South Italian Agro-Towns

Anton Blok

Comparative Studies in Society and History, Vol. 11, No. 2. (Apr., 1969), pp. 121-135.

Stable URL:

http://links.jstor.org/sici?sici=0010-4175%28196904%2911%3A2%3C121%3ASIA%3E2.0.CO%3B2-M

Comparative Studies in Society and History is currently published by Cambridge University Press.

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available athttp://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtainedprior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content inthe JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained athttp://www.jstor.org/journals/cup.html.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printedpage of such transmission.

The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academicjournals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers,and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community takeadvantage of advances in technology. For more information regarding JSTOR, please contact support@jstor.org.

http://www.jstor.orgThu Dec 13 15:27:24 2007

South Italian Agro-Towns" A N T O N B L O K University of Amsterdam

I

It has been observed by various authors that peasant agglomerations,' the so-called 'agro-towns' that may number several thousand inhabitants, are fairly common in southern Italy. This pattern, which geographically separates the people from their land, is not restricted to the south of Italy; it is typical for most countries along the Mediterranean.

What kind of explanations are in general presented for the appearance of this pattern? In most cases the authors confine themselves to a distinction between nucleated and scattered settlements and enumerate the various factors which are supposed to explain the difference between them. For example, Dickinson and Rochefort, both of whom carried out research in southern Italy recently, discuss the problem rather superficially (Dickin- son, 1955: passim; Rochefort, 1961: 101-3). It is true they both mention some conditions such as insecurity in the countryside, the organization of large estates, the prevalence of malaria, and the scarcity of drinking-water : conditions that they hold responsible for the presence of the peasant agglomeration. But they hardly treat these factors systematically. This is quite remarkable, because as early as 1927 the French geographer Demangeon pointed out in an instructive theoretical treatise on the subject the strongly variable character of these conditions. What we lack in most geographical descriptions and explanations of this settlement pattern is a distinction between historical, functional, and causal explanations. The purpose of this essay is to look for valid propositions concerning the nucleated settlement pattern in southern Italy and its natural and human

* Revised version of an Anthropology M.A. thesis (title: Enige aspecten van het geconcen- treerde nederzettingspatroon in Zuid-Italie) presented at the University of Amsterdam, November 1964. Although mainly based on literature, on some points I have drawn from my old field data gathered in western Sicily during the second half of 1961. I am indebted to the Stichting Morpurgo Studiereisfonds Italie and the Dutch Dolci-Committee for financial aid. Thanks are due to Professor A. J. F. Kobben for his critical suggestions. I owe a special debt to Professor J. F.Boissevain wh? corrected my English, a language not my own.

For the rather unwieldy term, nucleated settlement pattern', I use the brief term, 'agglomeration'.

I 2 2 ANTON B L O K

environment. The fact that scattered settlement is not entirely absent in this area may be useful in the comparisons that necessarily precede the formulation of generalizations. Here I shall consider the settlement pattern as the dependent variable and the environment as the independent variable.

Demangeon is one of the few geographers who tries systematically to verify a number of generalizations put forward by various authors in the course of time. He distinguishes between two main types of settlement: agglomera- tion and dispersed settlement, although he acknowledges that social reality is too complex to be satisfactorily expressed in only two terms. With emphasis he points out that in fact we are dealing here with a con- tinuum: 'Between the two terms, agglomeration and dispersion, reality shows us intermediary forms; between the village and the farm there is the hamlet' (Demangeon, 1927 : 112). Consequently, he distinguishes a number of sub-types with the ratio of agricultural land to agglomeration of houses as a principle. One of these sub-types is 'le village A champs dissociis', a pattern we find predominantly in southern Italy. Demangeon describes this 'extreme type' as follows:

But the most curious aspects of the nucleated settlement pattern are to be found in southern Italy and in Sicily; the peasants crowd in real towns which sometimes, as in Apulia, reach tens of thousands of inhabitants, such as Canosa, Andria, Corata, Biltonto, or as in Sicily, Caltanissetta and Caltagirone: large rural agglomerations swarming with life amidst the empty and desolate country-side. To reach the fields, the peasants must sometimes cover distances between twenty and thirty kilometers; often they must pass the week far from their homes, only to return on Sunday: example of a paradoxical settlement which alienates the agriculturalist from the land he cultivates (Demangeon, 1927: 4).

In his endeavour to explain the difference between agglomeration and dispersion, Demangeon distinguishes the following categories of factors: (a) physical geographical (relief, soil, hydrology); (b) social (conditions of security, health); (c) rural economic (agrarian system, mode of cultiva- tion). One by one he checks to what extent a number of dominant con- ceptions in the geographical literature may be valid. Through systematic comparisons he concludes that the next propositions (here presented summarily in the form of 'if A, then B' relations) cannot be accepted as true generalizations: in these forms they appear to be pseudo-generaliza- tions.

