1In 1931 the American geographer Isaac Bowman referred to Latin America as a « continent of pioneer fringes »1. Years later to-day, his words are still true but now hitherto neglected regions have become a matter of interest and concern for academic specialists, economists, sociologists, development planners, multi-national corporations, governments and even inarticulate pioneers themselves. This burgeoning interest in new frontiers of settlement stems from an awareness that empty regions are potentially rich in minerals and tropical products or can be developed to produce temperate crops or to graze cattle. With the disappointed expectations from the « green revolution » the supposedly high productivity potential of virgin tropical rain forest is regarded as a solution to imminent world food crisis. National governments wish to develop these areas as a solution to a number of internal problems. Frontier expansion can increase food supplies to expanding urban conglomerations without substantially disturbing existing landowning patterns and drain off surplus population from over-burdened cities and depressed rural areas. At the same time it serves as a profitable field of investment for both national and multi-national corporations.

2As new areas are opened up, the ways in which they are being exploited, or should be exploited, are becoming matters of international concern for environmentalists exercised about the exploitation of indigenous peoples, the waste of resources and the ecological consequences of the application of inappropriate technologies. Attitudes towards the « last frontier » swing from unbridled optimism to apocaleptic pessimism.

3Partly as a consequence of this wider international concern as well as the more direct interest of multi-nationals seeking new sources of raw materials and scope for investment, national governments have started to implement programmes designed to open up these areas for settlement and exploitation as part of a policy of national integration. It is not entirely coincidental that official interest in colonization schemes has coincided with the role of military governments concerned with geo-political considerations and the security of exposed political frontiers. A striking feature of South America is that these frontiers often run through the least populated parts of national territory. Hence national governments have been anxious to establish a national presence de facto as well as de jure. Pioneers have strategic uses – for example in the founding of Punta Arenas in the 1840s which stemmed from an acute awareness of the strategical importance of the straits of Magellan once regular steam packet services round Cape Horn had started, in the same way as in 18th-century Brazil the Portuguese planned to pre-empt Spanish territorial ambitions by encouraging immigrants to settle in exposed frontier regions and by subsidizing family emigration to Santa Catarina. Similar strategic considerations prompted Spanish expansion into California and Texas in the 18th century.

4Academic interest in the «last frontier» has been stimulated by the growth of world-system analysis and variations on the theme of metropolis and frontier or core and periphery, either within a non-ideological frame of reference, similar to W.P. Webb’s « Great Frontier » hypothesis or in neo-marxist categories following Gunder Frank and Immanuel Wallerstein2. From this perspective the ebb or flow of frontier expansion is intimately related to fluctuations in the international economic system as well as to the warping effects of the North-South dichotomy.

5The extent to which the concern of governments, official agencies and of academics is shared by non-specialists and the general public is a moot point. In Latin America, for a variety of reasons to be discussed later, the word « frontier » does not have the same resonance as in the United States. It does not have those connotations which have become part of the popular mythology of most North Americans who identify with the feats of collective endeavour by which the « West » and the frontier was tamed.

6In this respect, Latin American experience to-day differs markedly from the early phases both of Spanish New World settlement and from the 19th-century settlement of the United States. In the period of mass migration to South America from the 1880s cities rather than the rural wilderness were the magnet attracting the majority of immigrants. Governments wished to encourage immigration « gobernar es poblar » – preferably from northern Europe, in order to develop exposed frontiers and to offset what were considered to be the congenital deficiencies of a mixed population. But immigration was only desired in terms which would not constitute a threat to established interests. On many occasions private and official recruiters in Europe misled immigrants as to the extent of help they would receive on arrival. There was very little infrastructural support, in comparison with the United States (and with the exception of São Paulo, parts of the pampas and Chile). It is only recently with the growth of official interest in colonization programmes that roads have been built into the wilderness to stimulate migrant flows.

