The Savanna Untangled: A Critical Look into the Validity of the Nilo-Saharan Language Family
Advisor: Elizabeth Pittman
Internship Location: BASIS Ind. McLean
Onsite Mentor: Elizabeth Pittman
The Nilo-Saharan language family is a proposed African language family stretching from Algeria to Tanzania, encompassing over 200 languages and 70 million speakers. It was first proposed in 1963 by Joseph Greenberg in The Languages of Africa. Within this book, Greenberg sorted all the languages of continental Africa into four major language families: Afro-Asiatic, Niger-Congo, Khoisan, and Chari-Nile. Additionally, he identified six minor language groups and isolates, one of which, Kordofanian, was combined with Niger-Congo, while the other five, Songhai, Saharan, Maban, Coman, and Fur, were all combined with Chari-Nile to form the Nilo-Saharan family. In the past sixty years, this initial proposal has changed shape, with many of the family’s interior groupings being challenged, while the family at large continues to be used by linguists today. However, it is not a family without controversy, as Greenberg’s tendency to group together disparate languages from across the continent led to it being labeled a “wastebasket phylum” by Campbell and Poser, in which Greenberg could toss in all the “leftover languages” not accounted for within Africa’s more established families. The goal of my project is then to analyze and synthesize 60 years of academic inquiry into a comprehensive and defendable theory of the Nilo-Saharan languages and their relations, addressing both the groupings within the family and the validity of the family as a unified whole, all while utilizing linguistic techniques such as mass comparison and lexicostatistics, which use basic vocabulary to determine potential relationships between languages via the presence of cognates. By postulating an evidence-based thesis on the nature of the Nilo-Saharan languages, I hope to offer insight into the often overlooked field of African linguistics, a fundamental filament of the world’s linguistic fabric.
Project Blog
Week 1: The Genesis of the Theory
Where does one start with Nilo-Saharan? The natural answer is Greenberg. It was Joseph Greenberg who first proposed the concept of the family in 1963 and it is his initial postulations which the bulk of modern Nilo-Saharan studies rest atop. However, it is impossible not to recognize that Greenberg was a controversial figure in linguistics. He actively supported the idea of large, sweeping language families like Indo-Pacific and Eurasiatic, neither of which were ever seriously attested for by other linguists. Greenberg justified these decisions through his process of “mass comparison.” Mass comparison, also known as multilateral comparison, is a method of determining the relatedness of languages which importantly does not attempt to posit a proto-language which the modern languages could’ve theoretically evolved and diverged from. Rather, mass comparison starts with a pool of many languages’ vocabularies and attempts to draw comparison through the presence of similar-sounding words. There are multiple problems which can arise through this methodology, for example, similarities in vocabularies are often attributable to the proximity of one language to another, and do not inherently imply that they ultimately derived from the same language. German is an Indo-European language while Hungarian is a Uralic language, yet their words for math, ‘Mathematik’ and ‘matematika,’ sound similar because both are derived from the Ancient Greek ‘mathēmatikē.’ It is also possible for similarities in words to occur purely out of chance, such as the Portuguese ‘obrigado’ and Japanese ‘arigato,’ both meaning ‘thank you.’
One might naturally wonder why Greenberg’s Nilo-Saharan is still used by linguists today if Greenberg’s method were so unscrupulous. The answer to that question lies in the fact that while Greenberg’s overall body of literature is tinged with large assumptions, his work on African linguistics is actually quite respected. His book The Languages of Africa managed to completely alter the field in ways still respected by modern linguists. Greenberg successfully identified the Bantu languages as a part of the larger Atlantic-Congo family–as opposed to being an independent family–and also successfully refuted the eugenics-based theory of the ‘Hamitic’ languages, mapping them within the larger Afro-Asiatic family. His work on Africa was not completely accepted however, as the Khoisan family is generally accepted in the modern day as three distinct families with significant intermingling, still used today by linguists but as a helpful category and not a true family. This leaves Nilo-Saharan in a complicated position, as it is accepted by some linguists and rejected by others. There are some notable phonetic similarities in pronouns across the proposed family, and the prevalence of the otherwise rare singulative-collective-plurative number system–a trait in which a language has three different words for a singular noun, a plural noun, and the concept of said noun–lead some modern linguists to continue pursuing the theory. At the same time, other linguists believe that Greenberg was cherry-picking his words, as he didn't even use the Swadesh list–a standardized list of 207 concepts for which there is a word in almost every spoken language–to assemble his comparative vocabularies.
Overall, a consensus is yet to be met on the status of the theory, though I hope that my research can bring us the slightest bit closer to that goal.
Sources:
Greenberg, Joseph H. "The languages of Africa." International journal of American linguistics (1963).
Dimmendaal, Gerrit J. "On stable and unstable features in Nilo-Saharan." Nilo-Saharan issues and perspectives (2018): 9-24.
Week 2: The Linguistic Tapestry
In order to understand the Nilo-Saharan proposal in full, one first must peer into the details. In this blog post, I’ve compiled a list of all the subfamilies generally attributed to the Nilo-Saharan Phylum, a list which has changed quite a bit since Greenberg’s initial proposal, expanding outwards from his initial six subfamilies.
Eastern Sudanic
Eastern Sudanic has the oldest documentation of any Nilo-Saharan subfamily, with Old Nubian being recorded in the Coptic alphabet as early as the 700s. Stretching from Egypt down to Tanzania, it is also the most widely spoken group within the Nilo-Saharan family, with the Nilotic branch within Eastern Sudanic making up the better half of that number. Though, being the largest family in the group, it is also the most controversial, with some linguists not only refusing to recognize Nilo-Saharan as unified, but also refusing to recognize the Eastern Sudanic subfamily. It is a family with a significant amount of diversity, though some of that can be attributed to admixture, with most of the differences following a geographic north-south dichotomy, creating a sort of spectrum of change as one travels further in either direction.
Central Sudanic
One of the largest branches within the Nilo-Saharan label, stretching from Uganda to Cameroon. Generally, modifiers follow nouns within these languages, except for modifiers which indicate possession, which instead precede the noun. Many Central Sudanic languages have the quite unique feature of advanced-tongue-root vowel harmony. Vowel harmony is a feature of many languages where a vowel is altered to maintain the flow of a sentence, sort of like how ‘a’ can become ‘an’ in English but with a vowel changing as opposed to a consonant. The advanced-tongue-root refers to a process in which one pushes forward the base of their tongue, expanding their pharyngeal cavity, in order to change the sound of a vowel.
