Non-Indo-European language families

March 18, 2008 2:28 PM

LinguisticsFilter: Why are the many Indo-European languages considered to have a common ancestor, whereas Chinese, Thai, Vietnamese and other gramatically and lexically similar languages are grouped into separate language families? How did so many different language families arise in the Americas and Australia given that they were populated by the same groups of people? Much more inside.

I've always wondered about the monolithic Indo-European language family that groups together languages as different as Sanskrit, Farsi, and English, stretching from India, through Iran and up through Russia to continental Europe. Outside of this huge geographic swath, it seems the other languages are fragmented into several different language families, despite seeming to my lingustically untrained self to be very similar both gramatically and lexically.

For example, Chinese, Vietnamese, and Thai. They're all tonal monosyllabic SVO languages, don't conjugate verbs for person or tense, use measure words, and in general rely on particles and word order to express meaning. Chinese and Thai both have mei/mai as a negative particle, and le/laeo as a "completion" particle (I think Vietnamese has similar words in meaning and sound). I've also noticed that Chinese and Thai use similar set verb-object phrases, for example, chi-fan/gin-kaeo(both meaning "eat rice" and "to eat" in general), kai-che/kap-rot("drive car"-"to drive"), etc. And all of them have historically had several different pronouns to use depending on social rank (although it has disappeared from Chinese, it's still fundamental to Thai and Vietnamese).

Yet Chinese, Thai, and Vietnamese are grouped into entirely different language families - Sino-Tibetan, Tai-Kadai, and Austro-Asiatic, despite their geographic and linguistic closeness. Or what about Korean and Japanese, which are almost identical in grammar and sound similar (rapid staccato speech), yet both considered lingustic isolates (despite the geographic proximity and similar physical appearance and culture). Did Chinese, Thai, and Vietnamese really develop independently of each other, instead of stemming from a common ancestor? I'm aware of things like the sprachbund effect, but still have trouble seeing why Chinese and Thai are apparently unrelated when Chinese and Tibetan (with its completely different phonology and grammar) are. For what its worth, I remember reading a 1950s lingustic textbook grouping Chinese and Thai together in a "Sino-Thai" family, but that classification appears to be outdated. What did they discover that made them separate them?

I know very little about other languages, but I'm curious why there seems to be so many different language families in Australia or the (pre-Columbian) Americas, for example. Just look at the maps on this page and this page. If I'm not wrong, the people who populated those areas basically migrated as one group across the Bering land bridge, or through Melanesia in the case of Australia. How then did these people, ostensibly speaking the same language, end up developing entirely different language families? Is it even possible to do that? I realize there have been waves of migration, but I would guess that the people doing the migrating came from essentialy the same areas. For example, the Inuit and Polynesians all seem to speak closely related languages, despite living in vastly separated geographic areas.

What caused the diversity in language families outside of the Indo-European phylum? Is it just because linguistic research in the Indo-European languages is the most complete, and researchers haven't fully studied other languages? Or were non-Indo-Euoprean languages historically more fragmented? If the latter, how exactly were the Indo-Europeans so succesful in spreading their languages to areas that were originally non-Indo-European speaking?

posted by pravit to writing & language (46 comments total)

Clarification: Of course, I don't mean that the ancestors of Australian Aborigines and Native Americans were the same people. It's just two examples of the same phenomenon.
posted by pravit at 2:29 PM on March 18, 2008

I'm not sure, but I do remember my Japanese teacher trying to convince us that the differences in Japanese between the north and the south (what I would consider dialects) was much the same as the differences between Spanish and French, since both were descended from Latin.

This, of course, is a load of crap. I considered it nationalistic hubris on her part.

So I imagine it is due to lack of rigorous study.
posted by kpmcguire at 2:36 PM on March 18, 2008

While I can't say much as to the existance of the Indo-European Language which may have existed, I can say that, at the very least, Chinese and Japanese are very different, as I've done some study of both of them.

Chinese is a language which is completely monosyllabic an utilizes tone to also convey a difference in meaning between words. Japanese, however, utilizes tone in the same way that English does (think of how you would say "sure" annoyed versus actually OK with something). Japanese is also polysyllabic.

