France

In March 2012, after a viral video asked Western teenagers to help catch a Ugandan warlord, the writer Teju Cole published an essay in The Atlantic called “The White-Savior Industrial Complex”. His argument, as I understand it, is that the white saviour goes to Africa to have a great emotional experience, wants to be thanked for it, and rarely asks the people he is saving what they want, or why they needed saving in the first place. M.J. Hydara’s novel Songs of Dust and Water (2026) seems at first to tell exactly this story. Adam Sterling, a systems engineer for a foreign fund called Proxima, arrives in Madina, a village in the dry north of the fictional Republic of Awdagha, to build a solar and water plant, “the Glass House”. By “white saviour” I mean a figure from outside, usually Western, who is placed at the centre of a story about other people’s suffering and is given the power to solve it. The novel’s own word for him is “Tubab”, which the glossary defines as a foreigner from Europe or America. The plant burns, and Adam is blamed. Then, in Part Two, ten years later and six hundred kilometres south, the same man reappears in the village of Nafawol, but he no longer leads anything. We may therefore ask how the novel takes the Tubab saviour apart, and what it builds in the space he leaves. I will show first that Part One builds Adam as a saviour of systems who does not want to be touched; then that the novel dismantles him twice, through failure and through a deliberate silence in the narration; and finally that what is left is a story with no centre, in which the Tubab becomes one pair of hands among many, and which asks a French reader some questions as well.

The first image of Adam is of a man who wants his work to be untouchable. Presenting his “Madina Blueprint” to the Ministry in Bakoro, he promises the Minister that “Nobody in Madina will ever need to touch it” (p. 8). He means it as a kindness, since the village will carry “none of the burden”, but the sentence contains his whole idea of help: the people are “beneficiaries”, and help is something done to them. The narrator notes that Balla the griot remembers “the silence after the word touch” (p. 9), and Minister Khresay asks the question the rest of the novel answers: “What is your plan for the people of Madina?” (p. 7). Balla’s own verdict, that a system without roots “is a house built for the wind” (p. 7), comes with a look at Adam’s “clean, uncalloused hands” (p. 7). Hands are the key image of the whole book, and Adam starts with the cleanest pair in it. His arrival confirms the portrait. The convoy drives straight into the square, and Adam “gave a short polite nod to no one in particular. Then he turned his back on the ring of faces” (p. 12). He arrives as a man who has nothing to learn from the people watching him.

This is where the comparison with Joseph Conrad’s Heart of Darkness (1899), which we studied in Première, seems natural to me. Marlow’s aunt sees him as “an emissary of light”, and Kurtz writes a report full of noble words for a society meant to civilise Africa, with a postscript scrawled at the bottom: “Exterminate all the brutes!” Adam is obviously no Kurtz: he wants to light a clinic and has no wish to rule anyone. But the structure is similar: light and knowledge come from outside, and Africa is the place where they are applied. Chinua Achebe, in his 1975 lecture “An Image of Africa”, criticised Conrad for turning Africa into a background and for denying Africans almost any intelligible speech. What strikes me in Hydara is that the Africans of Madina talk all the time, and talk well, but Adam cannot hear them. When the certified bracket snaps, Malik the smith mends it in his forge, and Adam tells the contractor, “Log the delay,” and then, “We’ll wait for the certified replacement” (p. 16). Malik’s verdict, spoken to his apprentice in Nabari, a language Adam does not speak, is: “A thing you are not allowed to mend is not yours” (p. 16). The reader understands the proverb, and the saviour does not. In the same way, when Aisha’s box of sisal, charcoal and grass works better than his filters, Adam calls the data “inconclusive” (p. 31), and the box goes into the scrap store. His explanation to Talib, the weaver’s son who sweeps his floors, sums up his vision. The plant works, but the country around it does not: “None of it is rated. None of it is certified. I can’t put any of it in the model” (p. 29). Africa has become, as in Conrad, the background that spoils the picture.

