Thanks to OP Tipping at Wordorigins, I learn that Wikipedia has a list of extinct languages sorted by their time of extinction (mobile link). It is, needless to say, long. I learned about some interesting languages, like Zarphatic (which sounds like it should be spoken on the planet Zarph). And when I got to the 2nd millennium BCE I had the loss of Hattic thrown in my face yet again. (The latest entry at time of writing is Bering Aleut; I’m sure more languages have gone extinct since March 7…)
Very interesting. It’s worth skipping to the footnotes and perusing the sources. For the recent language extinctions, many of them are interesting reads: obituaries of the last native speaker of their language, typically with photos and personal details.
Also, it’s clear this list does not include even the majority of extinctions. It lists only 2-3 per year on average. One outfit that appears to track this more closely, the Language Conservancy, says that on average about 9 of the 7000+ languages disappear per year. (This is much better than some estimates made 15 or so years ago, which projected losing 90% of all languages by the end of the 21st C.)
So the list is capturing maybe one out of three extinctions. Most languages that die do not make the news, at least not in English.
Sumerian wins! First recorded, first to go …
There is absolutely no sense in which Egyptian “became extinct” in 600 BC. You might as well say that English became extinct in 1066. (The major structural changes in Egyptian took place between Old/Middle Egyptian and Late Egyptian/Demotic/Coptic and had happened by Akhenaten’s time; the distinction between Demotic and Old Coptic is script, not language.)
David – I noticed that too. If they meant the extinction of the written language (hieroglyphs) that would still be 4th century AD at the earliest.
Several ancestors of Uyghur are shown as “extinct” even though they are extinct in the same way Latin or ancient Greek are extinct.
You can sometimes buy surprising things in French supermarkets. One such thing was a facsimile reprint of Champollion’s Grammaire Égyptienne, offered at such a reasonable price (I forget how much, but less than 20€) that I bought it like a shot. What impressed me most was not so much the deciphering of hieroglyphics, because I more or less knew about that already, but the extremely short time (less than ten years) between that and a complete description of the language. Clearly Egyptian was very far from extinct.
Exactly that, in the sense that the Egyptian language was actually already known, albeit in the much later phase of Coptic. The recognition that Coptic actually is Egyptian goes back to Kircher:
https://en.wikipedia.org/wiki/Athanasius_Kircher
Although Champollion’s work certainly was remarkable, decipherment where both the script and the language are quite unknown is a much more difficult proposition. The way that successive decipherments of languages written in cuneiform have successfully bootstrapped on one another is in many ways more impressive (and much less generally known.)
(The thing that most impresses me about linguistic Egyptology is the sheer ingenuity which scholars have brought to bear on elucidating Egyptian syntax, with an exceedingly honorable mention going to Hans Polotsky.)
For a lot of the languages listed the “dates” given seem little better than guesses (unless they are wholly inaccurate, as in the case of Ancient Egyptian), which in many instances could be off by several centuries: in many instances the date of “death” seems to be the date when the language ceased to be written (or, to be more accurate, the date when writings in the language known to us seemingly cease to be produced…assuming those writings which are known to us have been dated accurately, naturally).
Even taking the above caveat into consideration, I do not understand how they can date the death of Oscan to 100 B.C.: there are graffiti in Oscan on the walls of Pompeii, which was buried in volcanic ash in 79 A.D. And if Oscan was still a living language within an urban center such as Pompeii, it must have been much more widely spoken in the Southern Italian countryside. I suspect its ultimate extinction took place centuries later, perhaps indeed after the fall of the Western Roman Empire.
Are there any attested medieval/early-modern assertions of the form “Priscian / Alcuin / Aquinas / Buridan / my prof was the last person to speak Latin”?
If Dante is any guide, the mediaeval understanding was that nobody had ever spoken Latin in the first place, in the sense that they supposed that the situation in Roman times was the same as it was for them, viz everyone actually spoke Tuscan (or whatever) as L1 but learnt Latin (possibly while very young, but not as a mother tongue) as an interlanguage and as a language of literacy and high culture.