Similarities in culture are the same reason why Europeans have similarities: conquest and mutual exchange. Cultural similarities can be through more than just language.

I can't say anything about Thai or Vietnamese, so I apologize for my lack of full answer.
posted by Gular at 2:51 PM on March 18, 2008

pravit: Your question is an excellent one, but it can't really be answered in the framework of AskMeFi; you need to study historical linguistics to truly understand what's going on. Short, wildly inadequate answer: the similarities you mention (tonal monosyllabic SVO languages, don't conjugate verbs for person or tense, use measure words, etc.) are not signs of genetic relationship, they're exactly the sorts of things (areal features) that tend to spread through unrelated languages in a given area—another example is the Balkans. To determine genetic relationship, what you need is consistent phonetic relationships. Indo-European was discovered because of examples like Sanskrit pitar-: Latin pater: Greek pater: English father, Sanskrit panca-: Latin quinque: Greek pente: English five, etc. etc. If you have that kind of consistent relationship (p-: f-), it doesn't matter whether the words sound alike, or whether the grammatical categories are the same in both languages—Latin, Greek, and Sanskrit have very complicated verbal morphology with lots of tenses and aspects, English has a comparatively simple one. (And one of my favorite examples of non-obvious relationship is Mandarin wu = Cantonese ng 'five.')

On the general history of languages and how they spread or shrank, I highly recommend Ostler's Empires of the Word. And for this:

how exactly were the Indo-Europeans so succesful in spreading their languages to areas that were originally non-Indo-European speaking?

you want Jared Diamond's Guns, Germs, and Steel.

Feel free to ask me further questions via e-mail if you're interested; I love discussing this stuff.

This, of course, is a load of crap. I considered it nationalistic hubris on her part.

What do you mean, "of course"? Have you done a comparative study of northern and southern Japanese dialects? Maybe it makes sense to consider them closely related dialects, maybe it doesn't, but I don't see how "of course" comes into it. And "nationalistic hubris" makes no sense—nationalists tend to insist that clearly separate languages are merely dialects of the "national language" (see: China).

So I imagine it is due to lack of rigorous study.

If you're talking about your Japanese teacher, that seems irrelevant to this thread. If it's in response to the original question, you don't have the faintest idea what you're talking about.
posted by languagehat at 2:58 PM on March 18, 2008 [9 favorites has favorites]

Is it just because linguistic research in the Indo-European languages is the most complete, and researchers haven't fully studied other languages?

Indo-European is kind of special because you have a variety of really, really old texts to compare: Sanskrit, Avestan, Homeric Greek, Hittite, as well as dozens of languages with extensive literatures to work with. And, those languages tended to be written mostly in alphabetic scripts where you can get an idea of the sounds of the language, as opposed to Chinese, which is in ideograms.

An alternate take on grouping languages has been tried by Greenberg, including putting most languages of the Americas into one superfamily, but he's controversial.

Note that neighboring languages can share lots of vocabulary, and colloquially you could say they're related. Languages that evolved near each other can tend to share what I'll call "habits"--for example Uralic languages (including Finnish) versus Altaic languages (including Turkish) do a lot of similar looking things--even though in the strictest sense they're not considered related by the experts, since hard evidence showing a common ancestor between them hasn't been reconstructed in a way that lots of people would agree with it.

I could probably come up with a more scholarly-sounding explanation if pressed, I'm trying to be accessible.
posted by gimonca at 3:09 PM on March 18, 2008

Note too that several attempts to describe wider language families may have been politically colored. It's been noted that the proposed Nostratic family, popularized by Russian scholars, conveniently seemed to lump together all the peoples of the Russian Empire/Soviet Union into one, big happy family.
posted by gimonca at 3:13 PM on March 18, 2008

I'm sure that languagehat will be along soon enough....

Is it just because linguistic research in the Indo-European languages is the most complete, and researchers haven't fully studied other languages?

