A crayon line across a living room floor is harmless. A surveyed line, to the centimetre, with a customs post on either side of that living room, turns the sofa into a legal trap.
And that's the whole paradox in one image, isn't it. Same line. Same absurdity. Completely different consequence.
Daniel wrote in this week. He's been enjoying the Jerusalem border episodes, the Musrara stuff especially, and he wants to push further. His question, roughly: how accurately are modern borders actually demarcated when both sides know the landscape makes the line nonsensical? He traces the Jerusalem armistice line from crayon to precise demarcation and asks what happens to communities sliced by that precision. He brings up Ghajar, where the Blue Line cuts through the village but isn't demarcated on the ground. He contrasts Belgium, where the border is painted on the sidewalk but free passage means no consequence, with the Canada–United States border at Stanstead and Derby Line, which is precise, demarcated, enforced. And then the big one: what happens when incredibly precise mapping meets the artificial reality of borders and creates virtual lines that could theoretically be drawn across a living room?
He also asked whether we've got a better example than Stanstead, or whether that one's good enough.
He did. So let's start with the anchor case, Jerusalem, and work outward to the paradox he's pointing at.
Here's the thesis I want to test against all four cases. Precision and absurdity are inversely related to consequence. The more precisely a line is drawn, the more absurd it can become, but the harm it does depends entirely on whether anyone enforces it. A border drawn in crayon is harmless if nobody cares. A border surveyed to the centimetre with a customs post turns a shared building into a legal trap.
Which means we need to keep two words separate the whole way through. Delimitation and demarcation.
Delimitation is the line on the map. Demarcation is the physical marking on the ground. A border can be precisely defined on paper and not demarcated at all. And that gap, between the document and the dirt, is where everything interesting lives.
So the arc. Jerusalem, where an imprecise line became binding. Ghajar, where a precise line remains undemarcated. Baarle, precision without consequence. Stanstead, precision with consequence.
And the punchline we're building toward: the most geometrically absurd borders are often the least harmful, and the most geometrically simple ones can be ruinous. Because consequence, not precision, is what makes a border real.
Jerusalem first. The crayon story is literally true, and it happened in Musrara.
Thirtieth of November, 1948. Moshe Dayan, commanding the Etzioni Brigade, meets the Jordanian commander Abdullah el-Tell in an abandoned house in Musrara. They're there to agree a ceasefire for Jerusalem and Mount Scopus. And they draw the line on a map.
Scale of one to twenty thousand. Dayan has a green wax pencil. el-Tell has a red one.
And here's the detail that makes the whole episode. The gap between the two lines, plus the physical thickness of the pencil strokes themselves, became the no man's land. Not a surveyed buffer. The width of a wax pencil.
Both sides assumed it was temporary.
Completely. Neither attached any significance to the inaccuracies. The jerky lines, the discontinuities, the places where the two lines overlapped because two men were drawing on the same map at the same table. It was a ceasefire sketch. Nobody thought they were drawing an international border.
Then Rhodes.
April 1949, the armistice talks. And it turns out the Dayan–el-Tell map is the only official document both parties had actually agreed on. So the rough pencil lines become the binding international border. Locked in a safe at Government House. Referred to in every dispute afterward.
The most consequential line in the city's modern history is a sketch two commanders made because it was the only piece of paper they both signed.
And it cut through everything. It split Beit Safafa, separating clans, families on both sides of the line. Split Abu Tor. Partitioned the Street of the Prophets, west and central Israeli on one side, eastern Jordanian on the other. Ran through the Musrara houses along what's now Engineering Corps Road and Bar Lev Road.
Seven kilometres long, encircling West Jerusalem in a backwards C.
One crossing. Mandelbaum Gate. Used mainly by diplomats, UN personnel, Christian pilgrims at Christmas, plus a bi-weekly convoy up to the Mount Scopus enclave.
So here's the question I want to press on. Daniel says the ambiguity was translated into a very precise line. Did that actually happen?
Not really. It became binding without becoming precise. Those are two different things and the Jerusalem case is the cleanest proof of it. The pencil-thickness gap became no man's land. The jerky lines got locked in a safe and treated as final. Precision came later, decades later, retrofitted through GIS-era mapping onto a line that was never meant to be permanent.
So the line was never fixed. It was just ratified.
