Talk:Language family

From Wikipedia, the free encyclopedia

Contents

[edit] Isolate language vs unclassified

What is the difference? Is it not very possible that unclassified ones are simply isolate?Disko 03:53, 16 May 2006 (UTC)

If a language is reasonably well documented and has been compared to other languages without success it is considered an isolate. Thus, Basque is considered an isolate since we have lots of data for it and quite a few people have tried to relate it to other languages. If, on the other hand, a language is very poorly known or has just not received much attention, it is considered unclassified. The intended distinction is that an isolate is a language that has had a chance to be related to other languages and still has not revealed such a relationship. Assuming that linguists have not been insufficiently perceptive, such a language is either not related to other languages or is related too distantly for that relationship to be recoverable. A language that is unclassified, on the other hand, is one that has not yet had the opportunity to be classified and whose current isolated status is therefore not taken to reflect anything significant.Bill 04:23, 16 May 2006 (UTC)
It's also possible for a language to be unclassified within a certain family. Pictish, for example, could be called "unclassified Celtic": most people are pretty sure it's a Celtic language, but not enough about it is known to say what its relationship to the other Celtic languages is. Angr (t • c) 06:59, 16 May 2006 (UTC)

[edit] Language counts from Ethnologue

Where are the updated counts from? Strange that the number of languages in many phyla went down - e.g. Niger-Kordofanian lost 500 languages. I computed it from the ethnologue data file from 1988. --Erauch 19:25, 19 Jan 2005 (UTC)

Yeah I noticed it too. These figures really need a reference; maybe we should pull them out till a reference is provided (otherwise we might forget it and end up with figures no one knows the exact source of). So my proposal is to delay the changes pending references (i.e. revert to the older version of which we at least know the source). — mark ✎ 01:10, 20 Jan 2005 (UTC)
I will wait a bit for user Wdshu to respond. --Erauch 04:40, 20 Jan 2005 (UTC)


[edit] Map for article

I added a language map that I adapted from the German Wikipedia.

Human Language Families Map
Enlarge
Human Language Families Map

Feel free to improve the map - it's by no means perfect. I'd especially like somebody to add Native American languages in Peru and Mexico.

I created the german version of the map. It is still not ready. Perú and central America will be the next countries on my list. Be patient and I will upload an english version, too! Stern 09:02, 14 Feb 2005 (UTC)

North American only has Indo-European languages. Rather ethnocentric. It is probably better to restrict Germanic languages to Europe and indicate only the American languages in North America.

- Ish ishwar 23:26, 2005 Feb 18 (UTC)

The map shows the distribution of language families today. Nonetheless, it does indicate the areas in North and South America where Native American languages are still spoken. --Industrius 23:53, 9 Mar 2005 (UTC)
I dont see any mention of the Athabaskan family or the Salishan family, etc. The map doesnt mention or show this in any way. It just show IE langs. So, I dont really understand your comment above.
Here are there pieces of an outdated & slightly inaccurate map of language families in the US. Not the best, but better than nothing.
Cheers! - Ish ishwar 06:27, 2005 Mar 10 (UTC)
The Native American languages are shown in the "other" category, like other small language families. It would be hard to show the distinctions between Algonquian, Athabascan, Uto-Aztecan, etc in the post-colonial world.
Our problem is that there are two different "language distribution schemes" that the map can display. Either it can show how the languages families are distributed today, or it can show how they were originally distributed. Both types of map are legitimate and useful. Maybe we should make two versions of this map, one called "Modern Language Families" and the other called "Historical Language Families". What do you think? --Industrius 05:20, 11 Mar 2005 (UTC)
The problem with the "original" distribution is that we don't know what it was. Our best guesses reflect where the languages were at first contact with Europeans, and this is centuries later in the west of what's now the USA and Canada than the east. In the meantime American societies had been profoundly affected by indirect contact, and there had been massive changes and movements of populations (population collapse of Mississippian civilization due to imported disease before de Soto arrived; movement out onto the Plains with imported horses; etc.) And in large areas of the eastern US (and eastern Brazil, for that matter), the map is a blank: we have no idea who the people were linguistically.
What I'd like to see is both. For the world at large, the current situation. For the Americas, both the current situation in more detail, and our best reconstruction of the situation at first contact.
There are more serious problems with this map, but I'm removing my comments and putting them on Stern's talk page. --kwami 20:13, 20 Mar 2005 (UTC)


Hi.

