▲ 54 points
back
75 comments
> political currents don’t have to move in one direction. It’s not inevitable that endangered languages become extinct.
There really isn't any good incentive to learn a language with very few speakers, since unless you already know a speaker, you really can't practice the real language. People were a few years ago at least hoping llms could help preserve the languages, but more most of them, there just aren't enough samples for any real preservation. I knew some dudes in college that were linguists that worked on recording native speakers of endangered languages, and while I hoped they would succeed, I never really had much hope of it working.
People learn Klingon or Esperanto.
You could probably convince me to join a hypothetical “Worshipful society of Gentlemen speaking Sumerian”.
There's a lot of disagreement about Sumerian grammar. If you like to argue, that might be a bonus.
Empirically, the best way to keep a language alive is for a user of that language to write a Great Work in it.
If there are 80 speaker/readers, how could a language get popular again? By interested readers of a translation?
But what good can come from preserving languages that already do not have enough samples?
Do languages have value in themselves?
Unless the language lacks concepts and like scientific recombinateability and the cultural Nexus of youth just moves.
> learn a language with very few speakers
Learn Sanskrit. You were writing in a descendant language just now. If you learnt the roots, you would understand the descendants better. And they are your tools, and knowing your tools is important.
Other more obscure languages? Well, as implied, languages are related... So studying the network clarifies the whole.
Or learn Hittite. It's got the oldest corpus in the Indo-European language family. You've got your standard assortment of Indo-European cases. Grammatical gender is animate/inanimate and the animate gender has yet to split into masculine and feminine. The script is tons of fun, especially if you write it with a stylus on clay. And as a bonus, you get to learn a smattering of Sumerian and Akkadian vocabulary.
Animate gender didn’t split into masculine and feminine — inanimate collective plural was repurposed as feminine.
That "repurposing" is more morphology than grammar. You also frequently find in Indo-European languages multiple cases having overlapping morphology, but that doesn't necessarily change the cases themselves. My point, in any case, is that Hittite has the older animate/inanimate genders from Proto-Indo-European (only attested in the Anatolian languages!) whereas Sanskrit has the later masculine/feminine/neuter genders.
Can you point us somewhere zo learn more about that?
Sanskrit is not an ancestor of English. Sanskrit is a satem language English centum. Proto Indo Iranian diverged long ago.
> Sanskrit is not an ancestor of English
Sanskrit entered into Greek, Greek into Latin, Latin into English.
Edit: downvoter, express yourself clearly. I cannot see how the above sentence could be considered false, apart from gross interpretations of the chosen terms.
Again: you have a relation with "jñāna, gnō̃́sis, gnosco, knowledge"; then you have relations in other dimensions (jña, jñāti, gnṓmōn, gnṓmōn...): the awareness of those details enriches the awareness of the language. That is through the relation between Sanskrit, Greek, Latin, post-Latin - it is just there. What is contended is not clear.
Sanskrit is no older than Kurdish; English Greek etc have the same relation to it as to Kurdish — they are all alike of Indo European descent.
Like Greek and Latin, Sanskrit is an anciently written language associated with major states and ‘civilizations’ - and, crucially, is like Greek and Latin in that it is still remembered if only by some thin intellectual or priestly stratum of societies.
So by use of those three, Indo European unity could be especially well validated by pure textual scholarship (especially if the scholar was British or German and thus on another branch). It is surprising as the result is intuitively scientific — how could I arrive at it if all I do is read and write?! - and was indeed of massive importance to the framing of Darwin’s quite different scientific enterprise, which needed different evidence .
Everyone should learn Sanskrit if only to read the great poets and dramatists in the original. The translations are often nice, but you do lose something in the process. Thanks to the incredible efforts of Pāṇini[1], the language was mostly frozen in time and is still just as understandable today as it was 2500 years ago.
I started learning it because it is India's civilizational language and I wanted to read the itihāsas in the original.
I have the Assimil Sanskrit course sitting on my shelf that I’ve been meaning to spend time with. Would love to be able to read old sutras one of these days.
An interesting tangent to the language discussion is the Epic of Gilgamesh => https://www.astralcodexten.com/p/your-book-review-the-epic-o...
You know how people learn Ancient Greek to study Homer in the original language?
Gilgamesh deserves that. I'm only a beginner but the refinement of the poetry is already obvious even as someone who doesn't know the language.
Euphony was very important to the Mesopotamian poets. They liked rhythms and alliteration and assonance. And the late editions of Gilgamesh reflect about two thousand years of iterative improvement:
> ilū işinu iriša. ilū kima zūmbē eli bēl niqi iptaḫru. ("The gods could smell the fragrance. They gathered around the sacrificiant like flies.")
The ş is probably an affricate like ts, š is sh, q like in Arabic, ḫ like German ch. Vowels like in Spanish or the phonetic alphabet. Macron means a long vowel. Stress on the penultimate. (Edit: fixed ş / š)
It really rolls off the tongue, doesn't it?