7North American frontier expansion occurred during the high-tide of laissez-faire capitalism. If the frontier did not exactly expand in a « fit of absence of mind » like the British Empire, at least it did so with the minimum of centralized direction, whereas to-day authoritarian technocrats, often military officers, attempt to encourage settlement in particular regions. However planners to-day cannot assume that frontier expansion will acquire its own momentum comparable to 19th-century American experience when millions of European immigrants sustained the rate of expansion. Nor do many contemporary migrants possess the farming skills of European peasant migrants of the 19th-century. Not only does the lack of impetus from mass immigration make the United States experience largely irrelevant but penetration into tropical rain forests raises novel problems for which there have hitherto been few precedents.

8In placing the frontier in Latin America in historical perspective it is first necessary to free ourselves of the influence of Turner’s frontier hypothesis. Ever since the original publication of his essay « The Significance of the Frontier in United States history » in 1893, United States historians have been sharply divided over how to interpret it. It has not been difficult to reveal its inconsistencies, and the lack of definitional precision has lessened its usefulness as an analytical tool for studying frontier experiences elsewhere. Nevertheless it provides a useful starting point3.

9The appeal and effectiveness of the thesis lay in its metaphorical re-formulation of an agrarian pastoral tradition which stemmed from the postulated of John Locke and 18th-century physiocratic notions, which were given a specific American twist by Jefferson’s views on the necessary relationship between democracy and an independent yeomanry. The Salvationists impulse at the core of Turner’s glorification of frontier virtues and his assertion that the frontier is the breeding-ground of democracy makes the thesis a turning-point in American culture and gives an insight into the interaction of myth and history but it makes it less useful as a heuristic device for analysing frontier movements elsewhere. Whatever the validity of the thesis that frontier societies promote democratic mores in a North American context, it would be difficult to transpose to Latin America where frontiers have more often been the breeding-grounds of caudillismo.

10In Turner’s formulation the concept of the frontier was used to refer to a number of related phenomena :

1. It was contrasted to political boundaries – those European frontiers which he described as « fortified boundary lines running through dense populations ». The European usage of the term was restrictive – it defined the limits of political control. Turner’s frontier – which, in fact, was not his invention but an adaptation of earlier usage – was non-restrictive and implied a continual process of expansion into « empty » space. In Latin America « frontera » may be used in both senses but the existence of political frontiers, often poorly de-limited through inadequate surveys at the time of independence, gave to frontiers between Latin American states a sensitivity which the United States-Canadian border and the Mexican-United States border (except for recent years) have never had.
2. The frontier as a region was defined by the demographic criterion of density of population. In the. 1890s land with less than two per square mile (and hence where the norms of centralized government could not be applied, so leaving much liberty to the local inhabitants to mould their own institutions) was deemed to be frontier territory. This definition, however, did not take into account the « carrying capacity » of the occupied land dependent on soil structure and on the particular commodity being produced.
3. The frontier was defined in terms of the dominant commodity exploited in a newly opened area of settlement, thus the agricultural frontier, mining frontier, etc.
4. The frontier could be defined by its dominant institution – thus the mission frontier throughout Spanish American or the military frontier in Mexico.
5. The frontier may be defined in terms of its dominant social type – thus the « trappers », « farmers », « miners » frontier, etc.
6. The frontier may be defined in terms of an area of cultural clash or exchange – as in the Indian frontier.

11Many of these frontiers may co-exist and overlap in a geographical location where settlement meets the wilderness but common to them all is the struggle for control over land – either with the land and its inhabitants or with an empty and intractable wilderness. Examples of the latter are hard to find, as even when land appears to be empty it may be being used for grazing or hunting but for settlers, hunters and nomads are « non-people ». They are described as living « beyond the frontier » and hence implicitly living in a barbaric state. It is, therefore, rare for an expanding frontier not to involve conflict between peoples of contrasting cultures, between « intruders » and indigenous inhabitants. A frontier opens at the point when the outside force intrudes, and closes once the indigenes have been subdued. At one extreme is extermination, as in the case of Argentina’s pampas Indians, annihilated in the War of the Desert as part of a deliberate policy of government – promoted frontier expansion, or extermination might occur fortuitously as a consequence of the transmission of disease as in the 16th century or with Amazonian Indians to-day. Alternatively, the indigenous inhabitants may be absorbed into the dominant economy as cheap labour as in the case of the incorporation of sedentary Indians whose hierarchical, priest-dominated and tribute-paying culture was transmuted into a parallel type of political, religious and social organization (although the sanctions may have differed widely). In rare cases the intruders might succumb to the indigenes by « going native » (the case of Portuguese in the Zambesi Valley in the 16th and 17th centuries is a striking example, while the coureurs de bois in Canada and the bandeirantes in Brazil were the nearest American equivalents), but Latin America is distinguished by its « frontier of inclusion » in which Europeans married with or bred with native women and produced a mixed population whose non-acceptance by intruders and Indians alike forced them into vagrancy and to seek a living beyond the fringes of ordered society. The centrifugal nature of dispersion of a mixed population is a major factor behind the spread of settlement in the early years of the Spanish Conquest and of the penetration of the Brazilian sertão.