Kunama
A small family in inland Eritrea, oftentimes grouped together with its larger neighbors in Eastern Sudanic, especially due to its tendency to form coverb-plus-light-verb constructions, though the fact that this isn't a feature present in the southern languages of Eastern Sudanic indicate that it may be an areal feature simply borrowed by Kunama and others.
Gumuz and Koman
Two language families that have often been grouped together, but are also often separated. Both are small and found on the border region between Sudan and Ethiopia. Gumuz is notable for its particularly large inventory of consonants, with the lowest number of consonants being 31. Koman has more variation in their consonant numbers, though the proposed proto-language only has 26 total. Both families are structured in “Subject-Verb-Object” order, and both families have the feature of deictic markers attached onto verbs that can indicate direction.
Kadu
A particularly disputed family in Sudan which is often grouped as a part of the Kordofanian family, which is further, tenuously grouped into the Niger-Congo family.
Kreish and Sinyar
Two small language isolates located in the general area of the Central Sudanic branch which are poorly researched, but often grouped into the Nilo-Saharan phylum. More research is necessary to confirm their status
Shabo
Another language isolate, spoken by only about 400 people in Ethiopia, occasionally grouped with Nilo-Saharan but oftentimes kept separate due to a lack of research.
Fur
Consists of two languages spoken, in Western Sudan and Eastern Chad, historically spread and developed by the sultanate of Darfur. The Fur languages have the unique feature of incredibly common metathesis, meaning the changing of a word’s internal phonemic structure by switching the placement of consonants. This often happens on a long-term scale, like the Latin ‘parabola’ becoming the Spanish ‘palabra,’ but in Fur these changes can happen within a sentence due to grammar alone.
Maban
Maban acts as quite an important pillar in the sustainment of the Nilo-Saharan theory, as it shows a significant lexical similarity to Eastern Sudanic, as well as to the more tenuous branches of Fur, Saharan, and even Songhay. Maban is a rather western family, existing primarily in Chad, which does leave some to believe that the lexical similarities it holds with languages further west are simply due to its location.
Songhay
The Songhay languages maintain a similar base vocabulary throughout the group. They have been significantly influenced by Afro-Asiatic (Berber, Arabic, Hausa) and Niger-Congo (Gur, specifically Bariba). Songhay languages were used by the Songhay Empire, and were able to spread as a result. It also hosts many similarities with the Mande family–a group typically included in Niger-Congo but occasionally left independent–including an almost identical phonology in the proto-language, as well as notable similarities with the Saharan family. Songhay is often considered to be one of the most tenuous connections to Nilo-Saharan.
Saharan
A group of Languages spoken in Chad, Niger, and Nigeria whose largest language, Kanuri, has over nine million speakers, and was partially spread by the historic Kanem-Bornu empire and its successors. While it is often grouped into the Nilo-Saharan family, its internal connections are quite inconsistent, with Blench stating that it forms a subfamily with Songhay and Dimmendal stating that it forms a subfamily with Eastern Sudanic.
Kuliak
Kuliak appears to be a relatively isolated group, but one which has also borrowed up to 70% of its lexicon from neighboring Nilotic languages. It generally is not considered Nilotic, however, as the remainder of its vocabulary is quite unique, though the grammatical structures of the language are in line enough with the family for it to be considered Nilo-Saharan, though that too may have arisen from the Nilotic admixture.
Mimi of Decorse and Mimi of Nachtigal
Two distinct languages attested to exist in Chad with the name Mimi during the years 1900 and 1870 respectively. Neither language appears to exist in the modern day and all that was ever attested of them were simple lists of vocabulary. It is possible that the languages died out, that they are still spoken but unattainable by researchers, that they were improperly recorded, or that they were simply a hoax, as faking languages was a not uncommon practice in that era. Regardless, Greenberg still incorporated these languages into the Nilo-Saharan phylum. It should also be noted that Amdang, a Fur language, is occasionally referred to by the name Mimi, but does not show any similarity to the vocabulary lists provided by Decorse and Nachtigal.
Sources:
Greenberg, Joseph H. "The languages of Africa." International journal of American linguistics (1963).
Dimmendaal, Gerrit J., et al. "Linguistic features and typologies in languages commonly referred to as ‘Nilo-Saharan’." Cambridge Handbook of African Languages (2019): 326-381.
Week 3: The Starostin Perspective
The world of linguistic taxonomics can generally be divided into two opposing camps: the ‘lumpers’ and the ‘splitters.’ Lumpers generally try to propose large, expansive language families while splitters try to argue for more granular classifications. Within this warzone of academia, however, there exists a man who toes the line between these conflicting factions, a man by the name of Georgiy Starostin. Starostin’s father, Sergei Starostin, was also a linguist, one who firmly planted his feet among the ranks of lumpers. He was one of the primary proponents of the Altaic hypothesis, the concept that the Turkic, Mongolic, and Tungusic language families were all part of a larger, Altaic family. This is a theory widely considered fringe by modern linguists, though it remained quite popular for quite a long time. Georgiy is not unlike his father in his tendency to lump things together. He himself is a supporter of the Proto-World language hypothesis, the idea that since all modern humans descended from a common ancestor, that naturally all spoken languages must ultimately be derived from one Proto-World language. Georgiy does not, however, believe that Proto-World can be proven within our present research capabilities. The idea is speculative, adored by many lumpers for its global connectivity but not in any way provable even by those operating on the fringe.
It would naturally seem to many that, being the lumper that he is, Georgiy Starostin would be attracted to the Nilo-Saharan theory, a concept developed by a fellow lumper which has since rocketed its way into the mainstream. Though, the interesting thing about Georgiy Starostin, is that while he is a lumper on the large scale, he can be quite the splitter in the realm of the minute. He may support the concept of a Proto-World language, but he would also like to ensure that one is constructing it correctly.