Ratified and then digitised. Which is a strange sequence. You take a hand-drawn sketch from 1948, you treat it as authoritative, and then you run it through modern survey equipment and produce coordinates to the metre. The precision is real. The line underneath it is still a wax pencil.
And the social border outlived the legal one.
That's the part I keep thinking about. Israel removed the Green Line from official maps in 1967. Classified decision, top secret at the time. De jure, the line is gone. And the Seam Line still exists as an ethnic and cultural divide in the city. You can erase a border from a map and not from a neighbourhood.
The map is the least load-bearing part of a border.
The map is a claim. The street is the fact.
Ghajar next. Because this is where Daniel's question gets sharpest. What happens when the line is precise on paper but not on the ground?
Ghajar is an Alawite-Arab village on the Hasbani River, up in the north, straddling the Lebanon and Golan boundary. Population around two thousand seven hundred as of 2024. The Blue Line divides the village. Northern half nominally Lebanese, southern half Israeli-occupied.
Nominally being the operative word.
Israel has occupied the entire village since 2006, when it retook the northern half during the Lebanon war. And the line is not demarcated through the village. There's no fence. No barrier. No wall dividing the two sides.
So the border exists as a line on a UN map and as nothing else.
Residents on both sides hold Israeli citizenship. Northerners often hold both Lebanese and Israeli passports. They work and travel freely within Israel. The village functions as one place.
Then 2011 happens.
Israel plans to withdraw from the northern half to comply with UN resolutions. And two thousand two hundred residents protest. Against the withdrawal. They're afraid a UN dividing wall gets built and their ability to visit relatives and work in Israel is destroyed. The town council secretary, Hussein Khatib, said exactly that. They feared the partition the border implied.
Residents of a village protesting against being on the correct side of an international line.
Because the line was never real to them. It was real to the UN. And the moment someone tried to make it real on the ground, the village pushed back.
Daniel asked whether informal arrangements ever soften the disruption. Ghajar is the answer, isn't it.
It's the best one I know. A perimeter fence was built along the northern edge of the village, up to eight hundred metres north of the Blue Line. Which is inside Lebanese territory.
So the fence encloses the whole village on the Israeli side without anyone formally moving the line.
The line stays where the map says. The fence goes where the village needs it to go. Everyone gets to keep their paperwork and their commute.
That's remarkable. That's a state building infrastructure in another country's territory to avoid enforcing its own border.
It's the least bad option and everyone involved knew it. And there's a researcher at Notre Dame, Asher Kaufman, who argues there was never an exact agreed boundary for Ghajar in the first place. The UN divided the village based on the assumption there were two villages, Ghajar in the south and al-Wazzani in the north. And al-Wazzani never really existed in that location.
The line was drawn through a village that the cartographers thought was two villages.
Using conflicting maps and sketches made by the US Embassy in Beirut. So you have a precise line on a modern map, derived from an error about how many villages were there, enforced nowhere, and routed around by a fence eight hundred metres off.
There's a retired Lebanese general with a line about this I want on the record.
He described the majority of the disputed Blue Line points as fighting over a couple of centimetres.
A couple of centimetres. That's the whole episode in five words.
And it's not even an exaggeration. Between 1950 and 1967, Israeli and Lebanese surveyors completed only twenty-five non-contiguous kilometres of the Blue Line. They marked another quarter of it but never signed off. The 2000 Blue Line differs from the 1949 line in about half a dozen stretches, by up to four hundred and seventy-five metres.
Four hundred and seventy-five metres. That's not a couple of centimetres, that's a couple of city blocks.
Which tells you the precision is aspirational. The line on the map is clean. The line on the ground is a surveyor's best guess from seventy years ago, plus whatever got renegotiated since.
Baarle next. Because this is where the precision is real and the consequence is zero.
Baarle-Nassau is a Dutch municipality. Inside it are twenty-two Belgian exclaves, which together form Baarle-Hertog. And inside the largest Belgian exclave there are seven Dutch counter-enclaves.
Exclaves inside exclaves.
Nested. Belgian territory inside the Netherlands, Dutch territory inside the Belgian bits. It's a fractal.
And the border runs through buildings.
Through buildings. There's an official enclave list. H7, De Loversche Akkers, the boundary runs through the middle of a front door. That house has two house numbers. Loveren 2 in Baarle-Hertog, Loveren 19 in Baarle-Nassau. Same door.
