Villagers in Bounou are nominally Muslim and celebrate some Islamic festivals (Image: Abbie Hantgan)
A remote village in Mali speaks a language that hides as much as it communicates. Did this unique tongue evolve to shield escaped slaves?
WHEN Westerners say 鈥淭imbuktu鈥�, it is as if we are talking about the ends of the earth. But the city鈥檚 remoteness is nothing compared to the small village of Bounou, tucked inside a rugged cul-de-sac valley 250 kilometres to the south. No European had ever visited the surrounding until French colonial officer Louis Desplagnes reached it in 1904 鈥� and even he didn鈥檛 get as far as Bounou.
Advertisement
is one of the few Westerners to have reached the village in recent years. She can still recall the last leg of her journey, after an arduous two-day bus trip to the small market town of Konna (see map). It was the height of the rainy season, meaning that a 5-hour journey by donkey cart was the only way to traverse the canyon where Bounou perches.
鈥淭he track was flooded waist-high,鈥� she says. 鈥淏ut the floodwater didn鈥檛 keep the cart from finding every rock and rut in the track along the way.鈥� Eventually, they reached a boulder marking the end of the track and she saw Bounou 鈥渉anging on the cliff side鈥�. It was, she says, 鈥渁 scene out of time鈥�.
For Hantgan, Bounou鈥檚 remoteness was one of its main attractions. She wanted to document the words spoken by its inhabitants, the Bangande. Although these people share much of their culture with the surrounding Dogon people, their language, called Bangime, is very different and has many unusual characteristics. Understanding its origins could therefore tell us a lot about the history of this little-explored area of Africa, while also offering a way to investigate the birth and evolution of languages.
As Hantgan embarked on her visit to the region, she knew it came with its share of risks. She was taking over research started by the young Dutch linguist Stefan Elders, who passed away while working in Bounou the previous year. He had contracted a stomach ailment and the isolation of the village meant he couldn鈥檛 reach a hospital in time.
Elders鈥檚 work was part of the US National Science Foundation鈥檚 , headed by linguist Jeffrey Heath at the University of Michigan. The project investigates relationships between the various languages spoken by the Dogon peoples living on the Bandiagara Escarpment and the adjacent Seno Plain. Some 80 named Dogon speech varieties exist, which Western linguists categorise as 22 separate languages and many more dialects.
Hantgan鈥檚 experience meant she was ideally qualified to take Elders鈥檚 place in the project. While volunteering with the US Peace Corps in Mali, she had learned Fulfulde and a called Bondu-so. Both would prove useful in her doctoral research into Bangime. Fulfulde, used as a lingua franca or bridge language in Bounou, provided her with a tool to talk to local people and elicit words in Bangime, while Bondu-so helped illustrate possible connections with the other Dogon languages.
Hantgan began by compiling a list of common words in Bangime 鈥� a task that often attracted derision from the locals. 鈥淓very day, villagers on the way to their day鈥檚 work in the fields would see me seated inside with my notebook and pen, asking a consultant to repeat the difference between 鈥榤oon鈥� and 鈥榳ater鈥� over and over again,鈥� she remembers. 鈥淲ith their hoes over their shoulders, they would make fun of me for spending another day sitting in the shade instead of going out to tend crops.鈥�
It was a lonely and frustrating time for her, cut off from contact with family and friends and without even a shortwave radio to remind her of home. But she soon found an ally in the village chief 鈥� although he had initially been anxious about her research. He said it upset him that visitors from other Dogon villages often asked why the Bangande have different surnames and don鈥檛 look like the rest of the Dogon, even though the Bangande consider themselves to be a Dogon people. Despite concerns that the research might emphasise those differences, he could see how much effort Hantgan was putting in. When villagers would chide her within the chief鈥檚 earshot, he would say: 鈥淪he is tending her crops! The pen is her hoe, and the notebook is her field.鈥�
Once Hantgan had compiled a suitable number of words, her next task was to identify any that were 鈥渃ognates鈥� with the other Dogon languages. Cognates are words originating from a common root. For instance, the word 鈥渓una鈥� in Italian is related to the word 鈥渓une鈥� in French, 鈥渓luna鈥� in Catalan and 鈥渓ua鈥� in Portuguese; all come from 鈥渓una鈥� in Latin, the mother tongue from which these Romance languages diverged. Identifying cognates can therefore help demonstrate whether two languages have a common origin.