Adam does have a human side, and Hydara shows that it is part of the problem. When Talib shows him his drawings of the plant, Adam asks his name, “though he knew it” (p. 30), and Talib answers that it “means student”. The saviour needs a disciple, and he chooses one who repeats his lessons back to him: “Nothing gets in. Nothing stops” (p. 31). It is a touching scene, but the reader knows that Talib has just been taught to distrust his own hands.

The novel could have stopped there and made Adam a villain. It goes further, and takes him apart piece by piece. The first stage is failure, and it comes through his own method. Talib follows the data, as Adam taught him, and finds that the plant’s money goes to shell companies linked to the brother of Minister Dunia. Adam refuses to see it: “Data can be misread,” he says, and “This is noise” (p. 51). The saviour, who believed only in numbers, refuses the numbers when they point at a power he has signed away. Later, when Dunia speaks to the village in Nabari after the fire, the reversal of Conrad is complete: “Adam could not follow the words, but he could see where the finger pointed” (p. 57). Now it is the white man who stands in the crowd without understanding, and the African minister who has the microphone and the cameras. The novel even shows us the caption written in advance: “foreigner restrained by Awdagha soldiers” (p. 57). The saviour has become an image in someone else’s story.

What I find most interesting is that Hydara does not let Adam keep the comfort of a villain either. He believes Dunia burned the plant on purpose. Aisha, the engineer, corrects Talib: Mr Sterling believes it “because it is easier to be beaten by an enemy” (p. 74). Dunia did not need the plant to burn; she only needed its money. Dunia herself has the strongest speech against the saviour in the book. She grew up in the north and “counted eleven plaques” (p. 61) left by charities, none of which still worked, and she concludes: “Your Mr Sterling is an island. So is every fund like his. I am trying to build a country” (p. 61). She is corrupt, and she is right about something. Even Aisha, in her exercise book, writes a judgement that is fairer than any of the ministers’: refusing the bracket was sound engineering, but “leaving it in the dust without learning the smith’s name had not” (p. 44). Aisha places the fault in his attention, which is harder to repair than a bracket.

The second stage is the confession. At the board meeting two years after the fire, Adam sends a single page, which ends: “I believed that if it ran well enough, it would not matter who owned it. I was wrong, and the village paid for it” (p. 82). When Dr Arramata, the engineer born in Nafawol, meets him at his review desk, he says of the danger he describes, “I’ve done it”, and “He did not look away when he said it, and he did not explain” (p. 83). Here the novel begins its most original move, which is to take away Adam’s voice. The man who presented the Madina Blueprint with metrics on page four now writes “local ownership” in pencil in other people’s margins, and tells Arramata to write her proposal “so it doesn’t need you. Then it won’t need me either” (p. 84). When she asks for him to carry the money, she adds one condition: he must learn Nabari, “Enough to be argued with” (p. 84).

When Adam arrives in Nafawol, he introduces himself with a sentence that sounds like a job description, even a demotion: “My name is Adam Sterling. I was an engineer. Now I carry the money, and I answer for how it is spent” (p. 90). The saviour has become a carrier. And Kadiatou, the Water-Keeper, answers him with the line that gives the chapter its title: “Words are wind, Tubab” (p. 91). From this point, and this is what I want to insist on, the narration almost never enters Adam’s mind about the present. In the terms of Gérard Genette, which our teacher made us learn, Part Two moves between internal focalisations through Kadiatou, Fatou, Hassan and Bintu, while Adam is seen only from outside, in external focalisation. We know what he does, never what he feels about it. The best example is the reading of the agreement under the baobab. Kadiatou tells him, “You will put a fence round the machine,” and then asks: “Who keeps the padlock?” (p. 101). He answers, “The committee,” and the narrator adds: “He said nothing else that afternoon” (p. 101). In Madina he would have explained the warranty, the protocol, the system. In Nafawol he says two words and then sits silent. The reader is not told whether this silence costs him something. We have to read it from outside, like the villagers.

