Reading the Traces of Time

关于时间、王朝与遗迹的一篇短文：历史学家如何从石头、骨头和日常器物里，读出一个已经消失的世界的形状。

The Wall Nobody Built Alone

The wall was old before anybody called it old. Two hundred years ago, farmers working the fields north of the river turned up stones that did not belong there. They were too heavy for houses and too square for tools, so the men piled them beside the road and forgot about them. A girl of about ten walked past every morning on her way to school. She put her foot on one block each day, as if she were counting, and the number never came out the same twice. Her mate, an old man who repaired boats along the bank, told her once that these stones were older than any king whose name she had learned. She did not believe him then. She believed the food she carried in a cloth, the color of the sky in winter, and nothing else at all.

It took a century and a half of quiet work before anyone asked the stones a real question. A history of the wall begins with that question, because a wall is not a fact. It is an argument that somebody once won, and then everybody forgot the losing side. Farmers kept the fields. The government kept the records. The stones sat in the middle, keeping the date of nothing in particular.

Look at the wall from the south and it appears to be one long period of careful work. Look from the north and you see three separate periods, built in a hurry, then patched, then built over. The age of the wall is not a single number. It is a stack of numbers, each one sitting on top of the last.

The stones stand in a loose circle now, though nobody living remembers whether they always did. That question is not idle. A change in the stone of a wall, or in the bond of the mortar, is a change in the people who built it, and those people build in traditional directions that leave marks on the ground. Follow the marks and the wall stops being one wall. It becomes a development after development, each layer attached to the last, and each one built for a reason we now have to guess.

The village adopted a name for the place early on, and nobody agrees what it should have been. The first version was simply the old word for the field, and it survives only in a register nobody has opened since 1902. The second version was a man's name, chosen because the man had paid for a well. Both are true in the way that a road is true: somebody walked it first, and then everybody else did.

Under it, the village keeps a shelf. Boys dig there after school and come home with a red tile, a green bead, a bent coin, and their mothers wash the pieces and set them on the shelf like small trophies. Nobody has ever counted them. A few are wonderful in the ordinary way: a thin ring no larger than a pea, a piece of glass the color of a rose. The boy who found the ring carried it in his pocket for a week and then left it on the table for the family to see. Nobody in that house has the faintest idea what it is. Everybody in that house would be unhappy to lose it.

Last spring a woman in her seventies came back to the field with a metal detector and found, forty metres from the shelf boy, a bronze pin with a spiral head. She took it home and wiped it with a cloth, and the shine came up under her hand, and she sat down in the grass and did not move for a while. She told nobody for two days. Then she told everybody, and within a month four more machines had appeared on that field, and by the end of the season the shelf boy had eleven new objects on it, each one with a story attached that the finder had already told six times.

Nobody minded the noise. The noise was, for three weeks, the most alive sound in the village, and it is worth recording that this is what an ordinary discovery sounds like: not a revelation but a crowd.

Dates Nobody Wrote Down

Every historical claim begins with a date, and every date begins with an argument. A king dies in a year, and the year is written down, and the writing is copied by hand for two hundred years. At some point a copyist is tired. A letter is rubbed thin. A building is burned by somebody who will never be named in a book. What is left is a death with no year, a birth with no day, and a king floating in era without an address.

So the historians build other clocks. A decade is guessed from the coins buried with a body. Timber is measured by its rings, one ring for one year, and a beam cut in the wrong season will give an answer off by two. Pottery changes its style in a way that spreads slowly outward from a capital, like ink through water. A door lintel is compared with other lintels, and the closest match is taken as a representative of a group. This is slow, and it is not exact, and everybody accepts it anyway.

In one town the local register of births survived, and it turned out to be the most valuable page in the whole publication. A girl in the last line had been born in the same year that the wall was finished. The trial of a thief two years later records a buyer of stolen goods, which is a strange sentence to find in a book about the dead. The man was ordinary. His name appears twice and then never again.

