Nursery School in Elfdalian.

BBC’s News from Elsewhere reports:

Elfdalian will be the sole language spoken to children attending the pre-school in the town of Alvdalen, which is the only community that still uses it, Radio Sweden reports. Elfdalian is believed to be the closest descendant of Old Norse, which was spoken by Scandinavians more than 1,000 years ago.

At the moment, only about 2,500 people can speak the language, fewer than 60 of them children, reports The Local website. To help preserve Elfdalian, councillors in Alvdalen on Tuesday voted unanimously to build the new nursery school.

The Local quotes the town’s mayor, Peter Egardt, as saying that officials were aware of their responsibility to “get a new generation to speak our unique language, thus giving the language more of a chance to survive in the long term”.

The new nursery school will be up and running in the autumn, and the site says that pupils will continue to learn the language until they turn 18.

I think I learned about Elfdalian from comments by Stefan Holm and Piotr Gąsiorowski; see those links for more links. Thanks, Trevor!


  1. A fight about the appropriate language tag for Elfdalian is in progress. The ISO Registration Authority (which is SIL) has turned it down for a separate language tag, mostly at the request of a Swedish government agency and despite plenty of testimony that it truly is an Abstandsprache despite having Swedish as a Dachsprache. The IETF language tagging list is trying to decide whether it is going to set a new precedent and override ISO, or simply register it as a variant of Swedish.

    Anyone with relevant professional or personal knowledge of Elfdalian is strongly encouraged to speak up now. You can sign up and read the archives at this page.

  2. “Elfdalian is believed to be the closest descendant of Old Norse”

    Is there any evidence for this? How does it stack up against Icelandic, Faroese or even Swedish?

  3. Trond Engen says

    I meant to comment on this yesterday. Closest descendant is of course meaningless. What it is is a remarkably conservative variety, preserving features that (all or most) other Modern Scandinavian varieties have lost. In morphology it has kept all four cases of the nominal paradigm and the personal inflections of the verbal paradigm. Phonologically it’s keeping sounds and phonemic distinctions lost elsewhere, most remarkably the nasal vowels that generally disappeared a millennium ago. But it does have its own innovations as well. Loss of initial j- is one (if I remember correctly).

  4. Since the item originates with Radio Sweden I assume there’s an implicit ‘within Sweden’. And that is true to a large extent, it does preserve things like Proto-Norse nasal vowels and a non-relict dative case — but it also has innovations that are found nowhere else.

    Comparing its distance to Old Norse with Icelandic or even conservative Nynorsk is probably futile, there is no well-ordered scale to use.

    As I understand it, Övdalską is the most conservative variety of the most conservative branch (Uvosiljan) of a group (Upper Dalecarlian) of very conservative dialects. All of Dalecarlian seems to be ambiguous between the east and west branches of North Germanic, with features of both — a case of incomplete lineage sorting — so arguably there is no ‘old’ continuum with the surrounding dialects. (Lower Dalecarlian shows a mixture of Dalecarlian and other-Swedish features, probably a result of mixture since historically it was less isolated).

    From a mutual intelligibility standpoint Dalecarlian dialects are generally not intelligible to speakers of Standard Swedish or other dialects, and largely not mutually intelligible with each other — but of course all speakers have Standard Swedish at least as L2 and are exposed to other regional variants on TV and radio, so the unintelligibility is not really mutual.

    So one problem with classification is that if Övdalską is a separate language there are at least three other forms of Dalecarlian that could make the same claim — but Övdalską is the one that has an active conservation movement and a self-image as a separate language.

    Another problem with classification is that Sweden understands a ‘minority language’ as the language of a minority, and happily provides official webpages in Yiddish because Jews are a minority in Sweden, not because Yiddish is in any danger of extinction. But the speakers of Övdalską are ethnic Swedes and not entitled to that protection.

    The EU seems to have another definition and has repeatedly asked for political action in the matter, but my guess is that the matter will remain referred to a researcher in a sub-department of a sub-institute of the Swedish Institute for Language and Folklore who is free to maintain her or his ‘detached’ scientific opinion because there is no political capital in 5000 speakers, and they aren’t otherwise marginalized or disadvantaged in a way that will upset the moral complacency of the average Swede and generate grassroots support.

  5. Trond Engen says

    Somewhat oddly, even further up the valley, in the former Norwegian parishes of Särna and Idre, the dialect is less conservative, traditionally being grouped with the Østerdalen dialects of Norway. Østerdalen is conspicuous for its general lack of deeply conservative linguistic features in spite of vaste distances of unsettled forest between significant settlements. Further west is Gudbrandsdalen, the northernmost valley in the belt of conservative “Midland” dialects along the mountain ridge dividing Eastern and Western Norway. My own hypothesis on this is that Østerdalen was (re-)settled fairly recently from the Norwegian southeast.

