Spiro Kostof, "What is a City?" in The City Shaped: Urban Patterns and Meanings Through History (Boston, 1991): 37-41

To conclude these introductory remarks, I think we can agree on some simple premises about cities, regardless of their origin, their birthplace, their form, their makers.  Two sensible definitions, both from 1938, would allow us a good starting point.  For L. Wirth, a city is "a relatively large, dense, and permanent settlement of socially heterogenous [sic] individuals."  For Mumford, a city is a "point or maximum concentration for the power and culture of a community."  Here is my gloss on these fundamental premises.

A.  Cities are places where a certain energized crowding of people takes place.  This has nothing to do with absolute size or with absolute numbers:  it has to do with settlement density.  The vast majority of towns in the pre-industrial world were small:  a population of 2,000 or less was not uncommon, and one of 10,000 would be noteworthy.  Of the almost 3,000 towns in the Holy Roman Empire only about 12 to 15 (Cologne and Luebeck among them) had over 10,000 inhabitants.

A few statistics will serve as future points of reference.  There were only a handful of genuine metropolises in antiquity, among them imperial Rome in the 2nd century AD and Shanghai in the 8th.  In the Middle Ages this prodigious size is matched by Constantinople, Cordoba and Palermo, the last two of which may have been in the 500,000 range in the 13th-14th centuries.  Baghdad may have had as many as 1,000,000 inhabitants before it was destroyed by the Mongols in 1258.  Again, we have Chinese parallels for such phenomenal concentrations as Nanjing in the 15th century, and, in the late imperial era, Beijing, Suzhou and Canton.  Beijing remained the world's largest city until 1800, with a population of 2,000,000-3,000,000, when it was overtaken by London.  Its only close rivals in the I7th century were Istanbul, Agra, and Delhi.  Behind every enormous city of this sort, at least in the pre-industrial era, there lies a vast, centralized state. Without its ruler, the city is bound to wither or collapse.

B.  Cities come in clusters.  A town never exists unaccompanied by other towns.  It is therefore inevitably locked in an urban system, an urban hierarchy.  Even the lowliest of townlets has its dependent villages.  As Braudel puts it, "The town only exists as a town in relation to a form of life lower than its own . . . It has to dominate an empire, however tiny, in order to exist."  In China, the urban hierarchy was expressed by suffixes added to the names of towns, like fu for a town of the first order, chu for one lower down, and hieu for one lower still.  Similarly, there was a clearly defined urban hierarchy in Ottoman Anatolia of the 16th century headed by Istanbul, and descending through a number of regional centers of about 20,000 to 40,000 inhabitants each to two sets of lower-order towns of under 10,000 and under 5,000.

C.  Cities are places that have some physical circumscription, whether material or symbolic, to separate those who belong in the urban order from those who do not.  "Une ville sans mur n'est pas une ville" (a city without walls is not a city), J.-F. Sobry wrote in his De l'architecture of 1776.  Even without any physical circumscription, there is a legal perimeter within which restrictions and privileges apply.

D.  Cities are places where there is a specialized differentiation of work where people are priests or craftsmen or soldiers and where wealth is not equally distributed among the citizens.  These distinctions create social hierarchies:  the rich are more powerful than the poor; the priest is more important than the artisan.  Social heterogeneity is also axiomatic.  The urban population contains different ethnic groups, races, religions.  Even in ethnically homogeneous cities, as the original Yoruba cities were intended to be, there might be slaves or transient readers.

E.  Cities are places favored by a source of income, trade, intensive agriculture and the possibility of surplus food, a physical resource like a metal or a spring (Bath), a geomorphic resource like a natural harbor, or a human resource like a king.

F.  Cities are places that must rely on written records.  It is through writing that they will tally their goods, put down the laws that will govern the community, and establish title to property which is extremely important, because in the final analysis a city rests on a construct of ownership.

G.  Cities are places that are intimately engaged with their countryside, that have a territory that feeds them and which they protect and provide services for.  The separation of town and country, as we shall see repeatedly in this book, is thoroughly injudicious.  Roman towns do not exist apart from the centuriated land roundabout; great Italian communes like Florence and Siena could not exist without their contado; and the same is true of New England towns and their fields and commons.  Polis, civitas, commune, township -- all these are terms that apply to an urban settlement and its region.