12However easy the conquest in the heartland of sedentary urbanized Indian cultures, the Spaniards – and even more the Portuguese – were faced with a continual running battle with nomadic Indians who, once they acquired the horse changed the military balance and established a stalemate as in New Mexico and Texas or on Southern Chile’s Araucanian frontier4 the pampas.

13In Latin America mestizos or ex-slaves became frontiersmen from harsh necessity, not necessarily choice. But nevertheless, the frontier still constitued a haven of freedom and so could provide the seeds from which a myth of the frontier could sprout as in the case of maroon communities, quilombos, palenques, cumbes, which evolved in complex symbiosis or in direct conflict with native Indians5.

14Whatever forms violence may have taken on the frontier and however « anarchical » or random frontier violence may have appeared to city-dwellers, it was often structured according to rules defined and understood by frontier society. The values generated in this society – equated, as in Sarmiento’s famous and influential dichtomy, with « barbarism » – sanctioned behaviour patterns which were not only rejected by « civilization » but which have been ignored and unstudied except by anthropologists and ethno-historians. The frontier as seen by Sarmiento was dark and menacing. It was not a paradisical Eden where man could commune with nature and find the secret of human happiness which had been denied him in the corrupting environment of « civilized » cities, in simple, honest toil6.

15Myths not only respond to deep and hidden hopes and fears : they can also be manufactured and manipulated by vested interests. Thus in the United States what we now call the Great Plains was once depicted as the « Great American Desert ». Being considered unfit for settlement by Europeans allowed President Jackson to move the five « civilized tribes » from the East to West of the Mississippi. Once, however, the potential of the grasslands was realized, and an appropriate technology developed, railroads and speculators worked to change the image from desert to garden forcing the Indians to move yet again7. Similarly, in the case of Amazonia to-day the image of a « Green Hell » is being replaced by that of a « Red Desert » by environmentalists or alternatively by the image of an endlessly fertile forest producing a cornucopia of riches promoted by government and corporations8.

16Without an agrarian myth it is difficult to persuade any but the foolhardy to venture into the wilderness. However potent the myth of El Dorado might have been in enticing Spaniards beyond the frontier, it did not encourage settlement. Latin American history has been characterized by the absence of those agrarian myths without which sustained and permanent settlement remains a forlorn hope. The potency of the Turnerian hypothesis in the United States derives from it being located within a long-standing pastoral tradition with roots in ideas about independent yeomen and the harmony of man and nature derived from classical allusion. Such ideas, derived from antiquity and revived during the Renaissance never enjoyed a popularity in Spain similar to the popularity they enjoyed in France or England, partly because of different and complex cultural responses to landscape and nature. In Central and South American the scale and strangeness of the environment was too daunting to transpose the pastoral idyll from a benign European environment, but in any case the pre-emption of available land by latifundistas seemed to reinforce the deeply rooted Roman-inspired view that civilization was essentially an urban creation thus inhibiting the development, in the New World, of a rural-based civilization.