The novel repeats this technique several times. When Kadiatou opens the academy, Hassan “saw the Tubab look away and press the back of his wrist briefly to his mouth” (p. 150). That is all: an emotion seen by another character and never explained. Later Hassan wonders whether Adam’s silence in council was fear, whether he had once been “the man a room waited for” and “had taught himself to keep his hands still” (p. 177). Again, this is Hassan’s hypothesis, not Adam’s thought. In my opinion this is the real dismantling of the saviour. Teju Cole says that the white saviour wants a big emotional experience; Hydara refuses to give Adam one, or at least refuses to give it to us. In Heart of Darkness, Kurtz dies with his famous “The horror! The horror!”, and Marlow builds the whole end of the story around those last words. Adam gets no last words. When the Teranu family leaves for the coast after the vote, “He stood on at the lorry park after the others had gone home” (p. 204), waiting until the dust has settled on the road, and the section ends there.

If the Tubab is no longer at the centre, who is? The answer of Part Two is: everyone else, and the novel is careful to show that the project was never Adam’s. Asked who sent them, Dr Arramata simply answers, “I did” (p. 89), and when Kadiatou asks her why she chose Adam, she replies, “I didn’t choose a good man, Yaara-Jaxo,” and explains that she chose “a man with a specific education” (p. 95). His education is his failure. Fatou insists on reading the agreement aloud herself, and when the alkalo hands her the key, “She put it in her purse, with her coins” (p. 101). On the Road of Hands, the boys of the youngest kafo put Adam at the back of the line and take the basket off his head “until he carried it their way” (p. 112). Asked about hidden accounts, he admits in public what he never admitted in Madina: “I told him he had misread them, and I kept the machine running” (p. 132). At the end, when he hands over Proxima’s second key, Fatou refuses even the gesture of a gift: “It was always ours. This is only the receipt” (p. 213).

For me, the scene that completes the dismantling is the last one with Adam, on the causeway at night, when his electric vehicle breaks down. A young mechanic, “not from Nafawol”, kneels, finds the loose connector and tells him to hold the lamp. “He held it where she told him to and did not let it move” (p. 215). This sentence echoes the night in Bakoro, in Part One, when the student nurse Hawa, Talib’s sister, held a torch for the midwife and “did not let it move” (p. 32). The engineer who once refused the smith’s repair is now the assistant holding the light for a young African technician. Then comes the sentence I find the most beautiful in the book: “She did not ask his name” (p. 215). In Madina, Adam asked Talib a name he already knew, in order to make him a disciple. Here, nobody needs his name at all. The chapter ends: “For a long way he watched the mirror more than the road” (p. 215). He looks back at what he has left behind, but the road ahead, whose soft stretch the crew have relaid with their own hands, belongs to others.

The final chapter confirms that the centre has moved. When the visitors from Madina arrive by train, it is Talib, now a weaver who mends pumps and radios, who asks the question that Adam’s Blueprint never asked: “Who is poor here now?” (p. 229). Hassan answers honestly that the farmers at the edges still are. When Halima asks her great-grandmother if the Tubab was like Hassan when he came, Kadiatou says, “He came from elsewhere” (p. 230), and to the question of whether she trusted him, only “Some days” (p. 230). There is no statue to the saviour, no plaque, just a measured judgement. Her last message to him is about someone else: “Tell him the weaver’s son has good hands” (p. 231). The phrase closes the circle that started with Adam’s “clean, uncalloused hands” in the Ministry.

Still, I do not think the novel pretends that the saviour can disappear completely, and this is its honesty. Proxima’s money remains. At the charter meeting, Dr Arramata admits that “The money bought steel” (p. 173), and Evelyn Reed, the businesswoman who brings TransGlobal’s offer, throws a difficult question at Adam about the villages that will wait ten years for help: “Have you thought about them?” (p. 195). She also tells him he is “the only person out here who has ever lost one of these” (p. 195). Adam does not answer that question. His earlier reply to her, “Madina didn’t get to vote” (p. 194), is the only defence he offers, and what he defends there is a procedure. He never defends himself.