It's certainly true that philology and historical/comparative linguistics has a particular cultural vector, one of those 'taxonomy / genealogy of X' disciplines developed with a nod to Linnaeus, and the people studying it -- western Europeans -- were looking back from where they were standing. There's also a lot of intellectual crud absorbed into the mainstream from when the Jesuits went east, tied to ideas of universal language, the Tower of Babel and suchlike, and you can see varieties of 'orientalist' speculation going on right up until the early 20th century.

It's also arguable that the rest of the 20th century shows the flipside, with a degree of hobbling by nationalistic influences: for example, it can be difficult for Chinese linguists to talk of Chinese 'languages' rahter than 'dialects', even when they fail the mutual intelligibility test.
posted by holgate at 3:15 PM on March 18, 2008

Yeah, there are a lot of people who are dying to lump everything together and reconstruct Proto-World. I come from a tradition (Yale School) that is very conservative about that sort of thing: if you can't find consistent relationships of the type I described, you can't talk about genetic connection.
posted by languagehat at 3:17 PM on March 18, 2008

"If I'm not wrong, the people who populated those areas basically migrated as one group across the Bering land bridge, or through Melanesia in the case of Australia. How then did these people, ostensibly speaking the same language, end up developing entirely different language families?"

Nobody knows whether those peoples originated in one group with a common language. It could be equally plausible that several groups, perhaps separated by hundreds or thousands of years, made it to Australia. And those groups need not have spoken a common language.

Further, there has been human settlement in Australia for a very long time; about 40,000 years, give or take. Without one dominant culture to impose/spread its language, and given a vast, sparsely populated area, it doesn't surprise me in the least that languages there are diverse.

"Polynesians all seem to speak closely related languages, despite living in vastly separated geographic areas."

Yeah, but they settled that vast area in less than a couple of thousand years, and managed to retain sea links between islands. That's why the Polynesian voyages are widely considered to be among the great human achievements of antiquity.

Take LH's advice and read Empires of the Word and you'll see that geography and speed of settlement are among the important factors.
posted by i_am_joe's_spleen at 3:20 PM on March 18, 2008

Er, that "Yeah" was to gimonca, though I also agree with holgate's example of Chinese (though not with his first paragraph, if that "cultural vector" and "looking back from where they were standing" is in any way casting doubt on the validity of historical linguistics—I despise postmodern attempts to downgrade science to just the Western way of looking at things, no better than any other).
posted by languagehat at 3:21 PM on March 18, 2008

In addition to agreeing with languagehat, I wanted to address some of your questions in a little more detail.

You write: How did so many different language families arise in the Americas and Australia given that they were populated by the same groups of people?
and
If I'm not wrong, the people who populated those areas basically migrated as one group across the Bering land bridge, or through Melanesia in the case of Australia. How then did these people, ostensibly speaking the same language, end up developing entirely different language families? Is it even possible to do that? I realize there have been waves of migration, but I would guess that the people doing the migrating came from essentialy the same areas. For example, the Inuit and Polynesians all seem to speak closely related languages, despite living in vastly separated geographic areas.

The answer is time depth. The Indo-European family is something like 6,000 years old, but the migrations to the Americas and Australia were much further in the past -- Wikipedia says at least 12,000 and 40-something thousand years ago, respectively. The methods of historical linguistics are able to reconstruct a common ancestor as old as Proto-Indo-European -- in part because we are lucky enough to have texts of ancient varieties of IE languages, e.g. Ancient Greek and Sanskrit -- but not (yet) able to demonstrate relationships between languages that split much further back.

Thousands of years of language evolution are enough to produce Hindi and English from Proto-Indo-European, languages that don't seem very similar at first glance. The languages of the Americas and Australia have had much more time to diverge, so it's not surprising that their degree of relationship remains obscure.

That doesn't mean that relationships of greater time depth won't ever be demonstrated. Just this month, work showing a genetic relationship between the Siberian language Ket and the Na-Dene family of North American languages has been announced (see here and here).

