Saturday, March 19, 2011

A Critique of the Grid, part II

Miletus' "Hippodamian" Plan
I've been reading a bit about Hippodamus of Miletus, the man credited with popularizing (if not inventing) the grid in the Western urban tradition.  I'd always considered the grid a reflection of a rational Classical Greek mindset, but Hippodamus seems to have been as much concerned with engineering utopian outcomes as simply laying out a physical plan — making himself out to be the Ebenezer Howard or Frank Lloyd Wright of his day. 

None other than Aristotle had rather uncomplimentary things to say about him and his ideas, calling him "a strange man, whose fondness for distinction led him into a general eccentricity of life," (Politics, Book II part VIII), and regarding Hippodamus' apparent plan for a third of city-dwellers to practice agriculture, reasonably asking "what use are farmers to cities?"

Aristotle did, however, tentatively endorse the so-called "Hippodamean" grid, writing that "[t]he arrangement of private houses is considered to be more agreeable and generally more convenient, if the streets are regularly laid out after the modern fashion which Hippodamus introduced."  Politics, Book VI, part XI.  Left unasked was the question of whether there was a relationship between Hippodamus' social proposals, which Aristotle denigrated, and his choice of the checkerboard grid.  Although some have credited the grid as an egalitarian choice for city planning, Hippodamus' evident plan for a top-down, class-stratified society run by a military caste (in a city designed by none other than himself) hardly reflects a democratic preference.  

In fact, rather than being an embodiment of egalitarianism, the grid's emergence is most closely associated with the rise of powerful and centralized states (see The Dark Side of the Grid), although other motives may factor in as well.  China's imperial capitals, for example, employed a grid not for convenience of wayfinding but as an imperatorial aesthetic of sorts intended to reinforce the order and centralized authority of the emperor.  In the Renaissance, the grid embodied an rarefied aesthetic of uniformity and rationality.  For others, such as the expansionist Greeks of the 6th century B.C. and 19th century Americans, the grid was an enabler of land speculation in newly-settled territory. 

The one purpose the grid has never served, evidently, is the desires and everyday needs of the citizens actually living in the city.  Where the centralizing power responsible for maintaining the grid disappears, the grid therefore frequently breaks down.  As Spiro Kostof writes in The City Shaped on the fate of gridded Roman cities in the medieval period:

"The grid is inflexible in terms of human movement.  We are not inclined to make right-angle turns as we go about from place to place unless we are forced to do so.  With the impairment of municipal controls in in the post-Roman city, natural movement soon carved shortcuts through the large rigid blocks of the grid." (p. 48).

This process is apparent in the ancient town of Rimini, at left, where the current street network, in black, is shown overlaying former Roman grid.  The grid's failure to provide direct routes to the central market (other than the Roman Cardo and Decumanus) has been partially addressed by diagonal pass-through alleys; overly large blocks have been bisected; sharp turns have been smoothed in an emerging concentric pattern, and (not visible in the diagram) many streets were narrowed from their Roman proportions.

The New Urbanists seem to have retreated from their initial praise for the "hypertropic" 19th century American grid and today advocate more complex street patterns with a hierarchy of streets.  The lessons of the grid, though, in a broad historical context, may not yet have been fully appreciated.


  1. I think what makes most urbanists nostalgic for the grid is its interconnectedness. Sure, a Cartesian grid plan is rigid and monotonous -- but it's got amazing redundancy and connectivity. From any given point A to any given point B there are multiple routes -- often multiple routes of the same length.

    Compare this with the cul-de-sac within cul-de-sac design of modern subdivisions. Half a dozen houses face onto one dead-end street, which plugs into a larger no-outlet street festooned with other dead ends, all opening out onto the old county road which is now choked with traffic from half a dozen other big unconnected developments.

    My favorite example of this was in Cary, North Carolina, where at one point something like ten thousand households were accessible through only a single intersection.

  2. Cambias: For many reasons as you point out the grid has advantages over the cul-de-sac plan. But there are good reasons the cul-de-sac typology emerged in the first place, reasons that in part have to do with the shortcomings of the 19th century grid (requirements for street width have distorted the economics as well).

    One of the most important of these is that a grid of broad streets lacks quiet, low-traffic ways of the kind most people would prefer to live on. Every street works as a possible route so every street becomes a thoroughfare. When cars arrived in the first half of the 20th century it became unpleasant and (for families with children) unsafe to live along these streets with traffic constantly speeding by. The cul-de-sac plans that eventually emerged can be seen as an understandable, if misguided, response to this.

    Fortunately it's not "either or" since there are ways to combine the connectivity of grid with the quiet and privacy of the cul-de-sac, and that's where I think some of these older organic street plans have a lot to teach us.

  3. Still, when there's a dense grid of short blocks, the amount of traffic on any one particular street is so small that it might just as well be a cul-de-sac. The slippery slope comes when sections start to be blocked off because someone only wants 3 cars per minute instead of 5 driving by. Suddenly the traffic starts concentrating on all the other streets, then they become more unpleasant and are subject to closing off too, and before you know it there's gridlock on the few through streets left.

    Better suburbs do provide through connections for non-motorized travel, but they still cause traffic problems when using a dendritic layout that restricts through routes.

  4. A most interesting topic - may I ask two questions:
    Why are there 2 grid proportions at Miletus? Was it Hippo having 2 bites at the cherry?
    The rectangular southern grid would appear to be more flexible in being able to offer plots for different building sizes compared to the (earlier?) square grid on the peninsular.
    Is the (older) Chinese city grid tradition derived from a manifestation of the proportion of the human body>bed>door>room>courtyard house>hotong>city?
    Maggie Keswick in "The Chinese Garden" says the plan of the city can be seen in the balcony rail of a Chinese garden verandah.