Reading this as a French student, I could not ignore one detail. Awdagha, the country of Part One, uses French alongside Nabari in its administration. Mustapha greets the Minister with “Madame la Ministre” (p. 38), Madina had a “chef de village” (p. 16), and the foreign contractor gives orders to Malik “in halting French”: “Dig over there. Là-bas” (p. 16). The language of power in Madina is partly ours. When I was in collège, the news was full of the end of Operation Barkhane, which had followed Operation Serval in Mali in 2013, and of French soldiers leaving Mali, Burkina Faso and Niger. I do not know enough to judge those operations, and the novel does not mention France at all. But it made me notice that the saviour figure is not only a character in American videos or in Conrad. The novel never names the city where Adam’s office is, but one of the few things we learn about it is that “a tram went by, lit and full of people going home” (p. 84). Lyon has trams too. It is a small detail, and I may be reading too much into it, but it reminded me that the “elsewhere” Kadiatou speaks of could easily be here.

To conclude, Songs of Dust and Water takes apart the Tubab saviour in two movements. In Part One it builds him in detail, as a man of clean hands and certified systems who cannot hear the people he wants to help, and lets him fail through his own method, without giving him the comfort of a simple villain. In Part Two it removes his voice: told from outside, silent about the padlock, carrying the money and a basket, holding a lamp for a mechanic who does not ask his name, he becomes one participant in a story that belongs to Kadiatou, Fatou, Hassan and, in the end, to Talib. The novel replaces the saviour with institutions and people, and it keeps a trace of his money as a reminder that the question is not fully solved. As an ouverture, I think of Ousmane Sembène’s novel Les Bouts de bois de Dieu (1960), about the strike of 1947–48 on the Dakar–Niger railway, a line built by the colonial power to carry goods to the coast. In Hydara’s last chapter, a railway line dead for forty years is reopened for the Union’s trade and carries the people of Madina south to Nafawol. It would be interesting to compare how the two novels imagine a railway when it stops belonging to those who built it.

In December 2023 the Burkinabè farmer Yacouba Sawadogo died. He was known as “the man who stopped the desert”, because from the 1980s, in the Yatenga region of northern Burkina Faso, he revived an old technique called zaï and grew a forest on land everyone thought was dead. He had no ministry behind him, and in 2018 he received the Right Livelihood Award, often called the “alternative Nobel”. His story raises a question that is at the heart of our programme on the environment between exploitation and protection: who protects nature best, the people who live on the land or the state that governs it? M.J. Hydara’s novel Songs of Dust and Water (2026) gives an interesting example. In its second part, Nafawol, a village on an abandoned floodplain in the fictional Republic of Tekuru, faces a well that has gone salty, a crusted soil and a river that has left. With money from a foreign fund, Proxima, and with an engineer born in the village, Dr Arramata, it restores its water, its soil and its trees, and ends up at the heart of a union of thirty-one villages.

First, the terms. By “community” I mean a group of people who share a territory and its resources and who organise themselves with their own institutions: here the village committee, the women’s kafo, the Coffer Fund and later the Senebian Economic Union. By “state” I mean the public authorities of the Republic: ministries, regulators, inspectors, the national rail authority. “Protecting the environment” means more than forbidding its use. In the Sahel it means keeping resources such as water, soil and trees alive while people continue to live from them. In 1968 the American ecologist Garrett Hardin argued in “The Tragedy of the Commons” that shared resources are always overexploited unless they are privatised or controlled by the state. In Governing the Commons (1990), Elinor Ostrom, who later won the Nobel Prize in economics in 2009, answered him with dozens of real cases where communities managed common resources well for centuries.