The methods have limits that are worth stating plainly. There is a maximum age a beam can be dated to, and there is a point below which a charcoal sample tells you nothing. The honest position is somewhat uncomfortable: the argument is usually unclear at exactly the moment when people most want it to be clear. A prompt decision, made in the first month, would have been wrong, and a decision made in the twentieth year is still probably wrong, only more slowly.

So the working rule is simple. Assume nothing without a date beside it. Not because the rule is scientific in a laboratory sense, but because a claim without a date cannot be checked, and a claim that cannot be checked cannot be convinced into evidence. The rule sounds dry. It has saved more than one career.

The word assumption does most of the work in every professional history ever written, and the word implication does the rest. The moment an author moves from what the evidence says to what it might reasonably suggest, the ground shifts under the argument. A good writer marks that shift on the page. A bad one simply walks across it as though the two rooms were one.

A graduate student named Behrens spent one summer with a tape measure, a notebook, and a bad temper, comparing every beam in the lower town with every beam in the upper town. He was twenty-six, he was paid almost nothing, and he nearly stopped nine times. What he produced was a table of slightly different numbers and one useful sentence: these two groups of buildings were not put up by the same group of people. The ability to hold a question open for twenty years is rarer than the equipment.

Behrens was not a genius and he never claimed to be. What he had was an everyday kind of stubbornness, the sort that cannot let a small unexplained thing go. He pursued one tile from one floor for a decade, following it wherever it led, and the trail ended in a warehouse nine hundred miles away in a crate marked with a number and nothing else. A stone that does not fit the pattern exists to be argued with, and he had hunted through two counties to find one.

What the Ground Keeps

Nothing from the past is kept in a way we can control. A preserver does not choose the shelf; the soil does. A thing survives if the ground around it is dry, cold, or extremely hard, and it disappears completely if the ground is wet and soft. That is why a bronze hook came out of a northern peat bog almost new, while a wooden shield from a wetter valley left nothing but a faint dark condition in the dirt: a stain that says something heavy was here once, and no more.

Under the northern wall, men found shells strung on cord, then metal components from a tool, then a bone needle with its hook still tied in place. Nearby, in a house that had burned, they found the shell money of a whole neighborhood, and beside it a wheel hub, a pack saddle, a coil of heavy rope, and a jar of fuel that had gone to tar. The most surprising element in that room was not the metal at all. It was the string. Somebody sat down in the last hour of a fire and threaded small shells onto a cord for a reason we will never recover.

Transportation has its own story. A road implies carts, and a cart implies a surface somebody worked on for years. A river implies boats, and boats imply a route that somebody piloted with the season in mind. A delivery of grain from one valley to the next implies taxes, a board of elders, and probably a long quarrel about the rate. A road like that is essential infrastructure, and infrastructure is the most honest kind of history, because nobody builds one for a poem.

Walls and ditches are the same story told louder. A fort is a strong place for a few soldiers. A wall running the length of a frontier is a long project of ordinary labour, done by ordinary people, and it is evidence of a state that could feed and pay them for years. The strength of a wall tells you about the strength of the government that built it, and the stretches where the wall suddenly stops tell you where the government's reach also stopped. On this wall the gap is four hundred metres wide, and the earth there is untouched. Nobody had to write that down. The absence is the sentence.

Some of the transport was stranger. A narrow-gauge track runs from the quarry to the river, and it was clearly designed for a single load, because the loading platform is a shallow roll of stone with a lip on one side. Somebody engineered a decade of freight down to the shape of a single crate. Wherever that crate was being taken, it mattered enough to reshape the landscape around it, and we still do not know what it was.

Digging is a craft with its own rules. A deep hole is worth more than a wide one, and a small square one more than a deep one, because a square hole keeps its own walls and a round one keeps nothing. A digger learns this slowly and by breaking gear. Winter is the correct season, because frozen ground is a wall you can lean on, and because stone can be moved when the mud is not arguing. Diggers in that generation were strong, patient, and paid by the meter, which made them careful about the walls of a hole and careless about the stories inside it.