[Or, in summary, what everybody else has said on preview.]
posted by The Tensor at 3:30 PM on March 18, 2008

Note that the Indo-Europeans didn't "spread their languages". They spread themselves. Most of the people of Europe are descended from invaders from the east. Probably the only significant language in Europe to survive from before the Indo-European invasion is Basque, which is not closely (or even distantly) related to any other known language.
posted by Class Goat at 3:33 PM on March 18, 2008

I'm not sure, but I do remember my Japanese teacher trying to convince us that the differences in Japanese between the north and the south (what I would consider dialects) was much the same as the differences between Spanish and French, since both were descended from Latin.

This, of course, is a load of crap. I considered it nationalistic hubris on her part.

Have you ever listened to 津軽弁 and 琉球語? I suspect they're mutually unintelligible. I can't say whether they're as different, more, or less than Spanish & French. And as they say, a language is just a dialect with an army.
posted by adamrice at 3:38 PM on March 18, 2008

And for this:

how exactly were the Indo-Europeans so succesful in spreading their languages to areas that were originally non-Indo-European speaking?

you want Jared Diamond's Guns, Germs, and Steel.

Hmm, languagehat, I think I might have to disagree. While Diamond has a good grasp of "big picture" history, his treatment of any given historical situation, like conflict between agricultural and non-agricultural peoples, is always overshadowed by his hypothesis, to the extent that many useful and interesting peculiarities are erased. If you're looking for the source of Indo-European dominance (as opposed to Asian linguistic fragmentation), you won't really find it in Diamond, who mainly talks about the Age of Discovery in that context--way after the languages were established. It was also my impression, which could be totally off-base, that the quest for a Primeval Indo-European Civilization is a dusty relic of the nineteenth century--in which case linguistic expansion won't be explainable very well in Diamond's terms.

(and, while I'm no expert on sixteenth-century colonization, I think his treatment of Pizarro and the Incas is absurd; the numbers given in his primary sources don't seem any more credible than Herodotus' OMG ONE ZILLION PERSIANS--yet he takes them totally for granted.

Collapse, for what it's worth, is a lot better.)


Instead, I recommend McNeill and McNeill's The Human Web. It's encumbered with its own not-very-clever hypothesis, but its treatment of cultural and societal interactions is a lot more subtle than anything Diamond provides. While I don't recall if they take up the linguistic question directly, the way they articulate the development of civilizational "webs" in Europe and Asia makes a lot of sense if applied to the evolution of language groupings. (I think. It's been a few years, and I don't know much about linguistics). It's also an easy and fun read.
posted by nasreddin at 4:15 PM on March 18, 2008 [1 favorite has favorites]

I'm prompted to wonder if the "differences in Japanese between the north and the south" is talking about the distinction between Japanese and Ryūkyūan. (upon Googling, Ryūkyūan is one of the things adamrice mentioned.)
posted by XMLicious at 4:18 PM on March 18, 2008

Excellent answers so far, thanks! I guess I really need to study historical linguistics more seriously to really understand anything, but what is it about areal features that makes them areal? That is, why are people who live near each other more likely to adopt similar grammar patterns or speak in tones, but less likely to adopt similar-sounding words?

Thanks letting me know about Empires of the World - when I saw the subtitle ("A Language History of the World") I knew this was something I have to read!

Could you say that the reason for the linguistic diversity in the Americas/Australia is because those groups that came over tens of thousands of years ago ended up settling in the same place and staying there, without really contacting other groups? I would have assumed that with migration/trade/wars that one language family would become dominant. This is getting out of the scope of this AskMe, but why didn't those regions see the same massive wars for territory that the Old World did?
posted by pravit at 4:21 PM on March 18, 2008

Japanese wise, I watch tons of anime... while not as common as the "hit the melon on the beach", or the "visit the onsen (hot springs)" trope, North vs. South and not being able to understand each other is pretty common. I'm not sure if it's dialect or different language.

In the US, I get the same thing going home (the South) after living in California for 20 years, it takes me a couple of days before I can understand my family (probably dialect here)... I go "what?" for a few days.

Different verb conjugations, different contractions, different pronounciations, the line between dialect and language is flexible.
posted by zengargoyle at 4:22 PM on March 18, 2008

Could you say that the reason for the linguistic diversity in the Americas/Australia is because those groups that came over tens of thousands of years ago ended up settling in the same place and staying there, without really contacting other groups? I would have assumed that with migration/trade/wars that one language family would become dominant. This is getting out of the scope of this AskMe, but why didn't those regions see the same massive wars for territory that the Old World did?

