Pierre Monbeig and the Formation of Geography in Brazil (1925-1957): Geohistory of Knowledge, Intellectual History and the Insertion of Brazilian Geography into the World-Science

This paper aims to present the exemplarity of an intellectual meeting between a French intellectual, trained in history and geography at the Sorbonne, France (before spending time in Spain during the beginning of his doctorate), and the “Brazilian terrain”. From his training to his work as a university professor in Brazil, what I want to characterize is a transnational intellectual context in the domain of the history of science, using geographical reasoning as a reference. However, before becoming aware of these intellectual processes, it should be said that at the base of this context lies the Brazilian space. This kind of reasoning as a proposed methodology is named here the geohistory of knowledge. In this paper, I seek to present this methodology and its theoretical and empirical results, focusing on how the construction of contextualization can be related to space.

Monbeig did not come into direct contact with Vidal de la Blache, but did with the first disciples trained by la Blache, such as Albert Demangeon. In 1930 and1931, Monbeig attended Casa Velázquez, in Spain, where he began his doctoral research. In 1935, however, he was invited by Henri Hauser and George Dumas to teach geography and history in Brazil, at the newly founded University of São Paulo, as part of a French university mission. Monbeig travelled to Brazil in 1935 and soon after he abandoned his thesis on the Balearic Islands, beginning a doctoral thesis on the "pioneering fringes" of São Paulo. 4 He would live in Brazil for 11 years.
In 1946, he returned to France and defended his main thesis, Marche du peuplement et zones pionnières de São Paulo (Brésil) in 1951, obtaining a doctoral degree from the Sorbonne. In the same year, he was appointed professor of Colonial Geography at the University of Strasbourg and, returning to Paris in 1952, he took the chair of Economic Geography at the Conservatoir National de Arts et Métier. In 1961, he was appointed professor of Human Geography at the Sorbonne, when he took up the post of Director of the Institut des Hautes Études de l'Amérique Latine (IHEAL) (Andrade 1994, 75-76). Shortly after, he became director of research at the Centre National de la Recherche Scientifique (CNRS), becoming one of the leading experts of his generation. Monbeig had always been close to the historians of the socalled School of the Annales, after having maintained long-standing friendships with Lévi-Strauss and Fernand Braudel (Salgueiro 2006a).
The purpose of this essay is to present Pierre Monbeig's trajectory,5 in relation to the challenges of the discipline as a whole, in the long term (longue durée), which appear in this context and that of Brazil, how it relates to geographical spaces and spaces of debates, and considering the context of a world-science (Polanco 1990). 6 I call this perspective the geohistory of knowledge. My objective is to clarify how this theoretical proposal relates in particular to the intellectual procedure of the construction of contextualization.

Geohistory of Knowledge
Particularly after the publication of David Livingstone's book, Putting Science in its place (2003), the proposal to produce a geographical history of geography and a geographical history of sciences was developed.
Derek Gregory exemplifies what this geographical history (Sanguin 1993, 108-109). La Blache had already shown an interest in Geography through his travels in the Mediterranean and the reading of the work of Alexander von Humboldt and the German geographer and historian Carl Ritter about Asia and the Inland Sea (Lira 2013). At the height of French colonialism, Vidal opts to lay the foundations of geography close to ethnography, elaborating the concept of genre de vie and establishing bridges with Haeckel's German ecology and geohistory. During the course of his career, he moves towards human geography (albeit still very ecological in its nature) and finally towards geopolitics and economic geography. He trained almost all the geographers who took up the geography chairs at French universities as they were inaugurated at the beginning of the 20th century (Berdoulay 1984). For updated information on Vidal de la Blache's career and the institutionalisation of Geography in France, see Robic (2000Robic ( , 2003Robic ( , 2004Robic ( , 2013. 4 The idea of a "pioneer zone" is related to the process of modernization of the areas of central-south Brazil as a result of international market influence, soon after the abolition of slavery. 5 Pierre Monbeig boasts a vast work on Brazil and Latin America. His books and articles were mostly published in France andBrazil between 1932 and1987. Much of his work is accessible in specialized journals. Another part of his work, such as letters, cuttings, notes, rare books, is placed in the Pierre Monbeig Fund, which is located at the IEB (Institute of Brazilian Studies) of the University of São Paulo. For a complete bibliography of Monbeig, see the work of Heliana Angotti Salgueiro on Pierre Monbeig (Salgueiro 2006b). 6 The way Xavier Polanco uses this concept of world-science is more associated with the geopolitics of knowledge, the way relations between the centre, semi-periphery and periphery structure scientific relations between countries (Polanco 1990). of geography consists of: "This essay considers a third thematic that also shaped productions of knowledge through productions of space: the modalities through which the routes described by travellers (their paths and circuits) shaped their descriptions of those routes" (Gregory 2000, 297).
From the point of view of the geography of science, Charles Withers employs the concepts of Geography (scale, place, traffic, etc.) to understand the advent of European Enlightenment (Withers 2007). I consider these works as more general parameters from which I aggregate the notion of plurality of historical time (Braudel 1992). Monbeig spent time in France, Spain and Brazil, and this geographical and global circulation was of particular interest to me.
