Northern Songhay Languages in Mali and Niger
A Sociolinguistic Survey

Michael J. Rueck and Niels Christiansen

Summer Institute of Linguistics

[A slightly abbreviated hard copy is to appear in: Trends in Nilo-Saharan Linguistics (Proceedings of the 1998 Nilo-Saharan Linguistics Conference), ed. by Norbert Cyffer (norbert.cyffer@univie.ac.at). Republished by permission.]

CONTENTS

Abstract

1. Introduction
   1. Objectives
   2. Appreciation
   3. Background
2. Demographics and description of the peoples
   1. Ingalkoyyu (Tasawaq-speakers)
   2. Igdalen and Iberogan (Tagdal-speakers)
      1. Igdalen (Tagdal-speakers)
      2. Iberogan (Tabarog-speakers)
3. Methodology
   1. Choice of test sites
   2. Choice of translators
   3. Interview schedules
   4. Word lists
   5. Dialect intelligibility testing
      1. The nominal method
      2. Deviations in this study
4. Results
   1. Dialect intelligibility testing
   2. Adaptation potential
5. Conclusions
   1. Tagdal
   2. Tasawaq
6. References
7. Appendix A. 380 item word list from each site
8. Appendix B: Intelligibility test texts
   1. B.1 Texte de Tamaya (tagdal)]
   2. Traduction du texte de Tamaya (tagdal)
   3. B.2 Texte de Tofabayogh (tabarog)
   4. Traduction du texte de Tofabayogh (tabarog)
   5. B.3 Texte de Ménaka (tadaksahak) [field transcription]
   6. Traduction du text de Ménaka (tadaksahak)
   7. B.4 Texte d'Ingal (tasawaq)
   8. Traduction du texte d'Ingal (tasawaq)
9. Appendix C: Intelligibility test data
   1. C.1 Ingal subject demographics and Recorded Text Test scores
   2. C.2 Tamaya subject demographics and Recorded Text Test scores
   3. C.3 Tofabayogh subject demographics and Recorded Text Test scores
10. Appendix D: Results of the Northern Songhay RTT

Abstract

Inherent intelligibility between Tadaksahak (Dausahaq), Tasawaq (Ingelsi), Tagdal and Tabarog was measured using the Recorded Text Test developed by the Summer Institute of Linguistics. Language attitudes of the speakers of the Northern Songhay varieties in Niger were also investigated using group interview schedules and by observation.

It was found that Tagdal and Tabarog are mutually intelligible and that speakers of both of these varieties consider that they speak the same language, for which all accept the name Tagdal. Also, Tadaksahak is not inherently intelligible to speakers of Tasawaq or Tagdal, and Tasawaq and Tagdal are not mutually inherently intelligible either. Although many speakers of Northern Songhay languages are bilingual to some extent, it is pragmatic concerns which drive them to use languages of wider communication and they reserve a more positive attitude toward their own languages.

1 Introduction

1.1 Objectives

From May to June 1998, the Summer Institute of Linguistics (SIL) conducted a sociolinguistic survey among speakers of Tasawaq, Tagdal, and Tabarog in Niger. The research team consisted of Michael J. Rueck, Niger survey coordinator, and Niels Christiansen, working in Menaka, Mali among the Idaksahak (who speak Tadaksahak) since 1992. Some of the goals of the survey were:

1. to collect some basic demographic and cultural information on the speakers of the target languages,
2. to measure the level of inherent intelligibility between the speech varieties in Niger,
3. to evaluate the potential for these groups to use written material produced in the Tadaksahak language in Mali.

1.2 Appreciation

We wish to express our gratitude to the Minister of Interior of the Republic of Niger, the Prefects of the arrondissements of Tahoua and Agadez, the Sub-Prefect of Abalak, and the Chief of the Administrative Post of Ingal for facilitating this study. We would also like to express our appreciation to the Chief of the 7th Group of the Tamajaq People of the Azawagh and the Village Chiefs of Tamaya, Tofabayogh, and Ingal for their permission to study their peoples' languages and for calling people together to work with us.

