August 26, 2009

Bronze Age origin of Semitic languages

Bayesian phylogenetic methods, originally developed for biology, have been increasingly -and successfully- applied to linguistic data in recent years (e.g., for Indo-Europeans, Melanesians, and Austronesian speakers from the Pacific).

The current paper proposes a Bronze Age origin for Semitic languages, ~3 thousand years after the split of European from Anatolian Indo-European speakers. I don't find this particularly surprising, as Semitic has been, until relatively recently, much more geographically constrained than Indo-European, and -due to the early literacy of the populations of the Near East, its post-Neolithic arrival can be observed in the archaeological record itself.

It also explains a facet of Y-chromosome distribution, that I have commented on before, namely the fact that the common Near Eastern haplogroup J2 extends from Europe to South Asia in a "horizontal zone" accompanied with little of its sister clade J1, but in the Near East itself, there is a "vertical zone" from the Black and Caspian seas to Arabia of high J1 frequency. As I have explained recently, the mixed J2/J1 frequency in the central Near East is due to an enrichment with J1 lineages of a population that had (in pre-Semitic times) a high J2/J1 ratio like those of Europe, Asia Minor, and Iran. J1 should not be seen as exclusively Semitic, but it can't be denied that the major factor affecting its current spread has been the arrival of Semites from the South, the latest episode of which involved the spread of Arab Muslims.

The current study also demonstrates that linguistic Bayesian phylogenetics (LBP) has no inherent bias to produce older dates for language dispersals; while the origin of the Indo-European (IE) language family has been dated to the early European Neolithic, and now Semitic to ~6,000 years, the spread of Melanesian languages to Pleistocene times, and of the Austronesian settlement of the Pacific to ~5,000 years. The congruence between LBP and traditional archaeology in all these cases should force IE exceptionalists who cling to the old theory of "steppe horse riders" to explain why, only in the dispersal of IE, it should LBP should have failed.

The paper also has free supplementary data, including a multistate phylogeny (pdf) of Semitic languages (reproduced top left of this post).

(More details to follow after I thoroughly read the paper)

UPDATE (Aug 27):

From the paper:
Furthermore, Eblaite (no Eblaite wordlists were available for our study), the closest relative of Akkadian and the only other member of East Semitic, was spoken in the Levant (specifically the northeast Levant or present-day Syria; Gordon 1997), which is also where some of the oldest West Semitic languages were spoken (Ugaritic, Aramaic and ancient Hebrew). The presence of ancient members of the two oldest Semitic groups (East andWest Semitic) in the same region of the Levant, combined with a possible long interval (100–3000 years) between the origin of Semitic and the appearance of Akkadian in Sumer, suggests a Semitic origin in the northeast Levant and a later movement of Akkadian eastward into Mesopotamia and Sumer (see figure 1 for a map of our proposed Semitic dispersals).
An origin of Semitic in northeast Levant (Syria) would be consistent with the observed east-west cline of decreasing J1 frequency in the Levant; the authors do, however, mention that the possibility for unknown extinct languages of the Semitic language may shift both the age of the language and its place of origin.
Lacking closely related non-Semitic languages to serve as out-groups in our phylogeny, we cannot estimate when or where the ancestor of all Semitic languages diverged from Afroasiatic. Furthermore, it is likely that some early Semitic languages became extinct and left no record of their existence. This is especially probable if early Semitic societies were pastoralist in nature (Blench 2006), as pastoralists are less likely to leave epigraphic and archaeological evidence of their languages.
A pastoralist association of Semitic languages is also consistent with the observed correlation of haplogroup J1 with herders and J2 with settled farmers in the Near East.


Proc. R. Soc. B 7 August 2009 vol. 276 no. 1668 2703-2710

Bayesian phylogenetic analysis of Semitic languages identifies an Early Bronze Age origin of Semitic in the Near East

Andrew Kitchen et al.

