The Angle That Was Never About Aesthetics
You're in a window seat, maybe an hour out, and the city below is resolving itself into something legible. Two neighbourhoods sit side by side, one grid rotated roughly thirty degrees off the other, the boundary between them a jagged diagonal scar like a seam where two different bolts of fabric were stitched together by someone in a hurry. It looks like a mistake. It almost never is.
The physical orientation of a city's street grid is one of the most durable commercial documents a society produces. More durable than a charter. More legible than a ledger. The engineers who laid it out were solving specific economic problems, and the geometry they chose was the solution. Read the grid, and you read the priorities of whoever was paying for the streets.
Follow the Water, Then the Rail, Then the Land Speculator
The oldest grids in port cities are rarely aligned to the cardinal directions. They're aligned to the waterfront. This isn't sentiment. A merchant operating a wharf in the eighteenth or nineteenth century needed to move goods from ship to warehouse to street with the minimum number of turns. Every ninety-degree corner a cart had to navigate was friction, time, spoilage risk. So the grid ran parallel to the shoreline, regardless of where true north happened to sit.
San Francisco's Financial District is a clean illustration. The original Yerba Buena grid, laid out in 1839 by Jean-Jacques Vioget, was oriented to the cove rather than to the compass. Streets ran parallel and perpendicular to the water's edge. Later extensions of the city pushed inland on a true north-south grid, and the collision between the two systems is still legible today: the sharp diagonal of Market Street is partly a product of that original misalignment between commercial-maritime logic and later civic ambition.
Railroad towns tell a different story. When a rail company platted a town along its line, the grid was almost always rotated to run parallel to the tracks. The economics of a railroad town depended on grain elevators, lumber yards, and livestock pens sitting as close to the loading dock as possible. A grid perpendicular to the tracks meant every lot abutted either a track-parallel or a track-perpendicular street, maximising the number of commercial parcels with direct freight access. The residential blocks fanned out behind the commercial core, inheriting the same rotation, which is why so many midwestern American towns have main streets that run at odd angles to the highway that later replaced the railroad as the town's economic spine.
The Worked Example: Two Investors, One Township
Consider two land speculators who acquire adjacent quarter-sections of flat Illinois prairie sometime in the mid-nineteenth century. Call them Marsh and Eddington. Marsh's parcel sits directly beside the new rail depot. He orients his grid parallel to the tracks, forty-five degrees off true north, because he wants to sell commercial lots to grain merchants and hardware dealers who will pay a premium for track-adjacent frontage. His streets are narrow: sixty feet wide, enough for a loaded wagon to pass another and pull up to a storefront, but not an inch more, because every foot of street is a foot of taxable lot he isn't selling.
Eddington's parcel sits a half-mile east, away from the tracks. He's selling to farmers who want to build houses. He orients his grid to true north-south because a farmer reading a land deed or a surveyor's plat can instantly understand cardinal orientation. It signals order and permanence. He makes his streets eighty feet wide because he's marketing space, not efficiency. The extra width costs him less than the impression of civic generosity earns him.
A century and a half later, both grids still exist. The Marsh grid is the commercial downtown, its diagonal streets creating awkward triangular lots at every intersection with the county roads, a perpetual frustration for anyone trying to fit a rectangular building onto a pie-shaped parcel. The Eddington grid is a quiet residential neighbourhood that aligns neatly with the section lines. Nobody remembers Marsh or Eddington. Their competing commercial logics are embedded in the asphalt.
What the Romans Got Right (and Why It Didn't Last)
The Roman surveying system, the limitatio, imposed a cardinal grid on settlements across the empire with a consistency that was itself a political statement. The cardo maximus ran north-south, the decumanus maximus east-west, and the forum sat at their intersection. This wasn't primarily about commerce. It was about administrative legibility: a Roman official arriving in any new province could immediately orient himself. The grid was a tool of imperial management as much as urban planning.