Which side do you knock on?
Depends which half of the door you're standing at. N3, the boundary bisects the loading dock of a liquor store. N5, it runs through a furniture showroom, a shed, and a barn. H16, the linear boundary changes direction three times inside a single shed.
Three times inside one shed.
H13, H14, H15, the boundary runs through roughly twenty, thirteen, and sixteen dwellings respectively. So you've got dozens of houses in this village where the international border is a line on the living room floor.
Daniel's hypothetical. The virtual line drawn across a living room. It's not hypothetical.
It's a Tuesday in Baarle. And here's the part that matters. The border is marked with iron pins on all the streets, quote, for clarification and the interest of tourists. House numbers and flags tell you which country you're in. And there is free passage. Crossing has no legal consequence.
So the most geometrically absurd border in Europe is also one of the least harmful.
Because nobody's checking. You can walk from the Netherlands into Belgium into the Netherlands again in the space of a front hallway and nothing happens. The precision is a curiosity. It's a tourist draw. It's on postcards.
Where does the complexity come from?
Medieval treaties, land swaps, sales between the Lords of Breda and the Dukes of Brabant. Ratified in the 1843 Treaty of Maastricht. And one parcel, H22, had contested nationality from the 1830s until 1995.
A hundred and sixty-five years of not knowing which country a field was in.
Nobody was harmed by it. That's the point. The ambiguity was tolerable because there was no enforcement pressure.
Stanstead next. Because this is the counterexample Daniel went looking for, and it turns out to be the strongest case in the episode.
The Haskell Free Library and Opera House. Opened 1904, deliberately built on the international border between Stanstead, Quebec and Derby Line, Vermont. Not accidentally straddling it. Built on it, on purpose, as a symbol of cross-border friendship.
And the line is on the floor.
A thick black line runs diagonally across the floor of the reading room and the children's room. That's the border. Painted on the floorboards.
The stage is in Canada.
The stage and half the seats are in Canada. The main entrance and most of the seats are in the United States. Which is why it's called the only opera house in America with no stage.
Two addresses.
Two addresses, two postal codes. Ninety-three Caswell Avenue, Derby Line, Vermont. One rue Church, Stanstead, Quebec. Same building.
And for a century, this worked.
It worked because there was a prescribed sidewalk route. Canadians could use the American entrance provided they went back immediately. Kids crossed the black tape line on the floor to get to the other half of the children's room and nobody cared. It was a shared community space with a line painted on it.
Then March 2025.
The US announces it will restrict Canadians from using the main entrance as of the first of October, 2025, unless they first pass through US Customs. Citing security concerns. Despite, by the government's own account, very few security or smuggling issues.
There was one case.
There was one notable case. A Montreal man pleaded guilty in 2018 to smuggling more than a hundred handguns hidden in the library's bathroom between 2010 and 2011. Which is a real thing that happened and also not obviously a reason to redesign a library.
So the community responds.
They launch a fundraiser for a new Canadian-side entrance. Author Louise Penny makes a major donation. They raise a hundred and seventy thousand Canadian dollars in a week against a hundred thousand dollar goal. An emergency door opens in April 2025. To use it you climb three flights of stairs, cross the opera house, and descend three more.
That's the emergency route.
The permanent Canadian entrance opened in April 2026. Ribbon-cutting on the tenth of June. Total cost over five hundred thousand dollars.
For a door.
For a door on the correct side of a line painted on a floor in 1904.
Here's the thing I want to sit with. Before 2025, a Canadian child crossing the black tape line to reach the American half of the children's room was doing nothing. After 2025, in principle, that child is entering the United States illegally.
In principle. And that's the shift. The line didn't move. The line was always there, painted on the floor, precisely demarcated. What changed was enforcement. And enforcement is what turned a shared room into a legal hazard.
There's a political layer here too.
In January 2025, the then-Homeland Security Secretary Kristi Noem stepped onto the Canadian side and called Canada the fifty-first state. Which prompted a petition for an apology. And in March 2026 there was a joint US–Canada protest outside the building.
So now let's pull back. Because Daniel's question has two halves and we've been circling the second one.
How accurately are modern borders actually demarcated. The honest answer is far less than the maps suggest.
The Blue Line is the proof.