In order to analyze the validity of the Nilo-Saharan family, Starostin utilized a process known as ‘lexicostatistics.’ This process relies on a linguistic tool known as the ‘Swadesh List,’ a list of 207 words composed by Morris Swadesh which are common enough to have an equivalent synonym in effectively every spoken language. Words in the Swadesh list tend to be some of the most commonly spoken words in a language, including simple nouns, such as ‘eye,’ simple verbs, such as ‘die,’ words used in questions, such as ‘what’ and ‘who,’ as well as proximal words, like ‘this’ and ‘that.’ Since all of these words are very commonly spoken, it is expected that these words will not have arrived into a language as a result of borrowing from another language, since linguistic borrowing often occurs when one language lacks a preexisting word for a given concept, but since all of the words in the Swadesh list are very simple, there isn't much of a reason for borrowing. This makes the Swadesh list very useful for comparing languages, because if two languages show similarities in vocabulary on their equivalent Swadesh lists, it means that both languages likely derived from an older language. Occasionally, a language might have many similarities to another in its general vocabulary but show effectively zero similarities within the Swadesh list, like how Hungarian has many Latin influences but has a base vocabulary which is entirely different from Latin’s. This difference in base vocabulary proves that the two languages do not share a language family, as Hungarian is a Uralic language which has been historically influenced by Latin due to the spread of Catholicism.
Lexicostatistics uses the Swadesh list to compare large swaths of possibly related languages. For example, if a linguist notices that a significant number of the words in the Swadesh list of Language A have a similar phonemic structure to their equivalent words in the Swadesh list of Language B, they might come to the conclusion that both languages evolved from a Proto-Language AB, and that the phonetically similar words are ultimately cognates derived from the vocabulary of Proto-Language AB. Further, if Language A and Language B lack a significant number of phonetically similar Swadesh words, they may still be related if both languages have phonetically similar Swadesh words in Language C. Oftentimes, different languages will change at different rates, so while Language A and B may have diverged quite significantly over time and no longer show their similarities with one another, they may still show similarities with the more conservative counterpart of Language C, which still bears many features of the Proto-Language ABC which all three languages derived from. Generally, linguists will avoid comparing just two languages with lexicostatistics, preferring to analyze a larger number, as occasionally there will be specific languages, such as the theoretical Language C, which can be very useful in connecting otherwise disparate languages.
In his lexicostatistical analysis, Georgiy Starostin approaches the Nilo-Saharan phylum with a multi-pronged methodology. First, partly as a proof of concept, he used lexicostatistics to back the already well-supported individual subfamilies of Nilo-Saharan. By doing so, he could prove the validity of his strategy, which has also in the past been used to connect the disparate families of Indo-European. He then lays out comparative evidence for the validity of four specific families: Saharan, Koman-Gumuz, which he chooses to classify as a singular group though one with many differences, Central Sudanic, which he groups together with the minor Kreish and Sinyar families, and Eastern Sudanic, which he argues is a united family.
However, Starostin has some difficulty in connecting these various families to one another through Swadesh-backed lexicostatistical methods. After developing the connected vocabularies of these families into proto-vocabularies, he finds effectively zero solidly potential cognates between Saharan and Koman-Gumuz with Eastern and Central Sudanic and only finds a small and tenuous connection between some of the words in Eastern Sudanic and Central Sudanic. The words for ‘ashes,’ ‘name,’ ‘two,’ and ‘who,’ all share phonetic similarities across the two families, but Starostin suspects that any relationship between the two families would require them to diverge at least 12,000 years ago. Otherwise, he asserts that Saharan and Koman-Gumuz should be separated from Nilo-Saharan under the present state of research.
Finally, Starostin tried to see if he could use lexicostatistics to group any of Nilo-Saharan’s smaller families into the four families he established. He suspected that Fur, Kadu, and Berta, which is often considered an isolate, could fit into the Eastern Sudanic family, while also suspecting that Maban and Kunama could hold a relation, and that the family they hold together could be grouped into Central Sudanic. However, Songhay, Kuliak, and little Shabo were unable to be sorted alongside any of the preexisting families within Starostin’s model.
Ultimately, Starostin proposes a more conservative model of Nilo-Saharan, which at its maximum, contains the established families of East Sudanic, Central Sudanic, Fur, Kadu, Berta, Kreish, Sinyar, Kunama, and Maban, while excluding the families of Koman, Gumuz, Saharan, Songhay, Kuliak, and Shabo.
Sources:
Starostin, George. "The Nilo-Saharan hypothesis tested through lexicostatistics: Current state of affairs." (2016).
Week 4: Linguistics and Prehistory
The field of linguistics is intrinsically linked to the field of history, but history can only reach so far. Eventually, any given researcher will run out of sources directly from the time period, and depending on one’s location of interest, those sources may be quite slim. In order to pursue a linguistic link further, one then must look beyond the surviving written records and dive into the often messy world of linguistic prehistory, a field characterized by the study of remnants, both physical remnants in the form of archaeology, and spoken remnants, in the substrates of our every word.
One linguist with a passion for the study of linguistic prehistory is a man by the name of Roger Blench, a British researcher who falls on the latter end of the splitter-lumper continuum. He is a strong supporter of the Nilo-Saharan theory and believes that it can be backed up through the study of the Sahara desert’s prehistory.
The African Humid Period was a geological phenomenon in the Earth’s prehistory which lasted from about 12500 to 3500 BCE. It’s been studied by scientists for over a century now and is strongly backed by both geological remains and anthropological remains in the form of cave art depicting savanna animals in a land which is now uninhabitable for them. During this period, the vast Sahara desert was not the wasteland pockmarked by oases which we see today, but rather a grassland teeming with life, much like the modern Serengeti. It was during this period, Blench claims, that the Proto-Nilo-Saharans spread out from their homeland and populated the Green Sahara, creating the Songhay, Saharan, and Maban families as they did so. The difficulty with proving this claim, however, is that history often blurs its predecessors.
In the modern era, if one excludes tertiary European languages spread due to colonization and globalism, the linguistic landscape of the Sahara can generally be split into two uneven halves, both of which are derivative of the Afro-Asiatic family. The first and larger half is of the Semitic branch of Afro-Asiatic, though that label is a little deceitful as effectively all of those speakers abide by a single language, Arabic, as well as its many–often divergent–dialects. The spread of Arabic, however, is quite recent and very recordable, as the language only entered and dominated the desert during the rise of Islam in the medieval era. Before Arabic, most of the Sahara’s speakers instead abided by the second, smaller half of languages, the Berber branch of Afro-Asiatic, which includes languages like Tuareg and Rif. While these two branches account for the majority of the desert’s modern lects, there are a couple outliers detectable in the historical record. Semitic had actually been present in North Africa before the rise of Islam in the form of various Phoenician colonies, as well as the Carthaginian empire which a group of them went on to form. Additionally, while modern Egypt speaks a dialect of Arabic, Ancient Egypt communicated in the Coptic languages, a now extinct branch of Afro-Asiatic which is still nominally preserved today as the liturgical language of the Egyptian Coptic Church. Finally, both Greek and Roman civilizations left their mark on North Africa, leaving a handful of dialects behind amid the coasts, though all of those languages were quickly swept up amid the Arab invasions.
