Foundation for Endangered Languages

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


5. Field Work Reports

Report on the Tarokoid languages from Roger Blench

Some time in 1995 I proposed fieldwork on the Tarokoid languages of Central Nigeria to establish their status and provide some initial documentation. According to recent classifications, Tarokoid consists of five languages, Tarok, Pai, Bashar, Turkwam and Arum-Chessu. Of these, only Tarok can be said to have any significant published material. To try and establish a solid basis for classification can be achieved only with the collection of new field materials. With this in mind it was decided to see whether a group such as Tarokoid actually existed and if so, could a justification be provided for including or excluding individual languages. In addition, information about the exact name of the languages or even the location and number of speakers is woefully inadequate. To fill these lacunae seemed a reasonable subsidiary goal.

In 1992, some material on Turkwam and Arum-Chessu were collected, while during May, 1996 I was able to complete fieldwork. The following is short report on the findings. The villages of speakers of each language were identified and substantial wordlists were taken as follows:

Arum-ChessuArum Kado10/11/92Musa Kado
TurkwamTurkwam10/11/92Yakubu Kos
Pe (a.k.a. Pai)Pai17/5/96Danjuma Torot
Yangkam (a.k.a. Bashar, Basherawa)Yuli19/5/96Salihu Mohammadu, Amadu Kondon Yuli
YangkamTukur22/5/96Abubakar Sulaimanu, Ibrahim Sale, Yushau and Idi Zuberu

Summaries of individual languages

1. Tarok. The Tarok live in and around Langtang in the southeast of Plateau State. There are probably some 150,000 speakers. An orthography has been developed together a literacy programme and the New Testament has been translated and published. The language is being well-maintained and there is no reason to believe it is under threat. Fieldwork took place within the framework of the Tarok dictionary project (Longtau and Blench, forthcoming).

2. Pe. The correct name of the Pai is Pe (singular uPe, plural aPe). There are some 2-3,000 speakers in seven villages. The main settlement is 17km south of the main road from Jos-Amper, turning a few kilometres before Amper. The settlement is extremely isolated with a single road that is cut during the wet season. The language is still well maintained with all the children encountered apparently fluent. Preliminary lexical material was recorded.

3. Yangkam. The standard reference on this people says that there are 20,000 speakers of the language located in and around Bashar, some 50 km east of Amper on the Muri road. This turned out to be entirely erroneous. The Bashar people seem to have been heavily affected by nineteenth century slave raids, perhaps by the Jukun as well as the Hausa. They were converted to Islam and a relatively powerful chiefs centre was established at Bashar. At the same time they began to switch to Hausa, while still retaining their Bashar identity.

In the Bashar region today, there are just two old men who remain reasonably fluent in the language, in the village of Yuli, some 15 km northwest of Bashar.



However, it turns out that at the time of the raids, the population split into two and another group sought refuge in a region west of Bashar, 25 km north from Jarme on the Amper-Bashar road. These people were not so heavily Islamised and have retained their speech to a greater degree. Bashar is spoken in some four villages, Tukur, Bayar, Pyaksam and Kiram. However, even here it only spoken by people over fifty and all the young people speak Hausa. The local estimate of the number of fluent speakers is 100, and falling every year.

The correct name of the Bashar language and people is Yangkam, plural aYangkam. Wordlists were taken from the two groups and only exhibited minor lexical differences. There seems to be no likelihood that Yangkam will be maintained as the speakers are quite content with the switch to Hausa, while remaining proud of their historical identity. Rescue linguistics therefore remains a high priority. Selbut Longtau is attempting to identify speakers for further data collection.

4. Arum-Chessu. Chessu is a single village and Arum a cluster of seven villages near Wamba in central Plateau State. The only difference between the two lects is said to be intonation patterns. The Arum are correctly called Alumu (plural Alumu-mbo). There are probably some 4-5000 speakers and the language appears to be still fluently spoken by young people.

5. Turkwam. Turkwam is a village northeast of Wamba with perhaps 3-4000 inhabitants. The correct name of the Turkwam is Toro (plural a-Toro-mbo). The Toro identify strongly culturally with the Kantana, a neighbour speaking a Jarawan Bantu language, but this is contradicted by their language which is clearly a Plateau language. Language maintenance still appears to be good.


Preliminary analysis of the lexical data suggests that the assignation of Toro and Alumu to Tarokoid is completely erroneous. They are clearly related to one another and probably form another subgroup of Plateau. Membership of Tarokoid should be restricted to Tarok, Pe and Yangkam. Within these, Pe and Yangkam are clearly more closely related to one another than to Tarok. A scholarly article is in preparation setting out the reasons for these conclusions in more detail.

Of the languages investigated, only Tarok is beginning to have an acceptable level of documentation. Yangkam is severely threatened and should be subject to an intensive investigation while speakers are still fluent. If a small sum of money could be found for transport and subsistence further work could probably be put under way. The other languages are clearly also in need of documentation and further research on their exact status.

Roger Blench

8, Guest Road, Cambridge, CB1 2AL, England
Tel/Ans/Fax +44-1223-560687
E-mail RMB5(at)