Foundation for Endangered Languages

Home | Manifesto | Membership details | Proceedings | Grant Applications | Newsletter | Links | Bibliography


3. Other Reports from the Field

Three African Field-Reports from Roger Blench: Ganang, Sambe and Tchumbula

Roger Blench,
Overseas Development Institute
111, Westminster Bridge Road
London, SE1 7JD, United Kingdom
ODI Web page:
Own Web Page:

The strange case of Ganang

Languages become threatened in different ways and occasionally gender rather than generation and ethnicity form part of the nexus. Ganang seems to represent a rather extreme case of gender differentiation in the process of language loss. Ganang or Gashish is often listed as one of the dialects of Izere, a significant Plateau language spoken north of Jos in Central Nigeria. No data on this language has ever been published and no Izere informants in Jos could tell us about the language. As a consequence I decided to go to try and resolve its status.

As we approached the Ganang-speaking area, we found that the Ganang, locally known as Gashish, are considered to be Berom, and indeed culturally they share much in common with their Berom neighbours. The Ganang language is spoken in Gashish Kuk village in Plateau State, Nigeria. Gashish Kuk is one hour's drive southeast of Jos, beyond Kura falls.

We encountered an old man sitting under a tree and requested him to help us fill in a wordlist. He readily agreed, but it very soon became clear that he did not speak the language, although he claimed to be Ganang. However, a group of women had gathered around us and began answering the questions in his stead. I soon switched to using them as the principal informants and Mrs. Cundung Bulus and Mrs. Cingun Mandong were able to help me complete a basic 400-word list on the 18th of January 2001.

Despite gathering quite a crowd it became clear that none of the men present could speak Ganang, despite the linguistic competence of their wives. However, the women were unable to produce vocabulary from the male world, particularly in relation to hunting, and so I was not able to elicit words for 'arrow' or for large mammals. The men speak principally Berom, and increasingly Ron, a Chadic language of recent migrants, as well as Hausa, the lingua franca of the area. The men said that these other languages were 'better' or 'more prestigious' than Ganang, while the women said they would continue to speak Ganang with their children. Indeed, young male children were heard speaking Ganang, so they must stop speaking it at a certain age. Husbands and wives communicate with each other in Berom, or increasingly in Hausa. Long-term bilingualism in Berom was later confirmed by the data analysis which indicated high levels of interference between the two languages. Linguistically, Ganang turns out to be a form of Izere that has been Beromised. The phonology and noun-class system have taken on features of Berom and it is for practical purposes unintelligible to mainstream Izere.

It turned out to be very hard to gauge the number of competent Ganang speakers, as most individuals are multi-lingual, also speaking Ron, Hausa and Berom. Almost all settlements are mixed, with Ron and other outsiders. The nearby settlements of Hye and I˜yø were reported to be principally Ganang but the same gender division of linguistic competence applies. Overall there are unlikely to be more than 3000 ethnic Ganang, but many fewer speakers. This unusual gender division makes it hard to predict the future of Ganang but it should definitely be regarded as threatened. A definite case for intensive sociolinguistic research.

A first record of Sambe, a language that is nearly extinct

On the 11th of February 2001, I was working on the Ayu language, with the assistance of Barau Kato. We asked if there were other languages spoken in the area, and the name of Sambe came up. No such language is listed in any reference source on Nigeria, so we went in search of it. After several wrong directions we came across the last speakers the Sambe language, spoken in a single village, Sambe, some 10 km. west of the town of Agamati, on the Fadan Karshe-Wamba road in Kaduna State, Nigeria. A short wordlist was collected by Roger Blench with the assistance of Atsar Musha and group of villagers. The name of the language and people, as well as the settlement where they live appears to be Sambe; we were unable to clarify this issue further. Sambe is a nearly extinct language and our informants were all very aged, hence the shortness of the list. We hope to return and extend the list at some future date.

Sambe is spoken by six people, three men and three women. All of these are extremely aged and the principal informant was said to be over 100 years old. Recall of the language is good and it is apparently still spoken between these individuals, though Ninzo is the usual language for communication with the rest of the village. Many other people of a slightly younger age have some knowledge of the language and can produce isolated words, but were apparently never fluent speakers. Sambe has given way to Ninzo and is effectively moribund; within 5-10 years it will be spoken no more.

