Dust, Water, and the Architecture of Consent

1.1 A novel about arrangements

Novels about aid are often praised for putting human faces on a subject usually left to statistics. Development has never been short of faces; its reports are full of them, down to the woman photographed beside her new pump. What it has been short of is a plain account of its own machinery: who signs, who holds the key, who reads the paper aloud, and what happens in the month after the funder’s staff fly home.

Songs of Dust and Water gives that account. Its turning points are administrative. A village elder signs alone under a meeting tree (ch2). A procurement portal answers a request for sisal with “ERROR: VENDOR NOT CERTIFIED” (ch3). An engineer signs an amendment in front of cameras without reading it to the end (ch4). A trader reads a contract aloud under a baobab and is made to read two of its clauses twice (ch11). A bank inspector finds a payment carrying two signatures where the charter asks for three (ch17). A young official, sorry to do it, hangs a lead seal on a boiler (ch20). A student finds a clause deep in an appendix (ch24). A stone drops into a clay pot (ch25). The fire at Madina is the loudest event in the book, and it is the product of smaller ones like these. The question the novel puts to each arrangement is the same: does it still hold when the people inside it are not good, or are no longer there?

This study argues that the novel’s answer can be stated as a single thesis. What decides whether a community keeps the benefit of its own labour is the architecture of consent: who decides, by what rule, recorded where, and whether the rule holds when the good people have gone. The technology a community receives matters less than this. So does the money that pays for it, and so does the goodness of the people who bring it. Every section that follows serves, tests or qualifies that claim.

1.2 The thesis in the novel’s own terms

Part One shows an arrangement in which none of the four questions has a good answer.

Who decides? One man. Mustapha is Madina’s signatory because the Ministry registered him, and when the old farmer Samba raises a grazing dispute three generations old over the very plot the plant will stand on, Mustapha calls the plot state land “under my authority as village signatory” (ch2). Samba’s reply is the first statement of the book’s case: “A signature tells you who held the pen,” he says. “It does not tell you who agreed.” (ch2). Adam Sterling takes the signature because it is quicker.

By what rule? Rules written somewhere else. The warranty holds “only if certified hands do the work” (ch1). The plant buys through a state portal whose approved vendors “did not sell anything that grew” (ch3). When Malik the smith mends a snapped bracket in his own forge, Adam waits a week for a certified one, and Malik draws the conclusion for his apprentice: “A thing you are not allowed to mend is not yours” (ch2).

Recorded where? Inside a sealed control hub and a portal that no villager can read. The only independent record in Part One is the exercise book Aisha begins when Minister Khresay tells her to “Write everything down” (ch5), and she keeps it wrapped in plastic at the bottom of her bag.

Does the rule hold when the good people go? It fails the first test. Khresay is a decent minister, and once the file passes to the Ministry of Energy he has “no hand on it now” (ch5). Adam is sincere, skilled and hard-working, and none of that stops him signing clause 4.B, which lets the Ministry take control of “all associated assets, revenues, accounts and credentials” (ch4). The plant burns. In the second year after the fire, Adam names his error in the one page he sends to the board meeting that hears Proxima’s inquiry: “I believed that if it ran well enough, it would not matter who owned it.” (bridge).

Part Two shows a village building answers, one at a time, and paying for each. The committee keeps the padlock (ch11). The common account goes up on a wall in the square, and Dr Arramata writes the borehole’s depth beside it every Friday (ch16). The charter is redrafted until its lawyer strikes out anything that sounds, in Nabari, like a man hiding something; when he writes disbursement, Isatou tells him, “Then write taking money out.” (ch16). No large sum leaves the fund without three names from three households, and the alkalo counts as one household like any other. The fund is registered under the banking law so that a minister cannot take it with a letter (ch17). A suspension clause separates a village that failed from one that was stopped (ch20). A union of villages gives each member two votes, one of them chosen by its young people, and sets a Council of Elders beside the Union Council that can warn but not forbid (ch21). The largest decision in the book is taken with stones counted in front of everyone (ch25).

The clearest statement of the thesis comes from an old herder. Asked how the fund should be built, Mamadou says: “Then build it so that it does not need me to be honest” (ch16). He wants the rule to bind him in particular, because one day “somebody I love will come to me with a good reason and a sad face. I want the rule to say no to him, so that I do not have to.” (ch16). Five chapters later the person he loves is his grandson Lamin, who has split payments to a carter who owns no cart so that each stays under the three-signature line. Mamadou argues for the boy from the doorway, then sits down in his usual place and asks for his own name to be written under Lamin’s, “as the one who asked for the rule” (ch21). The rule holds against the man who made it. That is the test the novel sets every arrangement, and the one Madina’s never faced.

