We’ve discussed the interesting and much-disputed subject of the origins of the Austronesian languages before: in 2009 in connection with “Language Phylogenies Reveal Expansion Pulses and Pauses in Pacific Settlement,” by Gray, Drummond, and Greenhill, which came down on the side of Taiwanese origin, and in 2014 in connection with Roger Blench’s paper “Suppose we are wrong about the Austronesian settlement of Taiwan?,” which came to the contrary conclusion that the Formosan languages “represent a continuing flow of pre-Austronesian languages from the mainland.” Now, according to a Phys.org news story, a team of archaeogenetic researchers led by Martin Richards “proposes a solution based on what has been the most comprehensive analysis so far of DNA from the region”:
The various branches of the Austronesian language can be traced back to a Taiwanese original, and DNA analysis does show that there was some expansion from Taiwan, about 4,000 years ago. But this accounted for a minority of the whole region’s population – no more than 20 per cent. An explanation for the spread of the language was that these Taiwanese migrants came to constitute an elite group, or became associated with a new religion or philosophy, according to Professor Richards.
I’ll be interested to see what those who know more than I do about the topic have to say. (Thanks for the link, Trevor!)
Original article here. Hurrah for open access!
Incidentally, the title of the original paper, “Resolving the ancestry of Austronesian‐speaking populations” is accurate. The Phys.org title, “New research into the origins of the Austronesian languages”, is not. Language is not genes, and it’s been known for a long time that many Austronesian languages are spoken by people of non-Austronesian genetics (‘Papuan’, ‘Negrito’, and others).
Open access in a Springer journal? The authors must have a huge grant. Well, they probably do, being geneticists…
I’ve downloaded the paper and will probably read it at some point. 🙂
Language is not genes
This is why I’m always loath to mention languages being “genetically related” in discussions with less informed people. It’s rather inapt to speak of genetic relationships without genes, isn’t it? It’s a shame that there isn’t a better, less potentially confusing term for us to use. Maybe “memetically related”? I guess that terminology would fail to adequately distinguish between descent and borrowing, but at least you can make a case that language is made of memes.
Not a very good case. It’s true that “genetic” can be confusing to the uninitiated, but it’s perfectly normal for technical terms to have different senses in different fields — the problem isn’t the terminology, it’s that people have zero exposure to the findings of linguistics, not even the basic exposure they have to the findings of physics and astronomy. They may not be able to explain the various subatomic particles, but they know such things exist and if they want to know more they will ask someone who has studied physics; not only do they have no concept of the bedrock facts of linguistics (to boil it down to Golden Rule level: language varies in space and time, and that’s both inevitable and perfectly OK), they have no concept that there is a class of scientists who deal with such things and know more than they do. Until that changes, there’s no point worrying about potentially misleading vocabulary.
Etymologically, no: genetic relationships are those within a gens ‘extended family’, and linguists had the term first.
Same Indo-European root, but genetic is Greek (adjective formed from genesis) while gens is Latin. Not that I know anything about the history of the use of the word genetic.
The first chapter of the bible is about genesis, so geneticists did not have the term first. They didn’t exist then.
That first chapter discusses the origin of species and languages, albeit without the more recent concepts of random mutation and selection. This morning I wondered if there is a relationship between origin and genesis but failed to find one during breakfast. http://www.acceity.org/2009/07/the-genetics-of-language/ was as close as I got.
Honestly, I don’t know how well researched this paper is.
In the conclusions they state,
“As both archaeologists and linguists have suggested, alluding to the spread of the early Metal Age in Europe, it may be that what began to spread across ISEA around 4000 years ago was primarily a new way of thinking—the adoption of a new ideology and perhaps even a new religion”
What??? The early Metal Age in Europe saw a massive migration of people. There was almost total genetic turnover in much of Northern Europe. How old are these “archaeologists and linguists”?
@David Marjanović
On the topic of open access papers…
Perhaps this article may be of interest.
http://tinyurl.com/hb23vqr
Sci-hub.io doesn’t seem to be working right now; we’ll see what happens.
Latin origo has to do with the springs where rivers (as we say in English) rise, and thereby sources in general: orient is closely related, as is Greek orchestra, the place on stage where people dance. No apparent connection with gens or genesis (whose Germanic counterpart, by the way, is kin(d).
Is the g in origo a part of the root (*ergh- according to etymonline) that disappeared in orior and so on, or is it derivational?
I know about that; it can’t possibly have motivated Springer (one of the Big Four science publishers) to set anything free. More likely, the journal is one of those where authors can make their paper open-access by paying 2500 to 3000 US$.
Note that although the view from the biology department looks like this, across scholarly publishing as a whole only about 30% (per Peter Suber’s statistics) of open-access journals require author fees. The rest are subsidized by universities, or accept advertising, or use other business models. In addition, some journals offer institutional arrangements, the reverse of institutional subscriptions, whereby scholars at that institution can publish all their papers in open access without per-paper fees.
