Oh, that changes everything! I don’t care about the mechanism–magic DuoLingo is fine–but the consequences were what I was focusing on.
With magic fluency, I’m far less concerned about consequences.
In this scenario most Nepali will probably pick up Hindi, which is very similar.
I think preserving the indigenous languages might prove useful in phycology and sociology someday, I would hate the thought of them just disappearing.
Ultimately, erasing language is a component of cultural genocide, and language suppression has been historically used for that purpose.* Our hypothetical will be enormously damaging in that context, especially since the languages will be fully erased from history rather than simply no longer being spoken.
*I’m in no way ascribing those motivations to the OP.
It’s not Moonrise doing it, it’s those damn aliens.
So it’s really about what writing exists in those languages. Without the writings it’s a bit pointless IMO. Random alien in 100,000AD can impress his friends by speaking English, but the last scrap of English writing has long since disappeared (and of course last native English speaker long before that), who cares? I’d choose a single surviving Maya codex and enough Mayan language to read it.
The OP specifies:
Admittedly, it doesn’t specify that they’ll be actively used by humans. Or even by some future species calling themselves humans, but evolved beyond easy recognition.
I’d be fine with kicking out French to make room for either one of those.
What weight do we assign to each of the following:
Number of speakers
Ease of learning
Historical significance (both written documents and oral history)
Change in language over time (old/middle/modern English, high/low German, etc.)
Vocabulary
Language family
If we go back to 1494 and the Treaty of Tordesillas we could have picked Spanish as #1 and Portuguese as #2. Go back a century or so and we’d choose French because it was “the language of diplomacy.” Certainly we’d have had no use for Hindi because India was under British rule and English would certainly drive Hindi into the ground. As for Japanese and Mandarin, maybe we would have considered them IF they’d adopt a Western written alphabet.
A civilization which would no longer have access to Latin or Aramaic, or be able to decipher the Rosetta Stone, or where the people of Scotland don’t have 421 words to describe winter conditions is poorer because of it.
It occurs to me that if the Sentinelese suddenly found themselves speaking English or Hindi or whatever, they might run into serious problems. Would the alternative language have all the words they need for what they need to know to stay alive?
I’d expect they’ve got words for the specific plants and animals they use and live with, and for the particular techniques they use to do so. If their language is to disappear leaving no trace – would knowledge essential for survival go with it?
Yes.
I didn’t feel the need to specify this but since you mention it, I’m a linguist by training, so I’m fascinated by the diversity of the world’s languages. The more there are to study, the better.
But I’m also a guy, so I like to make stupid rankings.
And the reason for that is… ?
Like it or not, French is a shoo-in, not just because of the demographics, but because of its historical importance, its centuries-long literary prominence and its academic interest (it’d be a stretch to call it a Germanic-Latin creole, but not a huge one).
That was the utimate goal of this thread.
- Number of speakers is probably criteria n°1 ;
- Ease of learning is both irrelevant in my opinion, and entirely dependent on your native language ;
- Historical significance would be my second most important criteria IF we include scientific significance ;
- Change in language over time would be included in the scientific significance criteria ;
- Vocabulary is not clear to me ;
- Language family is a very interesting criteria. Depending on how you rank them, you get different priorities.
In terms of total number of speakers, the top 5 looks like this :
- Indo-European
- Sino-Tibetan
- Niger-Congo
- Afro-Asiatic
- Austronesian
But in terms of total number of distinct languages, it looks like this :
- Niger-Congo
- Austronesian
- Sino-Tibetan
- Indo-European
- Afro-Asiatic
The weighing changes depending on whether you put the emphasis on demographics or linguistic diversity. Whichever is the most important, I think it’s essential to cover as many families as possible.
So do it by rotation - let’s say by family, then by number of speakers in that family - so
1 most spoken Indo-European language
2 most spoken Sino-Tibetan
3 … Niger-Congo
…
6 second most spoken Indo-European
…
40 8th most spoken Austronesian
and then 10 dead languages by whatever criteria make sense - number of texts/historical/religious significance/whatever.
Nice idea, except that it’d give an exaggerated weight to the five biggest language families to the exclusion of to the dozens of others. It’d be harsh to reject Dravidian (250+ million speakers), Turkic (180 million), Japonic (130 million), and Austroasiatic (115+ million).
Perhaps we could leave 10 spots for small but scientifically or historically relevant languages. Then, attribute points for the number of speakers and for the number of languages, add them up in the 9 language families mentioned above, then find a way to attribute a number of saved languages to each of the big families depending on that score to share the remaining 40 spots.
- Niger-Congo 7 + 9 = 16 points
- Indo-European 9 + 6 = 15 points
- Sino-Tibetan 8 + 7 = 15 points
- Austronesian 5 + 8 = 13 points
- Afro-Asiatic 6 + 5 = 11 points
- Dravidian 4 + 0 = 4 points (lots of speakers but only 84 distinct languages)
- Turkic 3 + 0 = 3 points (lots of speakers but only 33 distinct languages)
- Japonic 2 + 0 = 2 points (lots of speakers but only 13 distinct languages)
- Austroasiatic 1 + 1 = 2 points (lots of speakers and 169 distinct languages)
Treating the results as a percentage (rounded up for Indo-European and Sino-Tibetan), I get :
1. Niger-Congo - 8 saved languages
1. Indo-European - 8 saved languages
2. Sino-Tibetan - 8 saved languages
4. Austronesian - 6 saved languages
1. Afro-Asiatic - 5 saved languages
6. Dravidian - 2 saved languages
7. Turkic - 1 saved language
8. Japonic - 1 saved language
9. Austroasiatic - 1 saved language
I’d love to have some mathematicians weighing in on this, though. It may not be the best method.
Not a mathematician, but I think it looks like a good system. Only thing is that it does seem to completely exclude language groups from the Americas. That seems a shame.
Are we imagining that these chosen fifty will be static? Or will the fifty continue to evolve and adapt to future and even current needs? Borrowing from each other and coining new words and usages both of and independent of the interactions, even splintering into new dialects.
Hard to imagine any alien technology or fantasy magic that prevents such.
I would argue in that case that the analogy to preservation of biological diversity holds highest sway. Like the seed bank. Who knows what future needs may be?
Of course cover the bulk of current speakers and currently relevant corpus with the most spoken and read three to five. Then look for extreme linguistic diversity.
ǂNūkhoegowab is in, then…
I’d says sounds good to me, but I don’t know how that sounds … 
Like this…
So does sound good to me!
But the serious question: how linguistically diverse are the different Niger-Congo (I’m guessing that is one of them) languages? From each other and from the other families of languages?
I’d be more thinking that one of the click languages has to make the cut just because it is very different in structure than many of the more spoken ones.
No. This is a Khoe language, completely unrelated group.
A few Southern African Niger-Congo languages have picked up a few clicks from Khoe languages (Xhosa & Zulu, for instance) but most of them don’t have any click consonants at all.
And the N-C languages vary very widely - they are, after all, spoken across most of Sub-Saharan Africa. And Africa is a big place.