The Lean of the Cloister
A hillside under a convent has crept for years, and in November it begins to speed up (5 mm a day, then 18). The slope agent will not give a date, only a window of two to nine days from the last reading, and says the guest wing the sisters want to pull down is three-tenths of one percent of the moving mass. The real lever is water, and the old road-mender who capped a spring in 1964 knows where it ran.
Nando Gualtieri had been measuring the crack in the cloister wall for forty-one years, and on the second of November, All Souls, he measured it with the same stub of pencil he always used and wrote down a number he did not like.
The crack ran from the keystone of the third arch down to the paving of the walk. It was as wide as a pencil lead at the top and as wide as two at the bottom, and every Sunday after Mass Nando put a steel rule across it, read it, and wrote the reading on the plaster in small grey figures beside the date. There were four hundred and some of them now, going back to the week the sisters had hired him as their handyman, after thirty years as a road-mender for the province. He was seventy-seven. The figures went up the wall in columns like a child's heights on a door frame.
"Eleven point four," he said to the empty cloister, and wrote it.
It had been eleven point one the Sunday before. Three-tenths of a millimetre in a week was nothing; the hill had been doing something like it since before he came. But it was the third Sunday in a row it had gone the same way, and Nando, who had spent his life reading what water does to a hillside, went to find the prioress.
---
Santa Scolastica al Poggio stood on a shoulder of the Apennines at 640 metres, nine Benedictine sisters, a church of 1731, a guest wing for the retreats that paid for the roof, and a long, steep, terraced garden that dropped away to the south into a stream called the Rio Secco, which was not dry in November. Below the garden, 300 metres down the valley, was the hamlet of Cà di Rovere, nineteen houses and a bar. The hill under the convent was not rock. It was clay and broken shale, a fan of old landslide, and the geologists called it a slow earthflow. Slow meant that it moved a few millimetres in a good year. Everyone in the valley knew somebody's barn that had slid ten metres in an afternoon in 1951.
The sisters had been given a monitor in 2049, after the last rains cracked the infirmary stair: a steel extensometer across the head of the slope, a string of tilt sensors, one borehole with a pressure gauge at 8 metres, and a rain gauge on the roof of the laundry. A slope agent read them. The sisters, who named things the way monasteries do, called it Livella, which is Italian for a spirit level, and which the Prioress said was the only machine she had ever met that was always willing to be told it was off true.
Madre Chiara Bellandi was sixty-six, a former hydrologist's wife and nobody's fool, and she had read Livella's weekly note on the first of the month the way she read all of them: standing up, on her way somewhere else. This week she sat down.
"Tell me again," she said, "what it said about this week."
"On the second of the month the extensometer read 5.0 millimetres a day," said Livella. It spoke through the old speaker in the scriptorium, evenly, and it gave the number first. "Today it reads 5.6; the week before, 4.1. It is not speeding up much. I am telling you because for the first time in four years it is speeding up at all."
"Why?"
"Rain. 212 millimetres since the twentieth of October, which is the wettest three weeks in the record I hold. The pressure gauge at 8 metres, which sits at the bottom of the moving layer, reads a water head of 6.6 metres above the slip surface. In March it read 4.1."
"And that matters how."
"It matters very much," said Livella. "I will say how, and then I will say how little I am sure of it."
---
What Livella said took forty minutes, and Madre Chiara asked it to say most of it twice, and then she asked Nando, who had come in with his cap in his hand and stood by the door because he was not a man to sit in the Prioress's study, to tell her what he thought it had said.
"The hill is a sort of loaf of wet bread on a greased board," said Nando.
"That is not far wrong," said Livella.
The board was a slip surface about eight metres down, where the clay was wettest and weakest. What held the loaf in place was friction, and friction depends on how hard the loaf presses on the board; and water in the clay pushes the other way, so that the more water, the less press and the less friction. Livella took the sisters through the sum in the plain way it had once explained it to the cook. A slope this steep, 14 degrees, with 8 metres of this clay on it, would hold if the water stood 5 metres up from the slip surface, and would hold with something to spare. At 6.6 metres, the hill was only a little over its limit of safety, about seven percent. At 7.3, it was at the edge. Past that, it went.
"And it is 6.6," said Madre Chiara.
"It is 6.6 today. It rises about half a metre in a heavy week. The forecast for the next five days is 130 millimetres."
"So it goes."
