Blog

The first images came up on a screen on a boat off the coast of Benin, and the man watching them had spent years of his working life imagining this exact view without ever expecting to see it. Coffi Gerard Franck Zinzindohoue, a researcher at the Institut de Recherches Halieutiques et Oceanologiques du Benin in Cotonou, was looking at a seafloor about 54 metres down that his country’s scientific record described, on the only authority available, as dead. It was not dead. It was a garden. Soft corals stood thick enough to hide the rock they were anchored to, and fish moved between them in the low light, the kind of fish that only bother with a reef if the reef is working.

He has described the feeling since as a mixture of excitement and disbelief, which is an unusually honest thing for a scientist to admit about his own data. The disbelief is the part worth sitting with, because it tells you something about how this reef came to be written off in the first place, and about who has been able to go and look at the West African seabed over the last sixty years, and who has not.

The story starts in the early 1960s with a set of surveys that were never really about corals at all. French researchers working the Gulf of Guinea shelf, led in this stretch by Alain Crosnier, were doing the kind of work that governments paid for at the time: finding out where the fish were, and where a trawl could and could not be dragged. Somewhere off Benin and Togo their instruments kept hitting something hard at depths of roughly 52 to 56 metres, running in a line that looked like a barrier. The historical reports suggested the structure might extend something like 40 kilometres parallel to the coast. They recorded it as a coral reef barrier, noted it, and moved on with the assumption that it was dead.

That assumption was not lazy, exactly. It was the product of what the survey was for. A fisheries survey wants to know whether a given patch of seabed is hard or soft, snagging or clear. It has no particular reason to establish whether the hard thing down there is living tissue or the skeleton of something that died a long time ago, and in the early 1960s, at that depth, there was almost no practical way to find out. Nobody was going to put a diver down to 54 metres on a routine fisheries cruise. There were no drones. There were no cameras that could be lowered with any confidence of coming back with a usable picture. So the reef went into the file as probably dead, and the file sat there.

Then, for about sixty years, nobody went back.

Not nobody in a dramatic sense. Benin has a marine research institute, and that institute has been doing work on its waters throughout. But no coral research targeted the Benin continental shelf after those investigations of 1963 and 1964, which meant that the composition and distribution of whatever was living out there stayed undocumented. A line in an old report hardened into a fact. The country’s only known reef structure became a thing everyone had heard about and nobody had seen.

Part of why has to do with depth, and this is the piece that most people miss. The reef sits in what marine scientists call the mesophotic zone, the band below the bright shallows where light is present but thin. Zinzindohoue describes it as living deeper than the familiar shallow reefs, in a place where light is much more limited, and he makes a point that changes how you should picture it: a mesophotic system does not necessarily form one continuous wall of coral. It can be patchy, scattered, a run of separate communities spread across the seabed rather than a single structure you could swim along.

Fifty metres plus is an awkward depth for science. It is too deep for the easy recreational and research diving that has mapped shallow reefs around the world in exhaustive detail, where a team can be in the water within minutes of anchoring and back up with samples before lunch. Working safely at that depth means technical diving, specialised gas mixtures, decompression planning, a support team and a budget. At the same time it is far too shallow to interest the deep-sea expeditions that bring crewed submersibles and remotely operated vehicles with robotic arms to the continental slope and the abyssal plain. The mesophotic band falls between the two funding worlds, globally, and in a region where marine research money is thin to begin with, falling between two funding worlds means falling out of the schedule entirely.

What finally changed was not the question. The question had been sitting there in plain sight since 1964. What changed was the instrument.

Zinzindohoue’s team ran modern side-scan sonar over the suspect ground, covering a cumulative 11.5 kilometres of transects, which is a small amount of seabed and he says so himself. Sonar at this stage is a filter rather than an answer. It tells you where the hard, structured, reef-shaped things are, and it narrowed the search to two spatially separate zones that carried the right acoustic signature. He has said that the campaign up to this point was not easy, a run of long days at sea, technical problems and a great deal of uncertainty, and that the moment the sonar started throwing up shapes that could plausibly be reef was the moment the thing started to feel possible.

There is a wider reason people have started caring about this depth band in the last decade or so. Shallow reefs are the ones in trouble, bleaching in warm water years and taking the brunt of coastal pollution and physical damage, and researchers have spent a long time asking whether the communities sitting just below them are buffered enough by depth to act as some kind of reserve. The evidence on that is mixed and contested, and nobody serious treats mesophotic reefs as a guaranteed backup copy of the shallows. But it does mean that every one of these systems that gets found and described adds a data point to an argument that matters, and that a stretch of West African shelf with no description at all was a hole in the picture rather than a curiosity.

Then they put eyes down. An underwater drone went in first, followed by the Deep Sea Camera System built by the National Geographic Society’s Exploration Technology Lab, a lowered rig designed to be dropped onto a target and left to film. Suddenly, as he puts it, they could actually see the seafloor.

The camera system is worth a word, because it is the sort of kit that quietly decides which questions get answered. A lowered deep-sea camera rig is cheap compared with a crewed submersible, needs no pilot, and can be worked from a boat that is not a dedicated research vessel. It is the difference between a survey that needs an international expedition and one that a national institute can actually run.

What came back was a dense coral garden dominated by octocorals, the soft corals that bend in the current rather than building the stony architecture people picture when they hear the word reef. The team documented at least six octocoral morphotypes along with two black coral taxa, plus an associated crinoid, a feathery relative of the starfish that perches on high ground and filters the water going past. The corals were growing on sub-angular blocks of rock, consolidated hard substrate that gives a soft coral something to hold onto, which is exactly the kind of snagging, trawl-wrecking surface a fisheries survey would have flagged sixty years earlier without ever asking what was living on it.

And there were fish. Eight species, photographed on the reef and named in the paper: the robust butterflyfish Chaetodon robustus, the Monrovia doctorfish Acanthurus monroviae, the golden African snapper Lutjanus fulgens, the Guinean angelfish Holacanthus africanus, the blackbar soldierfish Myripristis jacobus, the West African goatfish Pseudupeneus prayensis, and two damselfish, Chromis limbata and Chromis cadenati. Several of them were clustered at the base of one of those rocky blocks in a mixed-species group, which is what fish do on a reef that has food and shelter on it.