So, in a region where the state is often far away and poor, are communities better placed than states to protect the environment, or does real protection need both? I will show first that Nafawol proves that a community can protect its environment better than the state at the local scale; then that the community has real limits, and that the state’s worries and tools are legitimate; and finally that the best protection comes from combining scales, as the novel and the real Sahel both suggest.

I. At the local scale, the community protects better than the state

A. The knowledge of people who live with the result

Nafawol’s first advantage is knowledge. When Dr Arramata washes salt down with her drip lines, it is Kadiatou, the village’s Water-Keeper, who tells her what she has noticed by her own compound wall: “Where I pour the bitter water on my okra … the ground goes white at the edge” (p. 105). When the demonstration harvester overheats, the weaver Bakary saves it with “Wet jute” (p. 97), the method the women use on their dye pots. The novel is precise here: Dr Arramata names it “Evaporative cooling” (p. 97), but the village already knew how to do it.

The techniques used on the floodplain are real ones from the Sahel. The crews build “low stone lines across the dead beds, knee-high, curved against the slope, so that when the rains came the water would slow and sit and sink instead of running off across the crust” (p. 126). These are the cordons pierreux developed on a large scale in the Yatenga in Burkina Faso in the 1980s. “On the higher ground they dug half-moon pits four paces across and filled them with manure and cut stalks” (p. 126): these are the demi-lunes used in Niger and Burkina Faso, close to the zaï of Sawadogo, small planting pits filled with manure. Finally, “the old stumps, no longer cut back for firewood, had sprouted again” (p. 126). This is what agronomists call farmer-managed natural regeneration (FMNR), promoted from 1983 in the Maradi region of Niger by the Australian agronomist Tony Rinaudo, who shared the Right Livelihood Award with Sawadogo in 2018. Instead of planting new trees, farmers protect and prune the living stumps already in their fields. Estimates often quoted say that by the mid-2000s it was practised on about five million hectares in Niger. Part of the explanation, according to Rinaudo and others, is that from the colonial period until the 1990s forestry laws made trees the responsibility of the state, so farmers had little reason to protect them. As a French student I find it uncomfortable that this system began under French colonial rule in the 1930s.

The novel also shows that knowledge belongs to several groups at once. Before the first stone line, Mamadou goes to the herders’ camp with kola, because their cattle “had walked those dead beds since before the river left, and nobody had asked them about stones” (p. 126). The herders send him home, and he goes back twice more before they agree where the cattle will cross. A ministry with a deadline would probably not have done so.

B. Rules that people write for themselves

Ostrom’s main point is that communities succeed when they create clear rules, watch each other and punish cheaters in a graduated way. Nafawol does exactly this. Mamadou asks for a fund built “so that it does not need me to be honest” (p. 128), and the Coffer Fund needs three signatures from three different households for any large sum, a rule Mamadou insists must apply to him as well. The accounts are posted on a wall in the square every Friday, because, as Isatou says, “A pot everyone can see into does not empty in the night” (p. 132). This transparency works. When Lamin, Mamadou’s own grandson, steals money in small payments, it is a child, Mariam, copying a column for her maths class, who asks: “Why does the carter always stop just under the line?” (p. 168). No inspector from Port Walada would have seen it.

Above all, the community limits its own use of the resource, which is the opposite of Hardin’s tragedy. Dr Arramata explains that “The harvester is there so that we take as little of it as we can” (p. 90) from the deep water, which belongs to the grandchildren. When the harvester settles at a little under seventy per cent of its capacity, she says, “It will not get better” and “We stop asking it to. We plan around it” (p. 125). From then on, every Friday, she “lowered a tape into the borehole and wrote the depth beside the curve” (p. 125), and in the hard eighth year, when “Every Friday the figure was lower” (p. 205), the village rations itself: “They opened the taps for two hours after dawn and two before dark” (p. 205). Nobody forced them. The result came in Year 10: the channel water “stood for twelve” days (p. 211), and by Year 15 “the borehole’s line on the wall had stopped going down” (p. 225).