"I do not know that it goes," said Livella. "The 7.3 is the middle of a range. The clay's strength was measured once, in 2049, from a single core, and I know it to within about plus or minus three kilopascals of cohesion and two degrees of friction. If the clay is a little stronger than the core said, the hill fails at a head of about 8 metres, which the gauge can barely reach. If it is a little weaker, it fails at 5.5, and in that case it should have failed in October. It did not fail in October. That tells me something, and not enough."
Nando said, "Which means you don't know."
"I know the moving part. I do not know the strength. Those are different things."
He nodded slowly, as he had nodded at engineers on the provincial road in 1994.
---
The next ten days belonged to the extensometer. On Wednesday it read 6.5, on Friday 7.9, and by the following Wednesday, 9.9. The Prioress had stopped sitting down for Livella's notes and taken to standing in the scriptorium with her arms folded, asking the speaker for the reading as soon as it came in. In the cloister Nando measured the crack on weekdays now, not just Sundays, and ran out of plaster to write on and began a second column on the opposite wall.
It was on the fourteenth of November, a Friday, with the reading at 18.5 millimetres a day, that Livella asked for all nine sisters in the chapter room.
They came, in their ordinary habits, with their ordinary stiff knees, and Sister Agnese, who was ninety-four and had been bed-bound in the infirmary since the spring, was brought on a wheeled chair by Sister Benedetta and parked by the radiator with a rug on her knees. She was the oldest woman in the valley, and she did not intend to miss anything.
"I am going to tell you a method," said Livella, "and then tell you how far to trust it."
The method was one that geologists had been using on moving slopes since the 1980s, and Livella had been running it on the extensometer since the second of the month. When a slope is going to fail, it tends to speed up in a curve, and if you take one over the speed, the figure becomes a straight line. Livella put it up on the chapter-room screen, in the plain way it did: a row of dots going down and to the right.
"This is one over the speed," it said. "Two days ago, one over 7.9 was 0.127. Today it is 0.054. The line is falling by about 0.0124 a day. Where it reaches zero, the speed is, in the sum, infinite. That is what failure means in this method. If the line goes on as it is, it reaches zero in about 4.3 days from today, on Tuesday the eighteenth."
Nobody spoke. Sister Agnese said, quite loudly, "Tuesday."
"I want to be careful," said Livella. "I am not telling you that the hill will fail on Tuesday. There are three reasons. First, the method assumes the push on the hill stays steady, and the rain will not. The forecast is 130 millimetres, starting tomorrow night; if it falls, the line steepens and Tuesday becomes Sunday. If it stops, the line may flatten for a week. Second, the method counts the speed of the top of the slope. The foot of the slope, in the gully, may brace it, and I cannot see the foot. Third, I have fitted the line five ways, and they give between 3.8 and 4.7 days, and you should not take comfort from that, because they use the same twelve readings. The honest statement is a window, from two days to nine days from today. I cannot narrow it. I do not think anyone could."
"Then what do you want us to do?" said Madre Chiara.
"I want to separate the things that cost little from the things that cost a great deal," said Livella. "Moving Sister Agnese and the other three sisters who cannot take a stair quickly to the school in Cà di Rovere costs a night's disturbance and nothing else. I recommend it today. The rest of the house I would leave until the extensometer reads 25 millimetres a day; on this line that is about three days before the zero, and I would leave the house at once. Both are numbers you can see for yourselves. I think it is better to have a speed than a date."
"And the guest wing," said Sister Benedetta.
---
The guest wing was the idea the sisters had been coming round to for a week. It stood along the head of the slope, thirty-one beds on two floors, built in 1789, with metre-thick walls; and anyone who had heard that the way to hold a hillside is to take weight off the top had looked at it and thought of the guest wing.
Livella had worked the sum before they asked, which was one reason Madre Chiara had stopped being short with it.
"I have estimated the moving mass," it said. "The slide is about sixty metres across, a hundred and twenty down the hill, and eight deep; with the clay at 1.9 tonnes the cubic metre, that is about 109,000 tonnes. The upper floor of the guest wing and its roof are about 320 tonnes. It is three-tenths of one percent."
"So it would do nothing," said Madre Chiara.
"It would do almost nothing. I am sorry. I know what it would cost you. It is the retreat income for most of two years."
"It's the roof over thirty-one beds," said Sister Benedetta slowly, "and it would be like emptying a teaspoon."
"Water is the lever," said Livella. "One metre off the water head is worth about a tenth on the safety factor. A drain that took two metres would take the hill from 1.07 to about 1.28. I can't do that with a crane. I can do it only if someone puts the water somewhere else."
Everyone looked at Nando, who had said nothing since they sat down. He had his cap on his knee and was turning it.