1.3 Background

Robert Nicolaï identifies Northern Songhay ( songhay septentrional ) as a subdivision of the Songhay language group of the Nilo-Saharan language family (1981a). (See Figure 1.) These speech forms are also referred to as "mixed languages" since they have adopted phonological and prosodic features and lexical items from Tamajaq, the Berber language spoken by the Kel Tamajaq or Tuareg (Lacroix, 1968). Nicolaï claims that intercomprehension "is possible" within the Northern Songhay group.

                     Northern Songhay

                 /                     \

          Nomadic                       Sedentary

         /       \                      /        \

   Tadaksahak   Tihishit          Tasawaq         Korandje

               /        \         /      \

            Tagdal  Tabarog  Ingelsi     Emghedeshie

                                           (extinct)

Northern Songhay language names are related to the names of the peoples who speak them in Berber fashion. Tadaksahak is spoken by the Idaksahak; Tagdal is spoken by the Igdalen; Tabarog is spoken by the Iberogan; and Tasawaq is spoken by the Issawaghan who also call themselves the Ingalkoyyu.

2 Demographics and description of the peoples

Three of the Northern Songhay speech varieties are unique to Niger. Tasawaq is spoken in Ingal, and Tagdal and Tabarog are spoken in a more widespread area centered around Abalak (see Map 1). A fourth variety, Tadaksahak (also known as Tadawsahaq or Dausahaq) is mainly spoken in the Menaka Circle of Mali, but individuals of this group can also be found in Niger and other neighboring countries.

We have found no reference to these people groups as distinct entities in official statistics, where the Igdalen and Iberogan tend to be grouped with the Tuareg. In the civil administration, they belong to the 7th Group of Tamajaq People of the Azawagh. The civil administrator in Ingal ( chef de poste ) reported that the people of Ingal claim to be "Songhay" on identity papers, while people in Teguidda-n-Tessoumt write "Tamajaq". He took this to mean that they see themselves as part of these two linguistic communities.

Our demographic statistics are tentative. Most of them come from responses to our questionnaires which were estimates or are a result of our observations on the field.

Click on the map to display a larger version (22k)

2.1 Ingalkoyyu (Tasawaq-speakers)

Tasawaq is spoken by the Ingalkoyyu [Lords of Ingal] (sg. Ingalumboro [person of Ingal]). This was the name they gave when we asked what they call their people group. We also heard them refer to themselves informally as "Bingali" (meaning the people of Ingal). Ingelsi / Ingelshi ("language of Ingal") is another term for the language, used in literature, and some use the term Issawaghan for the people. When we asked about the term "Issawaghan", they said this term came from Tamajaq. However, Issawaghan is also apparently the name of their most populous clan. As far as we know, this is the only Tasawaq-speaking community in the world. From official figures and their own estimate, we set the number of Tasawaq speakers at about 8,000.

The Ingalkoyyu cultivate date palms and vegetable gardens along the wadi where the town of Ingal is situated. They also exploit the salty springs of Teguidda n-Tessoumt, about 80 km to the north, where they evaporate salt which is used mainly by herders to keep their livestock healthy. Money from the sale of this salt is used to buy millet and other necessities. Finely woven and colorfully decorated mats of palm leaves are another source of cash income.

The Ingalkoyyu also play host to the nomads of the region (Tuareg, Arabs, Fulani, and Igdalen) for the annual Cure Salée in early September, when herds are driven to the area around Ingal to profit from the salty water and grass found there.

