http://www.economist.com/node/18557572?s...d=18557572
http://www.newscientist.com/article/dn20...frica.html
http://www.babelsdawn.com/babels_dawn/20...frica.html

His theory is that all languages come from Africa. It rests on the assumption that "as groups of people ventured ever farther from their African homeland, their phonemic repertoires should have dwindled, just as their genetic ones did."

So, for example Japanese uses less phonemes than Korean because it's further from Africa. And he sees a correlation in all the world languages where they have more diversity the closer they are to Africa.

I'm sure there's more to it, but it does not sound very convincing.

Firstly, that doesn't really look like a good 'correlation' to me. It's about as similar to me, as those who believe that star constellations actually resemble whatever thing they're meant to. But the whole thing seems a little baffling, relating a (supposedly/assumed) linear relationship in phoneme diversity in standard deviations to distance in kilometers? How is he so sure that the relationship in phoneme diversity is linear? There is no reason why it should be linear.

Secondly, as far as I know, I think most anthropologists stay well away from 'linguistic fossils' if you will. For nothing is more transitory in nature than phonetic structures in languages, so they'd be the last thing that one should actually use.

And aren't language groups/families grouped not by their lexical similarity, but rather their grammatical overlapping? nestor? help

liosama Wrote:And aren't language groups/families grouped not by their lexical similarity, but rather their grammatical overlapping?

Neither of these is the primary criterion for grouping. Language families are, quite simply, groups of languages that are descended from a common ancestor language. This common descent is determined by the methods of comparative linguistics. Both vocabulary and grammar can undergo large-scale changes over the centuries - for example, within the Indo-European family most languages are fusional and nominative-accusative, but German and Persian are somewhat agglutinative and English and Afrikaans are borderline isolating, while Hindi is split ergative. However, the systematic changes that led to the current state of these languages have been traced using the comparative method to a single ancestor language, so they are part of the same family.

Language Log had a post on this paper, if you want to read some discussion of it by other linguists.

I'm pretty sure the reason Japanese uses less phonemes has much more to do with geographic isolation, sakoku, and an inflexible writing system that heavily compresses any phonetic variation in loan words.

I imagine amount and type of contact with speakers of other languages would also account for phonemic variety in other languages too. A greater phonemic variety being preserved in Africa probably has more to do with less emergence of homogenising forces (large empires etc).

The emphasis on a quantitative and co-evolutionary approach is ideal, but in practice the paper and its supporting materials take far too many liberties and make too many presumptions to impress me.

Still, it's nice to see certain ideas permeating. I have some areas I'd like to see it go, in particular, but they're just part of my arbitrary pet theorizing, so. ^_^

@liosama

You might find this stuff interesting: http://simon.net.nz/ - That ‘phylolinguistics’ talk looks interesting, I'll have to see if there's a video of it.

Most recently: http://simon.net.nz/articles/evolved-str...niversals/

Here's an interesting piece from Gray, one of the authors, who's apparently known for championing certain methodologies in linguistics that have been so much fun in dismantling outdated models....

Language evolution and human history: what a difference a date makes

Bonus: Because there's a reference to Chater & Christiansen and it's relevant to the original post: http://www.replicatedtypo.com/science/mo...aphy/2100/ (& This earlier post on the same topic, because it references ‘niches’.)

Edit: Now that I'm noticing that Atkinson's been a mover and shaker in this field (I need to stop mentally abbreviating author list to ‘et al’ ;p), I'm willing to give him a bit more benefit of the doubt, though the paper definitely suffers, I think, from its brevity/narrow focus. I think perhaps if there'd been multiple authors it would've been more impressive and nuanced.

Thankyou so much guys.
Eikyu: I was meant to thankyou in my original post but I was late for the bus to work (I missed it actually).

Von: My understanding though is, a change in grammatical structure of a language is by far much more gradual and slow process versus a phonetic/lexical change which as far as I've read come quite rapidly. A very very crude and rough example to demonstrate what I mean: Do a quick google translate from language A to language B that are both in the same family both come out pretty well, then do the same thing to languages further and further apart from the language tree, the translations become worse and worse as you branch off into completely different trees. I haven't done exclusive testing on this. [I'll reemphasize *crude* here]

pm215: Thanks! I was going to check out langlog later, I would've figured they'd be the first to write about it.

And nest0r: Thankyou so much. As a matter of fact I'm actually quite interested about this all. Will read through them all at work


God it's good to be back



Edit: Reading through langlog now: I'm a prophetic genius.

