Martin Berger

Between Canon and Coincidence

Introduction

As outlined in the introduction to this volume, the study of the market for precolonial Latin American material (or “pre-Columbian art”) has grown slowly but steadily over the past three decades.1 These studies have largely focused on the centers of collecting and dealing in this material outside of Latin America: France, Germany, and the United States. Less attention has been paid to how these larger market centers influenced collecting in more peripheral locations in Europe. This article analyzes and compares the formation of the precolonial Latin American collections of the National Museum of Ethnology in Leiden and the Tropenmuseum in Amsterdam, both in the Netherlands. Through a numerical and statistical analysis of the acquisition of these objects, I aim to provide insight into how these collections were formed, who they were acquired from, when the main periods of collecting occurred, and how this maps onto patterns seen in other countries, trying to ascertain to what extent these museums were embedded in a global market for this material. Ultimately, this data-driven history of collecting should not only enhance our understanding of the historical processes affecting acquisition, but also give an indication of the extent to which these collections are based on strategic and calculated notions of value employed by museum staff in the acquisition of objects.

In this article, I take a data-driven approach, focusing on numbers and statistics, rather than detailed archival, archaeological, or (object) biographical research. Over the past years, several quantitative histories of collecting in museums have started to appear.2 These studies move their focus from the qualitative study of individual collectors or collections to large-scale institutional histories, concerned with sketching broader historic trends in the development of institutions and their collections. This article takes a similar approach, asking the basic questions “When?”, “From whom?”, “How?”, and “From where?”, in order to sketch a clearer picture of how these collections developed, who contributed to their formation, and how this focus on numbers can inform us about the development of a market for this material in Europe.

In this article, I prefer to use the term “precolonial Latin American material” over that of “pre-Columbian” art. This has two reasons. First, conceptually, “precolonial Latin American material” does not stress the aesthetic appreciation of these pieces as art, a development that only took place in the early decades of the twentieth century when a market for this material was already well-established. As such, the term captures both the trading, buying, and selling of these pieces as scholarly objects of study and as art. While the market naturally underwent significant changes under the influence of this new-found aesthetic appreciation of precolonial pieces, it should be noted that scholarly collections of artefacts (/archaeological material) continued to be traded and exchanged through these same commercial networks.3 A second reason is quite simply that, as has been noted by many indigenous activists and scholars, the use of Columbus as a primary reference point feels uncomfortable. While the term precolonial carries its own set of issues, it at least allows for the flexibility of recognizing that many indigenous communities resisted invasion and oppression for decades, if not centuries, into the Colonial Period, shifting the focus from one particular historical figure in 1492 to the introduction of a socio-political system in the centuries after.

The museums

The National Museum of Ethnology (NME) of the Netherlands (Museum Volkenkunde in Dutch) is located in the city of Leiden. Founded in 1837, it is one of the oldest museums of ethnography in the world and its collections span the globe. Its founding collection was acquired by Philipp Franz von Siebold (1796-1866) a German physician who worked at the Dutch trading post of Dejima in Nagasaki, Japan. As a result, the original holdings of the museum focused heavily on East Asia and Indonesia, a former Dutch colony. In 1883, the museum incorporated the extra-European collections of the former Royal Cabinet of Rarities, significantly diversifying its holdings beyond this earlier regional focus. Traditionally, the museum has had a strong relationship with Leiden University. Since 2014, the museum is part of the National Museum of World Cultures (NMWC), which also includes the Afrika Museum in Berg en Dal, the Wereldmuseum in Rotterdam, and the Tropenmuseum in Amsterdam.

The Tropenmuseum (TM) was founded in 1926 in Amsterdam as a continuation of the former Colonial Museum in the city of Haarlem. Its aim was to promote knowledge of the Dutch colonies, not just from an ethnographic but also from an investment and economic perspective. The museum was part of the Colonial Institute, which trained colonial servants for working overseas and included a division of physical anthropology. As such, the Colonial Museum was fully immersed in colonial politics and power structures, and more strongly under the influence of the Dutch government interests than the more academically-focused National Museum of Ethnology. After World War II, the name of the museum changed more than once, eventually being called Tropenmuseum (Tropical Museum). It continued to be a part of the Ministry of Foreign Affairs until its merger into the NMWC in 2014.

