City location

Why do cities often grow around river crossings, harbors, or crossroads?

The usual answer is that cities grow where the land is good. But good land is everywhere and cities are not, and some of the largest sit on marsh, sand, or rock nobody would farm. Notice instead what the ford, the harbor mouth, and the crossroads have in common: none is a place so much as a place where a journey changes. What could that do?

Try thinking about the traffic rather than the site. Carrying goods along a route is cheap and steady; the expensive moments are the ones where the carrying changes form. At a river the wagon waits for a ford or a ferry; at the head of navigation the boat can go no further and its cargo must move to land; at a harbor a ship's hold meets a road. Geographers call these break-of-bulk points, and the name is the whole idea: the load is broken open and handed to something else.

Now ask the small question. When cargo is broken open, who is standing there? Somebody must lift it, somewhere must hold it overnight, someone must guarantee that what was unloaded is what gets reloaded. And what does the traveller do while waiting for the tide? Eats, sleeps, borrows money, hears news, sells the surplus he did not want to carry further. Every one of those is a service, and every service is a person who now lives there rather than passing through.

So a handful of households. But why does it not stay a hamlet? Here is the turn: the services are themselves a reason to come. Once there are warehouses, moneylenders, and a market where prices are known, a merchant with a choice of routes prefers this one — which brings more traffic, which supports more services. Economists call the resulting stickiness agglomeration. Do you see that this second stage no longer depends on the river at all?

Two things get muddled here. A place's site is its local physical character — the ford is shallow, the harbor deep, the hill defensible. Its situation is its position relative to everywhere else. Site explains why the settlement is exactly here and not two miles upstream; situation usually explains why it grew large. And once the second stage has run for a century, the original reason can quietly expire without the city noticing: the fall-line towns of the American east coast sit where rivers drop off the piedmont, the furthest point boats could reach and a place to take water power. Neither fact matters much to them now, yet they are still there.

Two corrections. First, this is permissive, not deterministic: the overwhelming majority of fords and crossroads never became anything, so geography names the candidates and something else picks the winners — a charter, a garrison, a bishop, a merchant family's capital, or plain accident amplified by the feedback above. Historians argue seriously about the balance, and anyone claiming location alone explains a particular city is overreaching. Second, a crossing is a political chokepoint as well as an economic one: whoever holds the bridge can tax what passes, so fortresses and customs houses gather at the same spots, adding a second and quite different growth engine to the same map pin.

• Why some break-of-bulk cities died fast when containerization moved the work to a few deep-water ports.
• Central place theory, which asks not where a city is but why cities come in a nested range of sizes.
• Whether digital work weakens agglomeration or, as the evidence so far suggests, concentrates it further.

Every term the collection defines is gathered in the glossary.