zompist bboard

I've recently learned of the Indo-Pacific language family theory, which proposes links between the Dravidian languages and Korean (and apparently some other languages too). Apparently, there's a large number of very similar words between Tamil and Korean that are potential cognates, and that's where the theory came from.

Not knowing anything about this theory at all, I'm curious what you guys think! Is there any possibility of truth to this, or is it completely ridiculous? Or is this one of those theories that might be true, but is unprovable?

It's completely ridiculous. I like to joke that everybody likes to connect Dravidian languages with all kinds of other language families because our languages are just so damn weird.

Vijay wrote:It's completely ridiculous. I like to joke that everybody likes to connect Dravidian languages with all kinds of other language families because our languages are just so damn weird.

*our* languages?

There seem to be two completely different "Indo-Pacific" families on Wikipedia. At first I was assuming you were talking about the more easterly of the two, not having heard of the other one. What I wrote below was originally based on my assumption that you were talking about the Greenberg proposal, but I've edited it now that I understand what you're talking about. Still, my opinion of both proposals is very similar.

alynnidalar wrote: Is there any possibility of truth to this, or is it completely ridiculous? Or is this one of those theories that might be true, but is unprovable?

The maturation period required for proto-Japanese/Korean/Ainu to merely be a sub-branch of the family is so deep that the proto-Tamil/JKA language would likely have very few surviving cognates, no more than one would expect to appear by chance.

I'd like to see the list of cognates.

Wikipedia wrote:But recent genetic studies show an early South Asian genetic input in Korean and especially in Japanese people.

The genetic connections between the two might be attributable to a common substratum population that lived in southern India and Japan but not in northern Asia, from where the founders of China came. Most anthropologists believe that Asia was first settled by people who moved slowly west to east along the most tropical shorelines, so any humans who reached Japan would have had ancestors who had lived in India.

It is known that Japan remained an aboriginal (Jōmon) refuge until fairly late in history, with anatomically modern Japanese people only appearing after around 300 BC, very late compared to surrounding nations. (By contrast, the mainland of Southeast Asia seems to have had much the same diversity as today as far back as 26000 BC.) This would explain why the genetic signal is stronger in Japan than in Korea when it is commonly assumed that the ancestors of the modern Japanese people migrated from what is now Korea at a time when the two languages were already very far apart.

Once the aboriginals were absorbed, later genetic comparisons would thus show a resemblance between the Japanese and Tamils on the one hand (those who had mixed with the aboriginals) and the Chinese on the other hand (who had not), with Koreans somewhere in between. This does not provide any evidence of a direct migration from India to Japan.

GamerGeek wrote:

Vijay wrote:It's completely ridiculous. I like to joke that everybody likes to connect Dravidian languages with all kinds of other language families because our languages are just so damn weird.

*our* languages?

Yep. I'm Dravidian (Malayalee). Our languages have an inclusive vs. exclusive distinction, but that's not to say that yours does!

Soap wrote:There seem to be two completely different "Indo-Pacific" families on Wikipedia.

See? Told you everyone wants to lump other languages with ours 'cause we're so weird.

Yep. I'm Dravidian (Malayalee). Our languages have an inclusive vs. exclusive distinction, but that's not to say that yours does!

I don't know, I reinvented that distinction when I was ten or eleven, so maybe it's not weird at all.

Pole, the wrote:

Yep. I'm Dravidian (Malayalee). Our languages have an inclusive vs. exclusive distinction, but that's not to say that yours does!

I don't know, I reinvented that distinction when I was ten or eleven, so maybe it's not weird at all.

How?! I can't imagine such a thing in Polish.

ˈd̪ʲɛ.gɔ kɾuˑl̪ wrote:How?! I can't imagine such a thing in Polish.

The power of imagination.

We even have a term for this around these parts for the phenomenon of coming up with something and it turning out to already exist in a language (except usually in a "worse", more complicated form): ANADEW ("a natlang's already dunnit except worse").

ˈd̪ʲɛ.gɔ kɾuˑl̪ wrote: How?! I can't imagine such a thing in Polish.

Why in Polish?

Pole, the wrote:

ˈd̪ʲɛ.gɔ kɾuˑl̪ wrote: How?! I can't imagine such a thing in Polish.

Why in Polish?

Aren't you Polish, like me? I think you were talking about Polish. I just tried to come up with any idea but I can't imagine how it could be easily expressed in our (I think) language.

ˈd̪ʲɛ.gɔ kɾuˑl̪ wrote: Aren't you Polish, like me? I think you were talking about Polish. I just tried to come up with any idea but I can't imagine how it could be easily expressed in our (I think) language.

The short answer: it wasn't Polish, it was a conlang.

(A while of understanding this let-me-call-this-plottwist) Wow! The first thing I created was agreement of possessive pronouns with the gender of the possessor, not the possessed thing, I was about this age.

And seriously, nevermind . I just need to remember not to visit this forum when I'm busy.

Vijay wrote:

GamerGeek wrote:

Vijay wrote:It's completely ridiculous. I like to joke that everybody likes to connect Dravidian languages with all kinds of other language families because our languages are just so damn weird.

*our* languages?