Hantgan and her colleagues found that it was not unusual for at least 50 per cent of the vocabulary of a given Dogon language to be cognate with the vocabulary of another Dogon language 鈥� whereas just 10 per cent of Bangime鈥檚 vocabulary seemed to share roots with Dogon terms. Rather than reflecting a common mother language, this small shared vocabulary may simply be due to Bangime speakers borrowing a few words from their neighbours, in the same way that cultural ties resulted in English borrowing words like sushi, pergola and pyjamas.
In this way, Hantgan鈥檚 research seemed to mark out Bangime as the most recently discovered language isolate 鈥� a tongue not related to any other language. That is of interest to historical linguists like at the University of Hawaii in Honolulu, who points out that scholars tend to classify African languages as belonging to one of four major families: Afro-Asiatic, Nilo-Saharan, Niger-Congo or Khoisan. The recognition of Bangime as an isolate might suggest that the classification system needs a rethink, he says.
Orphaned tongues
Further evidence for Bangime鈥檚 uniqueness resides in the fact that its grammar is radically different from that of the other languages spoken by Dogon groups. To give an example: although the Dogon languages join words to form compounds, as does English (think football, rainstorm or driveway), Bangime doesn鈥檛. On the other hand, prefixes are found in Bangime, while being notable by their absence in the Dogon languages.
These differences are somewhat surprising, because in other ways, the Bangande and Dogon cultures are very similar. The Bangande wear the same clothing and jewellery as the Dogon people, and both use architecture 鈥� mud brick, coiled clay and stone masonry structures set into the cliff face 鈥� for granaries and burial grounds.
Looking at the archaeological record, it is easy to assume that people who share such material cultures are part of a single language community. This has been the basis for theories about the origins of the Indo-European languages spoken in Europe and Asia, for instance. Yet the unusual relationship between the Dogon and Bangande reminds us that we can鈥檛 rely on these assumptions.
What leads to a language becoming an isolate? Campbell notes that isolates may be the orphans of larger linguistic families whose other members have slowly died out 鈥� perhaps because the speakers adopted other languages. Many social, political and economic factors probably influence which languages survive, and which perish. Tongues like Bangime could represent a concerted effort to resist shifting to others鈥� words.
The first hint of this comes from the very name Bangande. Bang translates as secret, hidden, or furtive, and 鈥�ande is a plural suffix 鈥� like -s in English 鈥� so the combination translates as 鈥渇urtive ones鈥�. The word Bangime is formed in a similar fashion, with the suffix 鈥�ime signifying language; thus it translates as 鈥渟ecret language鈥�. Clearly, they were once keen to keep to themselves.
Hantgan discovered further clues as to why that might be when she moved from compiling words and phrases to collecting longer portions of continuous speech. Along the way, she documented oral histories of the Bangande villages as places of refuge for escapees from Fulani slave caravans, which served the internal and transatlantic slave trades. Peoples such as the Bobo, Samo and the Bangande themselves were commonly targeted by slave traders because Islamic law afforded non-Muslims no protection against enslavement.
The oral histories described many of these escapees as children who were seized while they were gathering firewood and water outside their villages. They had sacks placed over their heads for several days to make sure they were unable to orient themselves and attempt escape back to their home village. Some of those who did escape eventually found their way to the Bangande settlements, where they were integrated into the community and learned Bangime.
The integration of individuals from across the Sahel to the north and the Volta river basin to the south may explain the physical distinctiveness of the Bangande people. Being joined by runaways seeking sanctuary from slave raiders may be one reason the Bangande have come to refer to themselves as 鈥渢he furtive ones鈥� 鈥� and might explain why they have been determined to keep their own language.