It is important to note, however, that by splitting the Sahara into two groups of Afro-Asiatic origin, we ignore the thesis of Roger Blench, that before the Arab expansions, and that before even the Berber expansions, there existed a Sahara dominated by Nilo-Saharans. We see remnants of this era today, as while most Nilo-Saharan languages are located either on the Nile, south of the Sahara, or on the southern edge of the Sahara, the Korandje language is spoken in the northern half of Algeria, just south of the Atlas mountains, by about 3,000 people in the village of Tabelbala. The location of Korandje could very well be interpreted as a sign that, before Afro-Asiatic peoples migrated across the region, the Maghreb could’ve very well been host to a wider Songhay-speaking population, one which shrunk over time and now predominantly exists in the Sahel, but one which left remnants of its former expansiveness in the form of Korandje. However, the location of Korandje alone does not necessarily prove the thesis that Songhay stretched into northern Algeria prior to the Berber migrations. The presence of Korandje could just as easily be attributed to a more modern migration, as the village of Tabelbala had been an important Saharan oasis town during the reign of the Songhay empire, and while the empire itself never stretched so far as to subsume Tabelbala, it is entirely within reason that Songhay speakers originating from the empire could have migrated into the town.
Additionally, the study of remnants can be seen on a deeper level than entire languages if one chooses to peer into specific, individual words. While Greenberg’s methods of mass comparison often resulted in the establishment of false cognates, the Swadesh list isn’t perfect at establishing comparison either. By attempting to create a standardized list of the most important words in every language, one may ignore the cultural significance placed upon specific words. Therefore, one should not just compare any random word, but one also shouldn’t rely exclusively on a preset list. Roger Blench noted that a set of specific animals often appeared in the cave art attributed to the era of the Green Sahara, animals like crocodiles and hippos which can no longer survive in the deserted region. Blench figured that the ethnobotanical terminology for these animals could be useful in determining how the Proto-Nilo-Saharans may have migrated across the region.
It turned out that that terminology was very useful. While the various words for crocodile diverged quite noticeably, these differences could be explained by natural linguistic processes. Within Blench’s proposal, the root word for crocodile was likely something like ‘karam,’ which through end-dropping became ‘kaaray,’ through k-dropping became ‘arem,’ and through metathesis became ‘mara.’ The word for hippo seems even more supportive of the idea, with the Kokit language and Koyra Chiini language both using the term ‘banga’ despite being in different subfamilies and on opposite sides of the continent. Further, multiple southern Berber languages use a set of ethnobotanical terms which are otherwise unattested for in any other language of the region, potentially suggesting the presence of a substrate language which was subsumed into these lects during the Berber migrations, potentially one of Nilo-Saharan origin which has since been lost to time.
Sources:
Blench, Roger. "The linguistic prehistory of the Sahara." Burials, migration and identity in the ancient Sahara and beyond (2019): 431-463.
Week 5: Linguistic Comparison, Zarma & NKo
This week, I decided to expand the horizons of my linguistic research. So far, I’ve primarily been analyzing the work of others, but now I’d like to display some comparative linguistics of my own.
As previously mentioned, the Songhay language family–which is typically grouped with the Nilo-Saharan phylum–has a lot of phonetic similarities with the Mande family–which is typically grouped with Niger-Congo. However, both of these groupings are quite disputed, which leads me to wonder if these families could actually be related to each other, as opposed to their precarious positions in the larger families to which they’ve been ascribed.
In order for me to start answering the question of Mande and Songhay’s possible connection, I first had to establish a pair of basic vocabularies together. Ideally, I wanted to line up the Swadesh lists for multiple languages in both families, however–as I’ve stated before–African languages are under-researched, and as a result, I couldn't find any Swadesh lists for either family, meaning I’d have to assemble them myself. At this point, I decided to limit myself to comparing only two languages from either family, as it would prove very difficult to find online dictionaries for some of the smaller languages in these families, and of the ones that did exist many were only available in French–a language I cannot read.
Establishing a Swadesh list for my Mande language was easier than for my Songhay language, as there’s actually a machine translation available for the NKo language, a member of the Mande family, provided by Google Translate. NKo is a standardized dialect of Manding, used as an official language in Mali, which is importantly written in a non-Latin script. I cannot personally read this script, so after running the Swadesh list through Google Translate–a service which is often quite maligned at translating sentences but is very effective when it comes to translating individual words–I received a long list of symbols that I couldn't ascribe sounds to. While I began to manually transliterate the words into Latin, the Swadesh list is 217 items long, so I had to find an online transliterator to finish the process. I was able to find such a service, but unfortunately, NKo is a tonal language, and none of the services I could find added tone marks in their transliteration. However, because the language I’d be comparing NKo to was not a tonal language, I figured the presence of tone wouldn't really matter in the comparison, as tone would have to be a feature added after the two lects theoretically broke off from one another.
Establishing a Swadesh list for Songhay was a bit tougher. I chose to base my list after the largest of the Swadesh languages, Zarma, but unlike NKo, there was no machine translation for the language. There did exist a couple of online dictionaries for the language, but most of them were either in French or purely translated Zarma to English, and not the other way around. The only English to Zarma dictionary I could find was hosted on an old website published by the U.S. Peace Corps, so old that it still spelled ‘muslim’ as ‘moslem.’ Unfortunately, this dictionary was not quite complete enough for me to fill out the entire Swadesh list, though I was able to get most of the words filled in.
After lining up both translations of the Swadesh lists, I began to compare the Zarma and NKo vocabularies. Going into this process, I understood that any connection between the Songhay and Mande families would require a point of divergence very far in the past, meaning that many of the links between words would have to be rather tenuous.
