The residential differentiation of the urban population, a function of the axes of social differentiation extant in the society concerned, is reflected in a sifting and sorting of populations and locations. As the city develops typical patterns of differentiation become apparent. Different areas become associated with particular types of population and certain systematic relationships between geographical space and social space appear. The concern of the present chapter is with the spatial aspects of residential differentiation and, more particularly, with the validity of certain general models of this spatial structure which have appeared in the literature.

Discussions of the spatial aspects of urban structure generally concern themselves with three general models of urban form: the zonal, the sectoral, and the multiple nuclei. In the present case only the two former approaches will be examined. In contrast to the multiple nuclei model both the zonal and the sectoral analogies are concerned with the structural connotations of a particular set of differentiating processes, predict particular patterns of residential differentiation, and lend themselves readily to empirical test. The multiple nuclei ‘theory’ may be regarded as a caveat to the more general zonal and sectoral models.

THE ZONAL MODEL OF URBAN GROWTH AND STRUCTURE

The concern of the early Chicago ecologists with the differences in environment and in behaviour between different parts of the city led not only to descriptive studies of the ways of life to be found in particular natural areas, but also to a concern with the general features of urban structure.

The neighbourhoods which comprise the residential fabric of the city and form the framework for analyses of the relationship between locality and behaviour differ on innumerable grounds. In terms of physical structure they differ in density and age of development, in geographical position, and in types of dwellings. More importantly, from a behavioural perspective, their populations differ in age and sex composition, in occupations, incomes, and styles of life, in political, ethnic, and religious allegiances, and in a wide range of attitudes and behaviours. Each of these differences contributes its portion to the patterning of the urban residential system. In any attempt to understand this system it is necessary that each of these diverse differences should in some way be taken into account. The measurement problems which this involves form the major preoccupation of the present chapter.

UNIVERSES OF CONTENT, INDICANTS AND PROPERTIES

The conceptual tools which form the basic instruments of scientific endeavour may be classified under many rubrics. For our present purposes it is only necessary to distinguish between three such constructs: universes of content, indicants, and properties. In combination these three provide the basic tools for measurement and classification.

The universe of content

The universe of content forms the subject-matter of an analysis. It consists of the set of elements to which the procedures of measurement or classification are to be applied.

RESIDENTIAL DIFFERENTIATION

The urban community is neither an undifferentiated mass nor a haphazard collection of buildings and people. In the residential differentiation of the city the urban fabric comes to resemble a ‘mosaic of social worlds’. Similar populations cluster together and come to characterize their areas. As Park put it:

The residential differentiation of the urban population takes place in terms of many attributes and in many ways. Almost any criterion which can be used for differentiating between individuals and groups may become the basis for their physical separation. The process of separation may be accomplished through force, through a variety of sanctions, through a voluntary aggregation designed as a defence against unfamiliar ideas or customs or as an escape from persecution and discrimination, and through a selection of market forces. In much early town planning residential differentiation and segregation appear as prime characteristics.

It is characteristic of much work in urban ecology that while the fact of urban expansion is taken for granted, little attention is paid to other aspects of social or cultural change taking place in the wider society. All too frequently the urban community is taken as a phenomenon sui generis and considered in isolation. The hallmark of the macro-social approach to the study of residential differentiation within the city is that it attempts a rapprochement between urban ecology and the wider study of socio-cultural change. In the process an attempt is made to relate the bases of residential differentiation to the process of modernization.

Classical human ecology is a product of early twentieth-century Chicago. It is perhaps more than coincidental that the initial stimulus to the macrosocial approach to urban phenomena is a product of the burgeoning midtwentieth- century cities of the American West Coast, cities which have frequently been hailed as the harbingers of the ‘new urbanization’. There is an implicit assumption in much of the work produced under the aegis of classical human ecology that in some way the Chicago of the 1920s represents the typical case of the urban community. The experiences of the new communities of the West Coast and the growth of a comparative perspective in the social sciences, a development greatly speeded by the Second World War, have undermined the simplistic assumption that any single city can be typical of urban society and have caused a more searching inquiry to be made of the socio-cultural matrix in which urban communities are embedded.

Well-supported theories are still in short supply in the social sciences. Observations and hunches, on the other hand, are plentiful. The task of bringing theory and observation together is rendered difficult by the imprecision of many theoretical statements in the social scientific literature and by the dubious validity of many of the observations. It will be all too apparent that both faults are characteristic of much of the argument which forms the substance of the present work. Rather than an integrated theory of the residential differentiation of the urban population, complete with supporting evidence, what is presented should be regarded more as a series of essays towards such a goal. The generalizations offered frequently go beyond the evidence. This is particularly the case when attention is paid to the non-Western or pre-modern city. In the present state of the social sciences, however, it is felt that the dangers of over-generalization are less pronounced than are those of under-generalization. If nothing else, it is hoped that the essays will throw up a number of useful ‘Aunt Sallys’.

An overall summary of the book is provided in a brief last chapter. Chapters 1 and 2 are concerned with delimiting the nature and significance of residential differentiation. In the first chapter the notions of neighbourhood and of neighbourhood differences are explored in terms of a variety of human behaviours.