To build a geohistory of knowledge, I base my argument on the idea that spaces can frame specific evolutions at specific paces, in the circulation of ideas and such spaces would act as structuring filters in the process of the displacement of ideas and practices. It is also true that scales can vary, and that ideas evolve. These spaces conditioned evolutions, events and surface movements. Long and deep movements of ideas, that filled the constitution of science with events, conjectures, permanencies, ruptures and the circulation of ideas. It is therefore important to point out that my methodological handling of geohistory also includes reasoning about processes as if they also occurred in a deeper sphere. When I refer to processes in the history of science which are deeper than the others, I am referring to structures, layers, to a kind of "infra-mentality" that exists in scientific spaces and that can also circulate. 7 In the journey from France to Brazil, therefore, Monbeig's historical trajectory was touched in depth by long-lasting movements of scientific ideas that concerned the entire discipline, and all these movements were repositioned according to this and other displacements, and the spaces in which Monbeig is located: France, Spain, and Brazil (I do not analyse Monbeig's return to France).

Geohistory and Intellectual History
On the one hand, geohistory has led me to the spaces and plurality of times, in addition to the movements regarding the depth of ideas, which I call "infra-mentalities". On the other hand, intellectual history has led me to a general approach to the scholar's trajectory. Geohistory is associated here with a particular conception of intellectual history, which was previously established by François Dosse (2013). This means an intellectual history in a broad sense, involving actors, ideas, objects and practices. In this sense, Brazilian space was seen as one of the main structures of this circulation of ideas in Latin and global worlds, touched by deeply-rooted trends, and whose specific characteristics of space, such as its continentality, influenced the movement of displacement over the spaces.
An intellectual experience situated in the space, incarnated in characters and their concrete actions, beyond the very use and mobilization of objects, could not be ignored in the process of understanding the production of scientifical ideas. Besides, observing the spatial dimensions of the spaces of circulation, I seek to identify their unfolding in the sociological and political contexts (of the production of ideas), the affinities of the characters, the controversies, the associations, the strategies, and how these elements can help to understand the theories proposed by Pierre Monbeig and his colleagues.
However, I do not wish to encapsulate my reasoning in a geographical determinism, since it was not only the spatial dimension that conditioned the movement of ideas. I developed dimensions of analysis that started with geographic spaces and cultural traditions, the latter with different roots and geographic and regional frameworks (that the French geographical tradition developed in a specific way in relation to the German one, for example) that would be constructed by movements of depth, would pass through the mediations of circulations among other spaces and by the encounter of other movements of depth, only then to end up in individual, political, and then epistemological choices, even if within a framework of possibilities limited by the other spheres.

Geohistory and Epistemology
Finally, before proceeding with a geohistorical application to the contextual construction and the empirical results of framing primary and secondary sources, I will try to add to this approach the relationship between spaces, intellectual history and epistemology. That is, how space could focus on understanding an epistemological discourse and whether it could be geographically framed. I restrict here the understanding of the final form that the methodological path takes according to the analysis of its systems of causalities. Throughout my work (LIRA in press), I have assembled schemes, sometimes linear, sometimes circular, sometimes mixed, that give form to Monbeig's thought, inspired by Daniel Loi's methodology (1982, 1985. These forms could also be related to the spaces in which they were produced. In other words, a particular causal arrangement has the possibility of revealing individual methodological options, but they can also interact with the spatial situation and framework of which it is the fruit, in the case of geography.
For example, in geography, linear causal logics are usually rooted in deterministic reasonings, that is, the causal origin is usually a natural fact determining a human fact. On the contrary, circular causal logics are normally systemic, the result of the analysis of more humanized, economic or urban environments, and nature is included in the system (normally as a resource) but does not determine it. Thus, Daniel Lois had already elucidated that the use of different causal systems can come from methodological postures which are equally determined by the character of the object of the geographer, the spaces (Loi 1982(Loi , 1985. According to this author, causal structures reflect deterministic or possibilistic reasoning when considering the starting point of structure in natural or human phenomena (reverberating the duality of geographical science in the consideration of natural determinism); causal structures may also reflect the homogeneity of the reference space by taking into account internal or external causes of the space in question (revealing the duality of methodological considerations around isolation and circulation, as well as differences in vertical and horizontal, internal and external relations of the environment in question); synchronic and diachronic reasonings also reflect the characteristics of spaces, since they may be organized in linearity, normally revealing deterministic reasoning, or in cycles revealing independence in their relations with the environment (Loi 1982(Loi , 1985. Deterministic or semi-deterministic (mixed) reasoning is more common in agrarian monographs, while systemic reasoning is more common in urban monographs. Thus, causal structures result in aspects that are key to the understanding of methodological postures and are also structured according to the degree of humanization of the environment they reflect.
The intriguing thing about Daniel Loi's analysis is that the most recurrent causal structures in the classical works of geographers produced between 1900 and 1910 were those organized around linear linkages, which evoke a theoretical construction that is predominantly historical and ecological, based on linguistic and temporal linearity, revealing a "diachronic tendency of French geography" (own translation) (Loi 1982, 19) and which, as a rule, began with the consideration of natural factors. However, there is also the systemic, or synthetic structure: "It is the kind of [systemic] structure of which much is spoken, to which one aspires, but which is not practised, which cannot be practised" (own translation) (Loi 1982, 20). Pierre Monbeig uses five types of causal structure throughout his career: the simple type (linear and deterministic, since in its causal origin there is an aspect of the natural environment); a mixed linear type, involving a causal among both within nature and a constructed human space; the systemic type (circular and autonomous in relation to the characteristics of the environment, and which I try to model in the case of the advance of the pioneering fringes, its object of study); and the stratigraphic type (that is, the determinations would occur in a vertical ambit, in a relationship of structure with superstructure). Finally, it creates a mixed type between the stratigraphic and the systemic (the system at the top of a vertical evolution influenced by movement from below to above and vice versa), and which, moreover, proved to be specific to Brazil, as it was open and expanding, since it had a structural imbalance that led to the displacement and expansion of the cycle, since it never closed as a system. This is the causal model mobilized to explain the pioneer front, which is the purpose of Monbeig's study, that is, the movement of capitalist advancement toward the Brazilian hinterland ("sertão" in Portuguese) and the establishment of an agri-exporting economy in the transition from slavery to an industrial economy.

The Evolution of Long Durations in France
To consider the challenges of an entire discipline, which accompanied Monbeig's trajectory, I have to start in France, where he studied in his youth. Some of the long durations of French science would be reconfigured in the São Paulo region. On one hand, there is a conflict intrinsic to French scientific culture itself which, dating from the end of the 19th century, is beginning to emerge in the form of an antagonism between the systematisation of theoretical formulations, in the training of the social scientist, arising from the gaze of the renewed German natural sciences on reality and, on the other hand, the irreducibility with which the French literary tradition tended to adorn its narratives and which suggested assumptions about descriptions of totally unique situations (Lepenies 1996). These long-lasting French trends were first identified by Wolf Lepenies: "In the interruption of Buffon's career and the unstable reception of his Histoire naturelle, we can follow the process by which the sciences gradually distance themselves from literature, and how traditional values, which could be called literary, are excluded from the canon of accepted knowledge" (own translation) (Lepenies 1996, 13).
This conflict became increasingly explicit during the course of the institutionalisation of the social sciences in France, at the end of the 19th century, and was expressed by a supposed antagonism between the "literates", for whom it was necessary to describe, narrate and highlight particularities, and for the more "scientific" scholars, who argued that science should explain, systematise and highlight the universality of processes (Lepenies 1996).
These two trends were potentially conflicting and were rooted in the structures of French scientific infra-mentalities in the 19th century. Nonetheless, disciplines, such as geography, offered original solutions to reconcile these values. In agreement with Lepenies, I consider scientific values and literary values different academic postures in the face of the structuring and legitimization of knowledge, which cannot be reduced to epistemological logic, nor can they be explained by a relationship of a direct influence of the natural sciences or literature in the human sciences, since they imply taking positions in various spheres of society, in science, but also in culture and politics.
By characterizing this conflict even more, I consider it as a set of attitudes which are assumed around two poles and which can be expressed in various positions within society: the so-called "scientific" values, on the one hand, and "literary" values, on the other. Regarding scientists influenced by the renewal of the natural sciences, there is an "obsession with method", the primacy of "questions of classification" and "precise terminology", "specialized teaching", "gain in precision and objectivity", the primacy of "doctrine" over observation, a greater "methodological awareness" (Lepenies 1996), a higher degree of social commitment and application of the sciences to reality.
On the other hand, in relation to those who wagered on the power of sensibilities as a way of understanding the world, the "literates", they insist on the irreducibility of reality to systems, I see a critique of doctrines, highlighting the "illusion of dogmatism", the primacy of the importance of the "French spirit", of "brilliant rhetoric", of "linguistic sensitivity", of "style", of unselfish visions, of the detachment from the demands of politics. This play of forces has been part of the long tradition of French science (Lepenies 1996). In short, tension exists between forms of logical writing and expressive writing in France.
How is this tension expressed in Geography? In fact, such tensions have also greatly affected French geography, although in the coexistence of these values this discipline has found a positive and original "duality" in its epistemological constitution. This division was explained by the position occupied by the discipline at the time of its institutionalisation, with the result that this duality also constitutes a long duration of French geographical knowledge. Geography was divided, on the one hand, between the public success that previous geographical practices had generated (for example, the geography of Elisée Reclus), the insertion of a university position linked more to a literary discipline (History), and, on the other hand, between the practices that aroused an intention for exploitation, colonial interest and the planning for colonial areas (Robic 1991, 54). There is a "science" which, also immersed in the assumptions of "art", was structured around descriptions and narratives, on the one hand; on the other, a generalising appetite that allowed for action and "planning" of colonial spaces.
This formed an "epistemology of the mixed, or rather, of the between two, of the passage", according to Marie-Claire Robic (Robic 1991, 54). The expression "rational description" (an opposition in terms) condensed this "research strategy" (Robic 1991, 55) which incorporated the superiority of the explanation together with the description, whose very term "geo-graphy" etymologically highlighted the appreciation of it (Robic 1991, 55). Geography was situated between "empiricism" and "constructivism", between "pure description" and "explanation" (Robic 1991, 57), between the understanding of the literates and the logic of the planners. With little epistemological appetite or what Robic called "invisibility of the paradigm", geography was defined by the adoption of a "critical method" (Robic 1991, 59), under construction, a kind of passage between "the world of experience and scientific reason" (own translations) (Robic 1991, 59). On the one hand, the literary side of Vidal de la Blache's geography was closely linked to ecology and the description of the landscape based on the intuitive experience of the geographer in his fieldwork and his travels (based on perigenesis). On the other, a generalising appetite was not absent in the practices of geographers, that is, a systematic application from the point of view of a potentially abstract and general method (Robic 1991, 65). According to Robic, "the specificity of geography in relation to the autonomy of the social sciences is due to three structural features placed in the crux of tension between the natural sciences and the humanities. Paul Vidal de la Blache's success in founding the academic field can be accounted for by the way be utilizes writing styles and unpublished materials, both logical and expressional" (Robic 1991, 53).
However, in general, the human sciences are slowly moving away from literature and the canon of knowledge is gradually moving towards a greater systematisation of discourse and the applied use of the method, and France has also been a "victim" of this long duration (Lepenies 1996). The tension between literary and descriptive values, on the one hand, and general and utilitarian values, on the other, is long-standing and deeply rooted in scientifical thought of the geographical discipline.