(1) if plains, then agglomeration (2) if mountains, then dispersion (3) if marshland, then dispersion (4) if scarcity of drinking-water, then agglomeration (5) if wide availability of drinking-water, then dispersion

S O U T H ITALIAN A G R O - T O W N S I23

(6) if insecurity, then agglomeration (7) if large estates, then agglomeration (8) if hydro-agriculture, then agglomeration

In not one of these cases is it permissible to speak of 'imperative rules'. No one of these factors is individually responsible for the appearance of certain settlement forms. In this respect the number of exceptions is too large. Demangeon concludes the first part of his essay by pointing out that for an explanation of any settlement pattern, we must consider an extensive range of variables :

To explain how certain people are used to living in agglomerations and certain others on isolated farms or in small hamlets, we must take into account all natural, social and agricultural conditions (Demangeon, 1927 :23).

Demangeon does not arrive at final statements; such has not been his intention. He wants only to present a preliminary classification and to suggest some arguments for discussion (1927: 2). In the next sections I shall attempt to elaborate some of his points, restricting myself to the south Italian experience.

Although in various south Italian areas the peasant population lives concentrated around scarce spring-water, we must be careful in relating the two phenomena causally. Demangeon points to the fact that in some areas with the same hydrological structure different settlement patterns are to be found. On the other hand, Ahlmann and Semple ascribe to the subtropical Mediterranean climate with its long, hot, and dry summers, on one hand, and to the soil conditions, on the other, a fundamental significance for the dominant nucleated settlement pattern.

Climatic conditions, expressed in the rarity of a reliable supply of drinking-water, forbade in general the isolated farm, just as they do today. . . . But where the springs were more sparsely scattered, population had to concentrate about them in fewer but larger groups, no matter how far might be the way from the home village to the out- lying fields. . . . The longer the summer drought and the dryer the land, the sparser the villages in general (Semple, 1932: 539-40).

About Apulia, Ahlmann observes that the reason for the extreme concen- tration of the peasant population must be found (here as in Sicily and the rest of southern Italy) in the first place in the scarcity of water. He does mention other factors as well, but he considers them only as contributing agents and not of crucial importance (Ahlmann, 1926: 113-15). This physical-geographical determinism does not seem to fit his credo of evolutionism, according to which he postulates the oldest settlements must have been agglomerations (Ahlmann, 1925: 259-60). Hence the Swedish geographer calls the nucleated patt-ern predominating in southern

Italy 'primitive' (Ahlmann, 1926: 108). He states that only during a later stage would dispersion develop. Nota bene: we are here concerned with subsequent agglomeration and dispersion in the same physical environ- ment. No more are Ahlmann's ideas in agreement with the archaeological data about the prehistoric nuraghi in Sardinia, of which the oldest have been found in extreme dispersion (Le Lannou, 1938: 103).' Further difficulties arise when one tries to explain present-day dispersion in southern Italy in terms of Ahlmann's physical-geographical determinism.

Scattered settlement, although certainly not a general pattern in southern Italy, shows itself in two forms. In the first place, we find the so-called masserie (singular, masseria): big, often fortified, farms that appear in the same areas as the large agglomerations; such is the case in western Sicily where both forms go back as far as the fifteenth century. The masserie are more or less capitalistic enterprises of the large estate owners who, since they often live elsewhere, lease them to middlemen. In the second place we may find dispersed farmsteads of small-holders.

It is a mistake, however, to regard the dispersed farmstead situated in the midst of its own compact holding, and worked by the peasant cultivator, as absent in the South. There has been a steady increase in the areas of dispersed settlement over the last hundred years or more. Once the peasant is able to get secure tenure of a holding, adequate to support him, he ultimately moves out to it and builds his farm there. The crops he grows are naturally adjusted to the rhythm of winter rain and summer drought (Dickinson, 1955: 24).

This less frequent pattern is to be found in the regions of Apulia, Calabria, Lucania, Campania and Sicily (Maranelli, 1946; and Unger, 1953 :506-25).

From these considerations the following conclusions may be drawn. Socio-economic conditions are of as much importance as physical-geographical ones, if not clearly more significant. Hence it seems to be unwarranted to formulate generalizations regarding scarcity of water and agglomeration. We cannot, however, refute the thesis completely. Under certain (socio-economic) circumstances-in our case, extreme poverty of the peasant population-scarcity of water may force the peasants to crowd into agro-towns, since individual peasant families lack both the funds and the required mutual co-operation to dig wells in the country- side, and thus to leave the village and build up new homes in the country- side.

Being involved in the explanation of social phenomena we are more often than not concerned with complex causal linkages. Only rarely are we able to formulate simple, deterministic propositions of the type 'if A, then always By.It has been said that in the social sciences multivariate propo- sitions dominate over two-variate propositions (Zetterberg, 1965: 63 ff.).