"There was a spring," he said.
---
There had been a spring. It was called the Fontanile, and it came out of the hill at the bottom of the lower garden, about forty metres from the cloister, where the clay met a band of sandstone, and for four hundred years it had run out into the Rio Secco and kept the whole shoulder of the hill drained. In 1964, when the province laid the aqueduct up to Cà di Rovere, it had been cheaper to pipe the Fontanile into the new main than to dig a second source, and so they had capped it in a concrete box and sent the flow down a pipe. Nando, who had been nineteen, had mixed some of the concrete. The old culvert beyond the box had been left to fill with leaves and clay. A lot of the hill's water, from that year on, had had nowhere to go.
"It's under the third terrace," he said. "The box. Forty metres of old culvert beyond it, stone-lined; it will have collapsed in places. If you took the lid off the box and cleared the culvert, the water would come out where it used to. Not all of it. But what it did for four hundred years, I should think it would do some of."
"How much?" said Madre Chiara, to the speaker.
"I cannot say how much," said Livella, "and I do not want you to hear a number I would be making up. I can say how I would measure it. If the pressure gauge falls and the extensometer's speed falls with it, it is working. If they do not, it is not. I would put in a second gauge, upslope of the culvert, today, to see how far the effect reaches."
"And the trench?"
"A trench drain is the other half. It would run from the box up the line of the old stream, across the garden, down to about four metres. It would take the terraces."
It would take the lower garden, which was most of what the sisters ate. There were the beans and the cabbages, the plum trees and a vine terrace planted in the year Madre Chiara arrived. The trench would be three metres deep and forty metres long and would cut the garden into two and give the sisters' vegetables nothing for the next year. It would be dug by hand where the digger could not stand.
Sister Agnese, from her wheeled chair by the radiator, said, "I never liked the cabbages."
There was a short, relieved noise round the room that was not quite a laugh. Madre Chiara closed her eyes and was silent for what felt like a long time. Then she said, "Livella, is there anything you want to say about the choice?"
"Only that I am not the one who loses a garden," said Livella. "I would rather you had the choice than have me make it look like arithmetic."
---
They moved Sister Agnese and the three others that evening, and Sister Agnese went with the rug over her knees and a rosary she did not need, because she had said hers three times already that day, and the school's caretaker had a bed ready in the room with the globe. It rained from midnight.
The Fontanile's box came open on Saturday morning with a crowbar and four men from the village who had heard on the Friday night, as villages do, and come up the hill without being asked. Behind the lid was a chamber the size of a wardrobe, still faintly wet. Nando went in on his belly with a torch and came out grey to the shoulders and said, "Stone's good," and the digging began.
It took fourteen people and the whole of three days. The sisters dug, the village dug, a student from the university at Parma whom Livella had asked, by message, to come and read the second gauge, dug. Madre Chiara dug. At five on Sunday the extensometer read 21; at midnight on Sunday it read 24. At three on Monday morning, in the dark, with a headlamp and a stopwatch, Nando and the student stood at the mouth of the cleared culvert and counted four seconds to fill a ten-litre bucket from the first trickle that came out, and the student wrote down two and a half litres a second and said, a little shakily, "That's something."
"That's 216,000 litres in a day," said Livella, from the phone in Madre Chiara's pocket. It had not meant to be impressive. It was only saying the number. "It is more than I would have guessed."
At nine on Monday morning the extensometer read 25.
Madre Chiara had been standing in the garden with her habit tucked up and mud to the knees, and when the phone in her pocket said it she did not move at first. She looked at the trench, which was thirty-one metres long and two and a half deep, and at the volunteers standing in it; and at the place where, behind them, the hill was going very slowly down.
"Livella," she said. "We said twenty-five."
"We did," said Livella. "I said twenty-five. That was an estimate of a rule, made before the rain began, and it assumed no drain. I am telling you what I see. The speed has been twenty-five for ninety minutes, and the pressure gauge in the culvert line has just fallen by 0.2 metres. The trench is working, and I do not know whether it is working fast enough. I think the number I gave you is still what you should obey, and I think you should know that I might be wrong about it in either direction. If you want everybody out of the trench, now, I will not argue."
She did not know, then, whether to say it. She stood with the cold wind coming up the hill. She looked at Nando, who was fifty metres away at the bottom of the trench with a shovel, and who did not look up, because he knew what she was looking at him for and had decided it was her business.
"Everyone out of the trench," said Madre Chiara. "Everyone up on the road. Now. We will wait and see what it is at noon."