1.3 Four claims

Four further claims follow from the thesis, and the study returns to each.

The first is that the two parts form a controlled comparison. The novel holds the problem nearly steady and changes the institutions. Both villages are dry and short of water, both receive a machine from the same fund, and the fund commits far more to the second (“a hundred million dollars over ten years to this region”, ch10) than it ever spent on the first. The same engineer carries the money. Even the failure of the machine repeats: the harvester settles “at a little under seventy per cent of its design yield” (ch16) and never recovers, as Madina’s intakes choked on salt and clay. What changes is ownership, where the records are kept, and who may say no. Section 24 follows the pairings in detail.

The second claim concerns narration. Part One gives the reader Adam’s mind: his wish to “seal all of it out of his project” (ch1), his fear of the board, his reasons for signing. Part Two closes that door. From the bridge onwards Adam is seen only from outside, through Arramata, Kadiatou, Hassan and Fatou. He speaks of Madina in dialogue (“I have seen what happens when a system is owned only on paper. I built one.”, ch10), but the narration never again reports what he thinks. The inner lives belong now to the people who in most aid stories would be recipients, such as Kadiatou at the rope and Fatou over her ledger. Talib and Adam never meet again. When Talib comes south in the last chapter he is not named, and the only message bound for Adam is second-hand: “Tell him the weaver’s son has good hands.” (ch28). The novel withholds the reconciliation its plot appears to promise. Section 13 argues that this is how the book makes its argument in the grammar of narration.

The third claim is that the novel keeps the bill. Each answer Nafawol builds costs something, and the cost is named and left standing. A herd boy from Sirawol falls into a gully while two new night-watchmen chase him in the dark, and years later his arm has “healed crooked. He herds with it.” (ch20). The old widow Haja sells her goat so that Sirawol’s women can plant onions (ch20). After the vote, a collapse in world prices brings a year with no surplus; the Teranu family ask for the suspension clause, are refused because a price at a distant port is no act of the Republic, and leave for the coast (ch26). Mariam loses a year of study. The borehole’s line on the wall falls every hot season. Old Jankero, who asked Kadiatou how far the promised hospital would be, dies, and his son tells her: “It is as far as it was.” (ch26). Fatou, who wanted the offer, never changes her mind. The novel does not let the vote cancel these costs. Sections 9 and 25 weigh them.

The fourth claim is that the novel hands its question on. The last chapter ends with a girl of twelve helping her great-grandmother draw a bucket from a well that has, after fifteen years, come right: “The well was kept now by a girl who did not yet know it was hers.” (ch28). Then the coda takes away any comfort that ending might give. It returns to Wagadu, which the griot Balla says was lost four times and four times found, “always found by the people who had lost it” (ch8), and closes on a warning: “One pair of hands can always be made to let go.” (coda). The arrangements of Part Two exist so that no single pair of hands is all that holds. Whether they go on holding is left to Halima’s generation. Sections 15 and 16 read the well and Wagadu at length.

1.4 What the novel concedes

A thesis this clear invites a reading that is too clean, and the novel resists one. The minister who ruins Madina has a case. Dunia takes the plant’s money and lets the plant rot, and she also builds a grid that reaches the north. Her argument to Aisha is that the charities leave and the state stays: “Your Mr Sterling is an island. So is every fund like his. I am trying to build a country.” (ch8). Years later the street lamps of Madina are lit from that grid, and Hawa, who once delivered a child by torchlight, says: “We have not delivered a child by torchlight in years” (ch25). Aisha’s judgement of Dunia is sharper than Adam’s. The Minister never wanted the plant to burn: “Wanting it to burn would at least have been a decision. She never had to make one.” (ch9).

The state in Part Two is no simpler. Minister Sulayman’s distrust of the Coffer Fund rests on experience: of four community funds he has watched, “Two were emptied by the men who ran them.” (ch17). The novel also never pretends that Nafawol managed without outside money. Arramata tells the charter meeting what Proxima paid for: “The money bought steel. It paid for my time here. It bought us room to make mistakes without starving.” (ch21). The rules cost nothing to write, Fatou adds, and a great deal to keep.

So the thesis makes no claim that Nafawol was right and Madina wrong in some final reckoning. Its claim is narrower and harder: the difference between the two villages lies in the architecture of their consent, and that architecture has a price the novel counts to the last goat. The book lets a reader understand Fatou’s white stone and still see why the black ones won.

1.5 The shape of the study

The study has six parts.

• Part I, Foundations, states this argument, sets out the method of reading and gives a full account of the story.
• Part II, The Argument across Disciplines, puts the thesis to the novel from nine directions: philosophy, sociology, anthropology, psychology, politics, economics, technology, faith and the body.
• Part III, How the Novel Makes its Argument, studies its craft, its fictional world, the well at its centre, its use of Wagadu and its subtext.
• Part IV, The Novel in its Fields, tests its plausibility and places it among development paradigms, African thinkers and critics, the African literary tradition, and the utopian and speculative traditions.
• Part V, Commentary, reads the novel passage by passage and then gathers the pairings that join its two halves.
• Part VI, Judgement, states what the novel critiques and argues for, and speaks directly to the student of English literature.

25.1 What the novel critiques

The first target is the aid relationship itself, meaning the habit by which one side gives and decides while the other receives and waits. Adam’s pitch in Bakoro states it in two sentences: “Minister, the people are the beneficiaries,” and then, as a selling point, “Nobody in Madina will ever need to touch it.” (ch1). Khresay hears the flaw at once: “a man who does not help build the granary will never fight to defend it” (ch1). The novel makes the critique harder to dismiss by making Adam good at his work; in the hour before the plant is switched on he reads the commissioning curves for four minutes and sends one man to the right junction box (ch3). The fault lies in where he stands. Dunia, who is neither kind nor disinterested, gives the sharpest account of what aid leaves behind, plaque after plaque in her home district, where “Each one had a ribbon cut, and a photograph, and a report that went home and said success.” (ch8). In Part Two the critique turns on the village’s own daughter. When Dr Arramata proposes that Proxima keep control for ten years, for the village’s protection, Adam tells her what her goodness would do: “I think you’d hold it for them, and be good at it, and the day you left, it would fall down.” (bridge). The relationship survives a change of personnel, and that is the novel’s point about it.

The second target is the sealed system. The hub at Madina is built to keep the world out, and Adam’s creed is that “Inside the fence everything is tested, rated and covered.” (ch3). Talib learns it as a catechism: “Nothing gets in. Nothing stops.” (ch3). The seal is administrative as well as technical. The warranty holds only if certified hands do the work, and the procurement portal admits only three certified vendors, who “did not sell anything that grew” (ch3). The local fix exists, works, and is thrown away, and the novel lets it go on working where no one can use it: “In the scrap store behind them, in the dark, the charcoal in Aisha’s box drew the damp out of the air.” (ch3). Malik draws the moral that the book will spend its second half testing: “A thing you are not allowed to mend is not yours” (ch2). Section 10 follows the engineering in detail. For judgement, what matters is that the novel is candid about its own side. The harvester never rises above seventy per cent, and the mill’s boiler arrives certified from far away: “By the press, the little boiler hissed under its foreign stamp.” (ch18). Sirawol’s boiler carries the same stamp, and the want of a certificate from a recognised body is the lever the state later pulls. The Union objects to sealing and still depends on sealed goods.

The third target is the appointed signatory. Mustapha is registered by the Ministry, and his claim is that “My signature is all that is required.” Samba answers him under the meeting tree: “A signature tells you who held the pen.” (ch2). The novel gives Mustapha a daughter’s wedding, a son’s debts and a leaking roof, and then the line that measures him: “He told himself a number of things, and some of them were true.” (ch4). The office is built to be bought, and a man of ordinary weakness sits in it. Dunia sees this and uses it: “Every village has a Mustapha. A minister, at least, can be removed.” (ch8). Adam is a single signatory too, holding Proxima’s power of attorney alone; “It had seemed a convenience.” (ch4). And Nafawol begins the same way. When Proxima’s agreement is signed, “Nobody asked who should sign,” and Alkalo Sankung signs alone, as alkalos always had (ch11). The three-name rule of chapter 16 is the village’s answer to its own first habit.

