The Skeptic’s Guide to Reading Speech-to-Speech Translation News
Click here to close
Click here to close
Subscribe here

The Skeptic’s Guide to Reading Speech-to-Speech Translation News

The Skeptic’s Guide to Reading Speech-to-Speech Translation News

Speech-to-Speech Translation NewsEvery time that one of the big firms in the translation technology space demonstrates some rather impressive innovation, the technology media makers restart the Universal Translator / Babel Fish fantasy machine.

“Windows 8.1 users will be among the first to get a crack at Skype’s Star Trek-style universal translator,” crows PC World.

“SpeechTrans Wristband Watch : A small step towards the Babel Fish,” pronounces Network World.

“Google’s ‘Babel Fish Mobile’ Universal Translator Still Alive,” reads Tom’s Guide.

Well, I know I am relieved to hear that the fish has not been beamed up into the hereafter.

Speculation about whether we have finally reached the point where technology has overpowered the limits of different languages and different locales is more fantastic than the reality.

I remain skeptical. Here’s why:

Those French-only Frenchmen aren’t speaking to the English-only Americans

You know how the joke goes: What do you call people who speak several languages? Polyglots. What do you call people who speak two languages? Bilingual. And what do you call people who speak just one language? Well, those people are Americans.

It is not at all a joking matter. In a controversial opinion piece published earlier this year in the Financial Times, Oxford- and Harvard-educated journalist Simon Kuper writes, “Learning another language? Don’t bother. … Today, the tourist who stammers a few hard-earned German words is often answered in cheery English. This is increasingly true worldwide. English today is ‘more widely spoken and written, than any language has ever been’, wrote Robert McCrum et al in The Story of English. In addition, smartphone apps can now translate speech on the spot. Learning German (say) just for basic conversation probably isn’t worth the effort any more.”

Notably, Mr. Kuper lives in Paris.

Speaking of France? You better do it in French then. The French may be more adept at learning other languages than their American counterparts, but they do not have a reputation for being altogether welcoming to foreigners. Foreign words, I mean.

Back in 1994, the Toubon law mandated that universities teach in French if they wanted support from the government. While it comes as no surprise that business, science, and tech suffered, it does surprise to know that the struggle to overturn the law was met with considerable opposition from some of the country’s oldest and most esteemed institutions, including the Académie Française. Geneviève Fioraso, who served as France’s minister for education and research at the time, sought to remove the law from the books, in support of a multilingual delivery of curriculum. For her efforts, she was accused of undermining the nation’s sovereignty.

Cheeky comments about the French and the Americans aside, the conflation of a nation’s primary language with patriotic/nationalistic ideals remains a considerable barrier to the ideals of multilingualism. There’s room for us all under that one big tent, right?

Maybe not. Chalk another one up to skepticism here.

Open sourced and collaborative? Not in this lifetime, buddy.

Despite the showering of praise on Microsoft’s latest speech-to-speech technology initiative, it is just one in a long-line of developments in the space. Remember when Facebook acquired Mobile Technologies, the developers behind the Jibbigo speech translation app? That fueled speculation in the wonders of translation technology too, especially because Jibbigo development had the previous investment of the U.S. Military in a collaboration with Carnegie Mellon University on voice-to-voice translation from Arabic into English. At the same time, AT&T announced that it was testing its own the real-time, cloud-computing-backed speech translation system. “It will sound like you,” they pledged.

If all of this sounds exciting, it is … but only when you remove competition from the picture. Bring it back in and you will recognize that these folks aren’t talking to each other but past each other. Indeed, the lack of cooperation in the speech recognition space is recognized as undermining its potential.

“The people working on this can’t even decide on an acronym,” wrote Slate in May.



Okay, maybe I am being harsh by framing these as insurmountable barriers. Maybe people will pull back the nationalism curtain just a bit to recognize a multilingual audience that really does want to communicate with each other.

In the meantime, keep your enthusiasm high … and your skeptic’s cap tight on your head.