1 It has been assumed for a long time that the nuraghi were tombs, but Le Lannou feels on reasonable grounds that they were dwelling-places ('habitations des vivants').

S O U T H I T A L I A N A G R O - T O W N S I25

This is especially true, as I hope to show, for the problem of settlement patterns. The relation 'scarcity of water-poverty-agglomeration' may be called contingent: if scarcity of water, then agglomeration, but only if poverty. However, as we will see below, things may be still more com- plicated. Besides multivariate propositions, there are also situations in which different factors, or different combinations of factors, may inde- pendently have the same effect (here agglomeration). In such cases we can formulate substitutable propositions: if A, then B; but if C, then also B. And when we are concerned with combinations of factors :if A and B, then C ; but if D and E, then also C. And so on.

As I have indicated above, we often find in the geographical literature the idea that agglomeration may be explained by a set of 'possible relevant' factors. Although multivariability is thus recognized, only rarely do we find it specified: which conditions may be considered as necessary, which others are sufficient? In a word, there are many statements and descriptions, but few analyses. Mere enumerations evoke the impression that all factors mentioned are necessary conditions which in combination have the effect of agglomeration. As we shall see, this does not need to be the case. Character, number, combination and significance of the various condi- tions may vary in different situations.

High summer temperatures, abundant concentrated rainfall (which on impermeable soils result in swamps) and deforestation make the coastal rim of southern Italy a favourable area for the malaria mosquito (Le Lannou, 1936: 113-36). This factor too is related to the distribution of water. In this context, however, it is not the deficiency but the surplus of water that some authors, at least indirectly, consider to be of importance in settlement patterns.' Kish's observations for Calabria may, in this respect, be extended to large parts of the South.

The wars and invasions, the devastations and anarchy, that became the fate of Calabria when Roman imperial power was destroyed by the barbarian invasions left their mark on the land. The rivers ran unchecked to the sea, creating swamps in the lowlands; the forests having been cut on the mountains, erosion stripped the slopes bare; malaria made the seaside and lowland settlements unhealthful (Kish, 1953: 496).

Le Lannou especially has stressed this point: the population does not abandon the plains because of malaria, but the population abandons the plains (because of socio-political reasons) and only then malaria develops (Le Lannou, 1936: 121-6). Thus Le Lannou sees the presence of malaria emphatically as a consequence and not as an antecedent of socio-economic

In this respect it has been noted that 'southern agriculture suffers more from an abundance than from a shortage of water' (VBchting, 1951: 19).

126 A N T O N B L O K

backwardness. He states that the prevalence of malaria co-varies with the character of agriculture: where we find intensive agriculture ('rapports permanents et intimes entre l'homme et la terre') there malaria will be most unlikely.

It is not surprising that malaria is so important in the Mediterranean, an area where the relations between the people and the land are so eminently unstable. Malaria is the most common consequence of somewhat sudden modifications in the physical environment in countries which, like the Mediterranean and the tropics, have a season warm enough to permit the rapid development of the sickness. In this way innovations and abandonments present grave dangers (Le Lannou, 1936 : 134-5).

The bad upkeep of the fields and the large estates on which extensive agriculture and sheep-breeding are carried out are, according to the French geographer, in the whole of southern Italy, as elsewhere, the main causes of the development of malaria. This fact, that malaria has to be seen rather as an effect than as a cause of the moving of people into agglo- merations, has also been noted by other writers (cf. Maranelli, 1946: 26-8; and Compagna, 1963: 78). These ideas leave little room for the hypo- thesis: if malaria, then agglomeration. In this connection we may ask: given malaria, would then dispersion not be extremely difficult? From a logical point of view this is the same as maintaining agglomeration. Again, I think, other variables must be taken into account before we can answer this question definitely. Along the coast of Calabria, Kish came across a number of settlements (marine) which have originated in the course of the last hundred years. The peasant population has left the traditional hill-top villages in order to occupy the coastal zone. Apparently the wish to live where labour could be found and the desire for fertile land has overcome their fear of fever. Presumably the construction of a railway along this coast has also promoted this migration.

To summarize :malaria may obstruct dispersion, but does not necessarily become the cause of agglomeration. The case described by Kish demon- strates that socio-economic factors may turn out to be decisive. Besides, malaria has been wiped out in the South since the last war and is efficiently under control in most areas. Nevertheless agglomerations persist. Such circumstances should also make us cautious and must prevent us from seeing the relation between two variables in deterministic terms.

v Doubtless security, defence, and other strategic considerations have been of great importance in the construction of the nucleated hill-top towns. Virtually no author who writes of the geography of the Mediterranean neglects to mention this aspect.

