Felipe Álvarez de Toledo López-Herrera

Beyond Murillo: New Data-Driven Research on the Market for Paintings in Early Modern Seville

Writing on the Sevillian school in 1806, Juan Agustín Ceán Bermudez noted that “infinite paintings were once sent to America... [M]any painters of worth, when they did not have work, went to the Indies carriers, who always kept them busy.” This early art historian sensed the importance of the export industry and pointed to the intermediaries who once linked supply and demand on two sides of the Atlantic. He described an environment in which expediency and cost drove production: “[canvases] were made with such speed that subjects were often painted while the price was being negotiated...” Finally, he understood that a thriving mercantile culture was not contrary, but rather conducive to greatness, interpreting Murillo’s talent as the result of his involvement with a broader milieu operating in the Feria neighborhood. 1

These intuitions point to the concerns that must drive the analysis of historical art markets. A tradition of such research has yielded studies for various cities in Europe, providing the necessary conceptual lexicon and methodologies to also analyze Seville. Art market scholars have endowed agents and intermediaries with agency within artistic production. They have employed the notion of process innovations to describe the development of techniques that sped the production of paintings. They have used the term critical mass to describe an environment propitious to the development of extraordinary talents because of the presence of a large number of actors. On a methodological level, art market researchers have pioneered the aggregation of data for garnering conclusions unavailable in a single text, leading the way in producing an empirically-testable art history based on a statistical approach to databases of inventories, shipping registers and notarial information, among others. Finally, their case studies provide appropriate comparisons for a commercial city of the early modern period like Seville. 2

Though Cean Bermúdez saw the mass-market as integral to Sevillian production already in the early nineteenth century, little is known about its size, chronology and components. In this paper, we employ the conceptual framework of art market studies to reexamine its structures with new evidence. First, we review the state of affairs, presenting an alternative approach based on a database of archival documents pertaining to Sevillian painters. Using this information, we estimate the number of painters in the city from 1475 to 1700, concluding that rather than being insular and regressive as has often been claimed, Seville held an evolving mass-market comparable to those of northern export cities of the period. We then divide the market into primary, secondary and tertiary sectors, and describe the evidence of the presence of other actors and institutions vital to the art trade. Regarding the primary sector, we contend that guild regulations should be reframed not as opposed to the market or “flexible,” but in many ways as conducive to its development. A developed secondary sector existed with specialized dealers, active merchants and artists that transitioned into these roles. We conclude by presenting multiple avenues for research that might be pursued by scholars in the future.

A Database of Sevillian Painters

The Andalusian capital has often been described as a provincial artistic center: though important to Spanish art history, it was stunted by the weight of the church and guild. Pérez Sánchez described the city as simultaneously a “melting pot” and an isolated hinterland,3 a contradiction that results from a preoccupation with the rise of painting to the status of liberal art. In this area, Sevillian practice clashed with the modern ideal, but the paradigm of the court painter is hardly applicable in a city of commerce. More than on patronage, painters could rely on local, regional and transatlantic markets. From 1503 to 1680, Seville was the port of the Indies, channeling all legal trade between Europe and the Spanish colonies [Figure 1]. It became a veritable early modern metropolis, reaching a population peak of around 140,000 inhabitants.4 With a growing merchant class and access to foreign markets, the city fulfilled all the necessary prerequisites for supporting a sophisticated art trade.

Fig. 1: Unknown Artist,View of Seville, c. 1660, Oil on Canvas, 163x274 cm.

Sevilla, Fundación Focus (Hospital de los Venerables).

Wikipedia. This image is in the public domain.

Evidence for this proposition accumulates rapidly. In the 1980s, Duncan Kinkead discovered workshops specialized in the production of paintings for the Americas and suggested that it was a lucrative business for many masters.5 Following this lead, Pedro González García and Sandra van Ginhoven have used the shipping registers of the Archivo General de Indias to measure total exports of paintings in the sixteenth and late seventeenth centuries, respectively: 6,004 in the last decades of the sixteenth century, and 27,632 from 1630 to 1680.6 Inventory studies have also revealed that holdings of paintings in local homes were larger than previously assumed, at a level close to that of Holland.7 As Miguel Falomir accurately suggested in 2006, Sevillian painters faced a great, increasing local and international demand for their work, indicating the emergence of an art market in the city.8 But scholars have misconstrued Seville’s productive capacity, and we have little to no knowledge of commercial structures beyond the guild.

An evidentiary problem underlies this dearth of analysis. Extant paintings are problematic proof because of the fragmentary state of the corpus available to researchers: ecclesiastical and court commissions were more likely to withstand the trials of history than work made for the open market. Kinkead calculated that the existent body of work represents only 9% of known Sevillian painters from 1650-1699.9 Thus, survival bias – the logical error that results from focusing only on those objects that have survived some selection process – has led to many of the errors that plague the historiography of Sevillian art. Moreover, any attempt to employ documentation to study the art market quickly runs into problems. Because the archive of the guild of Saint Luke was lost in the nineteenth century, lists of masters, apprentices and guild-registered dealers are not available for study.

Kinkead surmounted this obstacle in 2009 by employing his own compendium of documents on Sevillian painters as the basis of a first analysis of Seville’s productive capacity between 1650-1699.10 Continuing in this line of thought, we have created a relational database that amasses the information contained in the city’s notarial, parish, cathedral, and municipal archives. In addition to Kinkead’s corpus, we include seventeen other volumes of documents on artists, published from 1898-2006, representing a long tradition of archival research in Seville.11 These books contain transcriptions and summaries of censuses, inventories, master’s exams and other contracts that can be analyzed in the aggregate. Transforming this information into data required scanning the volumes using Optical Character Recognition (OCR) software. Once digitized, the books were divided into documents with OpenRefine, and the texts mined for demographic, geospatial and output data and the relationships and events to which they testify. Using digital methods, long-published reference books were thus made into a database to be searched and queried.