One of the Swadesh words that had a particularly strong connection between the two languages was the word for ‘thick,’ which in Zarma was either ‘kom’ or ‘kassu’ and in NKo was ‘kamisɛ,’ a word with similar phonetics to both of the Zarma words. Additionally, one of the strongest connections was for the words ‘foot’ and ‘leg.’ In both languages, the concepts of feet and legs share a singular word, which in Zarma is ‘ce,’ and in NKo is the similar-sounding ‘se.’ The last of the particularly strong connections I found would be the word for ‘to suck,’ which in Zarma was ‘sunsum’ and in NKo was ‘soso.’ I also found a handful of weaker connections which I still feel like are worth mentioning as they follow general linguistic trends such as ‘l’ sounds shifting over time into ‘r’ sounds. These examples include the word for ‘to hear,’ which is ‘ma’ in Zarma and ‘mɛ’ in NKo, the word for ‘to live,’ which is ‘goro’ in Zarma and ‘balo’ in NKo, the word for ‘to spit,’ which is ‘tufa’ in Zarma and ‘ti’ in NKo, the word for ‘to blow,’ which is ‘funsu’ in Zarma and ‘fɛ’ in NKo, the word for ‘nose,’ which is ‘niine’ in Zarma and ‘nu’ in NKo, the word for ‘blood,’ which is ‘kuri’ in Zarma and ‘joli’ in NKo, the word for ‘fruit,’ which is ‘tuurize’ in Zarma and ‘yride’ in NKo, the word for ‘tree,’ which is ‘turi’ in Zarma and ‘yiri’ in NKo, the word for ‘small,’ which is ‘kayna’ in Zarma and ‘kadgbɔ’ in NKo, and the word for ‘you,’ which is ‘arang’ in Zarma and ‘alu bɛ’ in NKo.
In general, the similarities I found between Zarma and NKo were not enough for me to declare a connection between the two lects, but they were enough for me to seriously consider this possibility. It’s not unlikely that if I had a complete Swadesh list for Zarma, that I would have found more connections, and if I was comparing three to four languages from each family instead of just one, I would have wound up with a much more complete picture. I do believe, however, that these similarities between Zarma and NKo are enough to further place Songhay’s inclusion into Nilo-Saharan into question.
Week 6: The Sahelic Family
In my previous blog post, I compared the Swadesh lists for two African languages: Zarma, a Songhay language, and NKo, a Mande language, ultimately discovering a not insignificant quantity of commonalities between the two lects. In this post, I aim to take that comparison the next step further, incorporating the languages of Acholi, Dinka, Kanuri, and Dyula into my model. Acholi and Dinka are both Nilotic languages of the Eastern Sudanic branch of Nilo-Saharan, used in Uganda and South Sudan respectively, while Kanuri is the largest language of the tenuous Saharan branch of Nilo-Saharan, spoken in northeastern Nigeria. Dyula is an additional Mande language, used as a Lingua Franca across the Cote d’Ivoire.
The main reason as to why I chose these languages in particular was accessibility. These are all languages with millions of speakers, prominent enough for them to be available on Google Translate. Once again, I would never trust that service in the translation of sentences from these languages, but for individual words and phrases, it is quite effective. Furthermore, my concept behind adding these languages to my model was to see if the limited but non-zero number of correlations between Zarma and Mande would eclipse or fall short of the number of correlations between Zarma and other Nilo-Saharan branches. If it were to turn out that Zarma had more in common with Mande than it did with members of its own family, the idea that Songhay could ever be included in Nilo-Saharan would seriously be put into question. The results of my comparison forced me to ask that question.
However, before I go into the results of my research that made me question the validity of Nilo-Saharan, I’d like to speak on the one angle that seriously corroborated the theory, that being the word for the second-person pronoun. In English, this word is ‘you,’ while in Zarma, it is ‘ni.’ In NKo and Dyula, our Mande languages, the equivalent words are ‘i’ and ‘e’ respectively, both of which lack an ‘n’ sound. On the other hand, this word is ‘in’ in Acholi, ‘yin’ in Dinka, and ‘nyi’ in Kanuri, all of which share the ‘i’ sound and ‘n’ sound present in ‘ni.’ This word alone is not enough to prove the theory of Nilo-Saharan, especially as there really aren't any other words like it on the Swadesh list, but it absolutely should not be ignored.
Within the Nilotic languages, there were very few words with easy equivalents in Zarma. In Dinka, the only solid comparison I could find was the word for ‘to throw,’ which was ‘catu’ in Zarma and ‘cuɛt’ in Dinka. Acholi, however, had a larger number. There was the word for ‘tail,’ which was ‘dibba’ in Zarma and ‘yib’ in Acholi, the word for ‘knee,’ which was ‘kange’ in Zarma and ‘cong’ in Acholi, the word for ‘to stab,’ which was ‘gooru’ in Zarma and ‘goyo’ in Acholi, the word for ‘ash,’ which was ‘boosu’ in Zarma and ‘buru’ in Acholi, and the word for ‘earth,’ which was ‘laabu’ in Zarma and ‘lobo’ in Acholi. It should be mentioned, though, that due to the fact that Dinka had so few potential cognates with Zarma, combined with the fact that both Dinka and Acholi are both provably derived from proto-Nilotic, there is a high likelihood that the similarities between Acholi and Zarma are simply due to chance. If this were to be the case, we would have to separate the Songhay family from the greater Nilo-Saharan family, however, there is still one more subfamily we have to analyze.
Kanuri of Nigeria is the largest of the Saharan languages, and of the Nilo-Saharan languages it was the one that I found had the highest number of similarities with Songhay. Additionally, it should be mentioned that is also the Nilo-Saharan language geographically closest to Zarma of the ones I examined, though their range of speakers are not necessarily close. While I felt rather confident in the prospect of separating Songhay from Eastern Sudanic, the prospect of doing so with Saharan left me with a larger number of reservations. This is because, simply put, the languages have a significant number of potential cognates. The word for ‘one’ is ‘fo’ in Zarma and ‘fal’ in Kanuri, the word for ‘long’ is ‘kuuku’ in Zarma and ‘kuruwu’ in Kanuri, the word for ‘small’ is ‘kayna’ in Zarma and ‘gana’ in Kanuri, the word for ‘mother’ is ‘nya’ in Zarma and ‘yaa’ in Kanuri, the word for ‘to vomit’ is ‘yeeri’ in Zarma and ‘yiloktǝ’ in Kanuri, the word for ‘to blow’ is ‘fuusa’ in Zarma and ‘futǝ’ in Kanuri, the word for ‘to scratch’ is ‘yortu’ in Zarma and ‘kurtə’ in Kanuri, the word for ‘to swell’ is ‘fuusu’ in Zarma and ‘futǝ’ in Kanuri–which may indicate a shared etymology for ‘to blow’ and ‘to swell’–the word for ‘ash’ is ‘boosu’ in Zarma and ‘buwu’ in Kanuri, the word for ‘red’ is ‘ciray’ in Zarma and ‘cime’ in Kanuri, and most interestingly, the word for ‘to spit’ is ‘tufa’ in both languages. This comparison indicates to me that there very well may be a shared origin for the Songhay and Saharan families, but that that origin is not Nilo-Saharan, and that it may instead be adjoined with the Mande family, as corroborated by the evidence I found in the Dyula language. Whether or not this categorization would place Songhay and Saharan into the Niger-Congo family is not to be spoken of within this project, as the question of whether Mande should be included into Niger-Congo is one which could lead to an equally long project. For now, we may assume Mande-Songhay-Saharan to be an independent family, which I’ll refer to as Sahelic in the future, named after the Sahel region which these languages straddle.