I certainly didn’t intend to suggest that Sci-Hub.io influenced any publisher to make articles free. They will eventually all try to pry your last 35 dollars from you using their cold dead hands… ( or paypal, if it come to that ).
I was just meaning to let people know there are alternatives out there.
I don’t know what problems John Cowan is having, but I just checked with 3 different browsers from 3 different VPNs in various countries, and it seems to work fine. You might need to wait for a minute or two if you are looking for a super rare article that no one has requested before.
Yes, the problem I and others were seeing seems to have gone away.
Things seems to be pretty clear cut in Austronesian studies: most divergent languages are Formosan and 20% of genetic lineages in Austronesian-speaking populations are from Taiwan. Conclusion: elite dominance. Not so in Indo-European studies: most divergent languages are Anatolian but genetically Indo-Europeans are from the Pontic steppes (up to 50% of steppe legacy in Western Europe, if I’m not mistaken). Whence the discrepancy between Austronesian and Indo-European situations I wonder?
Austronesian is far from clear.
Is the g in origo a part of the root (*ergh- according to etymonline) that disappeared in orior and so on, or is it derivational?
It’s derivational: cf. vertigo from verto “turn”. There are various other Latin examples of derived nouns in -igo, -ago and -ugo.
Orchestra is probably unrelated; it’s from Greek orkheomai “dance”, whose etymology is unclear, but LIV cites a Hittite cognate arkatta “bespringt” and reconstructs a root *h₁erǵʰ-.
@TR, I did have a feeling that that was the case, but no examples came to mind. And I don’t know where to find a rhyming dictionary of Latin 🙂
Now I remembered virago as well, of course.
TR: Orchestra is probably unrelated; it’s from Greek orkheomai “dance”, whose etymology is unclear, but LIV cites a Hittite cognate arkatta “bespringt” and reconstructs a root *h₁erǵʰ-.
That’s the multifarious “arousal” word, the origin of scattered words for testicles and male animals in different branches of IE, whatever that might suggest about the origin of Greek drama. The Germanic descendant is the Scand./Ger. adjective arg “angry”. The modern meaning must come from metaphorical “fucked”. The older attestations mean “available to be taken sexual advantage of” and hence all sorts of macho extensions: “perverse; unmanly; weak; scared; dull”.
Arg in Danish has gone from “lascivious (of women)” over “wicked,” “immoral,” “bad,” to “obnoxious,” with only a brief, obsolete side track to “angry.” I like to use it, but it’s otherwise obsolescent.
Or rather it split, with an original spelling variant arrig taking on “grumpy,” “irascible,” and impotent-type “angry/mad.” (Righteously or acting-out “angry” is vred).
Forarge is “affront,” ærgre is “cause regret”.
@John Cowan
“Austronesian is far from clear.”
Do you have Roger Blench’s critique in mind?
-most divergent languages are Anatolian but genetically Indo-Europeans are from the Pontic steppes (up to 50% of steppe legacy in Western Europe, if I’m not mistaken). Whence the discrepancy between Austronesian and Indo-European situations I wonder?
Similarly to Armenian, speakers of Anatolian languages came into region dominated by speakers of non-Indo-European languages (Hattic, Hurrite, Urartian, etc.) They are thought to be related to modern languages of North Caucasus.
Anyway, local impact on newcomers who spoke Anatolian languages was extremely strong. In case of Hittites, they were completely assimilated to local Hattian culture, down to personal names, religion and even ethnic self-appellation.
Hence, Hittite – result of heavy Caucasisation of a perfectly normal Indo-European language.
Armenian (and I presume Phrygian) underwent same process, only a millenia later.
Not sure if Austronesian example fits this model.
@SFReader
So, essentially what you are saying is that Formosan is divergent due to an authentically earlier onset of a branching process, while Anatolian is divergent because of secondary influence of local non-Indo-European languages?
😀 “Humps”!
The German version means “extreme in a bad way”, by no means as precise as the Old Norse version, which is said to mean “sexually perverse”. Together, these probably account for the diversity of meanings in Danish.
Good that you remind me of ärgern “annoy”, Ärger “the state of being annoyed”, “low-level wrath”; I didn’t perceive them as related to arg.
The big difference here is that Anatolian is a single branch, while “Formosan” is not. The argument for Taiwan as the Austronesian homeland rests on the idea that 9 out of 10 (or 8 out of 9, I forgot) of the basic Austronesian branches are found on Taiwan and only there.
(This perception may change if Blench is right; we’ll have to see.)
I think he’s at most half right: Taiwan might well be a convergence zone for Austronesian varieties reshaped by rubbing up against one another, and the origin point might be somewhere else altogether, rather than the divergence point for Austronesian. I don’t think the archaeological evidence bears on that part of the hypothesis one way or the other.