17If it is not possible to speak in Latin America of the frontier in a similar sense in which North Americans do – as a myth which explains the distinctiveness of Americans, marking them off from Europeans, enshrining the values which moulded America’s destiny – individualism, democracy, hard-work, frugality, inventiveness, optimism, a belief in self-help and a religious simplicity – a relationship may nevertheless be discerned, in certain cases, between frontier values and the metropolis – especially in those instances where the cattle frontier was dominant – Argentina under Rosas, Venezuela under Paez, the Villa movement during the Mexican Revolution – and even Brazil under Vargas. Frontier caudillos, riding in triumph down Mexico City’s boulevards, riograndenses tethering their horses in the centre of Rio, gauchos importing the values of the pampas into the centre of Buenos Aires, are the symbolic images of frontiersmen conquering the metropolis. The residual legacies of pastoral despotism influenced governmental attitudes as well as provoking admiration for the vigour of caudillo rule. But such connections between the frontier and national myths are difficult to substantiate. In the process of national integration mythologies of rejuvenation, embodied in ideal-types fulfil an important function, but fifty years of post-Turnerian historiography in the United States have shown how difficult it is to prove a causal connection between an open frontier and the practice of democratic government. It may be that precise correlations are irrelevant and can never be proved and that what is important is that Americans believe that there is such a correlation and that the mythical force of that belief is more important than the substantive proof.

18If so, what correlations may be established in Latin America ? Have frontiers there played a similar mythic role or will they do so in the future as the need to integrate national territories assumes greater urgency ? And to what extent can it be said that the obverse of the North American case is true in Latin America where the absence of free land and Homestead Acts might seem to account for the absence of comparable frontier myths ? Is there, for example, a causal relationship between the perpetuation of traditional, hierarchically ordered societies and a closed frontier ? Denying access to land in the Brazilian Land Law of 1850 and bypassing the Avellaneda Law of 1876 in Argentina was one way of ensuring that frontier expansion would produce a « big man’s frontier ». By denying security of tenure and sale on realistic terms, poor immigrants were condemned to inequitable tenancy arrangements or to become squatters to be worn down in running fights with neighbouring powerful landowners over legal title.

19Far from stimulating democratic values and creating a democratic myth it could be argued that Latin American frontiers have perpetuated outworn forms of social and economic organization. It could be argued that it was the frontier regions which have engendered caudillismo ; that power and prestige derive from ownership of large landed property and the accompanying domination of a serf-like rural following, and that the absence of a rural middle class has been a direct consequence of policies deliberately conceived to keep the frontier closed, so ensuring that the social relations on the frontier were those of a two-class system. In the struggle between the countryside and the city it is the former which seems to have triumphed after independence and that instead of expanding under the continuous influx of immigrants frontiers have contracted, engulfing cities with a rising tide of peasant migrants escaping from rural immiseration. It might further be argued that in Latin America, cities have been the safety valve for rural discontent and not vice-versa as in the Turnerian hypothesis and that the failure of viable democratic frontiers to develop has been a contributory factor in the rise of authoritarian governments. Such might be the Latin American equivalent of the Turner hypothesis – a myth to console rather than to inspire.

20However, the multiplicity of frontiers in Latin America – where the variety is even greater than in the United States, makes such hypotheses inadvisable, and it might be more fruitful to give some idea of this variety before considering the historical background behind the current interest in the development of tropical rain forests9.

21The initial years of the Conquest were characterized by a thin, patchy, extended settlement pattern, a consequence of numerous probes by Spaniards seeking gold. The El Dorado myth was a powerful stimulus to this restless wandering in 16th-century Spanish America and in 17th-century Brazil when the bandeirantes set out on their Wanderjahre in search of emeralds, slaves and gold. Settlements resulted only when disappointed gold seekers finally gave up and began to put down roots (the foundation of Asunción and of towns in north west Argentina are cases in point). Within fifty years expeditions had criss-crossed the Americas from Kansas to Patagonia, leaving in their wake isolated towns. Whereas, therefore, in the United States there is often a clear-cut line between settlement and wilderness, in Latin America settlement has consisted of a slow filling-in process or as Richard Morse has written « settlement occured and still occurs in an archipelago pattern »10.