Week 7: On Varied Sudanics
According to Starostin’s analysis, Songhay and the greater Nilo-Saharan family couldn't possibly be related. In my last two blog posts, I corroborated this opinion, though my suspicion that Saharan and Songhay were related to one another was not reflected in Starostin’s work. It should be remembered, however, that Starostin’s use of glottochronology is not fully supported by many other linguists. It is a controversial technique, one still placed at a higher pedestal than Greenberg’s mass comparison, but not one to be immediately taken at face value. However, it remains a useful tool.
Starostin’s largest claim in his analysis was likely his proposal that Central Sudanic and Eastern Sudanic were, albeit distantly, related to one another. This is a claim that I’d like to look into, as it exists as one of the most fundamental pillars of the proposal itself, so fundamental that both subfamilies were not only included in Greenberg’s original Nilo-Saharan, but also in Greenberg’s now defunct subfamily of Chari-Nile. It is the general consensus among linguists today to split up the families of Chari-Nile, as there isn't very much they all share in common, and it is increasingly the opinion of linguists today to put the entire Nilo-Saharan family into question.
My goal this week was to incorporate a specific Central Sudanic language into my growing spreadsheet of languages. I already have two Eastern Sudanic languages, Acholi and Dinka, included on the list, and I wanted to add Lugbara, a Central Sudanic language of slightly over two million speakers from Uganda. It is important to note that there are significantly more Eastern Sudanic speakers than Central Sudanic speakers, meaning that I wouldn't have access to any machine translation software to assist me in my comparison. This would not prevent me from achieving my goals, it just meant I had to do things the old fashioned away. Using the ASJP (Automated Similarity Judgment Program) Database, I was able to compile a workable Swadesh list from the word lists of Viveka Velupillai, Guillame Segerer, and Søren Wichmann. Unfortunately, this would leave me with a slightly less satisfactory 100-word list, based on Swadesh’s original model, rather than my ideal 207-word list. This would make my comparison slightly less astute, but since BASIS refused to fly me to Uganda free of charge, I’d ultimately have to settle on this particular matter.
The first thing I’d like to point out when comparing the Lugbara to the Eastern Sudanic languages is the second-person pronoun. This feature, one actively noted by Greenberg, has proved to stand as the strongest piece of vocabulary in support of a united Nilo-Saharan. In Lugbara, the equivalent word is ‘mi,’while in Acholi it is ‘in,’ and in Dinka ‘yin.’ As a reminder, the second-person pronouns in Zarma and Kanuri are ‘ni’ and ‘nyi’ respectively. Lugbara seems to have the most distinct phonology of all of these words, being the only one without an ‘n’ sound, though the ‘m’ sound is quite similar and it still possesses an ‘i’ sound. Still, if we isolate our Eastern Sudanic and Central Sudanic languages, it can be seen how the two words could have a common root, through the process of metathesis, but they could just as easily be wholly unrelated.
Outside of the second-person pronoun, the support for Eastern Sudanic and Central Sudanic belonging to the same family becomes incredibly sparse. However, there does still exist a handful of words between Lugbara and Acholi that bear some resemblance, such as the word for ‘two,’ which is ‘aryo’ in Acholi and ‘iri’ in Lugbara, the word for ‘tree,’ which is ‘yat’ in Acholi and ‘pati’ in Lugbara, and the word for ‘to see,’ which is ‘neno’ in Acholi in ‘ne’ in Lugbara. Unfortunately, there exists a problem with all of these similarities. Acholi and Dinka are very similar languages, as they both belong to the well-supported subfamily of Eastern Sudanic called Nilotic. Likewise, they have plenty of similarities in vocabulary, a result of both carrying words from proto-Nilotic, the lect they’re both derived from. However, all of the Acholi words that I just listed had equivalents in Dinka which were completely phonetically different, meaning that there’s a high likelihood they did not come from proto-Nilotic but rather from elsewhere. Further, since Nilotic is a very well-supported language family, Lugbara can’t just be related to one of them, it would have to be related to both, yet it only shows similarities to Acholi. I suspect the reason for this dissonance is a matter of Geography. Acholi and Lugbara are both spoken in Uganda, and furthermore both in the north-west of Uganda, while Dinka is much further removed in South Sudan. The similarities between Acholi and Lugbara are likely then the result of interaction between the two cultures as instigated by Geography, rather than any ancient common root between the two lects. This could explain why there was only one word that was similar in all three languages, the word for nose, ‘um’ in Acholi, ‘nhom’ in Dinka, and ‘omvu’ in Lugbara.
Ultimately, like Starostin before me I find the links between Eastern and Central Sudanic to be very poorly supported. Starostin suspected that if they were to be related, their proto-language would’ve had to have diverged about 14,000 years ago, and in this number I detect a flaw in the method of glottochronology, as it is simply too large to postulate with any degree of provable accuracy. While families such as Afro-Asiatic have been provided with similar ages, Afro-Asiatic has the advantage of producing multiple ancient written languages, such as Egyption hieroglyphs, something which cannot be said about Nilo-Saharan save for a handful of potentially Nubian names. I believe that, if we are to recognize the inherent limitations of linguistics, that being that we cannot build a time machine, we must keep Central Sudanic and Eastern Sudanic separate for the time being.
Sources:
Boone, Douglas W. and (eds.), Richard L. Watson. 1992. Moru-Ma’di survey report. Nairobi: Sudan Branch, Summer Institute of Linguistics (SIL).
Tucker, A. N. 1967. The Eastern Sudanic languages I. London: International African Institute.