Experimental work fills the gaps where the direct evidence has run out. Someone will take a bundle of the same fibre, soak it in the same mineral water, and beat it for a week to see whether it can have been cord, or leather, or something else altogether. The results are reasonably convincing and they are not proof, because a trial that fails proves very little about the case it resembles. But it is a way of taking the word impossible out of the discussion and putting something there in its place, which is the only honest use of an experiment in a subject made mostly of things nobody can repeat.

The opposite is also true, and the profession has been scarred by it. Between the wars a number of extraordinary claims were made about a particular stretch of northern ground, supported by a private individual, a large fortune, and a genuinely gifted writer. Much of it was fraud. But some of it was not, and the fraud and the truth were mixed in the same published book, and afterwards nobody could separate them for a decade. The lesson is not that amateurs are fools. It is that a claim needs a person behind it who can be doubted, and that the individual scholar, however brilliant, has no mechanism at all for being wrong in public. The effect on the discipline was dramatic, and the effect on the individuals involved was worse, and the establishment that eventually absorbed the records had to spend twenty years untangling them.

And not all of it is ours to read. A grave in the old field is a family's, not a museum's. The rule is simple to state and hard to live by: what was buried on purpose stays buried. In a village where the ground is soft, digging is not a research method but an interruption, and the field belongs to the living as much as to the dead. The worth of a find is not measured in what it sells for. Most of it cannot be sold at all.

The ethical situation is harder than the technical one, because the technical rules have an answer and the ethical rules rarely do. A wall in a field that a farmer intends to plough is in no danger from a scholar. A town that is rebuilding its centre is. In three cases that I know of, an extension to a road was shortened by eleven metres because an archaeologist asked, and in two of those three cases the road was later moved again, for money, and the eleven metres were lost. It is not a comfortable story and there is no version of it in which somebody is entirely right. The most that can be said is that the ground beneath a city is finite, that the decisions about it are made by a handful of identifiable people, and that those people are answerable.

Excavation also transforms a place while it is happening. A trench changes the drainage. A spoil heap changes what grows. A trench left open over one winter can destroy the very layer it was opened to read, because frost gets into a cut wall and prises the stones apart. A good team therefore plans its own abandonment in advance: how the trench will be filled, what will be recorded before it closes, and which stones will be lifted out and kept somewhere with a roof. Most of the ground is only ever opened once.

Ordinary Lives Inside Big Events

Now come the public stories. An empire rises, a rival empire answers, and a war goes on for eleven years. The national version is clean and dramatic. The local version is not, and the difference between the two affects the answer dramatically.

In a farming village on the border, the records of the war years are almost all about grain. Whether the crop was planted. How much of it the soldiers took. What a female merchant paid in salt. A couple married in the third year of the war and never saw the fourth. The children of that year are counted in a national paper as a population percentage, which gives you a number and nothing at all about a girl of four who could not stop coughing.

A war is a sudden change in position, but for the people inside it the change is a long squeeze. A king declares independence. A governor modifies the law to fit the new border, and a counter-argument appears the same week. A supporter of the old order takes shelter in a hill town, and a supporter of the new one leaves for the coast within a month. Both families are entitled to a story, and the grand version is the one that survives in the capital first.

The military history of that war fills nine volumes, and the civilian history of it fits into forty pages that almost nobody cites. Those forty pages track food. A siege is, from the inside, a supply problem that lasts four months, and the moment the grain runs out the cavalry stops being interesting to everybody including the officers. A civilian court in the border town held its sessions in a building with no roof, for the entire season of the fourth winter, and the surviving minutes record nothing but the disputes over firewood. They constitute, in a way no battle narrative can, what a war was.

The economic records are more honest than the speeches, and they are also more boring. A town that had been self-sufficient in oil suddenly needed to buy it, and the price of bread doubled in a year. That is a massive change, and it happens rapidly enough that the older people in the house still describe it as something that happened in a single season. A change that fast leaves the mind behind, and the mind invents a story instead: a famine, a plague, a battle that never happened. The voice of an old person telling that story is worth listening to, and so is the silence in the middle of it, where the real explanation would go.

