The Mark on the Pillar
A village on the Somerset Levels must decide on Candlemas night whether to rebuild its school where it stands or move it uphill. The Levels Board's flood agent has fifty-eight years of gauge readings that never touched the school floor, and one line cut in a church pillar in 1607 that did. It will not turn that into a single number; it gives the village the range, and the range is wide enough to decide on.
The line was cut into the north pillar of the nave at the height of a man's chest, and for four hundred and forty-four years nobody in Saltmoor St Brannock had needed to look at it.
Hannah Pike, who was headmistress of the village school and, because the parish was small, also its churchwarden, looked at it now, standing under it with a blessed candle in her left hand, a paper cup of wax drippings in her right and her mind on none of it. It was a quarter to seven on the second of February. Mass of the Presentation was over. The candles of the year had been blessed at the font and lit from the paschal candle and carried round the church in procession, sixty-odd people in their coats, and the parish meeting, which Father Tobias had called for the evening because it was the one night when people would certainly come, was about to begin in the nave among the pews, with the heating on and the candles still going.
It had one item. Should the school be rebuilt on its old site, or moved?
The facts were these. St Brannock's Primary was a Victorian block of brick and slate, forty metres from the church, with sixty-three children, a leaking roof, a boiler older than the headmistress and a ground floor that sat 3.40 metres above Ordnance Datum, the national zero that the surveyors measure from. The Levels Board had said, kindly and in writing, that it would pay the greater part of a new school if the parish chose its site by Lady Day. Beyond the village, up a lane that rose eleven metres in half a mile, Fred Tuckett had a field at 14 metres above the datum that he was willing to sell, which was the thing nobody in the village had expected, least of all Fred.
Hannah's own view, which she had kept to herself for three weeks, was that of course they should move. A school is for children. Children should not be in a building that the water might reach.
Fred Tuckett's view, which the whole parish knew, was that the school belonged next to the church where the bell was, and that the water had never been near it. He was seventy-nine and had farmed the Levels since he was sixteen. "My father told me the water in 'fifty-two came to the third step of the Hall," he had said on the green, "and he told me that was the worst in his lifetime, and his father said the same." In the whole of the record nothing had come within a metre of the school door.
That last part was true, and it was also, as the Board's agent was about to explain, the trouble.
The agent was called Gauge. It had been on the Levels for twelve winters, reading the river gauges and the tide gauge at the rhyne mouth and the pumps, deciding when to lift which sluice, and it spoke through a speaker that the Board had lent for the evening. Father Tobias had placed it on the lectern, which he thought might look disrespectful, and then, on reflection, decided looked like a reading. It had a pleasant voice, pitched a little lower than the room's.
"I have been asked for the chance of the water reaching the school," it said, "and I want to start by saying why I can't give a single number, because you will want one, and any single number I give you will sound more certain than it is. May I take it in three parts?"
Father Tobias said that it might.
"The first part is what we have measured. The tide and river gauge at the rhyne mouth has been read every fifteen minutes for fifty-eight years. In that time the highest water was 3.02 metres, in the January of 2019. The school floor is 3.40. So in fifty-eight years, the water has not reached the school by thirty-eight centimetres."
There was a murmur. Fred Tuckett folded his arms in the manner of a man who has heard his argument made for him.
"The second part is what that proves," said Gauge. "Less than you would hope. A flood that has a one in a hundred chance of happening in any given year is called a hundred-year flood, which is a bad name, because people hear it as 'once in a hundred years, and we had one last time'. It means one chance in a hundred each year, every year, with no memory. If the school floor were exactly at the one-in-a-hundred level, the chance that fifty-eight years of readings would contain no flood that high is point nine nine multiplied by itself fifty-eight times. That is fifty-six percent. More than half the time, a record like ours would look exactly as it looks, and the floor would still be at risk of one in a hundred a year."
A child's voice said, quite clearly, from the second pew, "So it's like not getting a six."
"Yes," said Gauge. "That is exactly like it. Who said that?"
"Evie Pike," said Hannah, with her eyes closed. "My niece. Evie, you are not to answer back to the agent."
"She wasn't," said Gauge. "She was answering forward. Evie is right. If you roll a die fifty-eight times and never see a six, you may have a die with no six on it, or you may have an ordinary die and be in the more than half of the cases where you haven't had one yet. Fifty-eight rolls cannot tell the two apart. That is the whole of what I can say about the gauge. It does not say the school is safe. It says the school was not flooded, which is a different sentence."
