Geographies for Another History: Mapping the International Education of Architects from Colombia (1930–1970)

The history of Colombian architecture is poorly understood. This article maps the educational geographies of over 200 Colombian architects between the late 1920s and 1970, examining the historical, geopolitical, and disciplinary shifts that contributed to the international advancement of Colombian architecture. In the 1940s mobility was reoriented from Europe to the USA, while in the 1950s Brazil supplanted these destinations, becoming the main Latin American pole for Colombian student architects, and the Brazilian modernist repertoire was subsequently diffused in Colombia. This article revises long-held ideas about the architectural historiography of Colombia, expanding the geographical scope of the country’s leading architects to reveal the significance of the Americas in their education.


Introduction
In the 1950s, the historians Jorge Arango and Carlos Martínez acknowledged the diverse, international background of 20th-century Colombian architects as an essential feature of the country's architectural environment: As there were no architectural faculties in Colombia until 1936, the majority of the professionals who made up the country's contemporary movement, and who founded the country's first architectural programs, were trained in other countries. The schools of Perret and Le Corbusier arrived from France, accompanied by the reaction against them. Architects trained in Italy and England, in Belgium and Switzerland, in Germany and Spain also arrived in Colombia. Others came from North American schools influenced by Frank Lloyd Wright, Gropius, or Mies van der Rohe, and some came from Beaux-Arts affiliated schools or technical institutes, where construction remained the primary focus. Not a few also came from Chile, whose schools date back to the past century and where the profession has been subject to a continuous process of integration. (Arango and Martínez 1951: 30) 1 Of the 24 Bogotan architects and engineers responsible for the first significant contemporary architecture in Colombia, 16 were foreign-trained (Arango and Martínez 1951: 130-32). Although many architectural faculties were established in Colombia in the 1940s, students increasingly trained abroad in the next decades. However, the educational background of Colombian architects has been largely neglected. 2 Only recently has the subject of Colombian architectural education been addressed, in the work of Maarten Goossens (2013) regarding ties of Colombian architects with the USA.
This article maps the educational geographies of Colombia's architects 3 from the late 1920s to the end of the 1960s. 4 The architects' educational backgrounds, along with the source of that information, are presented in the three tables at the end of the article. The maps (in Figs. 1, 3 and 6) were generated from these tables (which are a work in progress). The numbers in the maps refer to degrees in architecture, urbanism, and urban/regional planning, and to shorter specialisation courses. Specialisation includes courses often offered by non-academic institutions (e.g. the Bouwcentrum in Rotterdam, or the many municipal or ministerial centres in France and the UK), or by universities, if the programme did not lead to such degrees as a bachelor of architecture or a master's of architecture (e.g. restoration studies at Rome's Scuola di Perfezionamento). Complete academic urbanism/planning programmes (e.g. those offered by the Institut d'urbanisme in Paris, or those in the US leading to master's degrees) are counted as degrees, whereas shorter courses are counted as specialisations. 5 This mapping and its proposed interpretation highlight the complex patterns of transfer, circulation, and reception that contributed to the modernist movement in Colombia, in which ideas were both drawn from international models and generated within the continent. As for the longstanding debate on internationalism and regionalism in Latin American architecture, this study suggests that international ideas complemented regional and local developments. 6 It is crucial to recognise Colombia as a fluid territory of artistic and technical exchange. To quote Del Real and Gyger (2013: 3), this means moving ' our understanding of history beyond the limitation of the national into more ambiguous territories, illuminating moments of intra-regional dialogue around specific challenges and solutions'.
As this discussion demonstrates, an internationally wellconnected, foreign-trained group of architects transferred theories, models, formal repertoires, and techniques to Colombia. The investigation, using the histoire croisée approach (see Werner and Zimmermann 2006), also reveals the different though intertwined dimensions of architectural transfer, such as the mobility of architects and the contents of their architectural magazines. 7 We can then move beyond the characteristic limits of transfer and comparative studies to better understand transactions among different aspects of the architectural profession. The present study thus diachronically explores geographies of artistic exchange influenced as much by historical and geopolitical reasons as by cultural and disciplinary motives, and delineates mid-20th-century architectural education.