22The discovery of silver at Zacatecas and Potosí drew settlers into hitherto unsettled regions and in the case of northern Mexico put into motion a moving mining frontier which was not to come to rest until it had penetrated California, Arizona and New Mexico. Mining settlements stimulated satellite economies, specializing in products for the mines. They also provoked an increase in state involvement for largely fiscal reasons. The advance of the mining frontier in northern Mexico was accompanied by an advancing mission frontier first with Franciscans, then Jesuits and finally after their expulsion, with Franciscans taking the initiative again. Elsewhere, mission frontiers expanded in response to the religious zeal of the orders (especially the Jesuits) and to the geo-political imperatives of the Spanish state which was prepared to benefit from the spearheading work of the missions. Both the mining and mission frontiers advanced in the face of Indian opposition. The Indian frontier can refer either to the cultural confrontation between settlers and Indians, or to a state of armed conflict where Indians obtained horses and so were able to bring about a military stalemate until the repeating rifle once again gave Spaniards the upper hand and enabled them to break Indian power in northern Mexico, in Southern Chile and on the pampas11. Indian frontiers were invariably in a state of endemic warfare as the weakness of an extended military frontier enabled Indians to carry out plundering raids with impunity at the same time as labour-starved settlers provoked conflicts as a means of legitimating the enslavement of captives. The reluctance of nomadic Indians to submit to christianization and civilization (becoming sedentary agriculturists) acted as a legitimation for the « imperialism of righteousness ». Until very recently, frontier historiography was written from the standpoint of the intruders. The « vision des vaincus » was ignored as being irrelevant. Indians were either destined to disappear before an irrevocably expanding frontier or to be incorporated into the intruding society as helots12. Except for those concerned to defend Indians’ right this attitude has not changed – least of all among settlers whose interests seem most directly threatened by the Indian presence.

23The most significant aspect of the economies which developed to support mining communities was the growth of a cattle industry, a unique contribution of Iberians to New World culture. Cattle expanded in response to the need to provide food, power, transport and clothing and rose to dominance in Mexico on the ruins of Indian agriculture and in the wake of the decline in the Indian population. Until the end of the 18th century, cattle was essentially a satellite economy producing goods for internal consumption. It was not until the 1780s that cattle production responded to external demand and then mainly in the La Plata region. Cattle and mule drives contributed to integrating regional economies, linking the cattle frontier of the sertao with the coastal plantations, the north of Mexico with Mexico City, Salta with Potosi, and Rio Grande do Sul with São Paulo. It is not for nothing that the tropeiro has been described as a culture-hero of colonial Brazil, whose modern counterpart is the long-distance lorry driver.

24The visibility of cattle cultures with their Iberian antecedents had an important effect on political attitudes once centralized control weakened with the demise of Spanish power. The « ruralization » of Latin American society in the first fifty years of independence was accompanied by the emergence of the caudillo, representing the values of cattle cultures and of pastoral despotism13.

25The early growth of cattle economies, combined with population scarcity and decline, inhibited agricultural development beyond the satisfaction of local needs. The agricultural frontier did not expand in response to external demand until rising European population, political stability and Indian defeat on the pampas enabled a cereal frontier to expand.

26Coffee frontiers clearly reveal the wide variety in the social relations of production which is one of the most significant features of Latin American frontier development and the one which marks it off most sharply from North American experience14. In the United States a free labour market fed by an increasing flow of immigrants guaranteed that frontier expansion would be sustained by a self-generating momentum. There, frontier expansion can be related to the ebb and flow of a free labour supply.

27In Latin America widely disparate man-land ratios as well as variations in the carrying capacity of soils coupled with the different labour demands of particular commodity production accounts for the differences in frontier labour systems. Labour shortages on the frontier did not mean that labour’s bargaining power was increased. It meant, on the contrary, that landowners with political influence were able to implement repressive labour systems and by concentrating land to force available labour to work on their estates. Peonage, share-cropping arrangements, limited tenancies and restrictions on the sale of public lands were all devices which perpetuated the domination of the large estate. Even when faced with the need to attract voluntary European immigration, land policy remained illiberal and many immigrants found their aspirations frustrated. This was one reason, among many, why so few immigrants took out citizenship and why, too, a higher percentage of the much smaller number who emigrated to Latin America returned to their country of origin that did their counterparts in the United States.