Week 8: Core Nilo-Saharan
In my past handful of blog posts, I’ve been discussing the areas in which the Nilo-Saharan phylum fails to hold together. Throughout this process, there has been a worry that I might be being a little too strict with my judgments. If I were to end up with a picture of Nilo-Saharan that was simply just every subfamily cordoned off into an independent family, I’d definitely develop suspicions that I was simply going about with the process the wrong way. However, my most recent findings have alleviated some of the worries I’ve been hosting inside of me. To be more specific, I analyzed the Fur and Kuliak subfamilies of Nilo-Saharan, the two families that Starostin positioned closest to Eastern Sudanic in his own analysis, and I found quite the number of cognates between the languages I compared.
However, before I start speaking on the specific similarities I detected in my research, I’d like to spend some time speaking on the other side of the senior project, that of my internship. I’ve been interning at RuReady.fyi, a blogging project dedicated to providing helpful frameworks for developing career readiness. Specifically, I’ve been working on a series of articles dedicated to an experience that many face which is quite personal and recent for me, the process of applying to colleges. The philosophy behind this series is that, because so much of what one does in their professional life is related to their experience in higher education, a series dedicated to approaching higher education was due in order to round out the website as a whole. I also found that, because I had just gone through a quite extensive college application process at a school that provides many resources to assist said process, that I might be able to pass down some of the wisdom I acquired to students at schools that don't put as much attention into college application. So far, I’ve found that my internship has really helped me with improving my ability to write informatively–as opposed to my preferred habit of writing narratively–which has overall assisted me in writing the document that will make up my final project on the prospect of Nilo-Saharan.
Now, back to the linguistics. I was able to locate a complete 100-word Swadesh list of the Fur language, published by the University of Khartoum, which I compared with my findings for Dinka and Acholi. I’ve put the list of potential cognates below inside a table, which I hope will be a bit more readable than some of my previous choices in formatting.
English | Acholi | Dinka | Fur
this | ene | kënë | in
all | weng | eben | nyet
small | tidi | kë nyot | ʔiti
human | dano | raan | duo
seed | kodi | kuët | kola
meat | ringo | miɔ̈c | nino
eye | nyin | shim | ungi
foot | tyen | cök | tar
heart | cwiny | puɔ̈u | kilma
eat | camo | cam | am
stand | cung | kɔ̈ɔ̈c | ker
say | waco | lueel | wa
moon | dwe | pɛɛi | dual
star | acer | kuɛl | ʔuri
cloud | kal cot | piny tɔ̈c | kutu
burn | me wango | cuäny | suing
mountain | got | gɔn | pugo
warm | ma lyet | atuöc | toke
round | olunge | ye rɔt | orol
Overall, I found enough valid potential cognates within the Swadesh list for me to be rather confident in declaring that Eastern Sudanic and Fur are languages related to one another.
The other language we’ll be looking at in this post is the Ik language of Uganda, the only non-moribund member of the Kuliak family. While the Ik language only has about 14,000 speakers, there has been quite a significant amount of anthropological study into their particular society. This is in part due to a particularly horrendous book by Colin Turnbull entitled The Mountain People which presented an incredibly distorted picture of the Ik culture that offended the anthropological community at large so much that many more experienced researchers began committing themselves to studying the Ik people in an effort to disprove Turnbull. As a result, I was able to find a complete English-Ik dictionary, which I used to develop a Swadesh list for the language. Below, I’ve charted the comparisons I drew between Acholi, Dinka, and Ik.
English | Acholi | Dinka | Ik
he | en | yen | ˈńtsḁ
this | eni | kënë | nā
that | ni | xɔn | nē
who | nga | de | n̩̄do᷆
three | adek | diäk | àɗīn
short | cek | kë chiëk | kūɗōn
narrow | ma ding | këcë nhïac yïc | ɪ̄ɗɪ́ŋɔ́n
thin | tidi | tɔ̄ɗɔ̄n
husband | laco | moc | e᷄ákw̥ḁ
dog | gwok | köŋ | ŋókʰ
bark | koko | kuär | ɓōɗókʰ
head | wik | nhom | ˈǐkḁ
knee | cong | tɛ̈k | kútúŋ
hand | cing | cin | kwɛ̄tʰ
back | cen | dhuk ciëën | kān
spit | laro | tuɔc | tātōn
vomit | ngoke | ŋök | hɛ̄nōn
see | neno | tïŋ | ɛ̀nɛ̄s
think | tamo | tak | tāmés
dig | tongo | wek | tōkōpḁ
stand | cung | kɔ̈ɔ̈c | ŋ̀kōːn
fall | poto | lööny | làka᷄méto᷆n
sand | kwo | kuɛt | ʄūmūʄūmás
dust | apuwa | abirguɔ̈p | búr
night | dye wor | waköu | mūkú
old | oti | kë ci dhiöp | dunétón
rotten | otop | nhany | māsān
sharp | bit | thɛ̈p | tsʼītsʼōn
dry | otwo | kë ci riεl | tsʼōːn
Ultimately, the list of potential cognates I found between Acholi, Dinka, Fur, and Ik leads me to believe that all four languages are related within a linguistic family I will be referring to with the name Core Nilo-Saharan.
Sources:
Waag, Christine. 2010. The Fur verb and its context (Nilo-Saharan: linguistic analyses and documentation 26). Cologne: Rüdiger Köppe.; McKeever, Ashley L. 2014. Sonority and Syllable Weight in Fur. University of Texas at Arlington doctoral dissertation.; Beaton, A. C. 1968. A Grammar of the Fur Language (Linguistics Monograph Series I). Khartoum: Sudan Research Unit, University of Khartoum.
Schrock, Terrill. The Ik language: Dictionary and grammar sketch. Language Science Press, 2017.
Week 9: Macro-Central Sudanic
While in a previous blog post, I had concluded that there wasn’t much support for a linguistic connection between the Central Sudanic and Eastern Sudanic families of Nilo-Saharan, that doesn't inherently mean that the families in question are not related to any of the other families included in Nilo-Saharan. In a recent post, I suggested that the Fur and Kuliak families were in all likelihood related to Eastern Sudanic, and in this post, I intend to take a look at the languages which could potentially be related to Central Sudanic.
The languages which I was most interested in examining were of the Kunama and Maban families. Both of these families were marked as being potentially related to Central Sudanic within Starostin’s model. When I first saw this comparison by Starostin, the inclusion of Maban made a lot of sense, as all of the Maban languages are located within the nation of Chad, and while they’re located further south, there do exist some Central Sudanic languages in Chad as well. The inclusion of Kunama, however, drew a bit more suspicion, as the Kunama family–which is made up of only one language, also called Kunama–is located entirely within the nation of Eritrea, quite a distance away from the confluence of most Central Sudanic languages. Nevertheless, I understood that geographical distance is not what makes or breaks the validity of linguistic grouping, what matters most is the language itself.