This research derives from a wide range of archival and bibliographic sources, verified through extensive cross-referencing. The records came from university archives that often contain resumés or student enrolment papers, university yearbooks, and digital databases by Colombian professional institutions. In addition, books and periodicals on architects were consulted for biographical information. The accuracy of sources appears to decline over time, making older publications more reliable than more recent texts based on second-hand tales and lacking references. A few original oral sources were also consulted. This data facilitated identifying over 200 architectural professionals, whose educational profiles are organised chronologically and geographically in the appended tables. An overview of architectural education by institution was also created according to bibliographic sources and previously recorded oral reports of Colombian students. 8 Beyond this, the examination of a limited number of theses, reproduced in magazines and conserved in one archive, enriches our understanding of the teaching environment of architectural programmes in the mid-20th century. The magazine Proa serves as an essential source, as it published projects from Colombian students abroad and documented their training, thus tracing the international foundation of important Colombian architects and projects. 9

Until 1936: A European Tradition
During the first two decades of the 20th century, architecture in Colombia was practised mostly by engineers or foreign architects, 10 although technical training programmes akin to architectural schools already existed (see González 2013). Few Colombians had received a foreign degree: in Bogotá, Mariano Santamaría, who studied in Weimar (Arango 1984: 11); in Medellín, Juan and Dioniso Lalinde, educated in the UK and New York, respectively (Molina 2001). By the late 1920s, aspiring architects began to study in Europe and, to a lesser extent, the USA, pursuing a trend already common in other professions. Since the mid-19th century, Colombian elites had travelled to France, the UK, and Germany to study medicine, engineering, and applied sciences (Martínez 2001: 212-18). 11 It may be argued that the presence of a few recognised foreign architects in Colombia in the 1930s, and the economic and urban growth the country then experienced, justified the investment of earning an architectural degree abroad. 12 Three quarters of Colombian architectural students who travelled abroad went to Europe, about equally to France and Belgium, where the main destinations were the École des Beaux-Arts in Paris and Brussels, the Saint-Lucas campuses of the Université catholique de Louvain in Ghent and Brussels, and the Institut d'urbanisme in Paris (Table 1 and Fig. 1).
The architect Antonio Mesa of Medellín divided his study between Belgium and France, while Ignacio Vieira studied at Brussels' Academie royale des beaux-arts. The Institut d'urbanisme also hosted many Colombians, even after World War II. Notable Institut graduates include Carlos Martínez, a leading figure within the Colombian architectural world who studied under the supervision of Marcel Poëte and in 1946 founded the architectural magazine Proa. The Institut appealed to Colombian students, who sought to learn urbanism in one of the few centres in the world providing such a degree. Its teaching theories and methods were far from the functionalist proposals of CIAM architects in that period, as evidenced by its graduates' final projects. For instance, the Colombian student Severo Reyes Gamboa submitted a thesis in which he designed a scheme for an industrial city on the outskirts of Cali similar to the Argentinian ideal city of La Plata, but which aligned public buildings along a ring as in Vienna (Fig. 2).
Other Colombian architecture students moved to Germany, Italy, and the US. Julio Bonilla, for instance, was the son of diplomats living in Europe. Bonilla graduated from the Technische Universität München in 1933. Upon his return to Colombia, Bonilla advanced a German language of architectural modernism, designing a dormitory building in Bogotá for the Universidad Nacional in a clearly Bauhaus fashion. At this time, the Universidad Nacional de Colombia (UNC) in Bogotá hosted many other notable architects with degrees from abroad. Jorge Camacho Fajardo, who graduated from the Politecnico di Milano in 1934, was among the first professors of architecture at UNC. Upon his return to Colombia, he submitted the design for a penitentiary showing many formal features typical of Italian Fascist architecture (Camacho 1936). Other Colombian students travelled to the US, although in fewer numbers. John Sierra and Nel Rodríguez from Medellín graduated from Columbia University, the latter studying either architecture or engineering. Some of the most relevant architects in Bogotá also studied in the US. The Carrizosa brothers, Guillermo and Hernando, graduated from the University of Michigan and Purdue University respectively, and Jorge Luzardo studied at the University of Kansas. Far fewer students remained in Latin America. Pablo de la Cruz, who    (Arango 2003: 52-53). The number of Colombian architects educated abroad in the 1930s was smaller than in the following decade, but not inconsequential. The faculty of architecture at the UNC, established in 1936, benefited not only from Colombians educated abroad but also from particular international and national circumstances. A skilled group of Spanish, German, and Italian architects immigrated to Colombia, away from the Spanish Civil War and rising totalitarianisms in Europe and encouraged by the immigration policies of the liberal governments of Alfonso López and Eduardo Santos (Palacio 1995: 161). By 1941, almost 50 percent of UNC's professors held a foreign degree. 13 The impact of UNC's international faculty has been widely recognised by historians: in particular the role played by Leopold Rother and Bruno Violi (Arango 1989), and Carlos Martínez, a major Corbusian propagandist. In the time they spent abroad, Colombian architects also had the opportunity to make personal connections with important European architects. For example, Antonio Mesa met Le Corbusier during his stay in Paris 14 and later contributed to an early reception of the Swiss architect's work in the faculty of architecture of the Universidad Pontificia Bolivariana (established in 1942) in Medellín. Mesa's relationship with Le Corbusier ensured that in advancing the architect's ideas, he needed no further intermediation from the professional, academic, and editorial world of the capital.

The 1940s: The Great Shift towards the US
In the 1940s, new faculties appeared in both Medellín and Cali (the one at Universidad del Valle was established 1947) and in the capital (where the architecture faculty of the Universidad de los Andes was founded in 1948). However, despite these developments and the success of Bogotá architecture school, more Colombians headed abroad. Moreover, educational geography changed substantially. The Second World War spurred massive emigration of architects and artists from Europe to the US (Kentgens-Craig 1999), and the landscape of modern architecture changed radically. As European masters took over in many North American universities, US schools (particularly the Catholic University of America in Washington DC, the University of Michigan, Cornell University, and Harvard) became the favoured destination of Colombian students ( Table 2 and Fig. 3).
It is important to understand that neither the academic environment of American architectural programmes nor their teaching methodologies were always progressive or cutting edge. Although many programmes welcomed foreign students and professors, the ideas and curricula of North American programmes often remained deeply traditional. In certain aspects, the Bogotá UNC faculty was more progressive than many North American ones. 15 Jorge Arango recalls that in the early 1940s, many Colombian architects trained in the USA were doing 'traditional' design, meaning Spanish or Georgian (Arango 2003: 169). Arango may have been referring to prominent figures of the Colombian Society of Architects such as Álvaro Hermida Guzmán or José Gnecco Fallón, both educated at the University of California, an epicentre of the neocolonial tradition. In Cali, major architects working in a neo-colonial manner were all trained in the US: Alfonso Garrido, from the Massachusetts Institute of Technology, and the Calero Tejada brothers, Álvaro and Hermann, respectively from the Rensselaer Polytechnic Institute 16 and California, where their Mexican brother-in-law Félix Mier y Terán was also educated (Ramírez, Gutiérrez and Uribe 2000: 77-105).
A great number of Colombians studied at the University of Michigan at Ann Arbor, one of the first American universities with an architectural curriculum. 17 The programme was formalised in 1906 by Emil Lorch who, in introducing the Pure Design method, favoured Beaux-Arts principles over abstract exercises. In 1923 Lorch invited Eliel Saarinen to teach at the College of Architecture and Design, and later, the Danish architect Knud Lönberg-Holm also joined the faculty (Alofsin 2012: 98;Bartlett 1995). After World War II, the College of Architecture and Design distinguished itself for the research it conducted in prefabrication, and for its merge of architecture and city planning according to Gropius's idea of teamwork (Ockman and Sachs: 2012 [Rodríguez] and I, with our studies at the Nacional, belonged to the vanguard, which the school did not fight and preferred to let free. Even outside the school the situation was so uncertain that a firm from Detroit contacted me to work in a project for a country club only because one of the firm members said that I used to make architecture like that seen at MoMA. (Rodríguez 2008: 24-25) In the Midwest, the Illinois Institute of Technology, then directed by Mies van der Rohe, pioneered progressive modern architecture. Its influence and Ann Arbor's proximity to the East Coast certainly favoured the spread of modernism and direct contact with many important figures  Cano, attended the Catholic University of America in Washington, DC, whose architecture faculty in the 1940s was directed by Frederick V. Murphy according to Beaux-Arts principles. There were many reasons for this choice: preference for a Catholic institution, proximity to the diplomatic environment, and personal recommendation (Goossens 2013). Other Colombians chose more elite schools, such as Syracuse, Cornell, and the Pennsylvania. The latter remained a Beaux-Arts stronghold until the end of the 1950s (Alofsin 2012: 115).
Also notable are the few students who took postgraduate courses at Yale and Harvard in the 1940s. Jorge Gaitán and Jaime Nieto studied at Yale, while at Harvard, Álvaro Ortega and Gabriel Solano studied with Walter Gropius and tightened their personal relationship with Marcel Breuer. Eduardo Mejía, among others, also attended Harvard. 18 The two schools shared little more than a great reputation. Harvard, together with the IIT, represented the bastion of European modernism transplanted to the US. Major European masters -Gropius, Marcel Breuer, Martin Wagner, and Sigfried Giedion -came to Harvard following Joseph Hudnut's reform of 1935, which unified the schools of architecture and landscape architecture with city planning under the umbrella of the new Graduate School of Design (Alofsin 2002 After a long journey from Barranquilla, Gaitán arrived in New Haven in April 1943. There, he met his future office partner Jaime Nieto, as well as several other prominent Colombians, such as former president Eduardo Santos. In a few letters to his family, Gaitán told of his first weeks in the country, relating with some astonishment the high living costs and expressing his relief at having received a scholarship for the first two months of study (JGCPA-1). But Gaitán embraced his time at Yale as an opportunity to expand his horizons. In New York, he encountered modern abstract art at the opening of Alexander Calder's Mobiles exhibition, where he also met Sert. 19 By May of his first year, Gaitán had started work on a hotel project, which later became his thesis and was published in Ingeniería y Arquitectura (JGCPA-2; Gaitán 1944). The project was designed in a modernist fashion with certain Cariocan reminiscences, such as the brise-soleil and organic shapes within regular blocks (Fig. 4). Although the outbreak of World War II undoubtedly represented the main impetus behind architects' abandonment of Europe in favour of North American universities, other factors contributed to the post-war continuity of this trend. The European shift to America should be read within the geopolitical framework of Franklin Delano Roosevelt's Good Neighbor Policy, as well as by specific economic and cultural strategies adopted during the War -the promotion of Brazilian modernism being the emblematic case (see Cavalcanti 2004;Del Real 2012). 