2.2 Igdalen and Iberogan (Tagdal-speakers)

The Igdalen and the Iberogan have for many purposes been treated as one group, and their speech forms are closely related. Nicolaï uses "tihishit" as a common designator for these two speech forms (1981b:306); however, this term is ambiguous. "Tihishit" is a term of Tamajaq origin meaning "the language of the blacks". The Igdalen and Iberogan used it to refer to all Northern Songhay speech forms. The Tuareg near the Niger River expand the term to include all Southern Songhay dialects as well. We observed that the Iberogan sometimes refer to their language as "Tagdal", which indicates that they think of Tabarog and Tagdal as the same language; thus, we have chosen to use "Tagdal" as a cover term for both speech forms. We will use "Tabarog" to designate the speech of the Iberogan in particular, and let context indicate when "Tagdal" refers to the speech of the Igdalen particularly.

The relationship between the Igdalen and the Iberogan on a daily basis is not clear to us from our short visit. It is often said that the Iberogan were formerly the slaves of the Igdalen; however, the apparent existence of a noble class of Iberogan makes this seem unlikely. There does seem to be a tendency for the Igdalen to want to speak for the Iberogan as well, without this necessarily being accepted by them.

Both the Igdalen and the Iberogan reported that there were individuals and villages of their ethnic groups living north and west of Tchin-Tabaradene who no longer spoke Tagdal. On Map 1, we have shown the approximate areas where the Northern Songhay varieties are spoken today.

2.2.1 Igdalen (Tagdal-speakers)

Tagdal is spoken by the Igdalen, a fair-skinned people, as well as by the descendants of their former slaves. Their territory centers around the three important sites with an Agdal (sg. of Igdalen) as chief, Tamaya, Mazababu and Tiguirwit, and extends to the north of Ingal. A total of 1,500 households (perhaps 6,000 individuals) was reported for the area around Tamaya. Tripling this number for a total, we estimate that there are 10-20,000 Igdalen. As far as we know this includes the black population of descendants of their former slaves. We mention the distinction here only because it seems important to the people themselves.

The Igdalen used to be primarily herders living a nomadic lifestyle, but some have abandoned the nomadic lifestyle in exchange for small-scale commerce and gardening around water holes as alternative sources of income. Those with substantial herds still drive them to the salty springs in the area of Ingal for the Cure Salée . The Igdalen also help to fill the role of Koranic experts in the greater Tuareg society. They are conservative Muslims.

The Igdalen use the Arabic script or the Tamajaq script, Shifinagh, for writing those languages - or Tagdal! A French primary school was established four years ago in Tamaya, and some Igdalen children are attending.

According to a man from the Tabaho family of Idaksahak in Mali who had contact with an Igdalen family on the border of Niger, the Igdalen have a common ancestor with the Idaksahak. The Igdalen themselves wouldn't tell us about their history; however, although they would never intermarry with the Tuareg, they do consider the Idaksahak marriageable. Even this is very rare though because, according to the Tabaho man, the Idaksahak women would never accept the isolated life expected of them among the Igdalen.

2.2.2 Iberogan (Tabarog-speakers)

Tabarog is spoken by the Iberogan, a group of black origin, possibly also with associated descendants of former slaves. This group seems to be economically inter-dependent with the Igdalen, but only marry within their own group. In the past they also were mainly herders, living a nomadic life, but now they have become sedentary and live from subsistence farming.

The homeland of the Iberogan is mainly in the area southeast of Abalak where rainfall normally is sufficient to support extended fields. They are spread over an area organized around nine villages, each with its chief, where they are the majority (>90%), and sporadically elsewhere, usually where farming is possible. We estimate their population at 7000.

Livestock still form an important part of their economy, for milk and cash income. The Iberogan are also the sole producers of a particular kind of sleeping mat, made of thin, straw-like sticks woven together with thin strips of leather. These mats usually form the sleeping surface on the portable wooden beds used by many nomadic peoples in the area (Tuareg, Igdalen, Fulani and others).

3 Methodology

3.1 Choice of test sites

Tasawaq is spoken only in Ingal and Teguidda-n-Tessoumt. The Bernuses (1972) indicate that most of the inhabitants of Teguidda-n-Tessoumt are just there seasonally to work the salt pans and have their permanent residences in Ingal. So, we went to Ingal.