Analysis of the language showed that it is of considerable importance linguistically. The external cognates show without doubt that the closest language to Sambe is Hasha, although the relationship is not that close and that the Arum-T\su and Toro languages are also related but further apart. Sambe is geographically between Hasha and Alumu and links together what were previously isolated Plateau languages.

The best guess for the internal structure of this group is:

Sambe is a nearly extinct language and our informants were all very aged, hence the shortness of the list. We hope to return and extend the list at some future date.

A recent report on Tchumbuli, Benin republic

A recent (June 2001) report on the Tchumbuli language of Benin (Schoch & Wolf 2001) provides some information on an otherwise barely-reported speech-form. Tchumbuli is a Northern Guang language spoken in three villages in the Departement de Collines between Savé and Ouessé. These villages are Okounfo, Gbede and Edaningbe and their total population is 3500 individuals.

The origin of the Tchumbuli is complex; they are closely related to the Chumburung of NE Ghana and oral tradition suggests that they migrated to their present site in the mid eighteenth century. However, while in Ghana they absorbed the 'Cobecha', mercenaries from Benin (the precolonial state in Nigeria) who had come to fight in the Ashanti wars and halted on their way home. This ethnic distinction is maintained in the Tchumbuli communities in Benin Republic today, despite the homogeneity of the spoken language. To add to the confusion, in the 1950s an expedition led by their Paramount Chief returned with a number of families back to Ghana and settled in Anyinamae, near the present-day Chumburung community. Their language has effectively been relexified and absorbed back into Chumburung.

Tchumbuli is slowly dying as a result of contact with two major neighbouring languages, Maxi and Cabe. Maxi is related to the Fon group while Cabe is a type of Yoruba, closely related to that spoken across the border in Nigeria. In Okounfo village, the switch to Cabe has occurred, with pervasive bilingualism and Tchumbuli only known to the older generation. In Edaningbe, Maxi is replacing Tchumbuli although a more complex ethnic mixture in the village means that the process of replacement is less straightforward. In Gbede, Tchumbuli remains widely spoken although Cabe is used to communicate with outsiders and appears to be spreading among younger children.

The total number of speakers of Tchumbuli was estimated at 1838, and although this is relatively high compared with many other threatened languages in West Africa, it conceals the fact that the language is largely confined to the older generation. Paradoxically, the Tchumbuli are proud of their historical traditions and their links with Ghana. Tchumbuli illustrates the problem of how much weight should be given to languages close to those that are not threatened. Tchumbuli is sufficiently close to Chumburung for linguists to classify it as a dialect. However, the results of complex interactions with Maxi and Cabe and the very different cultural traditions of the Tchumbuli have made the language quite distinct above the level of fundamental vocabulary.


Wolf, K. & G. Schoch 2001. A Sociolinguistic survey of the Tchumbuli language area. Unpublished report. SIL, Cotonou, Benin.

Outré but not “Out” in 1960s Britain: the Polari Language

Paul Baker, Lancaster Univ.

In the late 1960s, the BBC Radio programme Round The Horne was tremendously popular, attracting about 9 million listeners a week. A mishmash of comedy sketches, the most popular featured two “camp”, out of work actors called Julian and Sandy (Kenneth Williams and Hugh Paddick) who would greet the long suffering Mr Horne with “Oh hello, how bona to vada your dolly old eek!” Every week, thanks to Polari, Julian and Sandy made a mockery of the BBC's censors. For example, in one episode, they are domestic helps and have been shown into a kitchen where they are expected to get to work. "I can't work in 'ere," complains Julian. "All the dishes are dirty!" "Ooh speak for yourself, ducky!" retorts Sandy. The audience would probably get the use of the word dish as an attractive young man, as in "Isn't he dishy?", but hardened Polari speakers also know that dish refers to a person’s backside, which would afford them an extra special laugh.

Their use of Polari followed a long tradition - it had been known by actors and gay men in the U.K. for decades. But fast forward a few years and Polari has almost vanished from gay circles. Mention it now and you'll more likely than not to get a blank look, especially from anyone under 30. And those who do profess to have heard of it are likely to only know a handful of words.