The collections

Historically, the collections of both NME and TM have been strongly focused on Indonesia, due to the colonial ties that bound the Netherlands to Indonesia. Nonetheless, Latin American material began to be actively collected at the NME from the mid-nineteenth century onwards, as the museum aspired to become a museum of global (non-European) ethnography and archaeology. Earlier seventeenth and eighteenth-century Latin American collections also came into the NME through the 1883 transfer from the Royal Cabinet of Rarities, some of which can tentatively be traced to Johan Maurits van Nassau (1604-1679), governor of the short-lived colony of Dutch Brazil.4

In this study, the Guyanas are included as part of Latin America, as these countries’ collections have historically been grouped together with Latin America in the departments of “Middle and South America” (NME) and “Latin America and the Caribbean” (TM). This meant that the curator responsible for these regions was historically always the same person and their idea about acquisitions equally affected collections from Latin America and the Guyanas/Caribbean. Naturally, the Dutch colonial occupation of Surinam created sizeable collections for these institutions. While the majority of these Surinamese objects are ethnographic, some archaeological material is also present.

Together, TM and the NME hold around 10,000 pieces of precolonial Latin American material. This number is approximate, as it is spread over a total of 9,887 inventory numbers, some of which represent one piece, whereas others stand for a large number of ceramic sherds. These 10,000 pieces are spread relatively equally over the museums, with the Tropenmuseum counting 4,367 inventory numbers and the NME 5,521. In the following, I will analyze the acquisition of these objects based on the questions “When?”, “From whom?”, “How?”, and “From Where?”.

When?

Figure 1 shows an overview of the number of pieces that the NME acquired, arranged per decade. While the clear peaks in the 1950s and 60s reflect the global market boom and are to be expected based on the popularization of “pre-Columbian art” in this period, the fact that by far the largest number of pieces per decade came in before World War II is quite surprising. However, a closer look at the collections shows that these moments of activity are tied to two specific events. The first of these is the 1903 exchange of material between the NME and the National Museum of Antiquities (NMA), also in Leiden. In this exchange the NMA transferred all its material that was not of European, Mediterranean or Ancient Near Eastern origin to the NME, which in turn transferred its archaeological material from the aforementioned regions to the NMA. As a result, a total of 4,542 pieces came into the museum in 1903, 1,441 of which came from Latin America and the Caribbean. The majority of these objects came from the collections of AJ van Koolwijk (1836-1913), a Roman Catholic priest who conducted excavations in the Dutch Antilles in the 1880s. Since this collection consists of excavation material, the number of objects is high, with many inventory numbers represening pieces of shell and ceramic sherds. Similarly, the high number of objects acquired in the 1920s is explained by the acquisition of the material amassed by the Dutch-Danish Archaeological Expedition to the Caribbean. Its Dutch part was led by Leiden professor JPB de Josselin de Jong, which resulted in significant collections from the islands of St. Eustatius and Saba.

These excavation collections consist of large numbers and are unique events in the museum’s history. As such they are not comparable to other acquisition events and practices and can skew the analysis. Figure 2 shows an amended overview of the number of acquisitions per decade, excluding these excavation collections (as well as excavation material donated by museum curators HR van Heekeren and CJ du Ry in the 1960s). It is clear that the mid-twentieth century was the moment of most active collecting. As mentioned, this is to be expected, considering the “collecting boom” that took place in the 1950s-70s, especially in Mesoamerica and the Maya area. The 1890s and 1910s are the third and fourth most active period. This is rather striking because Latin American archaeology and ethnography were non-existent as a discipline in the Netherlands at this time. Nonetheless, it is clear that a significant number of pieces were already in circulation by the late nineteenth century, considering the fact that more than 550 pieces were acquired between 1880 and 1900, from fifteen different suppliers.

Comparing the number of pieces acquired with the number of transactions that took place in the same period (fig 3), it becomes clear that there is a significant difference in the size of individual acquisitions when comparing pre- and post-WWII patterns.

Before WWII a relatively small number of transactions is responsible for a relatively large number of objects. In contrast, in the period 1950-1970, the number of transactions increases enormously but the amount of objects acquired does not do so to the same degree.

The average size of acquisitions of this material before WWII is twenty pieces, dropping to just eight after 1945. In fact, forty-two percent of the acquisitions after 1945 consist of just one object. An explanation for this decrease in size may be that, prior to WWII, the material in this corpus was seen primarily as objects of study, whereas after this period it became seen as art. The nature of a study collection is that it includes many objects which allow for comparability and differentiation. In contrast, art pieces are generally acquired for their uniqueness, and similarity and replicability are not important characteristics when deciding what to acquire for a museum collection.

Two final points are important to note when looking at the distribution across time. Firstly, there is a notable lack of acquisitions in the period 1900-1909. This absence of activity is explained by the 1903 exchange with the NMA. Because of the enormous collection that needed to be inventoried as a result of this exchange, the museum did not have the resources to focus on acquiring additional material. A second observation is the fact that the introduction of the UNESCO 1970 Convention only seems to have had a limited impact on the collecting activities of the museum. While the number of acquisitions dropped by half in the 70s, compared to the 60s, active acquisition continued into the 1980s and only began to subside in the 1990s. This begs the question whether the reduction in the number of acquisitions in the 1970s is primarily due to the introduction of UNESCO regulations, or whether this is tied to other factors, such as different curatorial priorities or a diminishing availability of this kind of material on the global or regional market, itself possibly due to the introduction of the UNESCO convention.

When comparing these numbers to the Tropenmuseum (fig 4), both differences and similarities become apparent. Naturally, because the museum was founded quite a bit later, the first acquisitions also took place at a later date. Nonetheless, in the first decades of its existence, the museum seems to have had little interest in this material. This can be explained by the focus that the museum had on the Dutch colonies, especially Indonesia. Precolonial (indigenous) material did not play an important part in the narrative of the museum, which was aimed at showcasing the life of the inhabitants of the colonies and promoting the Dutch colonial endeavor. As a result, the few archaeological pieces entering the collections before WWII consist primarily of chance finds of precolonial stone axes from Surinam. The other part of the early collections of TM is made up of artificially deformed human skulls from Peru, which were collected for the Colonial Institute’s department of Physical Anthropology.

Only after WWII, when the museum became focused more broadly on “the Tropics” did the increasing global popularity of this material manifest itself in the collections in Amsterdam. As a result, the mid-twentieth century shows a peak in collecting which clearly coincides with the most active period at NME. In essence, this particular moment of popularity of “pre-Columbian art” is the only period in which TM actively collected such material. Part of this is due to the fact that the impact of the UNESCO 1970 treaty is much more pronounced at TM than it is at the NME. This becomes especially clear when comparing types of acquisition post-1970 (figs. 5/6). At Tropenmuseum the number of transactions dropped by sixty-three percent in the 1970s. Of the remaining transactions, around eighty percent were purchases that took place in the first five years of the decade. In the 1980s, the number of acquisitions dropped by another seventy-eight percent. Of these only one was a purchase and the rest were donations by individuals for whose collections the museum could possibly be argued to act as a safe haven, taking the pieces out of commercial circulation. In contrast, while there was a significant decline in acquisitions in the 1970s at the NME as well (of 52%), the number of acquisitions stayed roughly the same in the 1980s. Of the acquisitions in the 1970s, almost ninety percent were active purchases, and purchases still made up more than sixty percent of the acquisitions in the 1980s, showing that the museum actively pursued a targeted acquisition policy of “pre-Columbian art” despite the introduction of the UNESCO 1970 Convention. Regardless of the impact of the UNESCO Convention, it should of course be noted that many Latin American countries already had legislation prohibiting the export of archaeological material without a legal permit decades prior to 1970.

From whom?

In order to better understand the background of the actors supplying material, suppliers were categorized based on location, profession and sex. Since this project did not include a detailed biographical study, these characteristics were based on the available information in the museum’s database, which can include birth and death dates, short biographies, last known addresses, and sex, but does not do so for all individuals. This data was supplemented with readily available information in standard knowledge bases such as Wikipedia, RKD Explore, and online collections databases of other museums. In total, 229 individuals supplied material to TM and NME. Of these, 124 contributed to the Leiden collections, 89 to the museum in Amsterdam, and only 16 overlap between both museums, showing that the museums generally obtained pieces from different sources. Of these sixteen overlapping suppliers, eleven are art dealers, four are academics and one a Dutch filmmaker. This predominance of art dealers among overlapping suppliers is not surprising, as these would be the most likely individuals to have access to sufficient material to supply multiple museums and they would of course be interested in selling to as many institutions as possible.