The languages I decided to compare in this particular instance were Lugbara of Central Sudanic, Maba of Maban, which is the most spoken of the Maban languages and also their namesake, and Kunama of Kunama. I was able to once again utilize the ASJP database for the compiling of my Kunama word list, but unfortunately, the database was unable to yield any word list for Maba with more than sixty words. Luckily, I was able to find another word list for Maba which eclipsed 250 words. Unluckily, that word list had been compiled by one of my sworn linguistic enemies, Roger Blench, a proponent of Nilo-Saharan who apparently doesn't believe in the Swadesh list as it seems like he chose many of the words in the list at random. Nevertheless, the list was expansive enough for me to make an adequate use of it–and I must say, it was very useful. When initially comparing Lugbara and Kunama, I found a significant number of similarities, but I wasn't sure if it was quite enough to deem them a united family, especially considering their geographic discrepancy. However, when I included Maba in my comparison, the tables began to turn, as it was able to effectively act as a stepping stone between the more distant languages, sharing many similarities with both of them and leading me to believe that the three families may be one.
Below, I’ve provided a chart which displays many of the similarities in vocabulary I found between the selected languages. For the purpose of this project, I will be using the term ‘Macro-Central Sudanic’ to describe this united family.
English | Lugbara | Kunama | Maba
I | ma | aba | àm
you | mi | ena | mè
we | ama | kime | maŋ
one | alo | ela | tɔɔ
two | iri | bare | mbààr
small | ali | damada | ɲílì
man | agupi | ka | kámbà
bird | aria | shurka | kùʃi
meat | nya | ɲù
blood | ari | kokoba | àríi
bone | fala | sangga | kàɲjí
fat | oso | anya | ɲamu
egg | gbe | kokinya | kademi
nose | omvu | bobona | boiɲ
tooth | si | ma | sati
tongue | ede | ngema | delmi
knee | kumu | kokita | túmó
hand | d- | kona | kara
belly | ʔa | kasa | kùɲi
heart | culu | ulfa | kùlì
eat | nya | ngada | aɲ
bite | ci | ini | is
see | nde | nti | oko
die | da | bal | ɔy
water | yi | bia | bi-ta
smoke | acika | dud | kùuldaa
fire | aci | toma | wàsí
ash | haina | auun
road | gari | tabila | tarfu
green | dirga | drì-drí-yàk
white | emvero | ara | fà-fár-àk
Sources:
Blench, Roger. "The Maban languages and their place within Nilo-Saharan." (2021)
Bender, Marvin Lionel. 2001. English-Kunama lexicon. Afrikanistische Arbeitspapiere 65. 201-253.; Bender, M. Lionel. 1996. Kunama (Languages of the World/Materials 59). München: Lincom.; Bender, M. Lionel. 1971. The languages of Ethiopia: A new lexicostatistic classification and some problems of diffusion. Anthropological Linguistics 13. 165-288.
Week 10: Koman, Gumuz, Shabo, and Goodbyes
There is one final corner of Nilo-Saharan which I wish to analyze in this series of blogs. This corner consists of the Koman, Gumuz, and Shabo language families, all of which reside in the nation of Ethiopia. However, due to the coming deadline of the final project itself, I'd like to slightly change the format of this post and provide to you all a direct preview to my paper, On the Inconsistency of the Nilo-Saharan Phylum, provided in the paragraphs below.
While Greenberg’s initial proposal grouped the Koman and Gumuz families not only as subfamilies of Nilo-Saharan but further as subfamilies of the Coman subfamily of Nilo-Saharan, this particular instance of comparison has been highly debated by Greenberg’s successors. While the corpus of research on these particular families can be particularly slim, the shared base vocabulary between them and the languages of Core Nilo-Saharan is practically non-existent. However, this is a statement which must be tempered by the knowledge of the aforementioned lack of research. The available vocabulary lists for these languages tend to be minute, partially due to the political situations affecting the populations that speak them. For example, Uduk, the largest of the Koman languages at only about 20,000 speakers, has been in quite a fraught situation, as nearly the entire population of speakers has had to migrate from South Sudan to Ethiopia and back again due to various political crises within the former nation. Crises like these make it very difficult for research to proceed on the affected languages as very few linguists enjoy working in warzones, especially since there is little financial incentive to involve oneself in African linguistics to begin with. All of this ultimately leads to a drought of research when it comes to the vocabulary of these languages, making it difficult to ascribe conclusiveness to one’s research.
Similarly, the Shabo language is another instance of a lect facing a drought of research. It is not a Koman or a Gumuz language but rather a language isolate which is similarly spoken in Ethiopia. Shabo is not facing quite the same political crises as those of Uduk, but what it is facing is the perils of a very low number of speakers. Shabo has less than a thousand speakers, and any language of that quantity, though especially in Africa, will suffer from a drought of research. As a result, in Uduk, Gumuz, and Shabo, there were a very limited number of Swadesh words to compare, though I still found some notable matches, displayed in the chart below.
English | Gumuz | Uduk | Shabo
we | ile | am | an
dog | k"owa | ak”a | kan
louse | sukuna | Sokom | nEna
leaf | sindy$a | C”emen | C”am
ear | c”ea | C”e | k"iti
tooth | k"w~osa | Se7 | k”aw
breast | kuwa | ako | duh
drink | f3 | ph~i | who
hear | gEs | Cikh~ | ECEt
Ultimately, I do not believe that the comparisons shown are proof that the given languages are related, nor do I believe the lack of cognates to be proof that the languages are not related. If anything, I believe these results are proof that more research must be done into this particular line of study. Koman, Gumuz, and Shabo are very unlikely to be related to Core Nilo-Saharan, but their internal relations with one another could vary tremendously.
Sources:
Bender, M. Lionel. 1971. The languages of Ethiopia: A new lexicostatistic classification and some problems of diffusion. Anthropological Linguistics 13. 165-288.
Ehret, Christopher. 1995. Do Krongo and Shabo Belong in Nilo-Saharan?. In Nicolaï, Robert and Rottland, Franz (eds.), Actes du Cinquième Colloque de Linguistique Nilo-Saharienne / Proceedings of the Fifth Nilo-Saharan Linguistics Colloquium, Nice, 24-29 August 1992 Nilo-Saharan language studies, 169-194. Köln: Rüdiger Köppe.