It's impossible to pinpoint an exact date when Polari came into existence. It most likely arose from a type of 19th century slang called Parlyaree which was used by fairground and circus people as well as prostitutes, beggars and buskers. Many of these travelling people worked all over Europe, and as a result a fair number of the old Parlyaree words resembled Italian. The music halls of the 19th Century eventually replaced these wandering entertainers, and out of music halls developed the theatre. Parlyaree gradually morphed into Polari (or Palare as it was earlier known), being picked up by gay actors and dancers - who introduced it onto London's gay subculture.

But there were lots of other influences - The East End of London was full of vibrant communities, and so we find bits of Yiddish (schwartzer: black man, schnozzle: nose) coming into Polari. The docks were popular places for people who wanted to meet sailros - who used Lingua Franca. As a result, bits of Lingua Franca appear in Polari. Then throw in some Cockney Rhyming Slang and the less well-known backslang - the practice of saying a word as if it's spelt backwards (hair = riah, face=ecaf). Finally, in World War II add some American terms (butch, cruise) as gay men befriended and entertained homesick American G.I.s, and then throw in a few words stolen from 1960s drug culture (doobs: drugs, randy comedown: a desire for sex after taking drugs) for good measure. The result is a complex, constantly changing form of language which appears slightly different to whoever uses it. While many people used it as a lexicon, those who were most adept at it, began to invent their own grammar and morphology, making Polari more like a unique language. In researching Polari for a doctoral thesis I have collected a lexicon of over 400 items.

Polari flourished in the repressive 1950s, where the control of post-war sexual morality was viewed as a priority and prosecutions against gay men reached record levels. As being openly gay was dangerous, the need for a language that protected gay men, and at the same time acted as a kind of "gaydar" by allowing them to recognise others, was extremely useful.

By the 1960s, the political situation had begun to change. Polari was used less to cautiously "out" yourself, and more for chatting with friends. Its vocabulary - full of words to do with clothing (lally-drags: trousers, ogle-fakes: spectacles) and parts of the body (thews: arms, luppers: fingers) and evaluative adjectives (bona: good, cod: bad), reflects what it was most often used for - gossiping about potential sexual partners, while the target was in earshot. "Vada that bona omee ajax - the one with nanti riah!" translates to "Look at that nice man over there - the one with no hair!"

However, in the 1970s, Polari started to fade from people's memories. Julian and Sandy had represented a swan-song of sorts in any case. In 1967 (the same year that Round the Horne was at its peak, winning the award for best comedy radio programme), the legal situation for the average gay man was improved with the implementation of the Wolfenden Report’s recommendations of ten years earlier. Homosexuality was partially decriminalised, and as a result, there was less of a need for a secret language. In addition to that, Julian and Sandy gave Polari a kind of doomed respectability - they had inadvertently blurted out the secret via the radio, into 9 million homes a week.

And ultimately, there were political reasons for ditching Polari - it was associated with oppression, and the early Gay Liberationists wanted to put all of that behind them. It was rather easy to criticise Polari as being sexist, racist and brimming over with internalised homophobia. Writers of the early 70s are quick to cast Polari as ghettoising and politically incorrect. By the beginning of the 1980s, Polari had all but vanished.

However, in the 1990s, the situation changes again. Polari is undergoing a revival of interest. It's now possible to view it as part of gay heritage - a weapon that was used to fight oppression, and something to take pride in again.



For example, the London branch of the Sisters of Perpetual Indulgence use "High Polari" in their blessings, sermons and canonisations - adding religious mystique while also acknowledging the historical origins of their ceremonies.

It's unlikely that Polari will ever be revived to the extent that it was used in the 1950s - but that's no shame. And in any case, little bits of Polari have even been incorporated into mainstream slang. For example - the word naff was originally used as a Polari acronym meaning "Not Available For Fucking". Now it simply refers to something that's tasteless. Non-Polari speakers must have heard it - "oh don't bother with him, he's naff!", inferred it meant something bad, and started using it themselves - not realising that the word was originally an insult hurled at them.