Since we wanted our samples of Tagdal and Tabarog to be as pure as possible, we looked for villages which were inhabited only by Igdalen or Iberogan, but not both. Ideally, we would have chosen villages which were not on a main road (limiting their contact with other speech forms) and which had primary schools (for ease of test administration). However, we were constrained by the insecurity of the region between Abalak and Agadez to choose a village on the main road for our Tagdal sample. Tamaya was suggested to us by the Sub-Prefect of Abalak, and we had the good fortune to meet the village chief the following day in Abalak. A primary school was established there four years ago. The Igdalen chief, in turn, recommended Tofabayogh to us as an Iberogan village. Its relatively isolated location should ensure a stable linguistic community. There is no public school in any Iberogan village.

3.2 Choice of translators

Since no team members spoke any of the dialects in question, we had to rely on translators in order to accomplish this study. There are many young men in Ingal who have completed at least a primary education, so we had no trouble finding native Tasawaq speakers to translate from French for us. The chef de poste assigned one young man from his staff to help us during our stay, and other local men were available when we needed them.

The very conservative Igdalen and Iberogan, however, have resisted sending any of their children to public schools until very recently, so very few of them speak French. The Representative of the Chief of the 7th Group of Tamajaq Peoples in Abalak was able to call upon Alassane Ntinicar, a young ( approximately 25 years old), black Agdal (Kel Amdit) who had grown up in Agadez and had completed three years of middle school. He and his younger cousin, Abdoulaye Islamane, translated for us in Tamaya and Tofabayogh. We can't be sure whether their youth or their race influenced the data we collected. We did observe that, between themselves, they spoke more Hausa than Tagdal. Christiansen's knowledge of Tadaksahak also helped in the translation process.

3.3 Interview schedules

In each village we visited, we interviewed a group of people which was either called together by the chief or formed out of curiosity. In this way, we hoped that the answers and opinions expressed would be the consensus of the group rather than the beliefs of just one person. In Ingal, the chief invited 5 men and 5 women, all more than 40 years old, to the interview. In Tofabayogh, about 20 men, ranging in age from 16 years to more than 50, with the majority being around 30 years old, participated in the interview. In Tamaya, a group of men, aged approximately 25-55 years, assembled around the village chief to answer our questions. The size of the group varied between 6 and 15 during the course of our interview. The group interaction in Tamaya was restrained, and after the first day, the chief forbade us to return to work there unless he was present. Whether this was because he felt we would find people uncooperative without his support or because he wanted to control what we heard, we cannot be sure.

We asked questions from both a general demographic and a general sociolinguistic interview schedule. The answers to the former told us about nearby schools, markets, and government services; the latter dealt with perceived dialect differences, language attitudes, bilingualism and language use. The information gathered in this way provided a context for interpreting the results of the dialect intelligibility testing, and furnished background information.

3.4 Word lists

Christiansen, who has been living among the Idaksahak and studying their language, Tadaksahak, since 1992, elicited a 380-item word list in each test site. The list is a combination of the Swadesh 200-item list, the SIL Africa Area 200-item list, and the SIL Burkina Faso 270-item list. The glosses and data are presented in Appendix A.

3.5 Dialect intelligibility testing

3.5.1 The nominal method

Dialect intelligibility testing (Casad, 1974) essentially involves recording a 3 to 5 minute text in each of the speech forms of interest and testing comprehension of the text among speakers of the other speech forms. It is used primarily to measure the degree of intelligibility between various dialects inherent in their linguistic similarity. Below is a brief description of the steps in the preparation and administration of the test:

1. Two texts are elicited from a native speaker of Dialect A: one is a very short text used to "teach" the testing method and weed out unreliable subjects; the second is longer, approximately 3 to 5 minutes in length. It should be autobiographical in nature and rich in details. It should also be as free as possible from objectionable or predictable subject matter, proper nouns, and borrowed words. The longer text is transcribed and translated phrase-by-phrase into a language the test administrator understands.
2. A group of 15 or more basic comprehension questions in various semantic domains are developed based on the text and recorded. A test tape is made containing the text, in segments, with the questions inserted just after the portion of the text containing the appropriate response. Thus, a subject's performance will not depend on his ability to remember the details of the text, but rather on his comprehension of the text.
3. Ten native speakers of Dialect A, preferably including both men and women, listen individually to the text and respond to the questions so that any badly composed or misleading questions can be isolated and removed. Ten questions to which native speakers have responded with correct answers are chosen for the final form of the test. These three steps are repeated at each test site to produce a test in each dialect.
4. Before administering the tests in Village B, the comprehension questions for all of the texts are recorded in Dialect B and inserted into their respective texts. Thus, Village B test subjects will hear the texts in other dialects, but the questions in their own dialect, ensuring that their ability to respond is not impeded by a lack of comprehension of the questions themselves.
5. At least 10 native speakers of each target dialect, preferably including both men and women, are tested. They are screened against contact with the other dialects in question and then listen individually to the shorter "learning text" and the longer text in their own dialect. If they are able to perform well, they then listen to the texts in the other dialects and respond to the questions as they listen. They listen to each text only once, though they are allowed to listen to a section a second time if there was some distraction. Their responses are written down and scored as "right", "wrong", or "half-right". Their scores are interpreted as an indicator of the level of the inherent intelligibility of the other dialects.

A given speech form will be inherently intelligible to some degree to speakers of a related speech form simply because both forms have sprung from the same linguistic stock. As a property of the language itself, the level of inherent intelligibility of a given dialect to speakers of a second dialect should be constant throughout the entire population of the latter. Thus, it is not necessary to obtain a random sample to measure inherent intelligibility. Any ten speakers who have mastered their own dialect will suffice. Certain individuals, however, will normally have had contact with speakers of other dialects, and thus, will have learned to understand them better. This "learned intelligibility" could be considered a type of bilingualism, a related but separate phenomenon which naturally varies within a population. Test subjects are screened to eliminate those which have had much contact with the other dialects in question.

3.5.2 Deviations in this study

Christiansen recorded and transcribed a Tadaksahak text in Menaka as described in Step 1 above; however, circumstances beyond his control prevented him from completing the hometown validation of comprehension questions (Step 3 above) before we used the Tadaksahak test in Niger. We compensated for this by having all test subjects attempt to answer all 16 comprehension questions which Christiansen devised for the Tadaksahak text. Then, after we were able to return to Menaka and play the test for native Tadaksahak speakers, we knew which questions to discount and what range of answers to consider acceptable in scoring the Tadaksahak test.

While playing the Tadaksahak test for native speakers in Menaka, one person commented that the person asking the questions spoke better Tadaksahak than the person recounting the story. This could throw a shadow of doubt on the validity of the Tadaksahak test; however, all ten Idaksahak who listened to the test were able to answer 10 of the questions without any trouble, so we believe that the test is valid.

In all groups, but especially among the Igdalen and Iberogan, we were met by village leaders who were cautious of outsiders. In Tamaya, we were denied permission to speak with Igdalen women. The chief pointed out that no women were even visible in the village. On the last day of the survey, however, we did get to interview a young black woman there. The chief of Tofabayogh informed us that all of the women of his village were afraid to speak with us. We got the impression that this hesitation was due to the women generally being guarded against contact with outsiders. Perhaps a female surveyor would be able to interview them.

After we had recorded the Tagdal text and questions in Tamaya, we were not able to seek out ten people to validate the test. Only three of the men who had come to the tree under which the chief hosted us were willing to listen to the test, and at least one of them had been present when the story was recorded. All three of these fair-skinned men answered nearly all of the questions correctly, but we could not consider that they formed a big enough or naive enough sample to validate the test questions. We had to continue validation of the Tagdal text questions with Tagdal-speaking subjects as we found them, in Abalak, all black. Some of these referred to themselves as Iberogan, although our translator tried to assure us that this was due to the common misconception among outsiders that all Igdalen are fair-skinned and all Iberogan are black. A number of our validators were from Abouraya, a village inhabited by both Igdalen and Iberogan. Thus, it is possible that questions which might have distinguished between Tagdal and Taberog were excluded from the test.