That is the finding, published earlier this year in Frontiers in Marine Science after review, and the team states it as the first confirmed record of a living mesophotic coral ecosystem on the continental shelf of the Gulf of Guinea.

Now the part that a lot of coverage has skipped, and that Zinzindohoue and his colleagues put in writing themselves.

This is not a reef that came back from the dead. It cannot be, because nobody ever established that it was dead. The paper is explicit that whether the present observations reflect coral communities that have simply persisted down there all along, or changes in conditions since the 1960s, cannot be determined from the data available. Both readings are open. Either the corals were alive in 1964 and were recorded as dead by a survey that had no way of checking, or something died and something else grew. The honest version of this story is narrower than a resurrection and, read properly, more uncomfortable: a reef was written off without being looked at, and then sixty years later somebody looked.

Nor is the reef uniformly thriving. One of the published photographs shows a dead octocoral colony with hydroids and tube-building serpulid worms colonising it after death, which is an ordinary thing to find on any reef but is worth saying out loud in a story that is otherwise about abundance. Things die down there too.

The species names also come with a caveat the team flags clearly: because no physical samples were collected, the identifications remain provisional. Everything was done by camera. Visual identification of octocorals in particular is genuinely hard, and in a region this poorly surveyed there is a real chance that something on those rocks does not yet have a name in the literature at all. The paper is equally careful about the apparent scarcity of some groups, noting that whether that reflects a real ecological pattern or simply the limits of how they sampled remains an open question.

So what you have is not a tidy headline. You have a confirmed living mesophotic coral ecosystem where the record said there was nothing, described from images, with the taxonomy unfinished and the history unresolved. For a first look at a place nobody had looked at in six decades, that is about as much as anyone could expect, and the willingness to say it plainly is part of why the result is credible.

It is worth looking at the author list, because it explains how this got done. Zinzindohoue is the lead author and the survey lead, and he worked with three colleagues from his own institute in Cotonou: Houangninan Midinoudewa, Bola-Nle Christelle Akonde and Zacharie Sohou. Beyond that the names come from the Alfred Wegener Institute in Bremerhaven, GEOMAR in Kiel, the Faculty of Engineering at the University of Porto, and the National Geographic Society in Washington. That is a Beninese-led survey of Beninese waters, running on instruments and expertise assembled from four countries.

The argument underneath that is the one the paper gestures at and Zinzindohoue makes directly. He thinks what they found is probably not an exception. He thinks it is a reflection of how little is still known about coastal waters in West Africa and in plenty of other places, and he puts the question bluntly: if a reef could be found again after sixty years, how many others are sitting undocumented or simply forgotten? He is careful not to oversell it. He does not assume reefs are everywhere, and he says each one has to be confirmed, because without going back into the field everyone is just guessing.

That is the uncomfortable thought in all this. The reason this reef spent sixty years as a probably-dead line in an old report was not that the science was impossible or the question uninteresting. It was that the equipment needed to answer it, and the funding to put that equipment on a boat, were not available to the people whose coastline it was. The gap was never curiosity. It was access.

What Zinzindohoue says he wants next is not a press cycle. It is a proper scientific diving and sampling programme, back at the same site, with divers in the water and physical coral samples coming up, because that is the only way the provisional names become real ones. He notes that the historical reports pointed to a possible barrier running around 40 kilometres along the coast, and his team surveyed a small section of it with a limited number of visual transects. Almost all of it is still unexamined.

The part of this he seems most interested in is not the species count at all. It is the reef’s history, and what a structure like that can tell you about how the ocean in that region has changed over time, the way a stand of old coral holds a record of the water it grew in. He has compared it to opening an archive that had been sitting there for a very long time and reading the first page.

Which is roughly where the matter stands. One researcher, a borrowed camera system, 11.5 kilometres of sonar, and a country that can now say something true about its own reef instead of repeating somebody else’s guess.

The envelope was brown, the ordinary kind a school office keeps in a drawer, and the number written on it was 414,000 rupiah. At the exchange rate of the month she retired, that came to roughly twenty three dollars. It was the last money Atrianil was paid for teaching, and it arrived at the end of forty years at the same private vocational school in West Jakarta.

She is sixty three. Her pupils have called her Bu Ijah for decades, and it was Bu Ijah, not any reporter, who put the envelope on camera. She posted it to her own social account in June, late in the month, a few days after her final working day. Within about a week the clip had been picked up by national media, the teachers’ association had been asked to comment, and Indonesia’s largest daily had run an editorial whose headline was simply the two numbers side by side: 414,000, and forty years.

The first thing worth settling is whether the figure means what it looks like. It does. When the school was asked, it confirmed that 414,000 rupiah was the whole of her monthly pay, not a residual or a part payment. The teaching allowance was inside it. So was the transport money. There was no second envelope.

A second number makes the first one legible. The minimum monthly wage for the province of Jakarta this year is 5,729,876 rupiah, about three hundred and twenty dollars, set by the provincial government and binding on employers of ordinary workers. Her pay was a little over seven percent of it. She was not being underpaid against that floor, because the floor did not apply to her. That is the part of the story that takes some explaining, and it is the reason the story is worth more than a sad number.

Before going further, a word about where all of this comes from, because it matters here more than usual. There is one original. A Kompas reporter interviewed her and the school in the days after the clip spread, and effectively everything that has circulated since, across national outlets, aggregators and regional titles, traces back to that reporting. It reads like eight or ten independent confirmations and it is one. Kompas is not reachable from where this piece was researched, so the figures here were cross checked against pickups that reproduce them consistently, against the school’s confirmation as reported, and against the public response from the teachers’ association, which is independent of the original interview. Her own account includes one further episode from early in her career, concerning money she says she was asked for and refused. It is a striking claim and it appears in exactly one chain of reporting that could not be read at source, so it is left out here rather than repeated on trust. Everything below is either confirmable or flagged as her account.

The category she belonged to is called guru honorer. The English rendering, honorary teacher, is misleading, because nothing about it is ceremonial. An honorary teacher is a working teacher without civil service status. In a state school, the local education office or the school itself hires them outside the national establishment, usually because there are classes to cover and no funded civil service post to cover them with. In a private school, which is where Bu Ijah worked, the employer is a foundation, and the foundation decides what it pays out of whatever fees and donations it collects. There is no national pay scale in that arrangement. There is no accrual of pension rights. There is, in the plainest terms, no mechanism by which forty years of continuous service converts into anything at the end of it except a last envelope.

The scale is not marginal. According to ministry figures drawn from the national education database late last year, Indonesia had about 3.42 million educators, and 21.42 percent of them held non permanent status. That is roughly one in five of the people standing in front of classrooms, and the share is heaviest at primary level. A separate official count puts 237,196 non ASN teachers in state schools as registered and permitted to keep teaching through the end of this year, under a ministry circular issued earlier in 2026. The circular was widely read as a threat and the government insisted it was the opposite, a stabilising measure to stop regional authorities from cutting those teachers loose before replacements existed. Either way, the date at the end of it is real. The civil service law passed in 2023 set a horizon for clearing up non permanent employment, and the end of this year is when that horizon arrives for teachers.

The route out, for those who get one, is a scheme called PPPK, a contract form of government employment that carries a salary scale and benefits without full civil service tenure. Converting honorary teachers into PPPK posts is the government’s main answer, and the ambition is large, more than nine hundred thousand conversions across five years. Here is the mechanical reason it was never going to reach Bu Ijah. PPPK positions are government positions, allocated to government schools. A teacher employed by a private foundation is not competing for a slot that happens to be scarce. She is outside the pool. The education ministry’s own director general for teachers set out that distinction at a parliamentary hearing late last year, and the policy conversation since has been about whether teachers who pass PPPK selection might be posted into private schools, which is a different question entirely, about redistributing existing public employees rather than recruiting the people already doing the work.

So forty years in a private vocational school produced no eligibility for anything. That is not an oversight in her case. It is how the category is built.

Nor is her number an outlier at the bottom of a broadly decent distribution. A survey of teachers carried out in the first week of May 2024 by the research institute IDEAS with the charity Dompet Dhuafa, covering 403 teachers across twenty five provinces, found that 20.5 percent of honorary teachers were paid under 500,000 rupiah a month, around twenty eight dollars, and that 74.3 percent were paid under two million rupiah, around a hundred and twelve dollars. Fewer than one in a hundred cleared five million. The sample is modest and the figures are two years old, so they should be read as an indication rather than a census. The indication is that a fifth of the honorary workforce lives in the band Bu Ijah’s envelope sits in.

There is a detail in her last year that says more about the arrangement than any of the figures. By her own account, she had cut her teaching down to two days a week by the end, in order to reach the forty year mark at all. Forty years was not a term of employment she was serving out. It was a number she went after, on a timetable she had to shrink to afford, at a school that was paying her less per month than a single day of minimum wage work in her own city would have earned.

The pay also sat behind a gate she was never going to get through. The large allowances in the system, the professional allowance of two million rupiah a month and the remote region allowance of the same size, go to certified teachers, and certification runs through a professional training programme that more than seven hundred and fifty thousand teachers passed through across the last two academic years. The basic incentive, the smallest of the three, is the one aimed at the uncertified. An uncertified teacher at a private foundation school is therefore at the bottom of both ladders at once, too far outside the public establishment to be converted and too far from certification to be topped up.

What that produces, across the workforce rather than in one case, is a teaching profession that largely does something else as well. The same survey of 403 teachers found 55.8 percent holding a second job. Private tutoring was the commonest, which at least keeps them in their own trade, followed by trading, farming, manual labour, making online content and driving motorcycle taxis. The researchers also recorded what teachers pawn when something goes wrong in a given month, and the list ran from gold wedding gifts to motorbikes. None of that is about Bu Ijah specifically, and there is no reporting on whether she took outside work. It is about the shape of the thing she spent forty years inside.

The teachers’ association that responded to her case put harder cases on the record. Its national coordinator, Satriwan Salim, called the situation deeply sad and then named two he said his organisation had documented: a part time contract teacher in Dompu, in the eastern islands, paid 135,000 rupiah a month, about seven dollars and fifty cents, and another in West Java paid 50,000 rupiah, under three dollars. His association’s demand is narrow and unglamorous, which is usually a sign it is serious: a minimum wage floor for non civil service teachers, pegged to the regional labour minimum the way it is for every other kind of worker. It has also gone to the constitutional court over this year’s state budget law. The budget in question runs to about 769 trillion rupiah for education, something in the order of forty three billion dollars, which is the figure that makes the envelope hard to argue away.

There is a particular arithmetic in the government’s own relief scheme that lands without any need to raise the volume. From this year, non civil service teachers who qualify receive a monthly incentive raised from 300,000 rupiah to 400,000 rupiah, about twenty two dollars, paid to 377,143 recipients at a cost of around 1.8 trillion rupiah. Certified teachers get a professional allowance of two million rupiah a month, and teachers in remote regions get a special allowance of the same size, with the whole package passing fourteen trillion rupiah. The state’s top up for an underpaid teacher, in other words, is almost exactly the size of Bu Ijah’s entire salary. Whether she ever drew it is not documented anywhere that could be verified, and given the eligibility routes run through state school registration, it would be unsafe to assume she did.

It is tempting, at this point, to make her into an argument, and she should not be one. The honest shape of her forty years is not devotion to a wage. She has said, in the reporting that exists, that she stayed because of the pupils and because leaving felt like abandoning something she had built, and she has also been plain that the money was an injury rather than a sacrifice she was content with. By her own account she raised the practical grievance of an age limit on a Jakarta support payment for educators, which cut off at sixty and therefore cut off her. She has objected to the word honorer itself being thrown around in public as a label for a kind of person. Those are the complaints of someone who thought she was owed better, not of someone who had made peace with less. Treating her as a parable of cheerful endurance would get her wrong, and it would also let the system off.

Which brings this to an ending that is not satisfying, and should not be dressed up. Nothing happened to her. The clip spread, the association spoke, a national newspaper wrote its editorial, and there is no documented outcome for Atrianil: no reinstatement on better terms, no pension granted, no public fund, no official intervention on her behalf that can be sourced. It is possible something private was done and was never reported. It is not possible to claim it. Three months on, the only verifiable facts of her retirement are the ones she put up herself.

What she actually did, in the last week of a forty year career, was to document it. She did not write a manifesto or address a rally. She held up the envelope with the number on it, the one piece of paper in the whole story that nobody has disputed, and let the arithmetic do the work. That turned out to be the most effective thing anyone involved has done. The association’s wage floor has not been legislated. The deadline at the end of this year is still a deadline, with 237,196 people in state schools on the near side of it and an unknown larger number in private schools who are not even counted in that figure. The category that produced her envelope is still producing envelopes.

She taught in the same place for four decades and walked out with twenty three dollars and no pension. The number is not a scandal because it is small. It is a scandal because it is accurate, because the school was right that it was the total, and because it was arrived at legally, by a system that knows exactly how many people are inside it.

Donato Garcia Juan de Dios learned to read and write on a tablet computer at the age of seventy, while he was being treated for cancer, and when he stood up at a theatre in Lima to collect the certificate that said he had completed his basic education, he told the room something very simple. He said he had the strength to finish what he had started. Then he and the rest of his class were walked outside, put on an open-top tourist bus, and driven around the capital.

That is the part of the day that is hardest to explain to anyone who has not been in that position. The certificates were the achievement. The bus was the thing nobody had thought to ask for. Eighty people, the youngest in their twenties and the oldest in their eighties, were driven past the archaeological mound at Huaca Pucllana, along the clifftop viewpoints above the Pacific, through the parks, in a city many of them had lived in or travelled to for years without ever once seeing it from the top deck of anything.

The theatre was the Teatro NOS in San Isidro. The ceremony was called Reescribiendo Historias, which translates roughly as Rewriting Stories, and it was run by two private organisations rather than by a government ministry: a Peruvian business movement called Es Hoy, which works on education among other things, and Fundacion Dispurse, a foundation that runs literacy circles and what it calls open community schools. The graduates had come from eight regions, including Cajamarca, Cusco, Pasco, Huanuco, Ancash, La Libertad and Piura, as well as Lima itself.

Before going any further, there is something about the reporting on this day that has to be said out loud rather than buried, because it is the sort of thing that produces a wrong story if nobody checks it.

Two separate counts were published about this ceremony within twenty four hours of each other, and they are not the same number counting the same people. One account, from the Peruvian state news agency, dated the ceremony to the seventeenth and reported that more than six hundred users of Pension 65, the state pension programme for older adults run by the Ministry of Development and Social Inclusion, had completed their basic education through digital literacy over the course of the year, and that the programme was honoured at the event. The other account, published the following day by a national newspaper, described eighty young people and adults graduating through the adult education programme run by Es Hoy and through Dispurse’s own literacy and community schools, followed by the bus tour.

Six hundred and eighty are not two versions of one figure. The six hundred is a national annual total for a state programme. The eighty is the group that graduated through two private routes and climbed onto the bus. They were honoured in the same room on the same day, which is why both numbers attach to one ceremony, and anyone who adds them together or swaps one for the other has written a different article about a different thing. Donato Garcia is the point where the two accounts touch: he appears in both, named and aged seventy in each, which is how it became possible to be confident they describe one event rather than two.

It also means something about how he learned. The state strand teaches reading and writing to older adults through an application called Focus, installed on tablets supplied by local governments. That is the tablet. A man of seventy, in oncological treatment, learning letters by swiping at a screen, in a programme that reached more than six hundred people like him in a single year.

What makes the eighty worth two thousand words is not that they are inspiring, a word that gets used to flatten people. It is that each of them failed to finish school for a completely different structural reason, and you can see the whole shape of a national problem by lining up four of them side by side.

Epifania Juanita Huaman Guevara is fifty eight. By the account published about her, she learned to read in order to manage her own accounts and to support her children. That is the least sentimental motive in the group and the most revealing one. Illiteracy is not primarily experienced as a sadness about books. It is experienced as a daily, grinding vulnerability at the point of transaction: not being able to check a figure someone else has written down for you, not being able to read what you are signing, having to take another person’s word for the arithmetic of your own livelihood. People who learn to read late very often say some version of what she said. They did it to stop being at a disadvantage in rooms where numbers were being written down.

Luciana Flores Lopez is sixty, a Quechua speaking artisan from Cusco, and the reason she did not go to school is that there was no school. Not a school she was kept out of, not a school she dropped out of, simply no school in the community she grew up in. That is a geography problem rather than a willingness problem, and it is the explanation behind a large share of Peru’s older rural illiteracy. It is also why a Quechua speaker arriving at adult education is not starting from the same line as a Spanish speaking city dweller: she is being asked to become literate in a second language as well as literate at all.

Lidia Mendoza Pacaya is forty four and went back to secondary school with the backing of her employer, finishing earlier in the year. The detail that matters in her case came afterwards. Her daughter resumed her own studies. That is the mechanism nobody budgets for when they fund adult education, and it is the one with the longest tail. A parent who goes back makes going back thinkable for the child who had quietly stopped.

And Donato, who is the reason the story has a seventy year old at the centre of it rather than a statistic.

The scale of the thing they stepped out of is genuinely large. Peru has more than eight million people aged fifteen and over who have not completed secondary education, which is close to thirty percent of that age group. Around fifty five percent of them are women, roughly four and a half million. Something like fifty seven percent are adults between thirty and fifty nine, meaning this is overwhelmingly not a story about teenagers who recently dropped out but about working adults in the middle of their lives. About fifteen percent of the group cannot read or write at all. Around a quarter never completed sixth grade.

Where you live changes almost everything. In rural areas roughly fifty seven percent of the population carries this educational lag. In urban areas it is about seventeen percent. In Huanuco, Cajamarca, Loreto and Ayacucho, more than forty percent of the population over seventeen has not finished school.

Now the part that explains why a business movement, rather than a ministry, was handing out certificates in a theatre. Peru does have a public route for this. It is called alternative basic education, delivered through roughly fourteen hundred centres nationwide. Enrolment in those centres sits at around a hundred and ninety thousand people. Against a target population of eight million, that is less than three percent. The route exists. It is simply not reaching the people it was built for, which is a different and more awkward failure than not existing.