28The only possible access to land for those without capital or influence was by squatting. Squatting seems to have been a perennial feature of the Latin American frontier although it is a topic on which little historical work has yet been done, but it is clearly a major problem to-day. Even with restrictive legislation sealing off the frontier in Brazil it seems that posseiros were numerous in the 19th century but as the frontier expanded so the posseiro was either forced to move again or to become a hired hand15. Lacking a legal title the squatter cannot benefit from any improvements he may make. For the legal owner (however the legal title was acquired) the task of clearing the jungle is done at no cost to himself. Co-ordinated resistance on the part of squatters is difficult in the face of capangas or because of distance although as the Sierra Maestra experience showed squatters can become guerrillas under outside influence.

29Sugar had been the first substantial agricultural export crop, but it was confined mainly to North East Brazil and from the late 18th century to Cuba where abundant virgin forests made it possible to fill the vacuum left by the collapse of Haitian sugar production. However, coffee provides the most spectacular and extensive agricultural frontier in 19th-century Latin America for which there was no contemporary parallel. In Cuba, Central America, Colombia, Venezuela and Brazil, coffee production expanded to meet an unlimited demand in North American and Europe. Unlike sugar, coffee frontiers spanned the whole gamut of labour systems. In Guatemala, the coffee plantation, largely under the control of German merchants was worked by Indian peon labour and shared many of the characteristics of the traditional hacienda. In Costa Rica, coffee produced a medium sized peasantry. In Cuba the cafetales were worked by slave labour on small profit margins. In Venezuela, much of the labour was casual. In Colombia coffee expanded in response to the initiatives of Medellin merchants and developed into an economy of prosperous smallholders. In Brazil, the Land Law of 1850 ensured a virtual monopoly of access to the limitless virgin land west of São Paulo to the established plantocracy. At first coffee production depended on slaves recruited via the interprovincial slave trade from the declining north-east, but it finally came to rely on contracted European labour.

30However varied the impetus behind frontier expansion, settlement tended to avoid tropical rain forests16. Wherever these were encountered by Europeans they were skirted or destroyed – as da Cunha tells us happened in North East Brazil.

31Tropical rain forests rarely attract the voluntary migrant – only those driven to desperation by push factors such as the Ceará drought of the 1870s or maroons fleeing slave plantations. Imaginative literature with the jungle as its theme reinforces the popular image of it as a place of malevolence and mystery where man is overwhelmed by a nature too powerful to be harnessed and where the normal restraints of civilized living break down. Rivera’s La Voragine, Rangel’s Inferno Verde and Ferreira de Castro’s A Selva show how in Latin America literature the jungle has become a symbol of evil and corruption17. This image takes root at a time when Europeans were penetrating the jungle in considerable numbers in search of quick fortunes, and replaces the idyllic image of earlier scientific travellers such as Darwin, Bates and Wallace, writing in the mid-19th century, who were captivated by the Amazon’s unrivalled fauna and flora18. In spite of offering a refuge from the civilized world jungles have rarely become locations for Utopian communities. The macabre case of Jonestown in Guyana was very much an exception.

32Expeditions might enter the forest – searching for drogas de sertão, malocas seeking slaves or bandeiras searching for emeralds and gold, but it was not until the rubber boom years between 1970 and 1914 that large scale penetration occured but this was a caricature of a boom or bust frontier, leaving no integrated economy in its wake and few centres of population apart from the jungle emporia of Manaus and Iquitos. It bequeathed as its main legacy and unfulfilled dream of limitless potentialities, as well as an impetus towards precise definition of disputed political frontiers in anticipation of the day when the potentialities of development could be realized. The rubber boom attracted adventurers from all the Amazonian countries, and roused the interest of their respective governments in regions which they had previously ignored.

33Although Venezuela, Colombia, Ecuador, Peru, Bolivia and the Guayanas share common ecological problems with Brazil in the exploitation of their jungle territories, their respective attitude towards Amazonia have been conditioned by differing historical and geographical factors. Geography provides the immediate contrast. Penetration into the heart of the jungle is comparatively simple from the east but extremely difficult from the west as Bolivia found to her cost during the Acre dispute in 1900. The « selva » has never captured the imagination of Spanish Americans in a comparable way to the Brazilians, with their domination of the greater proportion of the Amazon river system. Nothing illustrates this more graphically than the failure of the Spaniards to follow up the early Amazonian expeditions by Orellana, Gonzalo Pizarro and other, or the penetration by Spanish Jesuit missions in the 17th century which collapsed without state support in the 18th. The Treaties of Madrid (1750) and San Ildefonso (1771) by which Spain surrendered her Amazonian rights for hegemony in the La Plata underlined very clearly the priorities of the day and the recognition of the easy exploitation of the potential wealth of the pampas in comparison with the putative riches of Amazonia19.