These plains are charged with history; the permanent insecurity during thousands of years has forced the agglomeration of people on fortified places (Sorre, 1952: 74).

In many Mediterranean areas such as Sicily a man living outside the urban walls as rural worker and country resident is almost unknown. This is a product of century-long insecurity (Weber, 1962: 82).

I t would not seem too difficult to collect a large number of similar state- ments. I t often seems tempting to explain a social phenomenon by referring to the past. I would not maintain here that such historical explanations have no right to exist. I only wish to draw attention to the fact that, in spite of their plausibility, they are often less satisfactory for the logician and sociologist, because no clear distinction is made between the origin and the persistence of the phenomenon. Casu quo :what keeps people in those large (paradoxical, as has been said) settlements? This question appears to be legitimate because insecurity has decreased significantly over the last hundred years, although in certain areas like Sicily and Sardinia it has not disappeared completely. For an illustration I may again refer to Kish.

After the fall of the Roman Empire, Calabria (like other parts of the South) was characterized by invasions, raids, and anarchy. The peasants took to the mountains. A map, made in 1759, shows only two main coastal towns : Reggio, protected by Messina, and Crotone, lying in the shade of a strong fortress. After the return of political quiet, a part of the mountain population moved to the plains in the course of the nineteenth century (Kish, 1953: 499). In this respect it must be noted that between 1861 and 1901 scattered peasant settlements in the South increased, not only in an absolute sense, but also in proportion to the increase of the urban popula- tion (Maranelli, 1946: 28). Nevertheless, we still find considerable agglo- merations in largely pacified areas. Before we deal with this problem, I wish first to consider the situation in which insecurity is still relevant, as it is for instance in western Sicily.

As I have observed elsewhere, scattered settlement as proposed and facilitated by the government (construction of small farmsteads with allocation of small adjacent holdings) has proved to be a failure in a number of west Sicilian villages (Blok, 1966). There were various reasons for the negligible results of this land reform, and one of them, I believe, was that public order was not efficiently guaranteed. Before and at the time of the reform (about 1950) cattle-rustling, extortion, theft of crops, and even homicide were fairly common.' So people had good reason to prefer the traditional nucleated village. Yet, scattered settlement was not entirely absent and in fact has always been present in this area. For centuries there have been the masserie, large outlying farms, the centres of the latifondi (large estates).' The owners or managers kept large herds

1 I have dealt with this problem in a paper: The Use of Violence: a Case-study from Western Sicily. MS., 1967.

a A similar settlement pattern has been noted in the Great Hungarian Plain (Alfold). See den Hollander (1960-61 : 74-88, 155-69). In this area we also find isolated farmsteads

of cattle and sheep. These enterprises were the most prosperous cases in this poverty-stricken area, and were only rarely pillaged. The paradox may be explained as follows. These estates had their own private police, feared armed men on horseback who more often than not extended their super- vision to brigandage and protection of bandits, whom they then organized to carry out large-scale rustling operations at the expense of less protected farmers and peasants. Against exactions from the law, these supervisors received protection from the estate owners.'

Insecurity does not invariably result in agglomeration. This holds true for other areas as well. Demangeon shows that in northern and central Italy isolated farms were fortified (1927: 16). So we cannot accept Konig's statement: 'the isolated farm may only (iiberhaupt erst) appear after pacification of the countryside and this is the reason that the farm is everywhere a relatively recent phenomenon' (1958 : 35). Nevertheless, although insecurity does not seem to be a sufficient condition for agglo- meration (if insecurity, then agglomeration, regardless of anything else), it is doubtless a very important onen2 On the other hand, it may be argued that the relation between insecurity and agglomeration is also reversible; the mere fact of agglomeration creates favourable conditions for in- security. This is at least obvious in the desolate inland region of Sicily.

To conclude this section let us dwell for a moment on the persistence of agglomerations in largely pacified areas. In such cases we must take other factors into account. It might be, as I have indicated in my review of the factors of scarcity of water and malaria, that the distribution of resources, that is, poverty of the peasant population, forms serious impediments to dispersion. On the other hand, and this seems to me a point of view entirely neglected in the literature on settlement patterns, agglomeration, for whatever reasons called into existence, may in the course of time generate conditions which make dispersion unattractive for the population. For example, an urban pattern of culture may generate not only a contempt for labour on the land but for the countryman as well. This is clearly

1 There exists an extensive literature on mafia and brigandage. For two neat historical surveys see S. F. Romano, Storia della Mafia (Milano: Sugar Editore, 1963); and Domenico Novacco, Inchiesta sulla Majia (Milano: Feltrinelli Editore, 1963).

a The relationship between insecurity and peasant agglomerations has also been observed in some African societies. See Watson, 1958: 75; and Bascom, 1955.