So the question would have made no sense for them.
In a yet earlier period, people simply supposed that they were speaking Latin, and wrote their Extremely Early Old French (or whatever) as if it were classical Latin, but pronounced it as EEOF. They would have regarded writing what they actually spoke just as bad spelling, not as a novel linguistic departure.
For recent entries on the list, maybe the takeaway is that the U.S. and Australia aren’t better than other countries at stopping language extinction but they’re better at noticing / documenting it when it happens? I imagine there are lots of poorer countries in Africa and South/Southeast Asia where if the total number of different languages spoken drops from 98 to 97 no one is doing a news story about it.
Yes, I think that’s a reasonable conclusion.
And of course the list is full of errors, some idiotic, and one wishes it were done better, but the same is true of Wikipedia as a whole.
It’s interesting that the first languages with a “last speaker” recorded by name, and therefore with an official “extinction date” set when they died, all start poping up in the late 1700s: Polabian (1756), Galwegian Gaelic (1760), Cuman (1770), Cornish (1777). That must reflect the beginning of European popular and scholarly interest in language diversity, as a strain of antiquarianism, which would soon evolve into the rigorous science of comparative linguistics.
It happens again and again that a language is falsely declared without speakers, and then nobody bothers to look for any more. Ainu has suffered that discourse several times, declared dead again and again since the 1960s at least.
In recent times, more and more communities are taking charge of how their languages are represented to the world at large. One side of that is the use of the word “extinct”, which to a lot of people signifies something going out of the hands of the community and forever into the public formaldehyde jar. The recent trend of speaking of languages as “dormant” rather than “extinct” reflects, among other things, the potential for revitalization. This is not superficial woo. It has very real implications to the way minority communities reclaim their place in the world at large. I still use the word “extinct”, but prefer to use more neutral alternatives, and hope that in this context the word itself will be extinguished.
Most of the recent entries are timed to the death of the last known L1 speaker, which makes sense, but realistically some of these languages were moribund decades before that death.
Martin: “One outfit that appears to track this more closely, the Language Conservancy, says that on average about 9 of the 7000+ languages disappear per year. (This is much better than some estimates made 15 or so years ago, which projected losing 90% of all languages by the end of the 21st C.)”
90% might be pessimistic but there is a long list of languages that are probably beyond the point of rescue. There are over a hundred languages in Australia alone that cannot be expected to survive til 2050: only spoken by seniors, not being taught to children. UNESCO lists some 1700 languages as either “definitely endangered”, “severely endangered” or “critically endangered”.
A bureaucrat in me is tempted to tie support to indigenous communities to ability to speak their indigenous language.
“You are not allowed to hunt whales since whaling is prohibited to everyone who is not indigenous”
“But I am an indigenous person, I am an Itelmen, we always hunted whales”
“The Government instruction No.1123245/501 dated 11.12.2022 says that a member of indigenous group must demonstrate ability to speak the language of that indigenous group. Here is an Itelmen language proficiency test for you. You will be issued a whale hunting permit once you pass the test”
I wonder how the number of speakers will grow with such policy.
If there are any deadset errors in the list and you don’t feel like editing Wikipedia, just post them here and I’ll get them fixed.
No one ever claimed Beothuk is Algic, certainly not Algonquian. It’s been suggested that it’s remotely related to Algic, not convincingly.
Cupeño is Uto-Aztecan, not Chumashan.
Corrected Cupeño, Beothuk, Oscan.
I left Egyptian as is: I take your points, the comment column does mention that it evolved into Demotic, so I don’t think the current entry is actually misleading even though the validity of Egyptian’s presence in the list is moot.
Thanks indeed for doing the work we’re too lazy (or too jaded) to do, OPT …
However:
Demotic is Egyptian. Saying that Egyptian became “extinct” at that point is exactly as sensible as having an entry for the extinction of Early Modern English. (Demotic is, properly speaking, the name of a writing system, though the term is, by extension, used for the Egyptian of that period; but the latest inscriptions in hieroglyphic are in fact later than the latest texts written in Demotic script.)