A little under seventy percent of the individuals supplying pieces came from the Netherlands. Clearly, for both museums, the main source was local. However, it should be noted that there is significant difference between TM and NME, with the former having around twenty percent foreign suppliers and the latter slightly over forty percent. In total, suppliers from nineteen countries (including NL), supplied material. While the countries from which these suppliers came overlap to a certain extent, there are some important differences between the two museums (fig. 7). For example, many German individuals contributed to the Leiden collections. In contrast, Tropenmuseum only had one interaction with a German institution. This discrepancy can be explained by looking at the dates of these transactions. All but one interaction with a German supplier at NME took place before WWII. After the Second World War, the focus of international collecting for NME shifted to Belgium and the US. Since Tropenmuseum only started collecting this material in earnest after 1945, the absence of Germany can be explained by a shift in focus after the end of the war. A similar pattern – though starting at an earlier date – has been noted by Léa Saint Raymond and Élodie Vaudry, who point to the disappearance of German actors on the Parisian auction market for pre-Columbian art after Hitler’s ascent to power in the early 1930s.5

Conversely, Belgian and American suppliers only start playing a significant role at both museums after 1945. All but one of the transactions with a US supplier took place after 1945, the same goes for Belgium. Importantly, nearly all Belgian, Swiss and North American suppliers are professional art/antiquities dealers. This stands in marked contrast to suppliers from Latin America, of which most are Dutch expats who lived and worked in the region. This is especially true for suppliers to the Tropenmuseum and should not be surprising considering the ties that the museum had to the Ministry of Foreign Affairs. Only a limited number of dealers from Latin America (including Guillermo Schmidt Pizarro, see Orsini, Antonini, and Villa in this volume) seem to have had direct access to Leiden and Amsterdam. All of these came from South America and sold Ancient Andean material. This seems to suggest that dealers from Mexico and Central America may not have sold directly to European customers or institutions. While this cannot be concluded definitively based on this small a sample, the relative proximity of Mexico to the United States may have led to a situation in which Mexican traders sold primarily to dealers in the US, who in turn sold pieces to museums in Europe.

There is a marked difference between NME and TM when looking at the presence of professional dealers. In Leiden, twenty-nine percent of the suppliers are art dealers, whereas in Amsterdam this is only seventeen percent. In total, at least forty-four art and antiquities dealers from ten different countries contributed to the collections of both museums. This is a minimum number, as some of the suppliers may not have been identified as art dealers because of a lack of biographical information. Of these forty-four, twenty-nine interacted with NME, eleven overlapped and only four dealt solely with TM. The Leiden network of dealers was clearly more extensive. Interestingly, of the overlapping dealers just two contacted TM before contacting NME, a possible indication that NME may have been used as a “jumping-off” point by foreign dealers when establishing contact with museums in the Netherlands.

The international orientation of Museum Volkenkunde is apparent already in the earliest stages of collecting Latin American material. Between 1880 and 1900, twenty-seven transactions took place. Of these, only seven were with Dutch suppliers, the remaining twenty were with international suppliers from ten different countries. It should be noted that many of these suppliers were connected to museums abroad as curators or directors. An especially important connection existed between the NME and the German city of Hamburg, a city that was later also fundamental in the Parisian interwar auction market.6 Not only did the NME deal directly with the Museum Goddefroy in that city, it also acquired pieces from dealers and collectors based there. In fact, all German transactions before 1900 were with individuals in Hamburg. This can be explained by the presence of Johann Schmeltz (1839-1909), who was hired as curator in Leiden in 1882 and became the museum’s director in 1896. Before coming to Leiden, Schmeltz had been a long-time curator at the Museum Goddefroy, securing close ties not only to the museum (which closed and whose collections were sold in 1885) but also with collectors and dealers in Hamburg, highlighting the importance of personal ties and the presence of certain individuals in determining where, what, and from whom material was acquired.