    I wasn’t aware that Elvdalian has some West Scandinavian features. Any features not shared by its immediate neighbours? To me it feels distinctly Mid-Scandinavian, i.e. Central Swedish/Eastern Norwegian, however conservative.

  6. @Trond, there’s a list in section 4 of this. Things like not merging *au and *ey.

    But the ‘immediate’ neighbours are Uvosiljan dialects too and probably share a lot of that. I don’t know where the isoglosses are, the stuff I am able to find mostly treats Övdalską as an isolate.

  7. Trond Engen says

    Thanks! Well. It took part in monophtongization, one of the defining characteristics of East Scandinavian, including the easternmost dialects of Norway, although not merging the two rounded dipthongs. It has vowel balance, and initial hw- became h- (> Ø-) or w- depending on the following vowel, both typical of Central Scandinavian (while WSc. has hv- > k(v)-). But this doesn’t really say much. The east-west divide was never clearcut and is mainly an artefact of documentation. The early scribal centers were in the Øresund and Vättern regions versus Iceland and Trondheim, but the isoglosses between them crossed the peninsula in all sorts of directions.

    (I misremembered loss of initial j-. It’s initial h- that was lost,)

  8. Trond Engen says

    Erratum: (while WNo. has hw- > kv-)

    Keeping h- before -w- is a defining feature of WSc. Only WNo. hardened it to k-

  9. The language taggers have asked SIL to reconsider their decision on an expedited schedule (normally they consider language tags once a year only, and some tags languish in limbo for many years). If that doesn’t happen, the five-letter tag may well get approved after all.

  10. Aaaaaand the appeal to SIL succeeded, probably thanks to the huge bibliography of works either in or about Elfdalian assembled by Mats Blakstad. The tag will be “ovd”.

    You heard it here first!

  11. Hooray! Thanks for the hot-off-the-presses update.

  12. Interesting to see how that will play out internally in Sweden — if it’s internationally recognized as a separate language, that should give some leverage.

  13. That’s exactly what the Swedish government has been afraid of all along: that the existence of a minority language implies the existence of a minority with minority rights according to the treaties Sweden has signed. It is no secret that when a modified version of the Ethnologue’s list of languages was being adopted as ISO 639-3, the Swedish delegate to the ISO working group insisted that Dalecarlian, Jamtska, and Scanian be removed from the list before it could be standardized. They are still visible as languages in the superseded 15th edition of the Ethnologue (2005) from the Languages of Sweden page, along with Swedish, Standard and Tornedalian Finnish, various Saami and Romani languages, and Swedish Sign Language, to which the current Ethnologue adds Danish (non-indigenous, but not solely immigrant either).

  14. Man, I hate the influence of politics on linguistics. I know it’s inevitable, because politics influences everything, but I still hate it.

  15. @John, do you have a link for the SIL decision / IETF code allocation that I can refer to on the Swedish WiPe page?

  16. Also, Danish is the second-largest minority language in Sweden, but it’s much harder to get mother-tongue instruction in Danish than in Kiswahili or Thai — but the reverse is probably true as well, except that Swedish may be the largest one in Denmark. “What’s your problem? It’s so easy for a Dane to learn Swedish!”

    That said, Danish families in Sweden generally have lots of native Danish speakers in their social circles and given the tiniest smidgen of motivation it’s easy for children to maintain native competence — though specialized vocabulary tends to suffer once they get to gymnasium. It’s not like being the only Thai family within 400km as can easily happen for immigrants who end up in the colder parts of the country.

    And of course no Danish person would ever suffer the delusion that it would be better to use Swedish as the home language. Mixed Dano-Swedish couples may end up that way by accident, though.

  17. Here’s the SIL decision (PDF). The IETF registration (it’s SIL that allocates codes) will follow in about two weeks and will appear at the Language Subtag Registry per usual.

    Oddly, the current Ethnologue pages on Denmark make absolutely no mention of Swedish or Norwegian. The only local languages listed are Danish, Danish SL, Greenlandic, and Standard German. The immigrant languages listed are Croatian, English, Faroese, Icelandic, Iranian Persian, Iu Mien, Northern Kurdish, Polish, Serbian, Somali, Standard Arabic, Tamil, Thai, Turkish, Urdu, Vietnamese, but such lists are inherently incomplete.

  18. Swedish Wikipedia updated.

    The status of Nordic nationals in other Nordic countries is regulated by treaties from the 1920’s and totally short-circuits the usual immigration procedures. You basically move in, show your passport and register your address to get an ID number, and that’s it. After 5 years you gain the right to become a citizen of the new country.

    That’s about 10 times less hassle than citizens of other EU countries experience, and 100 times less than a refugee or an immigrant from somewhere like India gets.

    But that also means that Swedes and Norwegians living in Denmark aren’t listed in the same places as ‘real’ immigrants, and perhaps the SIL researcher missed that.


  1. […] Hat notes the creation of a nursery school functioning in the Elfdalian dialect, in […]

Speak Your Mind