> ibri kudanu ťardu akkannu ša šadi nimru ša şēri ("my friend the swift driven mule, wild ass of the mountains, panther of the wild")
> ibri ša arammušu danniš ittīa ittallaku kalu marşāti ("my friend, whom I loved greatly, who journeyed with me through every hardship")
> ikšudāššu šimat ("fate overtook [him...]" the tablet breaks off here)
>The ş is like sh, š is regular s
That's not correct, in transcriptions of Akkadian š is pronounced sh (Heb. shin), ṣ (not ş) is pronounced ts (Heb. tsade), and there's also s, pronounced as regular s.
Is it possible that the Hebrew consonants' pronunciations don't align fully with their cognate consonants' pronunciations in Akkadian? This is certainly true among other Semitic languages, that there are drifts over time in the values of consonants.
> Gilgamesh deserves that
The very moral in Gilgamesh deserves that. "I went in a quest for immortality; I did not achieve it and came back, and while returning I saw in the distance the City I built, and I understood what my immortality was...".
(A "must" of an insight for engineers, architects, builders...)
You're possibly misreading that.
In the Bible the city builder was Cain, and he built it because he was a criminal and needed protection from the world when he lost God's grace.
I.e., the message is that a city is a kind of protective exoskeleton or "safe space" when man becomes estranged from God (the Universe).
It's a story about man's fragility and ultimate futility.
The longevity of a city is difficult to achieve as an engineer/architect/builder nowadays. One might have better chances as a scientist.
The Astralcodex link has a YouTube URL to a woman purportedly reading the Epic in the original tongue.
The golden age of languages is when there is only one language left, and everyone can understand everyone without cunning translators and hallucinating llms.
Isn't that a collapse in cognitive diversity?
Not to be "gotcha", but out of true curiosity: may I ask if you speak multiple languages fluently?
Different languages afford different ways of thinking and speaking? Just talk to a translator in works of literature or poetry. Some languages are more verb-y or noun-y, and so are more about getting between things than being something. Or have more or less ambiguity due to tonal systems.
And while ambiguity can create miscommunication when misused, it also allows for different minds to hear one thing, but interpret it in ways more personal to them. When used pro-socially (and not manipulatively), this can afford cooperation that might not otherwise be possible. Saying things while leaving some things unsaid, that has utility if our collective project is certain forms of synchrony -- we can sync on some things while the ambiguity allows asynchrony to remain unresolved in less important dimensions, etc
Imho monoculture is toward death, not toward something golden.
>Not to be "gotcha", but out of true curiosity: may I ask if you speak multiple languages fluently?
yeah, I speak 4 languages fluently, and can get through more with a dictionary
>Imho monoculture is toward death, not toward something golden
Same language does not mean same culture. Singapore and USA both speak English, yet the cultures are different.
The same culture can be represented in different languages, as translators prove by their existence.
Unfortunately, even if we had a single language, I doubt we’d be free of misunderstandings. Jargon, double entendres, shibboleths and other barriers to understanding all still exist even when everyone is nominally speaking the same language
A big practical difficulty is the pressure for languages to continue to evolve in different directions when speaker communities are even slightly separated (hence the difference in present-day varieties of English, which is probably growing larger over time rather than smaller, based on (almost, approximately) a single common ancestor just a couple of centuries ago in the time of the British Empire and colonial expansion).
People can decide to try to prevent dialects from forming and diverging
https://en.wikipedia.org/wiki/Language_planning
but they have to actively work against it if they want to prevent it, and a lot of the common means of doing so historically have been pretty harsh (e.g. corporal punishment in schools for using disfavored varieties). I guess this is partly because the consensus that we should centralize and unify and standardize language, for all of its logistical and economic and cultural benefits, is mostly a central government/imperial metropole impulse rather than something that the general public is super-attracted to.
Standardizing language and letting it diverge both have cultural benefits! Just different ones. The standardization benefit is that more people can access a much larger body of culture without linguistic barriers and without needing translation or explanations, and cultural conversations can grow larger, with more direct participants in a given conversation or genre. The diversity benefit is that the language differences themselves can acquire, or mediate, cultural phenomena and cultural creativity.
English doesn't have a significant tradition of language planning the way that some of the continental European languages do, and this is posing a (minor) challenge to English's role as a global language, as there are sometimes genuine difficulties of intelligibility between English varieties, and again I think those differences are getting wider on average rather than narrower. Then again, if we did have such a tradition, it doesn't mean everyone would care to follow it... it's not as though everyone in Latin America cares about what the RAE says (or that they refrain from having meaningfully different vocabulary and grammar between Latin American regions!), or even that everyone in Austria or Switzerland necessary cares what Duden says.
If we got down to only one language, that natural human tendency would be for new languages to grow out of that. They naturally change all the time, and suppressing that is a loss, not a gain.
Hmmm maybe. There are obviously benefits to adapting languages to better describe the environment the speaker finds themselves in. But the utility of being able to communicate with others (especially if they are the overwhelming majority) is probably higher. In the absence of geographical separation, I don’t think new languages would have become mutually incomprehensible.
