Many Mouths
As was usual with First Contact, communication had been the problem. The problem we were facing here was that the Nitkas had 56 mouths around their gigantic heads, 28 on each side.
When they spoke, each mouth spoke one word but all the mouths spoke at the same time. If their sentence had less than 56 words, the unused mouths would hoot when talking so as not to be left out. Each sentence was one big crowded shout, like a whole orchestra playing one note for one second. If a Nitka barked several times, that was a paragraph. It was a very efficient method of communication.
We were there for six Earth months trying to put together a translator. It was frustrating because we talked at an obscenely slow speed compared to them with our one lonely mouth. We said our words in a linear order taking forever to meander to the end of a sentence. Only the most patient Nitkas partook in our studies.
Seeing them learn English had been humbling. They broke the dictionary into groups of 56 words and shouted them staccato-blast at each other in their classroom. That took an hour. After that, they blasted rules of syntax to each other in the same way. They could learn our language in a day.
But they couldn’t speak it one word at a time. The one trick we’d been able to teach them was to treat each word as a sentence. They could say ‘the’ with one mouth while the other 55 mouths droned. Then they could say ‘cat’ with one mouth while the other 55 mouths droned. Then they could say “went” with one mouth while the other 55 mouths droned. And so on. The Nitkas were uninterested in that as a solution because it took so long and it was hard for us to hear what was being said by the one mouth with the other mouths droning. In a way, that led to our solution.
With a Nitka standing in a spherical cage of directional microphones pointed at each mouth, we could isolate the one word being spoken. With that discovery, we realized we could isolate all the words. With speech recognition programs, we could recognize all 56 words but then we had to order them. The computer could work out the versions of the sentence that the Nitka probably meant and show them on a screen. The Nitka could point to the right sentence. That let them talk to us fairly quickly.
Speaking back was a challenge. We could dictate words to a small bank of 56 speakers that would say them all at once. We had to be careful to make sure not to say sentences longer than 56 words or the Nitka would get confused. The result was us speaking in a straight line, one word after one word, and then pressing a trigger and the sentence was barked to the Nitka by the speakers. After that, they’d respond and then point to the sentence on the screen that was closest to what they meant. On our side, there were still embarrassing pauses as we spoke but it worked. It encouraged us to be succinct.
The result was a lightweight net of microphones worn by the Nitka ambassadors around their heads with an accompanying datapad for clarity and the humans wore a small bank of speakers on their chests. It remains one of the hardest challenges I’ve faced as a translator designer.
tags
When they spoke, each mouth spoke one word but all the mouths spoke at the same time. If their sentence had less than 56 words, the unused mouths would hoot when talking so as not to be left out. Each sentence was one big crowded shout, like a whole orchestra playing one note for one second. If a Nitka barked several times, that was a paragraph. It was a very efficient method of communication.
We were there for six Earth months trying to put together a translator. It was frustrating because we talked at an obscenely slow speed compared to them with our one lonely mouth. We said our words in a linear order taking forever to meander to the end of a sentence. Only the most patient Nitkas partook in our studies.
Seeing them learn English had been humbling. They broke the dictionary into groups of 56 words and shouted them staccato-blast at each other in their classroom. That took an hour. After that, they blasted rules of syntax to each other in the same way. They could learn our language in a day.
But they couldn’t speak it one word at a time. The one trick we’d been able to teach them was to treat each word as a sentence. They could say ‘the’ with one mouth while the other 55 mouths droned. Then they could say ‘cat’ with one mouth while the other 55 mouths droned. Then they could say “went” with one mouth while the other 55 mouths droned. And so on. The Nitkas were uninterested in that as a solution because it took so long and it was hard for us to hear what was being said by the one mouth with the other mouths droning. In a way, that led to our solution.
With a Nitka standing in a spherical cage of directional microphones pointed at each mouth, we could isolate the one word being spoken. With that discovery, we realized we could isolate all the words. With speech recognition programs, we could recognize all 56 words but then we had to order them. The computer could work out the versions of the sentence that the Nitka probably meant and show them on a screen. The Nitka could point to the right sentence. That let them talk to us fairly quickly.
Speaking back was a challenge. We could dictate words to a small bank of 56 speakers that would say them all at once. We had to be careful to make sure not to say sentences longer than 56 words or the Nitka would get confused. The result was us speaking in a straight line, one word after one word, and then pressing a trigger and the sentence was barked to the Nitka by the speakers. After that, they’d respond and then point to the sentence on the screen that was closest to what they meant. On our side, there were still embarrassing pauses as we spoke but it worked. It encouraged us to be succinct.
The result was a lightweight net of microphones worn by the Nitka ambassadors around their heads with an accompanying datapad for clarity and the humans wore a small bank of speakers on their chests. It remains one of the hardest challenges I’ve faced as a translator designer.
tags