File talk:Langs N.Amer.png

White areas
What's with the big white chunk centered around, I guess, the Ohio Valley? "Uninhabited, unknown, out of area" ?! Surely not? TresÁrboles 04:15, 15 August 2005 (UTC)

Oops, just noticed the commentary below the image. Still hard to believe there's not enough data on this area. TresÁrboles 04:19, 15 August 2005 (UTC)


 * Yeah, amazing, but by the time records started being kept, I don't believe anyone lived there permanently. Iroquois, Sioux, and Algonkian were all found in the area, or parts of it, but I don't believe they considered it their ancestral land. Of course, settlers may have simply exterminated whoever did live there. The Atlantic coast is different; we know bits about people who lived there, but for the most part don't even have names. One language on the Georgian coast, Guale, was recorded, but all records were lost. kwami 20:45, 2005 August 29 (UTC)

Comments
Hi Ishwar,

Wow, what a spectacular map! I like how you used hatchings with the discontinuous families. That makes it very easy to follow which languages are Na-Dene, for example.

A couple comments:


 * 1) An "isolate" is a language well known enough to be classified, but which doesn't have any demonstrable relatives. I don't believe that Cayuse and Beothuk qualify; rather, I think you should have a separate color (maybe a lighter shade of grey?) for "unclassified languages". If you consider them to be isolates, why not add Calusa in Florida and Aranama in Texas as well? I think this is an important distinction: Zuni and Kutenai are obviously not in the same family as any other recorded languages. The same cannot be said about Cayuse, which might easily turn out to be Plateau if we had more data. (There are others, of course, such as Guale, which are simply unknown and are best left blank, as you have them.)
 * 2) Timucua as you have it should be a family. However, Tawasa is basically unclassified, and a lot of people feel Goddard is a little wacko for making such a strong claim for their relationship.
 * 3) Normal wording (or at least the wording I'm familiar with) is "Yuki-Wappo".
 * 4) Why not add Seri? You already have the boundary drawn, and it's in cultural North America (not Mesoamerica, I mean). It's also a likely Hokan language, and so is more of interest here than on a Mesoamerican map.
 * 5) Since recent studies have proposed that Takélma and Kalapuyan wouldn't form a group, even if both turned out to be Penutian, I think they should be broken up. Hardly any work here, not even another color!
 * 6) If you're going to group Plateau Penutian together, why not Oregon Coast Penutian too? I'm not advocating anything either way here, just raising a question of consistancy.

One final comment, purely esthetic: The color for Algonkian is hideous! Wouldn't matter if it were a tiny family, but since it covers a quarter of the continent, I think it deserves a pleasing color. I know it's hard to juggle all those colors, but the one for Wintu might be nice.

Great work! kwami 20:45, 2005 August 29 (UTC)

P.S. If you add Seri, then N. Mexico would be included in your map too. The white areas would then all be unrecorded or uninhabited, and none 'out of area'. kwami 20:51, 2005 August 29 (UTC)


 * hi. Thanks for the comments.


 * isolate-unclassified: Yeah, I wasnt going to make an evaluation of this. Mostly, because it would introduce yet another color/shade into the map which I was trying to keep simple. I dont know, maybe you can convince me to do this, but I would have to read more about the other isolates.
 * If not unclassified, then maybe just leave them blank. You're making a claim that these languages constitute their own families. Yes, perhaps a couple of the other isolates should be verified, but as far as I know these are the two worst attested "isolates" on your map.
 * Since you're only using one shade of grey, another shouldn't cause problems with your color scheme. And light grey is iconically motivated, since it implies that the languages aren't far from being unattested. Or maybe you could blank out the area, but still leave a number. No new colors, and it would get the point across.
 * Calusa & Aranama: I deemed Calusa and Aranama as being essentially undocumented (Calusa: 10 words, Aranama: 1 word + 1 phrase) and thus unknown. But, yeah, where do I draw the line? But, it certainly wont hurt to put these 2 on the map.
 * Not as isolates. Better just to leave them blank. You're right, Aranama is basically unattested. For Calusa that depends on which words were recorded. Greenberg thought he had enough to classify Calusa as Arawakan, so maybe there's enough to call it unclassified.
 * Timucua-Tawasa: I have mostly seen Tawasa mentioned as either a dialect of Timucua or a language related to it. Can you point me toward some source(s) that discuss (& disagree with) this? This is outside my knowledge at any rate. Additionally, I see Timucua described as having several dialects but not a family. But, if we consider Tawasa a separate language, then it would be a family. Goddard's map & classification & Jack Martin (in the Handbook, vol. 14) go with a Timucuan family including Tawasa. Mithun (1999) and Jackson et al. (Handbook, vol. 14) say language or family (including Tawasa).
 * Maybe you're right. I don't have a reference to fall back on. It was mainly water-cooler chitchat. Several of the other "families" (such as in Oregonian) are basically languages with divergent dialects, so you could justify it either way. As it is, it's the only isolate with a discontinuous distribution, which color coding helps to disambiguate.
 * Yuki-Wappo: I have seen both of these and also Yukian (in Goddard's map). Since you have a "feel" for the better name, I will change.
 * "Yuki-Wappo" I believe was used in Voegelin & Voegelin, as well as in Mithun and Campbell. I personally don't like "Yukian" because the validity of the family is still questioned, and "Yukian" implies that Wappo isn't terribly significant.
 * Seri: Yes, I should probably add that, too. (I also need to draw an island there.)
 * Takelman: I was following Goddard & ignoring that research. I will break up.
 * Plateau Penutian: I was following Goddard (and also Campbell's high probability rating). Although I havent read much about Penutian, it seemed that this grouping was more certain than Oregon Penutian (or at least certain enough to Goddard).
 * That could very well be the case. Also, Wintu might belong in Oregonian, which would throw a wrench in things.
 * color: I'll switch Wintu & Algic: tell me if it looks nicer.
 * Will do.


 * peace – ishwar  (speak)  20:05, 2005 August 30 (UTC)

update
after taking a long break, i have updated the map following user:kwami's suggestions, with the exception of (1) keeping Plateau Penutian (which i think has adequate support now) and (2) identifying Tawasa separately from Timucua (of course, as i mention above, i hardly know the literature). i still need to draw 2 islands off the western Mexico coast.

additional changes include adding hatching to Algic and switching the colors of Yuman-Cochimí & Wintuan. – ishwar  (speak)  17:42, 6 December 2005 (UTC)

Mexico
why isn't the entire country of Mexico shown? --Revolución hablar ver 22:20, 1 April 2006 (UTC)

Vector?
Has anybody put any thought into doing this map as svg?

Now that mediawiki will automatically generate pngs from svg data, I would think that svg would be ideal... Mobius 05:26, 4 April 2006 (UTC)