Talk:Language isolate

What about Hungarian? I seem to remember that Hungarian is thought to be related only (and this is controversial) to Japanese, and if Japanese is an isolate, then Hungarian would necessarily be also (at least until the two were proven to be related). Kadin2048 04:59, 13 Apr 2004 (UTC)

No, Hungarian is related to Finnish and Estonian. Adam Bishop 04:57, 13 Apr 2004 (UTC)

Uralic languages (Hungarian, Finnish etc.) have been popularily connected with Altaic languages (Japanese, Turkish etc.) forming a Ural-Altaic language group. This theory did propose a genetic connection between Japanese and Hungarian, but is now defunct. Please refer to the respective articles for more info. himasaram

Contents

1 Japanese 2 Hadza and Sandawe 3 first two paragraphs contratict 4 Native American languages

Japanese

Strictly speaking japanese is not isolate, but belongs to Japonic language family together with Ryukyuan and Okinawan languages.--Kulkuri 23:02, 3 Jul 2004 (UTC)

But the Japonic languages page includes in its first paragraph the admission that it may be merely a sprachbund, and Japanese is generally considered to be an isolate in the general linguistic community. UnDeadGoat

No, the sprachbund cited is Altaic-Japonic, not Japonic itself. Japanese is considered an isolate because it's often considered a single language with "dialects", much as Chinese is, for social rather than linguistic reasons. Japanese proper has enough diversity to be considered more than one language based on mutual intelligibility tests; the Luchuan (Ryukyuan) languages are even more diverse, and even the most conservative linguistic analysis would classify these two as separate languages. And no one questions the relationship between Japanese and Luchuan, it's just too transparent. I would hazard a guess that Japonic has approximately the diversity of one of the younger branches of Indo-European. --kwami 06:26, 4 Apr 2005 (UTC)

Good point Kwamikagami.... Japanese really can not be considered a language isolate, not with language such as Okinawan still alive and spoken by thousands on a daily basis. I suggest Japanese be removed from this article.

--User:harald 07:15, 7 Jun 2005 (UTC)

Hadza and Sandawe

These are listed on their own pages as Khoisan languages, but the Khoisan page says they are other languages. Anyone know for sure? -phma 14:55, 30 Jul 2004 (UTC)

The Khoisan family is rather tentative. Most linguists expect that the three (or perhaps four) Southern African families will prove to be related to each other, but Sandawe and especially Hadza are iffier. There really isn't any good reason for including Hadza in Khoisan; it was put there primarily because it has clicks and isn't obviously related to anything else. Part of the problem is that these langages are very poorly described. --kwami 06:42, 4 Apr 2005 (UTC)

first two paragraphs contratict

It seems like the definition in the 1st paragraph contradicts with the statement in the 2nd that some languages became isolates in recent times. I hope someone can sort this out. What is the critical criterium - 1) not related to any living language, or 2) no known genetic relationship to any other language (living or dead)? ike9898 15:33, Jan 30, 2005 (UTC)

"Language isolate" has no precise definition. Context will have to disambiguate. People classify Basque as an isolate despite the fact that it's most likely related to (extinct) Aquitanian; similarly Etruscan and Rhaetian, though both of those are extinct. However, if a language was classified as part of a family and only later its relatives died out, as is the case with Piraha in the Muran family, then it might not be considered an isolate. Consider also Japanese, which is frequently called an isolate despite being a small language family. The term "isolate" is as fuzzy as the term "language" itself. --kwami 06:36, 4 Apr 2005 (UTC)

Another example of a case like Japanese, is the Chimakuan family which is often called an isolate, even though it consists of two languages, Chemakum and Quileute.