Paul Baker is writing book on Polari, which will be published by Routledge towards the end of the year. Note on the Kusunda Language

- B. K. Rana

This note has appeared in JANAJATI, 2. 2 - a governmental journal of nationalities or the ethnic peoples of Nepal.

Ethnologue survey of languages in Nepal has painstakingly dug out more numbers of languages than they actually are there, offering independent nomenclature to them to increase unreal number of languages in the country.

For example: it mentions - Byangsi, Chaudangsi and Darmiya as three different languages spoken in Darchula District, of far west Nepal but they are dialects of Shauka language which I had an opportunity to study last year. Likewise, the survey report presents Tarali Kham known as Kaike, Kham Gamale, Kham Maikoti, Kham Nishi, Kham Sheshi and Kham Takale as different languages of the area which should also have been introduced as Magar language of Karnali area. The Magar language of that area is publicized as Kham Magar Kura but in fact, Kham does not mean any language category. It refers to an administrative unit set by then Yumila kingdom to rule over the indigenous Magar peoples of that area.

Now, the practice of offering a nomenclature as Kham and Kaike for Magar languages of Karnali area requires linguistic redefinition and new recognition as well.

Kusunda is not a Dead Language
Kusunda is one of the endangered languages in Nepal. At the moment, there are only three speakers of this language: Prem Bahadur Shahi Thakuri of Dang, Lil Bahadur Kusunda of Pyuthan and Raja Mama of Tanahu.

Prof. Sueyoshi Toba, one of the Kusunda authorities, who first analyzed the language scientifically, in association with Johan Reinhard, now believes that Kusunda is not a dead language and further states that ‘we do not call a language ‘dead’ or ‘extinct’ as long as there is anyone alive who knows even a little of the language in question’ (Toba 2000).

Kusunda had already been declared extinct following the death of Raja Mama's mother, the presumed last speaker; who died of diarrhoea few years ago in Damauli of Tanahu District, west Nepal. Although, there are very limited noun phrases and a remarkable loss of major word classes including verbs and their patterns, yet Kusunda is not a dead language because there are at least three Kusunda speakers physically alive in different parts of the country, which I have mentioned above.

Kusunda is one of the unique languages found in the southern Himalayan region, primarily in Nepal, which was recorded and published, for the first time, by Brian Houghton Hodgson. The Hodgson wordlist of 1857 ( Hodgson 1992 reprint ) contains only 223 words and fifteen sentences collected through supposedly available trained-hands of those days. It is understandable that Nepali was even then the lingua franca.

The Rana Regime (1846-1950) had barred Hodgson from visiting Kusunda areas in rural Nepal. It is believed that he could not have any opportunity to listen to Kusunda utterances by himself. Researchers in Linguistic Survey of India Team carried over his works. But, ‘one is to argue that Hodgson (from whose article the Linguistic Survey of India drew its Kusunda vocabulary) was a well-meaning Victorian amateur whose data are worthless, whereas those of Reinhard and Toba are the reliable findings of modern professionals’ (Whitehouse 2000).

Following Hodgson’s return to his country, Kusundas and their language remained ignored for a long time until Narahari Nath Yogi tried to write something on them in 1955. And in 1970, the anthropologist Johan Reinhard from Austria arrived here and took an interest in them. He recorded some sample sentences and hundreds of Kusunda words, brought them to Katmandu for analysis, at a time when the language was only spoken by few Kusundas of central hills Nepal. Prof. Sueyoshi Toba, a linguist from Japan worked together with Reinhard, analyzing the record in a standard linguistic framework. Both of these scholars’ contribution to Kusunda community is essentially very great for their reports are the only authentic source of information on Kusundas, their language, their plight and other sorts of things related to them.(Reinhard & Toba 1970).

Kusunda is Tibeto-Burman
Kusunda language, along with its unique morphology and syntax is both interesting and intriguing as linguists, until today, have not been that able to define exactly what form of language is this and in which family it falls. Meanwhile, it is prevalently believed that this language is a ‘single, isolated, traditionally hardest to classify’ - not matching with any of the major language families found in Nepal and also has external relationships with Nihali, a language that has some 5000 speakers in the Indian state of Madhya Pradesh (Whitehouse 2000).