Older people in that village had a second memory that no record contains, and it is a memory of sequence rather than of event. They remembered which families left first, and in which direction, and roughly how long the road took. That order is enormously useful and almost impossible to obtain from a document, because nobody thought to write down the order. It merely survives, and a historian who sits in a kitchen for three afternoons can recover something that an entire archive of letters would not have given. The technique is embarrassingly simple. It consists of asking a specific question, waiting, and writing down the answer in the family's own words rather than your own.

The surprise, if it can be called that, is how often the people who know least about the theory know most about the detail. An elderly man I spoke to in the coastal resort town remembered, without any prompting, that the old road came in at a different angle and that there had been a gate on the north side, and he had built a model of it as a schoolboy. The model was in his attic, in a tea chest, made of a cardboard tube and about forty inches of string. The initial reading was that the tube was the gate. The initial reading was wrong; the tube was a well, and the string measured a depth. Both features belonged to the same small building, and the rest of it was in a photograph taken in 1904 by somebody who had no idea what he was documenting.

Nobody is born a slave and nobody is born free. In the first years of the empire the law is written by a man who has never met one, and the boundary moves as the tax moves. A buyer in the port writes his letters in a merchant hand and sells more grain in a bad year than in a good one. The system is capable of enormous improvement and of terrible suffering, and the same decade shows both, which is the part that a national story finds hardest to say out loud. Everyone in that port suffered; almost nobody was in a position to say so in a sentence that anybody printed.

The registers of the port are the most sobering pages in the whole business. They are also the most ordinary. A line reads: six barrels, one male, two women, one child, price agreed, no birth date entered. Whoever wrote it was in a hurry and would have said he was doing nothing remarkable. An infant recorded without a birth date is not a mystery to be solved. She is a piece of paperwork, and the fact that a system could treat her that way for a century is the finding, and it is enormous, and no wall has ever announced it so clearly.

What survives is often the smallest thing. A tale told by a grandmother about a fish that was too large, repeated in a dialog where nobody knows who is speaking. A piece of period clothing, normal cotton, a dozen sizes. A wall that never came down, standing in the middle of a modern street beside a neighborhood that now has a favorite coffee shop. Every historical layer is a position somebody fought for, and the coffee shop is a position too. Nobody wins that argument permanently, which is the point of The Weight of a Choice.

A single episode survives in that neighborhood better than any dynasty does. A woman born in the last year of the war used to say that the whole street went silent for an hour on the day the news came, and that afterwards nobody could agree who had spoken first. She died in 1998. The story occurs in print exactly once, in a small local book of no distinction whatever, and it is now the only source for a day that changed the street. A stroke of luck, a cheap pamphlet, and a memory that somebody finally wrote down. That is the whole accident, and it has occurred to perhaps one episode in ten thousand. The interval between a thing happening and a thing being written down is sometimes a generation, and that interval is where most of the past has gone.

How Historians Argue

Historians do not investigate the way detectives do. They investigate the way people who must be wrong in public do. They start with a notion, a draft of an idea, and they try as hard as they can to break it. A study must be technical enough to be repeated and plain enough to be familiar to a reader who has never seen a kiln.

The procedure is boring on purpose. Gather what previously nobody thought worth keeping. Establish the scope of the claim, which is exactly how much the writer is trying to explain. Separate what is obvious from what is implied. Put the two on a scale and watch the balance. Ask which words in the source are evidence and which are the narrator addressing a listener rather than a judge. Then declare the finding, briefly, and check it against something from a different place.

It is a slow craft, and it looks, from the outside, like a series of people in a cold building arguing over paper. It is worth being seriously clear about why the arguing matters. A result published too fast, however dramatic, becomes the version that every later writer copies, and the copy becomes the fact. People believe what they have heard repeatedly, and the effect of one early dramatic error on a field is enormously greater than the effect of the hundred quiet corrections that follow it. Almost nobody in the field is deeply upset by a mistake. Almost everybody is quietly relieved when it is found and printed. The strange-looking things that do not fit the story are the most valuable part of the account, precisely because they are strange. A writer who hides them has exchanged a small embarrassment for a permanent error.