C. A state that does not come

By contrast, the state in Part Two is mostly absent. Dr Arramata measured the aquifer for eleven years and published her results twice, but “Nobody funds a study of a place nobody intends to visit” (p. 89). Promises of new wells were “all of them swallowed by paper in Port Walada” (p. 89), and when the village asks for a road, the government “sent a surveyor who drank our tea and vanished” (p. 111). In its first years the village asks the state for nothing beyond “the water permit a clerk had stamped” (p. 134). The community also has a much longer time horizon than a minister. Asked why their grandfathers built stone lines they never saw pay, an old man answers “For us” (p. 126), and Dr Arramata predicts the channels will run in “ten if they were lucky, and fifteen if they were not” (p. 126). Minister Sulayman, by comparison, has lasted nine years, “which in Tekuru was a long time” (p. 134).

So at the scale of a floodplain, the community seems to do better than the state. But the novel is careful not to make Nafawol a perfect utopia, and the state’s point of view is not simply that of a predator.

II. The limits of the community and the legitimacy of the state

A. Sulayman’s worry is reasonable

When Minister Sulayman tries to take temporary control of the Coffer Fund, the narrator gives him a selfish motive: a fund outside his ministry “was a pool of money that bought him nothing” (p. 134). But his argument in the meeting is serious. In nine years he has seen four community funds grow, and “Two were emptied by the men who ran them” (p. 136). The depositors came to him and he could offer nothing but regret. “You will forgive me if a ledger on a wall does not settle the matter for me” (p. 136). The novel actually proves him partly right, because Nafawol makes mistakes: a payment entered twice by Fatou, a culvert paid with two signatures instead of three, Lamin’s fraud. Communities can also cause harm. The young Guardians, who are supposed to watch and not to chase, run after a figure in the dark, and a herd boy from Sirawol falls and breaks his arm. Alkalo Musa says bitterly: “And you use them to chase our children into the stones” (p. 124). And a community can simply leave: Kerewol repaid its loan, “sold its mill to a company from the coast and left” (p. 225).

B. The tools of the state protect the community too

What is interesting is that several state institutions help Nafawol. To protect the fund from the ministry, the lawyer registers it as a savings society under the banking law, “so that the Ministry could not simply seize it without the regulator” (p. 135). The regulator’s inspector, a woman who arrives by bus, finds the two errors and refuses to sign until they are corrected: “If you let it be two when it is convenient, it will be two when it is not” (p. 138). She is stricter than the village, and she makes the fund stronger. She even admits: “I wish half of them kept books like this” (p. 138). The carbon credits, which pay for trees and soil, can only be sold because “The Republic had authorised the carbon credits earlier that year, through its environment ministry, and asked for nothing but the tax” (p. 175). The village sells carbon on a market that only states can guarantee.

The boiler episode is more ambiguous. An inspector seals Sirawol’s mill because its boiler lacks a certificate from one of three recognised bodies. “This is a good file,” he says. “It is not the certificate” (p. 159). The rule “had never been enforced north of the river” (p. 160), and Fatou suspects that the minister, unable to take the fund, has taken the mill. Kadiatou corrects her: “He has taken the year” (p. 160). Here the state uses a legitimate rule in an unfair way. But a pressure vessel is dangerous, and I think a reader should not forget that safety standards exist because boilers can explode, whoever owns them.

C. Some things are too big for a village

Finally, some environmental and economic questions are beyond the scale of a village. The train that carries the Union’s goods runs on an old freight line “dead for forty years until the national rail authority reopened it for the Union’s trade” (p. 226). Rail transport is much less polluting than lorries on dirt roads, and no village could have built it. The state charges for it, “a levy on the Union’s own freight for thirty years, which Fatou paid each month and called the rent” (p. 226), but it is still a public investment. In the same way, the falling crop price in Year 8 comes from “a subsidy renewed on another continent” (p. 203), and the Teranu family, who are not at fault, cannot use the suspension clause, which only covers acts of the Republic. A village fund cannot protect anyone from the world market. Part One of the novel offers the strongest defence of the state, in the mouth of Minister Dunia in Awdagha: after all the charities had gone home, “The only thing still there, when everything else had gone, was the state” (p. 61). It is true that Dunia’s state also destroyed the Madina plant by taking its money, which shows that a state can be a predator as well as a protector.

