Language and Identity
Reality and Belief



June 20, 2005

A Language is a Dialect with an Army


Raymond G. Gordon, Jr. (ed):

Ethnologue - Languages of the World

2005, Summer Institute of Linguistics
1.272 pages, 208 maps in colours, Price 80 $

Reviewed by Ole Stig Andersen.
Abbreviated version broadcast June 20th 2005 by Danmarks Radio (Danish National Public Service Radio), Programme 1 Morning: "Fagbogen" ("Non-fiction")
Listen to the review here: (6'07 mins in Danish)

The threat against biological diversity is a frequent issue in the public discussion, but the same thing can certainly not be said of the ongoing enormous destruction of cultures - a language dies every fourteen days or so, and in a hundred years less than half the world's languages will have survived. At the same time as the linguistic diversity is being drastically reduced, science gets an increasingly better idea of the situation. Ole Stig Andersen has taken a look at a newly published survey of all the languages of the world.
According to this hefty book there are 6.912 languages in the world, and Ethnologue, as it is called, lists them each and every one, with information on which countries a language is spoken in, number of speakers, literacy, dialect names, deaf population and more.

Since the first mimeographed edition of only ten pages appeared in 1951, Ethnologue has grown to become the world's most complete and authoritative survey of the world's languages. It is the 15th edition that has now been published: 1272 pages. 80$. And the whole thing is on the Web too:

Ethnologue is fully cross-referenced. You can check a country, a language or a language family, and there are special lists, like the one of the world's most endangered languages, 497 languages spoken by less than 50 people. There are several statistics and more than 200 maps in colours showing where the different languages are spoken. The front-runner in the discipline of linguistic diversity is Papua Niugini where a population the size of Denmark's (5 mio) apparently speaks 830 languages. Other countries with hundreds of (indigenous) languages are India, Indonesia, the Philippines and Nigeria. But they have hundreds of millions of inhabitants.

It is not a book you would imagine translated into Danish. Only a few of these tens of thousands of unknown languages, dialects and place names have a Danish name, and there is no reason they should, too.

But let us check out Denmark. According to Ethnologue eight languages are spoken in the kingdom. Five of them: Danish, German, Faroese, Greenlandic and Sign are no surprise. But how do you like Traveller Danish aka Rotwelsch? And then there are Jutish and Scanian, which are counted as independent languages on a par with Danish. The more detailed descriptions of these two languages reveal that by "Jutish" is meant South Jutish, and by "Scanian" is meant Eastern Bornholm. (The facts)

There are two more tiny languages in Southern Jutland which have strangely eluded Ethnologue: North Frisian and Platt (Low Saxon). Both are very near extinction in Denmark, or already gone. But so is Traveller Danish which IS listed.

Ethnologue's claims about the immigrant languages are also not too correct, they border on the invented. And the cross-referencing makes it even more absurd. E. g. Ethnologue claims that Iu Mien (a Yao language of Southern China) is spoken in Denmark, and this piece of misinformation - it has absolutely no bearing on reality - appears both in the article on languages spoken in Denmark and the article on countries where Iu Mien is spoken. Because of the authority of Ethnologue this Iu-Mien-in-Denmark-myth is repeated all over the net, my last googling gave 1.280 hits. (Cf. a rather better-informed list of Denmark's languages).

Such mysterious, misunderstood and outright faulty information raises serious questions about the reliability of Ethnologue. One fears that many other articles on languages and countries one knows less about may be of the same problematic quality.

An information supplied by Ethnologue on all languages, is whether a Bible translation exists. Danish, German, Faroese and Greenlandic are well supplied, but information is lacking on "Jutish" and "Scanian".

And this is the crux of the matter. In its core Ethnologue is not a survey of the world's languges, but a survey of the world's need for Bible translations. It is published by Summer Institute of Linguistics which from its headquarters in Dallas, Texas, strives to spread its variety of Christianity all over the world, especially the Third. The researchers from Summer Institute of Linguistics work as missionaries among indigenous populations for years, and they produce language descriptions of the highest academic quality with the end purpose of getting the Bible translated.

Ethnologue was created as an instrument of American missionary thrusts that were at first directed against Latin American Catholicism which works in Spanish. So the evangelical missionaries directed their attention towards the Indian languages. Since then they have expanded their acivities considerably and are now proselytising in e. g. West Africa, India, South-East Asia, New Guinea and exSoviet.

The people who study languages scientifically at Summer Institute of Linguistics also form a parallel organization, Wycliffe Bible Translators, where they work to translate the Bible into as many languages as possible. In a number of cases larger or smaller portions of this old Middle Eastern myth collection will be the only texts that exist in a number of languages that are dying. The cultures these languages have passed on are considered by many missionaries to be the Devil's delusions (the cultures, that is, not the languages). They strive to enrich or supplant the local cultures with their own version of the American Bible Belt's Christian fundamentalism.

