherited understanding of the world, their historical struggles and accomplishments, their sense of the possible and the necessary.

While the statistical model discussed in the piece suggests translations that are appropriate vis a vis the “common wisdom”, the best case scenario is one where multiple translations are provided so that the nuances of meaning are made available to the non-speaker.

That kind of translation, where multiple options are provided, is inappropriate for the kinds of scenarios which will most likely favor software translation. Most medical, military, commercial and pop culture exchanges wouldn’t benefit from the added friction of users having to weigh sometimes arcane or parochial variations.

But, yes, as with poetry, there will be many instances where “most people mean X when they say Y” won’t cut it. Psychiatry and hardball negotiations come to mind.

Second update

Hours later, I wrote again to RC: I worry that my response to your response wasn’t very direct and thus unfair. I think I was trying to balance out my original post with some of the negative implications you described and/or hinted at. So, more directly: I think technology will allow for more responsible translations but not all scenarios will require this full “toolkit.” (Not every photographer needs Photoshop, etc.)

I completely agree that language structures the way you think and is thus a good reason to learn multiple languages. But, perhaps, we’re both circling around another possibility: that interacting with *people* who speak another language can also change the way you see the world – even if you’re talking through an interpreter, living or simulated.

Another possible future scenario is everyone speaking just one language, say English. Now *that* would be bad. (cf. language extinction.) Perhaps, allowing users to speak in their native tongue while still making themselves understood to non-speakers could be a boon for diversity.