Es Hoy’s answer to that is a programme called EPA, short for education for adults, and the design is worth understanding because it is unusually practical. EPA does not try to recruit individuals off the street. It offers free support to companies so that those companies can help their own workers, their workers’ relatives and their suppliers finish basic education. In other words it uses the workplace as the point of contact, on the reasoning that if someone has a job, their employer already knows where they are every day, which is the single hardest problem in adult education. By the account published late last year, that approach had carried more than nine hundred young adults through to completion across around seventy companies. An education projects manager at the organisation, Jessica Guimac, made the economic case plainly: the lower the level of education, the higher the probability of working informally.

There is a second edge to that design which the programme itself does not advertise, and it is worth naming. Routing adult education through employers means an adult’s chance of finishing school can depend on whether the company they happen to work for has signed up. Lidia Mendoza Pacaya finished secondary school with her employer’s backing, and that is genuinely good news for her, but it also means the opposite case exists: the identical worker at the firm down the road, with the same gap in her schooling and the same determination, gets nothing. A workplace is an efficient place to find people. It is not a fair way to decide who is reachable.

Dispurse comes at the same gap from the opposite direction, from the countryside rather than the payroll. The foundation runs literacy circles and reading spaces, and its open community schools are built around a specific and clever premise: that adults who have spent decades acquiring real knowledge outside a classroom should be able to have that knowledge validated and converted into official recognition within the alternative basic education system. It works mainly in Cajamarca and Cusco, which are two of the highest-lag regions in the country, and it has also worked in Bolivia and developed its digital literacy application in Sweden. Earlier this year it expanded an agreement with Peru’s education ministry in Cajamarca, and it had already been recognised for its digital literacy work with older adults alongside the state pension programme the year before.

That validation idea deserves a moment, because it is the least obvious and possibly the most useful thing in the whole arrangement. An artisan of sixty who has spent four decades costing materials, pricing work, managing orders and running a household has an enormous amount of applied knowledge. What she has never had is a piece of paper saying so. The conventional response is to sit her down at the beginning of a syllabus designed for eleven year olds. The alternative is to assess what she already knows, convert it into formal credit, and teach only what is actually missing, which for most adults in her position is literacy itself plus whatever the certificate formally requires. It is faster, it is less humiliating, and it treats a late learner as someone with a gap rather than someone with a deficit.

The tablets matter for a similar reason. Handing an older adult in a rural district a printed primer asks them to begin with the exact skill they do not have. A device that can be practised on alone, as often as needed, with nobody watching the ninth attempt at the same letter, removes the single biggest obstacle to adult literacy, which is not difficulty but embarrassment. Grown adults will not practise reading badly in front of other grown adults. They will practise in front of a screen.

So the ceremony was not a one-off act of charity. It was the visible end of three separate pipelines, one state and two private, converging on a stage in San Isidro.

It is worth being careful about Donato Garcia at this point, out of straightforward respect. He was seventy and in cancer treatment at the time of the ceremony, he was present, he collected his certificate, and he spoke. Beyond that date there is no further published information about his health that could be found, which means nothing should be asserted about where he is now, in either direction. What is documented is a man who, by his own account, decided that a diagnosis was not a reason to abandon a goal he had carried for most of his life, and who finished the thing.

What stays with you about the whole day, though, is the bus.

If you wanted to design a gesture that understood exactly what being illiterate does to a person, you could hardly do better. Illiteracy is not only about words on a page. It is about the quiet daily decision to stay inside a small, known circuit of streets, because the risk of being lost somewhere unfamiliar, unable to read a sign or a timetable or ask for help from a written map, is not worth taking. It shrinks the size of the world a person is willing to move through. People who cannot read often know their own city only as a series of memorised routes.

The certificate changes the paperwork. The bus changes the circuit. An open-top tour of Lima, taking in a fifteen-hundred-year-old adobe pyramid in the middle of a modern district, the long line of viewpoints above the ocean, the parks, is a city handed back to eighty people as something to look at rather than something to get through. The oldest of them was in her eighties. Some of them had come in from Cusco, from Piura, from Huanuco. For a number of them it was presumably the first time they had seen the capital of their own country from above the traffic.

That is where the day ended, and it is a better ending than the stage. A seventy year old man with cancer who learned his letters on a borrowed tablet, a fifty eight year old woman who learned to read so nobody could fiddle her figures again, a Quechua speaking artisan who never had a school to attend, and a mother of forty four whose daughter went back to studying because she did, all sitting on the top deck of a tourist bus in a city that had always been there.

Nobody on that bus is a symbol. They are eighty people who finished something late, and then got taken sightseeing.

The swim is 3.8 kilometres in open sea off Jurere Internacional, on the northern edge of Florianopolis. Then 180 kilometres on a bike. Then a full marathon, 42.2 kilometres, on legs that have already done the first two. Everyone gets seventeen hours. There is no separate clock.

Gabriel Ferrari did all of it towing an inflatable float through the swell, then pedalling an adapted tricycle, then pushing a racing chair with something close to 60 kilograms of weight in it for the length of a marathon. About eighteen months before he started, he could not swim at all.

That is the part of this story that is easy to skip past, and it is the part that actually took the work. Ferrari was already a runner. Running he had. Swimming he did not have, in any form, and cycling he had only just picked up. So the project was not a father deciding to do something hard. It was a man in his thirties walking into a pool to learn the single discipline in endurance sport that cannot be faked, and then doing it again with a boat tied to him.

In the racing chair, and in the float, and on the back of the tricycle, was his son Lucca, who is six.

Lucca was born early. He spent 46 days in intensive care. During that period he suffered a cardiorespiratory arrest, and the family says he has cerebral palsy. That is as much as the family has put in public about the medical side, and it is as much as belongs here. What matters for the rest of this is simpler: Lucca is six, he goes where his father goes, and on race day several thousand people along a coast road knew his name.

Ferrari has a line he uses when people congratulate him, and he uses it so consistently that it reads less like modesty and more like a correction he has had to make a lot. He jokes that he is not the one pushing Lucca. Lucca, he says, carries him the whole way.

