ON THE ORIGIN OF SOME NORTHERN SONGHAY MIXED LANGUAGES. Abstract
This paper discusses the origins of linguistic elements in three Northern Songhay languages of Niger and Mali: Tadaksahak, Tagdal and Tasawaq. Northern Songhay languages combine elements from Berber languages, principally Tuareg forms, and from Songhay; the latter provides inflectional morphology and much of the basic vocabulary, while the former is the source of most of the rest of the vocabulary, especially less basic elements. Subsets of features of Northern Songhay languages are compared with those of several stable mixed languages and mixed-lexicon creoles, and in accounting for the origin of these languages the kind of language mixing found in Northern Songhay languages is compared with that found in the (Algonquian) Montagnais dialect of Betsiamites, Quebec. The study shows that Tagdal and the other Northern Songhay languages could be construed as mixed languages, although the proportion of Berber and Songhay elements varieties somewhat between these languages, and also indicates that the definition of ‘mixed language’ is labile because different mixed languages combine their components in different ways, so that different kinds of mixed languages need to be recognized. NS languages seem to belong to the category of Core-Periphery languages with respect to the origins of more versus less basic morphemes.
Tagdal, Tadaksahak, and Tasawaq are three Northern Songhay languages located in the Azawagh valley in modern-day republics of Niger and Mali. Today, all of them function as in-group vernaculars with stable structures.1 Tagdal and Tadaksahak are spoken by the Igdalen and Idaksahak, most likely descendants of semi-nomadic Tuareg-Berbers, located in both Niger and Mali, respectively. Tasawaq, spoken by a sedentary non-Berber people called the Isawaghan, is spoken in the villages of Ingal and Teggida-n-Tesumt in Niger. The following map, taken from Benítez-Torres (2009), demonstrates the approximate locations of Tagdal, Tadaksahak and Tasawaq.
[INSERT MAP 1 HERE]
Another sedentary Northern Songhay language, Korandje or Kwarandzyey,2 is located at Tabelbala in the far western region of Algeria (Cancel 1908; Champault 1969; Tilmatine 1991, 1996; Kossmann 2004, 2007 a, b; Souag 2009, 2010). According to Souag (2015), speakers of Korandje migrated from modern-day Mali around 1200 CE in order to practice modes of agriculture best suited to dry lands.
Yet another Northern Songhay language, Emghedeshie,3 was at one time the language of wider communication in the city of Agadez, in modern-day Niger, but has been extinct since the early twentieth century (Barth 1851, Hamani 1989: 208).4 Surviving data collected by the German explorer Heinrich Barth demonstrate that it was very similar to modern-day Tasawaq.5 The followig diagram, based on Nicolai (1979), shows the relationship of the languages from a genetic (or rather genealogical) standpoint.6
The feature that most distinguishes Northern Songhay languages from their mainstream Songhay cousins is their combining both Songhay and Berber features, with the nomadic varieties, Tagdal and Tadaksahak, having, relatively-speaking, more Berber features than either Tasawaq7 or even Kwarandzyey/Korandje in Algeria.8
The goal of this paper is, first, to demonstrate that Nicolai’s (1979) dichotomy of nomadic versus sedentary Northern Songhay languages bears out robustly both from a socio-cultural/historical,9 as well as from a grammatical, standpoint, though the Songhay material in these languages bespeaks a sedentary lifestyle (Souag 2012).Second, we wish to examine whether bilingualism could have played a role in the development of Northern Songhay languages,10 and to come to some tentative conclusions about whether the same process(es) might have been responsible for both the development of the nomadic and sedentary Northern Songhay branches.
Sections 1.1 and 1.2 provide background information. In section 2.1 we compare the inflectional morphologies of Tagdal, Tadaksahak and Tasawaq,11 focusing on subsets which demonstrate the Berber-Songhay mixture and the sedentary/nomadic distinction especially. First, we will describe the pronominal subsystems, the negation and TAM affixes, in that order. Second, in section 2.2 we look at their derivational subsystems of the nomadic varieties, Tagdal and Tadaksahak, along with some syntactic structures with similar functions to those of the derivational affixes, and compare them with the derivational structures of sedentary Tasawaq. Throughout, we will contrast Northern Songhay structures with both Tayart and Tǝwǝllǝmmǝt, the Tuareg-Berber languages with which modern-day speakers of Northern Songhay languages are most in contact; and with Zarma and Songhay from Gao and Timbuktu, the only remaining vehicular Songhay languages in Niger and Mali. By the end of section 2, it will be clear, first, that all Northern Songhay languages have similar inflectional subsystems of Songhay origin. Second, it will be evident that Tagdal and Tadaksahak, the nomadic varieties, have default Berber derivational subsystems, while the sedentary Tasawaq’s default derivational subsystem is of Songhay origin.