20 During the 1940s, ties with the US became visible in the academic field. Already in 1939, Gabriel Serrano visited the New York World's Fair and various faculties of architecture along the East Coast. The same year, Jorge Triana (1940), dean of the Bogotá faculty of mathematics and engineering, undertook a similar journey. Serrano's tour included four renowned institutions: Columbia, Harvard, MIT, and Yale. At Columbia, as Serrano noted, students produced complete projects in full detail within integrated workshops while under Gropius's guidance: The Harvard school completely broke with the traditionalist school; their problems are real, and their solutions fit the most modern technical advancements; with plenty of reason, much importance is given in projects to issues of construction, so its tendency is to make a functional architecture. (Serrano 1939: 24) In 1940, Roberto Ancízar, dean of the faculty of architecture, participated in a similar trip throughout many US universities (Ingeniería y Arquitectura 1940). In Chile, Lincoln Kirstein, a member of MoMA's Advisory Committee, approached Arango and the Department of State invited him to the US (Arango 2003: 111-112), while at the end of the 1940s Hernán Vieco, according to L'Architecture d'aujourd'hui (AA 1958a), visited different architectural schools in the US during a trip funded by the Rockefeller Foundation. Later, in the 1960s, the Rockefeller Foundation sponsored exchanges between Cornell and the Universidad del Valle for students and professors of architecture. But more than bringing Colombians to the US, this programme increased the number of Latin American professors in the Americas. 21 During the 1940s every UNC staff member or student returning from the United States was asked to submit a report about architectural pedagogy in the schools they visited. 22 In the Bogotá faculty, reference to North American programmes was also a constant theme in discussing the school's own architectural curriculum. A real effect became visible only in 1948, when Eduardo Mejía became dean after returning from Harvard, and a new curriculum was introduced. This reform has been labelled as 'Bauhausian', as it diminished the role of subjects such as architectural history (Niño 1987: 52-53). But the principles underlying the programme's reform did not entirely derive from the Bauhaus model. The school adopted the precepts of realism that characterised Hudnut's reforms at Columbia and Harvard (see Alofsin 2002). In a report about the study plan, Mejía wrote, 'in the part concerning urbanism, works on actual Colombian problems are undertaken, with all possible closeness to reality'. 23 Furthermore, the integration of workshops with different subjects recalled Gropius's ideal of teamwork. However, the curriculum still promoted a broad training of architects -in building design, urban planning, and engineering -contrary to the principles of specialisation promoted by US schools, and indeed, the very structure of architectural studies in these programmes.
The tie between leading Colombian architects and the US also paved the way for the international reception of Colombian modernism. Francis Violich observed that 'more functional design is to be found in Bogotá, Colombia, where a group of younger men, several of whom have been trained in North American architectural schools, have had an opportunity to build office buildings, apartments, and residences (Violich 1944: 128). The direct ties of Ortega, Solano and Gaitán with North America favoured the diffusion of their architecture; in particular, the 11 de Noviembre baseball stadium in Cartagena and the bus workshop and station in Bogotá. These works circulated within major international magazines like The Architectural Forum, Architectural Record, and L'Architecture d'aujourd'hui, defining the first international reception of Colombian modernism. In a special issue of the French journal dedicated to Gropius's influence, such as his teaching at Harvard, the baseball stadium in Cartagena became the first Colombian building to receive international press coverage (AA 1950: 93). 24 Later on, Ortega, Solano and Gaitán, together with Gabriel Serrano and his firm Cusego, gained even more visibility in Henry-Russell Hitchcock's book Latin American Architecture Since 1945, where Colombia was the third most represented country.
For Hitchcock, the real North American influence on Latin architects came not from the work, scattered throughout the continent, of US firms but from the simple fact that many Colombian architects had trained in the US: A very considerable proportion of the best Latin American architects, therefore, particularly those under forty, owe at least the final stages of their professional education to the architectural schools of the United States. It is not alone the more famous and old established schools or those that have been headed by world famous architects like Gropius and Mies, not just Harvard, Illinois Institute of Technology, Yale, Cornell, and Columbia, but less internationally known schools such as the University of Michigan, Georgia Institute of Technology and the Universities of Oregon and Florida […]. Nowhere [than in Colombia] are there more architects whose training is in North America and many of the characteristic problems they have faced are more familiar in North America then elsewhere in South America. (Hitchcock 1955: 20-37) Nonetheless, despite the predominant shift towards North America in the 1950s, Europe's influence had not entirely faded. But within Europe, architectural culture had evolved. The Beaux-Arts schools were no longer a valued destination; interest was restricted mostly to British universities and some French specialist programmes. Many Colombians headed for Europe, mostly for specialisation courses in topics such as urban planning and prefabrication. Often, these students were already recognised architects and academics, such as Hernando Pinzón Isaza, a professor at the UNC in Bogotá, 25 who attended some courses on urban planning, construction, and mathematics at Durham University between 1947 and 1949. While at Durham, Pinzón wrote to the UNC dean: Professor Allen, dean of the department of urbanism of Durham University, my director and supervisor, is preparing for me a series of visits in other universities and different sectors of England and Scotland, where I will be able to see in practice not only the ongoing reconstruction in the cities destroyed by the War, but also the planning and construction of new satellite cities and the solution of the housing problem by means of prefabricated houses. 26 In his final report submitted to the UNC faculty, Pinzón explained his intention of modifying his courses in geometry, trigonometry, and construction once in Colombia, and adding a section on prefabrication. He had visited a factory where houses were produced with concrete and a minimum use of steel, making this technology potentially suitable for Colombia, a country that lacked this material. In England, Pinzón also attended courses, organised by other institutions, on urban and rural planning and on theory of construction. He also visited various cities, discussing new urban design programmes with those in charge, as well as factories of prefabricated houses and neighbourhoods built with this technology. He noted that reconstruction problems of the 1950s were the same as 50 years earlier, but were now tackled according to CIAM's four functions: Nowadays, the plans are not to achieve an appealing drawing on a paper sheet, following axial symmetry or shaping circles and semicircles, or tracing radial avenues around a monument or a cathedral; [reconstruction has] been designed for people who live and work in the city, with a logical relation between home and workplace; it has been designed to provide a happy and hygienic life to a population adjusted within certain limits, with an adequate coefficient of density and in an area where the main factors are not limited to air, light, beauty, space, and vegetation; but also pedestrian safety, free vehicular movement and transports' efficacy and rapidity. 27 Pinzón's report ended with some suggestions, such as funding a faculty of urban planning, and sending a group of architects to the United Kingdom to study urbanism. 28 This interest in British planning appears to have developed in many Latin American countries at the same time, as a panoramic reading of major architectural magazines proves. Not only Proa and Ingeniería y Arquitectura, but also the Cuban and Uruguayan Arquitectura, the Mexican Arquitectura México, the Argentinian Nuestra Arquitectura and Revista de Arquitectura, and to a lesser extent the Chilean Arquitectura y Construcción, all reveal a common attention towards Great Britain. All over the continent, articles on the reconstruction of British cities and the development of prefabricated housing and school buildings were circulating, thanks to both the magazines' networks and British propaganda efforts. In Colombia, Ingeniería y Arquitectura and Proa devoted a significant number of articles to the same themes. In Bogotá, the first exhibition on international architecture was dedicated to British production, 29 while the Spanish architect Santiago de la Mora, a professor of urbanism at the UNC since 1946, promoted Patrick Abercrombie's theories, whose book Planeamiento de la ciudad y del campo he had translated into Spanish in 1936.
Although the decade of the 1940s seems to have been dominated by the relationship between Colombia and the US, given the large number of architects who trained abroad, historians of Colombian architecture have been reluctant to acknowledge these circumstances. Instead, their attention has focussed on the contribution of European immigrants in Colombia and Le Corbusier's arrival in the country in 1947. In that year, however, 32 of the 72 (44%) professionals registered at the Sociedad Colombiana de Arquitectos (SCA 1947) had spent at least a period of study abroad. Moreover, 18 of the 72 (25%) had gained experience in the United States. Therefore, Maarten Goossens's claim regarding the necessity of refocusing on the Colombo-US relationships, and eventually redefining the 'generation Proa' (Arango 1989: 211) as the 'generation USA' (Goossens 2013), seems convincing. The many consequences of this rich dialogue have still not been sufficiently explored. Numerous projects completed in Colombia show a direct link with the works studied by Colombians during their stay in the USA. The hospitals with flat roofs and white surfaces, designed by Serrano's students in 1939, reveal a clear Gropius influence (Goossens 2013), while Felipe Rolnik's project for a church in Quiroga, Bogotá, is a clear homage to Eduardo Catalano, his professor at North Carolina State University (Proa 1955a ; Fig. 5). The entire residential production of firms such as Obregón & Valenzuela or Pizano, Pradilla & Caro also shows a great affinity with the houses of Mies, Breuer, and above all, Richard Neutra.