The term is also used refer to single members of a particular branch of a larger grouping (i.e. a stock or phylum) which has families at its other branches. So, for example, Bella Coola is an isolate within the Salish(an) family (which has also the Tsamsoan subfamily, Interior Salish family, etc.). For better or for worse, all of these usages exist. — ishwar  (SPEAK) 00:25, 2005 Jun 4 (UTC)

Native American languages

Am I the only one reacting to all native American languages listed as language isolates? Wouldn't they be classified as isolated languages (as being spoken only in reservations) rather than language isolates with no genetic connection to any other languages? I'm not sure because I'm no expert on these languages. It'd be nice if somebody could comment on this. ;-) himasaram

Hi.

No, all of the languages on the list are, as of the most current research, isolates with no known relations to any other family or language. The North American ones are according to Goddard (1996) & Mithun (1999). Central & South are according to Campbell (1997).

Many of these languages became extinct a good while ago before accurate data was collected (due to the unfortunate political practices of the past). Often times having only a small list words is not enough to determine any kind of genetic relationship. In other cases, such as with many of the South American isolates, enough data has simply not been collected yet (at least to the knowledge of Campbell). You can probably expect the number of isolates in South America to decrease over time.

As the article says, language isolates may be related but the relationship is undemonstrated. This is true of any language or language family. So, English may be related to Arabic, but whatever connection there may be between these two (and this connection would have to be very, very remote) has not been convincingly "proven" to the linguistic community.

Regarding North America, there are three big language family proposals that you should be aware of:

• Penutian
• Hokan
• Amerind

Penutian and Hokan are due to Edward Sapir and his predecessors. Amerind is a proposal by Joseph H. Greenberg that considers all languages in America to a part of one stock excepting the Eskimo-Aleut and Na-Dene families. Greenberg's proposal would, then, consider there to only three families with probably no isolates at all. Penutian is currently undemonstrated although most linguists find this hypothesis to be very promising. Hokan is less certain—some languages may be related, but others are probably not. If Penutian and Hokan are valid, then the number of isolates may reduced slightly. Amerind is not accepted by most Native American specialists.

I would guess that if you wait 25 years or so the issue of Penutian & Hokan will be better known, as will the situation of South American languages. However, I should also mention that the Americas are extremely linguistically diverse areas (for instance, only in the Americas does there occur all 6 word order types). The diversity of the Americas is rivaled only by places like Southeast Asia and Indonesia. Peace. — ishwar  (SPEAK) 14:52, 2005 Apr 11 (UTC)

Superb explanation, Ish_ishwar. Keep up the good work! -Himasaram 21:38, 3 Jun 2005 (UTC)

Wouldn't unclassified language be a more appropriate place for the ones for which inadequate data (or none) is the only reason? - Mustafaa 23:02, 3 Jun 2005 (UTC)

Good point. But I personally don't feel confident enough to make the distinction between an unclassified language and a language isolate... -Himasaram 23:26, 3 Jun 2005 (UTC)

Hi. Yes, I think you can make a distinction between these two categories. I am using a kind of loose definition (an as-yet genetically unaffiliated single-member language family ?). So, making the distinction would involve an evaluation of the documentation status of these languages, and I guess that we could have a kind of confidence rating of the status. For example, Zuni could an isolate with a 80% (?) certainity. Baenan (a.k.a. Baeña) could be an isolate with a 0% certainty which would be, in other words, an unclassified language (Kaufman 1994:70 says that "this language is too poorly known for even Gr[eenberg] to dare classifying it"; I cant find it in Ethnologue). As far as evaluating all of the languages, I find it easier to just say that they are not known to be related to other languages. Anyway, I was just trying to put (quickly) all of these language names somewhere on the internet since many are probably unrepresented here in cyberspace. That's all. I guess, you are requesting me to work harder :) peace. — ishwar  (SPEAK) 23:57, 2005 Jun 3 (UTC)

Your intentions are admirable, Ish_ishwar. My only concern was that the list on this page was totally dominated by Native American languages. Perhaps the easiest way for you, is to create a separate list on this page or on a separate page, for these isolated/unclassified American languages? -Himasaram 18:27, 4 Jun 2005 (UTC)