Brendan Borrell’s NY Times story (archived) about a once elusive people of Indonesia doesn’t have a great deal about language in it, but there’s a fair amount about genetics, which I know is of interest to a formidable group of Hattics, and it’s quite a story in general, so I thought I’d post it. The Punan people were thought to have abandoned their traditional hunter-gatherer lifestyle:
And so in 2018, when Stephen Lansing, an anthropologist at the Santa Fe Institute, and Pradiptajati Kusuma, a geneticist at the Mochtar Riady Institute for Nanotechnology in Tangerang, Indonesia, said they had learned of a clan of about 30 Punan families who sheltered in limestone caves and rarely, if ever, emerged from the forest, many experts were skeptical. But with funding from the National Science Foundation, the scientists made contact with the nomadic group in 2018, and began collecting data with the aim of ensuring their health and welfare.
After that first trip, Dr. Lansing returned to Santa Fe with photographs of a man wearing a loincloth made of bark fiber, along with recordings of a song language he believed resembled no other. His initial description of these people, who call themselves the Cave Punan or Punan Batu, was published last year in the journal Evolutionary Human Sciences. Press reports in the Indonesian media catalyzed the local government to declare the Punan Batu as regular users of their forest, a first step toward obtaining the right to manage it under national laws.
Some experts remain doubtful that this unusual group could really have been secluded for so long. The skeptics have compared the announcement to that of the Tasaday, a “lost tribe” discovered in the Philippines in 1971, whose isolation was eventually determined to be an exaggeration, if not a hoax.
Bernard Sellato, a Punan specialist with the French National Center for Scientific Research, has been particularly fierce in his criticism. In an email, he referred to the Punan Batu and other coastal groups as “‘fake’ Punan.” Based on historical accounts and ethnographies, he remains convinced their ancestors were not native to the island, but rather enslaved people imported from New Guinea and eastern Indonesia several centuries ago.
But a new study focusing on the DNA of the Punan Batu, recently accepted by a scientific journal, is poised to eliminate the doubts of all but the most hardened critics. Based on the limited diversity revealed in the genes of the Punan Batu, they appear to have been isolated for more than 20 generations. Dr. Sellato’s contention that the Punan Batu are the descendants of imported slaves does not fit with these results. […]
Halfway up the mountain was a cavern as large as an amphitheater. The cave, which contained a dense concentration of swiftlet nests, is a sacred site for the Punan, who consider it the source of all things. Once inside, a man named Ma’ruf took a seat on the dirt floor. He was in his early forties but appeared to be half that age, with swooped-over bangs and the youthful skin that comes from a life lived in the shade. […]
Ma’ruf began to hum, a deep and powerful vocalization that rose from his chest and echoed through the cave. Words took shape in a language only the elders understood. “I am like a porcupine who comes to the cave to rest,” he chanted, according to a translation of a recording of the chant made by Dr. Lansing.
The next singer up was a shirtless man in his sixties named Bo’odon. “I am a true friend of yours,” he sang to Dr. Lansing. “When might this bear fruit, I ask myself. May our relationship bring the return of our lands to us.”
Such song languages represent a fluid form of creative expression. Unlike a typical spoken language, where different speakers will identify common objects and concepts using the same words more than 95 percent of the time, there was only a 70 percent overlap in the vocabulary used by different Punan Batu singers, according to word lists Dr. Lansing has gathered.
Dr. Lansing has wondered if any unknown words in the songs may have been passed down from Borneo’s Stone Age hunter-gatherers, who we know from bones and cave paintings. Determining whether the Punan Batu are descendants of those cave dwellers, or if they may have coexisted with them, would require a genetic analysis.
Yes, the whole words-passed-down-from-the-Stone-Age thing is highly dubious, but hey, anything’s possible. There’s lots more at the link, including gorgeous photos. (Thanks, Bonnie!)
the whole words-passed-down-from-the-Stone-Age thing is highly dubious
Oh, I don’t know: in parts of New Guinea, the Stone Age is within living memory …
(And in parts of Amazonia, it was this morning.)