鈥淭he slave trade may explain why the Bangande were determined to keep their own language鈥欌��
The Bangande鈥檚 eagerness to retain their secrecy may even have led Bangime to develop what British linguist Michael Halliday calls an anti-language. That鈥檚 a distinct 鈥渄ialect that serves to mark off a group of speakers from the larger society鈥�, resulting in an 鈥渁nti-society鈥�. Jargon is one common element of such dialects, but Bangime鈥檚 anti-language also uses more elliptical tactics.
Hantgan didn鈥檛 become aware of the existence of the anti-language until near the end of her third year of work in Bounou, when she had gained some conversational proficiency in Bangime. She started to see a pattern in which some terms were the polar opposites of the things they described. For example, a particular white-barked tree was referred to as 鈥渂lack-eyed,鈥� and a particular black-barked tree as 鈥渨hite-eyed鈥�.
As her mastery of the language improved even more, Hantgan began to notice that many words she had asked the villagers for didn鈥檛 regularly appear in natural speech, where circumlocutions were often preferred. For example, she had previously recorded the term 蝉谩脿苍 for fence. Yet one day, she heard a garden fence being referred to as 鈥渟tick(s) put into the ground so that people may pass next to the rice鈥�. Similarly, cakes were sometimes called 鈥減owder which has been sweetened鈥�, while sunglasses were 鈥渂lack things to hide the eyes鈥�.
This sort of linguistic theatricality and deception are an example of what Mark Pagel at the University of Reading, UK, calls 鈥渁 powerful social anchor鈥�. He has argued that languages evolve to deceive and exclude others, as much as to ease communication. A roundabout way of describing objects is just one strategy that helps the Bangande set themselves apart from other group 鈥� and perhaps helped them to distance themselves from the passing traders who may have begun to pick up their everyday words.
Nuances and exceptions
The slave trade also seems to have left its mark in the way Bangime distinguishes social class. The 鈥渁ristocracy鈥�, who claim to descend from the families who harboured the escaped slaves, speak in a high register associated with a more complex tonal system, compared with the speech of the 鈥渟erf鈥� population, who are thought to be descended from those escapees.
A process known as over-regularisation may account for the distinction. Learners tend to assume regular patterns in a language until a wealth of exposure or being corrected shows them the nuances and exceptions. For instance, non-native speakers of English may say 鈥渃atched鈥� instead of 鈥渃aught鈥�.
Such errors can be difficult to overcome, and they sometimes feed back into the native language. Indeed, many linguists now believe this can explain why grammar gets simpler over time for languages that have a lot of contact with outsiders, like English. It is easy to imagine that the escapees learning Bangime as a second language over-regularised its tonal system 鈥� leading to patterns that are distinct from those used by people descended from the native inhabitants.
The ongoing conflict in Mali means that fieldwork has been halted for the foreseeable future 鈥� yet there is much more to discover. One of Hantgan鈥檚 long-term research goals is to investigate links between the origin of the Bangande people and the Dogon cultures.
Previous researchers had suggested that when the Dogon arrived about 600 years ago, they displaced the existing populations in the region. As evidence, they pointed out that historical Tellem structures and funerary remains don鈥檛 seem to correspond to present-day Dogon material cultures.
The Ounjougou research project at the University of Geneva, Switzerland, however, has revealed how pre-Dogon and Dogon material culture and funerary practices subtly influenced each other. It could be that the Bangande were those people who lived in the region before the Dogon arrived and shared some of their cultures with the newcomers, explaining the similarities we see today.
Alternatively, the ancestors of the Bangande may have arrived along with those of today鈥檚 Dogon, but speaking an unrelated language. Other groups may have also moved to the area, with only the Bangande resisting the shift to using a Dogon language. Until the security situation in Mali improves, it won鈥檛 be possible to gather fresh data related to these hypotheses.
At present, Hantgan is eagerly working as a newly minted postdoctoral fellow at the School of Oriental and African Studies in London. Her position will see her beginning field research soon in rural Senegal, but she also hopes to return to her friends and research in Bounou. Despite the hardships, her enthusiasm is as strong as ever. 鈥淚nvestigating the warp and weft of tone, the rainbow of vowel harmony and the ladder of consonant mutation, these are the intricacies that make human speech so fascinating to me,鈥� she says.
This article appeared in print under the headline 鈥淭he secret ones鈥�
