Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 82, 1954-55
This text is also available in PDF
(825 KB) Opens in new window
– 913 –

Section F—Chairman's Address
Rural Settlement, with Examples from Canterbury and Southland.

The systematic study of settlement is one of the recognised fields for geographic research. During the past half century a substantial literature and a well-developed discipline have been achieved in respect of urban settlement studies, but in the related, though distinct study of rural settlement, there is still a relatively sparse literature and no general measure of agreement on classifications; attempts at the formulation of general laws, or even generalizations, have been followed by the discovery of so many exceptions that one is forced to reflect upon the complexity of the subject, and regret the lack of an adequate assembly of material to date.

The founder of the modern French school of Geography, Vidal de la Blache, made many inspiring and fruitful initial contributions to the geography of population and settlement, including that of rural settlement. He noted particularly that regions of long-established settlement exhibit the most complex and composite picture in this respect, and that newly settled regions (amongst which he was familiar with North America), tended to show no range of settlement types comparable to those developed through successive stages of occupance in European countries. Villages in Europe, he noted, commonly occur in series, as along lines of contact between physically contrasting units of the landscape. He concluded that “man is a being of habit, rather than of initiative,”1 and he saw this reflected in the types of settlement established. He believed that group-types of settlement, both rural and urban, are constantly evolving, and says pertinently “It would be most interesting to trace this evolution, not only in Mediterranean countries, where it can be observed at the present time, but in recently settled areas like America and the temperate zones of the southern hemisphere, where it is just beginning.…” We in New Zealand have had a number of studies produced dealing with urban forms, but very few studies on the far more numerous, and perhaps lowlier forms, of rural group-settlement.

Since that early period of modern geographic study, considerable interest has been shown by geographers in the subject of the rural population and its groupings, but an enormous amount remains to be done before it can be said that the world patterns of rural settlement are adequately known. There is as yet insufficient reliable information for a discussion of settlement types in terms of world distributions. It is to draw attention to this subject as a worthy topic for geographic investigation in New Zealand that I have chosen the subject of this address.

[Footnote] 1 P. Vidal de la Blanche “Principles of Human Geography” N. York, 1926.

– 914 –

At the end of the nineteenth century, Meitzen's exhaustive treatise,2 published in 1895, considered rural settlement forms in the German lands. In this work, he attempted to correlate rural types of settlement with ethnic groupings—in particular the agglomerated village with the Germanic, and the isolated dwelling with the Celtic mode of life. Such a simplified explanation was at once criticised, notably by historians, and since that time it has become generally recognised that nucleation or dispersion is a far more complex phenomenon; Meitzen's theories are discredited, although ethnic influences cannot be entirely ruled out as irrelevant amongst the many significant influences to be noticed upon rural settlement forms.

Jean Brunhes, in his “Human Geography,” the first edition of which was published in 1910, devoted much space to a consideration of rural settlement forms, reviewing individual house-types. settlement types, and their patterns of distribution. He called for first-hand investigation rather than a search for classifications and laws,—” It is better to describe dissimilar, but co-existing types,—thus explaining what it is that actually makes up the synthesis of regional life.”3 He pointed out the value of topographical maps, as the most exact and faithful representation of the distribution of the population in all its details, and showed how such maps, “fragments of our soil,” illustrate the extraordinary variety of human adaptation, sprung from long centuries of history. “It is not the houses alone that need consideration,” he stated, “nor even the towns; it is the pattern, the ‘fundamental seed-bed’ of human settlement, a more or less closely woven network on which is built up the entire range of centres of population” He noted that no one country or portion of a country, possesses a single type of settlement, although there are rare cases where this situation is approached—as in the fertile loess plain of Alsace around Strasburg, where all the houses crowd into villages, forming small islands in an uninhabited countryside; an association of several types is more generally characteristic of an area,—as in Limousin, where small villages are interspersed with scattered farms, or in Flanders where an extreme form of dispersal of dwellings is evident, and though many communes have no real concentration of houses, there are small centres to be seen relatively far apart. He commented that “there are really no isolated houses—‘einzelhöfe’—in our very crowded civilization, but only houses more or less far apart.” There has since been a great deal of controversy, over the criteria to be adopted in measuring dispersal and agglomeration in the field; some standard formula has been sought by several investigators, notably by Demangeon, but no mathematical yardstick has proved widely acceptable, and the conclusion is forced upon one that each regional pattern must provide in its own terms, the key to definition of nucleation or dispersal.