Often the city-form is locked into rural systems of land division.  The Romans commonly correlated the main north-south and east-west coordinates of the centuriation, the division of rural land into squares that were supposed to be the theoretical equivalent of one hundred small holdings, with the cross-axes of the city.  The National Survey that regulated two-thirds of the United States territory determined the placement and size of many towns.  It is, furthermore, of great interest to us that pre-existing rural property bounds will often influence subsequent urban lines and determine the shape of urban development.

The question of which came first, town or country, is not simple. The first towns in the Middle East or China controlled and organized an already functioning countryside.  In the opening of the American West, the towns preceded the farms and made their operations possible.  By the same token, the strains of a deeply felt disagreement about the relative superiority of town and country can be heard throughout history.  Two examples, as distant as I can make them.  In China, the Confucian view that the proper function of the elite was to govern, that government presumed cities, and that the purpose of government was to civilize the countryside, clashes with the ultimately Taoist and Buddhist ideal of rural existence.  Thomas Jefferson's agrarian republicanism had no use for cities in the structure of the young nation. Cities were "sores on the body politic."

H.  Cities are places distinguished by some kind of monumental definition, that is, where the fabric is more than a blanket of residences.  This means a set of public buildings that give the city scale, and the citizenry landmarks of a common identity.  Technological monuments are also important:  Rome had its aqueducts; Tikal, a large manmade reservoir; Anuradhapura in Sri Lanka, a hydraulic system of monumental proportions.  In the public realm early cities under central authority chose to emphasize the palace and the temple.  In the people's city, the princely palace disappears, or is translated into a palace of the people, and the temple is "secularized." That is the case of the Greek polis, and that too is the case of the European commune in the Middle Ages with its palazzo pubblico or Rathaus and its "civic" cathedral.

I.  Finally, cities are places made up of buildings and people.  I agree with Kevin Lynch:  "City forms, their actual function, and the ideas and values that people attach to them make up a single phenomenon."  Hundreds of new towns in every age of history were still-born, or died young.  The majority of the grids laid out by railroad companies along their lines in the 19th century never fleshed out into real towns.  Conversely, we can discount scholarly claims for the fully established and long-lived Mayan sites or for places like Angkor Thom and Nakhon Pathom that they were not real cities because they had no resident population.  These spectacular ceremonial settings and the priests and the builders and the artisans and the people selling them things belonged together.  We will be well served, in reading this book to recognize that there have been cityless societies, and times when cities were vestigial marks in a predominantly rural landscape.  Let us recognize, too, that the urban and pastoral ways of life were at times contending social systems; and further, that the history of human settlement must be predicated on a rural-urban continuum, and that the city as a self-contained unit of analysis must be seen as conditional enterprise.  For all that, the city is one of the most remarkable, one of the most enduring of human artifacts and human institutions.  Its fascination is inevitable: its study is both duty and homage.

Ours is certainly not a story restricted to the past. At this very moment cities are being born ab ovo, either through the legal instrument of incorporation, or through parthenogenesis.  Since 1950 more than 30 new towns were created in England.  In France, along two  preferential axes following the Seine Valley, several new towns for 300,000 to 500,000 each are in the process of building -- St.-Quentin-en-Yvelines, Evry, Marne-la-Valle.  Others have been started outside the Paris region. In the Soviet Union we have many hundreds of recent new towns, closely associated with the spread of industrialization.  In the United States, the "new communities" program introduced in the Sixties has had its own considerable offspring.

It is a long haul, from Jericho to Marne-la-Valle.  My ambition to encompass all of urban history through this thematic approach rests on a paradox.  Cities in their physical aspect are stubbornly long-lived. As Vance put it, "the most enduring feature of the city is its physical build, which remains with remarkable persistence, gaining increments that are responsive to the most recent economic demand and reflective of the latest stylistic vogue, but conserving evidence of past urban culture for present and future generations."  At the same time "urban society changes more than any other human grouping, economic innovation comes usually most rapidly and boldly in cities, immigration aims first at the urban core forcing upon cities the critical role of acculturating refugees from many countrysides, and the winds of intellectual advance blow strong in cities. . ."

The challenge in this book and its companion volume will be to seize upon and reconcile this vital contest between socio-economic change and the persistence of the artifact.