Repositioning of the Long Durations in Brazil
Pierre Monbeig did not live with Vidal de la Blache, but his memory in France in the 1920s was very much alive. But for now, to reposition Pierre Monbeig spatially in Brazil, it is worth noting that the initial priority scale of analysis was the national scale, even if other scales were incorporated along with the evolution of t h e long durations. The characteristics of t h e Brazilian space matter. Emptiness, vastness, situation, tropicality and the opening of internal borders are the main Brazilian geographic arrangements that condition the themes, negotiations and adaptations of the theoretical subjects and methods of French geography in Brazil. Thus, the literary geography that in France was focused on the historical understanding of a certain equilibrium reached by the country, of a relatively small space, had to be placed at the service of the planning of the occupation of a continental and "empty" space in Brazil, 8 thus, an altogether different task.
While it may be assumed that this new demand would lead to a total shift in the scientifical paradigm, this was in fact not what happened. Despite the desire expressed by intellectuals and politicians to occupy this broad space, Brazil had technical and capital constraints that prevented it from understanding continental space and planning it with the same statistical instruments of the sciences that were already being applied in countries like Germany and the United States in the same period (the 1930s and 1940s). Therefore, the method applied to the intellectual instrumentalization of the Brazilian reality is a mix of a literary conception of geography, supported by field work, perigenetics, and the values of literature, and an applied conception, supported by the values of projection and statistics.
As far as conjunctures and displacements are concerned, it had to be understood that they also play a role: they accelerated the processes and revealed the cracks of long-lasting tensions, favouring the ruptures of the initial arrangements of the Vidalian paradigm. The main Brazilian conjuncture to which these long-lasting arrangements are submitted brings together the following simultaneous situations: an effort of diffusion, propaganda, the irradiation of French geography in Brazil; an invitation from a group of very important Brazilians in national politics (the so-called "Paulista elite") to French intellectuals to contribute to the processes of renewing science and the formation of a Brazilian school, also using this intellectual capital for political means; the consecutive tension between the autonomy of the group of young French scholars in the face of their sponsors; the pressures from other sectors of Brazilian society to change the Brazilian space and from the intervening state which was beginning to take shape; a context of the advance of capitalism in Brazilian territory, which would lead geographers to a hesitant but early tendency to apply and insert themselves into the practical field of intervention (Brazil is a country experiencing the advance of colonization over the hinterland); a feeling of urgency and need to boost development; a process of heightening awareness of social inequalities in the country. 9 In 8 Despite the existence of indigenous people. 9 More specifically, the events alluded refer to the conflict between the new populism and the oligarchies, which appeared in the Revolution of 1930 and the conflict for Paulista "autonomy" in 1932. After relations began to be pacified again, the Paulista elite had drawn up a plan to deepen the spread of their political objectives through the organisation of higher education (Cardoso 1982). However, in 1937, the regime closes further and soon after the coup d'état nouvelle (Estado Novo) took place, with the aim of combating regional discrepancies in Brazil and promoting national development through the action of a strong state (Penha 1993). In 1938, the Instituto Brasileiro de Geografia Estatística (Brazilian Institute of Statistical Geography) was created by the Brazilian State, an essentially geographic and planning institution, with which the generation of all Brazilian and French geographers remain linked. other words, I understand these movements as being inserted in a process of formation of the modern national state that aimed to pursue the insertion of a continental country in the modern nations, which tended to accelerate the movement of transformation of its spaces and, along with this, the potential transformation of a literary paradigm into an applied paradigm. Pierre Monbeig's leadership in the formation of the Brazilian geography had been subject to these tensions, posed by space and the historical context experienced in Brazil.
From the point of view of the intellectual field, the creation of a militant intellectuality in Brazil in the 1930s is observed, concerned with drawing "portraits of Brazil" (own translation) (Miceli 2001), and a tendency towards applied science, focused on the question of the valuation of the territory and, later, its development. These intellectual provisions are also linked to space, though not exclusively. To make the reasoning proposed here clearer, I reaffirm the argument that space is also an essential component for understanding the militant aspect of Brazilian intellectuality at the time, since the urgency of development is felt in Brazil, considering its continentality and tropicality, but also the fact that these intellectuals make their own ideological choices. An intellectual history embodied in the actors cannot ignore the spatial dimension of politics in Brazil.

The Evolution of Monbeig's Trajectory before Brazil: Training in France and Doctorate in Spain
Having explained these methodological presuppositions, I will proceed to their actual application by first focusing on the intellectual formation and circulation of Pierre Monbeig. As stated, he studied at the Sorbonne as a graduate student in Geography and History between 1925 and 1929. He is inserted in a moment of renewed French geography after the death of Vidal de la Blache, in 1918, and in the movement of knowledge of new research terrains by geographers after the First World War and important institutional conquests of geography such as the foundation of the Institute of Geography at the Sorbonne in 1925.
These three examples will be analysed.
First, I observe the tension between literary studies and applied studies materialising mainly around the tension between the French scientific model and the renewed German scientific model, after the rapid German industrial boom, relations which were reconfigured after the First World War.
The First World War, however, also gave French geographers, as experts on the subject, the opportunity to take part in the debate on the new borders in Europe and the consolidation of peace after the War, at a time when an autonomous discourse was being built. This French expertise takes the academic debate in the direction of a more technical and less monographic or literary language. The conflicts created by the War also break some old epistemological balances (Ginsbourger 2010). Thus, while there was, on the one hand, a strong influence of the German model, on the other, the ancient admiration for German science gives way to a spirit of independence and French geographers do not want to see themselves as tributaries of German science.