The fourth target is the ministry that makes itself the operator. Khresay warns Adam early that “this Ministry stands on fine, loose sand, and the sand is always moving.” (ch1). Dunia’s amendment provides that “the Ministry may assume direct administrative control of the Installation” at its own discretion (ch4), and once the state runs the plant it can also write its history. Adam’s worst discovery at the tribunal is that the facts read out are true: “They had simply started the story where it suited them.” (ch8). In Tekuru the state takes a quieter route. Sulayman asks the Fund’s signatories to sign over “temporary administrative control”, and when Kadiatou asks what supervision means, the last answer she gets is “It means order.” (ch17). When that fails, a safety rule that “had never been enforced north of the river” is enforced once, on Sirawol’s boiler (ch20). Kadiatou names the cost exactly: “He has taken the year.” (ch20). The novel lets the state put its own case. Dunia speaks of a country built against islands of foreign money; Sulayman has watched two community funds emptied by the men who ran them (ch17); and in Madina, years later, Hawa says of Dunia’s grid, “We have not delivered a child by torchlight in years” (ch25). What the novel attacks is the state as owner and operator of what others built and paid for (section 8).

The fifth target is the language of partnership in commercial terms. The word runs through both parts. Adam calls his arrangement a partnership in chapter 1; Dunia asks him, “Surely you trust us as partners, Mr Sterling?” (ch4); and Mark’s message records that “the cameras have you calling it partnership” (ch8). When Evelyn Reed calls the deal “a partnership” (ch23), Fatou corrects the noun: “It would be a sale,” and then, “Call it one, and we can talk about the price.” (ch23). The perpetuity clause is the obvious target, and the novel has Reed strike it out within two days, with a decent letter of apology, so that the reader has to look at what is left. What survives is the veto and the first right to buy: “The Union’s leaders would still sit at the head of the table. TransGlobal would still decide what the table was for.” (ch24). Adam’s summary to Kadiatou is exact: “They have left in the part that matters.” (ch24). The same test is applied to Proxima’s own board, where a contributor in a good suit asks to hold the rights to the Union’s model, and Adam’s reply would serve for TransGlobal too: “If we own the model, every village that uses it is renting it from us.” (ch22).

25.2 What the novel argues for

Against these five, the novel sets four answers.

The first is the rule that binds the people who wrote it. Mamadou asks for the three-name clause because he knows his own weakness: “I want the rule to say no to him, so that I do not have to.” (ch16). He has already asked the committee to “build it so that it does not need me to be honest” (ch16). The regulator’s inspector supplies the principle: “If you let it be two when it is convenient, it will be two when it is not.” (ch17). The rules are written in the language of the people they bind; the lawyer learns to strike out anything that sounds “like a man trying to hide something” (ch16). They can change, but only in public. The suspension clause is written into the charter after Sirawol’s boiler is sealed, and Mamadou’s clause about adding payments across the year follows his own grandson’s theft (ch20, ch21).

The second is the public record. Part One’s records are private and hidden. Khresay tells Aisha to “Keep a copy somewhere they cannot reach.” (ch5), and her exercise book spends the rest of Part One at the bottom of a duffel bag. Part Two puts the record on a wall in the square. Isatou gives the reason: “A pot everyone can see into does not empty in the night.” (ch16). Her list of the women’s cooking is entered “so that the grandsons can read it for themselves” (ch13). It binds the outsider too: Proxima’s papers go up beside the Fund’s, with Adam’s salary on the second page (ch16). The test of the record is what happens when it shows its keeper in the wrong. Fatou finds her own double entry and writes the correction in full; “She did not cross either of them out.” (ch17). The novel’s last chapter carries Aisha’s private book into the academy library at Nafawol, so that a record of failure kept in secret ends up open on a shelf.

The third is ownership understood as the right to correct. Malik’s saying, turned the other way, gives the novel’s definition: a thing you may mend is yours. Proxima’s new rule says the same in a funder’s language: no money “unless the people who would live with the result owned it, and could stop it” (bridge). The seventh clause of Nafawol’s agreement turns it into law: “Nafawol might build from them, mend from them and copy them, and nothing in the paper let Proxima stop it.” (ch11). When Adam hands over the last key, Fatou declines to treat it as a gift: “It was always ours. This is only the receipt.” (ch26). Ownership in this sense also includes the right to leave. Kerewol repays its loan, sells its mill and goes, and “The charter allowed it, and nobody had been able to tell Kerewol it was wrong.” (ch28). Hirschman’s pairing of voice and exit is present in the design, and the novel accepts its price (section 5).

