Ethnologue update comments

Ethnologue recently announced an update. Since, at least online, they are the easiest to find and hence most commonly cited authority on worldwide language distribution, this merits some comment. They've made some improvements in North Africa, including a much prettier map. However, there still remain a fair number of errors. This is perhaps natural given that SIL, which produces Ethnologue, is basically a Christian missionary organisation and as such is unwelcome in a number of countries. (Why is there no neutral source doing something comparable, one might ask? UNESCO has attempted a language endangerment atlas, but sadly that one is both less complete and far less reliable.) However, quite a few errors could have been avoided simply by closer attention to sources.

In Algeria, they've updated the Korandjé entry with my population estimate and endangerment classification, and corrected the Tashelhiyt one based on my thesis - although they apparently couldn't be bothered to cite the source of this estimate! They've reclassified the Berber dialects of the Southwest (Boussemghoun, Igli, etc.) as Taznatit, along with the Berber of Timimoun; in previous editions, the former were classed as part of Moroccan Tashelhiyt. The new classification is rather more tenable than the old one, at any rate; the former dialects are not called "Taznatit" by their speakers, but they are rather closely related to it, and are not at all closely related to Moroccan Tashelhiyt.

There are still a fair number of errors, though: the Tarifit of Arzew became extinct a century ago (and there is no such place as "Alteria"); "Tamazight de l’Atlas blidéen" is rather more like Kabyle than like Chenoua, and is missing from their map; the boundaries they give for Kabyle are rather inflated, annexing the Arabic-speaking areas of the Boumerdes coast to the west and half of Jijel in the east; the enormous circle they draw for Teggargarent contrasts oddly with the tiny ones they draw for Korandjé and Temacine Tamazight, which in reality occupy comparable areas (perhaps this was to take into account Ngouça, but the latter oasis is even smaller than Ouargla); Hassaniyya Arabic is still missing even though it's the primary language of Tindouf; and the curiously precise boundaries they give for "Saharan Arabic" leave me baffled as to what this "language" is supposed to be. The "Algerian Arabic" dialect of Jijel or Skikda is far more different from the Algiers koine than the "Algerian Saharan Arabic" I heard in Bechar or Timimoun or Abadla; perhaps the dialect further east is more different, but the boundaries shown are indefensible. They also seem to have misunderstood the constitutional amendment: it's Tamazight in general that is legally a national language now, although it's admittedly Kabyle that benefits the most in practice.

In Morocco, they've belatedly figured out that Ghomara and Senhaja are still spoken - for the past several editions they've been labelled extinct - but according to the entry French has no presence in the Moroccan linguistic landscape, unlike Algeria or Tunisia. For Egypt, their map ignores Gara (the other oasis where Siwi is spoken), and their population figures for Siwi are extremely optimistic. For Libya, their map (unlike their text) ignores Zuwara, while grossly inflating Ghadames and Sokna (especially in contrast to Siwi, which actually occupies a much larger oasis). For Mauritania, they still include the apocryphal Imraguen "language", whose reported existence Catherine Taine-Cheikh deconstructs in a forthcoming article, and their population figures of Zenaga are mutually inconsistent bad guesses which should have been updated based on the introduction to the latter's Dictionnaire zénaga-français.

"Political complexity predicts the spread of ethnolinguistic groups"

An interesting paper: Political complexity predicts the spread of ethnolinguistic groups. Two basically unsurprising claims that it's good to have calculations supporting: "pastoralists were found to have larger language areas than agriculturalists" and "languages associated with more politically complex societies cover significantly larger areas than those of less complex societies". They also present arguments that "although regions of high biological and cultural diversity do overlap to a striking degree, it is unlikely that biological diversity has any direct effect on cultural diversity on a global scale." Surprisingly, mountainousness was found to correlate with larger language areas, not smaller ones - seems a little suspicious that, though some mountainous areas are pretty un-diverse. Flaws: well, it relies on Ethnologue data and GMI maps, both of which are often unreliable, and systematically more splittist in some areas than in others; but it's not obvious that that would substantially affect the result. Also, ethnic groups, languages, and political units very often don't match up, and their measure of political complexity is based on data for ethnic groups rather than for languages.

(Via GNXP.)

Baccouche on Darja

Just found an interview with Taieb Baccouche, preparing a Linguistic Atlas of Tunisia - the interview isn't actually that interesting, but I definitely want to see that atlas.

