#5 The prudence of translation

Translations have always been a little questionable, so I guess it's only prudent (haha) to dedicate this entry attempting to solve the central question of whether, if at all, it's appropriate to translate languages.
I ran into quite an opinion piece here:
http://lfpress.ca/newsstand/News/Columnists/Gillespie_Ian/2007/10/13/4572738-sun.html

The columnist here offers a "quick fix" to the language barrier between the Peruvian oil companies and the Amazonians. He claims in his solution that to befriend the native Amazonians (and we all know that sharing a common language is almost a direct segway into friendship and intimacy) the Peruvian oil companies should hire a U.S. company to create a "language guide" that would be comprehensible to the natives. Already, I see several things wrong with this concept.
First, language is being used as a facilitator of communication, and not THE channel of communication. I wonder if in a way, the Peruvian oil companies can take adavantage of this langauge barrier and exploit the Amazonians and their resources, and then reassign responsibility by claiming that there was a misunderstanding between the capitalists and the natives. Also, why is the intermediate US company (third party) even necessary in this process. For a more intimate connection, shouldn't the companies take time to practice cultural immersion (like this author did here: http://www.todayszaman.com/tz-web/detaylar.do?load=detay&link=124576&bolum=132, if not as a way to thank the Amazonians for all they will take, then at least for some shared ground of communication?) and actually learn the language and the cultural implications to minimize damage to both parties?
The columnist then goes on to state that by simply translating a few phrases that are purportedly "native" in sound and concept (ie, "How many moons or suns have you walked for," or "We come in peace and are people just like you") the companies will simply gain entrance into Amazonian society. First off, even if these questions were culturally valid, the mere fact that an unaboriginal individual speaking it, and doing so not from innate knowledge but in a stilted form with it's source in a foreign creation (the language guide) will create an immediate sense of inauthenticity and general falsity.
So we ask ourselves, does TRULY learning a language require consequent cultural immersion, like the author of the second article did in Turkey, or is it even marginally feasible to learn a language, be accepted by the natives of that language, without undergoing cultural immersion. Or, would the ideal solution be to develop a "universal language" that transcends all these barriers and provide an immediate forum of interaction for all societies. But it that's the case, can individual cultures still exist, or will there be more of a global culture, which is a stir-fry of every culture?
This is a little remniscent of the internet culture that's developing: it transcends cultural implications and brings everyone together in a global forum with a shared language. This however, seems to be the closest to an universal language, which in truth, isn't really a language at all.