(tanydk)and large concentrated villages, 'big enough to be called urban yet with a rustic air: the streets are unpaved, the houses are of one storey only, and the occupations of the people are largely agrarian. These peasant towns are a feature almost peculiar to the Alfold; no- where else in the world are there so many large agglomerations of this type' (den Hollander, 1960-61: 74-5). The south Italian experience sets limits to the claim to uniqueness of the Alfold settlement type. Yet two minor differences between the forms must be mentioned. The Hungarian tanydk arose as a temporary home during the busy season, only to become a fixed peasant family settlement unit afterwards. The Sicilian masseria was never permanently inhabited by peasant families: it was a big farm, situated at the centre of a latifondo, where supervisors and agricultural personnel (who had their homes in the agro-town) stayed temporarily. Like the tanydk, it included stores and stables.

the case in large areas in southern Italy. We will turn to this subject in a later section.

Demangeon quotes several authors who contend that there exists an unequivocal causal relationship between the organization of large estates and the appearance of nucleated settlements. The large agro-towns of the South owe their existence to the fact that large estate owners wanted to concentrate their tenants in order to control them more efficiently (Deman- geon, 1927: 17). Moreover, the large estate owners were interested in avoiding permanent settlement of the peasants in the countryside for fear the peasant family would lay claim to the farmstead. Demangeon points to an exception: the situation in the Baltic areas in the early nineteenth century, described by Woeikof. This geographer came across large estates in combination with dispersed settlement of the peasants. He offers the following explanation: The Lettish, particularly in Courland, have long lived on isolated farms and not in villages; there are maps and plans from the 17th century proving this. After the abolition of serfdom (1816-1819), the noble estate owners judged that it was in their interests to encourage the peasants to live on farms, instead of in small villages which existed in Estonia and in the North and East of Lithuania. They believed with reason that the peasants living on farms would prosper better and would pay revenues due to their former lords more punctually. The Estonians did not like dispersion, but the nobility employed a strongly German tenacity to achieve this end, and succeeded entirely in it (Woeikof, 1909: 15-16).

It seeins that we are dealing here with two ad hoc explanations. On the one hand large estates would lead to agglomerations (southern Italy) and on the other hand to dispersion (Baltic countries). Doubtless, other conditions have to be taken into account. I suspect that relations between both variables may be contingent: if A (large estates), then B (agglomera- tion), but only if C (?).And the same would be true for dispersion. Now I shall try to establish what may be read for C.

One of the important differences between the areas is the density of population. As is apparent from Woeikof's description, we have 'open resources' (cf. Nieboer, 1910: 385) in large parts of the Baltic countries: apart from the large estates there was land on which the peasants could practice shifting cultivation. The existence of these open resources may have induced the landlords to force the peasants into fixed places, in this case scattered homesteads. Besides, according to Woeikof, shifting cultiva- tion is not a favourable condition of agglomeration (1916: 16). On the contrary, in southern Italy pressure of population prevails: all arable land is occupied ('closed resources') and there is a population surplus, which finds outlets in overseas emigration. In such circumstances, that is closed resources, 'a landlord can always find free tenants who are willing

to pay him a rent, and free labourers who are willing to work for him' (Nieboer, 1910: 349). This is indeed clearly the case in southern Italy where large estates prevail. The agglomeration or, more precisely, the piazza of the agro-town serves as a labour market, where the unemployed masses crowd to be considered for a contract on terms dictated by the managers of the large estates. In this respect it must be noted that the persistence of agglomeration in the South also depends on the nature of the tenancy systems: short-term leases, which involve a yearly or two- yearly circulation of tenaats, predominate.' Such a system of tenancy is functionally consistent with the prevailing mode of crop rotation: after the grain is removed, the stubble-fields serve to pasture the flocks of the owner or manager.

It seems now that we know what may be read for factor C: density of population, tenancy system, and mode of cultivation. However, we cannot yet formulate propositions that assume contingent relations between these variables. We must take into account the fact that population pressure in the South is a rather recent phenomenon. During the fifteenth, sixteenth and seventeenth centuries a considerable number of nucleated villages were built in western Sicily by large landowners because there was at the time a shortage of labour: they wished to colonize their unfarmed and underfarmed land (Garufi, 1946: 81-3 ; and La Mantia, 1904). In view of these facts, density of population does not appear to be a factor of crucial importance for settlement forms. To explain the difference in settlement in the Baltic areas and in the South we must focus on variables such as modes of agriculture, tenancy system, insecurity and perhaps other conditions not yet clear to me. Only a more complete analysis of Baltic society (which, as far as I know, is not available) may bring forward suggestions about relevant variables.

VII

In southern Italy as in other parts of the Mediterranean area, labour on the land is rather negatively valued. This has been observed in many studies. Banfield remarks: 'Manual labor is degrading in the southern Italian ethos, and labor on the land, together with personal service, is especially so' (1958: 69). Pitt-Rivers notes that Andalusian peasants lack 'a mystical attitude towards the land; they dwell in towns from which they go out to cultivate the earth, but do not love it. This characteristic is typical of the whole Mediterranean' (1961 : 46-7). Harris, in his book on a small town in Brazil, believes that aversions to agriculture and to manual labour in general are elements of a complex of values, an urban ethos which 'is

1 Cf. Maranelli, 1946: 27-8. McDonald observes that 'Italian literature is rich with accounts of cu1tiv;rtors of the "typical" South abasing themselves before the more wealthy in order to get share-cropping contracts' (1956: 446).

S O U T H I T A L I A N A G R O - T O W N S I31

probably a fundamental part of Mediterranean and Latin American cul- ture' (1956: 280-1). In a recent article by Pitkin we find similar statements concerning the attitude toward the land.

In Mediterranean Europe, among those who work the soil, there is not the strong tradition of independent peasantry in the sense of the individual cultivator closely attached to the soil and living according to the traditions of a folk culture that has been described for other parts of the world. In fact, a great many who live by the land have been day-laborers, not owner cultivators, and their attitude toward the land they work has often been more negative than positive. Even the small owner often attempts to relegate the labor of a part of his property to others so that he may enjoy the status of employer. And the large landowners generally absent themselves entirely from the scene expending their energies on urban living (Pitkin, 1963: 127-8).

This cultural emphasis on an urban way of life brings us back to a question briefly indicated in an earlier section of this essay: what conditions may account for the persistence of agglomeration, for whatever reasons it may have come into existence? It is generally accepted that certain circum- stances in the past, especially insecurity, which was widespread through large parts of the Mediterranean, caused agglomeration. Now the very fact that large numbers of people were living closely together over long periods may have generated an urban pattern of culture which was strongly enough developed by the time these areas became pacified to make disper- sion unattractive.

As has been noted in various places in this article, dispersion is not entirely absent in the South; in some areas there is even an increase of scattered settlement patterns. There is some evidence for a correlation between dispersion and farm status. In this respect, the information of Banfield, who made a study of what he calls a 'fairly typical southern town' (1958 : 1 1), is most illuminating.

Three-fourths of the farmers live on outlying farms; the others live in town. Those who live in the country have much larger farms on the average (17 acres as against 5); indeed, it is because their farms are so small that the town-dwelling farmers do not live on them (1958: 46). The country-dwelling peasant is generally much better off than the town-dwelling peasant, in fact he lives in the country because he has enough land and livestock to require his presence there (1958: 71).'

The village near the Pontine Marshes, studied by Pitkin, shows a similar pattern: poor peasants tend to live in the nucleated village and landowning peasants tend to live on farmsteads (Pitkin, 1959: 172). Thus we have to consider the distribution of resources as an indispensable variable in explaining agglomeration. The prestige of the town, the polis, carrying with it the peasant's distaste for agricultural life (cf. Redfield, 1960: 66)

The fact that so large a proportion of the peasants of this village lives outside its walls weakens somewhat Banfield's statement that the village is typical for southern Italy (cf. 1958: 11).

has, so to speak, a strong economic determinant. Yet these variables, both cultural and economic, must be kept separated since the mere fact of living on the land lowers the social status of the peasant, regardless of economic benefits (cf. Moss and Cappannari, 1962 : 289-90; Silverman, 1965 : 175; Silverman, 1966 : 908-9).

In the west Sicilian village where I studied the results of land reform, the new settlers showed little wish to live permanently on the small farm- steads that had been built for them. But one of the reasons that dispersion failed was that these farms were too small and unviable (to support life most colonists had to turn to landlords for contracts on outlying fields). Where the allocated new farms were substantially larger, as for instance in Campania Felix (Unger, 1953: 522), dispersion did succeed.

Therefore we can propose the following hypothesis: in pacified areas agglomeration will persist, if labour on the land is frowned upon and if resources are strongly limited.

VIII

Closely related to the previous argument is the thesis put forward by Ahlmann and Demangeon. Both geographers state that there exists a reversible relation between dispersion and intensive cultivation (Ahlmann, 1926; Demangeon, 1927: 23). It is obvious that in cases where the peasant lives on the land he will be able to devote more attention and energy to his land and crops. At least, he will not lose time and energy going from his town to his land as does the peasant from the agro-town. Intensive culti- vation, which involves a significant capital input, is in many coastal areas of southern Italy (where we find irrigated fields and crops such as lemons and sweet oranges) coupled rather with dispersion than with distinct agglomeration. On the other hand, extensive grain cultivation and nomadic cattle- and sheep-breeding, which predominate in the inland regions where agro-towns are prevalent, do not necessitate the permanent residence of peasants on the land. This mode of agriculture seems to be functionally consistent with nucleated settlement. Yet, again we must take into account the way in which resources are distributed. Dispersion then seems to make sense only if the peasant is either owner of a more or less united plot, or if as a tenant he enjoys a certain degree of independence regarding a similar piece of land. Traditionally, however, south Italian peasants are poor, and if they own land it is often scattered over great distances, a circumstance principally due to the bilateral law of inheritance. As has been noted before, short-term leases prevail, an additional factor that discourages fixed scattered settlement.