Tamghazinut

I'm writing my core chapter at the moment on Kwarandzie (Korandje), the Northern Songhay language of Tabelbala. (Ethnologue and basic historical common sense notwithstanding, it is specifically Northern Songhay in ancestry, sharing common innovations with the language of places like In-Gall rather than the city it used to trade extensively with, Timbuktu.) It is very heavily influenced by Berber, like other Northern Songhay languages, and I found a great example the other day: the word for "old woman" is tamghazinut. Amghar is a Berber word meaning "old man"; zinu is a Songhay word meaning "old"; and ta-...-t is a Berber circumfix forming the feminine, which, even though Kwarandzie doesn't have gender agreement of any kind, seems (judging by this remarkable case) to be marginally productive as a derivational affix. (Postvocalic r is regularly lost in Korandje.)

On the map below (which I put together for my thesis using Google Earth and GIMP), you can see something of the geographic improbability of the situation:

Saudi Dialect Map

An item of passing interest that caught my eye, and seems quite hard to find online: a Saudi Dialect Map. Unfortunately, there is little information on what (if any?) data this is based on, but there are other interesting materials on the author's website. Of the dialects pictured, the Najdi dialect is rather remarkable for its conservatism; it retains both the classical passive and tanwin.

Speaking Arabic in public; or: don't say Yallah

You may have heard about the imams who got taken off a plane in the US (Minneapolis) because some passenger thought they were suspicious. Apparently:

Before passengers boarded, one became alarmed by an overheard discussion. "They seemed angry," he wrote in a police statement. "Mentioned 'U.S.' and 'killing Saddam.' Two men then swore slightly under their breath/mumbled. They spoke Arabic again. The gate called boarding for the flight. The men then chanted 'Allah, Allah, Allah.'"

It's bizarre the way that "They spoke Arabic again" seems to be characterised as somehow suspicious in itself - but the part that really makes me think "duhhhh!" is "The gate called boarding for the flight. The men then chanted 'Allah, Allah, Allah.'" It's obvious what they must really have been saying (although I haven't seen any paper point this out): Yallah, yallah, meaning "come on! let's go!". If even the use of the word "Allah" alarms some paranoid passengers, then Arabs will be hard-pressed to speak at all - between inshallah, hamdulillah, and bismillah alone (let alone yallah or wallah) you could easily reach at least one mention of "Allah" every couple of sentences in a completely mundane conversation! I hope this is an isolated instance rather than a trend.

In related news, speaking Yan-Nhangu is apparently suspicious as well...

In unrelated news, I thought Tulugaq's Google Map of Inupiaq was pretty cool, as is Sydney Place Names - I hope this is a trend.

Google Earth for linguists - and more Tunisian Berber

I've been playing around with Google Earth lately, and apart from all the obvious things you do when you get a satellite picture of the earth to play with - find your house, places you've been, etc. - it became clear that the ability to create and save placemark files opened up some interesting applications for linguists. To make a linguistic map, all you have to do is:
* create a new folder for the linguistic map (menu Add > Folder);
* list villages and towns that you know speak the language;
* look up their coordinates (where necessary) on sites like FallingRain - or better yet, record them with a GPS while you're there;
* go to them in Google Earth (you can type in rather than placename) and create placemarks for them (the pin button near the bottom right corner);
* change the icons for the placemarks if you have distinctions you want to make;
* add text to the placemarks (or folder names) in the Comments field;
* save the resulting folder as a KMZ file to be reopened in Google Earth.

Google Maps won't let you draw borders in, but (where relevant) this can be handled easily enough: File > Save Image, open it in Photoshop or GIMP, add a layer (so you can see the original at any time if you mess up), and draw the borders which, if you've plotted enough points, should be pretty obvious by then anyway. Filled in in suitable monochrome, this will look nicer in print, but has disadvantages: you lose the ability to attribute lengthy text to individual points (which shows up in Google Earth if you click on them), not to mention the ability to zoom in, or see the overall topography and environment.

By way of an example (possibly relevant to my PhD plans), here's one I did earlier: Tunisian Berber - Shilha. It has a bibliography of everything I could find on Tunisian Berber under the main folder, with works on individual villages cited under their placemarks, along with quotes on the vitality of Berber there. Berber is highly endangered in Tunisia, so I used four icons to represent different stages: a ghostly grey square for places where it disappeared shortly before 1900, a small bluish square for ones where it was still spoken in the 1930s, a white and blue circle for places where it is probably still spoken, and a larger white and blue square for places where it is still spoken by almost the whole population. It is divided into four subfolders, corresponding to different regions. As you will see, these varieties, in addition to being confined to less than thirteen villages in the whole country, are rather inadequately investigated - contrast the wealth of literature on and in Kabyle, or even Tashlhiyt. I hope this "cartographic bibliography" is found to be useful.