In section 3 we demonstrate if and how Northern Songhay languages fit into a typology of mixed languages, and also observe that not all forms of Berber origin in these languages must or can be from Tuareg varieties. Finally, in section 4 we will briefly review the data, then come to some tentative conclusions concerning the linguistic processes which could have led to the development of the nomadic and the sedentary varieties. We conclude with some theoretical considerations for future research.
Most of the data on Northern Songhay languages in this article were collected by one of the writers, who lived in Niger on and off from 1999 to 2012. These data came primarily from recordings by mother-tongue Tagdal speakers, including some 42 texts of varying lengths and genres, including narrative, hortatory (e.g. sermons), instructional and 12 folk tales. Unless otherwise noted, data from Tadaksahak came from Christiansen-Bolli (2010), though some were elicited from mother-tongue Tadaksahak speakers residing in Niamey, Niger. Unless otherwise explicitly stated, all of the Tasawaq data came from various places in Alidou (1988), though some were elicited from Tasawaq speakers residing in Niamey, Niger. All data from Tǝwǝllǝmmǝt and Zarma were elicited from mother-tongue speakers in Niamey, Niger.12 All data from Tayart came from Kossmann (2011), and all data from Gao Songhay came from Heath 1999b.
Overview, sociolinguistic history of the sub-region
Due to the scarcity of accessible written documents13 from the region of northern Niger during the late middle ages, it is difficult to know with absolute certainty what might have led to the origin of Northern Songhay languages, probably some 300-600 years ago. Therefore, any conclusions would be based on circumstantial evidence. Nevertheless, we can make some educated guesses. Most of the material in this section is taken from Bernus (1972), Adamou (1979) and Hamani (1989).14
The ancestors of modern-day Tagdal and Tadaksahak speakers were North African Berbers.15Today, their descendants participate in the general semi-nomadic Tuareg-Berber milieu (Lacroix 1968: 93), mostly as pastoralists and religious experts for the Tuareg-Berber speakers around them.
Tasawaq speakers, on the other hand, are primarily of sub-saharan descent,16 leading mostly sedentary lifestyles in the villages of Ingal and Tegidda-n-Tesumt in northern Niger. Bernus (1972: 15-17) suggests that the ancestors of modern-day Tasawaq speakers, at least among their aristocracy, are descendants of Berbers who migrated to the region around the same time as the ancestors of modern-day Igdalen and Idaksahak, and were subsequently absorbed into the local population. Adamou (1979: 25) suggests instead that they are descended from a Songhay-speaking colony established in the oasis near Ingal sometime during the 1500s-1600s CE, the period that the Songhay Empire ruled the area. Regardless of the origins of Tasawaq speakers, today there are clear ethnic and cultural differences between sedentary Tasawaq speakers on the one hand and nomadic Tagdal and Tadaksahak speakers on the other.
The following can be more or less confirmed historically: 1) the ancestors of the Igdalen and Idaksahak most likely arrived in modern-day Mali and Niger sometime between the 7th and 9th centuries CE, as part of the first recorded Berber migration into the region.17 2) Most other Tuareg groups arrived in modern-day Niger and Mali several hundred years later, sometime around the 12th and 17th centuries CE. 3) The Songhay Empire annexed the area in the late 16th century CE. This brings up the question of what exactly was/were the lingua francas in the region of modern-day northern Niger before the 16h century. Adamou (1979) and Hamani (1989) suggest a variety of Hausa. Then again, both Hamani and Adamou wrote their historical treatises in order to refute Tuareg-Berber claims that Sub-Saharans in modern-day northern Niger are only recent arrivals. Regardless of the contrasting viewpoints, these questions are beyond the scope of this study. What seems to be relatively clear, however, is that Songhay remained the lingua franca of the region until the early 20th century, and still remains so in parts of modern-day Mali. 4) After the fall of the Songhay Empire, the region experienced almost constant warfare until the arrival of French colonisers in the late 19th century CE.
Given the history of the region, Benítez-Torres (2009) suggests that the ancestors of the Igdalen, bilingual in both the vehicular Songhayof the time and their vernacular Berber, for whatever reason became motivated to differentiate themselves from the wider Tuareg population,18 as well as from the more urban mainstream Songhay-speaking populations. In order to do this, they turned to their already-existing speech patterns as a means of highlighting these differences. In other words, if the above scenario were the case, unlike those mixed languages that arise in situations of intended language maintenance,19 Tagdal, and by extension Tadaksahak, would have arisen in a situation of intended language shift (Benítez-Torres 2009), as a means of underlining the identity of a new group or sub-group.20
The degree of mixture in NS languages can be seen from an examination of the languages’ segmental phonology. One such consideration is the presence of lexical tone in Tasawaq, absent in other NS languages. In addition, words from each major source seem largely to preserve their phonological shapes and the sounds and also the syllabic canons which they contain. As a result, the phonemes /e o/ occur much more frequently in vocabulary of Songhay than of Berber origin, while relatively speaking words from Berber languages can contain more instances of pharyngeal consonants, schwa and complex consonantal clusters.