34For the first hundred years of independence Amazonia made little impact on the successor states. In Venezuela, early missionary penetration up the Orinoco in the 18th century came to an abrupt end when the missions collapsed during the wars of independence. Throughout the 19th century Venezuelans were distracted by inter-faction conflicts for control of the rich cattle lands of the llanos. Nor was there any demographic push to open up new territory. In any case, the border with British Guiana was a more pressing concern. Thus the Amazon frontier remained a petty-traders’ trans-border smuggling frontier. Its potential has only recently impinged on the national consciousness, prompted by a mingled fear of Brazilian expansionism and, most recently, by the possibilities of reciprocal trade between the two countries once the road systems are completed.

35In Colombia, the situation has been similar except that the development of the llanos and the selva has always been inhibited by the greater obstacles of access from the Andean uplands. The Colombian Oriente has often been described by politicians as the country’s El Dorado and the land of the future, but in the words of one of its historians it has remained a « static frontier »20.

36Historically Ecuador has shown more consciousness of its Amazonian territory. Most of the expeditions to explore the Amazon started from Quito and the Jesuits’ missionary activity was organized from an Ecuadorian base. But it was the loss of its Amazonian provinces in the war with Peru in 1942 which has stimulated an acute consciousness of the country’s Amazonian heritage21. Ecuador’s irridentist claims against Peru have not been lessened by the development of a trans-Andine oil frontier, and have been an important factor influencing the views of the Peruvian military.

37Apart from coca and chinchona exploitation, and involvement in the rubber boom, Peru’s trans-Andine frontier remained largely static until the 1950s. Peru illustrates even more clearly than Ecuador the difficulty of making a choice between differing development options, as revealed in the contrasted policies of Apra and Acción Popular. Both Haya de la Torre and Fernando Belaunde used historical precedents to legitimate their respective strategies22. The former argued that development should be inspired by pre-Incan coastal civilizations. Underpinning Haya’s view was his belief in the need to create a favourable man-land ratio (influenced by Japanese intensive agriculture) by concentrating resources on developing the costa by irrigation and migration from the Sierra. Only when wealth had been generated from the intensive cultivation of reclaimed desert should resources be diverted to the Sierra and beyond. One should not be misled by Apra’s indianist pretensions. Its appeal has always been to socially mobile mestizos and in one sense Haya’s development strategy of a contracting frontier was a rationalization of the need to marry his economic views with his power base.

38In contrast, Belaunde’s strategy was an expanding frontier conception, based on reversing the colonial and post-colonial drift to the costa and the preponderance of Lima (reflecting an arequipeños hostility to the capital) by resurrecting Indian highland communities. Inspired by Inca example, Belaunde’s Cooperación Popular was a self-help organization, subsidized by state funds and technical aid, aimed at utilizing communal traditions. Far from encouraging migration to the coast, Belaunde planned to stimulate movement to the selva by means of an ambitious programme of building feeder roads from the spinal carretera de la selva running from north to south on the slopes of the Eastern Andes. The difficulties and expense of this programme contributed to Belaunde’s downfall and to his replacement by the military with whom he had originally been in alliance.

