So how exactly does the translator know when to work and when not to work? A typical example of what I’m talking about:
A Klingon and human are conversing. Human says something to piss the Klingon off and all the sudden the Klingon starts shouting expletives at the human. Only except, the translator isn’t translating the words.
Or another example is when two Klingons want to say something to each other but don’t want the human to understand what the are saying. They just start speaking Klingon to each other. You never see them deactivate the translator.
The concept behind it was explained in the TOS episode “Metamorphosis”: There are certain concepts that are common to all intelligent life. The UT detects brainwaves, identifies the concepts it recognizes, and provides the appropriate translation. Not 100% efficient, but nothing ever is.
The example shown in the episode was the handheld device Kirk used to converse with The Companion. There was no mention of the UT used in ship-to-ship communication, referred to in at least one other episode.
Judging from how UTs were used in subsequent iterations of the show, one would assume they were interfacing directly with the brains of both parties, rather than just verbally. I presume they used electromagnetic waves for this purpose.
It just knows. 23rd century tech, you know. It’s not unlike the Heisenberg compensator on the transporter. When asked “How does the Heisenberg compensator work?” by Time magazine, Star Trek technical adviser Michael Okuda responded: "It works very well, thank you.
It really wasn’t that fancy. It’s just an app on a standard communicator, a computer uses a predictive model translator for most of the words, at the same time bio alien language specialists are listening in on the conversation at locations around the universe. Whenever the computer falters on some language one of the bios fills in the gap. It works sort of like Uber, when the conversation starts the first bio to correctly identify the language gets to work the call, but they only get paid a nominal rate for listening, the only real money comes when the communicator software fails. However, the communicator has to skip anything that sounds like a profanity because a mistranslation could have disastrous results, so the bios make a lot from that when anyone is talking to Klingons.
Which works great unless there is a multi-hour time lag for subspace messages like in Balance of Terror. And watch out for those congestion prices.
The communications software of the ship should be able to do instant translations of known languages. 300 years more work on Google translate and it should be perfect. And you’d think they’d teach Klingonese at the Academy, so it won’t even be necessary lots of the time.
As for new cultures - magic - unless the plot demands that it doesn’t work, like the excellent TNG example we all know and love.
Nichelle Nichols brought up a similar point (that as a communications officer she should be able to speak Klingon) during Star Trek: VI when Uhura had to thump through actual paper dictionaries to come up mangled Klingon sentences, but was overruled by the director.
I also thought it was ridiculous for Uhura to not understand Klingon. I’m glad Nichelle brought it up, even though she was overruled. At least Abrams let Zoe Saldana speak it in Into Darkness.
Both Uhura and Scotty were made to look ridiculous in Star Trek VI.
I liked that episode with the Skrrea on DS9 in which the UT didn’t work at first, and they had to keep talking to them until it developed a translation matrix.