Abstract

The evolution of languages provides a unique opportunity to study human population history. The origin of Semitic and the nature of dispersals by Semitic-speaking populations are of great importance to our understanding of the ancient history of the Middle East and Horn of Africa. Semitic populations are associated with the oldest written languages and urban civilizations in the region, which gave rise to some of the world's first major religious and literary traditions. In this study, we employ Bayesian computational phylogenetic techniques recently developed in evolutionary biology to analyse Semitic lexical data by modelling language evolution and explicitly testing alternative hypotheses of Semitic history. We implement a relaxed linguistic clock to date language divergences and use epigraphic evidence for the sampling dates of extinct Semitic languages to calibrate the rate of language evolution. Our statistical tests of alternative Semitic histories support an initial divergence of Akkadian from ancestral Semitic over competing hypotheses (e.g. an African origin of Semitic). We estimate an Early Bronze Age origin for Semitic approximately 5750 years ago in the Levant, and further propose that contemporary Ethiosemitic languages of Africa reflect a single introduction of early Ethiosemitic from southern Arabia approximately 2800 years ago.

Link

204 comments:

«Oldest   ‹Older   201 – 204 of 204
Ardagastus said...

Errata to my last reply: Bithynians instead of Bythinians and civitates (pl.) instead of civitas (sg.)

I agree with Maju that this thread is too long, so I'll leave some scholars to make their points about language shifts, demic difussion and the Indo-Europeanization of Europe.

Here's the review of a book on the spread of farming and language, of particular interest are the chapters signed by Comrie and Zvelebil. The latter, though a supporter of the Anatolian hypothesis, speaks of Celtic, Germanic, Baltic and Slavic lingua francas replacing the languages of the local communities.

J. Nichols is one of those scholars supporting language shift as a main drive of IE expansion. She authored a nice study in Archaeology and Language, here's the introduction and the conclusion (I don't think Google Books will let you read all those pages between).

And finally a good chapter by Roger Blench, though not focused on IE, but showing that language shift is an important phenomenon, a process which should be taken in account by "any convincing model of the relation between language and prehistory"

Ardagastus said...

Are you deliberately obtuse? Hungarian and Tatar did not affect most of Europe, and I specifically spoke about the impact of steppe elements from the east
You're out of arguments and you started insulting, so this will be my last set of replies on your blog. You actually said "The point is that none of the languages that were introduced to Europe from the east [...] had a lasting influence even in a small region of Europe, let alone the entire continent." (emphasis mine). Both the Middle Danube plain and the north-Pontic steppes are at least some "small regions of Europe".

Celtic languages did not originate in the steppe. Celts were not steppe pastoralists. Nor did they impose their language on Anatolia by elite dominance.
You claimed that no languages "that were introduced to Europe from the east " had a lasting influence on Europe and all IE languages (Celtic included) came from the east (even in Renfrew's hypothesis, if you know where Anatolia is).

Irrelevant, since all these were within-Europe population movements of farmers/stockbreeders
If you don't follow the discussion it doesn't make my replies of no relevance. You earlier suggested many IE spreads and also the Celtic invasions (in Balkans and Anatolia in particular) were "folk migrations" and obviously there's no evidence for that.

Nonsense, these were small tribal units that attracted attention because they were troublesome to the Roman Empire. Most Eastern Europeans spoke Indo-European languages (principally Slavic ones) and the various intruders disappeared without a trace.
There's no evidence whatsoever that Northern and North-western Pontic steppes were Slavic speaking until late in the Middle Ages and the only evidence we have for earlier periods shows Iranian, Germanic (Gothic), Turkic languages. We can't all believe in some "invisible Slavs" just to defend PCT-like positions.

It is not a preposterous claim at all as neither today nor at any time in its know history has any great part of Europe spoken a language that originated outside the continent. This is not true for the Eurasiatic steppe, or for the Levant, or for North Africa, but it is true for Europe.
Unless humans invented languagse in Europe and one of the languages born there became PIE after millenia of evolution, obviously the languages most spoken in Europe today originate outside it(for the record, Anatolia is also outside Europe) and consequently Europe shows no particular resilence to changes from outside. Actually we know nothing about all the languages spoken in Europe in prehistory and this Eurocentrist religion of continuity is an intellectual deadend.