Thus, in the counterweight of this movement, a context emerges, in the intellectual field of geography, of profound nationalism, highlighting the French style of doing geography. However, the institutional renewal of geography, expressed in the case of the foundation of the Institute of Geography at the Sorbonne in 1925, continues and its central influence is the model of the German institute, more concerned with the practical interests of society than with scholastics of knowledge: practical exercises were privileged in relation to the great oratory exercises, modern languages in relation to ancient languages, projections in relation to history. However, the importance of not abandoning the French style continues to be stressed, in the words of de Martonne, one of the geographers who carried out (years after that declaration I quote above) the organization of the Institute of Geography in 1925: It can be seen from the concern for the practical side that the teaching of German geography tends to offer students not only ideas, general principles, but above all, ways of [practical] work, instruments of personal scientific research. It is a teaching method that is perhaps less for the past and more for the future (...). One sees practical spirits forming rather than method-oriented profiles. These tendencies (...) are generally opposed to the tendencies of our teaching (...). Predisposed by their literary and philosophical education to face facts from above, to connect themselves to ideas rather than to the facts which are the origin of these ideas, to the great relationships of cause and effect which are intertwined in geographical phenomena rather than to these phenomena themselves (...). It will be able to bring together the advantages of these two ways of seeing things and of completely different types of teaching, and, to put it better, which complement each other. (own translation) (De Martonne 1898, 259) At the time when Pierre Monbeig is at the Sorbonne as a student, important controversies about scientific and literary values are already taking place between three poles of understanding geography, within the direct disciples of Vidal de la Blache. In one way or another, the nature of these controversies will be reproduced in other intellectual situations in which Monbeig will see himself throughout his career and will push him by different forces and senses of the geographic poles and social situations in which these sets of values are aggregated.
In France, Jean Brunhes represents the total, and almost retrograde, preservation of Vidal's own methodological developments (Jean Brunhes advocates a geo-ethnography with strong literary assumptions); Lucien Gallois is the most faithful of the disciples in proposing the continuity of a regional geography; Albert Demangeon and Lucien Febvre defend a kind of aggionamento of Vidal de la Blache's conception of geography (from human ecology to the logic of understanding the economy and globalisation, without ignoring the analysis of the natural or ecological basis and the description of the landscape). However, unlike subsequent controversies, no one at the time doubted the capacity of a geography based on the values of literature to be able to explain modernity (or its roughness) to some extent. The nature of these controversies will shift its centre towards a growing negation of the geography of literary bias, as I have already stated, and with it the social and spatial poles that tension these ideas, and around which different scientific values and attitudes are added.
In his early work as a student, Pierre Monbeig allied himself with the reformers (Albert Demangeon and Lucien Febvre) in the way he embodied his studies at the Sorbonne. But the work that Pierre Monbeig (1929) does at that time also shows a strong duality. He studies the conflict between the "personality" of a small French region ("personality" is a "metaphor" much used by Vidal de la Blache), which progressively becomes the "banlieu" of Paris. In this work, he traces out two orders of causal linearity that are separate. One that starts from the natural reality, imposing transformations in the landscapes (the natural conditions that worsen) and the other in which the landscape changes according to the proximity of Yvelines de Paris (Monbeig 1929).
Monbeig uses a temporal retrospective to corroborate that Yvelines progressively loses its "personality" to become the "banlieu" of Paris: "Once again, the Yvelines region shows how it tends to lose its personality to be a part of the Parisian suburbs" (Monbeig 1929, 390). "Small and medium-sized owners are increasingly constrained to sell their goods to the Parisians and castle owners" (Monbeig 1929, 387). This new situation, in turn, is characterised in the first place by the modernisation of forest exploitation (Monbeig 1929, 385). Through the incorporation of machinery ("local labour is constantly diminishing") (Monbeig 1929, 385), family work begins to disappear ("since the end of the 19th century, these family métiers have practically disappeared") (own translations) (Monbeig 1929, 385).
The management of Vidal de la Blache's epistemological heritage keeps the master very much alive in the memory of his disciples, with priority contact with the spaces of France in general and Paris in particular. But the decade of 1920 would also be marked by the movement to export the French legacy to other countries. With a degree in Geography and History from the University of Paris, Monbeig leaves Paris for Spain to begin his doctoral thesis.
In 1930, Pierre Monbeig moved to Spain to do a two-year research internship at Casa Velázquez, with the aim of writing a thesis under the guidance of Albert Demangeon, a direct disciple of Vidal de la Blache. When Emmanuel de Martonne ascends to the position of new head of school, French geography is at a turning point with its capacity for intellectual export, so that geopolitics of knowledge and the understanding of the effect of globalisation on new research terrains offer the possibility of new intellectual and epistemological balances. In order to understand this new period, I am mobilizing a kind of reasoning that tries to perceive successive adjustments and a gradual erosion of literary values, opposing each other between poles where these values are agglutinated: France and Spain, which is equivalent to saying, on the one hand, a territory that is more slowly integrated with globalisation, France, and, on the other hand, a territory that is more rapidly integrated (albeit as a semi-periphery), Spain. This moment is part of a process of becoming aware of the crisis of 1929 and of a policy of association with French cultural diplomacy in favour of the expansion of its areas of cultural influence and the export of its intellectuals (Delfosse 2001), given also the already relative closure of the French university system which presents difficulties in creating new posts.