Yep. I'm Dravidian (Malayalee). Our languages have an inclusive vs. exclusive distinction, but that's not to say that yours does!

Touché

ˈd̪ʲɛ.gɔ kɾuˑl̪ wrote:The first thing I created was agreement of possessive pronouns with the gender of the possessor, not the possessed thing

So, basically his, her, its, yeah?

Imralu wrote:

ˈd̪ʲɛ.gɔ kɾuˑl̪ wrote:The first thing I created was agreement of possessive pronouns with the gender of the possessor, not the possessed thing

So, basically his, her, its, yeah?

I think it would extend to other persons as well.

Yes, eight-year-old I wondered why when I wanted to say "my mom" in Polish ("moja mama") the pronoun took the gender of "mom" ("mama") and not of "I" ("ja"). This was my first morphological thought.

Pole, the wrote:

Yep. I'm Dravidian (Malayalee). Our languages have an inclusive vs. exclusive distinction, but that's not to say that yours does!

I don't know, I reinvented that distinction when I was ten or eleven, so maybe it's not weird at all.

Well, no, that's not the weird part about Dravidian languages! They just have a bunch of features that aren't all that common in Eurasia (or sometimes elsewhere), I guess, like a whole bunch of retroflexes and anywhere from eleven phonemic fricatives (Toda) to none at all (some varieties of Tamil).

Vijay wrote:Well, no, that's not the weird part about Dravidian languages! They just have a bunch of features that aren't all that common in Eurasia (or sometimes elsewhere), I guess, like a whole bunch of retroflexes and anywhere from eleven phonemic fricatives (Toda) to none at all (some varieties of Tamil).

Retroflexes are just typically 'Indian', extending as far west as Pashto.

Eleven fricatives to none is just evidence of variability. I had been going to remark that having very few fricatives feels very Eurasiatic.

I can't help wondering if this "Indo-Pacific" means "subject to Eurasiatic influence".

Richard W wrote:Retroflexes are just typically 'Indian', extending as far west as Pashto.

Yeah, but they come from our languages. (In fact, I think a lot of "typically 'Indian'" things do).

Eleven fricatives to none is just evidence of variability.

Yeah, but where do you find that much variability, especially between what are basically neighboring languages?

I had been going to remark that having very few fricatives feels very Eurasiatic.

Why specifically Eurasiatic? Isn't it more common cross-linguistically to have relatively small numbers of fricatives?

Linking Dravidian to Korean is IMHO random and unwarranted. Sure, these two units are both characterized by agglutination and SOV word order, but that's far too unspecific.

But "Indo-Pacific", as defined by Greenberg (and Wikipedia understands it the same way), is a completely different group, consisting of the indigenous languages of the Andaman Islands, the "Papuan" languages and the extinct indigenous languages of Tasmania. And that is absurd. These three areas are so far from each other that there is no reason to assume that their indigenous languages form a single family. And indeed, "Andamanese" seems to be at least two families (Great Andamanese and Little Andamanese), probably even three, depending on what the completely unexplored Sentinelese language is like; "Papuan" just means "anything non-Austronesian indigenous to New Guinea and neighbouring islands" and consists of at least a dozen families; and the Tasmanian languages are completely unknown as they went extinct before any linguists found the occasion to do field work on them. This simply does not warrant any discussion.

WeepingElf wrote:(and Wikipedia understands it the same way)

Actually, Wikipedia has articles about both "Indo-Pacific" hypotheses: https://en.wikipedia.org/wiki/Indo-Paci ... age_family and https://en.wikipedia.org/wiki/Indo-Pacific_languages.

I really should've linked the Wikipedia article in the first post to avoid confusion! I was indeed referring to the former: https://en.wikipedia.org/wiki/Indo-Paci ... age_family

It's silly in any case; you need a lot more than a small list (and a contrived one at that - 네가 nega has a freaking case marker in it and actually resembles both Tamil pronouns less than the Mandarin Chinese equivalent lol; uḷḷa (உள்ள) does not even mean 'in' but rather something more like 'with' or 'having', and in any case, a postposition derived from the verb for 'to exist' sharing a common heritage with a form of the verb in another language for 'to climb' doesn't exactly make for a strong case; "'mom' and 'pop' words" are infamously similar cross-linguistically; and in India, AFAIK it's common for every family to have its own set of kinship terms, so it's not terribly surprising that you can cobble together a small list of some kinship terms that happen to vaguely resemble any given language) of chance similarities to get anywhere close to proving a genetic relationship. You can find chance similarities with all sorts of language combinations; this is why historical linguists are really supposed to follow the comparative method and find as many similarities as possible with regular sound correspondences before making any claims about genetic relationship.

Vijay wrote:It's silly in any case; you need a lot more than a small list (and a contrived one at that - 네가 nega has a freaking case marker in it

Two, in fact. The stem form is 너 /ne/. This contracts with the subject particle 이 /i/ to yield 네 /ney/, which then takes pleonastic 가 /ka/ (the form of the subject particle preferred after vowels).

It's a laughable list. Back in my sci.lang days, these were dime a dozen. Every kook had one. I think my favourite was Zuni-Latvian.

linguoboy wrote:I think my favourite was Zuni-Latvian.

I would love to hear the logical acrobatics it would take to justify that one.