Ideally, written texts in these endangered and moribund languages should be about the cultures that the languages transmitted, their contribution to Mankind's cultural heritage. But the history of writing shows that missionary activities of different religions generally have been the most important factor whenever a language has acquired written form, so in that respect there is nothing special about SIL. Many of SIL's languages will not survive, though, and their written heritage will not be indigenous.

It is also in the scientists' religious motivation that one should look for the reason why Ethnologue splits up languages. For since their mission is to spread the Word in all languages, they wouldn't risk missing a single one. Therefore they are - as far as lingustic classification goes - extreme splitters. As a matter of course they elevate dialects to independent language status. Where others hear a dialect, they envisage an independent Bible translation.

And when a language once has ended up in Ethnologue's survey, they don't rashly take it off again, as the persistent inclusion of Rotwelsch as a language of Denmark shows. So there is good reason to expect that Jutish and Scanian will stay and that Ethnologue eventually will discover Vendelbomål, Amber Jutish and East Greenlandic and include them in the list.

For there is the curious thing about Ethnologue that at the same time as the number of languages in the world is steadily falling, the number of languages in Ethnologue is just as steadily rising. It has grown with 103 languages since the previous edition four years ago, almost all of them by promotion of dialects. In the same period humanity has probably lost over a hundred languages.

Ethnologue's strange information about the languages of Denmark is nothing exceptional. It is the norm, and for many other languages it is much worse. E. g. Chinese is partitioned into 14 languages and Arabic into no less than 35.

Of course, the splitting has a political side, too. I find it more ridiculous than offensive that Ethnologue elevates South Jutish and East Bornholm to language status, and under misleading labels at that. But an Arab acquaintance of mine does certainly not laugh over the fact that his and the Quran's precious language in the world's most widely used and authoritative survey has been split into 35. On a symbolic level he considers it part of some Western divide-and-rule tactics against Islam! (Comment on mutual intelligibility)

Language surveys of the Ethnologue type necessarily share several drawbacks with language atlasses: It is difficult to depict non-geographical information such as functional domains, different languages for different purposes, effects of immigration and urbanisation, bi- and multilingualism, second languages, high- and low status varieties, and writing.

Missionaries don't arrive with the Word alone, they also bring a dominant culture's material, organizational and ideological ressources. Missionaries are agents for so much more than the Word, just like converts have very complex motives which may be far from spiritual. In Denmark mission is no problem, but rich missionary organisations in poor cultures are far from being politically neutral.

There are examples of missionaries in the 60s and 70s who cooperated with the fascist dictatorships installed and supported in Latin America by the US government. ("There is no government on Earth that God hath not permitted to come to power." Romans 13:1). Some have also let themselves be used by international companies' attempts to take over natural ressources from indigenous peoples, land, forest, oil etc. That is what one hears from indigenous representatives and anthropologists. (Document).

But Wycliffe/SIL have also changed with time. Even though Ethnologue did develop as a tool for American cultural export, the religious aspect has been toned down, thank God, in this the 15th edition of the most detailed and accessible survey of the world's languages available. For linguists it is an indispensable overview, but its data should be taken with a load of salt.

... was the warning from Ole Stig Andersen who had studied Ethnologue, a survey of all the languages of the world, and many more, it appears.


Søren Hvalkof and Peter Aaby (eds): Is God an American?
An Anthropological Perspective on the Missionary Work of the Summer Institute of Linguistics.
Published by IWGIA (International Work Group for Indigenous Affairs) and SI (Survival International),
Copenhagen 1981.   ISSN 0105-4503   ISBN 87-980717-2-6

Mutual Intelligibility, Comment, July 24, 2005

I am not implying that all dialects of Chinese resp Arabic are mutually intelligible. They certainly are not, and Ethnologue may even underestimate how many "languages" they can be split into, using their criterion. In my work with immigrants in Denmark I have noticed that uneducated Maroccans and Palestinians/Iraqis do not communicate easily in Arabic. And South Jutish and other fringe dialects of Danish are indeed not readily understable to all Danes. But does that make them different languages?

My Arab acquaintance certainly doesn't believe (nor wish) that all dialects of Arabic are mutually intelligible. Living in Denmark where he encounters other kinds of Arabic, he knows full well they are not. But he (and all other Arabs I know of, except possibly the Hasaniya of Western Sahara (and of course the non-Arab, but kind-of-Arabic-speaking Maltese)) still insist that these dialects all are (inferior) instances of one Arabic language. And they behave as if.

All the modern mutually unintelligible dialects of Arabic stem - with or without substrates from earlier/local languages - from Classical Arabic, an instance of which is Quranic Arabic, a dialect that is respected by the whole Arabic language community, be they Muslim or Christian or whatever. It is everybody's written language and nobody's mother tongue. It is indeed spoken today in all Arabic communities, although not by everybody, and in different - mutually highly intelligible - dialects. It is everybody's ideal. Therefore Arabic is one language community and not 35.