In this context, a theory of economic colonization resurges as an important forgotten research agenda of Vidal de la Blache. The curious thing is that the topic of colonization had already been the subject of disagreement between Vidal de la Blache and Marcel Dubois, who advocated an applied colonial geography of the French colonies. 10 This also reveals the gradual and specific movement of French Human Geography in the construction of its theories, its capacity to interpret the world through processes that go before and after the processes of economic globalization, as well as to export these tools of analysis to other territories were again in conflict. At the same time, I see an initial shift and an erosion of important literary principles of regional geography. Thus, if Vidal de la Blache represents one of the legacies of classical French geography on Monbeig's reasoning, a second pole of influences is represented by Henri Hauser, Lucien Febvre and Albert Demangeon, in a counterpoint to Brunhes even more marked than previously took place, when Monbeig was still a student. Before Braudel, Demangeon had managed to understand the geographical rhythm of the expansion of the world market in the geographical thought (Demangeon 1927(Demangeon 1929(Demangeon 1932. His conceptions would have profound effects on Monbeig's studies in Spain and later in Brazil. One aspect of Albert Demangeon's work that had and would continue to have an impact on Monbeig's work in Spain and then in Brazil, with the deep involvement in the methodological transformations of geographical discipline in the context of globalisation and internationalisation, is the formulation, in the context of geography, of the different temporalities of the local market (linked to nature) and the world market (linked to transport networks). In fact, it is Pierre Monbeig who expresses this understanding in an article published in 1934 entitled "Les importations de fruits frais en Grande-Bretagne" (1934).
Demangeon explains the imbalance caused by the 1929 crisis from a twofold ecological and economic point of view, stressing that the contradiction of these dimensions was what led to the imbalance of the global economy (Demangeon 1929). Demangeon expresses the view that agriculture continues to be commanded by natural forces in the places of origin. But the propensity of imperial mercantile circulation strives to annul the rhythms of the seasons that command the local market and produces an excess supply (Demangeon 1929). The world market cancels out the imbalance of local markets by varying the areas in which it drains its products. Thus, the world market, geographically, maintains an almost constant demand and distribution of products. Local markets, when closely linked to the propensities of global markets, are driven to annul the natural rhythm of their seasons, or the natural dynamics of the soil in terms of productivity (Monbeig 1934).
Thus, in studying Spain, Monbeig again reasons under a double type of separate causalities, which do not merge; this time he organizes one of these causal chains in the form of a systemic reasoning (Monbeig 1930(Monbeig 1931(Monbeig 1932: one is the deterministic and ecological causality, relative to a period or area barely affected by globalization, and the other a cyclical and systemic causality, in circuit, which explains the circuit of regions in modernization. These two orders also do not merge in their discourse. In my view, this shows the difficulty of French geography in merging the literary paradigm, which is more ecological, with the applied paradigm, which is more economic, demonstrating a slow transition. Monbeig notes that "(...) the huertas [Spanish] participate in modern activity: they tend to have a perfected and specialized agricultural production, thanks to the irrigation systems which constitute a beautiful technical undertaking (...). Industries also evolve; they tend to work for the universal market" (own translation) (Monbeig 1930, 606).
Moreover, of course, the literary descriptions of landscapes are always present in his works.

The Evolution of the Long Epistemological Trends of French Geography through Pierre Monbeig's Trajectory in Brazil
Monbeig arrived in Brazil in 1935 as part of a French mission specifically formed to contribute to the foundation of the first so-called "modern" university in Brazil, the University of São Paulo. He immediately became involved in the construction processes of this university. Also within the organizational process, I observe Monbeig's practices as a way of expressing his double engagement with France and Brazil, transported in the form of an engagement around a Paulista pole and a "Brazilian" pole, adopting at this stage of my work two other points of polarization. The conflict between a French model and a "Brazilian model" in the formation of the university now appears in a series of examples that make up the history of USP's formation. In fact, since Pierre Monbeig takes root in Brazil and changes the subject of his thesis to a Brazilian one about the pioneering fringe (and thus begins to build a personal project of a greater commitment to the country), what can be observed is that his research takes root in São Paulo, but with a vison of intellectual and organisational action in the whole of Brazil.
On the one hand, what was at stake was the formation of Brazilian geography, which, if its original pole was Paulista, maintained, at the same time, a national aspiration. On the other hand, the variation in scale between the poles of diverse intellectual environments shows that, as Monbeig distances himself from a Paulista pole, which is, in turn, "more French" in the Brazilian context (and without ever losing his own particular roots), he approaches a Brazilian pole -which is a crossroads of influences. And so, at the same time, he transforms his conception of science, which takes on features that are increasingly engaged and practical in style. Through the notion of literary and scientific values that vary in scale according to new polarizations, I observe successive and gradual adjustments of the longterm trends -here lies the key to contextualization as I reconstruct it. Simon Schwartzman explains how USP was primarily inspired by the French university organisation model: But there would be yet another very strong reason for preserving the preference on the French model [for the foundation of the university], disregarding the scientific importance mainly of the English and American models, where experimental science (and not "principalist" like the French and the German ones) (...). In the view of the founders of the University of São Paulo, the Faculty of Philosophy, Sciences and Letters would bring together all the branches of science in one single institution, giving it a high power of generation and irradiation of a new mentality. The paulistas were in a hurry (own translation). (Interview with Paulo Duarte in Schwartzman 1979, 200) I note that in the initial stages of institutional action, in the circle of affinities of Monbeig in Brazil, there is a strong resilience of the old balances of the French Geography through the resilience of Paulista values in the organization of the university. However, as Monbeig's Brazil becomes "more Brazilian", either due to the influence of other global and American currents, or to the very discovery of new landscapes revealed by the march towards the west and the entrance into the hinterland by Getúlio Vargas' government, or even by the entrance of low-income students (usually children of immigrants and migrants in the FFCL, in general with more pragmatic behaviour), there is a strong tension in these old epistemological balances of French geography. 11 The question is whether, as Monbeig spends more time in Brazil, he submits himself to this tendency towards fragmentation that the Brazilian territory seemed to be able to displace the French Geography still very much anchored in literary values, and whether it changes its geography gradually or radically. Both possibilities existed in Brazil.