A young researcher will tell you that her first dig changed everything. She is not wrong and she is also not finished. It is easy to catch the moment when a site stops being a set of fragments and becomes a place where people lived, and it is hard to stay there afterwards, because once a field has a person in it you can hardly go on treating it as a problem in the distribution of pots. The professional temperament is largely a matter of resisting that feeling for about ten years, which is not a small thing to ask, and which hardly anybody talks about in public. The extra effort it costs is invisible in the publication and is the entire difference between a book that people cite and a book that people read.

Reading a good monograph is a lesson in this. Watch the author turn an odd little number over and over, closely, the way you would turn a coin you had found in a coat pocket. What looks like fussiness is the method. Nobody has ever won an argument by being right faster than everybody else, and the eventual winner in a long scholarly dispute is very often simply the person who stayed with the problem longest and said less.

The failures are more instructive than the successes. An argument built on one unusual object can collapse. A trial of ideas goes exactly like a trial of a theory about pottery, and the losing version should be isolated in a paragraph where it can be seen from all sides. A good writer integrates the exception into the account instead of dropping it, because an account with no exceptions is not an account, it is a wall. A selection of sources is a selection, and the missing part is doing silent work in the conclusion, where nobody will look for it.

Twice in my own reading the exceptions turned out to be right and the rule was wrong, and both times the proof arrived differently from what anyone expected. The first was a coin of an ordinary sort, found on the far side of the frontier where it had no business being. The second was a stretch of wall that everyone called late, until somebody noticed a twist in the metal clamps that only an earlier workshop produced. In both cases the standing explanation had the merit of being defended by people who had spent twenty years on it, which is the strongest defense there is. Scientific training is not a body of facts. It is the habit of noticing what the facts refuse to do, and of modifying the argument when they refuse long enough.

Careful preparation also protects the small sites. Before a trench goes in, the plan is checked twice, and somebody whose whole job is to stop the work writes down what might be lost. That person is occasionally unpopular. The intervention of a local society can turn a routine dig into a two-year delay, and the delay regularly turns out to be the right call. Ancient ground is not renewable on any human scale, and this is a double sentence: it is dramatic to say that, and it is also, apparently, true in the ordinary administrative sense of budgets and permits and letters.

This is why the people who do the work seem so odd. They are excited by a bucket of mud. They watch the color of a surface for an hour. They record the position of every stone in a grid that will be covered by a tarpaulin by Friday, and they are right, because when the tarpaulin goes, somebody has to know what was under it. Nobody thanks them. The satisfaction of the work is entirely in the doing.

Consider how a date gets corrected. Somebody finds a small site on a farm, initially written off as a sheep fold, and hunts through the parish records for the name of the family that held the land in that century. The name turns up in a rent book, and the rent book mentions a dovecote, and there is a dovecote on the farm, and that changes the reading of the whole site. It is now a house, and a substantial one. Three years of work turned a fold into a house, and the result was published as four pages with a diagram, and the person who did it went back to surveying because she says, quite reasonably, that surveying pays.

Nobody in that story is a hero. The person who noticed that a fold was too large for sheep was not more intelligent than her neighbours; she was merely the kind of person who catches the small discrepancy and follows it, and the discipline is built almost entirely out of such people. It is not a romance. It is also the only method that has ever found anything, and it has found a great deal.

Two things follow from all of this, and neither is a slogan. First, an extraordinary result requires extraordinary evidence, and almost nothing in a village field rises to that level. Second, the ordinary result is the one that accumulates. A century of competent, unremarkable documentation will transform our understanding of an ordinary valley, while a single spectacular claim about a spectacular site will teach us almost nothing, because it will be overturned and the overturning will be the only part anybody reads. The professional evaluation of a dig is therefore mostly a matter of proportion, and it requires the writer to describe the drudgery in detail, which is the least dramatic possible way to justify the money.