Demangeon has perhaps been the greatest scholar of rural population, and produced a long succession of publications on the subject, chiefly in the “Annales de Géographie,” from 1920, when he first proposed a classification of the principal types of rural habitations in France, until the outbreak of War in 1939. He looked to agricultural function as the key to rural settlement classification, since the rural dwelling is primarily a phenomenon of the

[Footnote] 2 A. Meitzen. “Siedelung und Agrarwesen der Westgermanen und Ostgermanen.” 1895. Berlin.

[Footnote] 3 J. Brunhes. “Human Geography.” Trans. from Fr. Edition by E. F. Row. London, 1952.

– 915 –

agricultural economy, and which the farmer regards as a working implement, adapted to the conditions under which he cultivates the soil. Demangeon's enthusiasm was largely responsible for the setting up of a Commission to investigate the subject, for the International Geographic Union, and the “Report of the Commission on Types of Rural Settlement” was published in 1928. This stimulated much interest amongst geographers, and has led to a considerable literature on some aspects of the subject. The historical approach was stressed in one of the earliest contributions—that of Ahlmann,4 who suggested, as a result of examples studied in Italy, Germany, Denmark, and Norway, that an historical succession of major stages of settlement might be recognized, as for instance, in the Mediterranean lands, where early settlements had been all strongly nucleated on well-defended, and healthy, hill-top sites, from which a later movement down to the plains took place, accompanied by a pattern of dispersal. More recently this led on to the development of towns on the plain, and the growth of a few large cities. All stages contribute to the total settlement pattern to-day.

Each of these major considerations has some force, and in any study of rural settlements, historical, ethno-cultural, and economic evidence must be taken into account in relation to the permissive factors of the physical landscape, and reviewed in respect of the stage of occupance reached in a region. The particular realities of the local setting must first be appreciated and settlement features mapped; in considering explanatory descriptions no isolated factors, whether of the physical conditions of the area, or of the cultural traditions of the people, are likely to be alone responsible for the patterns found. Pierre George has recently stressed,5 in a study of what he calls “the elemental distribution of population—i. e., the types and character of rural and urban settlement,” that “it is essential to view the subject as linked with modes of life, land utilisation and density and migration of population, in order to give a full understanding of the human pattern.”

Studies of settlement have inevitably been closely related to studies of population, and there has been a tendency in the last two decades to move away from the direct study of settlements in the landscape, as approached by Brunhes and Demangeon. towards the study of population distributions and densities. Brunhes's “Essential Facts” were the marks left by man upon the landscape, and as the visible aspects of the landscape, they represented the features associated with a particular population grouping; landscape was viewed as almost synonymous with geography. Brunhes examined the question of the distribution of population in connection with the house, the village, and the urban centre. On the other hand, most population studies have tended to be approached from another viewpoint—that of number distribution and densities. Numerical distribution of population lends itself particularly well to cartographic representation by standard symbols, leading to patterns of relative densities, and such examples of research are numerous. However, in a recent strongly-worded address to the Association of American Geographers,6 Trewartha deprecates this almost exclusive emphasis, and claims that geographical literature “has seriously slighted that other important aspect of distribution

[Footnote] 4 H. W. Ahlmann “The Geographic Study of Settlements” Geogr Rev. Vol. 18. 1928.

[Footnote] 5 Pierre George. “Geographie de Population et Demographie.” Population V. 1950.

[Footnote] 6 G. T. Trewartha “A Case for Population Geography.” A. A. A. G, June, 1953.

– 916 –

which seeks to understand the characteristic size groupings of population as exhibited by settlement units. It is this aspect of distribution,” he says, “which is most conspicuous when population is observed in the field rather than by means of the census.” Mapping should be concerned with absolute densities and not merely with relative density distributions. “It is my belief,” he continues, “that an analysis of population distribution in terms of settlement size and spacing is a neglected aspect of population geography worthy of serious attention.” Much of this work, he suggests, may take the form of representative locality studies.

This comparative neglect, pointed out by Trewartha, had also been noted from another point of view—that of planning—in Britain the year before, in a paper by Vince,7 a member of the Research Division of the Ministry of Housing and Local Government, who in reporting on a survey of the composition of the rural population of Britain, commented that the numerous questions posed by such a survey as his own, could not be solved without considering the form and pattern of settlement. He pointed out that outside geographical circles it is still too often assumed that the typical rural grouping in Britain is the nucleated village, and that this natural tendency towards centralization can be profitably accelerated, in order to facilitate the provision of the various amenities and services, so difficult to supply to rural areas because of the small numbers of consumers. “There is a clear and urgent need,” he concludes, “for the geographer's contribution to our knowledge of settlement patterns and groupings as a basis for rural planning.”

