Wednesday, February 06, 2008

Lost in Translation—Men, Ferengi, Countrymen, Lend Me Your Lobes

Little Green Men” (DS9) gives us insight into the Ferengi Universal Translator, and in the process illustrates some of the logistical problems with any such device. According to this episode, Ferengi have their UTs implanted in their ears. This echoes the Babel Fish of the Hitchhiker’s Guide to the Galaxy. (There’s a slightly more serious, in-ear device in the Michael Crichton novel Timeline, although it didn’t make it into the movie.)

There’s a logic to the in-ear approach that we don’t have with other incarnations of the Star Trek UT. Consider the scene in “The 37s” (VOY). In that scene the revived Japanese man from 1937 marvels that he is hearing everyone speak Japanese. How? Because Janeway and Kes are wearing little gold pins? How is it that there is no cacophony of overlapping voices speaking English, Japanese, Ocampa, and whatever other languages may have been represented? This is a basic flaw in the presentation of the Universal Translator.

On the other hand, having the device out of the way, in your ear, would allow you to hear the translation you need without broadcasting it to everyone around you. This is the basic logic behind the Klingon walkie-talkies that Kirk and McCoy had in Star Trek:VI, or the earphones delegates use at the United Nations.

So the Ferengi really seem to be onto something here. But there’s one major difference between the Babel Fish and the Ferengi UT. The Babel Fish lets you understand other people, but it doesn’t let other people understand you. They would need a Babel Fish of their own. In theory, the Ferengi UT should let Quark, Rom, and Nog understand the hu-mons but not the other way around.

But that’s not what happens at all. The Ferengi UT somehow let’s everyone in the room understand everyone in the room. So we’re back to where we started. How do you make multiple overlapping voices make any sense to the listener? Either these little devices must have incredible acoustic control—able to project specific sound waves to specific locations around them—or they are somehow projecting the translation directly into the person’s mind, creating the illusion of hearing a single voice (and maybe even lip-synching while they're at it).

However, we quickly run into the same problem here that we did with scanning brain activity: if the UT we’re really that powerful it could be used in all manner of unscrupulous ways. If you can flawlessly project sounds and visual stimulation into someone’s mind you could . . . fill in the blank with whatever devious plan you can think of.

So the basic problem remains unaddressed. Once we overcome the challenges of creating a device that can translate, we have to find a way to provide each listener with the translation they need. Ship to ship communication is one thing, since you’re already listening though a communication channel that could be modified. But for face to face communication, no solution has been offered.

Next: Lost in Translation - Um . . . Disfluencies and Translation

Sunday, February 03, 2008

Lost in Translation—"The Universal Translator Would Be Recognized"

"We must respond personally. The Universal Translator would be recognized." - Pavel Chekov

There is no doubt that this scene is mostly played for laughs, and seeing Starfleet officers frantically searching through books and speaking Klingon badly is an entertaining sight. But it's problematic if you think it through. At the same time, Chekov's statement highlights a significant issue for the Universal Translator.

First, some of what makes this scene difficult to accept. To start with - and this is really just nitpicking - why does Chekov get the disembodied voice over about the UT? Wouldn't that be Uhura's line?

Would a Starfleet vessel really have a library of books about the Klingon language? Would Uhura really speak Klingon like a gringo? Hoshi Sato was probably spinning in her grave.The Klingon translators at Kirk and McCoy's trial speak better English than she speaks Klingon. (Supposedly Nichelle Nichols objected to the scene because she believed Uhura would have been able to speak better Klingon.)

So there are some oddities here we just have to be willing to accept for the sake of the gag, but we have to ask, how could this little ruse have been at all convincing? If the Universal Translator would be recognized, what about Uhura's poorly spoken Klingon? Bernd Schneider's article "Alien Monocultures in Star Trek" offers an intriguing possibility. It suggests that not all Klingons necessarily share the same native language and continues: "Maybe this was even the reason why Uhura's awkward attempt to speak Klingon in 'Star Trek: The Undiscovered Country' did not raise any suspicion."

In any event, what does Chekov's statement mean? In what sense would the Universal Translator be recognizable? Evidently he wasn't talking about not replying with proper pronunciation and grammar, since they didn't do that anyways. And he evidently wasn't worried about the delay in translation, since their human efforts to translate certainly took longer than the UT would have taken. So it must primarily have been a concern over the quality of the synthesized voice.

We might assume that the quality of the UT's voice would improve over time, as the technology improves in general and as it became more familiar with a given language. The question becomes, would a 23rd century UT really have such trouble with Klingon when there had already been extensive contact between these two races? By this time there were individual Klingons that were fluent in English. They had an ambassador that could address the Federation Council and the Federation President. They sent a whole group of delegates to the Khitomer peace conference.

But the problem of the Universal Translator being recognizable over an audio channel only multiplies with face to face communication. The obstacles we have identified so far would only compound: Beyond the basic challenge of providing an accurate and nuanced translation and dealing with a potential artificial sounding voice, there would necessarily be a delay and lip-syncing issues.

'The Universal Translator would be recognized' indeed.

So how is that Starfleet officers - even of Archer's time - could seamlessly intermingle with alien populations and not be picked out as impostors?

Even if we accept the seemingly magical ability of the UT to learn alien languages instantly and basically perfectly, such a device would always be recognizable while in use. Storytelling convention allows us to gloss over this for the most part. We don't have to be subjected to the delays, the strange sounding voices, the lip-syncing issues every time. But then by slight-of-hand we slip from storytelling convention to simply ignoring these issues altogether.

Well, maybe the Ferengi have a solution . . .