Take the swim first, because it is where the learning curve was steepest. Open water swimming is not pool swimming with a view. You cannot see the bottom, you cannot touch a wall, the horizon moves, and you have to lift your head every few strokes to find out whether you are still going in the right direction. Adults who learn to swim late almost always learn it in a pool, where the water is flat and the end is visible. The sea takes that away.

Now attach a float. A towed load does not sit still behind a swimmer. It yaws. It catches chop. Every time you accelerate it resists, and every time you glide it pulls your hips sideways. Swimmers talk about drag as if it were a constant and it is not; towing is an argument you have with the water several thousand times. Learning to swim badly takes a month. Learning to swim well enough to tow for nearly four kilometres in the sea, having started from nothing, is most of a year of mornings.

The bike is a different problem. An adapted tricycle is heavier than a road bike and steers nothing like one. A standard racing bicycle leans into a corner because the rider does; a trike with a load does not lean, it resists, and the rider has to take the whole corner on steering input and judgement. On a climb you cannot stand up out of the saddle, which is how cyclists give their legs a change of angle and recruit different muscle. Everything is done seated, at whatever cadence the gearing allows, for 180 kilometres.

Then the marathon. The running chair is the part of the kit that looks most ordinary and is probably the most brutal, because the figure that matters is weight. By one account the chair with Lucca in it came to somewhere around 60 to 64 kilograms. That is a load the size of an adult, on three wheels, pushed uphill and braked downhill, by a man who at that point has been going since early morning.

There is a fourth leg nobody counts, and it is the one that explains why this format is so rare. An Ironman has two transitions. For an ordinary competitor they are a matter of pulling off a wetsuit, jamming on a helmet and running out with a bike, and the fast ones do it in a couple of minutes. For a pair racing adapted, each transition means getting a child out of one piece of equipment and securely into another, correctly positioned, strapped, padded and protected from the sun, while your own hands are shaking from exertion and your own heart rate is still somewhere near the top of its range. The clock does not pause for it. Whatever care that takes comes straight out of the seventeen hours.

Ironman events do have a division for athletes who compete in adapted formats, and Ferrari and Lucca were classified in it rather than running as a side exhibition. That distinction is worth noticing. There is a long tradition in endurance sport of letting a pair like this go off the front or the back of the field with a nice photograph at the end and no place in the results, and the effect, however kindly intended, is that they were never actually in the race. A classified division means the pair are in the standings, measured against the same cut off as everyone else.

And the cut off is the real test. Seventeen hours is seventeen hours. Nobody gets extra time for towing. Which means that the moment Ferrari entered, the arithmetic became unforgiving in a very specific way. A pair moving at towed pace in the water and loaded pace on the bike has to bank time it does not have, and then hold a marathon pace that would be respectable unloaded.

Work the arithmetic and you can see the shape of the day. Four kilometres of towed sea swimming is not going to be quick, and a loaded tricycle over 180 kilometres of coastal road with real elevation on it is slower again, at a cadence you cannot vary much. Which leaves the marathon, the discipline where the load is heaviest and the legs are already finished, as the leg where the day is either saved or lost. Most competitors in an Ironman walk some of the marathon. A man pushing a chair who walks too much of it does not finish inside the limit. Ferrari had to arrive at the start of the run with enough margin to keep moving, and enough left in him to keep pushing uphill, after more than ten hours of work.

They made it. Published accounts put the finish somewhere past fifteen hours, with one report describing it as around sixteen; the individual official result could not be retrieved, and the honest statement is that they came in inside the seventeen hour limit and were classified. They crossed together, which was the entire point of the exercise.

The course itself is not a gentle one. The swim is a genuine sea swim rather than a sheltered lake, in a bay on the north of the island where the water has somewhere to come from. The bike leg runs out along the coast and back with rolling elevation rather than a flat loop, which for a seated rider on a heavy trike means the hills come at you repeatedly and you absorb all of them in one position. The marathon is run in the afternoon and evening, which on that coast means heat for the first half and darkness for the last of it.

This was the twenty fourth running of Ironman Brasil, which has been held in Florianopolis for most of its existence and which brought in a field reported at roughly 2,000 athletes from 26 countries. At the front there was prize money and a handful of world championship slots. Professional triathletes were done and showered and eating by mid afternoon. Hours later the pair were still out on the marathon course, which is the part of a long distance race that the cameras have usually left.

Then there is the question this kind of story normally walks straight past, so let us not. What does a six year old consent to?

It is a fair question and it does not have a tidy answer. A six year old cannot evaluate a seventeen hour endurance event, and nobody should pretend otherwise. What can be said is this. The family has not treated the race as a private matter it then had to be asked about. Ferrari talks about it publicly. He accepted a public award for it, at a ceremony where the citation named Lucca and described the distance. The decision to put this in front of people was the family’s own, taken in the open. And adapted racing with a young child is not an improvised stunt but an established format with its own equipment industry, its own division at major events and its own long running internal argument about exactly this question, which the people in it have had far more thoroughly than any outside observer is about to resolve in an afternoon.

The useful thing about that award is not sentiment. It is verification.

Ten weeks after the race, the Instituto Guga Kuerten, a long established Santa Catarina foundation set up by the retired tennis player, gave Ferrari its Special Prize at a ceremony in Florianopolis. That matters because the institute is not the family and not the race organiser. Its citation independently records the race, the distance, the three pieces of adapted equipment and Lucca by name. For a story that otherwise rests largely on one interview with the father, having an outside body with no stake in the telling put the same facts on its own record is the difference between an account and a corroborated account.

The company Ferrari was in that evening is also worth knowing, because it places the family inside something rather than making them a novelty. The overcoming prize went to a three time world champion in para jiu jitsu who rebuilt his life through sport. A social project award went to a padel programme for people with Down syndrome, run by a trainer in a coastal town. A volunteer award went to four women who provide sign language interpretation for deaf women during pregnancy and childbirth. An education award went to two schoolteachers using Minecraft to teach neurodivergent students. Five entries, five different problems, five sets of people who built something because the existing version did not work for someone they knew.