This is partly a matter of genre: the sort of texts that were written in hieroglyphic during the period when Demotic was the ordinary writing system for Egyptian were also written in a deliberately extremely archaising style: the scribes attempted to write in the language of the Middle Kingdom, which was by then as remote from contemporary Egyptian as Latin from Italian; but it’s all Egyptian, and the major changes that turned Middle Egyptian into the later kind of Egyptian had been completed centuries before the “Demotic” period: they are first seen clearly at the time of Akhenaten, when that heretic Pharaoh engineered a deliberate radical break with traditiion which (among other things) led to the language being written more or less as spoken at the time, instead of in (attempted) Middle Egyptian, which had long been a “dead” language – in the sense that Latin was “dead” in mediaeval Europe, that is (i.e., not.)
Egyptian really became extinct when Coptic was finally supplanted altogether by Arabic; the date of that is unclear (and the issue is unfortunately affected by politics and wishful thinking), but it was probably well into the second millennium CE, at any rate.
All of the 15 languages marked as “evolved into…” suffer from the same problem as Egyptian/Demotic. All languages have evolved from earlier forms, and you can divide them arbitrarily into as many stages as you like. That modern linguists call an earlier stage by a different name doesn’t mean tha tthere was a sharp divide involving language replacement.
Also: Jie links to the Nilotic, not the Yeniseian language.
I’m not disagreeing, but unilaterally deleting the evolved languages would be a bit rude so I’ll start a discussion in the article’s talk page to see if I can obtain a consensus to get rid of such entries.
Assuming that you are “Ordinary Person” then I see that you have already acted on your proposal. Excellent idea, and I have added my two cents. Let us hope that others join in.
there are graffiti in Oscan on the walls of Pompeii, which was buried in volcanic ash in 79 A.D.
This doesn’t necessarily prove anything; AFAIK the only dated piece of graffiti in Pompeii turns out to be dated to something like 78 BC. It does suggest a later date than 100 BC, though.
Of course, in general, for a lot of the ancient languages the extinction date is a guess at best; notably, for several of them, there are much later reports to the effect of “the people over in those mountains speak something weird we do not understand”.
to the death of the last known L1 speaker, which makes sense
…though of course even otherwise extinct languages can be transmitted through a community of L2 speakers, as had happened to Hebrew for roughly eighteen centuries, and to Sumerian possibly even longer.
@January First-of-May
Well by some standards then Egyptian ain’t even dead yet because Coptic is still in use as a liturgical language. Real live humans are still reading it, speaking it…
A language which has large written corpus (and a dictionary) can’t really go extinct.
On that basis, Akkadian* is not extinct …
Although the concept of “extinction” is a good bit less straightforward than it initially appears, I don’t think it’s all that problematic in general; a language is extinct when no longer acquired by children without any special deliberate effort being made to teach them the language. It has been supposed that this was the case with Hebrew in the period before the “revival” in modern times; but whether or not that is the case, Hebrew would certainly be very exceptional, so uncertainty on this point does not vitiate the concept of language death in general.
The great difficulty is the Latin versus Romance one (et sic de similibus.) The only linguistically reasonable way I can think of to apply the concept of extinction in cases like this is to say that a language has become “extinct” when it has developed into more than one mutually incomprehensible form: but even this is impossible to apply consistently. Mutual comprehensibility is not all-or-nothing, and is hard to be sure of even with modern spoken languages; moreover, how many speakers do we need of an aberrant variant in order to declare that the older form of the language is now dead? Is German dead because Swiss German is incomprehensible to people from Cologne? Is Arabic extinct? Is Chinese? Is Modern Greek extinct?
Accordingly, It is not possible to tag languages like Latin as “extinct” without leading to manifest absurdity; declaring Egyptian extinct in the first millennium BC is even more absurd. It’s not a matter of terminology: it’s just outright wrong.