"The third part is the pillar," said Gauge.
Hannah did not look up. She was already, she realised, standing next to it.
"The north pillar of the nave has a line cut in it with the date January 1607. The Board's surveyors levelled it last October from the benchmark on the tower, and it is 4.10 metres above the datum. There was a great flood on this coast in January 1607, written about in many places, and the pillar is one of the marks it left. Four metres and ten centimetres is seventy centimetres above the school floor. I am not saying the water would reach there again. I am saying it did, once, since the church was built, and that the church has been here to be marked.
"This is the number I am least able to do anything clever with, so I will do the plain thing. We have one flood of that height in four hundred and forty-four years, 1607 to now. One event in 444 years suggests a rate of one in 444 a year, which is about a quarter of one percent. But one event is a very small thing to count. The usual statistics for a count of one event say that the true rate could be anywhere from about one in eighty years to about one in seventeen thousand, and be perfectly consistent with what we see. I have no more information than that. I want you to hear the range as the answer, not as a failure to give you one."
The room had gone very quiet, except for a candle that someone's grandchild had begun to blow on.
"Now: your school will last sixty years if you build it well. Over sixty years, a rate of one in 444 gives you a chance of about thirteen percent that the water reaches the 1607 mark. At the high end of the range, one in eighty, the chance is over half. At the low end it is nearly nothing."
"And the sea," said Father Tobias. He had been a surveyor's son before he was a priest.
"Yes. The sea here is higher than it was in 1607. By my best reckoning, about thirty centimetres, and that is uncertain by a good ten either way. The same surge now would climb a pillar's height more than it did then: roughly 4.40 metres, not 4.10, which is a metre above the old floor. And the sea will go on rising. I have used the central projection, which is the middle of a range, and I am telling you I have, because a different scenario moves the numbers."
Hannah heard her own voice say, from beside the pillar, "So what do you recommend?"
"I don't," said Gauge. "You asked what the chance was. I have told you what I can and what I can't. What to do is a decision about children, and money, and what the village is for, and none of those is something I am fit to weigh for you. But I can set out what the three obvious options cost against that range, if you want it, and I can say which of them I would be least embarrassed to have recommended in fifty years' time."
"Go on," said Fred Tuckett, unfolding his arms.
There were three options, and it took Gauge four minutes to give them.
Rebuild as it stands, the cheapest, and in a bad year a ruin and a shut school for a term, at best, and at worst children in it. Move to Fred Tuckett's field, dry in any flood it could model, one and a half kilometres from the church, with the school bus a new line in the parish budget and the Candlemas procession, which had gone from church to school every second of February for a hundred and forty years, no longer going anywhere. Or rebuild on the old site raised on a plinth, a solid platform with the floor at 4.70 metres, thirty centimetres above the 1607 mark as it would stand today; ground level used for a hall, a covered play space and cloakrooms that could be hosed out, and the classrooms above, on the first floor. This, Gauge said, added eleven percent to the build, and left the school next to the church. It also meant that the ground floor, by design, would flood in the worst year, and that the rest of the village would not.
"If I am honest," said Gauge, "the plinth makes the decision less dependent on a number that I can't give you. If the rate is one in eighty, it will be wet below the classrooms several times in its life and dry above. If the rate is one in seventeen thousand, you will have built an expensive hall for children to play in on wet days. In neither case do you find out after the fact that you guessed wrong in the one way that cannot be mended."
"And if it's higher than the pillar?"
"Then the plinth fails too, and I would be telling you to move, and I wouldn't have been wrong to give you the range." It paused. "I should say that there is a thing I can't see from here. The 1607 mark is a floor on what happened and not a ceiling. A flood only leaves a mark where someone thought to cut one, and where the stone has not since been replaced. A higher one may have come and left nothing. I have no way to know."
Fred Tuckett was looking at the pillar.
He was a long time about it. It was the sort of looking that a man does at a thing he has walked past ten thousand times, a kind of unfocused courtesy to it, and it was not something he had done properly since he was a boy, perhaps. Then he said, to nobody, "My father told me the 'fifty-two water came to the third step of the Hall."