Quote:But there's something about Atkinson's "Total Phoneme Diversity" that should strike you as odd. Tone, vowel, and consonant "diversity" are weighted equally, although the numbers of alternatives and the contribution to syllable- or word-level "diversity" are radically different in the three cases. Thus losing a single tone would generally reduce "Total Phoneme Diversity" by as much as losing about 10 consonants would. Worse, the lost tonal feature would probably give us only one pair of phonological "alleles" per syllable, at most, while the consonant choices would probably be available (to different extents in different languages) in several places in a syllable, so that the equivalent reduction in "phonemic variants" — seen as alternative symbols in different functional locations – might be numbered in the hundreds.

So let's take a closer look at how "Total Phoneme Diversity" was calculated. The data comes from M. Haspelmath, M. S. Dryer, D. Gil, B. Comrie, Eds., The World Atlas of Language Structures Online (Max Planck Digital Library, Munich, 2008), where (for plausible typological reasons) the phonological inventories of up to 567 languages are treated in a coarsely granular way.

liosama Wrote:Von: My understanding though is, a change in grammatical structure of a language is by far much more gradual and slow process versus a phonetic/lexical change which as far as I've read come quite rapidly. A very very crude and rough example to demonstrate what I mean: Do a quick google translate from language A to language B that are both in the same family both come out pretty well, then do the same thing to languages further and further apart from the language tree, the translations become worse and worse as you branch off into completely different trees. I haven't done exclusive testing on this. [I'll reemphasize *crude* here]

Funny that you mention Google Translate: in fact I have spent most of last night and some of the time in class today translating a huge article from English into Russian by copying paragraphs into GT, then taking their translations and correcting them. Even though I consider myself fluent in English I have a hard time translating, the biggest problem being choosing the most appropriate word. GT has been a great help with that, most of the time the terms that I cannot think of in Russian are translated spot on, and whenever the word is translated wrong it's so ridiculously wrong it's obvious. On the other hand, grammar is a real pain in the arse - not a single sentence comes out without a wrong case, a missing preposition or with an awkward word order (unlike English, the Russian word order is pretty much free, but there are still stylistic limitations). I seem to be spending roughly equal amounts of time correcting the grammar and replacing the badly chosen words, and Russian and English are still in the same Indo-European family.

Okay, the above paragraph was more about venting my frustration over this pointless task than about countering your point, but I still think that Russian and English are great examples of how grammatically different two languages within the same family can get (although Indo-Iranian languages might be even more different, but I don't know much about their grammars). I get your point though, I do realize that if it was Russian to Ukrainian I probably would have been done last night and if it was English to Japanese translating whole sentences in GT would actually slow the process down, as opposed to just translating the individual words I have trouble with. Still, the fact remains that grammatical similarity alone is not an indicator of relatedness - after all, Korean and Japanese grammars are said to be almost eerily similar, and yet their relationship still has not been conclusively proven by the comparative method. And BTW, words are not the only thing that languages can borrow from each other, grammatical forms can also travel between unrelated or distantly related languages within a Sprachbund.

Now, back to my article. Still fifteen pages to go and the deadline is tomorrow evening

Yes but Indo-European is a pretty broad language family and encompasses so many languages. And Russian falls under the Slavic family with all those other Slavic languages while English is under Germanic, try English with any other Germanic language, Dutch, Norwegian. It comes out pretty well.

Also excuse my ignorance but isn't Ukranian basically Russian but with a few different words? As for Korean and Japanese, they're both in huge debate till today. From what I know, only the elementary parts of the languages are (almost) identical. According to some places I skimmed, and some people I know, they say the difference is huge when you get to the meat.


Thanks for the link!

I'm somewhat impressed with the level of linguistics competence that has come to the forum

re: lack of Japanese phonemic diversity.
It used to be a lot more diverse. I read an interesting article a few weeks back about how a historical linguist was finding multiple pronunciations for what are now considered the same kana (or more correctly, what were formerly considered to be simply alternative manyogana for the same sound, and not a different sound). Of course, all we know now is that there was a distinction and not how they actually pronounced it. Why didn't those silly ancient Japanese use IPA?

re: Japanese is Korean
Japanese and Korean probably share a common, long-dead ancestor, but subsequent development and influence (such as from contact between Yayoi and Jomon) has caused a lot of drift, aside from mutual adoption of Chinese compounds much later on. I'm guessing the lack of a written language for most of Japanese history accelerated that drift.

I used google translate to translate some French articles to English; of course it needed a lot of cleaning up, but it took a lot of the grunt work out of looking up words and it did a decent job.