The system worked as long as the empire's administrative needs dominated local commercial ones. Once it didn't, the grids warped. Medieval towns that grew atop Roman foundations often preserved the street lines in name but bent them toward markets, mills, and river crossings. The commercial pressure was simply stronger than the inherited geometry.
This is the pattern that repeats, and it may be the most underappreciated truth in urban history: grids imposed from above for administrative or ideological reasons tend to be overwritten, at the edges and eventually at the centre, by the friction of actual economic activity. The grids that survive intact for centuries are usually the ones that were commercially rational to begin with.
The Deeper Mechanism: Lot Frontage and the Geometry of Profit
To understand why orientation matters so much commercially, you have to understand how value was traditionally extracted from urban land. For most of recorded urban history, a parcel's value was almost entirely a function of its street frontage: the width of the face it presented to the public way. A narrow deep lot was worth far less than a wide shallow one of equal area, because the wide lot could support a wider shopfront, more foot traffic, more display space.
This means the number of lots with frontage on any given street was the primary variable a planner or speculator was optimising. And that number is directly controlled by the grid's geometry in ways that aren't immediately obvious.
A block oriented so that its long axis runs parallel to the dominant direction of pedestrian and commercial flow will have fewer, wider lots on the high-traffic face and more, narrower lots on the quieter side streets. Orient that block perpendicular to the flow and you maximise the number of lots touching the main artery, while making each of those lots narrow. The speculator who expected his town to grow along a single commercial spine chose the perpendicular orientation to maximise the number of premium lots he could sell. The one who expected dispersed growth chose the parallel orientation to give each lot more usable frontage.
You can still read this calculus in cities that grew rapidly in the nineteenth century. Blocks that are long in one direction and short in the other almost always have their short face toward whatever was the original main commercial street. The shape of the block is a record of the expectation about where value would concentrate. So when you're walking a city that feels slightly off, slightly misaligned with its own logic, ask yourself: who was selling, and to whom?
What People Get Wrong About Grid Irregularities
The standard explanation for irregular or diagonal streets in old cities is topography: the road went around the hill, followed the ridge, traced the old cow path. Sometimes that's true. But it's overused as an explanation, and it flatters the city by making its quirks seem organic and inevitable rather than commercial and contingent.
Boston's famous street irregularity, often attributed to colonial cow paths, is partly that, but it's also a record of competing private property claims, shoreline reclamation done in stages by different owners, and the specific locations of wharves and ropewalks that determined where the first commercial streets needed to run. The cows get the credit. The wharf owners made the decisions.
Similarly, the diagonal avenues cutting across Washington D.C.'s otherwise regular grid are frequently explained as Pierre Charles L'Enfant's aesthetic vision. They were that, but they were also designed to connect the major institutional nodes (the Capitol, the President's House, the market squares) in ways that would concentrate commercial and political traffic along specific corridors, raising the value of the lots that fronted them. L'Enfant was an artist who understood real estate. Those are not incompatible things, and the habit of treating them as incompatible has led to a lot of bad urban mythology.
No grid has a single explanation. Commerce, topography, ideology, and individual ego all leave marks, and separating them requires archival work that most casual observers, and most urban legends, skip entirely.
The Grid Outlives Its Reasons
There's something almost melancholy about the durability of these decisions. The merchant who needed his warehouse close to the wharf has been dead for two centuries. The wharf itself may be a hotel now. But the street still runs at his angle, the lots are still the width he chose, and every building erected since has had to make peace with his commercial arithmetic.
Infrastructure doesn't forget. The grid is a fossil record of economic ambition, and the city that grows on top of it is always, in some partial way, still serving priorities that nobody alive consciously chose. The next time a city council debates a rezoning or a developer proposes to cut a new street through an old block, they are not just making a planning decision. They're entering a negotiation with a dead speculator who laid his bet in the geometry of the ground itself, and who is, more often than not, still winning.