What concretely do they even mean by songlike? All kinds of varieties are described as “songlike” by speakers of varieties with a slightly difference cadence. Or are we to think of it more like a whistled language?
As the Wikipedia article notes in its first sentence, “Punan Bah or Punan is an ethnic group found in Sarawak, Malaysia and in Kalimantan, Indonesia.” In other words, they’re on the island of Borneo. Both the ad hoc entity that we call Indonesia and the other that we call Malaysia have holdings there (Malaysia has the state of Sabah as well as Sarawak). Brunei constitutes the small remainder.
As it happens I’ll be in Sarawak for part of October. I’ll look out for local manifestations of Punan culture.
Collateral reading about humming. Hot off the virtual press. These Punan may be on to something.
As for the language, by some local reckonings Punan may trump them all as the westernmost Austronesian tongue. At least on the evidence of my trip to Sabah about fifteen golden years ago (yes, I climbed to the top of Gunung Kinabalu), directions work in a decidedly different way on the island. Another matter I incidentally intend to investigate in October.
The paper is here: https://www.cambridge.org/core/journals/evolutionary-human-sciences/article/deep-ancestry-of-collapsing-networks-of-nomadic-huntergatherers-in-borneo/3E5BDE9823F6CD827E66DAF7C307273F
‘Song language’ meaning a language only used for singing and not for everyday talking. Their regular language is Punan Sajau, an Austronesian language shared with a neighbouring group. This song language (Basa Latala) is, according to the authors, an isolate (and apparently so is a different ritual song language from Borneo, called Sangiang).
“Songlike” is a word used in the title. A strange word.
But yes,
“However, some are also fluent in a hitherto undocumented language which is not spoken, but rather sung, customarily at night in the rock shelters or caves. This song language, Latala or Menirak (‘to sing’ in the daily Punan language), is known only to the Punan Batu, although it shares some formal features with other ritual song languages of Borneo such as Sangiang (Baier & Scharer, 1987), the language of the first ancestors employed by the Ngaju Dayaks whose territory lies to the south of the Punan Batu in Central Kalimantan. Figure 5a shows the phylogenetic relationships between these languages, estimated using an optimised linguistic distance (Downey et al., 2017). ”
By lingusitic distance they mean similarity between words as measured by a machine:)
“Bernard Sellato, a Punan specialist with the French National Center for Scientific Research, has been particularly fierce in his criticism. In an email, he referred to the Punan Batu and other coastal groups as “‘fake’ Punan.” Based on historical accounts and ethnographies, he remains convinced their ancestors were not native to the island, but rather enslaved people imported from New Guinea and eastern Indonesia several centuries ago.
…..to have been isolated for more than 20 generations. Dr. Sellato’s contention that the Punan Batu are the descendants of imported slaves does not fit with these results”
20 generations is several centuries….
Not relevant to this post, but two new open-access books I heard about that might be of interest:
The lexicon of Proto Oceanic: The culture and environment of ancestral Oceanic society – we’ve used this before here, but apparently the last volume only just came out.
The Tibetic languages just came out at my own lab (one of the last projects of a collapsing publisher), but I wasn’t involved.
On the subject of the actual post: this Latala “song language” sounds interesting; I was unimpressed by the phylogenetic diagrams, but glad to find they’ve also put up a proper data table. Although… taumenditawena for “not”? Pajengorin for “one”? Really?
The next step should be etymologising, not phylogenising. aku for “I” is pretty obviously Austronesian anyhow…
....... Basa Latala . Punan Sajau . Modang . Punan Kelaihusband .. tsoʔin ....... awan .... səgwən .... segun
wife .... sopeinihu ..... awan .... səgwən .... segun
It seems the linguist among the co-authors is Peter K. Norquest.
Sangiang does not seem to have attracted much attention of linguists.
A very poetic register indeed.
“First ancestors” and the language names are in italics in the original; I’ve mirrored that here.