However, this observation lacks sufficient comparative studies with other Tibeto-Burman languages found around the traditional home of Kusundas in the central hills of Nepal. Without having any comprehensive work, as required, Kusunda language should not have been labeled as an isolated language. Nevertheless, I am fully convinced that it falls in Tibeto-Burman category because there are a great number of Tibeto-Burman cognates and phrases present elsewhere in this language and remarkable similarities in grammar of Magar Language of Karnali area. Below is a functional explanation of Kusunda cognates and their comparison with other Tibeto-Burman languages found in Nepal:

a) Kusundas have ‘tang’ [ta+h] for water, Shaukas and Chepang have [ti] and Magars say it [di]. In ‘tang’ we have voiceless alveolar ‘t’ of Shauka and Chepang ‘ti’. And, Shauka and Chepang ‘ti’ is voiceless representation of Magar ‘di’.
b) For fish Kusundas say ‘ngsa’= [h+sa], Magars of Karnali area say ‘nga+sya‘ , Chepangs say it ‘nya or nga’ ,Baram say it ‘nanga’ and Magars of Gandaki area say it [di+sya]. The Kusunda segment ‘h’ of ‘ta+ng’ i.e. ‘water’ stands here to denote ‘water related object’ and ‘sa’ for meat > ‘meat from water = fish’. These segments: [h+sa], [nga+sya] and [di+sya] have same meaning and morphologically, the formation of these words are distinctly similar.
c) Blood is ‘yu+ei’ in Kusunda, whereas, it is ‘chyu+huei’ in ‘Balkura’ and ‘wei’ in Chepang language. These three words are phonetically similar in these three languages.
d) ‘Aagai’ is a dog in Kusunda, whereas in Baram language it is ‘aakyo’ and here voiced velar ‘g’ is present in Kusunda ‘aagai’ and voiceless ‘k’ in Balkura.
e) Generally speaking Kusunda phonology is that it has initial ‘ h’ distribution in at least two words, so far found, for example: fish = [h+sa] and wife = [h+yang+di]. The initial ‘ng’ segments in these words suggest their root from Tibeto-Burman language. One of the major characteristics of Tibeto-Burman languages is that they have initial ‘ng’ segment in their cognates.
f) In Kusunda words like ‘ngsa’, ‘ngyangdi’, ‘dimtang’ lahang’, mangmi’, ‘kapang’ , ‘gelang’, ‘pinjang’ ,‘ghinga’ , natang’ , ‘chining’ ‘chan’ ‘iping jing’ ‘ing’ etc. we can find [h] in initial, middle and final distribution. These all words have Tibeto-Burman characteristics.
g) Tamangs say ‘moje’ for banana and ‘kakhare’ for a crab. Kusundas say them ‘mucha’ and kakchi’ respectively,
h) Bhotes say ‘manjya’ for millet and ‘du’ for a snake and Kusundas have ‘mangmi’and ‘tu’ for them respectively.
i) As concerns numerals in Kusunda, there is ‘ghinga’ for count number ‘2’ and in Baram it is ‘ni’ and ‘ngi’ in Thaksya [Thakalis (Bhotes?) who lived around Thak Satsaya Khola] (Grierson 1909). The nasal [h] in these words is very similar to Magar ‘nish’ for numeral 2. Similarly, Chepang say ‘ploin-zho’ for ‘4’ - Kusunda say it ‘pinjang’ and in Baram it is ‘bi’ - here too, we can find bilabials [p] and [b] present in these three words.
j) Chepangs say ‘micha’ for a goat and in Kusunda and Baram languages it is ‘mijha’ and ‘michha’ respectively. And, a road is ‘un’ in Kusunda and ‘ungma’ in Baram language.

i) Let us again consider the Kusunda pronominalized sentences and phrases below: I eat rice = ‘ci kaadi taamaanan’ [ it should have been ‘chaamaanan’. Raja Mama says ‘gaamaanam’];
You eat rice = ‘ nu kaadi naamaanan’
He eats rice = ‘ git kaadi gaamaanan’
My stomach = ‘cie cimat’;
Your stomach = ‘nie nimat’;
His stomach = ‘gidie gimat’ (Bandhu 1999).