Previous literature on NS languages
There is a small amount of literature on Tadaksahak (most notably Christiansen and Christiansen 2002, Christiansen-Bolli 2010, and Heath 2004) and on Tasawaq, Kossmann (2007a,b, 2012) and Wolff and Alidou (2001) being the most readily accessible materials; note also the Master’s thesis by Alidou (1988), but there is rather less on Tagdal. Rueck and Christiansen (1999) provides a 360-item wordlist containing most of the items on the longer and shorter Swadesh lists along with various texts in several NS varieties translated into French, including Tsadaksahak, Tasawaq, Tagdal and Tabarog. Benítez-Torres (2009), looking at what historical and sociolinguistic data exists, suggested some possible scenarios for the genesis of Tagdal. There is also quite a large amount of unpublished material21 on these languages from the 1960s onwards, which is not accessible to the general public; the authors have not seen all of this.
A note is needed here concerning Nicolaï’s research. Robert Nicolaï has dedicated much of his work to exploring the origins of Songhay languages. He has suggested a number of possible scenarios. (c.f. Nicolaï 1990a, 1990b; 2003; 2006a, 2006b; Nicolaï ms.) First, an extinct, Koinéized, Afro-Asiatic language could have served as the lingua franca in the region of modern-day Niger and Mali. After lexifying surrounding languages and giving rise to Songhay, this language subsequently disappeared. A second possible scenario, the ancient lingua franca of the region was an ancestor of modern-day of Songhay, in which case modern varieties are a vernacularisation of this ancient Proto-Songhay – for lack of a better term. Most recently (2009), Nicolaï suggested that modern-day Songhay developed through contact with Mandé, with an extinct Afro-Asiatic language providing much of its lexicon. Whatever one thinks of these theories (see Dimmendaal 1992, 1995, Benítez-Torres 2005, Kossmann 2005), no one disputes that Nicolaï has been influential in the area of Songhay and genetic linguistics. Nevertheless, it would seem that Songhay may not have even been the language of wider communication in modern-day northern Niger until after the 1500s, when the Songhay Empire spread eastward and established colonies there (Adamou 1979: 26). Therefore, unless one can demonstrate convincingly that the Songhay language was a major factor in the area before this, questions about the development of Proto-Songhay are beyond the scope of this paper.
A comparison of several typologically salient grammatical structures and a note on basic lexicon.
All Northern Songhay languages have relatively complex morphological systems, most of whose elements can be traced to either Berber or Songhay origin. As we will see below, all of the languages in question have inflectional sub-systems of Songhay origin, while some of those in the nomadic varieties Tagdal and Tadaksahak have derivational sub-systems of Berber origin. Emphasis here is on subsystems which differ between nomadic Tagdal and Tadaksahak on the one hand and sedentary Tasawaq on the other, though other topics are also covered.
Aspects of inflectional morphology
In this section, we describe the inflectional clitics and prefixes of Tagdal, Tadaksahak and Tasawaq22 – the pronominal systems, negation and Tense-Aspect-Mode, in that order. Since all of these bound morphemes in all three languages are of Songhay origin, almost any sentence in any of the languages will have some very key Songhay elements included.
We begin with the independent pronouns in Table 1. Though not, per-se, part of the inflectional morphology of Northern Songhay languages, they help demonstrate some of the similarities and differences between the three languages in question. We also include for comparison the pronouns of Songhay from Gao (Stauffer 1997) and Tayart Tuareg-Berber (Kossmann 2011).
As Table 2 below shows, the subject markers in each Northern Songhay language are bound morphemes, which attach onto the verb root.24
[INSERT TABLE 2 HERE]
Table 2: Subject clitics in NS languages of Niger and Mali
Tagdal, Tadaksahak and Tasawaq have both a completive and an incompletive negation. In Nothern Songhay languages, these are bound to the verb root as prefixes. The prefix nǝ-, or ni- in Tasawaq, the default choice for negation in spoken discourse, functions as a completive negation, indicating something which did not occur in the past, or in the case of stative verbs, something which is not true or is not presently the case. Data in Table 3 indicate this.