Seen in that room, Ferrari is not an inspirational outlier. He is one of a category of people who looked at a thing their child was excluded from and set about engineering a way in.

There is also a practical point that the romance of these stories tends to erase. Nobody does this alone. An adapted entry needs the equipment to exist, which means somebody specified and built a float rig, a tricycle and a running chair that fit one particular six year old. It needs a race organisation willing to accommodate the format on a closed course. And it needs people on the day handling the things a solo athlete handles for himself, from food to sunscreen to the hundred small checks on a passenger who cannot adjust his own position. The reachable reporting does not name who built the kit or who crewed, which is a genuine gap in the record rather than evidence that there was nobody.

What he had to learn, in the end, was not really three sports. It was a fourth skill that has no governing body, which is how to move a load a long way without wrecking yourself or your passenger. Every triathlete learns pacing. Ferrari had to learn pacing for two bodies, one of which is not generating any of the power and all of which has to be comfortable, fed, shaded and safe for most of a day. There is no coaching manual for that. There is the trike, the chair, the float, and whatever you worked out on the last long Saturday.

It is worth saying plainly that this did not come out of a sporting background, because that is the claim that makes the rest of it make sense. By his own account, reported in the one long interview the story rests on, he was a runner who could not swim and had only recently learned to ride, and he gave himself roughly a year and a half of serious preparation. That is a short runway. Most people who finish an Ironman at all have been swimming since childhood. Learning two of the three disciplines from scratch as an adult, while also learning to do them with equipment that makes each one harder, is the kind of timeline a coach would talk somebody out of.

He has said, in the form that gets reported, that he never wanted to make history. He wanted to change his son’s story. It is a neat line, and the reason it holds up is that it describes the sequence correctly. The history was incidental. The thing he set out to alter was the list of places a six year old gets taken.

Which is why the swim is the detail to end on, rather than the finish line. A finish line is a photograph. The pool was a decision, made by a grown man who could not float, repeated at an unsociable hour for most of a year, with nobody watching and nothing yet to show for it. The sea off Jurere does not care who you are doing it for. It only asks whether you learned.

He learned. Then he went and got the boat.

Six of the trees were more than 140 centimetres across the trunk, which in New South Wales is the measurement that makes a tree legally untouchable. Three more were hollow bearing, the kind of old timber that gliders and owls live inside because nothing younger has holes big enough. Conservationists put their ages at between 300 and 500 years. All nine came down in the middle of 2020, in Wild Cattle Creek State Forest inland from Coffs Harbour, felled by crews working for the Forestry Corporation of New South Wales, a company owned by the state itself.

What happened six years later is the part worth sitting with. In March of this year the Land and Environment Court convicted the Forestry Corporation of four offences and ordered it to pay 450,000 Australian dollars, roughly 312,000 US dollars at current rates. The money does not go to the court. It does not go into the state’s general revenue, which is where a fine normally lands and where it stops being traceable. It goes directly to the Yurruungga Aboriginal Corporation, a Gumbaynggirr organisation based in Bellingen, with a list of things it has to be spent on.

That is a different shape of penalty, and the people involved have been careful about saying exactly how different.

Start with how a state agency ends up cutting inside its own exclusion rules, because it is less dramatic than it sounds and more worrying. Under the approval that governs coastal forestry operations in the region, a tree over 140 centimetres in diameter has to be left standing. There are two species exempt from that rule, blackbutt and alpine ash. Everything else over the threshold stays. The rule is simple. Measuring is where it falls apart.

The judgment found that in late 2019 an officer from the state Environment Protection Authority sat in on Forestry Corporation training that contained incorrect guidance on how to measure a giant tree. The next year, crews in Wild Cattle Creek used that incorrect method to decide which trees were protected, and nine trees that should have been flagged and left were not. Tony Chappel, the chief executive of the Environment Protection Authority, acknowledged afterwards that the offence may well have been avoided if his own officer had spotted the error and said something at the time.

So the chain was not a rogue contractor. It was a training document, a classroom, a regulator in the room who did not catch it, and then a work crew with a tape measure applying the rules as they had been taught. Nobody set out to take a 400 year old tree. The system simply did not have a step in it where someone checked.

It is also worth being precise about what was standing there, because nine trees is not nine units of timber. The area sits inside the footprint of the proposed Great Koala National Park, around 315,000 hectares of forest on the mid north coast. The hollow bearing trees were habitat for hollow dependent species including the greater glider. A tree that has been alive for four or five centuries is the only thing in a forest that offers a cavity of a certain size, and nothing growing now will offer one for another two hundred years. You cannot plant your way out of that gap.

Measuring a standing tree sounds trivial and is not. Diameter rules are written as a measurement at a set height above the ground, which means the person holding the tape has to agree with the rulebook about where the ground is. On a slope, on a buttressed trunk, on a tree with a swelling or a burl at chest height, two honest people with the same tape get different numbers. That is precisely the kind of detail a training document is supposed to settle, and precisely the kind of detail that, settled wrongly, propagates into every tree a crew looks at for the next season.

For the Gumbaynggirr people, whose country runs from the Nambucca River to the Clarence River and from the Great Dividing Range to the sea, the trees were doing other work as well. In evidence before the court they were described as habitat for totem species, as ancestral connection points, and as navigation markers, the landmarks that tell you where you are in a forest that otherwise looks the same in every direction. Remove the markers and you have not only lost nine trees. You have made a piece of country harder to read.

Dean Kelly, of the Budaabang clan and chief executive of Yurruungga, gave evidence about going out to look at the stumps and the distress that caused him. His organisation is not a legal shell set up for the case. It manages the Bellinger Valley Indigenous Protected Area, runs cultural burning and restoration work, and delivers ceremonies and community programs out of an office on a main street in Bellingen.

Dr Chels Marshall, a traditional owner, knowledge holder and cultural ecologist, put the loss to the court as death by a thousand cuts, the point being that no single felling ever looks like the one that matters. What goes with each of them is intergenerational knowledge, the things you can only teach someone while standing in front of the thing itself.