Well, Mesoamerica had savage wars for territory until the Spanish showed up. And I suspect a lot of the linguistic diversity in South America was due to the Incas, who never had serious military challengers and who maintained a non-assimilationist policy with respect to the tribes they ruled over. (Unlike, say, the Romans).

And, of course, inaccessible terrain and small-scale non-agricultural societies probably helped.
posted by nasreddin at 4:27 PM on March 18, 2008

I realize there are many Japanophiles eager to share their knowledge, but the whole north-south Japanese thing is a derail, so I'd appreciate if we could drop the topic. I'm not asking about dialectical differences within a single language (and certainly not Japanese for that matter), but why language families outside of Indo-European seem so fragmented. I'd appreciate comments on why Japanese and Korean are classified as unrelated, however - is it really just areal features?

If you look at those maps I linked to, you notice huge linguistic diversity in California and northern Australia. Why is that? Are there geographical barriers to prevent groups from contacting each other (e.g. the mountainous regions of SE China)? I'm not an expert in Californian geography, but I don't think the area is really covered with impassible mountains or anything.

Also, thanks for the book recs, Nasreddin!
posted by pravit at 4:36 PM on March 18, 2008

In the New Guinea highlands, which are right up there for sheer profusion of tiny, different languages, extremely difficult mountainous terrain is the probable reason, because low-level warfare is endemic as tribes compete for the very scarce resources (game, horticultural land). Before the intervention of modern technology from colonists, no one could maintain control over a bigger area than the local valley.

In Australia, I don't know what the level of intertribal conflict was historically, but it's pretty clear that no group secured the sort of access to natural or technological resources that enabled them to dominate others. And ideas spread pretty slowly when you can only travel on foot or by paddled canoes.
posted by i_am_joe's_spleen at 4:49 PM on March 18, 2008

One of the top results for "linguistic diversity in California".
posted by i_am_joe's_spleen at 4:56 PM on March 18, 2008

Could you say that the reason for the linguistic diversity in the Americas/Australia is because those groups that came over tens of thousands of years ago ended up settling in the same place and staying there, without really contacting other groups? I would have assumed that with migration/trade/wars that one language family would become dominant.

It doesn't usually work that way. You underestimate the extent to which having a separate language is part of tribal identity and ethnocentricism. (Think "Quebecois French", for instance.) It's also difficult to learn a new language, and there's really no significant incentive to do so and then use that other language in place of the one you grew up with.

It isn't even always the case that conquered people come to adopt the language of their conquerers. The Normans learned Saxon; the Saxons didn't really learn Norman. Old English (the result) was mainly Saxon (an Germanic language) with a thick coat of Norman vocabulary on top. English is not Romance, even though Norman was Romance.

There's a lot of inertia, simply due to the difficulty and expense of learning a new language. It can take centuries for a conquerer to kill off a conquered people's native language. (Welsh is still spoken in Wales. You can still find people in Scotland and Ireland who don't speak English.)
posted by Class Goat at 5:57 PM on March 18, 2008 [1 favorite has favorites]

You underestimate the extent to which having a separate language is part of tribal identity and ethnocentricism. (Think "Quebecois French", for instance.) It's also difficult to learn a new language, and there's really no significant incentive to do so and then use that other language in place of the one you grew up with.

Right - but I'm talking about language families, not just individual languages/dialects - groups of languages which are supposedly completely unrelated from each other, and stem from different ancestors. For example, most of sub-Saharan Africa is dominated by Bantu languages, despite the people in these areas historically living similar lifestyles to Native Americans and indigenous Australians. Each tribal group may speak its own language, and it may be very important to their identity, but the language is still related to the other Bantu languages that surround it. You don't see a dozen or so little independent language families coexisting in an area the size of California like you do in the Americas or Australia. The existence of so many language families(again - groups of languages that are entirely unrelated) implies that there must have been something to keep them from mixing with each other for thousands of years. But what?