* On reflexion, the reason you could sensibly call Akkadian extinct, but not mediaeval Latin or mediaeval Hebrew, is that Latin and Hebrew continued to play major roles in their societies even though no longer transmitted “naturally.” If there were a community that used Akkadian as an everyday means of communication, then you could perhaps claim that Akkadian was not extinct. But even this criterion cannot be applied consistently: how much of a role in everyday life is needed for a language not to be extinct? Is a treasured role in a beloved liturgy (along with a very necessary Arabic crib) enough? Does it mean using the language to write to your mother, or only to write your school essays?
When you have texts and a dictionary of an existing language, you automatically get L2 speakers (or L2 users).
Because on this planet there has to be someone interested enough in this language or rather several someones and they will go to conferences and meet each other and discuss it and maybe even converse in it.
In short, it’s no longer quite dead.
Stupid autocorrect.
Changed “extinct” to “existing”.
@de
German and Italian (to some extent also English) are “umbrella” languages; even where a proscriptive standard exists, individuals are mainly diglossic, or believe themselves to be using the standard when they are not. Swiss German is just an extreme version, compare the intelligibility of some Scottish accents. So I think “German” has not become extinct; it is just a label for dialect continuum (on linguistic grounds probably including Dutch and Flemish, perhaps also Frisian) plus diglossia.
compare the intelligibility of some Scottish accents
All pairfickly intelligible.
Let us hope that others join in.
I have done so. Let reason prevail!
@David Eddyshaw, funnily, I misunderstood (or at least misread) your earlier point, and was going to ask you several hours ago, how Akkadian – Latin – Old English – Early Middle English – Arabic – Hebrew are different. They look like a spectrum.
(along with a very necessary Arabic crib) – what do you mean?
The Coptic liturgy includes some passages in Coptic, but with an Arabic translation.
Ah, I am an idiot:)
I was thinking about Latin and then Akkadian, and could not connect them to Arabic cribs:)
Now I want to see Latin with Akkadian cribs.
I was thinking about […] Akkadian
I rather like the idea of a society of loyal Marduk-worshippers out there somewhere in Iraq, with their lovlngly preserved Babylonian hymns and prayers (for which, however, the laity need the help of a running Arabic commentary …)
I think I spoke too hastily in saying that if Hebrew had been transmitted continuously, without specific language instruction, up until modern times, that would be very exceptional.
Ghanaian schoolchildren acquire English at school, not because they learn it the way French is taught in schools in the UK, but because that is the language used at school, and they’re still young enough to still be able to pick up Ghanaian English with native-speaker competence just from that. It would be silly to deny that that is “natural transmission” too.
I don’t think that is an exceptional situation world-wide, now I think about it. I don’t know if mediaeval schoolboys (probably not girls, eheu) acquired effective native-speaker competence in Mediaeval Latin, but it’s certainly conceivable. It would actually have been the Renaissance that finished Latin off finally, on that hypothesis, by pushing the unequivocally dead usage of Cicero as the only acceptable form of the language at the expense of any living oral tradition.
I rather like the idea of a society of loyal Marduk-worshippers out there somewhere in Iraq
Well, the Harranians were such a group.
So I googled Harran.
The people there – in Turkey on the Syrian border – worshipped precisely not Marduk, but Sin, the moon. And indeed they kept going for a long time.
The English Wikipedia article on Harran has a section “End of the Sabians”, which says: “In 1032 or 1033, the temple of the Sabians was destroyed and the urban community extinguished by an uprising of the rural ‘Alid-Shiite population and impoverished Muslim militias.” Then the Mongols came in 1260 and flattened the place entirely, it wasn’t rebuilt for at least 100 years.
But the German Wikipedia article on the Sabians says they merged into the Shamsi Alawites in the 12th/13th century. And on those I can find nothing.
Dutch incl. Flemish, yes; Frisian (or English), nope.
I wonder if there are villages in Egypt where Coptic is sill used in everyday life. When I went to Birmingham in 1970 there was a Coptic Christian PhD student in the department. I only knew her slightly (I knew her supervisor much better). I could have asked her if she knew the Coptic language. Too late now.