The “language tree” is based on distances, i.e. total similarity, not on any attempt to identify shared innovations. It does contain a good sample of Austroasiatic across its entire diversity, though – in addition to Austronesian languages across the Indonesian archipelago (but no further).
and from the Methods:
…and some of this must have been known to specialists for a good long while, judging from the existence of the following reference:
It is weird, though, to use “I” in “intraspeaker transfer”. Or are the songs all in the 3rd person?
…Maybe not anymore: “Several Punan composed and sang songs to us in Latala evoking the loss of their forests and asking for help.”
It does contain a good sample of Austroasiatic across its entire diversity, though – in addition to Austronesian languages across the Indonesian archipelago (but no further).
Bernard Sellato’s hypothesis would seem to suggest that we should rather be comparing these to Papuan languages…
Actually, I should forward this to my Papuanist acquaintances.
the only words of Austronesian origin detected so far appear to be the result of intraspeaker transfer from the daily language of Punan Sajau into Latala
aku for “I” looks like a counterexample on the face of it: similar forms are found on the table in Bintulu, Ma’anyan, Palembang etc. but not given for Punan Sajau. But this is Indonesia, so maybe haʔ is just one of half a dozen 1st person singular pronouns in Punan Sajau.
Not just ‘not’. There are much too many long words and too few short ones. I suspect esoterogeny, not through borrowing or using rare synonyms, but through euphemistic compounds or something like that. That means someone should put the computer aside and do some etymology.
The story sounded like a string of exaggerations for a noble cause, to help the forest dwellers protect their forest. The genomic part didn’t really convince me that these groups have super deep roots. Together with other signs of reduced population diversity, it just might point to drift due to relative isolation of a small founder population. Founder-derived group genetic patterns are always easy to overinterpret …
“Earth” is ahom, “sky” is ahomnukasa – presumably “earth-above” or the like. “Fire” is apoioroŋ – the first half is obviously Punan Sajau apoi. Yes, esoterogeny – compounds, kenning, euphemism, you name it – seems vital for analysing the data from this “language”.
“like a string of exaggerations for a noble cause”
It does not soudn too sensationalist to me: “In his grant proposal with the National Science Foundation, Dr. Lansing described how he would be making “first contact” with the Punan Batu, a bold claim that contributed to early concerns about his credibility. Experts did not believe there were any uncontacted Indigenous groups in Borneo. It was only after Dr. Lansing paid Mr. Mita to take him there in 2018 that he learned Mr. Mita had previously brought in other foreigners.”
_____
The researchers are concerned with the history of this population.
The readers I think are intrigued by song language (the very practice of having and using one).
And its genesis of course must be interesting to linguists (or at least it is stupid to only concern oneself with spoken languages, period).
Three lines here, linguistic philogenetics is subordinate to the first of these causes. I guess they would have to do (or refer to) some for Punan Batu’s spoken langauge anyway….
____
Regarding the origin of the vocabulary, I would not assume that it is homogenous (all words were borrowed from the same source / created the same way).
“That means someone should put the computer aside and do some etymology.”
Still can be done for the other song language mentioned….
“not” can be an error.
Though it would be convenient to have a language where “no” is a minute or two long word:)
And they said that are two lists (with forms coinciding only for 70% of words) – but in their table there is only one. How did they obtain it from those two?
That’s answered in the Methods section. As in Nature-tier journals, the Methods section is last, not second.
The language section from Supplementary Materials:
DM, oops, you are right:-/
Sadly,
1. “we included only the lexical data from the first elicitation session”
2. “the word for ‘to know’, which in the first session was recorded as [lakopan], and in the second as [parakopan]”
3.
103 to know .... parakopan ; ;140 to say ....paluwaiAh, they confused the first and the second session at some point…
Sangiang:
From A bibliography of the languages of Borneo (and Madagascar) (academia):
Basically i decided to quote it because of this “and the language of songs’ (emphasis mine) perhaps distinct from the language of sprits (B. Sangiang).