It can also work the other way round: Serbian, Croatian and Bosnian are certainly mutually intelligible and were previously considered a single language labeled Serbo-Croatian, but now it is considered three languages by the natives who are eagerly striving to make them as different as possible, especially in vocabulary.

The problems arise where the Ethnologuers use their structuralist linguistic criterion for language/dialect classification to create and label separate linguistic identities ("Ethno"-logue), also in cases where the populations concerned think otherwise. There is no doubt that all speakers of Eastern Bornholm and Southern Jutish consider their languages to be dialects of Danish, not independent languages on a par with Danish. As does in fact every informed linguist but Ethnologue.

The idea of mutual intelligibility presupposes that languages enjoy equal status and have the same value and importance. But they almost never do. There is always dominance and privilige of different sorts involved, since Man is an hierarchical ape.

Intelligibility is not mutual, but asymmetrical. Speakers of minority variants understand and speak the dominant language or dialect, if necessary, whereas the opposite seldom is the case, because it is seldom necessary. Very few Danes speak Bornholm, but all Borholmians speak a variety of standard Danish beside their own. There are loads of "Danish" loanwords in Bornholm and almost none the other way round.

Intelligibility is also a case of beliefs, of imagined communities. Standard Danish, Norwegian and Swedish are generally considered separate languages, because there are three states, but they are in fact mutually intelligible, and thus one language according to Ethnologue's criterion, but not their listing. But we Scandinavians hold the belief that we are different but understand the two other languages, or at least that we ought to. (More than once have I seen people get ashamed in situations where they did not understand a fellow Scandinavian).

(An interesting - and sad - contemporary development is that the reciprocal intelligibility of Scandinavian dialects seems to be declining, presumably due to all speakers' growing competence in English. A quite special example of domain loss.)

The situation is highly asymmetrical, with Danes understanding Norwegian best, Swedish next. Swedes understand Norwegian best, Danish less (many not at all). Norwegians understand Danish best, Swedish next, both easily. Some fringe dialects of these three languges are not mutually intelligible at all. How many languages will one count in such a situation? Ethnologues solution is absurd and even by its own standards inconsequent.

The language aspect of ethnicity can not be judged solely by the abstract linguistic criterion of mutual intelligibility, it must be weighed against an evaluation of the asymmetrical social value and economic usefulness of the languages in question, and the native groups' resulting own beliefs and imagined communities.

Like Arabic and Chinese. And, in all modesty, like Danish.

Ole Stig Andersen

Danish Dialects

The dialects of Danish originated around 11-1200. The last 150 years of extreme standardisation of Danish society (democracy, industrialisation, urbanisation, roads, public school, massmedia) have taken a heavy toll on the dialects, and the general picture today is one of a quite uniform Standard Danish that comes in a handful of regional flavours.

Historically, Danish split into three main dialects:

The Central dialect originally spoken on the island of Sjælland where the capital København is situated, has of course become the basis of Modern Standard Danish, generally called Rigsdansk ("Realm Danish").

The main dialect of Eastern Danish was Skånsk (Scanian) spoken in Skåne, which was conquered from Denmark by Sweden in 1658. A harsh Swedification process immediately ensued and today Skånsk is no longer a dialect of Danish, rather a dialect of Swedish on an Eastern Danish substrate. There may be a few old people who still know Danish Skånsk. It is interesting to note that Swedish and Danish are mutually intelligible to this day, and of course they were even more so 350 years ago.

Another dialect of Eastern Danish is Bornholmsk which has survived, more or less, on the Eastern part of that island. The Western part with the seat of the administration was Copenhagenized and is now not an Eastern Danish dialect any more, but a central Danish dialect on an Eastern Danish substrate. So there are two kinds of Bornholmsk, quite different, but both recognizable by their similar peculiar "singing". Eastern Bornholmsk may require some getting-used-to for other speakers of Danish, but not much.

Western Danish, i.e. Jysk (Jutish), is spoken on the peninsula of Jutland in several varieties with a major dialect divide between North and Sønderjysk (South Jutish). Sønderjysk developed on Danish' interface with Low Saxon and Frisian. The westernmost version of Sønderjysk does require some getting-used-to for other speakers of Danish, as do a couple of western and northern dialects of Northern Jutish. Sønderjysk is seldomly written, it has no standard. Interestingly, Sønderjysk is also the home language of the German minority in Sønderjylland.

'Rejsende dansk', Traveller Danish, was a blend of Danish and Plattysk (Low Saxon) with some admixture of vocabulary from Rotwelsch and Gypsy. It originated with the dismissal of hired soldiers in Holsten in the early 1800s and was the language of an economical niche of itinerant families travelling around 1850-1950 with circuses, tivolis, markets and the like, never exceeding a few hundreds. This jargon/argot has hardly survived but in parody: "cirkusdansk", Circus Danish.








til top
© Ole Stig Andersen, June 20, 2005 (uploadet July 23, rev Jan 16, 2006)