A Quiet Ending

A country that has existed a long time develops two ways of looking back. One is official, dated, and printed in a publication with a national cover. The other is everywhere else: a grand doorway nobody can explain, a hill that people avoid for reasons they no longer have, a word in a village dialect that means nothing in the capital.

Archaeology works in the space between the two, and it is a distinct kind of conversation. It adds what was written down; it does not replace it. It gives weight to a name. It says: there was a person here, and here is what they were holding. It is more ordinary than the old stories, and for that reason more convincing. It has no heroes, only the continuous work of unknown people who kept a wall up long enough for the wall to matter.

The two accounts transform each other, and this happens more often than the official version likes. It is the same kind of argument that runs in How We Know What We Know, only slower and with fewer instruments. An old local name for a hill gets taken seriously by a linguist, who finds eleven variants of it across four neighbouring valleys, which suggests a settlement pattern nobody had predicted. The official map is redrawn. Nobody in the field is paid for it. The art of drawing a map by hand, incidentally, is itself a kind of historical tool, since it forces the drawer to decide what counts as a road and what counts as an accident of ground, and every such decision is an argument that somebody else can dispute. A survey is a theory with a ruler in its hand.

The last two centuries are a strange place to end. The past century saw almost every large social change in Europe begin with a rapid change in something technical, and the changes were driven by wars and empires. In the first half of the nineteenth century, the old order of life was overturned in almost every country. Whether this was progress or only change is a question historians still argue about, and each generation answers it for itself, and none of the answers is final.

I once watched a boy in a museum say hi to a case of old objects. He said it the way you say hi to a cat behind a window, and then he asked whether the objects were real. His teacher said no, they were copies, and the boy's face closed, and I felt sorry for the copies. The originals are in storage, in a condition that air and light will damage. A real ancient thing is more fragile than the photograph of it. We preserve a pale impression of the past, and we forget how much work it took to earn even that.

History is ultimately a discipline of wonder, and wonder is the one thing that cannot be organized. A serious book can contain thirty pages of tables, six of them on agricultural output in a valley nobody has visited, and the number that changes your life will be something else: a child's height at ten, a wage for one season of weaving, a doctor visiting the fever ward on a Tuesday. The figure with the most decimals is rarely the figure with the most light in it.

It is a discipline of evolution, in the honest sense: one slow adjustment after another, each defensible, none final. And it is a discipline of limit. The peak of certainty in a well-run dig on a well-run site reaches perhaps eighty per cent, and the remaining twenty per cent is not ignorance that a bigger machine will remove, it is a permanent part of what the evidence is. A historian who promises you the whole of it is selling something, and the person buying it is usually the person who most wants to be told a story. I know this because I have wanted to be told one.

Two corrections from readers are worth repeating here, because they are the kind of thing nobody argues about. An elderly schoolteacher wrote to say that the fields she grew up in were ploughed over in the fifties and that no one had ever asked her what was under them; she had a photograph of the stone that marked the boundary between two farms, and she was the only person alive who could say which side was which. And a retired engineer wrote, entirely without bitterness, that the wall on the road south is not a fortification at all, merely the foundation of a row of houses that were moved when the valley was flooded. His grandfather had told him so. Nobody had listened. Both letters made the argument look small, and both were correct, and I have quietly used each of them to modify a paragraph in the version above.

So the past arrives among us in fragments and nowhere else. Whether it is a wall in a field or a word nobody uses any more, it stimulates a curiosity that a good museum would gladly feed and a good history would sooner complicate. The finest historians of the last century wrote like that, and the field is different for it, and I do not think it has been a bad trade. The tender detail is always somewhere in the margin. A thirsty courier stops at a spring. A rose is pressed in a book that is otherwise a list of grain prices. A girl asks her grandfather where the road went, and he answers with a story that is forty years off and unmistakably true about the family.