The advantages and disadvantages of dispersion and agglomeration respectively are set forth by a number of geographers. Besides, in land- reform programmes dispersion is often stressed for developmental ends.

SOUTH I T A L I A N AGRO-TOWNS 133

Less frequently has it been pointed out that agglomeration forms a func- tional requirement of, and is adjusted to, existing patterns of land tenure, lease systems, inheritance and extensive modes of agriculture. Unless resources are reasonably redistributed for the benefit of the peasants living in agro-towns, dispersion may not be attractive. Hence in such cases it would be unwarranted to speak of the 'clear disadvantages of agglomeration'. To call this settlement type paradoxical (Demangeon, 1927: 4) is to overlook its main functional attributes.

IX

It is sometimes argued that dispersion is impeded by the fact that women are not accustomed to work on the land in southern Italy (Sartorius von Waltershausen, 19 13 : 23-7 ; Carlyle, 1962 : 106). Especially in the inland region of Sicily, for women to till the fields is a rare phenomenon and restricted to only the very poor families. It has been pointed out to me by Sicilian peasants that, given difficult communications and hard labour, female labour in the country would be unfeasible and unwise. However true this may be, there are, I think, still other reasons why Sicilian peasants are unwilling to expose their women to labour on the land. The South Italian family has been characterized as 'father-dominated and mother- centred' (Moss and Thomson, 1959: 38). Masculine dominance and prestige find expression in and are enhanced by the male's economic role. As the head of the family he is the main breadwinner. In this task he is usually assisted by his unmarried sons, but the responsibility is entirely his. The economic activities of his wife and daughters are confined to the home. Strong restrictions exist for the sexual activities of his wife and daughters, which he is supposed to control rigidly. A man who allows his women to work outside the home would jeopardize his honour in two ways: directly, by showing that he is not himself capable of supporting his family, and thus failing in his culturally defined role of the superior male versus the inferior females of the family; and indirectly, because he will be less able to control their sexual activities. Hence there is cultural emphasis on seclusion of women: most peasant women are not only not allowed to go out into the fields, they are rarely seen in public. As has been said, very poor families cannot live up to this ideal (though most peasant families try to do so by making important sacrifices) and in order to make a living at all, must send their wives and daughters into the fields for light tasks such as gathering wood, or harvesting grapes and olives. In this way labour on the land by women becomes associated with low status. I think we are dealing here with a variant or an implication of the previously noted urban pattern of culture in which labour on the land is negatively valued; a pattern so strongly articulated in south Italian peasant villages.

I34 ANTON BLOK

X

The problem with which we have been concerned in this essay deals with the raison d'2tre of the south Italian concentrated settlements, the so-called agro-towns. Conditions for the appearance of this settlement type have been often noted and described, but seldom systematically analysed and tested. One of the few exceptions was an essay by the French geographer, Demangeon, published in 1927.

It proved to be useful in distinguishing between conditions that con- tributed to the origin on the one hand, and conditions that may account for the persistence of agglomerations on the other. As might have been expected, deterministic and sufficient relations, so rare in the social sciences, could not be formulated. In most cases a range of variables had to be considered. Several of these multivariate propositions turned out to be contingent relations and substitutable relations. Of the main variables it may suffice here to mention the unequal distribution of resources and the urban culture pattern. These conditions are not characteristic for southern Italy only, but may be found in large areas of the Mediterranean, a region typical for the large concentrated settlement.

R E F E R E N C E S

Ahlmann, Hans W. (1925), 'Etudes de Geographie Humaine sur 1'Italie Subtropicale. La Sicile', Geografiska Annaler, 7, 257-322.

Ahlmann, Hans W. (1926), 'Etudes de GCographie Humaine sur I'Italie Subtropicale. Calabre, Basilicate et Apulie', Geografiska Annaler, 8, 74-124.

Banfield, Edward C. (1958), The Moral Basis of a Backward Society. Glencoe: The Free Press.

Bascom, William (1955), 'Urbanization among the Yoruba', American Journal o, Sociology, 60, 446-54.

Blok, Anton (1966), 'Land Reform in a West Sicilian Latifondo Village: The Persistence of a Feudal Structure', Anthropological Quarterly, 39, 1-16.

