The client's scifi project will be a lot like our your experience listening to Middle English in high school, except what you and I speak is barely-intelligible and old-fashioned, and the characters in the story speak some crazy 31st-century babble.
It's true that people in 1890s London didn't sound like people in 2020s Philly, but English speakers did make a bigger difference between public and private speech. And their private speech was closer to what we speak now. There are examples (from America a generation later) of Hollywood actors dropping out of their official Transatlantic accents: "Why, Mahthah, I cahnnot think of any reason why- hey, get that dog outta here!" If you're curious about it, John McWhorter has recorded a whole bunch of podcasts about English sociolinguistics.
I didn't think I'd be able to find that wikipedia article in Frisian. But you know, it just occurred to me I could ask ChatGPT to make one for me. I also made texts for Bulgarian and Japanese. I'll post them when I have time.
Considering the future trajectory of spoken and written language evolution, I wonder about the effect of real-time machine translation. It seems like we're approaching a future in which two people can speak any language they want to one another and have personal computers automatically translate each other's meaning in their earpieces. This future is already more or less here in the textual medium; I regularly see commenters carrying on conversations in with each other in multiple languages on Youtube thanks to the translate feature on the platform.
Long term, would this arrest the tendency of everyone to switch from their native language to the highest-prestige common standard, like English or Chinese? There might even be some prestige attached to learning and communicating in an especially obscure language, given that it has no practical drawbacks. Maybe there'll be a fad for parents raising their kids as monolingual Greek or Latin speakers, for example.
Even longer term, machine translation may be so good and the need for common speech standards so low that language transmission itself is degraded, with every family or even every individual having their own non-standard idiolect. People would effectively retain baby-talk into adulthood, with any infelicities in their speech smoothed over by machine translation.
Yes, I've been thinking about what good, real-time machine translation will do. My personal guess is the same as yours: languages will splinter into thousands of dialects, with individuals learning a home language from their parents and family. Normally, any non-standard ways you household speaks get smoothed away when you go to kindergarten or start understand TV (kids at that age seem to have an instinct for mimicking high-status people outside their families, such as school teachers and heroic characters in videos). But if kindergartners are fitted with translators, the language at school or from the media will be translated into their home language. This assumes that AIs are constantly listening to private conversations and are smart enough to keep up with each individual's idiosyncratic speech tendencies. So...plausible!
But I don't think you'd get adults speaking baby-talk. The earliest limit would really have to be about six years old. Earlier than that and there isn't anything to translate *into.* Even in kindergarten, what happens when a 6-year old gets exposed to a new word? Teacher says "stallion" and kid hears...what? "Daddy horse"? Or maybe the kid hears the unfiltered word, and that's how he builds vocabulary. Maybe you can change that in the settings :)
In that case, you'd have something like Medieval Europe, with a huge amount of dialectical variation for common, daily-use words, but more uniformity for rare vocabulary you learn in school.
I need to ask the client about how he wants this to work in the language he commissioned. He mentioned that everyone on the colony thought the kids were just speaking normally until there was a wifi outage and the translators stopped working.
This is really cool. Reminds of two things.
1) my high school English teacher reading Canterbury tales to us in Middle English
2) (apocryphal) Victorian English was an affect. Englanders in the 18th century sounded like more Philadelphians now.
The client's scifi project will be a lot like our your experience listening to Middle English in high school, except what you and I speak is barely-intelligible and old-fashioned, and the characters in the story speak some crazy 31st-century babble.
It's true that people in 1890s London didn't sound like people in 2020s Philly, but English speakers did make a bigger difference between public and private speech. And their private speech was closer to what we speak now. There are examples (from America a generation later) of Hollywood actors dropping out of their official Transatlantic accents: "Why, Mahthah, I cahnnot think of any reason why- hey, get that dog outta here!" If you're curious about it, John McWhorter has recorded a whole bunch of podcasts about English sociolinguistics.
Please send! Also, I love the 1940s radio voice.
https://open.spotify.com/show/4u9Jx8cLE4ouDuHpoaYXP7?si=a8f64e104b0b4581
Even the first one is surprisingly easy to read. I imagine Frisian would be a tiny bit easier.
I didn't think I'd be able to find that wikipedia article in Frisian. But you know, it just occurred to me I could ask ChatGPT to make one for me. I also made texts for Bulgarian and Japanese. I'll post them when I have time.
I love this exercise!
Considering the future trajectory of spoken and written language evolution, I wonder about the effect of real-time machine translation. It seems like we're approaching a future in which two people can speak any language they want to one another and have personal computers automatically translate each other's meaning in their earpieces. This future is already more or less here in the textual medium; I regularly see commenters carrying on conversations in with each other in multiple languages on Youtube thanks to the translate feature on the platform.
Long term, would this arrest the tendency of everyone to switch from their native language to the highest-prestige common standard, like English or Chinese? There might even be some prestige attached to learning and communicating in an especially obscure language, given that it has no practical drawbacks. Maybe there'll be a fad for parents raising their kids as monolingual Greek or Latin speakers, for example.
Even longer term, machine translation may be so good and the need for common speech standards so low that language transmission itself is degraded, with every family or even every individual having their own non-standard idiolect. People would effectively retain baby-talk into adulthood, with any infelicities in their speech smoothed over by machine translation.
Yes, I've been thinking about what good, real-time machine translation will do. My personal guess is the same as yours: languages will splinter into thousands of dialects, with individuals learning a home language from their parents and family. Normally, any non-standard ways you household speaks get smoothed away when you go to kindergarten or start understand TV (kids at that age seem to have an instinct for mimicking high-status people outside their families, such as school teachers and heroic characters in videos). But if kindergartners are fitted with translators, the language at school or from the media will be translated into their home language. This assumes that AIs are constantly listening to private conversations and are smart enough to keep up with each individual's idiosyncratic speech tendencies. So...plausible!
But I don't think you'd get adults speaking baby-talk. The earliest limit would really have to be about six years old. Earlier than that and there isn't anything to translate *into.* Even in kindergarten, what happens when a 6-year old gets exposed to a new word? Teacher says "stallion" and kid hears...what? "Daddy horse"? Or maybe the kid hears the unfiltered word, and that's how he builds vocabulary. Maybe you can change that in the settings :)
In that case, you'd have something like Medieval Europe, with a huge amount of dialectical variation for common, daily-use words, but more uniformity for rare vocabulary you learn in school.
I need to ask the client about how he wants this to work in the language he commissioned. He mentioned that everyone on the colony thought the kids were just speaking normally until there was a wifi outage and the translators stopped working.