The fourth is the institution that outlasts its founders. Dr Arramata learns the principle from Adam: “Write it so it doesn’t need you. Then it won’t need me either.” (bridge). The mill that “runs on two people” produces the academy (ch19). The founders give up their stones at the vote (ch25). Hassan leaves the technical lead because he is needed: “People had told him he was needed. He had told them that was why he was going.” (ch28). The same change reaches Madina, where “Mustapha’s son kept the village’s papers, as his father had, and read every one aloud under the tree before he signed it.” (ch27). The coda states what all this is for: “One pair of hands can always be made to let go.” No single keeper, however good, is safe enough.

25.3 Where the argument has limits, or leaves questions open

The novel’s case is strongest where it admits what it has not settled. Three open questions stand out, and a fourth group of smaller ones follows them.

The first is Halima. Every office in Part Two is chosen, counted, rotated or bound by rule, down to the clause that no clerk may keep the same book two years running (ch21). The office at the centre of the book is inherited. The title of Yaara-Jaxo passes through the women of Kadiatou’s line, “an inheritance no one boasted of and no one refused” (ch10), and the phrase leaves open what would happen if someone did refuse. Kadiatou releases Mariam from it in her own thoughts, “Then the rope is not for you” (ch21), and the title then moves down a generation to a child who has had no say. At the close she is named keeper by narration: “The well was kept now by a girl who did not yet know it was hers.” (ch28). There is a second strain here. The keeper’s knowledge is the one record in Nafawol that never goes on the wall: Kadiatou “had tasted the water every season, and told nobody” (ch28). One can defend this. By the end nobody drinks from the well, and custodianship of that kind belongs to faith and memory (sections 11 and 15). One can also read it as the place where the novel’s argument stops short. A book that asks every power to show its warrant ends by handing an old power to someone too young to accept or decline it. The novel does not decide between these readings, and an honest reading should not decide for it.

The second is transparency without any way to answer an entry. The wall says what happened. It has no column for the reply. Lamin’s defence, “Each one was small,” and “Nobody went without because of any of them.” (ch21), is spoken in a doorway and recorded nowhere. What follows is handled by custom, through a Tobaski bowl that waits on a shelf through the rains. The buyer from the ginnery is named in two villages’ records on the strength of a boast at a naming ceremony, reported by Musa, and the novel says outright that “there was nothing anyone could prove in a court” (ch20). Within a year he has lost his trade. That is close to what happened to Adam in Bakoro, where he concluded that “There would be no hearing here.” (ch8). The Union’s record is honest where Dunia’s tribunal was not, but neither offers a hearing. Isatou’s experience shows a related gap. Her question about the cooking fires is entered and deferred: “The question was written down for a later meeting.” (ch22). Then it is deferred again, and in the end she has to let the fires go cold to be heard. The one entry that answers for itself is the Teranu family’s, because the eldest son asks for it: “Write on the wall that we paid what we could” (ch26). People in the novel improvise a right of reply where they need one, and the novel shows the need without building the remedy.

The third is the national state, which the novel leaves underdrawn. Tekuru appears as one minister, an inspector or two, clerks at counters and a rail authority. There are no elections, no police and no national politics beyond Sulayman’s career. Yet the Union survives through the state’s institutions. The banking regulator stalls the minister (ch17), the registry stamps the cooperatives, and the rail authority reopens the line and charges a levy for thirty years, “which Fatou paid each month and called the rent” (ch28). The Union grows to thirty-one villages and a university, and its relation to a Republic that “knew of no place called Senebia” (ch21) is never worked out. The novel marks the gap itself. When Fatou notices that the courts named in TransGlobal’s revised contract belong to that Republic, the narration records: “She said so. Nobody answered.” (ch24). Awdagha is left equally unsettled. Dunia’s grid stands and the torch stays in its drawer, while Nabil is in prison “for the biofuel money” (ch25). Whether the state’s building redeems its taking is a question the book raises and leaves alone. The reader is shown a federation of villages beside a state, and is not told what each owes the other, or what the state might do if it set out to fight properly (section 8).

Smaller limits complete the account. The three-name rule counts households, and Isatou observes, “I know whose names the households go by.” (ch18); the question of whose hand receives household money returns with TransGlobal’s fifty thousand and is never settled. The gap between fitter and farmer is still open at the end: “We have not solved it.” (ch28). Reed’s strongest argument, about the next village still drinking bad water, gets no reply: “He did not answer that, and she did not wait for him to.” (ch24). And Part Two rests on a large outside fund, a premise Dr Arramata answers only in part when she separates what the money bought from what it could not (ch21; section 19). The novel names each of these limits in its own scenes. In a book about public records, that may be the most consistent thing it does.

