The hierarchy is then like a pyramid; the few large and complex cities are at the top, and the many smaller simpler ones are at the bottom. There are always more smaller cities than larger ones. One can envisage, say, seven-level hierarchy where the complexity of cities increases as one rises in the pyramid. When a spatial dimension is added to the hierarchy, it becomes clear that a spatial system of metropolitan centres, large cities, small cities and towns exists. Goods, services, communications and people flow up and down the hierarchy according to the location of the urban areas.
The few high-level metropolitan complexes provided specialized functions for larger regions, while the smaller cities serve smaller districts. It may be pertinent to mention that the cities interact with the area around them, but since cities of the same level provide roughly the same services, they tend not to serve each other unless each provides some very specialized activity, such as housing the political capital of a region or a major university. Thus, the cities of given level in the hierarchy are not independent but interrelated with cities of other levels in the hierarchy. Together, all cities at all levels in the hierarchy constitute an urban system.
The concept of Rank-Size Rule was developed by G.K. Zipf in 1949. The observation that there are many more small than large cities within an urban system (‘the larger the fewer’) is a statement about hierarchy. For many large countries of great regional diversity and advanced economy, the city size hierarchy is summarized by the rank-size rule.
It tells us that the ninth largest city of a national system of cities will be 1/n the size of the largest city. That is, the second largest settlement will be half the size of the largest, and the 10th biggest city will be 1/10th the size of the largest. The rank-size ordering may describe the size patterning of cities in complex economies where urban history is long and urbanizing forces are many and widely distributed. Although no National City system exactly meets the requirements of the rank-size rule that of the Russia or the USA closely approximates it. The rule is less applicable to countries with developing economies or where the urban size hierarchy has been distorted through concentration of functions in a single paramount centre. In some countries, the urban system is dominated by a primate city, one that is far more than twice the size (and therefore functional complexity) of the second-ranked city.
In fact, there may be no clearly recognized ‘second city’ at all, for a characteristic of a primate city hierarchy is one very large city, few or no intermediate-sized cities, and many subordinate smaller urban settlements. The capital cities of many developing countries display that kind of overwhelming primacy. In part, their primate city pattern is heritage of their colonial past when economic development, colonial administration and trade activities were concentrated at a single point. Kenya, Mozambique, Senegal and several other African nations are examples. In other instances, Egypt and Mexico, for example, development and population growth have tended to concentrate disproportionately in a capital city whose very size attracts further development and growth. The cities of Srinagar in the valley of Kashmir and Shillong in Meghalaya are the typical examples of primate cities. Where central places dominate, of course, the urban size hierarchy will be one of a relatively few classes of cities with distinct size differences separating the classes. Land uses and populations they house.
An understanding of the nature of cities is incomplete without knowledge of their internal characteristics. In those characteristics, however, the world’s cities are marked more by regional variation than by global uniformity. Like other built environments, settlements reflect not only their own histories but also the cultures and technologies of the societies that create them.
The concentric zone and sector models assume urban growth and development outward from a single central core, the site of original urban settlement that later developed into the CBD. These ‘single-node’ models are countered by a multiple-nuclei model which postulates that large cities develop by peripheral spread from several nodes of growth, not just one. Individual nuclei of special function commercial, industrial, port, residential are originally developed in response to the benefits accruing from the spatial association of like activities.
Peripheral expansion of the separate nuclei eventually leads to coalescence and the juxtaposition of incompatible land uses along the lines of juncture. Therefore, the urban land use pattern is not regularly structured from a single centre in a sequence of circles or a series of sectors but based on separately expanding clusters of contrasting activities.