Barbarian warrior peoples were often invited into Roman territory, but being warriors is only one part of the equation of "elite dominance". The other part is "dominance". The Galatians did not dominate Roman society despite being "warriors".
Galatians did not dominate Roman society because there were no Romans in 3rd century BCE Anatolia, they dominated Phyrgians, Cappadocians and other Hellenistic peoples.
Galatians were not invited into Roman territory but they were eventually conquered by Romans.

Nonsense, Celts had no special position in the Roman Empire. Certainly no evidence at all that they "dominated" either the Roman Empire itself or any part thereof.
But they had, as most aboriginals in their provinces (which is also why Celtic, Basque, Albanian and such languages survived Romanization) except for the more troublesome ones such as Dacians. They had their local elites, they could preserve their language (being bilinguals), they could preserve their religion, etc. After all, Celtic outcompeted Latin in Britain.

terryt said...

"if we exclude Hittite, the rest of the IE languages have more recent common ancestry than the initial Neolithic colonization of Europe".

Doesn't that rather negate the Anatolian hypothesis? After all the original Neolithic colonisation is the only one that could have spread the languages so widely. Speaking of Anatolia, we know that a steppe language, Turkic, has replaced previous languages there, including several branches of IE related to Greek. Surely earlier replacements further west are therefore quite posssible.

"Afroasiatic is believed to be Neolithic or even earlier".

Unlikely to be earlier, and even then we find arguments over whether many particular languages should or should not be included in it. And Semitic could well have begun its split more recently than the Bronze Age. Although Semitic languages in the horn of Africa may be Bronze age Phoenician and Hebrew for example share a common ancestry merely 3000 years ago, and Arabic may have split from those languages just a little before then.

"There is no known span of 700 years in which any great part of Europe spoke anything other than Indo-European languages".

In spite of your protestations it seems a fair bet that IE was introduced to Europe, so there you have it: a long period when languages other than IE were spoken there.

I understand completely where you are coming from. You are Greek and are convinced that Greeks are totally separate from even their near neighbours, a common belief, and their history goes back a very long time. I saw a program last night about Chinese anthropologists. Most of them are convinced that the Chinese too are separate, descended from a branch of H. erectus that split off long ago. Same belief, perhaps more extreme.

The best bet for the original spread of IE languages is still that they spread with the ability to domesticate and control the horse, presumably before the chariot's invention although that did aid the later elements. But of course the language moved beyond the genetic expansion and the horse eventually spread beyond the IE languages, for example into the Mongolian steppe and so to the Turkic and Mongolian-speaking people.

Anonymous said...

It is strange to me how people want to cling to their racist theories of the origin of J1, and the language the J1ers spoke.

You are referring to a haplogroup whose origin point is the north of the Middle East, many thousands of years older than proto I.E or proto Semitic. J2 is thousands of years older than both those language groups. So why go on about haplogroups, farmers, pastoralists when those haplogroups predate farmers, pastoralists and I.E or Semitic languages by thousands of years. Not logical, Joyce.

Your theory about I.E languages being spoken in Europe in the early Neolithic is somewhat fanciful. I.E languages were not spoken in many parts of Europe in the historic era. Every heard about the Iberians, Pelasgians and other folks who did not speak I.E languages. Even the Etruscans did not speak I.E languages yet they came from the I.E speaking Lydians. No one has answered that anomaly. Frankly it would not take I.E languages some many thousands of years just to cover Europe. A Bronze Age origin to I.E languages in Europe sounds logical so by the time of the Greeks or Romans, their were pockets of non I.E language speakers still present in Mediterranean Europe. The rest of Europe is basically not known as it still was unrecorded and barbarian.

There is an anomaly in that report. All the old Semitic languages are in the north. All the young ones in the south. So both languages and haplogroups originated in the same parts of the Middle East in the ideal zone for farming and about the same time. There must be other reasons, far more logical ones, for the distribution of J2 and J1 in Eurasia. By the way both those haplogroups exist in the three continents which says one thing, they are old, older than languages, older than agriculture and older than fanciful racist theories.

«Oldest ‹Older   201 – 204 of 204   Newer› Newest»