When we returned to Tamaya to administer the test, we had the good fortune of drawing the interest of five 15 to 16 year-old students who were on their lunch break and who became willing and capable test subjects. About half of these were black, and since the Igdalen tend to pay close attention to race, once again we cannot be sure our testing did not obscure any differences in intelligibility that may exist between the Tagdal- and Tabarog-speaking groups. We were able to develop a valid test in Tofabayogh, but every time we went to this village it rained in the evening. The Iberogan are subsistence farmers rather than nomadic pastoralists as we had expected, so, of course, the day following a good rainfall, every able-bodied person among them went out to his fields to plant millet. We are glad that after two years of below-average rainfall, Niger has received abundant rains this year, but it meant that people were only available to work with us after nightfall. We were also able to test some boys early in the morning before they were called away to other duties. Depending on the testing conditions and how well a subject catches onto the method, it can take from 40 to 60 minutes to administer the test in four dialects. After spending two nights in the village, we had completed testing with only seven subjects, but we judged this was as good as we were going to be able to do, and we didn't want to wear out our welcome.

4 Results

4.1 Dialect intelligibility testing

The Northern Songhay texts used in this study are shown in interlinearized form in Appendix B along with their related comprehension questions.

The results of our Northern Songhay dialect intelligibility testing are summarized in Table 1 below. The columns correspond to the village test sites, and the rows correspond to the texts in each speech form. Demographic information on all subjects and individual test scores are tabulated in Appendix C. Means, standard deviations, and sample sizes are tabulated in Appendix D.

In Ingal, the subjects averaged 94% on the Tasawaq text, showing that they understand their own form of speech very well, of course. However, they scored only 40% and 49% on the Tabarog and Tagdal texts respectively, indicating insufficient intelligibility of these speech forms to suggest they could easily use Tabarog or Tagdal literature. Their comprehension of the Tadaksahak text was even lower at 25%.

For the seven subjects we were able to test in Tofabayogh, the Wilcoxon t-test shows no significant difference between their scores on their hometown text (Tabarog) and the Tagdal text. The Iberogan and the Igdalen seem to live in such close contact that it may not be possible to find a significant number of Iberogan men who do not have regular contact with Igdalen men. We certainly weren't able to work with any during our brief visits to Tofabayogh. Two of our test subjects identified the Igdalen story-teller by name! This means that it was practically impossible for us to measure the inherent intelligibility of Tagdal to Iberogan men because they could be considered bilingual in Tagdal. Perhaps there are Iberogan women who do not have contact with Tagdal speakers, but no women were willing to talk with us. Our best attempt for the moment is to consider only the scores of our four youngest Iberogan subjects, who were between 10 and 16 years old. It may be that they have not had extensive contact with Igdalen simply because of their youth. Their mean score, however, is also 88%. Thus, Tagdal appears to be inherently intelligible to the Iberogan. ^Endnote The Iberogan subjects averaged only 73% on the Tasawaq test, which is too low to suggest that they could profit from development of the Tasawaq language. Their average score of 54% on the Tadaksahak test shows clearly that this language is not inherently intelligible to them.

In Tamaya, once again, there is no statistically significant difference between our nine subjects' scores on their hometown test (Tagdal) and the Tabarog test. It appears that Tagdal and Tabarog are mutually intelligible. ^Endnote The Igdalen appear to understand less Tasawaq (62%) than the Iberogan. Their score of 50% on the Tadaksahak test shows unequivocally that Tadaksahak is not inherently intelligible to them.

4.2 Adaptation potential

While we have not attempted to build adaptation tables showing the consistent phonological and morphological changes between any of these dialects, we did find a Tasawaq text transcribed and glossed by P. F. Lacroix (Bernus, pp. 109-114). Christiansen studied this text and estimated that by simply making regular sound changes and morphological substitutions, one would have a good Tadaksahak text. Although we did not try the reverse, we assume that a Tadaksahak text could easily be adapted to yield a good draft of a Tasawaq text.