In his critique of the general approach to modern geography referred to above, Trewartha goes so far as to suggest that the entire field of population geography, of which settlement forms one aspect, has been neglected, and that this neglect constitutes a fundamental weakness in our attitude to the subject. While we remain wedded to the idea that geography is to be viewed as a subject organised in a two-fold structure—physical and cultural, population cannot be adequately considered. We are therefore invited to review our approach to our subject, and to think rather in terms of a three-fold approach, in which man is placed first, as the creator and originator of the cultural landscape, thus giving man his deservedly important position within the unitary geographical structure. Under such an organisation of the subject, settlement geography would take its place as one aspect of the field of population geography, within which we might study the historical processes involved in populating the earth to its present stage; the numbers, dynamics, and patterns of population to-day, population clusters,—their size and spacing in settlement units; migrations and movements of population; and also qualities of population, including physical (i.e., health) and social-economic qualities, viewed in regional patterns.

It is of interest to note a somewhat parallel position in New Zealand geography to date. Outside a comparatively small list of thesis topics for Honours degrees, scarcely any literature exists on the detailed rural patterns of settlement, and population groupings have been considered within the framework of wider investigations or holistic regional studies; some study has been made

[Footnote] 7 S. W. E. Vince. “Reflections on the Structure and Distribution of Rural Population in England and Wales, 1921–31.” Transactions and Papers, 1952. Institute of British Geographers. London.

– 917 –

by sociologists, as in “Littledene—a New Zealand Rural Community,” but it has not been dealt with as a deserving study in its own right. Amongst University of New Zealand theses, a number of titles indicate research into population and settlement per se, but only one such investigation has been devoted primarily to an enquiry into rural settlement forms over a considerable area—the three Northland counties of Rodney, Otamatea and Waitemata. Several enquiries into the development of areas originally covered by large runs, have included work on the successive stages of settlement to be recognised over the last near-century of occupation, but this theme has been subordinate to that of agricultural changes and development. Two theses on the Distribution of Population in North and South Canterbury considered their topics very much from the viewpoint criticised last year as inadequate by Trewartha, and further research is required in order to obtain precise information on the morphology of habitation and settlement forms, or to appreciate the characteristic details of settlement units.

It might be noted in passing that Australian geographers have profered little more, although a valuable contribution on the subject by Andrews in 19338 makes one wish that many more such investigations might appear. He was concerned in studying the settlement net in a part of rural New South Wales, where changes in the complex of influences had induced during the last generation, changes in settlement forms, notably a decrease in the importance of the small sized rural agglomerations, and an increase in the influence and importance of the bigger units. Many of the smaller nineteenth century townships were found to have lost a number of their former functions, their buildings were being rarely renewed, and an air of decadence pervaded them. He interpretated the situation at that time as one of incomplete adjustment to new cultural influences; further contemporary studies of this type would be welcome.

It is not altogether through lack of basic material that rural settlement in New Zealand has received little detailed attention to date. For almost all the more populous districts the one inch to one mile topographical maps provide an invaluable source of information as a foundation for cartographic work on the topic, though there is an unfortunately marked measure of variation from one sheet to another in respect of the details recorded on rural settlement. Revisions with the aid of the excellent air photographs now covering almost all the areas already mapped on the one inch scale, will be very much welcomed. Small hamlets and rural townships are commonly shaded by cross-hatching to represent built-up areas, where in fact very few houses are found to exist to-day; moreover there is no uniform standard for representing details of the farm-buildings, and some sheets indicate by only one symbol, the group of sheds, and shelters adjacent to the farms. Details may be seen in most cases on the local air photographs, but it is regrettable that only at the Head Office of the Lands and Survey Department in Wellington is there a full set of prints for New Zealand, as only the cover for each Land District is held at the other centres. Functional data can only be recorded as a result of field observations.

[Footnote] 8 J. Andrews. “The Settlement Net and the Regional Factor” Australian Geographer, Vol. 2, 1933–34.