26.1 The novel’s vision

Stated in one sentence, the novel’s vision is of a world in which help is judged by what it leaves behind in rules, records and the power to mend, and in which a community’s keeping of its own gains depends on arrangements that hold after the good people have gone. That vision can be established from the book’s construction without any appeal to its author.

The first evidence is structural. The novel tells its argument twice, once as failure and once as a slow and costly success, and it builds the second half out of answers to the first (section 24). Mustapha’s lone signature is answered by three names from three households. The tribunal that read only the convenient facts is answered by the inspector who reads the whole ledger and makes the village correct two lines. Aisha’s box on the scrap heap is answered by Bakary’s wet jute and the K-Purifier. A reader can trace these pairs without guessing at anyone’s purpose.

The second is where the narration places the reader’s sympathy and attention. In Part Two Adam is seen only from outside, while Kadiatou’s thoughts are given in her own voice, in italics, from the first chapter of the part. The last chapter ends inside her senses, at the rim of the well, and the book’s last words belong to a short coda on Wagadu (section 16). A book that wanted the foreigner to be the hero of its second half would be built differently.

The third is what the novel refuses to resolve. The vote is fourteen to ten, “No one cheered” (ch25), and Kadiatou’s answer to Mariam is that she will not call the vote right or wrong: “I will not tell you it did not. I will not tell you it was wrong either.” (ch26). Fatou, at the end, “is allowed to say it” (ch28). The vision includes disagreement as a permanent feature of a free community. A reader who reports the vote as the book’s verdict on TransGlobal has misread what the book is careful to keep open.

26.2 Intention without a statement from the author

The book carries no preface or afterword in which M.J. Hydara explains it, and this study draws on no interview or essay by the author. A student cannot report what the author meant, and should not write as though she could. That limit is less serious than it sounds, and a good essay can say so in its introduction.

The relevant debate in criticism is long. W.K. Wimsatt and Monroe Beardsley argued in the 1940s that an author’s design is neither available nor a proper standard for judging a work. Roland Barthes went further and moved meaning from the writer to the reader and to the shared conventions of the text. Michel Foucault treated the author as a function that organises a body of writing. E.D. Hirsch pushed back, holding that only the author’s meaning gives interpretation a stable object. Umberto Eco offered a workable middle position: the intention of the work, which can be read off the text and which rules out arbitrary readings without claiming access to anyone’s mind. Section 2 sets out the hermeneutic method this study has followed, and Eco’s middle position is the one a student will find most useful here.

In practice, a few things can be said with confidence. A student can say what the novel repeats, what it pairs, what it withholds and what it rewards. She can say that the note at the front of the book calls Awdagha and Tekuru “two fictional countries”, and can describe what that choice does: it lets the argument be tested on two republics of the novel’s own making, one leaning to French and one to English administration, without being tied to the record of any one real government. She can say that the narration gives Kadiatou an inner voice and denies one to Adam in Part Two. Each of these claims can be checked against the text, and each would remain true whatever the author later said.

Other things cannot be said. A student cannot claim that the author shares Kadiatou’s view of the vote, that the book was written in answer to a particular project or country, or that the author’s name or background gives the novel authority on West African life. Arguments of that kind replace evidence with biography. A safe habit is to make the novel or the narration the subject of every interpretive verb: the novel argues, the narration withholds, the chapter pairs. Where a sentence would have to begin Hydara wanted, it usually belongs in a footnote as a guess, or should be cut.

One further caution applies to this book in particular. Because its argument is clear, it is tempting to treat the characters’ best lines as the author’s opinions. Kadiatou’s “With whose hand on the key” (ch23) is a strong line, and the novel gives Fatou, Reed and Jankero’s sons strong lines against her. A character’s speech is evidence of what the novel puts in play, and the novel’s position has to be read from how such speeches are arranged.

26.3 Implications for readers and practitioners

The novel is fiction, and it proves nothing about the real world. What it can do is run a thought experiment that no evaluation can: hold the dust, the money and the outsider constant, change the arrangements, and show what follows. Read that way, it has implications for anyone who funds, builds or governs projects in other people’s places.