What if the Austronsian were not develpoed in Taiwan, but were brought to Taiwan as per Roger Blench’s comments. The ancestors might be a group of people who spoke Proto Tai-Kadai+Austronesian and separated from each other in the south China coastal area. One migrated to Taiwan and the other to inland of south China with multiple movements.
As the relationship between Tai-Kadai and Austronesian is becoming accepted despite a dispute among the idea of Tai-Kadai is a branch of Austronesian leaded by Laurent Sagart and the idea of Tai-Kadai is a sister language to Austronesian leaded by Weera Ostapirat.
Where the sister language idea seems to get a stronger position, the Proto Tai-Kdai+Austronesian speakers might slowly migrate from the further south – Land of Sunda at the end of ice age – 8,000 years ago as per Dr. Richards study.
The idea of “Sunda (via South China) Origin” is totally different from “Mainland China (via Taiwan) Origin” which would open a new chapter of Austronesian history if it is proved to be sound.
One simple word “fish” could be a good indicator.
Some of Formosa aboriginal people of Taiwan call fish as “vulaw”, “quleh”, ”qcurux”, “baute”, “kuraw”, “pulaw”, “alao”, “ʔalaw”, “ciqaw”, “vutukulu” and “ʔælaw” and most of Tai-Kadai call fish as “pla”. The interesting is that Indonesian also has word called “pulau” << pu+laut = something arising + sea = island. Both fish and island then share the common meaning ‘something arising in the sea’.
It reflexes 3 important messages; Firstly, “pu” and “laut” are the oldest words, then “pulau” is the second and “vulaw” and “pla” of Formosa and Tai-Kadai are the youngest. Secondly, “vulaw” and “pla” have lost the form of island and shifted to the form of fish while still maintained the ground meaning of ‘something arising in the sea’. Thirdly, “fish” implies the primitive way of living near to or at the sea with island existed.
It is an indicator that it was not originated from Out of Taiwan but older than Taiwan and not from Mainland China but from South China Sea, the mother land of Malayo Polynesian (and Tai-Kadai) as per Roger Blench critic; suppose we are wrong.
Another indicator is that the basic counting system of Austronesian 1-10.
Austronesian basic counting number; *esa/isa-1, *duSa-2, *telu-3, *Sepat-4, *lima-5, *enem-6, *pitu-7, *walu-8, *Siwa-9 and *sa-puluq-10 (Robert Blust2013)
The numbers might be reconstructed from a group of living fossil words including kali, lu, nem, pat, pu, sa, tu and wa. By applying a simple counting system, it can be divided into two parts; number 1-5 as a core part and number 6-10 as a surrounded part. Number 1-5 represent a family body and number 6-10 represent a way of living in this case meaning to water.
The philosophy;
starting point >> having direction >> getting through >> almost there >> be united
The meaning of each living fossil word;
“kali” means ‘something used for counting’ (finger) (preserved in Tai)
“lu” means ‘pass through’ (preserved in Tai)
“nem” means ‘water’ (preserved in Tai)
“pat” means ‘thin, tight or tiny’
“pu” means ‘something arising’ (preserved in Tai)
“sa” means ‘my body or a body’ (preserved in Tai)
“tu” means ‘direction’
“wa” means ‘spreading or expanding’ (preserved in Tai)
Reconstruction;
The core part refers to a story of self-family.
*esa/isa-1 (staring point)
This word is put weight on “sa”. Affix “sa” has been composed in many Indonesian words such as “saya” << sa+ayah = myself + is belong to my farther = me, “sakit” << sa+kit = myself + is horned = get sick, “sarung” << sa+rung = myself + lives in my tiny sack = my sarung, see the connection with Indonesian word “satu” << sa+tu = myself + is caused to exist = ‘single’.
*duSa-2 (having direction)
This word is rather put weight on “du” than “Sa”. It is acting as ‘moving out of home’- “Sa” on one hand and acting as ‘a mirror effect’ on the other hand. “*duSa” can be compared to a word “dulu” in Indonesia which means the past.
*telu-3 (getting through)
“lu” is a very old affix in Indonesian as well as in Tai. It shares the same meaning as ‘something has traveled with a distance from one point through another point’. In Indonesian, many words are the result of suffix “lu” such as “lalu” = passing or then, “bulu” = hair, “dahulu” = before, “hulu” = head.
If “*esa/isa” is the first existence or a farther and “*duSa” is the mirror effect of “*esa/isa” or a mother, then “*telu” would imply something has departed from farther (passed) through mother or ‘become a child’, a reproduction of “*esa/isa” and “duSa”.