I wonder if there are villages in Egypt where Coptic is sill used in everyday life
Apparently not, and not for centuries; though there have been a few enthusiasts in recent times who have attempted to raise their children as Coptic-speaking.
I did once have a Copt colleague who was gratifyingly impressed that I could recite the Lord’s Prayer in Bohairic Coptic. (He couldn’t. He did recognise it, though.)
@DM:
i may or may not be right about this, but i think “shamsi alawites” would be the folks known as ‘alawi’ or ‘nusayri’, centered along the coast of what’s now syria, as opposed to “anatolian alawites”, who get called ‘alevi’, centered in inland eastern anatolia. they’re different but often confused, because the terminology is always confusing, and because they’re both complicatedly related to shi’a islam (as hatters can tell from the “ali” in their names).
Oh, if they’re simply the Alawites-as-opposed-to-Alevites, then they’re in charge of Syria! The Ancient Mesopotamian moon-worshippers are among us, gradually modified beyond recognition in a Ship-of-Perseus way. 🙂
…I should have recognized al-Shams. Prompted by the moon, I confused it with the “sun” word.
All pairfickly intelligible.
That’s pushing it. “Deep” Shaetlan is barely intelligible to other Scots, never mind non-Scots.
pushing the unequivocally dead usage of Cicero as the only acceptable form of the language
That happens around Charlemagne’s time, I think, with his missi dominici. However, it doesn’t reach Romance-speaking Iberia until the 10C or so.
“…, but Sin, the moon. ”
Celestial bodies as I remember.
I wanted to write about Harran. when Belti (“my lady”, Venus, friday) was mentioned here.
The Nusayri are just the Alawites, broadly construed; the name is after the founder of the sect, Abu Shuʿayb Muḥammad ibn Nuṣayr al-Numayri, who was a close associate of the last few Twelver imams. I believe the Alawites claim that ibn Nusayr was (one of) the appointed messengers of the Madhi during the early days of occultation (in which the twelfth imam has supposedly been in divinely assisted hiding since the tenth century).
If you accept “evolved into”s, you could start a separate list of first speakers, like Modern English, ca., idk, 1500, alongside Esperanto, 1887, etc.
Celestial bodies as I remember. Correction:
there is Harran of Akkadian times, Hellenistic and Christian times, when it was described as pagan.
And then then there is an anecdote cited by al-Nadīm:
from (also available in libgen and other pirate collections).
It is not the only anecdote about them (in al-Nadīm, and other books, Arabic and others). Al-Biruni calls them Buddhists. But I think there is not a strong association with Moon there as there was in Akkadian times.
And of course, accuracy of the account above have been questioned:-)
—
Let reason prevail! it won’t. January First-of-May says “a lot”. It is not a lot, it is all.
There is no way to tell if there are speakers other than a complete census-like linguistic survey.
If there is no way to tell, why pretend?
It is absolute complete brutal shameless pseudo-science, pardon my ordering. Totally, utterly.
@de, jc (yesterday)
What I mean is something like this speaker:
https://m.youtube.com/watch?v=AWYZTbWcBPg
For me the problem is more unusual stress (or lack of stress) of syllables within words and words within sentences rather than pronunciation. For example he says “give it a shot” and it took me a second to realise that was what he had said.
A Russian and Bokelji speakers:
“ɪntɛrˈnɛt!” “what?” “ɪntɛrˈnɛt!!” “what??!” “ɪntɛrˈnɛt!!!!!” ….”ah, you mean înternet!!!!”
Instead of Proper Bokelji I’m giving “Serbo-Croatian” transcription from WIktionary. I have no idea how Bokelji say this word. Their accent is quite difficult for me.
‘We are Ṣābians (Ṣābʾūn),’ for this is the name of a religion which Allah, may His name be exalted, mentioned in the Qurʾān. Profess it and you will be saved by it.”
