There are 100 or so languages, living or known extinct ones, that allegedly have no known relatives (i.e., nothing proven to be related to them by the techniques of comparative linguistics). Basque is an obvious example; Japanese, Ainu, the Paleo-Siberian languages, several American Indian languages, a couple of Caucasus languages, and some aboriginal tongues in PNG and Australia, are others that come to mind. However, there is a lot of theorizing going on among philologists.
Basically, this is a GQ with a breadth of accurate answers and nothing that is a definite “right answer” as to what has reasonably been posited for connections between those language isolates and language families.
Burushaski (one of my favorite languages) is hypothesized to be related to Ket, which is a very small palaeo-Siberian language.
Which Caucasian language isolates did you have in mind? I was under the impression that they could all be classified into at least very small families.
Oops, I didn’t mean to post that. What I meant to post:
This link lists a few attempts that have been made to relate Basque to other language families, but it seems that nothing but the possible link to Caucasian languages is even slightly credible.
Note that Japanese might well better be considered a member of a very small family of languages, including the Ryukyuan languages (e.g. Okinawan), but it’s a very small, very closely related family. Japanese and Korean have fairly similar grammar and are often linked into the hypothetical Altaic superfamily which includes the Turkic, Mongolic, and Tungusic languages. The existence of Altaic is tenuous, of course, and the membership of Korean and the Japonic languages even moreso.
If the latter, I think it’s fairly undisputed that human language probably developed only once, in Africa, maybe 50,000 years ago, and so all languages are related. An isolate is a language that some linguist has posited cannot be shown to be more closely related to one known language (or group) than another. “Lumpers” such as Merritt Ruhlen would include, say, Japanese within the Altaic languages, while “splitters” would not, and might consider Japanese an isolate. If we consider that the farthest back two languages or groups could share a common ancestral language and still have demonstrable connections to be about 15,000 years (this is farther back than most linguists would consider possible, but some such as Ruhlen would say we can), then an isolate would be a language with no living (or documented) relative whose common ancestral tongue was spoken more recently than 15,000 years.
Even Ruhlen considered the following languages isolates in A Guide to the World’s Languages, though his inclination is to try to find relatives for them anyway:
Basque, Burushaski, Ket, Gilyak, Nahali, Sumerian, Etruscan, Hurrian, Meroitic (the last four are no longer spoken)
If the former, you will notice that no American Indian language is on Ruhlen’s list of isolates. That is because each is considered by the “lumpers” to belong to one of three groups: Amerind (representing the earliest migration from Asia, and encompassing the vast majority of the languages), Na-Dene, the Athabaskan and Navajo language groups, representing the second migration – and posited by a few to be related to certain Caucasian languages, as well as Sino_tibetan ones like Chinese! --, and Inuit-Aleut, representing the most recent migration.
Of Hazel-Rah’s alternatives, I was particularly interested in the examples, but would certainly welcome knowing how language isolates came to exist (or what is theorized about that question.)
Regarding Amerind languages, it was my understanding that there was no consensus on the Amerind/Athabascan-NaDene/Aleut-Inuit trichotomy theory, and that the “classical” phyla like Macro-Siouan, Ge-Pano-Carib, Uto-Aztecan were still accepted by the more conservative students of Native American languages. And that there were a number of isolates among them, particularly in the Caucasus-like melange of pre-Columbian California.
IANALinguist, but I’ve always been under the impression that Amerind was not a terribly widely-believed grouping. The real lumpers may believe it, but it’s not something with real solid evidence to my knowledge. Does anyone know a little more about it?
Apparently there is some corroboration between the proposed Amerind lingustic grouping, and genetic studies such as those of Cavalli-Sforza. But I think it’s as you say, accepted by a minority of researchers – though from what I can tell a sizeable minority, not just a few mavericks.
The OP mentions California’s linguistic diversity. It seems that several languages formerly considered isolates there are now being recognized by at least some linguists as having affinities with other, sometimes even distant languages. For eaxmple, Miwok and a couple pf other Californian languages were found, decades ago, to form a grouping, one of the “traditional” groups as mentioned an earlier post – namely, Penutian. A few years ago, both lingustic and way-of-life parallels were found between Penutian and the languages of (mainly formerly of) the Gulf Coast (Louisiana, etc.) Somewhat more controversial is the most recent hypothesis, that this expanded Penutian language group should include the Maya languages of Yucatan, Guatemala, and beyond!
Perhaps this is a subject for another thread, but if this theory is shown to be likely true, how did it happen? Assuming some (at least coastal) marine navigation, the Gulf - Maya expansion doesn’t seem too far-fetched to me, but the link to Miwok, etc. seems stranger. However, it does seem similar to the proposed history of other dispersed language groups: Hokan (mainly northern Mexico, but one language in Honduras) and even Na-Dene seem to have originated in small, but linguistically diverse, sections of the Pacific coast. It seems perhaps that the original inhabitants of the New World first made their way down the California coast, settling into small pockets, then some folks from each pocket quickly migrating to distant , disparate locations from which they spread. A bit like the Austronesian languages (Madagascar to Easter Island) “originating” in Taiwan, where their lingustic diversity is still greatest.
Highly anectodal, but I was watching a documentary about mongolians, and while I couldn’t understand a word they said it sounded a heck of a lot like Japanese.
(And I wouldn’t be surprised if there’s a linguistic link. It’s not like people there were descended from Amaterasu or anything. :P)
Of course, the Altaic hypothesis is highly disputed – it’s not even clear that there is such a thing as an Altaic family, to say nothing of Japanese being part of it. Nevertheless, most of the languages that would fall into such a family if it did exist do have fairly similar sounds – long words, few consonant clusters, and little or no stress.
I’m surprised no one has yet mentioned Joseph Greenberg who has addressed the OP’s question directly by proposing a theory which ties all langauges together. That theory, needless to say, is highly controversial.
But I think **JKellyMap **pretty much hit the nail on the head:
In defense of Greenberg, whatever one thinks of his loopy first-language ideas, he did bring sense, order, and now-widely-accepted insight to African comparative historical lingustics.
Thanks! Though if my PhD advisor (not in lingusitics) finds out how much time I’ve been spending lately reading some of this stuff, I am in deep doo doo. So – mum’s the word, 'kay?