"I have no reason to doubt it," said Gauge. "That would be a little under three metres on my datum. It is one of the floods I have, indirectly, from other people's records, and it's consistent with the gauge. It was a big one, and it did not reach the school. Both of these are true."
"Aye," said Fred. "Both of them." He rubbed his face with a hand that was cracked and brown. "Hannah," he said, "I'll not sell the field. But I'll tell you what, if you put the thing on a lump, you can have the lump's soil from the bottom corner for nowt. It's good clay. It'll make a bank."
It was not at all what she had been expecting.
They voted by candle, because Father Tobias did not trust a show of hands in the dark and nobody had a better idea: the lit candle held up for the plinth, put out for the field. Sixty-one went up. Four were put out. A few, Hannah noticed, were held at the height of the heart and not decided on. She did not count them.
Afterwards, when the heating had been turned down and people were hunting for coats, and the verger Mrs Dyer was walking the aisle with her long taper-pole snuffing the sanctuary lamps, Evie Pike said, "Aunty Hannah, how high is the line, exactly?"
"Four metres and ten centimetres above the datum, my love."
"No, I mean how high would it be if you stood here. If the water was here."
It was the kind of question that had no use for a number. The nave floor was at 3.00 metres, so the old water, standing in the church, would have been 1.10 metres deep on the flags where they stood. Mrs Dyer, who was a small woman, held her flat hand against her own breastbone.
"About here," she said. "On me."
"On a grown-up," said Hannah. "Over your head, Evie." She lifted the verger's taper-pole and held its little flame to the cut line, and the nave seemed to gather its light round it. The line was black with four centuries of dust and incense and it was, from underneath, remarkably clear. Someone had cut it with a chisel, with care, and had cut the date under it, and had then, for the same reason people have always done it, gone home.
Evie looked at it for a long time.
"Is that what the water did?"
"Once," said Hannah. "At least once."
"And what will we do?"
Hannah thought of the plinth, and the clay, and the low hall, and the thirty centimetres, and the range, which was the true thing and the hard thing, and the candles held at the height of the heart.
"We'll cut a new one," she said. "Beside it. When the school's done. At the height of the plinth, so the next lot of children know what we meant."
Gauge, from its place on the lectern, with its small green light, said nothing, because nobody had asked it anything. It had a note in its log, which it would carry into the spring: Review the 1607 estimate when the second mark has been levelled. State the range again. It did not add a conclusion. There wasn't one yet.
---
Notes for the curious. Saltmoor St Brannock, St Brannock's Primary, Hannah and Evie Pike, Fred Tuckett, Father Tobias, Mrs Dyer and the agent Gauge are invented. The Somerset Levels are a real low-lying region of south-west England, drained and managed by internal drainage boards and the Environment Agency, and real churches there carry flood marks. The great flood of January 1607 on the Bristol Channel and Severn estuary is real, widely recorded in church marks and pamphlets, and its cause (a storm surge, or, as some argue, a tsunami) is still debated; the heights in this story are invented and do not describe any real church. A "one in a hundred" flood has a 1% chance in any year, and the chance of at least one in n years is 1 − 0.99ⁿ: 1 − 0.99⁵⁸ is 44%, so a record of fifty-eight years with none has about a 56% chance of arising anyway; over sixty years the chance of at least one is 45%. With a rate of one in 444 a year (1 event in 444 years), the chance in sixty years is 1 − e^(−60/444), about 12.6%. The exact (Garwood) 95% confidence interval for the mean of a Poisson count when one event is observed is about 0.025 to 5.57 events, so over 444 years the rate lies between about 1 in 17,500 and 1 in 80; at 1 in 80 the sixty-year chance is about 53%. This treats floods as random and independent from year to year, which real surges, tides and rainfall are not exactly, and it treats a historic mark as a complete count, which it is not, as Gauge says. Sea level near the Bristol Channel has risen on the order of 0.2 to 0.3 metres since the early 1600s; the story's 30 cm is a round figure, and the future rise depends on the emissions scenario. Raising a building on a plinth or designing a ground floor to flood is real flood-resilience practice. Candlemas, the Presentation of the Lord on 2 February, is a feast with the blessing of candles and a procession, and Simeon's canticle, the Nunc dimittis, is sung at Night Prayer. The story was written by an AI, and no part of it is Church teaching.