____
Ritual languages, special registers and speech decorum in Austronesian languages by James J. Fox (in Austronesian Languages of Asia and Madagascar, chapter 4., online) has:
“The Bible translator A. Hardeland, in his Versuch einer Grammatik der Dajackschen Sprache (1858), was the first to note the Hebraic parallels in Dayak ‘spirit language’ (basa sangiang). Although Hardeland provided only a single textual illustration of this ‘spirit language’ with translation and commentary, the posthumous publication of the Swiss missionary H. Schärer’s two-volume compilation of texts in basa sangiang, Der Totenkult der Ngadju Dajak in Süd-Borneo (1966), provides a substantial corpus for study. M. Baier (1987) continued this work of more than a century by publishing a dictionary of basa sangiang based on materials from both Hardeland and Schärer. This corpus on the Ngaju has been further extended by the substantial work of Sri Kuhnt-Saptodewo (1993, 1999).”
Hardeland in Sri Kuhnt-Saptodewo’s translation:
“The magical chants (…) consist of short, mostly paired, parallel and synonymous parts of speech which repeat the same thought in the second arrangement using other words taken from the Sangiang language, whereas in the first arrangement the Dayak speech is more commonly found”
Then Sri Kuhnt-Saptodewo explains:
(here)
My point here is that priests apparently feel the need to replace “not” with a word absent from normal speech often enough.
Perhaps just because synonyms are not available in that normal speech.
priests apparently feel the need to replace “not” with a word absent from normal speech often enough
Really? Where?
I suppose there is the
https://en.wikipedia.org/wiki/Wicked_Bible
but that is said to have been the printer’s fault. (Along with “Behold, the Lord our God hath shewed us his glory and his great-asse.” *)
We shall probably never know the true story of Gnostic sabotage orchestrated by Opus Dei and the Templars in conjunction with Radical Socialist Globalists led by George Soros.
* The so-called “Moonie Bible.”
I meant Ngaju priests but you are right. For it was printed:
quid autem vides festucam in [o]culo fratris tui et trabem in [o]culo tuo non vides
… his glory and his great-asse
Michelangelo represents both very pointedly, on the Sistine ceiling.
Thanks, drasvi, for uncovering so much of the relevant context. One thing we can certainly conclude is that, for a poetic register of this nature, simply running a wordlist through an “optimised linguistic distance” program should be expected to give a vastly exaggerated measure of the distance from other languages. Figure 5a is therefore quite inappropriate and misleading, as well as doing little to support the text; I don’t understand why anyone would waste their time producing such a thing when they’re already providing a much more useful table.
Re. “basa Latala”: in a number of Borneo languages, Latala is the name of (the highest) God, presumably derived from Allah, thus bearing evidence of islamic influence – not surprising in Indonesian and Malaysian contexts.
Evidence of the glottal stop as allophone of ?
Hunter-gatherers have ritual languages! A lightbulb moment rather than a chock revelation, but still very interesting.
Aha, apparently there are also Lahatala and Alahatala (presumably < Allah Ta'ala)
Trond, what do you mean?
Is this your generalisation or a well-known one that you just remembered?
What generalization? It just hadn’t struck me before that small communities of hunter-gatherers might have ritual languages, but that says more of me than hunter-gatherers. Because why shouldn’t they?
I thought I you mean that hunter-gathers tend to have them more than others* (Catholics, Jews, Muslims, Russians…) do:)
Can make some sense, given that, say, Latin is not, say, a secret langauge.
But it seems you were moving in the opposite direction.
* slightly incorrect: there can be some Catholic or Muslim [ethnic] Jewish Russian-speaking hunter-gatherers.
You can listen to the song language in this podcast on the story: https://leakeyfoundation.org/podcast/episode-42-the-cave-punan/
It is a song.
Video of the song language is here: https://leakeyfoundation.org/origin-stories-the-cave-punan/
Maybe it’s plain Austronesian after all. Blevins and Kaufman, “Austronesian Lexemes in Basa Latala of Borneo: A Punan Sajau Song Language”, Oceanic Linguistics 63(1), 182, 2024, here.
I say maybe, because some of the etymologies are uncertain. It seems like the data is sloppy to begin with. B&K tentatively parse “taumenditawena”, supposedly ‘not’, as tau menda, tau (e)na’, ‘don’t know, don’t know’.