Brazil in the 1930s was a country that wanted to modernize and transform its landscapes in the same march of modern civilizations, with one of the outcomes of this movement being the creation of the National Council of Geography (CNG) and the Brazilian Institute of Geography and Statistics (IBGE), in 1938, in the country's capital, Rio de Janeiro, besides many other actions carried out by the Vargas government. It is also worth remembering that São Paulo and Rio de Janeiro were the main opponents of an almost separatist war, when São Paulo fought for a supposed autonomy of the nation in 1932. However, the desire to transform the Brazilian space did not live up to reality.
By around 1940, the scientific institutions that sponsor the Brazilian geography are already consolidated, and Pierre Monbeig makes a personal effort, together with his disciples from São Paulo, to nationalize the so-called modern geography throughout the Brazilian territory, also involving the IBGE. Thus, I first used the intellectual poles, that opposed the different disciples of Vidal from their understandings about globalization. I then used the different poles surrounding the space of France and Spain as expressions of the agglutination poles of the long-term tensions of geography. I will now transpose Brazil and France as poles of repositioning tensions, transported to the cities of São Paulo and Rio de Janeiro, since Pierre Monbeig is already relatively immersed in Brazil, as well as the institutional poles and collective groupings that involve French and Paulista intellectuals, on one side, and Cariocas and Brazilians on the other. This was the way in which I began to polarize Monbeig's heritage when he was located in São Paulo, parallel to a process of national irradiation of his geography.
In this context, by nationalizing his way of practising geography, Monbeig's work ends up developing a geography of the colonization of Brazil, relating it to the genre de vie of the remote and old areas of Brazilian territory (the genre de vie is an ethnographic concept of Vidal de la Blache) and understanding the processes of the spatiality of the economic forces in the context of colonization. Therefore, he elaborated a regional geography of colonization that could be understood and commanded by the national state. Again, there is a shift from the centre of the controversies surrounding the erosion of the literary values of geography. But there is also a widening of the range of positions related to the poles furthest from the new controversy. Thus, while, on the one hand, he already adhered to the idea of colonisation (which, it should be repeated, was outside the hard core of regional thinking in the early 19th century), on the other, he refused to draw up an applied theory of market-led colonisation, another pillar through which the colonial geography of Marcel Dubois, in conflict with Vidal de la Blache, revolved in the past. In this sense, the fact that he dedicated himself to the subject of colonization in the way he practised geography in Brazil also revealed a hybridism of perspectives. The topic of colonization was a preference already defined in Spain and continued in Brazil. Thus, although Monbeig strongly preserves his inheritances, Brazil was also able to represent the need to review the French geography. It represented a space of colonization, in the face of the expansion of capitalism in the mid-20th century, or late capitalism, where the force of humankind's action in the face of nature is even more important but directed by the state. It also represented the concrete impossibilities of abandoning Vidal de la Blache's geography in depth, in Monbeig's eyes, because of the very constraints that the great emptiness placed, this time, on applied studies.
Ephemeral wealth, that of the pioneer fringe, may object. But there were solid fortunes built. (...) for Brazil, the hour has passed when all production is destined for export trade; there is a national market for products of the land that widens year by year. The pioneering economy can only profit from this development, because it will be less exposed to the crises of the world economy. (own translation) (Monbeig 1894, 391) Thus, the incorporation of Brazil and of a whole regional geography into the process of the universalization of French geography can be seen as a stimulus to its modernization, not only of its methods, but also of its institutional and political strategies. In 1940, Robic speaks of a return to the paradigm of Marcel Dubois' colonial geography (Robic 1996, 46) (with all the nuances and specificities of this movement). Colonial and political geography that were observed to be marginalised at the beginning of the century at the core of the French geography and that represented another paradigmatic arrangement of modern geography (Clerc 2017) was renewing. As can be seen, the tension between literary and scientific values, now called "applied" geography, is at this time being played out by other intellectual and political attitudes, that is, it has new injunctions, reformulated by intellectual devices of successive approaches to new realities and spatial and sociological poles of agglutination. I observed at this moment that the essential tension between the paradigms in Brazil surrounded a liberal conception of geography or a nationalist conception, guided by the State, which sought to evoke the irreducible singularities and particularities of the nation and its embarrassments, at the same time as intervening in space and society. In other words, it is no longer a totally disinterested conception, as literary values could valorise, nor a cosmopolitan, highly abstract and planned, theoretical and liberal application that would make a blank slate of the singularity and constraints imposed by nature on man's actions. Monbeig's organisational presence andtheoretical elaboration in Brazil between 1946 and1952 (year of publication of his thesis) reached its peak in 1952. In theoretical terms, Pierre Monbeig was responsible for the formulation of a geohistory of the colonial expansion of capitalism in Brazil and from Brazil. At that time, Monbeig's proposal involved a reformulation of the cognitive tools of the human sciences, mobilizing synchronic, diachronic and stratigraphic reasoning (articulating systemic, cyclical and vertical causal relationships). Moreover, Pierre Monbeig is increasingly taking upon himself the task of understanding the condition of social inequality in Brazilian society and of considering Brazil within the framework of its development.