Christiansen's observations while using Tadaksahak for limited communication with Tagdal-, Tabarog-, and Tasawaq-speakers, and test subjects' comments about the texts they listened to indicate that among the dialects studied, Tadaksahak has borrowed the most Tamajaq features and Tasawaq the least. Since Tagdal and Tabarog are in the middle of this spectrum, we assume that if Tadaksahak texts could be adapted into Tasawaq, they could also be adapted into Tagdal and Tabarog.

5 Conclusions

5.1 Tagdal

Intelligibility testing results show clearly that neither Tadaksahak nor Tasawaq are inherently intelligible to the Igdalen or the Iberogan. The isolation of their women makes it unlikely that they are highly bilingual in any other language. We observed that there were many monolingual Iberogan men. Both groups reported that Tamajaq was the most widely-spoken second language in their villages, but their attitude toward Tamajaq seemed to be one of acceptance by force of necessity rather than eagerness to use it. This rather negative attitude towards Tamajaq may have been accentuated by a dispute between the Tuaregs in the 7th Group and the Igdalen and Iberogan at the time of our survey (Jeff Woodke, personal communication). Both the Igdalen and the Iberogan are proud of their language and it appears that it will continue to be their primary language for the foreseeable future. Thus, it appears that the Igdalen and Iberogan could benefit from the development of a Tagdal literature in a way that development of another language would not afford them.

As noted in Section 4.2, there appears to be great potential for adaptation of literature from other Northern Songhay languages (Tadaksahak or Tasawaq) into Tagdal.

5.2 Tasawaq

Intelligibility testing results show clearly that neither Tadaksahak nor Tagdal is inherently intelligible to the Ingalkoyyu. However, since the economy of Ingal depends to a large extent on interacting with non-Tasawaq-speaking peoples, one has to assume that at least certain segments of the community must be bilingual at FSI Level 3 or higher in the other languages of the region (Hausa, Tamajaq, and Tagdal). Contrary to our test results, the Ingalkoyyu reported that they and the Igdalen/Iberogen understood each other very well. Regardless, since this has been the situation for several generations at least, this is clearly a case of stable bilingualism, where the use of other languages is not a threat to the vitality of the primary language.

The Ingalkoyyu have a very positive attitude towards Tasawaq. Their language is one of the main things which distinguishes them as a people and they are proud of it. Tasawaq seems likely to remain the primary language of the Ingalkoyyu as long as the town of Ingal survives, and the prospects for this seem good since the Ingalkoyyu appear to be a happy, healthy, and industrious people.

Endnotes

1. Several difficulties in preparing and administering the Tagdal and Tabarog tests (see Section 3.5.2) may have masked any difference in intelligibility between these two populations; however the Igdalen and Iberogan claim to have the same language, and we have not been able to show any lack of comprehension between them.

References

Bernus, Edmond et Suzanne (1972) Du sel et des dattes: introduction à l'étude de la communauté d'In Gall et de Tegidda-n-tesemt. Etudes Nigériennes n. 31. Niamey: Centre Nigérien de Recherches en Sciences Humaines.

Casad, Eugene H. (1974) Dialect intelligibility testing. SIL publications in linguistics and related fields, No. 38. Lacroix, P.F. (1968)

Nicolaï, Robert (1981a) Les dialectes du songhay. Contribution à l'étude des changements linguistiques. (étude phonologique). Paris: Selaf.

Nicolaï, Robert (1981b) Le songhay septentrional (études phonématiques): Extrait du Bulletin de l'Institut Fondamental d'Afrique Noire. Tome 41 (1979), série B, nos 2, 3, et 4. Dakar: IFAN.

Woodke, Jeff (8 May, 1998) Interview. Niamey.

Appendix A. 380 item word list from each site