Knowing multiple languages, in particular less related or unrelated ones, is mentally enriching. If you only know a single language, you don’t know what you are missing in terms of diversity of thought and expression. Languages aren’t fungible. Translation is lossy, often considerably so.
>Knowing multiple languages, in particular less related or unrelated ones, is mentally enriching
Just as being able to play an instrument or dance or actually reading literature from other cultures, even in English.
A thought experiment I sometimes think about is if it'd ever be possible for the world to somehow, in some way, agree that every country would switch over to the same constructed language over the course of generation or two. A modern conlang without all the idiosyncrasies, irregular verbs, exceptions to rules, etc. that's endemic in current organic languages. Of course, how long a language like that could stay relatively clean is an open question.
It's interesting to think about the next generation being taught both their home language and the new language, they're the intermediate generation that can bridge the two worlds, and the generation after that being only taught the new language, and how that process might work.
We are already switching to English slowly, but steadily.
Live translators exist! It has become much better than even a couple years ago
Between one and three thousand years ago.
Less ambiguously: between 1000 and 3000 years ago. ;)
So definitely not either 1000 years ago or 3000 years ago, but in the range 975BC to 1025AD?
Cuz if it's got to be between 1/3000, it can't be 1000 or 3000, right?
Between can of course mean (1000,3000) or [1000,3000], right?
Just poking fun at the attempt to remove ambiguity including ambiguity.
Why are you assuming that the boundaries are whole calendrical years? Clearly, I meant exactly between the precise moment 1000 years ago and the precise moment 3000 years ago. ;)
I think that we are entering a new golden age of language, personally.
It has never been easier for a human being to take a corpus and refactor it in a fashion that they can understand it without making an energetic investment in constructing a new ontology; or that is to say, humans can use AI now to grasp text and words - language, specifically - in extraordinarily industrious, productive ways.
I know it is a crackpot event horizon, but I truly wonder if Szukalskis’ Protong isn’t just waiting there, tip of the iceberg, to be discovered. Seriously.
Vibe coding is doing its part to keep linguistic diversity alive: My wife went to sign up for a doctors appointment, and the form offered choices for "preferred language" that included Old English, ancient Egyptian and Elamite.
I don't suppose the age ended with a particular tower falling down?
A minor point, but the colour scheme on that chart infuriates me. When the data can be seen a progression (not endangered -> threatened -> few known speakers -> no known speakers), why use red to green but in a random order?
The order is alphabetical, not random. ;)
> We show that language isn’t isolated from political currents and that this dynamic has been a facet of human history for much longer than previously thought.
Um, when did anyone think it wasn't?
The entire history of linguistics shows this. Proto-Indo European had no common word (for that family) for sea/ocean [1] from which we can deduce something about the people who spoke it. That's not political obviously but environment and culture are inseparable from language is the ponit.
Languages really only grew for political reasons. Look at the spread of Latin throughout Roman territories.
Likewise, languages splinter and diverge in isolation. Populations that remain where they are and separate from others will drift linguistically. You can look at most of Europe for this. As just one example, the drift in Swiss Germany in Switzerland, which is a relatively small country.
But more than that, it takes some kind of centralization, which is to say, political force, to maintain a language. English had massive changes from Old English (which is essentially a separate language from Modern English) to Middle English (which is largely comprehensible by the modern speaker) when French was the court language of England in the centuries after 1066. And the Normans left their own imprints on English, which is where a ton of loan words came from [2]. But English grammar also greatly simplified without a central authority telling everyone what is proper, basically [3].
One of the fascinating things I remember learning was that there are between half a million and million cuneiform tablets int he world and fewwer than 10% of them have been translated. Now cuneiform is a writing style but one of the difficulties is that it encompasses multiple languages. And this was one of the earliest recorded examples of language displacement from Sumerian to Akkadian over thousands of years. But tablets could be written in part in both languages for a very long time. This is where we have the earliest documented human name, Kushim [4].
But the point is that the rise and fall of Sumer and its displacement by Akkad is recorded in the language.
And of course we had the spread of "colonial" languages in the colonial period (ie Spanish, French, Portugese, English).
Pragmatism also is a factor here. Different languages are inconvenient between populations that interact. The natural tendency is to homogenize on language given sufficient time, even without political pressure. it's a little sad when languages die but it's kind of inevitable.
[1]: https://talageri.blogspot.com/2020/05/the-proto-indo-europea...
[2]: https://www.krysstal.com/display_borrowlang.php?lang=Norman+...
[3]: https://www.sciscanpub.com/index/journals/ainfo/alr/7943.htm...
humanity's natural state is an impressive diversity of ideas and behavior and then inevitable the most violent of them marginalize the rest. Humanity has already found many social structure that prevent concentration of power but those have been lost not to time, but to intentional destruction.
Often natives look at our way of life and say hell no that's retarded. But they are more polite about it. Europeans on the other hand just defined their flawed structure as "progress" and went on killing anyone who disagreed. In a way the worst parts of AGI are already here, we already killed all the real humans and we are the soulless automatons bounded to our possessions.