When looking at the number of interactions that suppliers have with museums, it becomes clear that the overwhelming majority only interacted with the museums once. Nearly eighty percent of all suppliers has only one interaction with a museum. Only relatively few individuals had multiple transactions with one or both institutions. Most of these “recurrent individuals” are well-known antiquities dealers, such as Belgian dealers Edgar Beer, Emile Deletaille, and Henri Schouten, the Amsterdam-based Galeries Lemaire and Menist, and the US art dealer Robert Stolper, who also had a gallery in Amsterdam in the mid-twentieth century. Two individuals in the “most transactions top ten” come from Latin America. These are Leopoldo Gómez, an entomologist (and apparent antiquities dealer) from Ecuador and a certain B. Jones, supposedly an antiquities dealer from Argentina about whom no further information could be located in the museum database. A surprising name to appear in the top ten is that of Michael Francis Podulke, an American artist who co-owned a gallery for contemporary art in Amsterdam in the 1960s. While his artistic work is relatively well-documented, his involvement with pre-Columbian art is not mentioned in any publicly available biographies.

The individual with the largest number of interactions with the museum is Hans Feriz. His twenty-seven interactions with the Tropenmuseum make him the largest single supplier of material to either of the museums. In fact, as his collection makes up more than three thousand pieces, making Feriz personally responsible for almost two-thirds of the Tropenmuseum’s collection of Latin American archaeology. After studying medicine in his native Austria, Feriz moved to Amsterdam where he worked and taught at the local university. In addition to his work in Amsterdam, Feriz worked as a ship’s doctor, visiting large parts of the Americas and becoming interested in the indigenous cultures of the continent. Feriz acquired objects from dealers but also excavated in several Latin American countries. In the 1950s, he became an honorary curator of American Archaeology at the Tropenmuseum, maintaining this affiliation until his passing in 1970. His collections are diverse and come from many different countries in Latin America. To an extent, one could say that TM would not have a significant collection of this kind of material, were it not for Hans Feriz.

Finally, a look at the sex of the suppliers for both museum shows that the overwhelming majority was male. It should be noted that sex could not be established for twenty-seven percent of the suppliers due to a lack of biographical information. However, of those for whom sex was known, seventy-nine percent were male. As might be expected, most female suppliers were active after WWII. Still, several women supplied material before 1940. There is no significant difference between both museums, though the percentage of women contributing to TM is slightly higher (eighteen percent), compared to NME (twelve percent).

How?

Figs. 5 and 6 show the types of transactions that took place, divided into purchases, donations (including gifts and bequests), and exchanges. As these figures clearly show, there was a significantly different pattern of collecting between the two museums, in which TM received more donations and NME pursued more active purchases. In total, sixty-seven percent of the transactions at NME were purchases, while at TM this was only forty-two percent. However, there is a significant difference at the NME when comparing pre- and post-WWII acquisitions. Whereas, before the war, only twenty-eight percent of the transactions were purchases, fifty-four percent were donations and fifteen percent were exchanges, after the war most transactions (seventy-seven percent) were purchases, with the rest being donations, and exchanges disappearing altogether at the NME.

These patterns clearly show that the NME pursued an active acquisition policy. The Tropenmuseum, in contrast, was dependent on donations for most of the material it acquired. Naturally, the dominant presence of Hans Feriz – all of whose collections were donated – plays an important role in this. However, even when taking Feriz’s collections out of the equation, fifty-two percent of the transactions concerned donations. In fact, when looking at the number of pieces, rather number of interactions, the difference between NME and TM becomes even more pronounced. Of the 4,367 inventory numbers at TM only fifteen percent were acquired through purchases. In contrast, fifty-nine percent of the pieces at NME were bought from private collectors or art dealers. These figures clearly show that the NME was much more active on the global art market than TM.

