Microsoft has just announced that its researchers have developed, after a decade of work, with the help of Microsoft's Skype and Translator team, speech recognition software that can decode languages in real-time, called Skype Translator. This software, that will enable users previously barred by a language barrier to instantly communicate with one another, has only been tested from English to German and Chinese for the moment, by way of subtitles at video conferences.
This project connects with the utopian ideal of allowing humans to bridge geographic and language boundaries, in the global village, without resulting in the demise of the diversity of languages, as it once was before the Tower of Babel that separated men; such is still the dream of science fiction writers, like Gene Roddenberry, a power he gave to Starship Enterprise's Captain Kirk, on the Star Trek universe.
Indeed, what good will it really do us? That is hard to say: History tells us that men have always erred as to the real scope and function of the new communication technologies. Thus everyone thought, when they first appeared, that the printing press, gramophone, radio, television or computer would only entrench the powers of the strong of the time. This was never the case.
Today some also believe that simultaneous interpretation will be primarily used to empower the market economy, and in particular for commerce and playing online with people from around the world; that this will mostly benefit areas such as marketing, advertising, serving legal or criminal markets, and some preachers, doomsday terrorists, those attracting immense masses of faithful and followers, populist buffoons or politicians who will thus be able to escape the containment of their mother tongue, to send electric shock waves running through the entire planet.
Furthermore, assuming that it could really be implemented, such a technology will enclose every people in their own language, sharing their perspectives and experiences but through automatic voice translation. Languages will complement and nurture one another less; each language will freeze and die, or will evolve only within itself. And we will lose a major source of the awakening of intelligence, which is language learning and multi-lingualism. All that will delight the strong and powerful, whoever they may be.
Therefore, it will be a long time before such a technology will be available massively. First of all because it is not yet known how to translate automatically a written text in a convincing way, which presents a less difficult linguistic and cognitive challenge to address than that of simultaneous translation. And then because there would be a need, in principle, to translate on the fly in language pair the 8,000 existing languages, what would represent approximately 64 millions of possible combinations and therefore as many softwares. We will therefore without a doubt, if we are not careful, limit ourselves to just translate the main language pairs, there are about a hundred of them. And the languages that will not have such simultaneous interpretation softwares into or from the dominant languages will continue to disappear, as is the case currently, but this time at an ever increasing velocity.
Nonetheless, I firmly believe there is no point trying to oppose it. On the contrary, it should be encouraged, in particular from the French language; and into all languages even the rare or less common languages, and to do this we need to encourage the development of open source translation softwares that, as illustrated by Wikipedia, will lead to widespread development of translation softwares, more or less reliable, in translations involving rare languages. It will then be possible to get to know cultures with languages not widely spoken. The so-called "gravity" laws, (according to which a language makes a big impact on the economy when it is widely spoken), will no longer hold; any nation, however small it will be, thin on the ground as the speakers of its language will be, will be able to network with all the other countries and play an important economic, political and cultural role. This will finally contribute to the emergence of a myriad of new services, outside the realm of markets and politics, in order to teach, provide care, entertain, guide, advise, share, create, help, console and love.
In a world where the force of ideas will remain at least as powerful as arms, simultaneous automatic translation will be one day an essential part of an important power, the oldest one maybe, that of words.