I think we're talking about different things - I'm talking about change on the order of thousands of years. As people upthread have pointed out, massive Indo-European migration wiped out the native languages of Europe (like Etruscan and Iberian), save for Basque. I imagine something similar happened in Africa - given that it's the ancestral starting point of all humanity, you would expect to see more language families represented there, but Bantu seems to have gained the dominant position.
posted by pravit at 6:14 PM on March 18, 2008

"massive Indo-European migration wiped out the native languages of Europe"

That's one theory. There are others, somewhat controversial and perhaps discredited, that suggest that the spread of IE was based in the spread of IE-speaking farmers, not necessarily by conquest, but generation by generation extending the borders of their farms, teaching their neighbours, and exporting their language and culture. Colin Renfrew is the archaeologist who proposed that in some detail.

Some of this stuff is literally pre-historic and therefore subject to continual review and debate as new evidence comes to light. The idea of heroic mounted Aryans spreading their horsey patriarchy all over the place waxes and wanes.

This is why people who really know this stuff are hesitant about making definite, unqualified statements, even though they may personally regard a particular view as the most sound, and then they have to grind their teeth while less competent but more confident lay people like me rush in.
posted by i_am_joe's_spleen at 6:38 PM on March 18, 2008

That is, why are people who live near each other more likely to adopt similar grammar patterns or speak in tones, but less likely to adopt similar-sounding words?

If you use devices that other nearby languages use, it makes it easier to learn them and/or for their speakers to learn yours. There may be an acquisition thing going on: If two languages are doing similar things, and enough people encounter just enough of the other language to assume they're doing the same thing, they may learn it that way, speak it that way, interact with its native speakers that way, and gradually pressure it into being the same.

In fact, when languages are in contact, they do tend to borrow similar-sounding words as well. But vocabulary borrowing does not make for family membership. As languagehat pointed out, families are reconstructed on the basis of (largely) consistent relationships between the sounds of the daughter languages, which isn't always the case with borrowings. Also some word classes are much more resistant to borrowing than others: Regularly inflected verbs, nouns, adjectives, and adverbs are easy-come, easy-go, but irregulars, auxiliaries, modals, pronouns, prepositions, articles, and the like are much more strongly conserved. When people adopt those words from their neighbors as well, it's generally because they're switching languages entirely.

Still, the question of why some features go areal is a fascinating one, not nearly answered, so you raise a good point.

why didn't those regions see the same massive wars for territory that the Old World did?

The Eurasian steppe is huge, it's flat, it's horse- and wheel-friendly, and it's at the same latitude (therefore similar climate and biology, similar ecological and cultural opportunities) all the way across. I suspect this has something to do with it.
posted by eritain at 8:02 PM on March 18, 2008

I despise postmodern attempts to downgrade science to just the Western way of looking at things, no better than any other

Then you're greatly overinterpreting my point, which was just to say that historical linguistics originated in a particular time and place, and its early focus grew outward from the stuff of local expertise. Grimm's Law was an early, seminal breakthrough assisted by the fact that its documenters were German; it's not 'postmodern' to suggest that had modern historical linguistics come out of China, that consonant shift would have been arrived upon by a different path of intellectual discovery. The genealogical impulse in European biology got people looking at language in a similar way, and when they looked closely enough in the right way, they discovered things. That's scientific method.
posted by holgate at 8:09 PM on March 18, 2008

"You can still find people in Scotland and Ireland who don't speak English."

Unless "don't" means "unwilling to", or you're counting children who haven't hit the education system yet, that isn't true.

According to Wikipedia, there haven't been any monoglot Gaelic speakers in Scotland since 1981 or before.
posted by i_am_joe's_spleen at 8:31 PM on March 18, 2008

It seems my information is a bit out of date. I thought there were still a handful of very old people in Scotland who didn't speak English.
posted by Class Goat at 9:18 PM on March 18, 2008

I imagine something similar happened in Africa - given that it's the ancestral starting point of all humanity, you would expect to see more language families represented there, but Bantu seems to have gained the dominant position.