However, the training received by many Colombian architects in US schools affected Colombian architecture in more than just formal terms; it also defined its very mode of production. More than quickly consumed formal repertoires, these experiences proved to be rooted in the way architecture was conceived and built. Architects began to work according to precise models and technologies adapted to local material conditions. Concrete construction developed in response to the lack of structural steel (Téllez 1979: 395;Vargas 2009), while typical North American elements such as the curtain wall were successfully adopted (as Hitchcock's chapter 'Urban Façade' proved;1955: 191-97) and were visible in the work of many firms, from Cusego to Sudarsky & Menéndez. 30 Goossens (2013) highlights the diffusion in Colombia of US installations and construction products, and, by the late 1930s, the development of many integrated firms of design, engineering, and construction akin to North American models. In the meantime, Harvard-trained Ortega and Solano became pioneers of prefabricated thin concrete barrel vaults with the Vacuum Concrete patent (Proa 1955b;Ortega 1989) following Gropius's and Breuer's ideals, and realised three projects in Philadelphia (The Architectural Forum 1955). They were also among those interested in manufacturing modern furniturethey attempted to produce some of Breuer's chairs -as well as prefabricated modular kitchens and bathrooms (MBDA-1 and 2). Furniture design, in fact, represents another field strongly affected by Colombo-US relations, as most of the Colombian designers and architects involved in furniture production were trained in North America. Jorge Arango, who was well acquainted with US furniture design, 31 set up the modern manufacturing firm Artecto in partnership with Ernesto Vivas Puyana (Gómez 2008). Artecto was the initiator of modern style furniture in Colombia, and quickly 'became synonymous with functional furniture' (Proa 1952d: 18). Many cases follow similar paths. The US-trained industrial engineer Guillermo Durana pioneered serial manufacturing of classic furniture for Camacho Roldán (Gómez 2008: 196), which later shifted towards 'modern and functional' design under the direction of Juan Manuel García, who was educated at the Chicago Art Institute (Proa 1952d). Álvaro Sáenz Camacho, who worked for both Camacho Roldán and Artecto, trained at Cornell (Aragón and Espinosa 1965). Later, another US-educated designer, Roberto Bermúdez Santamaría, introduced modern design in Fabrex (Gómez 2008: 195). Last but not least, Enrique Triana and Santiago Vargas Rocha, respectively from the University of Michigan and Stanford, played a fundamental role in the field during the 1950s (Fríes and Saldarriaga 2001).
The contribution of US-educated architects to the transfer of modern ideas of planning also appears crucial, as Goossens (2014) implicitly suggests in regard to the introduction to Colombia of elements such as neighbourhood units and separated networks of pedestrian circulation. The very masterplan for Bogotá was marked by a long competition between the team of Le Corbusier, Paul Lester Wiener and Sert, supported by Carlos Martínez and others,and Breuer,promoted by Solano and Ortega. 32 The very diffusion of North American architecture in Colombian magazines seems to be due to these study experiences and the subsequent personal and professional connections. One of the clearest cases is that of Marcel Breuer. In Proa, the Hungarian was probably the most followed architect active in the USA, more so than Gropius, Mies, Neutra, or Wright. His work began to be featured in 1949, when his house in Connecticut was published (Proa 1949), and continued throughout the early 1950s with four projects (Proa 1951b(Proa , 1953a(Proa , b, 1954.

The 1950s: The Fragmentation of Routes and the Rise of Latin American Schools
The 1950s marked a new shift in the international geographies of Colombian architects. By the 1940s, the number of Colombians travelling to other Latin American countries had increased dramatically for the first time.
In that decade, Chile became the major continental pole, with students divided between the Universidad de Chile and the Pontificia Universidad Católica de Chile.
Movement to Brazil also began by the end of the decade, although this trend became evident only in the 1950s. From then on, Colombians started to move to other South American countries almost as much as they did to the US. In this changed context, the most important phenomenon was the rise of Brazil as the main continental destination (Table 3 and Fig. 6).
The relationships between Colombia and Brazil, including the reception of Brazilian modernism in Colombia, have not yet been studied in depth. What history presents is a limited vision of the relationship between the two countries, focusing on shared 'formal influences' (Arango 1989: 219) and 'fashions' (Téllez 1998: 92-93). However, the presence of many Colombian students in Brazil produced a widespread diffusion of Cariocan modernism in Colombia. The amount of space Proa dedicated to Brazilian architects during the early 1950s, starting with the monographic issue 47 (Flórez and Riaño 1951), was largely due to these direct connections. In particular, Humberto Flórez Álvarez and Luis José Riaño, who graduated in Rio de Janeiro and São Paulo, respectively, continued sending graphic material to the magazine after editing this special issue. Jacobo Kuperman, who studied in Rio de Janeiro, also acted as correspondent from Brazil (Proa 1953c ; Fig. 7). Thanks to these sources, not only was the work of the renowned masters acknowledged in Proa, but also that of less well-known architects; 33 for the first time, schools like Carioca and the Paulista found an international audience (Proa 1952c).
In relation to Brazilian education, two things must be emphasised. Firstly, movements towards Brazil started at the end of the 1940s, when architects in Colombia began to look to the country after Serrano's trip there, the report of which was partially released (Serrano 1948). The diffusion of Goodwin's Brazil Builds (1943) was also significant. This mobility gained momentum in the early 1950s, after the international reception of Brazilian modernism had reached its apogee, and continued up to the 1960s despite growing critiques addressed to Cariocan architects (see Bill 1953). Secondly, with few exceptions, Colombians who went to Brazil attended the Universidade Federal do Rio de Janeiro, where figures like Sérgio Bernardes, Affonso Eduardo Reidy, and Ernâni Mendes de Vasconcelos had teaching positions. This meant there was a clear choice of schools. Rio was the city of Niemeyer, and the faculty, notwithstanding his absence, was devoted to his achievements. 34 Most Colombian students in Rio came from Bogotá and Cali. About 25 have been identified so far, but the number may have been much higher, as an article in Proa (1952a) suggests. How these experiences contributed to spreading the Brazilian-Cariocan formal repertoire is exemplified in the work of the firm Noguera y Santander, whose collaborator, Álvaro Larreamendy, had graduated in Rio de Janeiro. Projects such as the Sefair House, which includes multiple Cariocan elements, including two ramps, an organic-shaped garden, and a curved roof inscribed within a thin regular structure, as well as many others, prove this tendency (examples can be found in Suramericana Editores 1960: 78-87).