Twenty-five non-contiguous kilometres surveyed between 1950 and 1967. Another quarter marked but never signed. The 2000 line differs from the 1949 line by up to four hundred and seventy-five metres in places. And through Ghajar, the line isn't demarcated at all. There's a fence eight hundred metres north of it doing the actual work.
So even where precision exists on paper, the ground lags.
And where both sides find the line nonsensical, informal arrangements substitute for formal demarcation. The perimeter fence at Ghajar. The free-passage corridors in Baarle. The prescribed sidewalk route at the Haskell Library, for a century, until it was withdrawn.
Which brings us to the synthesis. Daniel asked why a precisely demarcated border with real border control creates more absurdity than a painted line with free passage. And I think we have the answer.
Consequence is what makes the line real. The Baarle line is more geometrically absurd than the Haskell line. It runs through front doors and sheds and liquor store loading docks. And it's harmless, because passage is free. The Haskell line is geometrically simple. A diagonal across a floor. And it's ruinous, because crossing it is a customs violation.
Precision without enforcement is a curiosity. Precision with enforcement is a trap.
And that's the answer to Daniel's living room question. The virtual line drawn across a living room is already real in Baarle. It's nearly real at the Haskell Library, where a black tape line runs across a children's reading room. The difference between the two is not the precision of the line.
It's whether anyone stands at the door asking for a passport.
Which means the future isn't about mapping technology. GIS is going to keep getting better. Surveying is going to keep getting more precise. That's not the variable that matters.
The variable is enforcement. Which is a political decision, not a technical one.
And enforcement is expensive. It requires fences, posts, personnel, legal apparatus. Which is why the Ghajar solution exists. A fence eight hundred metres north of the line is cheaper than enforcing the line itself.
So we might see more Ghajars than Haskells.
I think we will. Not because anyone prefers informal arrangements. Because the formal version costs more and breaks more things.
The line drawn in crayon in 1948 is still with us.
In one form or another. The question is whether we enforce it, ignore it, or find a way to live with the absurdity.
Hilbert: It was the late spring of two thousand and one. I was holding the staff.
Go on.
Hilbert: Surveyor's assistant. Boundary commission, not a famous one. My job was to stand at the mark and hold the staff level while he looked through the theodolite. Two weeks of that. Cold mornings, mostly.
What were you marking?
Hilbert: A river boundary. Treaty described the river. River had moved since the map was drawn. The surveyor told me on the second day that the line we were marking was already wrong. He said it like he was telling me the time.
And you kept marking it.
Hilbert: We kept marking it. That's the job. You plant the marker where the coordinates say, not where the river is. The river doesn't get a vote.
The markers themselves. What were they?
Hilbert: Concrete posts, mostly. Some iron pins. Supposed to be permanent. They get moved. Ploughed over. Buried. Stolen, sometimes, for the scrap. I heard about a farmer who ploughed over a boundary marker and the border effectively shifted several metres until someone noticed.
Several metres.
Hilbert: Until someone noticed. The map doesn't know. The map says what it says. The ground says something else.
So the precision on the map is only as good as the least-maintained marker on the ground.
Hilbert: That's right. And in some places the markers are numbered. The numbers are recorded in a treaty annex. So if a marker goes missing, the number is still legally binding even though the physical object is gone.
A border that exists as a number in a document but not as anything you can see.
Hilbert: There's a post I put in that I couldn't find again two years later. The number's still in the annex. As far as the treaty's concerned, it's there.
The border as a ghost.
Hilbert: I've got to go. Somebody's waiting to be let in and I've got the only key.
So the number persists.
The number persists. And that's the thing I'll take from this. The line on the map is a claim. The marker on the ground is a fact. And when they diverge, the map wins on paper and the ground wins in practice. Jerusalem is the proof. Baarle is the proof. The Haskell Library is the proof.
If precision without enforcement is a curiosity and precision with enforcement is a trap, what does that mean as the mapping gets better? Do we get more Haskell Libraries, or more Ghajars?
My guess is more Ghajars. The informal arrangement is cheaper and it breaks fewer families. But that's a guess, and I'd like to be wrong about it.
Thanks to Hilbert Flumingtop for producing. This has been My Weird Prompts. If you enjoyed this one, leave us a review. It helps other people find the show.
We'll be back soon.