Reflecting on Tacye Philipson’s work on the collections of the National Museum of Scotland, Isobel MacDonald7 has discussed the difference between “active” and “reactive” collecting and how this could map onto the difference between purchases and donations. MacDonald argues that, because museums may have actively cultivated relationships with potential donors, donations do not necessarily imply that the agency in deciding what enters the collections lies with the giver rather than the museum. MacDonald also stresses that the museum ultimately decided which donations it accepted or declined. While these caveats are valuable observations, the high number of donations at TM does seem to reflect a particularly “reactive” form of collecting. This becomes especially clear when looking at the percentage of donations in combination with the number of interactions with suppliers and the location of these suppliers. As mentioned, around seventy percent of suppliers were based in the Netherlands and nearly eighty percent of these had only one interaction with the museum. The combination of these two facts with the large amounts of (small) donations suggests that TM was primarily dependent on the benevolence of local collectors. With the exception of Feriz, rather than courting private collectors with large collections to donate their material to the museum, the TM received material from Dutch individuals who collected (or inherited) a small number of pieces and were no longer interested in keeping them in their homes. As such, Philipson’s division between active and reactive collecting does seem to largely hold up in this particular case.

From where?

Fig. 8 shows the national origin of the material in the two museums. It is not surprising to see Mexico and Peru as the two most collected countries, considering the long history of archaeological studies in these countries and their overall predominance in museums across the globe. When comparing the two museums, there is a pronounced difference between the presence of Ecuadorian, Costa Rican and Panamanian collections in Leiden and Amsterdam. This difference can be explained by the fact that these countries were of special interest to Hans Feriz and places where he conducted his own fieldwork.

The prominent place that Costa Rica, Panama, and especially Ecuador have at TM seems to also be an indication that this museum was less market-dependent (or market-focused?), as it primarily followed the acquisition interests of its most important donor/honorary curator. While pieces from these countries were definitely available on the global market, they were largely outnumbered by material from Mexico and Peru, as well as (Classic) Maya art from Central America. At the more market-oriented NME, seventy-six percentof the material collected after 1945 comes from Mexico or Peru. In contrast, only forty-two percent came from these two countries at TM during the same period. This percentage is significantly higher (sixty-five percent) when not including Feriz’ collections, another indication that the overall TM collection diverges from the “market standards”. In comparison, only fifty-five percent of the pre-WWII collections at the NME comes from Mexico and Peru. This increasing focus after the Second World War is probably linked to the advent of an international market for pre-Columbian art which resulted in market-making (or even branding) for certain kinds of pre-Columbian art by art dealers. The extent to which these “canons of pre-Columbian art” were formed by the interplay between commercialization and academic research is a topic that cannot be fully addressed here. However, as Rosemary Joyce has suggested “the capacity we have to identify types of objects [...] as fashionable in the market is partly a reflection of the scholarly attention the same categories of objects have attracted.”8 Still, as Turner and Phillips show elsewhere in this volume, the reverse is equally true.

Discussion and Conclusions

Even though the museums discussed in this paper are only around forty kilometers apart, their histories of collecting precolonial Latin American material are radically different. Important suppliers do not overlap and the institutions seem to move in networks of their own. To an extent, these different networks can be explained by the fact that NME has historically been an academic museum linked to Leiden University, whereas TM fell under the auspices of the Ministry of Foreign Affairs. More important however, is the fact that, after WWII, NME seems to have had a much more targeted acquisition in buying this kind of material. Not only is the amount of purchases much higher at NME (fifty-nine percent of objects acquired versus fifteen percent at TM). The fact that the NME had a dedicated curator for the Americas from 1960 onwards is undoubtedly part of the explanation of why NME’s policy was more specific. However, the total number of pieces in both museums does not differ enormously, suggesting that the presence of a specialist curator influenced the “quality” of the collection more than the quantity.

When comparing these museums to broader global patterns and flows, it is clear that acquisitions at NME and TM largely follow patterns that are seen in other countries as well. The most important decades of collecting were the 1950s and 60s, as a result of the boom in both interest in this material and availability of it. However, significant acquisitions in the late nineteenth century show that a market for this material was well-established long before the commercial explosion in the mid-twentieth century. While this market has been studied for France,9 the NME only acquired two objects from France before WWII. Saint-Raymond and Vaudry10 have suggested that the trade in this material was triangular between Latin America, Northern Europe, and France. It is clear that the NME was focused on the Northern European – German and, to a lesser extent, British – part of this triangle. This changed after the Second World War, when pieces seem to have been acquired from almost anywhere but Germany. While it has been argued that the dominant market for pre-Columbian antiquities was North America,11 this is only true for the post-WWII period. After all, even Robert Woods Bliss bought his first pre-Columbian piece in Paris.12 Still, the market for this kind of material remained highly international after 1945, as shown by the fact that even the two relatively peripheral museums discussed here acquired pieces from nineteen different countries. While most foreign suppliers (post-WWII) did indeed come from the US, suppliers from France, Belgium, the UK, and Ecuador were almost as important.