Like the society in which it exists, the modern city is highly differentiated. Different parts of the city are associated with different populations, with different opportunity structures, and with different reputations. The geographical framework of the city provides the basis for the emergence of a mosaic of social worlds. The increasing movement characteristic of modern society has almost certainly lessened the salience of location in the day-to-day lives of city-dwellers, but it remains the case that residence in one part of the city rather than in another has implication for a wide range of behaviours and biographies. The effects of location may be expected to be most pronounced on those whose daily movements are more or less confined to the bounds of their immediate neighbourhood – the young, the old, and the ‘care-takers’ – but the role of the local community in the initial socialization process and in the provision of a reference for social comparison purposes, ensures that the influence of the ‘neighbourhood effect’ is felt across a variety of activities and groups. Different populations have different relationships to their locale and areas which are suitable for one group may be quite unsuitable for another. The diversity of urban life and society both reflect and demand a diversified community.

The application of factor analytic techniques to urban residential differentiation has produced relatively consistent results. It seems rather well established that, at least in the Western world, much of the detailed variation in the characteristics of urban sub-communities may be interpreted in terms of three or four underlying constructs relating to differences in socio-economic status, in family composition, in ethnicity and in mobility. The demonstration of factorial invariance, however, is not to be equated with an explanation of factor structures. Having established the empirical validity of the major axes of residential differentiation the next step is to attempt the explanation of their significance and to examine the relationship they exhibit to other facets of human behaviour and social structure. It is to this task that the attention of this and of the next chapter are directed.

As is true in the case of many other sociological phenomena, the explanation of ecological structure may be attempted at either of two levels: a micro-social and a macro-social. In the micro-social approach attention is focused on the relationship between residential differentiation and patterns of individual decisions and behaviour. In the macro-social approach attention is focused on the relationship between residential differentiation and certain global characteristics of the encompassing society. A satisfactory theory of residential differentiation must include both approaches.

THIS book is an attempt at combining a record of exploration with the teaching of a little elementary physiography. Facts newly observed and theories of my own are also woven into the general description of a characteristic region. May the difficulties of the experiment in some measure reconcile the general reader, desirous of hearing something interesting or amusing, and the critical scholar in search of information. The style of the book sufficiently accounts for the omission of an incubus of footnotes and literary references. I have absorbed the writings of others, not quoted them. The bibliographical list at the end will, I hope, form the necessary acknowledgment of intellectual property. Scientific discussion has been reserved for the Appendix, where I have laid stress on the intricacies of many problems, also drawing attention to various questions to be cleared up by future travellers. The student of physical geography for whom this book has chiefly been written need not despair when faced by frequent doubts. Scepticism tempered with enthusiasm is the attitude of true science.

I express my deepest gratitude to the Council of the Royal Geographical Society and to the Syndics of the University Press who, by their munificent help, have both enabled me to publish this work. Nor can I forget that invaluable guide behind the scenes, the Corrector.

My sincere thanks are due to the Imperial Russian Government and its political representatives, as well as to His Royal Highness the Amir of Bokhara, for permission to travel in their dominions and for the invaluable assistance they always gave me. I also include many kind friends and helpers in Russian Turkestan and the Caucasus.

Wishing to draw a picture of the Zarafshan valley in its unbroken length, I have kept until now our fortnight's excursion to the Fan district or Hazrat-sultan, as this part of the Hissar range is sometimes called (see Maps 110 and 115). We left Varziminar on the 26th of July, and soon found ourselves on the dangerous cornice-paths of the Fan-darya. The Russian soldier, fond of pet names even for pet aversions, has given the name of “balkonchiki” to these giddy ledges. All bridges and paths are emergency structures, for our friend the native only builds emergency roads and emergency houses, his whole life being evidently one long emergency in view of a more permanent state in paradise. The “balconettes” are characteristic of most Asiatic mountain roads, and reflect great credit upon the improvising skill of the inhabitants, having been made without the use of a single ounce of powder for blasting or a single inch of rope for tying. We have more eye for their defects, but explorers and Cossacks have horses, whereas the humble pedestrian praises Allah for having caused such a comfortable passage to be made by man. Steep rock walls rise from the foaming torrent and along their face runs the narrow shelf stuck together out of crooked sticks and rubble. Twisted trees and branches are jammed into clefts or supported by friction on pads of brushwood ; little walls are raised on tiny ledges, or alternate layers of blocks and fagots formed into a coping. On this projecting scaffold is spread a mixture of stones and bits of wood, the surface of the overhanging road.

We left Pamak on the 2nd of October, proceeding down the Yakhsu, where numerous kishlaks line the great diluvial terraces on either bank. The openings of side valleys on the left (south) ever disclose new forms and arrangements of conglomeratic fancy. Below Saripul the spurs of conglomerate begin to thin out, and near Sinji there is a last revelry in the shape of a comb of cylinders and slender finials. Here I shall describe an excursion which my wife and I made to the province of Kuliab towards the end of October, 1898. Below Saripul we struck out for a region of wavy hills among which lies the provincial town of Muminabad. When some years later I saw the loess landscape of the Kaiserstuhl and Kenzingen in the Rhine valley I was much impressed by the close and remarkable likeness it bore to the yellow slopes of Eastern Bokhara. There is the same general aspect of rounded swellings dotted with trees singly or in files, amid dry fields and powdered vineyards (compare Fig. 185). There are the same tray-like terraces, water cuts, and dust; and there is also in proud evidence the lime-loving walnut-tree, which may be called a characteristic denizen of loess hills and dry mountain slopes. It is a noteworthy fact that it avoids the plains, perhaps because they are too moist, or because it cannot gain a footing in the thick and spongy deposits of the irrigated country. It may also be that, as a wild tree, it has long ago been cut down in the lowlands, and is not cultivated owing to the slowness of its growth.