I suppose I should have posted about this on Thanksgiving (turkey, get it?), but I didn’t have the heart. But people seem to expect me to write about it (the e-mails and comments keep pouring in), so I’ve been rolling up my sleeves and preparing a volley of sarcasm. Fortunately, the ever-dependable Des of Desbladet has saved me the trouble:
A lesser journal than Nature might have invited some historical linguists to review the submitted article, but why let being utter bollocks get in the way of announcing newly-broken ground in a shiny new interdiscipline? “Turkish farmers” means “farmers in what is now Turkey,” of course, and we can put that one down to the journalistes, but we will note with not inconsiderable hilarity that:
Gray was encouraged that his research had been supported in the United States by Stanford University’s eminent geneticist Luigi Luca Cavalli-Sforza.
When evolutionary psychologists and prominent geneticists agree, what business could historical linguists possibly have objecting?
Why are people so obsessed with trying to prove unprovable things about the prehistory of Indo-European using nonlinguistic methods (necessarily, since linguistic methods have gone about as fur as they can go)? Until we dig up a six-thousand-year-old account of “Our Journey to the West” in PIE, we’re not going to know. Get used to it and start doing something useful with your scholarly apparatus, Varied Nonlinguists.
Addendum. MM, in the comments, has directed me to an invigorating rant on the subject over at Phluzein, which I happily commend to your attention.
Is it just me, or do these yahoos never turn their methods on language families with well-established dating?
I bet turning ’em loose on various varieties of Romance would be hilarious.
Yes, it would be great if they could prove Proto-Romance was brought to Europe eight thousand years ago via Madagascar.
Phluzein had a nice rant about it too.
Well, if we re-invent the wheel, it might be rounder the second time.
“Turkey” in Spanish can be translated “peru”, so it is also possible that there’s a translation problem from New Zealandish, which is closely related to Spanish, and that what they’re saying is that the English language actually originated in Peru and then migrated north when the land bridge formed in Panama (which perhaps is the Spanish word for “kurgan”). That makes sense to me, since Peru is closer than Turkey to the US, where English was first recorded in writing.
My favorite quote from the article is this:
“The researchers then used sophisticated computer programs to do the analysis and build language trees.”
A geneticist friend once informed me that most people who use these computer programs don’t understand the algorithms used to generate the results, and are therefore unqualified to interpret the results. A sign of this is the presenting of the result as a “tree” implying concrete links all the way back to the root of the “tree” rather than understanding them as “probability distributions” which implies a much fuzzier and less certain connection between the branches.
Curious ? Alibaba and his band of ? what is the true arabic/english name ? The English and Arab sounds do conflick so much. Like Lawrence in Arabic is not understood by the ordinaire Limey when said by an “Arab”
Maybe Language could retaliate by having a creation “science” special issue, with a glowing endorsement from the Pope. It would as much sense…
Here’s the money paragraph of their discussion:
“Eleven nodes corresponding to the points of initial divergence in all of the major language subfamilies were given minimum and/or maximum ages on the basis of known historical information (see Supplementary Information). The ages of all terminal nodes on the tree, representing languages spoken today, were set to zero by default. Hittite and the Tocharic languages were constrained in accordance with estimated ages of the source texts. Relatively broad date ranges were chosen to avoid making disputable, a priori assumptions about Indo-European history.”
Basically, that means all their time estimates come from fourteen data points, deliberately chosen to have wide error bars.
Here’s a recap of exactly what those dates and how wide those error bars are:
Iberian-French [sic]: 450AD-800AD
Italic-Romanian [sic]: 150AD-300AD
Germanic: 50AD-250AD
Welsh/Breton: 400AD-550AD
Irish/Welsh: before 300AD
Indic: before 200BC
Iranian: before 500BC
Indo-Iranian: before 1,000BC
Slavic: before 700AD
Balto-Slavic: 1,400BC-100AD
Greek split: before 1,500BC
Tocharic: 140BC-350AD
Tocharian A & B: 500AD-750AD
Hittite: 1,800BC-1,300BC
Notice all the “before” qualifiers? Those automatically bias the calculations to earlier dates.
I should note that the selection of languages is dodgy as well. Three Swedish dialects, but only one for Hindi? Pennsylvanian Dutch? Koine Greek, which by their own admission was given a default age of zero? And the heavy emphasis on closely related dialects in the tree biases the forward probabilities.
Amazingly enough, the tree itself is bad. They did come up with an Italo-Celtic relationship, but then discarded it in favor of an Italo-Germanic branch, with a more distant relationship to Celtic. So it doesn’t even pass the undergraduate sense test.
Finally, I note that their paper does not give a date for the split between Greek and Armenian, which is rather odd, considering they claim their results support the Anatolian hypothesis.
(I also like the Albanian – Indo-Iranian split they come up with. Also undated.)
Enough crap like this, and they might discredit the usefulness of the method in *genetics*.
C.
All words flow from Karelia, the heart of human civilization. All putting together of words originates in small villages between big Finnish and wide Russian. Time center going both ways such that past was made in there, and future comes from same place. Then each one went speaking over to other places and became new. Verb forms show this. Even the clicks of Xhosa show this. The banjo comes from there, before the deserts of Timbuktu. Princes of polynesian extraction wear ceremonial markings in early photographs, signals that were first used to mark the intonations of Karelian oratory. Cheap masonic imitations also. The Nazca lines begin this alphabetic rune. Amundsen’s letters mention Inuit reverence for seeming nonsense later proved to be verbatim chants preserved. All our rhetoric, all our happy talk, children’s rhymes and drinking songs, carry the tales still. Turkey is a ruse, though a well-intentioned one. Out of the ground of Karelia, like spring water from a spring, in spring, springs our gift this languages, all of them at once in seed there, into the air, spontaneous, like a quack from a duck.
msg, I think there’s a job at the Times waiting for you!
MSG’s post is fairly good, except that he may be unaware of recent research showing that Esthonian is the origin of all languages, Finnish being a late and corrupt dialect of Esthonian.
Yes, yes, but who put the “Honi” in “Esthonian” (East Honia)? That’s right, the Huns. As in Hungary. As in Hungarian.
Which, as anyone would tell you, is Austrian for Finnish.
Karelia 1, Esthonia O.
Question: is Nimrod a Finnish word? Because that whole Tower of Babel thing is either completely wrong, or Moses was confused about the location of said event.
My son was amazed that “Nimrod” was a real word in any language. In his dialect it was a more-colloquial equivalent of “dumshit”.
It’s ba-ack! See Nicholas Wade’s article in today’s NYT. Argh.
C.
msg, you rock. 🙂
Tolkien’s universe is also nothing but a collection of well-ornamented Karelian myths – in case anyone was wondering about the elvish dialects.