Yes, the points that user kwami states is very true. So, as usual, when you make a Native American map you have to state this disclaimer. The most recent map is the one made by Goddard (1999) (the updated & revised version of the one he made in Goddard 1996 for the Handbook of North American Indians'). You should check this out (it is very nice map). I think that National Geographic had an issue where they copied the Goddard map & added spiffy stuff to it (I havent compared NG to Goddard though). Re South & Central Am.: I just dont know about a map for these there must be something mentioned in Campbell (1997)....

It will be hard to make a map of the current distribution of speakers. Industrius, you seem to be implying that you & Stern plan to ignore them for the most part? I dont want to suggest that (since perhaps this will be interpreted as further insult & injury toward the many speakers who are not represented). But it will be a very big challenge to find out current distribution. I think that it would entail a lot of research—ideally a lot of phone calls/emails to lang family specialists. Also since many of these languages have been profounded affected and are now spoken by only a few person, there is a practical challenge of even displaying them on the map. I just dont want a casual reader to think briefly look over the map & come to the conclusion that none of the indigenous languages are still spoken. I want them to get full exposure since they are beautiful and make the Americas linguistically & culturally unique. Peace. — ishwar (SPEAK) 00:40, 2005 Mar 22 (UTC)


The label "Turkish" is incorrect. Part of the teal area may be "Turkic" or "Altaic," but the Paleosiberian languages are not seriously considered to be candidates for this language family at this point in time. Perhaps if Ainu may have a place in this grouping, though all of them are almost always regarded as language isolates, and the naming is for convenience. JH 2005, April 08

Going just from memory, it looks like the Paleosiberian families have been submerged under Slavic. Or is the azure on the Arctic coast more than just Yakut? "Turkic" would of course be preferable. But comments should probably go on Stern's home page, [[1]]. --kwami 11:06, 8 Apr 2005 (UTC)

Added another kind of map where there are some languages and language groups named and put on the map where they are currently spoken. Tried to divide larger language families according to some articles, but there seems to be debate over these. So the map has "East Siberian Languages" which is to my understanding no kind of a language family. Other sorts of inaccurasies have been bound to have happened, sorry, feel free to improve the map (f.e. the colours?). Dreg743 13:26, 2 June 2006 (UTC)

Check oout image:languengl.gif —The preceding unsigned comment was added by 84.230.251.132 (talk • contribs) 2006-07-26 18:28:28 (UTC)

[edit] Number of Speakers

Maybe there should be a rough number of 1st language speakers for different families?

[edit] Ainu

Kwamigakami just added "one surviving language out of perhaps a dozen or more". Now I am aware of the significant differences between, in particular, Karafuto and Hokkaido Ainu, and could envision perhaps splitting it into two languages, although that is not the typical solution; but a dozen? Which dozen are these? - Mustafaa 03:10, 25 Mar 2005 (UTC)

The records are of poor quality, but the people I've spoken to who've worked with Ainu believe it to be a small family, like, say, Japanese, or at least Yukaghir. Similar enough to be considered a single language when glossing over "obscure" languages, but diverse enough to warrant separate identities by the standards of Ethnologue. I take it that's reading a fair amount into limited data. Sorry, I don't have any refs with me to back this up, so go ahead and take it out if you like. kwami 04:02, 25 Mar 2005 (UTC)
Ah - I wasn't sure if that was what you meant. In that case, your point is well taken, but we shouldn't favor splitters over lumpers or vice versa. Is "like Arabic or Japonic, the diversity within Ainu is large enough that some consider it to be perhaps up to a dozen languages while others consider it a single language with high dialectal diversity" a fair compromise? - Mustafaa 04:12, 25 Mar 2005 (UTC)
Sounds good, but since I don't actually know the data, I'm starting to feel unsure of saying anything. I don't know how the time depth would compare to Arabic. Places to look might be Tamura’s The Ainu Language, or Hiroshi Nakagawa at Chiba University. If the Ainu only spread north to Hokkaido with the introduction of Yayoi agriculture c. 500-700 AD, and to southern Sahalin and the Kuriles after that (quite possible, given the archaeology), it's reasonable to suppose they might be similar. I think we should have something to indicate that there is (or was) some diversity, though. kwami 10:19, 25 Mar 2005 (UTC)

[edit] Article title

The article is about language families and not languages per se, yet the title implies that it should be encroaching on the scope of the article entitled language. Why? The title is long, complicated, not terribly enlightening to non-editors and newbie editors alike and fairly non-encyclopedic. Could someone give a reason that we shouldn't move this back to the far more appropriate language family?