The 1950s seemed to be characterised by an increased circulation of architects within Latin American countries, not just Brazil. Surprisingly, not many went to Argentina, Uruguay, México, or even Venezuela -a growing neighbouring country linked to Colombia by a developed system of academic and professional relations. A significant number of Colombian students graduated in Ecuador, although perhaps most were Ecuadorians who later emigrated to Colombia, rather than Colombian natives. At the same time, students continued to immigrate to Chile (Fig. 8).
But it is worth emphasising that the Pontificia Universidad Católica and the Universidad de Chile in Santiago were far from being modernist strongholds in the 1940s. In the latter, a major reform took place only in 1946, following pressure from both students and professors (Mondragón 2006: 32). In Católica, leadership of the programme was the object of competition between Alberto Risopatrón and Alfredo Johnson, who supported continuity according to Beaux-Arts principles, and Sergio Larraín and Emilio Duhart, who pushed for the introduction of a Bauhaus pedagogy. In the 1940s, the faculty at Católica reached a compromise. The curriculum was divided, with the first two years given to classic teaching, and the remaining three guided by modernist architects. Only in 1949, after various resignations and a student strike, was new reform at Católica undertaken, resulting in Larraín's election as dean in 1952 (Mondragón 2006: 38). Like many American schools, the Chilean case shows that Colombian architects' study abroad did not necessary imply an encounter with a more progressive architectural environment.
In Europe, geographies changed again. France remained a major destination, but Italy emerged with both Milan and Rome as the main continental pole. This was likely due to both the interest in preservation studies 35 and to the appeal of Bruno Zevi's ideas, which were then well known all over South America (Rueda 2012: 89-90). In the 1940s and '50s, the Bouwcentrum in Rotterdam also received a significant number of Colombian students, who came to take specialised courses. 36 In the US, on the other hand, the most remarkable change was in its internal geography, with the rise of some schools located in the southern part of the country, such as Tulane University and the Georgia Institute of Technology, and the universities of Florida, Miami, and Texas.

Conclusions: Geographies for Another History
Studying the educational geographies of Colombian architects represents a first and partial attempt to build a different cultural geography of architecture in Colombia and to expand its existing history. As the historian Carlos Niño (1991: 110) points out, there is still much work to do on the broader context of the architectural profession in mid-20th-century Colombia.   The mainstream narrative cites many 'influences', 37 such as Gropius, Mies, Le Corbusier, Aalto, Wright, and Niemeyer (Saldarriaga 1983; Téllez 1979) without exploring their reception paths. In addition, Colombian historiography has emphasised the contribution of some European immigrants, particularly Violi and Rother, and  all of whom trained in the US. This also produced a very interesting shift in the international reception of Colombian architecture, which, in less than two decades, moved from being identified as influenced by North American practitioners (Violich 1944;Hitchcock 1955) to becoming entirely associated with Salmona's brickwork poetics (see Liernur 2015: 76). 38 As a new generation of Latin American historians emerged in the 1970s who used history as a tool to affirm regional values (Segre 1999: 21), Colombian architecture began to be identified with this tendency, and historiographic problems increasingly acquired a regionalist perspective (see Mondragón and Lanuza 2008). The same happened with Brazil-inspired architecture, whose buildings scarcely fit this dominant narrative. The history of Colombian architecture, therefore, has neglected a large group of architects, ignoring their international experiences and the multiple transfer routes contributing to the modernisation of the profession in that country. More importantly, extending the study of educational geographies to the whole continent may help overcome another well-established topos: the lack of dialogue among Latin American countries. In the 1920s, this idea was already widespread among intellectuals and architects, and it has permeated Latin American historiography until recently (see, for example, Niño 1991: 234), despite some early concerns. Hitchcock, for instance, recognised, though without questioning the phenomenon, the increasing circulation of architects across the continent: 'Building materials rarely travel by air, but most architects do and their ideas as well' (1955: 11-12).
As for the circulation of people and ideas within Latin America, Carranza and Lara's (2014) provocative final chapter, 'Islands No More?' warrants discussion. The Colombian case proves that in the 20th century architectural students regularly studied in Chile, and also new educational hubs evolved, such as Brazil. Moreover, figures show that by the 1960s, the number of architects heading for the US almost equalled those moving to other Latin American countries (Fig. 10), and circulation along certain Latin routes became increasingly congested. To maintain Carranza and Lara's metaphor, as in any archipelago, certain islands are better connected than others, despite geographical proximity.