As mentioned above, the number of foreign suppliers for NME is double that of TM (forty percent vs. twenty percent). This shows that the NME was much more directly embedded in the global market for precolonial Latin American material. This is not only reflected in the fact that there are more international suppliers in Leiden than in Amsterdam, but also by the number of professional dealers that contributed to the collections, which is much higher in Leiden. Whereas Hans Feriz’s presence might be part of the explanation here (again), several other factors may also have contributed. First of all, the NME was internationally more renowned, due in part to the presence of the Leiden Plate, a Classic Maya jade pendant that had already been an object of study for specialists since the late nineteenth century.13 In addition, from 1960 onwards the NME had a specialist curator for Latin America with a background in archaeology. Since networks of curators, academics, and dealers were much more closely integrated in the mid-twentieth century, these Leiden specialists had more direct access to the international networks in which pieces moved. Finally, the NME’s acquisition policy specifically targeted certain kinds of archaeological material, as opposed to TM which focused more on ethnographic and folk collections. In order to achieve this targeted policy, the NME relied strongly on (international) professional dealers to find those objects that were need to ‘fill the gaps’ in the collection.14

Still, despite the significant influence of foreign suppliers on the collections, the majority of suppliers to both TM and NME were local Dutch citizens. In addition, the overwhelming majority of these individuals only had one interaction with a museum. This pattern of many smaller donations creating a larger collection is not unique to the museums under study here. On the contrary, it seems to be the standard, rather than the exception. For example, at the British Museum, more than seventy percent of acquisitions have consisted of donations of less than ten pieces.15 Similarly, in the English collections of the Pitt-Rivers Museum (PRM), donations make up around seventy-seven percent and more than half of the suppliers only contributed one or two objects.16 More broadly at the PRM, nearly ninety percent of suppliers donated less than ten objects.17 In a slightly different context, one third of the sellers at Parisian auctions in the interwar period sold only one or two artefacts.18 Wingfield has argued that the relationship between these small-scale suppliers and the museums they gave to was relatively insignificant to these individuals and that “their agency in relation to shaping and forming the collection of material from England held by the PRM was [...] fairly minor.”19

However, if the bulk of most museum collections seems to be formed by “small-scale suppliers”, is it fair to say that their agency in shaping these collections is not of major significance? Indeed, if these small-scale suppliers are responsible for more than seventy percent of our collections, doesn’t that show that these smaller suppliers were important in driving what the collections look like today? As Gosden, Larson and Petch have argued, the large percentage of small, individual donations “suggests that the usual ‘great man’ approach to the history of archaeology and anthropology needs to be balanced somewhat by a consideration of the thousands of ordinary people whose small, individual contributions were cumulatively vital.”20 It is here that the value of a numerical/statistical approach to collections (histories) becomes most apparent. Traditional histories of collections have had to focus on those suppliers for which there was ample biographical and contextual information. These suppliers tended to be those individuals who provided museums with large quantities of objects. Statistical approaches to collections can put these individuals in their institutional context and treat the “big hitters” the same way as the smaller-scale suppliers.