Peter Isotalo 12:54, 6 August 2005 (UTC)

Language isolates. — mark ✎ 17:14, 6 August 2005 (UTC)
Well, then logic dictates that the title should be "language families and (language) isolates". And language isolates is inconsistently enough not titled "language isolates and families". The point is that the title is quite at odds with any proper naming, not that it would somehow ignore the isolates. It's a lot more likely that someone will be blasphemous enough to start a separate language family article than to demand that all mention of language isolates should be removed.
Peter Isotalo 17:47, 6 August 2005 (UTC)
I'm not dead set on keeping the current title, which I admit is a little clumsy; I just wanted to point out that they're not subsumed under the term 'language family'. — mark ✎ 17:53, 6 August 2005 (UTC)
Why not just point out that language isolates are in effect families of their own? Japonic was until a few years ago called an isolate (and often still is); and many of the "families" (such as Papuan) contain lots of isolates. Yukaghir is an isolate or a family depending on your definition of "dialect". Ket is both a member of the Yeniseian family and an isolate since the extinction of Yugh. How long ago must a language's relatives have to be extinct before it's considered an isolate? Is Basque an isolate, with its extinct relative Aquitanian? What about Etruscan, when both it and its two known (or suspected) relatives are extinct? All other isolates presumably had recognizable relatives at one time; we're merely ignorant of what they were. Würm in his Papuan classification speaks of family-level isolates, phylum-level isolates, and stock-level isolates. There's no clear dividing line between a 'family' and an 'isolate'.
Besides, even if we agree on how to separate them, an article on X should also discuss what's not X. In an article on language families, it's entirely appropriate to say, 'by the way, the following languages don't seem to belong to any family'.
If countries contain cities, should the Vatican, Singapore, and Monaco be removed from the list of countries, because they're single cities? If not, why should language isolates be removed from the list of language families?
I agree with Peter that this article should have the simple and straightforward name 'Language families'. kwami 08:17, 2005 August 7 (UTC)
Convincing. Thanks, kwamigami. I support the move to Language family (singular in accordance with the naming conventions). — mark ✎ 10:33, 7 August 2005 (UTC)
That was indeed more convincing than I was, kwami. Let's wait a week or two to see if anyone has any relevant objections. If none arise we'll move it.
Peter Isotalo 11:31, 9 August 2005 (UTC)

[edit] Small Map Legend

The legend of the map on this article is too small to read. Leon Trotsky 9:31 30 October 2005

You can click the image to enlarge it. --24.21.101.75 10:04, 4 November 2005 (UTC)


[edit] sign languages and creoles

This page currently contains the entire list of sign languages from the list of sign languages page. However this article is a list of language families, not a list of languages, so I would like to truncate the sign language section. While many sign languages are probably isolates, some are certainly related to others — the problem is that there seems to be very little linguistic research into these relationships. So do we leave the explanatory note and delete the list?

Also, creoles languages and the like are listed individually. Could they not be grouped according to the superstrate and/or substrate languages from which they are formed? American Sign Language is sometimes described as a creole arising from French Sign Language, Martha's Vineyard Sign Language and a number of home sign systems. Should it be listed with the creole (etc) languages? --ntennis 23:47, 17 November 2005 (UTC)

Since the SL list merely duplicates another article, I deleted it. The same should be done with creoles, but I didn't want to lose any info. Different creole families like Bislamic could remain, but there is a lot of uncertainty as to which creoles are related. Same for SLs: if we list the Britannic family, we imply that other languages are isolates. Is ASL part of the FSL family? etc. Maybe this could be covered in the SL and creole articles rather than here. kwami 00:41, 18 November 2005 (UTC)

[edit] time depth

Would it be a good idea to give some kind of (tentative) time depth for those language families (i.e. for their protolanguage) listed in this article?128.214.205.4 15:01, 7 March 2006 (UTC)

[edit] Term "Genetic" extremely confusing

As a non-linguist, the term "genetic" (and related terms) used in these articles is extremely confusing. Could we please state very clearly that "genetic" in language has nothing to do with "genetic" in DNA, genes, etc.?
(IMHO this is especially important because I've occasionally run across more-or-less racist comments that "they talk that way because of their race". I think that we need to make plain that that's not what linguists are saying.) -- 201.50.249.78 15:51, 18 October 2006 (UTC)

Retrieved from "http://en.wikipedia.org../../../l/a/n/Talk%7ELanguage_family_ed25.html"