– 918 –

The isolated farmstead is the most widely typical unit of rural settlement, and on the Southland and Canterbury Plains is represented in a great variety of individual forms. The same spirit of individualism that prompted the movement forwards isolated holdings, seems to be reflected in the development of such variety in the clusters of farm buildings to be observed. This is seldom evident from the topographical maps. The broad functional division between dairy farmsteads and sheep farms is a basic one, but within each division there is great variety, from the simple unit of house and cow-bail for dairying to the well-established groups of over half-a-dozen buildings clustered around the roughly rectangular sheltered enclosure to be found south of Edendale; the sheep farm may have only a small wooden wool-shed, or at the other extreme a large modern concrete one, set a little apart from a modern suburban-style house, surrounded by ornamental exotic trees and shrubs. New homes, in fact, were observed on several recent traverses across the Southland Plain, to be so frequent as to form a notable characteristic of settlement to-day—brick and tile homes, built to a good suburban bungalow design, standing away from the old wooden house in its tall shelter hedges, and the new home commanding in many cases, wide views and a sunny northern aspect; the old house commonly houses a second family unit on the property, and is vivid evidence of the processes that have been rapidly getting under way in the last decade towards intensification of farm production. Since 1920, this audience will scarcely need to be reminded, there has been little increase in total acreage of agricultural land in New Zealand, and the substantial increases in production have been due rather to intensification of output on existing developed land. It appears that in Southland, outside the southern dairying districts, the point is now being reached at which further intensification requires the addition of further labour units on established holdings. Such new homes are less frequent on the smaller dairy holdings of the southern parts of the Plain. In the Canterbury landscapes, such new farmsteads were not observed; among the older farms of both regions, several house designs recur again and again.

The first indication of accretion in the settlement pattern is afforded by the small hamlet. Such groupings are not recorded separately in census returns, but are included in the figures given for the wider surrounding localities, but they are nevertheless significant, settlement forms. One or two farm homes may be included, with school, hall and store, the latter frequently providing a multiplicity of services, including post office and petrol pump. Many such units are noteworthy before achieving the minimum size suggested by Trewartha as the criterion for North American study—i. e., six functionally active units, including at least two non-farm units; the spacing of such buildings must give an appearance of compactness exceeding that of the ordinary farmstead spacing. Southland has, in its southern dairying districts, the nucleus of a hamlet in its frequent cheese factories and the variety of settlement form linked with these is great; the hamlet may comprise the factory, manager's home and two smaller homes for his assistants, as at Otara. From this minimum, one may observe successive stages of accretion. The number of cheese factories, however, has declined in recent years, partly following the trend from dairying to fat lambing, and partly because the close pattern of such establishments reflected the need for ready access from farm to factory in horse-transport days; as a result a number of deserted factories and hamlet sites is to be seen.

– 919 –

At Titiroa, a local transport company is using the old factory building as shelter and garage. Some hamlets, with their small factories, are definitely declining, while others are being extended; a more widely-spaced pattern of larger, more efficient units, is coming into being. The Canterbury Plains have no corresponding feature inducing accretion, and at a number of nodal points, hamlets have been reduced to virtual non-existence, as at Charing Cross, the junction of nine roads, but possessing only its school, school-house, and one farm to-day. A generation ago, there was a need for closely-spaced provision of basic services; to-day the farmer prefers to drive to the nearest township. However, no general rule can be deduced, as there are thriving hamlets to be observed, as at Lochiel, in Southland, even quite close to Invercargill, while others, like Thornbury, (a small township with initial advantages such as its railway and distance from the city) take on an air of decay with abandoned shops and empty store buildings. Other factors beside mere distance and local economic needs, evidently play a part in reaching the results seen in the landscape to-day.

This smallest group-settlement—the hamlet—appears to-day to be reduced to a core of basic cultural and servicing units, i.e., general store, school, garage, church and hall. It is unlikely to expand, except where some special local activity is present—such as that demanded by main road motorists in search of refreshments and services.

There is no clear-cut division between the hamlet and the next group-settlement, the small township. Many similar influences have been brought to bear over the past 25 years upon this form; in some cases to-day, the same activities that were noted along the main roads through hamlets, have helped to develop or to maintain the small township. A succession of prosperous and growing townships is to be observed from Timaru to Christchurch, indicating the concentration of rural services along this axis, which also concentrates the more elaborate services provided by the string of urban centres sited along the same road. Modern means of transport has eliminated the need for a second line of townships on the inner margin of the plains, and those away from the major road are, as a group, showing relative decline. Even a local county administrative centre, like Balcairn, has lost most of its other functions as a township. The somewhat similar situation of townships between Gore and Invercargill is noteworthy and several prosperous settlements are tending to show a new focus of settlement by the roadside as well as an older one, often beside the railway, with its earlier shops.