Jarvik: Are you talking about the kou/otsu vowel distinctions or is this something newer? (The huge changes in phonology can really be seen by looking at some of the romanization transcription methods people have used for the Man'yoshu and other works from that period.)

liosama Wrote:Yes but Indo-European is a pretty broad language family and encompasses so many languages. And Russian falls under the Slavic family with all those other Slavic languages while English is under Germanic, try English with any other Germanic language, Dutch, Norwegian. It comes out pretty well.

Also excuse my ignorance but isn't Ukranian basically Russian but with a few different words? As for Korean and Japanese, they're both in huge debate till today. From what I know, only the elementary parts of the languages are (almost) identical. According to some places I skimmed, and some people I know, they say the difference is huge when you get to the meat.

I was using the term "language family" in its older sense, which is still used in Russian and some other languages, of the broadest set of languages conclusively proven to have descended from a single language. In that terminology Slavic and Germanic languages are groups or branches of the IE family and not families themselves, which is why I used English and Russian as examples of divergence within a family and Russian and Ukrainian as examples of a relatively small divergence.

As for Russian and Ukrainian, the issue is hotly contested and the science is often diluted with politics and patriotic fervour, but the general consensus is that they are separate languages. Personally, army and navy arguments aside, I tend to agree with the consensus. The reasons are to do with grammar and mutual intelligibility. The level of mutual intelligibility is rather low. Case in point: my father spent a part of his early childhood in rural eastern Ukraine, where the people speak Surzhyk, which is basically heavily Ukrainianized Russian. He and his Ukrainian friends could not understand standard Ukrainian that they heard on the radio, even though their dialect borrowed heavily from it and their accents were unmistakeably Ukrainian. The grammars also have some differences that are significant for the Slavic group (Russian has six cases, Ukrainian has seven, Ukrainian has the -mo 1st person plural verb ending, which is also used in Serbian, but Russian does not; I never studied Ukrainian, so there may be others that I am not aware of). In fact there might be a stronger case for considering Belarusian a dialect, considering that these grammatical features are missing in it and that mutual intelligibility with Belarusian is much higher: there are a couple of Belarusian bands that I like and when I listen to their songs I get the overall gist of pretty much every line, while with Ukrainian songs my understanding varies greatly. The only problem is that Belarusian has about as much in common with Russian as it does with Ukrainian, so there won’t be an agreement on whose dialect it is.

Even if we do accept Russian and Ukrainian as two dialects of the same language, saying that "Russian is basically Ukrainian with a few different words" would be just as acceptable as what you said (but if I said that to my fellow Russians I would probably get lynched). After all, the Kievan Rus', the predecessor nation to Russia, Ukraine and Belarus, had its capital in what is now Ukraine, and before the 19th century Ukrainians referred to their language as "Rusian" (from Rus'; I am not sure where we got the second "s" from) and to the standard imperial Russian as "Muscovite". There are Ukrainian nationalists arguing that it’s the Russians who are speaking a vulgar, bastardized version of the language with a substantial Turkic and Finno-Ugric admixture, just as there are Russian nationalists arguing that Surzhyk is the actual dialect of Ukraine and that 19th century Ukrainian intellectuals who standardized Ukrainian invented the more distinctive features of the language (or imported them from Polish) to make it appear less similar to Russian. I have no stake in this debate, so I’ll just add that if Danish, Swedish and Norwegian can be considered three separate languages, then so can Russian, Ukrainian and Belarusian.

Ahhhh v. interesting. I say this because of the battle with Urdu and Hindi. I've asked countless (staunchly) patriotic Indians, and they refuse to admit that they're the same language but with different writing systems and some words influenced from their Islamic/Hindu origins respectively. The level of mutual intelligibility is really high, so much so that they all love their Bollywood.

And yeah, why am I not surprised that language/dialect is also used as a basis of degradation? The human condition will never cease to amaze me. 'They' speak a 'vulgar', 'crude', 'dirty', 'wrong', 'lower-class' dialect, whereas 'We' speak a 'clean', 'intelligent', 'upper-class', 'proper' dialect. Like they say, a Language is merely a dialect that has power.

I find it quite appalling that they have to go to that extreme. Sure hate each other, that's understandable, but don't go denying science. I don't know Hebrew, but Arabic and Hebrew both being semitic languages are only slightly similar ,they're both related to different aspects of their parent Aramaic more than they are to each other if that makes sense, and people on both sides admit this. There is no "we aren't of the same origin" crap that you see going on with Urdu and Hindi.

Anyway have been at the hospital all day ttyl