Page:American Journal of Sociology Volume 15.djvu/555

THE DEFINITION OF THE CITY 541 All these definitions whether simple or complex, have certain vices in common. They rest upon characteristics which are too special, many of which are also superficial and secondary, and many more are not universal but apply to certain types of cities only. The definition should partake of a sociological character, which ought to present the following qualities :

1. It ought to be universal (constant), i. e., be common to all the types of cities. This condition goes without discussion.

2. It ought to be uniform (fixe), i. e., it ought to be found everywhere relatively unchanged, and vary as little as possible in degree. For a modality which, while existing in the various forms of its object, varies among these forms, would not answer to the purpose of the definition, which is to allow easy recognition of the object defined, and to enable it to be distinguished at a glance.

It is seen at once that the characteristic in question ought not to be a functional characteristic; no characteristic of this kind can be universal and uniform. Specialization in activity, considered abstractly and in itself, is true of only certain types of cities. It is wanting in what has been called "urban industry" (économie urbaine). And likewise the concrete nature of the activity varies greatly according to the cities specialized (commercial cities, industrial cities, or even those at different times dependent upon different industries, etc.).

The criteria which we employ ought then to be of the morphological order; and since we have already eliminated certain criteria of that type, the field of our choice will be rather limited. Neither the territory covered by the establishment, nor the condition of things which it implies (rampart, construction of the houses) can suffice for our needs. The external form of the aggregation is not any more specific. There is left to us then only the characteristics relative to its internal structure. The