Early attempts to found village settlements, on the old British model, have generally failed; several sites in Canterbury and Southland were built up as nuclei for labourers' homes, in the hope of providing a pool of labour for local farms; this form of economy did not take root. Barrhill, west of Rakaia, was laid out to serve the Corwar estate in this way, with a planned octagon, oak-lined roads, and the whole settlement surrounded by other English trees. The oaks shelter a small school and church to-day, but they are alone in the octagon; two small farmhouses are occupied in the all-but deserted shady, grassy “streets”. Peel Forest, with its milling industry, on the western edge of the Plains, and with camping attractions more recently to augment its station work, has survived as a village unit—a rare New Zealand settlement type.

However, the problem of a continued increase in farm productivity poses the question of the need for more rural labour There is a tendency for farmers

– 920 –

to prefer full mechanization plus the use of casual and contract labour for particular seasonal tasks. This does not favour full-time farm labourer accommodation on holdings, but a revival in some modern form of village-units, offering some social life to the members of such labour pools in rural areas, has been advocated in some quarters. If New Zealand's third million by 1975 or thereabouts, is to be achieved without any fall in general living standards, agricultural production must be also increased by a similar rate—probably by about 2 per cent. per annum. Although such an increase will not require a proportionate increase in agricultural labour, nevertheless an increase of considerable dimensions in the agricultural population will be vital, and the associated problem of rural settlement needs to be considered therefore, as a point of national policy.

A parallel phenomenon of early settlement—the large station, which employed its own skilled tradesmen, is also no longer to be seen in its earlier form as a unit of settlement. The “big house” survives in many areas, and contributed a noteworthy feature of the landscape in both Southland and Canterbury; a few “cottages” used by estate labourers, lend an associated air of landlord-and-tenant economy to some localities, while in fact this relationship no longer exists.

Even the larger townships, with populations above approximately 400, up to 1,000, are giving way in relative rate of development to the larger urban centres. Few shops in such townships are to be seen undergoing expansion; some reconstruct their frontages, to give a more modern air to the small “commercial core” which is evident in the plan of townships of this size. They do, however, appear to be holding their own economically, and several new motor-servicing establishments are to be noted in most such settlements to-day, with a few new houses of recent construction. The new era of motor transport is showing that there is an economic demand for a fairly wide range of services in such townships some 25 to 30 miles apart on the Canterbury Plains, and new building is to be noted at such centres as Amberley, Leeston and Darfield, while Methven and Rakaia townships exhibit features that tell a story of economic progress also. The Southland centres of Wyndham and Otautau, correspond in this category.

Neither Plain is extensive enough to show development towards the theoretical hexigons suggested in Crystaller's diagrams of rural and urban settlement patterns.

Apart from the normal servicing centres discussed, there are also special functions to be observed in several instances. An example of a railway settlement is provided by one of the two foci in Springfield, where cars are loaded and trains set off for the mountain section of the West Coast line from Christchurch. Springfield also has the modern advantage of main road situation and exhibits both early settlement and modern settlement features along the road. In Southland, rural settlement types include townships attached to limeworks, such as at Browns, north of Winton. To east and west of the Southland plain, at its bush fringes, milling settlements are to be observed, with characteristic pioneering features and unkempt appearance. Timber townships are essentially ephemeral, though characteristic of a few districts, and therefore geographically significant.

– 921 –

It is by means of cartographic recording in the field that geographic investigation of such settlement types as those outlined must proceed, perhaps in sample areas. It is a somewhat time-consuming task, but no alternative all-embracing source of information is to be found. Census reports include all small rural hamlets in the totals of population for the surrounding locality, and no guide to the true extent of nucleation of settlement can be found from this source. By seeking to recognise the areal extent of combinations of settlement types, such as those outlined and illustrated* in this address from the several field traverses undertaken, conclusions of geographic value should be reached. The assessment of settlement data within a regional context must be the aim. Such conclusions, and the data obtained in the field, are not only of value to academic geographers, in their search for knowledge on the forms of the occupation of the earth's surface, but also will be of value to others concerned with the economic progress of New Zealand's rural communities. Although perhaps 70 per cent. of New Zealanders now dwell in urban or suburban surroundings, it is nevertheless upon the economic well-being of the rural areas that the national economy fundamentally rests.

[Footnote] * A series of slides illustrated the settlement types mentioned in Canterbury and Southland.