Carlyle, Margaret (1962), The Awakening of Southern Italy. London: Oxford University Press.

Compagna, Francesco (1963), La Questione Meridionale. Milano: Garzanti. Demangeon, A. (1927), 'La Geographie de l'Habitat Rural,' Annales de Giographie,

36, 1-24 and 97-115. Dickinson, Robert E. (1955), The Population Problem of Southern Italy. An Essay in

Social Geography. Syracuse: Syracuse University Press. Garufi, C. A. (1946), 'Patti Agrari e Comuni Feudali di Nuova Fondazione in Sicilia'

Archivio Storico Siciliano, serie 3, vol. 2, 7-131. Harris, Marvin (1956), Town and Country in Brasil. New York: Columbia University

Contributions to Anthropology. Hollander, A. N. J. den (1960-61), 'The Great Hungarian Plain: A European Frontier

Area', Comparative Studies in Society and History, 111, 74-88 and 155-69. Kish, George (1953), 'The Marine of Calabria', The Geographical Review, 43, 495-505. Konig, RenC (1958), Grundformen der Gesellschaft. Die Gemeinde. Hamburg: Rowohlt. La Mantia, G. (1904), I Capitoli delle Colonie Greco-Albanesi di Sicilia dei Secoli XV e

XVI. Palermo: Tip. A. Giannitrapani.

SOUTH I T A L I A N A G R O - T O W N S 135

Le Lannou, Maurice (1936), 'Le Role Gkographique de la Malaria', Annales de Ge'o- graphie, 45, 113-36.

Le Lannou, Maurice (1938), 'Les Nuraghes de Sardaigne. Contribution de la Prdhistoire B 1'Etude du Peuplement Rural de l'Ile', Comptes Rendus du Congr2s International de Ge'ographie (Amsterdam). Leyden: Brill, 2, 101-6.

Maranelli, Carlo (1946), Considerazioni Geografiche sulla Questione Meridionale Bari: Laterza.

McDonald, J. S. (1956), 'Italy's Rural Social Structure and Emigration', Occidente, 12, 432-57.

Moss, Leonard W. and Stephen C. Cappannari (1962), 'Estate and Class in a South Italian Hill Village', American Anthropologist, 64, 287-300.

Moss, Leonard W. and Walter H. Thomson (1959), 'The South Italian Family: Litera- ture and Observation,' Human Organization, 18, 35-41.

Nieboer, H. J. (1910), Slavery as an Industrial System: Ethnological Researches. The Hague: Martinus Nijhoff.

Pitkin, Donald S. (1959), 'Land Tenure and Family Organization in an Italian Village', Human Organization, 18, 169-73.

Pitkin, Donald S. (1963), 'Mediterranean Europe', Anthropological Quarterly, 36, 120-30.

Pitt-Rivers, J. A. (1961), The People of the Sierra. Chicago: The University of Chicago Press, Phoenix Books.

Redfield, Robert (1960), Peasant Society and Culture. Chicago: The University of Chicago Press, Phoenix Books.

Rochefort, Renee (1961), Le Travail en Sicile. Etude de Gkographie Sociale. Paris: Presses Universitaires de France.

Sartorius von Waltershausen, A. (1913), Die Sizilianische Agrarverfassung und ihre Wandlungen 1780-1 91 2. Eine Sozialpolitische und Wirtschaftliche Untersuchung. Leipzig: A. Deichert'sche Verlagsbuchhandlung.

Semple, Ellen Churchill (1932), The Geography of the Mediterranean Region. Its Relation to Ancient History. London: Constable & Company.

Silverman, Sydel F. (1965), 'Patronage and Community-Nation Relationships in Central Italy', Ethnology, IV, 172-90.

Silverman, Sydel F. (1966), 'An Ethnographic Approach to Social Stratification: Prestige in a Central Italian Community', American Anthropologist, 68, 899-921.

Sorre, Max (1952), Les Fondements de la Ge'ographie Humaine. L'Habitat (vol. III). Paris: Librairie Armand Colin.

Unger, Leonard (1953), 'Rural Settlements in the Campania', The Geographical Review, 43, 506-25.

Vochting, Friedrich (1 951), Die Italienische Sudfrage. Entstehung und Problematik eines Wirtschaftlichen Notstandsgebietes. Berlin: Duncker & Humblot.

Watson, William (1958), Tribal Cohesion in a Money Economy. A Study of the Mambwe People of Northern Rhodesia. Manchester: Manchester University Press.

Weber, Max (1962), The City. New York: Collier Books Edition. Woeikof, A. (1909), 'Le Groupement de la Population Rurale en Russie', Annales de

Ge'ographie,18, 13-24. Zetterberg, Hans L. (1965), On Theory and Verification in Sociology. Totowa, N.J.

The Bedminster Press, third enlarged edition.