The historian's real preparation is daily and slightly ridiculous: a habit of looking at the ground. It is odd, in a life full of devices, to feel that the closest thing to certainty lies in a layer of dirt, but that is where the certainty is. The ground does not summarize. It does not make an argument about progress, and it does not know the name of any king. It simply keeps what was put into it, and it returns it slowly, and the person who returns it to us is usually not famous and never will be.

The exchange between past and present is not a transfer of property. The past does not belong to us, and neither, necessarily, do we belong to it. We are the middle of a long sentence that began before anybody was listening and will continue after the room is empty. The interesting question is not whether the past was great, but what it can still do for the people living inside it.

I can justify a good deal of it and I cannot justify all of it. A visitor to the coastal resort three valleys away asked me, with real politeness, why anyone would care about a heap of stones on a farmer's field, and I gave the standard answer, and it did not satisfy either of us. What finally worked was smaller than my argument. I told her about the woman with the metal detector sitting in the grass. My listener's face changed, and she said, almost in irritation, that she has a box like that at home and has never once looked inside it. We sat down and opened the box. There was a jade bead, three keys, and a saucer. No explanation whatsoever. It was, without any argument at all, a good afternoon.

This is the part of the subject that never appears in a professional journal. Historians are not unusually moved and they are not unusually cold. They are people who spend their lives in a room where the past is a sequence of small physical facts, and that room changes you. You come out and cannot stop measuring things.

We are also, whether we like it or not, a civilization with a memory problem. Somewhere in every institution there is a portion of paper that is not stored, catalogued, or protected, and that portion is not a neutral sample. It is a sample of whatever nobody at the time thought would ever be needed. The establishment of a national archive in the nineteenth century saved millions of documents and altered nothing whatever about which ones got thrown away. To take a snap decision in that direction is to decide, without noticing, what the century after you will be able to know. I mention it here not to make anybody guilty, which is a waste, but because the choice is real and is made in an ordinary office by an ordinary person on an ordinary afternoon, and it is the same choice, in miniature, that a single farmer makes when he knocks down a wall to widen a gate.

I would like to leave one small story. In the last village school, the children were asked to write down the oldest thing they knew of. A boy of eleven wrote that the wall was eleven years old, because his father had told him so. A girl of eleven wrote that the wall was older than her father's father, which was obviously wrong, and then crossed it out and wrote something different instead. Her teacher kept the paper. Somewhere in a drawer in the school, folded twice, there is a version of local history written by a child who decided that a fact she could not check was not worth putting in the answer. That decision is older than any dynasty, and every generation of that village has made it, and every generation has been wrong in a different way, and each of them was right to begin again.

The past is a house with most of the doors locked. The tools are ordinary: a shovel, a brush, a grid, a notebook, patience that nobody praises. What you get back is never the whole truth. It is a partial voice, a fragment of an ordinary day, a place where somebody once stood, and it is true that some of it is wrong, and it is also true that we are building this house of locked doors ourselves, one question at a time, and that the questions are the only part of it we get to keep.

The Water and the Wall

Two more pieces, and then I will stop.

The first is about a spring. In the high field there is a wellhead that has been stone since the seventeenth century, and it is fed by a channel that someone cut, and the channel is what keeps the well alive. A parish record from 1711 mentions the channel twice and a dispute once, which tells you what mattered. In the 1970s the channel was straightened, for efficiency, and the well declined and nobody connected the two things for forty years. A thirst in a valley is a historical document just as much as a document is.

The second is about a wall. I once heard a young guide explain a Norman earthwork by saying that the people who built it did so because they were told to, and that the work was done by farmers who had no idea what It Was For. That is a very modern and very comfortable story, and it is not quite true. The men who built that bank were governed, yes, but they were not fools, and there are signs in the ditch that they improved the design twice as they went, which is what people do when they care.

A society, it turns out, has never been governed entirely from above. Even the most autocratic regimes have left room for the worker to be better at the job than the person who ordered it, and that gap is where most of human history actually happened.

生词与提示