For funders and agencies, the first implication is to ask who holds the key, and to plan the handover from the first day. Proxima’s rule, that money flows only where the people affected own the work and can stop it, would be a demanding standard for any real fund. The novel also shows the cost of keeping it. The rule slows everything down, and when a contributor leaves the following year Adam’s title gains the word deputy (ch22). The second is to put one’s own terms on the wall. Adam’s salary in Nabari, posted in the square, is a small scene with a large point about who is entitled to read whose paperwork (ch16).

For engineers, the implication is to design for repair. The charcoal box, the wet jute and the purifier made from parts “bought nearby or made here” (ch26) belong to what Robert Chambers called putting the last first: building on what local people already know and can maintain. Section 10 sets this beside the literature on appropriate technology.

For people who run community institutions, the implications are the novel’s plainest. Three signatures from three households. A ledger read aloud or posted. Rules written in the language of those they bind. A treasurer who corrects her own error in public. Elinor Ostrom’s design principles for governing common resources are visible in almost every one of these (section 5), and the novel adds what case studies tend to leave out: the feeling of watching a rule you asked for land on your own grandson.

There is a trap here for anyone who reads the novel as a manual. The plan its Union grows from is called the Senebia Blueprint, and the word should make a careful reader uneasy, since Part One shows what happens when a perfect plan is carried in from elsewhere. The novel’s account of how good practice spreads is more modest. The charter goes out free, a village three borders away paints its own ledger on a wall, and when Halima says they copied Nafawol, Kadiatou corrects her: “They wrote their own” (ch28). Nobody from the Union had been there or been asked. James Ferguson’s account of development as a machine that turns political questions into technical ones is a fair warning against turning this novel into a toolkit (section 19). What it offers practitioners is a set of questions to ask of any arrangement, and the expectation that the people who live under the answers will write them.

For readers in general, the novel asks for a change of habit. It invites us to judge an outside actor by what is written and whether it will hold, instead of by sincerity. That is hard to accept for anyone raised on appeals that ask us to trust a face, and the request is all the more striking coming from a novel, a form that usually rewards attention to inner life.

26.4 How to write about the novel

A few practical rules will keep an essay sound. Quote briefly and exactly, and give the chapter every time; the novel’s force often lies in a single line, and paraphrase blunts it. Keep the two parts distinct. Madina is in Awdagha and Nafawol in Tekuru; the village has a committee and the Union a council; Adam in Part Two is seen from outside. Treat Senebia, Nabari, the Coffer Fund and the Union as fictional, and the peoples, the empires and the faith as real. Do not summarise for more than a sentence at a time. An examiner will want to know what a scene argues, and why it is placed where it is.

Four kinds of essay suit the book. A comparative essay might set it beside Bessie Head’s When Rain Clouds Gather, where a cooperative in a dry place meets a sceptical chief and a well-meaning outsider, or beside Ousmane Sembène’s God’s Bits of Wood, where collective action is organised by people who are not its leaders (sections 21 and 22). A formal essay might follow focalisation, and argue that the novel deals with the outsider at the centre through its narration, by deciding whose mind the reader may enter. An interdisciplinary essay might test the novel against Scott, Ostrom or Sen, and ask what fiction can show that a case study cannot. The hardest and best essay takes one of the open questions in section 25.3 and works it through: whether the public ledger is a different kind of thing from Dunia’s portal or the same instrument in other hands, given that “They had made themselves impossible to rob quietly” and also “very easy to study” (ch23); or whether naming Halima keeper is the novel’s last image of continuity or its one unexamined inheritance.

A sound shape for any of these is simple. The introduction states a claim about the work, not the author, and says how it will be tested. Each body paragraph makes one claim, gives a short quotation, and explains what the scene does and why it sits where it does; the pairings across the bridge are good material, because they let one quotation answer another. One paragraph should take the strongest reading against the essay’s own and give it a fair hearing, using the novel’s own dissenters: Fatou on the vote, Reed on the next village, Jankero’s sons on the hospital. The conclusion should return to the claim and say what the essay has shown that a summary could not.

A worked example may help. A weak sentence about the ending would say that Hydara shows Halima will be a good keeper. It claims knowledge of the author’s mind and of a future the book does not show. A stronger sentence would say that the novel names Halima keeper in narration, at twelve, before she knows it, and so places its one inherited office with the person least able to consent to it. That sentence can be checked against chapter 28, and it opens an argument instead of closing one.

Whatever the essay, the best ones stay with the novel’s own method. They state the claim, show the scene, note the cost and leave open what the book leaves open.