39The military too were obsessed by historical example but of much more recent duration23. Peru’s defeat in the War of the Pacific had underlined the need for national social integration and had bequeathed a sensitivity towards exposed political frontiers. Victory against Ecuador in 1942 only served to increase this sensitivity, given Ecuador’s irridentist ambitions to recover the lost provinces. The Ecuadorian border dispute was only one factor conditioning the Peruvian military’s Enkreisung’ mentality. On the Chilean border the Tacna-Arica dispute had also left feelings of insecurity. On the Bolivian border where a political frontier runs through Indian communities sharing similar languages and culture there is a potential « submerged nationality » problem where a roused Indian consciousness could weaken loyalties to the national states of Bolivia and Peru. Finally, there is the threat of Brazilian economic penetration, facilitated by the completion of the trans-Amazonica highway. Faced with the historical example of Brazilian expansion at the time of the Acre dispute with Bolivia and with the contemporary example of the economic penetration of Paraguay and the Bolivian Oriente, the Peruvian military gave a high priority to colonization projects in the selva. But the poor performance of the Peruvian economy has made it difficult to implement high-cost colonization.

40Bolivia has also had to choose between policy options24. The loss of its outlet to the Pacific has removed a coastal option (although irridentist claims against Chile can still rouse nationalist feeling) but the memory of the loss of Acre has made governments sensitive to a repetition of a « spill-over frontier » phenomenon. Population pressure combined with increased mobility after the agrarian reform has stimulated spontaneous and sponsored colonization of the Yungas valleys and beyond under the stimulus of supplying the La Paz market. But the mineral and agricultural potential of the Oriente is increasingly shifting the centre of economic gravity eastward. How far the links between the altiplano and the Oriente can survive the economic pull of Brazil will be a key question of the coming years.

41As internal frontiers of settlement move closer to political frontiers with the filling in of empty spaces so disputes over territorial sovreignty can be expected to sharpen, underlining the need for international co-operation, as in the Amazon Pact.

42To what extent is it possible to distinguish between continuities and discontinuities in the history of frontier expansion in Latin America ? Where frontiers are expanding in plateau zones such as the Brazilian cerrado continuities may be discerned with similar developments in the past and regions elsewhere. But where the frontier is moving into tropical rain forests there is discontinuity. Examples of tropical colonization, except by Indians, (in Yucatan for example) have been extremely rare and where it did occur tended to be short-lived.

43Frontiersmen are destroyers – both of native peoples and the environment25. In the confrontation between « civilization » and « barbarism », between sedentary agriculturalists and nomads, between Christians and non-Chirstians, native peoples are destroyed. They have been victims of the « inevitable » march of Progress and if they do not die in the medical struggle for survival by succumbing to disease they are decultured and marginalized by frontier processes they do not understand and which do not seem very different from those in the past.

44Frontiersmen are tempted by abundance of land to be prodigal in their use of it as in the São Paulo coffee frontier in the late 19th century, and to pay little attention to, (or be able to afford), conservation measures. This is particularly true of boom-or-bust frontiers. It is difficult to avoid the conclusion that the contemporary process of destroying valuable, renewable timber resources in order to achieve quick profits from cattle ranching even at the price of soil exhaustion, differs markedly from earlier predatory exploitation in the 19th century. At least then the « lazy industry », as Spaniards called cattle ranching, could be rationalized in terms of labour shortages. No such justification or rationalization is possible to-day at a time of spiralling population growth. Cattle frontiers share the brutalizing tendencies of all « big mens » frontiers’ where profit is the major motive.

45The slowness of the Brazilian response in the past to the Amazonian challenge is explicable in terms of labour scarcity, geographical isolation, the easy exploitation of accessible staples elsewhere, and communications difficulties26. The far-sightedness of the Portuguese crown in its geopolitical struggle with Spain bequeathed a political rather than an economic problem to independent Brazil and it was mainly a revival of geopolitical considerations which underlined the need for more effective political control over empty territory in order to avoid the area becoming « internationalized ». The apparent urgency of this need led to a number of false starts in planning and investment policies in which historical patterns seemed to be repeated. Ecological imbalances were paralleled by the emergence of a two-class structure in newly-opened regions. The speed of development initially obscured the need for appropriate technolocies for solving new and complex problems27.

46The real challenge of the « last frontier » is whether frontiersmen and planners have the humility, cultural flexibility and imagination to learn from those who have lived and survived in a jungle habitat for centuries of time28. There are always two sides to a frontier and it should be the task of historians to help in changing the perspective from which frontiers have habitually been viewed. In doing so they might be able to add some small wisdom to offset the arrogant assumptions which have hitherto underpinned past approaches to frontier historiography29.