*Sepat-4 (almost there)
This word is put weight on “pat”. There are a few Indonesian words to be taken into account; “sempat” << sem+pat = step forward + nearly there = still be able to or still have a chance, “dapat” = be able or may be, “tepat” = precise, “tempat” = point or spot. Therefore, “pat” can be summarized as ‘something slim or tiny’ which reflexes in a word “*Sepat”, ‘we are almost there’.
*lima-5 (be united)
As most scholars known very well, this word comes from “kalima” << kali+ma which means 5 fingers and hand, a symbol of struggle ‘raise your fist’. On the other hand, “*lima” points to ‘something getting together or a family’.
The surrounded part refers to a story of living course.
*enem-6 (starting point)
Based on the interpretation that the people who first created their owned counting system should live in a land of plenty water. It is, then, become the reason to start Number 6 with a word “water or source of water”, the same meaning with “*daNum” –water.
*pitu-7 (having direction)
The word “tu” in Indonesian means ‘something pointing out or very strong’ such as “butuh” = need or penis, “tulang” << tu+lang = strong and long shape + atmosphere = skeleton, “batu” = rock, “jatu” = fall down. Therefore, “*pitu” reflexes ‘a water is flowing out of the source with a certain direction or the upper course of river’. Furthermore, its meaning can well be compared with the Number 7 of Indonesian “tuju(h)” = heading or direction.
*walu-8 (getting through)
“*walu” comes from 2 elder words as “wa+lu” where both are the ancient affixes of Indonesian and Tai. “wa” means ‘spreading over something’ such as “bawa(h)” << ba+wa(h) = shoulder + something spreading over = bring along or under and “sawah” << sa+wah = myself + something spreading over = rice field.
“bawa(h)” and “sawah” are very important words because they reflex the ancient marriage cultures of southeast Asian in particular Tai-Kadai and some places in Indonesian. Traditionally, a man should not receive any asset from his parent but all of asset including rice field should be given to a woman. The groom has to work for the bride until her parent accepted (sujit wongthes2013). In Tai, “bawa(h)” is shorten to “baw” –บ่าว = a man who is slave or a groom and “sawah” is shorten to “saw” –สาว = a bride or a single woman.
Therefore, “*walu” means ‘something spreading over + something pass through = the lower course of river where spreading over the flooding plain area’. The meaning is not much differ from the Number 8 in Indonesian “lapan” where “pan” means a sheet-like body such as “papan” = board, “pantai” = shoreline, “panjang” = long or extensive.
“*Siwa-9 (almost there)
Although the meaning of “Si” could not be identified for sure but not “wa”, it can be assumed that “*Siwa” reflexes ‘the delta plain area, a place where the river will meet with a mother sea’. It is very similar to Indonesian word for Number 9 “sembilan” where “sembil” or “sambil” means a moment or in between. It should be noted the coincidence in common meaning between “sempat”-4 and “sambil”-9, something in between and almost there.
“*sa-puluq”-10 (be united)
“*sa-puluq” comes from 3 words as sa+pu+lu(q). “pu” is one of the oldest word of Tai-Kadai and Austronesian. Its meaning is a very simple ‘something arising’. Tai-Kadai still use “pu” in daily living, meaning to a man –ผู้, mountain –ภู, spring water- พุ, something arising –ผุด ปุด.
Therefore, “*sa-puluq” means ‘myself + something arising + something pass through = the place where all rivers get together as a mother sea or 10’.
As explained, 8 words are all living fossil preserved in Indonesian (a part of Malayo Polynesian) and 5 preserved in Tai (a part of Tai-Kadai), nothing to be done with diversity.
We can use 8 basic living fossil words to construct the numbers 1-10 without applying any complicated calculation as Laurent Sagart has tried to develop. It is a strong reason that why many of Austronesian can preserve only numbers 1-5 and can not preserve numbers 6-10 due to changing of living environment.
Tai monosyllables are themselves simplified from older polysyllables, with rather good sound correspondences. They cannot be taken as roots from which polysyllabic AN forms are derived.
Natural languages aren’t constructed according to a philosophy. They are much more random than that, often with several processes of more or less philosophy-like analogy operating on the same word – in different directions!
@Minus273
Yes, it is a typical characteristics of people speaking native Tai in particular Thai to simplify polysyllabic words, even loan words. You can find many loan words from Sanskrit were simplified by Thai, but not Indonesian who can preserve the loan word’s form for a much longer period of time. Apart there are many of proto monosyllabic words which Tai and Indonesian share together until todays as shown in the counting numbers.
@David Marjanović
Yes, natural languages are constructed randomly but can not apply for all. The AN basic numbers 1-10 is a good sample, it implies that there is a philosophy behide the scene. They were constructed systematically by a people who had a knowledge and knew wording very well. The numbers 1-5 might be developed first and followed by 6-10 on the same philosophy.
No; your hypothesis requires a huge amount of unrealistic assumptions. Learn more about sound changes.