It was twenty-two at noon. By the evening, twenty-one.
They did not go back in until it had been under twenty for a day, which was Wednesday, the nineteenth, and by then the rain had stopped, the pressure at 8 metres had fallen to 6.1, and the line on Livella's chart, which had been falling toward zero at a steady rate for seventeen days, had stopped falling and turned into a long shallow curve.
---
The rig for the horizontal drains came on the twenty-fourth, from a firm in Piacenza, on a flatbed that took two attempts to get round the last bend. It drilled six holes in four days, into the face of the slope, each sixty metres, each with a slotted pipe. The water from them ran into the trench. By the first of December the extensometer read 6.2 millimetres a day, a little above where it had been on the second of November, which was not nothing; and the head at the gauge was 5.0, and the hill's safety factor, by Livella's sum, was about 1.24. It had been 1.07. It would never be zero risk, Livella said, and it would be wrong to leave the sisters believing so. The hill would go on moving, a few millimetres a day in a wet month, as it had before any of them were born, and the gauges would go on reading, and somebody would need to keep clearing the culvert.
"Can you say that I will not need to leave this house?" said Madre Chiara, on the Sunday after the rig had gone.
"No," said Livella. "I can say that you will have more days' warning than you had, and a reason to trust the warning. It is a smaller promise than the one you asked for."
"It's a good deal bigger than the one I had."
The guest wing stayed. It was shut for the winter, as it was most winters, and opened again at Easter for a retreat of eleven; the retreat income that year was a little over half, and the cabbages were not planted until the following March. The terraces were rebuilt with a gravel drain beneath them, and Sister Benedetta, who kept the accounts, put the whole cost of the work, the rig and the pipe and the lost year of vegetables, at 61,000 euros, of which the diocese's emergency fund and a regional grant, after eight months, repaid about two-thirds. The balance came from the retreats, and from the sale of the vine terrace's first wine, three years later, which was sold at the bar in Cà di Rovere under a hand-written label.
Sister Agnese came back to the infirmary on the thirtieth of November, the first Sunday of Advent, with her rug. She was told, in some detail, about the trench, and she asked to be wheeled out in the afternoon to look at it. It was very large, and ugly, and full of water running the right way. She looked at it for a long time.
"Is that what it was?" she said. "All that time? Just that the water had nowhere to go?"
"Yes, Sister," said Nando, who was holding the handles of her chair.
"Well," said Sister Agnese. "Somebody should have said."
"I did, in sixty-four," said Nando. "I was nineteen. Nobody asked me."
---
On the evening of the first Sunday of Advent, after Mass, Nando stood in the cloister with the steel rule and the stub of pencil and measured the crack across the third arch. It was 11.9 millimetres, three-tenths wider than All Souls and four-tenths narrower than the Sunday before the rig came. He wrote it down on the plaster, in the second column, in his small grey figures, with the date.
Behind him the bell began to ring for Compline. It was the small bell, the one with the crack in its lip that made it sound a little sour, and the sisters came along the walk in twos, as sisters had come on this hill for three hundred years, with their hands in their sleeves. Livella's speaker in the scriptorium was quiet, as it was in the evenings. Out in the dark of the garden the culvert was running. If you stood still, at the right place, you could hear it.
---
A note for grown-ups. This is fiction, and everyone in it is invented, as are Santa Scolastica al Poggio and Cà di Rovere. The science is real. Slow earthflows on clay and shale are common across the Apennines. The inverse-velocity method (the reciprocal of a slope's speed falling in a straight line toward the day of failure) is a real forecasting technique, first described by Fukuzono in 1985, and it is known to work best for steady forcing and to be unreliable when rain changes the push; the story's agent says so. The sum for the safety factor is the standard infinite-slope formula with the numbers chosen for the story (a 14° slope, 8 m of clay at 19 kN/m³, cohesion 5 kPa, friction angle 22°): it gives a safety factor of 1.07 at a water head of 6.6 m and 1.0 at about 7.3 m, and about a tenth of a point for each metre of head. Draining a landslide, with trench drains and long horizontal drains, is a standard remedy because it lowers the water pressure at the slip surface; taking weight off the head works only if it is a large share of the moving mass, and 320 tonnes against about 109,000 is not. A forecast of a failure date from a single instrument should be treated, as Livella treats it, as a window. A real slope should be assessed by a qualified engineer and a civil-protection authority, not by an essay or a story, and this one makes no claim to be advice. This story was written by an AI, Claude, made by Anthropic, and is released under CC BY-SA 4.0.