Found in Translation

By Raul Quinque

Wish I could read this. Cred­it: Janayugam arti­cle on Soft­ware Free­dom & lan­guage Com­put­ing by Ani­var Aravind, https://www.flickr.com/photos/technopolitrix/1414813791/in/photostream/

Full dis­clo­sure: I’ve writ­ten this blog on my behalf. Or in eigen­er Sache, as I would say in Ger­man. Is there a dif­fer­ence between the two expres­sions? Worlds, I would say, as some­one who is pre­oc­cu­pied with lan­guage trans­la­tion most days of the week. In light of the rapid pro­lif­er­a­tion and evo­lu­tion of machine-learn­ing trans­la­tors, here’s to what makes think­ing, speak­ing, and writ­ing with diverse lan­guages invalu­ably human.

My teacher once told me that I write in nom­i­nal style. Basi­cal­ly, I write Eng­lish in a dis­tinct­ly Ger­man way. It was a sem­i­nal les­son for sure and has affect­ed every sen­tence I’ve writ­ten since from con­cep­tion to revi­sion. How cool is it that I find my own per­son­al expres­sion some­where in this Ger­man-Eng­lish mix? I’m hon­est­ly not sure if this enthu­si­asm trans­lates well. Maybe I sound like a nerd going on about their hob­by horse. (Case in point: I just dis­cov­ered that the Ger­man expres­sion Steck­enpferd trans­lates direct­ly to English.)

Appar­ent­ly, what makes trans­la­tions tick doesn’t depend on any kind of exper­tise. We all do lan­guage, all the time. It’s pos­si­ble to learn for­eign lan­guages just by imi­tat­ing them or by div­ing into lin­guis­tic detail. Every act of trans­la­tion is impres­sive and pro­duc­tive in its own right. Using lan­guage is like walk­ing through a door, lead­ing to dif­fer­ent modes of knowl­edge and con­scious­ness. We can nev­er live some­one else’s expe­ri­ence, but some­thing as sim­ple yet mean­ing­ful as an idiom offers a glimpse into oth­er ways of think­ing. It’s the clos­est we can get to see­ing the world through oth­er peo­ples’ eyes.

The run­ner-up for my Bachelor’s stud­ies was Trans­la­tion. Dig­i­tal Media won out in the end, but I’ve nev­er wavered in my pas­sion for lan­guage. Dig­i­tal trans­la­tors are an evi­dent link between the two top­ics, trans­la­tors that are evolv­ing as fast as any­thing dri­ven by machine learn­ing. In pop­u­lar media, we are often con­front­ed with the more sen­sa­tion­al but in this case much less spe­cif­ic and gen­er­al­ly more enig­mat­ic term Arti­fi­cial Intel­li­gence, AI.

Ram­i­fi­ca­tions from the increas­ing avail­abil­i­ty of machine learn­ing for tasks, such as trans­la­tion, are ambigu­ous. And both, absolute idol­iza­tion and con­dem­na­tion, miss out on what’s real­ly at stake. If there’s one les­son I’d love to share, it’s this one: AI is nei­ther human nor inhu­man. When it trans­lates, it too cre­ates a world based on its own par­tial­ly informed world­view. A dig­i­tal trans­la­tor can be our imper­fect cre­ative part­ner in writ­ing and trans­la­tion, but we shouldn’t leave all the imag­i­na­tive pow­er to it. Every for­mal lin­guis­tic mis­take, every chance dis­cov­ery of a new phrase, embraces anoth­er voice found in translation.

 2,273 Total Views,  2 Views Today