Thus, later, around 1957, he began to think in terms of "development". For him, the challenge of Brazilian development is, first of all, to install infrastructures that allow the entire population to live under the same temporality, promoting a transformation of the planteur [planter, or "plantador" in Portuguese] into a peasant, more rooted in the soil and creating an internal market in which goods and capital circulate, and where the peasant takes care of the land. Moreover, strengthening the internal market would be a way to cushion the setbacks suffered as a result of international economic cycles. For Monbeig, the creation of the domestic market is a direct consequence of the care of the "natural processes" and they can only be favoured by "closing" the border (the idea of an internal border seen as constant advancement, a horizontal dynamism). Pierre Monbeig himself calls this entire intellectual process in Brazil "neo-regionalism", revealing an awareness of its intrinsic originality: For this neo-regionalism to be likely to succeed, it was necessary to choose the branches of production to be intensified or created in the different regions wisely, which would presuppose perfect knowledge of natural resources and no less perfect knowledge of economic and human problems. Here, too, geographical research was called upon to collaborate in this eminently practical work. (...). It would be unfair not to refer to Brazil, where geographers are used to conducting practical surveys. The very creation of the National Council of Geography aims to provide the best knowledge of the country, indispensable for its good administration. (own translation) (Monbeig 1957, 22) In addition, he proposes standard forms of thinking about the economic movement on the territory (which he named "urban cycles" in 1948) on the advance of modernizationdriven by international demands, which, by clashing with traditional tropical structures, produces a structural imbalance of the territory in continual expansion, and which, moreover, slows down the process of securing the worker and renews conditions for his exploitation, as well as delaying the formation of an internal market through natural processes. From my point of view, he anticipates a geographical understanding of underdevelopment, although the word will perhaps be consolidated in the international scientific arena a little later. This idea, he later expresses as: Agriculture and rural society in tropical countries have two radically opposed aspects, at least in most cases. On the one hand, traditional subsistence agriculture carried out by human groups with archaic techniques, old structures and a local or domestic economy. On the other, a sector that is less interested in national than international markets (...). Thus, 'developed' agriculture and 'underdeveloped' agriculture coexist. It is precisely this duality which constitutes one of the forms of underdevelopment of the intertropical zone, while at the same time being one of the essential causes. (own translation) (IEB, PIERRE MONBEIG FUND. Untitled. On malnutrition in the intertropical world. Box 5, doc 14. SD) In Brazil, Monbeig's thinking leads to a geography of development and I think this process of constructing his way of thinking is situated. Moreover, this process of transforming French geography into Brazilian terrain contributes to advancing the erosion of literary values, tending to place geography at the service of state human action, within a re-founding conception of the structure of Brazilian society, and normatizes the geographical method in the form of establishing reasoning that conveys standard forms of functioning. Finally, in 1957, Monbeig was involved in an important dialogue with another Brazilian historiographic current, the economic theories on the development, which proved to be one of the great critical fortunes of Brazilian historiography. However, he remained loyal to his heritage.
Monbeig's thesis is almost a literary book. He strives not to create a model about the functioning of the pioneering fringes, not to create a theory, but to describe individuals and landscapes, to dilute almost all systematic thinking amid beautiful descriptions, although, in my view, his thesis has the objective of building a model, without actually doing so. A "model" of the advance of capitalism through the territory of Brazil explained, moreover, under the injunctions of the resistance of a geography that advocates literary values. The idea of building models on the basis of description may be a potential contradiction in global epistemological reasoning, but it seems to be a contradiction which geographical tradition has internalized as the driving movement.
Pierre Monbeig was not one of the actors that endorsed the detachment of the literary paradigm from the applied one, which would take place in terms of a deeper appreciation of the agency of man, the elimination of natural determinism, the tendency to understand the generalities of global landscapes, to applied geography and a political and planning bias. Thus, he developed a regional thought on colonization (regional because it is constrained by nature), which originally brought together elements of geography that were epistemologically different in the past (Vidal de la Blache -regional -and Marcel Dupois -colonial) and finally ended in a geography of development. In this final movement, he seems to have diluted his main theoretical originalities in a discourse in which the backdrop was the reaffirmation of his legacies. This, because he seems to have been carried away by the long durations, the spaces and the conjunctures, but also by his own vision of a realistic temporality, whether of the occupation of continental spaces like Brazil; of the intellectual thought that accompanies him as a whole; or of the difficultly to overcome colonial and territorial historical liabilities. Thus, Monbeig had his own convictions. They were based on his deep individual predispositions around the geography of Vidal de la Blache and his disciples. Brazilian geography is also the fruit of Monbeig's individual choices, despite the Brazilian space. Monbeig chose to remain within the framework of a literary and regional geography, but the way in which long periods of geography evolved, and repositioned themselves in Brazil, led him to make necessary adaptations to the original paradigm towards a geography of development.

Conclusion
This paper addressed the way I use a geographical method in intellectual history and the history of science, the geohistory of knowledge, and the problem of the specificity of Brazil in the field of world-science. The trajectory of the French geographer Pierre Monbeig acts as the axis of the analysis. For me, the Brazilian space and the subsequent injunctions and mediations ended up tensioning Monbeig's originally literary geography to a geography of development, promoting hybridism between a historical and projective, ecological and economic, descriptive and practical vision. A geohistory of knowledge, enriched by the contributions of intellectual history, was the method of analysis used to reveal the globality and specificity of Pierre Monbeig's trajectory in Brazil. At the same time, it was able to reveal how the Brazilian space and context itself guided Pierre Monbeig's geography in a certain direction. As a result, I believe I have succeeded in highlighting a certain specificity of Brazilian geography in world-science, which I believe is to provide answers to the problem of development.