One possible way to theorize this different focus is to think about (archaeological) collections as oscillating between canon and coincidence. Michael Coe has argued that there is a “profound randomness in the kinds of [precolonial Latin American] objects” that end up in museums. As Coe says, “exploitation for commercial ends of the remains of antiquity [has] introduced a strong element of chance into the process by which objects are transmitted to museums: one year it might be the Olmec jades of Guerrero, the next Teotihuacan murals, and the next gold from a northern Peruvian valley”.21 As such, museum collections of “pre-Columbian art” can be considered “coincidental”: whatever is available on the market comes to the museum. In a sense, we could see the impact of small-scale suppliers in a similar way: whatever is offered to the museum by local individuals ends up in the institutional collection. However, these “coincidental” contributions are still “canonically influenced”. In Coe’s case it is clear that Olmec jade, Teotihuacan murals, and Peruvian gold are among the perennial favorites of collectors and museums and that they can be considered part of “the pre-Columbian art canon”. It is no coincidence that Coe does not take Marajoara ceramics, Jama-Coaque figurines, or some other more regionally-known cultures/styles as his examples. Similarly, many of the small donations made to the museums under study here consist of “well-known” cultures or styles, such as Moche, Chimu, Chancay, or Classic Maya. Clearly, whatever is acquired is not only influenced by what is available, but also by what is deemed to be desirable. In turn, what is considered desirable is determined by an interplay of commercial, academic, and aesthetic factors that, together, create a canon of what kinds of pre-Columbian art “serious museums” should minimally be considered to own and display. While this canon of “what pre-Columbian art is” has not been clearly defined (and it is beyond the purposes of this article to attempt do so), it is important to bear in mind that valuations of objects vary over time. As Achim has noted, the value of pre-Columbian antiquities was created “over time through the choices of individuals navigating through a web of calculations, negotiations, and transactions.”22 As a result, the most expensive pieces to be sold at auction in Paris during the interbellum – a set of eight items from Venezuela23 – would arguably not have fetched a high price at auction after WWII as their value did not survive “the canonical shift”.24 In this regard, the predominance of Ecuadorian and Costa Rican material at TM is a good example of how idiosyncrasies/coincidences can measure up to canonization. Similarly, TM’s dependence on a network of expats and diplomats working in different parts of Latin America may have contributed to a more “coincidental” make-up of the collections, as these were often non-specialists who collected whatever was available to them in the places that they lived.

When thinking about the NME and TM collections as canonical vs. coincidental, one could argue that the NME collection has more canonical aspects than TM. The most obvious reason for this difference is the presence of Hans Feriz, an essential donor who contributed the majority of the collections of the Tropenmuseum. Because TM mainly followed Feriz’s interests in collecting, the collection is primarily a reflection of his own (regional) interests. This has resulted in a collection in which certain countries/cultures are abundantly represented that are generally not as well-known/present. While all collections have their idiosyncrasies, TM’s collection can be productively viewed as primarily a private collection-turned-public, rather than an institutional collection. In contrast, Leiden’s stronger market-oriented acquisition policies led to a more “canonical” collection, shaped by contemporary (mid-century) academic research, as well as marketing by professional art dealers. In the end, what this analysis shows is that the National Museum of Ethnology was much more deeply embedded in the global art market for “pre-Columbian art” than the Tropenmuseum. However, at the same time, what this analysis also shows is that there were many channels besides the market through which museums in the Netherlands acquired this material. Local collectors, as well as Dutch expats and diplomats, played a fundamental role in building the collections of the NME and TM.

One final point that bears mentioning is the influence the display of pieces from art dealers had on creating canons of pre-Columbian art, both among scholars and the general public. When looking at the pieces presented in permanent exhibitions at TM and NME, it is clear that art dealers had a significant influence on what was presented as pre-Columbian art. Whereas pieces from dealers make up only around thirteen percent of the total pieces in the corpus, this material was more likely to be displayed in the museums’ permanent galleries. At TM, the difference is relatively minor, with sixteen percent of the pieces in the permanent galleries (1979 – 2016) coming from art dealers. At the NME, the difference is much more pronounced. Unfortunately, the museum only registered which pieces were on display in the permanent galleries after the year 2000. In the period 2000-2010, forty percent of the archaeological material on display came from dealers. In the present permanent gallery (opened in 2011), that percentage has gone up to forty-four percent. When looking specifically at Mesoamerica, the percentage is even more extreme, as eighty-six percent comes from dealers. Likewise, at TM fifty percent of the Mesoamerican material exhibited came from dealers, as opposed to eight percent for Andean material and zero for material from Central America, an indication that the collecting of Andean material may have been less strongly-controlled by market forces than pieces from Mesoamerica. What is clear, at least for Mesoamerican material, is that, because their material was so often exhibited, commercial art market actors were not only supplying museums with precolonial Latin American material. They were also (indirectly?) “training the eye” of the visitor in understanding what pre-Columbian art should look like.

Martin Berger is an assistant professor at the Faculty of Archaeology of the Universiteit Leiden and a member of the interdisciplinary research group Museums, Collections, and Society.

---