One of the Hausa terms for “Pagan Hausa” is Maguzawa, “Magians” (i.e. Zoroastrians); this came about because it was inexpedient (or impossible) for the Emirs to order the conversion of millions of their subjects on pain of death, and a pretext was necessary for not doing so, in good Muslim conscience. Hence “Magians.”
I recall that the heathen Norse were also called Majus by the Muslims that encountered them, or that they encountered.
After Muslim invaders came to rule much of South Asia they (as in the Hausa-speaking regions) found themselves in the awkward position of having many too many Hindu subjects for conversion-or-massacre to be a practical approach, although Hindus did not fit into the “People of the Book” category like Magians/Sabians etc. particularly naturally. So it may have taken considerable time and ingenuity to come up with appropriate rationales.
Another addition to the list
Southern Tsimshian
British Colombia (Canada)
date: early 2013
last speaker: Violet Neasloss (1914-2013)
reliable sources: Marie-Lucie Tarpent, Dale McCreery
@JWB
I think you are overestimating the zeal of secular rulers (some of whom had only converted in order to retain their positions) for following abstract precepts in a way that would lead to conflict and disaffection. In Bengal, where mass conversion occurred, this seems to have happened very gradually under Mughal rule in a transformation like Hindu gods and spirits -> Allah + Hindu gods + Muslim saints and angels/demons -> Allah /Hindu god (dual identification) + Muslim saints -> Islam
No one ever claimed Beothuk is Algic, certainly not Algonquian. It’s been suggested that it’s remotely related to Algic, not convincingly.
Well, it’s been *claimed* — a lot of things have been claimed, including that Beothuk was Athapaskan or Austronesian. Sapir put it, with a question mark, as a sister group to Algonquian (is that who you mean by “not convincingly”?). A little more recently, John Hewson did claim in the 1960s-70s that the evidence supported Algonquian based on his reconstruction of Proto-Algonquian; as far as I can tell from a little googling, that didn’t get much traction, though Hewson did go on to publish a Proto-Algonquian dictionary. Obviously Goddard and Campbell are right that the question is unanswerable.
I don’t think even Hewson would say that it was closer to Algonquian that Yurok/Wiyot.
Marie-Lucie, did she work with you?
Yes, during several summers. After the first time I stayed in her house and was treated like a member of the family. She considered ST as her language, as that was her mother’s, but she had been raised bilingual in Heiltsuk, which was her father’s language, still spoken by a number of elders there at the time (as had been by her husband). She also spoke Coast Tsimshian, along with some relatives who had migrated elsewhere. And of course the local brand of English. She had outlived all the other ST speakers.
After I was unable to return, the village hired Dale McCreery, then a graduate student, to work on a school program. I think Violet was thrilled to have a young man working with her! I later met Dale, who is a very talented linguist. He later moved to Bella Coola where he still works and has put down roots.
My original interest in ST was comparative-historical. The only person who had worked on it (actually “discovered” it) was John Dunn, who declared that ST was as different from Coast Tsimshian as the latter was of Nisqa’a (which I know best). I can say categorically that it is closest to CT. The differences are mostly in phonology and lexicon. ST, which is geographically farthest and used to be even more so 200 years ago, preserves phonological contrasts which do not exist in the others, for instance ST “wan” ‘to sit (pl), “xwan” ‘deer”, all others “wan” (both meanings). This means that it is crucial for Proto-Tsimshianic reconstruction and comparison with other Penutian languages.
m-l, I hope you’ll be able to reconstruct your lost comment in the De Castries thread — I’m dying of curiosity to know what you have to say about that odd name!
If you accept “evolved into”s, you could start a separate list of first speakers, like Modern English, ca., idk, 1500, alongside Esperanto, 1887, etc.
I wonder if Itamar Ben-Avi qualifies as “first speaker” of Modern Hebrew.
Mr Hat, it is taking me longer than I expected, but yes, I will post more in the Castries thread.
Just to clarify: “pseudo-science” above referred to historical languages. We have no slightest idea if the language of Meroitic inscriptions is still spoken today or was spoken back then.