There's quite a bit of research (Ehret, Sutton, Holl, Petite-Maire, Muzzolini) that ties in climate change in the Sahara and Sahel between 6000 and 4000 BCE, changes in subsistence strategies (pastoralism spreading), and the eventual Bantu migrations (that may have been influenced by climate change) with the dominance of Niger-Congo languages (to which Bantu belongs).
posted by anansi at 9:56 PM on March 18, 2008 [1 favorite has favorites]

*Full disclosure: I'm currently working with Ehret and Posnansky, Sutton's mentor. So I might be a wee bit biased. But seriously, its compelling stuff.
posted by anansi at 10:06 PM on March 18, 2008

That was supposed to be really tiny font. It looked good in preview, but apparently my HTML skills suck.
posted by anansi at 10:07 PM on March 18, 2008

Pravit: re. Chinese and Asian languages.
Since the 1920s, 1930s, up till today, there has been deep and enlightening work on the Asian, esp. E. Asian languages. Much of it's in Chinese and Japanese, so I won't bother mentioning the names of the famous ones, since I assume you don't read those langs. But in English, check the works of William Boltz, Edwin Pulleyblank, Jerry Norman, and Roy Andrew Miller. They utilize the Chinese and Japanese scholars and go forth in their own right. Also, they cite important European linguists on this topic as well. Much of this literature is in journal articles and book chapters, so you'll need some research aids to find authors' bibliographies. Certain essentials (and theories of origins) of south and southeast Asian languages, and of Japanese and Korean, remain difficult and are unresolved. Also, ditto re. Central Asian langs. -- much will remain unresolved, since very little of the proto- and ancient-Turc and proto-Mongolian, and early Manchu, Khitan, Jurchen, etc. languages was ever written. Good luck.
posted by yazi at 11:07 PM on March 18, 2008

Again, great great answers all around. Thanks everyone for pointing me in the right direction. Also, yazi, I do read Chinese (no Japanese though), so I'd appreciate if you could mention a few of those works you mentioned.

I'm still very curious about the many different language families in northern Australia and California - can anyone shed more light onto this?

It's times like these that really make me wish I went into historical linguistics instead, but I can always be a hobbyist, at least...
posted by pravit at 11:43 PM on March 18, 2008

"The Normans learned Saxon; the Saxons didn't really learn Norman. Old English (the result) was mainly Saxon (an Germanic language) with a thick coat of Norman vocabulary on top.">

You mean Middle English.
posted by litlnemo at 11:55 PM on March 18, 2008 [1 favorite has favorites]

I'm still very curious about the many different language families in northern Australia and California - can anyone shed more light onto this?

Look at it this way: it takes evidence to prove two languages are related. When you don't have enough evidence to settle the question either way, the default is to assume they're unrelated.

So it's not like we've got proof the dozen or so language families in California are unrelated. In fact, just the opposite — we've got no proof at all in either direction. For all we know, they had a common ancestor 6,000 years back, or 10,000, or 30,000, and there's just been too much change since then for us to see the resemblance.

(Both these areas got writing fairly recently, by the way, and I think that's significant. Without the help we got from Hittite, Sanskrit, Greek, Latin and so on, we wouldn't have been able to reconstruct proto-IE either, and we'd be stuck with the default hypothesis that the languages of Europe and Asia belonged to a dozen different families.)
posted by nebulawindphone at 1:17 AM on March 19, 2008

you want Jared Diamond's Guns, Germs, and Steel.

I remember in particular a couple of maps of language groups and such in Asia in his discussion of language in that book (I'm 90% sure it was that book, at least) that address some of your questions here. Worth a look, if only for that, for a readable and interesting discussion of the subject of East Asian languages.
posted by stavrosthewonderchicken at 2:26 AM on March 19, 2008

For what it's worth, in terms of Korean in particular, an estimated 60% (some say more) of Korean vocabulary is of Chinese origin (although often there are Korean equivalents), and although 한글, the Korean alphabet, has been in use to one degree or another since and Chinese is only rarely used (mostly newspapers and such, although Korean school kids are still required (or were until recently, I can't recall if they've changed this or not) to learn 2000 common Chinese characters), Korean was written with Chinese characters before that time, and the influence not only in culture but in language cannot be underestimated.
posted by stavrosthewonderchicken at 2:29 AM on March 19, 2008

(It might be interesting to compare classification in botany. Linnaeus did the best he could with the information he had at hand. Today, we have access to deeper information, and the classification of plants looks very different in some places, thanks to people like the APG.)
posted by gimonca at 5:41 AM on March 19, 2008

Then you're greatly overinterpreting my point

No, no, I was just saying I hoped that wasn't what you were saying. Your clarification is perfectly sensible and I agree with it, so I'm glad I asked.