The community and the state therefore both have strengths and weaknesses. The real question becomes how to combine them.

III. Protection works best when the scales cooperate

A. Nafawol builds an intermediate scale

What Nafawol finally builds is neither the village alone nor the state. The Senebian Economic Union “was registered under the Republic’s law as a union of village cooperatives” (p. 173): it accepts the law of the state, while governing itself. It grows from four villages to twelve, then thirty-one, with two councils, one for elected representatives and one for elders. Ostrom calls this “nested enterprises”: local rules inside larger institutions. Kadiatou first refuses a charter with Sirawol, saying “Neighbours do not need paper” (p. 156), but after the boiler episode she changes her mind: “A rope of four strands is not four ropes” (p. 165). The traditional authorities also play a part, since the seyfo, the district chief, settles damages when a herd breaks into the cotton: “the seyfo settled the damage” (p. 140). Dr Arramata insists that the most important parts of the model cost nothing: “Any village committee in this country could write them before the next dry season” (p. 173). Fatou adds the condition: “Writing them is free … Keeping them is not” (p. 173).

B. The Great Green Wall: a lesson in scale

The real Sahel offers a comparison. In 2007 the African Union launched the Great Green Wall initiative, a project to restore a belt of land across the continent from Senegal to Djibouti, with eleven countries in its core zone and a target of 100 million hectares of restored land by 2030. At the beginning, it was often imagined as a line of planted trees decided by states. Progress has been slow, and many observers now say that the most successful parts are those where farmers themselves protect trees and build stone lines and pits, as in the Yatenga and in Maradi. In other words, the states set the goal and find the money, but the work succeeds when communities own it. Nafawol follows the same logic: Dr Arramata plants a green belt of neem and gmelina, but the old stumps that sprout again are protected by the villagers; within a season of the first planting, “The dust no longer struck Nafawol from an open plain. It met green first” (p. 108).

C. The same question at home

This question is not only African. Near Toulouse, in the Tarn, the Sivens dam project, which was meant to store water for irrigation and was supported by local and national authorities, provoked a violent conflict; a young protester, Rémi Fraisse, was killed in October 2014, and the project was abandoned in its original form in 2015. In my grandfather’s valley in the Ariège, the summer pastures are shared by several farmers in a groupement pastoral, a form created by the pastoral law of 1972, which is a community institution recognised by the state. Like Nafawol, these examples suggest that people accept rules about water and land better when they help to write them, and that the state works best when it guarantees the rules and leaves the ownership to the people who live there.

In conclusion, the example of Nafawol shows that a community can protect its environment better than a distant state at the local scale, because it knows the land, lives with the results and can watch its own members. Its stone lines, half-moon pits and regenerated stumps are taken from real Sahelian practice, and its ledger on the wall answers Hardin’s pessimism in the way Ostrom described. But the novel also gives the state its due: Sulayman’s fear of emptied funds is justified, the regulator strengthens the Coffer Fund, the environment ministry authorises the carbon credits and the national rail authority reopens a line no village could build. My answer is therefore that communities protect better when the state gives them rules, guarantees and infrastructure, and lets them own what they build. As an ouverture, the novel ends with a problem of scale. At thirty-one villages, the Union already meets in a lecture theatre where Alkalo Yaya “could no longer hear the small villages at the back” (p. 225), and Kadiatou “had not yet found the proverb for thirty-one” (p. 226). The Great Green Wall, with its eleven core countries, faces the same difficulty on a far larger map: how to grow without losing the voice of the people who carry the stones.