There is a timing problem built into all of this that no penalty can fix. A hollow large enough for a greater glider does not form because a tree is big. It forms because a tree has been alive long enough to be injured, to rot in places, to lose limbs and heal badly. That process takes well over a century, and in many eucalypt forests considerably more. Which means the replacement for a 400 year old hollow bearing tree is not a sapling. It is time, measured in multiple human lifetimes, and the only way to have one in two hundred years is to be protecting a middle aged tree today.

The six years between the chainsaws and the sentence were not empty. There was an unusual procedural fight along the way. In 2024 Justice Nicola Pain allowed the Bellingen Environment Centre to take part in the sentencing hearings as amicus curiae, a friend of the court, so that it could make submissions on how serious the offence was and convey community anger about it. That decision was itself written up in a law journal, where the author argued that framing the arrangement as restorative justice was a stretch, since restorative justice is supposed to rest on an offender choosing to take part rather than being made to sit through submissions by a third party.

That critique matters for how we read what happened next, because it is the difference between a label and a mechanism. In April of last year there was an actual restorative justice conference, with Gumbaynggirr elders and Yurruungga on one side and the Forestry Corporation on the other. The Forestry Corporation took part. Out of that came the list of things the community said would help, and that list is what the court later turned into orders.

So here is what is genuinely new, stated carefully rather than asserted. Restorative justice is not a novelty in Australian environmental law. New South Wales courts have run conferences in environmental prosecutions before, and the sentencing provisions that let a court order an offender to fund environmental work rather than simply pay a fine have existed for years. What had not happened before is this: the Forestry Corporation, an organisation with 14 prior criminal convictions behind it, had never been through a restorative justice process at all, and had never been ordered to make financial restitution specifically for cultural harm. The harm being compensated is not timber value and not ecological damage measured in hectares. It is the damage to a people’s relationship with a place, and the court priced it and assigned the payment to the people who hold that relationship.

The structural difference is in where the money sits afterwards. A fine goes to consolidated revenue. It is a real penalty, it hurts the balance sheet, and it also disappears. Nothing connects the dollars to the forest they came from. Dean Kelly made exactly that point after the judgment, noting that penalties in cases like this had rarely returned to country, and rarely to the people carrying the loss. A payment to a community body with a defined worklist is traceable. You can go and look at whether the work happened.

The worklist is specific. The money funds a healing ceremony. It funds repair plans for the sites where the trees stood and ecological assessments to go with them. It funds cultural mapping of the area. It funds a Gumbaynggirr forestry working group, and it funds training local rangers as Gumbaynggirr Guardians. It also covers the unglamorous costs that community work actually runs on, transport, venues and catering, which are the line items that quietly kill projects funded on paper.

The court ordered more than money. The Forestry Corporation has to engage in genuine and meaningful consultation with Gumbaynggirr people about future harvesting in the area. It has to hire an independent consultant to audit its resourcing, its training, its procedures and its policies for identifying and retaining giant and hollow bearing trees, which goes straight back to the 2019 classroom. And it has to publish notice of the offences and the sentence in national, state, local and industry publications, which is a penalty of a different kind for an organisation that would rather this were a technical matter.

Three days after the judgment the Forestry Corporation issued a statement. Its chief executive, Anshul Chaudhary, apologised for removing nine trees that were required to be protected and said the organisation had taken significant steps to stop it happening again. The statement lists them. Thirty six additional trees have been permanently protected on expert ecological advice, a replacement ratio of four to one. Twelve nest boxes have gone in. More than 400 hollow bearing trees already identified in the operational area have been retained. Standard operating procedures and staff training have been reviewed and rewritten so the rules are communicated clearly. The harvested area, the statement notes, has regenerated.

How much is an apology worth next to a court order? Some of it is worth quite a lot, in the sense that nest boxes and 36 protected trees are physical objects in a forest and will be there whether anyone is watching or not. But the timing tells you something too. The apology came after six years of litigation, after a conviction on four counts, and after the court had already decided where the money was going. Richard Kelly, a Gumbaynggirr elder and Dean Kelly’s uncle, had been arguing for years that compensation paid to community should replace fines in cases of cultural damage. He was making that argument long before a court made it an order.

There is a bill attached to all of this that nobody set out to run up. The Nature Conservation Council of New South Wales put the total cost to taxpayers at more than 1.18 million Australian dollars, around 817,000 US dollars, once legal fees, the audit, the publication orders and the restorative justice process itself are added to the 450,000. And this was not the first time. In 2022 the same organisation paid 285,600 Australian dollars, close to 198,000 US dollars, in fines and costs for logging in a koala high use area and rainforest in the same forest.

The question everyone involved is now circling is whether this becomes a template. Jonathon Captain Webb, First Nations director at the Environmental Defenders Office, called the decision significant and rare and said it was one of the first times restorative justice had been applied in an environmental context, with the court effectively accepting that a fine was not the most appropriate way to deliver justice. Dr John Corkill, a member of the Bellingen Environment Centre who appeared in the case, was blunter about where it leads, warning that the Forestry Corporation has a long history of breaking the law and predicting further claims from First Nations groups now that one has succeeded.

Whether that happens depends on things no single judgment controls. Another group would need a corporation capable of receiving and spending the money, evidence of cultural harm specific enough for a court to price, and an offender willing to sit in a room. Yurruungga had all three. Most of the hard work in this case was done years before the hearing, by people building an organisation that could stand up in court and say what had been lost.

It also helps that Yurruungga was not starting from nothing on the practical side. An organisation that already manages an Indigenous Protected Area, runs cultural burns and employs people to be out on country has the one thing a court order cannot create, which is the ability to actually do the work. Money arriving at a body with no field capacity would have funded consultants. Money arriving here funds rangers who were already walking the same ground.

Which brings it back to the mapping, because that is the part of the list that outlasts the payment. A healing ceremony happens once. A repair plan runs its course. A cultural map is a record of what is standing in a particular patch of forest and what it means, written down before anyone arrives with a measuring tape and a work order. It is the step the system did not have in 2019. If it exists next time, a crew cannot mismeasure a tree nobody told them about, because somebody will already have told them.

Nine trees that stood for five centuries came down because of a mistake in a training document. The money is going to pay for the thing that would have prevented it.