Look at it this way: it takes evidence to prove two languages are related. When you don't have enough evidence to settle the question either way, the default is to assume they're unrelated.

Exactly. Some people seem to have an emotional need to believe We're All Connected, so they twist and overinterpret the evidence to "prove" what they want to believe, but that ain't science. The sad fact is that outside of a few well-documented families like IE and Semitic, there will never be enough evidence to prove distant relationships; all we can say is "maybe." Unsatisfying to romantic spirits, but that's science.

This is a great discussion!
posted by languagehat at 7:17 AM on March 19, 2008

Nebulawindphone, excellent answer - that really clears things up for me and kind of confirmed my suspicions - that all those language families may very well be related, but we have no way to prove it. Perhaps those groups of indigenous Australians have been living in semi-close proximity for tens of thousands of years and the relationships between their languages have been blurred so much that they don't really fit into our language taxonomy.

And stavros, I've always been fascinated with the Chinese influence on the other East Asian languages - it's amazing to see how much Chinese vocabulary Korean, Vietnamese, and even Japanese have absorbed and worked into their grammatically very different languages.

Anyhow, I'll pick up Empire of the Word at the library today or tomorrow - looking forward to reading it!
posted by pravit at 11:37 AM on March 19, 2008

"Perhaps those groups of indigenous Australians have been living in semi-close proximity for tens of thousands of years and the relationships between their languages have been blurred so much that they don't really fit into our language taxonomy."

That is one common view of the Australian situation, where various features like ergativity are widespread and may be what's called "areal" features. I'm just racking my brains trying to think of the name of the Australian linguist who wrote a good short text on this. Might have been Dixon, although I remember reading this book in 99 or so, and his book came out in 2002.
posted by i_am_joe's_spleen at 12:24 PM on March 19, 2008

Re the Japanese thing: There actually is some interesting evidence to suggest that Japanese is related to a Korean language -- just not the one that survived to dominate the peninsula and become the "Korean" dominant there today. Try googling "Goguryeo".

There's also a more general theory that Japanese and Korean, and maybe also Ainu, are part of an "Altaic" family, which would also include the Turkic, Mongolic, and Tungusic languages.

And one of the greatest living scholars of Old Japanese, Ono Susumu, not a crank by anyone's standards, is absolutely convinced that Japanese is related to Tamil.

So, as wiser posters than me have already noted, it's not that linguists are refusing to admit relationships out of tradition or bigotry or anything. Quite the opposite -- they're so eager to find relationships that several mutually contradictory theories are on the table. But no-one's managed to prove anything. Yet.
posted by No-sword at 4:51 PM on March 19, 2008

Sorry, just noticed that I missed a bit. Should have said '한글, the Korean alphabet, has been in use to one degree or another since its creation in the 15th century.'
posted by stavrosthewonderchicken at 10:28 PM on March 19, 2008

I found a very large version, in color and everything, of that map I was talking about, and uploaded it to my site. Here you go.
posted by stavrosthewonderchicken at 3:56 AM on March 22, 2008 [2 favorites has favorites]

At least 10 years ago there was an absolutely amazing article on linguistics and the spread of the Indo-european languages in the Atlantic Monthly. It was mostly about how linguists identify the root words (and the codes they use to indicate what's a root, what's borrowed, etc.) but also dealt with cultural and migration issues. I have been looking for this article again ever since, and would be in the debt of anyone here who can find it. Help me, obi wan languagehat, you're my only hope!
posted by nax at 6:30 PM on March 22, 2008

stavrostwc-- whoa. Gorgeous.
posted by nax at 6:32 PM on March 22, 2008