This agrees with recent studies that demonstrate the fluidity of ideas circulating within Latin America in the 1920s through broad cultural networks (see Pini and Ramírez 2012;Bernal 2015). Extending this review to the architectural profession would provide complementary tools for a more ' connected' architectural history of the continent. 39 Furthermore, framing these movements geopolitically may add another layer to the existing history. Scholarship on actors and knowledge circulation has developed in the field of architecture, focussing on global empires (on the British, see Bremner 2016 and Willis 2016; on the Portuguese, with an actor-network theory approach, see Faria 2016). Similarly, the race of Colombian architects towards the US seems a clear expression of cultural dependency (see Carnoy 1974) within an informal empire. However, the rise of some Latin American architectural schools as educational destinations may be read in two ways: as a further consequence of imperial patronage -not only in the sponsorship of Latin American modernisms, but also in inter-American cooperation that fostered internal mobility -and also as an emancipatory response to this patronage. Finally, the scrutiny of architects' mobility and their networks in the Colombian case suggests how the role of magazines should be specifically read. Periodicals functioned as complementary tools in the diffusion of international models (Torrent 2011 and, which reflected current trends and became spaces for the representation of wider cultural dialogues.  S. Reyes Gamboa (1905-?) [?]      (Arango 1984: 16). 3 The definition of these geographies on the basis of national categories may appear questionable, as the existence of typically transnational networks such as the Beaux-Arts academic system proves. However, here nations are used as a first and intuitive category, either referring to particular institutions related to specific academic or professional milieus or to groups of architects, which may share nationality or place of activity, but never presuming the existence of any national architectural identity. 4 As studies in architecture may take up to five years, clearly it is tricky to fix chronological watersheds to describe trends and gradual changes (for the tables, graduation date is used as the term of reference). As for periodisation, 1936 is considered a turning point for both endogenous factors (the foundation of Colombia's first architecture faculty) and exogenous (the outbreak of the Spanish Civil War and a rapidly deteriorating European situation), while 1950 approximates the closing of the transitional phase of Colombian modernism (1930-40s) leading to its 'golden age', according to most architectural historians (with different shades, see Arango 1989;Samper 2000). A closing date of 1970 serves to elucidate 1960s trends. 5 Finally, civil engineers are displayed in the tables but not counted in the maps. In general, most listed engineers are either prominent in construction or illustrate specific networks, as in the case of the Universidad de los Andes and the University of Illinois. 6 This article moves beyond certain historiographic patterns, distancing itself from the unifying tendency initiated by Hitchcock (1955) and later endorsed by Bullrich (1969). On the difference between the two interpretations see Torrent (2015). At the same time, this text moves away from the regionalist rhetoric that arose in Latin America according to the categories of ' appropriateness' (Fernández 1987) and ' otherness' (Browne 1988) after Frampton's introduction of ' critical regionalism' in the architectural debate (1983). 7 A type of source increasingly analysed in the last two decades (see Gutiérrez 1995;Gutiérrez et al. 2001;Aguirre 2013; for Colombia see Mondragón 2008). 8 These accounts sometimes seem to contradict the image recounted in most literature about the teaching environment of certain schools. Nevertheless, they appear as useful complementary sources in the understanding of student experiences abroad. 9 At least two school works designed by Colombian students abroad were also published in Ingeniería y Arquitectura. Regrettably, securing permission to reproduce these works is extremely difficult nowadays -this and other journals having already disappeared, and the authors are unreachable. Another fundamental source on the architectural production of the 1950-60s is represented by a couple of anthologies published at the time: Lo mejor del urbanismo y de la arquitectura en Colombia and Moderna Bogotá arquitectónica. 10 To name a few: Julio Casanovas and Raúl Mannheim (Chilean), Victor Schmidt (Swiss), Herbert Raupricht, Erich Lange, and Ernst Blumenthal (German), Urbanist Karl Brunner (Austrian) and Vicente Nasi (Italian) in Bogotá; Alberto Dotheé and Augustine Goovaerts (Belgian) in Medellín; Manuel Carrerá (Cuban) in Barranquilla. 11 Up to the 1940s most Colombian architects came from an inner circle, as Carlos Rueda (2012: 51) noted: Obregón, Pizano, Sáenz, Samper, and Urdaneta were all typical surnames of traditional well-known families. 12 As Frédéric Martínez underlines, due to the high costs involved, disciplines studied abroad were commonly those providing good professional perspectives. 13 Of the 26 professors, 11 graduated in Europe and one in the USA. 14 Although his reputed internship at 35, Rue de Sèvres (referred to in many bibliographic sources) appears untrue. FLC, S1-3 Atelier Le Corbusier, Demandes d'emplois ou de stages refusées, 1937. 15 The same could also be said in comparison with other Latin American schools, with the exception of Montevideo where, by the 1930s, the faculty of architecture and urban planning represented an extremely advanced reality, as Violich (1944: 166-67) noted. But its avant-garde disappeared from international historiography due to its distance from the International Style canons (Arango 2012: 271-72