ii) The Magars of Karnali area say ‘ge +pang’ for ‘their own language’ and ;’ rangpang’ for Khas Kura . Kusundas also say ‘gi+pan’ for their language. The Magar language of Karnali area is one of the Tibeto-Burman languages which also has pronominalized sentences and phrases as outlined below:
I eat rice = ‘nga [ha] yai/kang jyonga’
You eat rice = ‘nang [nh] yai/kang jyona’
He eats rice = ‘wola yai/kang jyowa’

My stomach = ‘nga phu’
Your stomach = ‘na phu’
His stomach = ‘wa phu’

The above Kusunda sample cognates, phrases and sentences are enough to establish that Kusunda is not an isolated language. It offers me a great sense of satisfaction and happiness to put forward this proposition that there should be a language which has cognates and grammar of a match with certain language family found close to its geographical boundaries; undoubtedly, the language belongs to any of the major families of languages that other languages belong thereto; and eventually evolves itself in its own pattern sharing with the existing principal ingredients and characteristics of other languages following the courses of different time intervals. It is therefore, I have a firm belief in these findings that having in it numerous cognates, morphemes and phonemes very much matching with Magar, Shauka, Baram, Chepang Tamang, Thaksya (?), Bhote, Bhujel (Rana 2000) and other languages; Kusunda is definitely a Tibeto-Burman language and not an isolate.

Kusunda can be reintroduced
Kusunda is not a dead language, therefore, we might be able to reintroduce it by opening a language council, at central level, which would have linguistic and cultural planning as required by the nation. To fulfil this objective the present National Committee for Development of Nationalities - NCDN or any entities that are formed later by legislation, should be entrusted with extra resources: both financial and legislative; so as to bring in those three Kusunda speakers together in one place, re-establish their natural living with lesser modernisation - so that they would have a natural environment to slowly recall the forgotten language to reintroduce among themselves.

References: Bhandhu, C. M. ( 1999): Keynote Address to the Fifth Himalayan Language Symposium Kathmandu Sept. 13-15, 1999.
Grierson, G. A. ed. ( 1909): Linguistic Survey of India: New Delhi, India.
Grimes, Barbara ed. (1996) Ethnologue: languages of the World/SIL.
Hodgson, Brian H. ( 1848) On the Chepang and Kusunda Tribes of Nepal. Journal of the Asiatic Society of Bengal.
Krauss, Michael (1992 ) The World’s s in Crisis, Language.
Rana, B. K. (2000): An Ethnographic Study on the Shaukas of Byas Valley, Janajati Vol. 2 No. 1, Journal of Nationalities of Nepal 2057, Kathmandu.
Rana, B. K. ( 2000) : Eklo Jivan Jiudaichhan Kusunda Raja, Samadristi Weekly, Vol.3 Nos.11, 12 and 13 - February 15, 22 and 29, 2000. Kathmandu.
Rana, B.K. (2000): On Genetic Preservation of Kusundas and Reintroducing Their Language. (Paper presented in a workshop on preserving endangered languages, Kathmandu, May 19 -21, 2000).
Rana, B. K. (2000): Writing on a Dead Language, The Kathmandu Post, April 16, 2000.
Rana, B. K. (2001): Kusunda Bhasa: Thapa Addhyan (Kusunda Language: an Additive Study ) Samadristi Weekly Vol. 4, Nos. 3,4 and 5 -January. 7,14 and 21 2001, Kathmandu.
Reinhard & Toba (1970): A Preliminary Linguistic Analysis and Vocabulary of the Kusunda Language, University of Vienna and Summer Institute of Linguistics, Tribhuvan University, Kathmandu, Nepal.
Toba, Sueyoshi (2000): Kusunda Wordlists Viewed Diachronically, Central Department of Linguistics, Tribhuvan University, Kathmandu.
Whitehouse, Paul (2000):The External Relationships of the Nihali and Kusunda Languages. London.
Yogi, Narahari Nath (1955): Itihas Prakashma Sandhipatra Sanghra, Ratna Pustak Bhandar, Kathmandu Nepal.