The first text processed, Duncan T. Kinkead’s Pintores y Doradores en Sevilla, 1650-1699: Documentos, yielded 3,493 documents containing 4,276 references to a population of 345 trained painters and 310 apprentices. The series Documentos para la historia del arte en Andalucía added an additional 2,368 documents. Other information employed in this article has been inputted manually for the purposes of this text. Extending the chronology to 1450-1699, we find 1,295 painters, the vast majority of whom are unknown today. In the next sections, we will demonstrate how this data can revise previous preconceptions about art production in Seville.

Re-scaling Seville

Citing the signatures on a document from 1599, several prominent scholars have given the figure of only twenty-seven registered masters in Seville.12 Consequently, the city has been described as a place with a restrictive guild, where a small group of painters was limited to the patronage of the church. Jonathan Brown suggested that “the energy and vitality of commercial life failed to affect the artistic sector… until after 1600” and that “the church in all its aspects dominated the artistic scene.”13 And this characterization has infiltrated wider art markets research. In 2008, Maarten Prak stated that “[there were] fewer than 30 guild members in a city of 100,000, extremely small by Dutch standards; Delft, with only a quarter of Seville’s population, provided work for 36 painters in 1650.”14 Kinkead proved this number untrue for the latter half of the seventeenth century, finding 51 painters at the city’s nadir in 1650, and 140 at its peak in 1675.15 We have revised his estimates to factor in fragmentary data and extended his analysis to include the sixteenth and early seventeenth centuries, demonstrating that the digit of twenty-seven masters grossly underestimates Seville’s productive capacity throughout the period.

Fig. 2: Number of painters in Seville, 1475-1700.

Elaborated by the author based on published archival sources.

Figure 2 represents two estimates for the number of trained painters in Seville from 1475 to 1700. The first series (dark curve) was culled by treating the years between the first and last documents signed by a painter as the first and final years of their presence. This curve represents a lower bound for the number of painters because there tend to be few documents associated to any particular individual. For instance, of 345 painters found between 1650 and 1699, 107 appear in only one document, and hence for only one year. We have compensated for this deficiency, which increases with time, by creating an upper estimate (light curve). This helps ameliorate the bias resulting from varying data quality over time, thus correcting chronological trends. For its calculation we made an assumption along the line of Montias and other scholars by taking painters to have been active for twenty years.16 When a painter was present for a shorter period of time, we took the midpoint between the two extremes of evidence and added and subtracted 10 years to create said range. The only control we undertook was to shorten the estimates for those artists who were last evidenced prior to 1649 but whose estimated final year went beyond that date: because a plague then eliminated one-third to one-half of the population, it was more prudent to assume that those painters had passed away during the crisis; these estimates were truncated at the year 1649.

The numbers allow us to conclude that the painters’ workforce was much larger than previously stated. For the oft-cited year of 1599, there is evidence for at least 45, and up to 109, trained painters in Seville. The periods before and after reveal much larger populations of painters. To give concrete examples, there were at least 72 (up to 163) practitioners in 1530, and 142 (213) in 1677. These numbers exclude apprentices, which greatly increased the workforce – i.e., there were 81 apprentices in 1677, bringing the active labor force up to 223 (294) workers. In light of such evidence, the aforementioned suggestion that Seville paled in comparison to Delft is unjustified. What is more, these numbers place Seville on the same scale as the cities that have long been considered strong painting export centers. In 1584-5, Antwerp had 108 painters, and Amsterdam had an estimated 175 painters in 1650.17 Precise values are not a key concern; the takeaway is that the difference between these cities is not one of category but one of degree. Seville was undoubtedly a place with a large productive capacity. Arguments around its insularity lose force, and the more fruitful task is the analysis of the regulations, structures and commercial culture that shaped its market.

Additional notes must also be made on time trends (key dates have been marked on Figure 2). The sixteenth century has been understood as a period of stagnation for Sevillian art. Instead, the data gives evidence of dynamism, which much more comfortably fits what we know of Seville’s history and the evolution of demand for paintings in the sixteenth century. The guild of Saint Luke was established in 1480 and the Casa de Contratación, the organism that controlled American trade, was created in 1503. It follows that the number of painters would have grown with the city’s economy, as they profited from the commissions from new religious foundations in the city and in the colonies, as well as from enriched elites. This growth was to be followed by the development of a broader demand for canvases in the latter half of the century, as tastes for painting developed in Spain.18 The years between 1522 and 1573 show a dip in our curve which coincides with a disruption in Seville’s trade monopoly, as emperor Charles V opened Indies trade up to eight other Spanish cities.19 After Seville recovered monopolistic control, trade increased in volume until 1620, as did the number of painters. The movements of the corrected upper curve follow the progression of events closely.

The years following the plague of 1649 have been interpreted by Fernando Quiles as a moment of regeneration, in which younger masters replaced a waning older generation.20 The data support this view, even suggesting that it may have been a boom time for the Sevillian art market as the population of painters reached unprecedented heights. This unlikely event flies against local events, including the gradual loss of control of trade to Cádiz after 1680, though Kinkead already pointed out the contradiction that the dependence on exports of paintings seemingly increased between 1650 and 1699 despite a reduction in trade overall.21 Future research must determine whether growth was spurred by changes in painting consumption in Spain or the Americas or whether differences in data quality between the period of Kinkead’s data and 1500-1650 belie the apparent trend.

Primary, Secondary and Tertiary Market Sectors

The above considerations imply that a reexamination of Seville as a large marketplace for art should yield new insights, but we must consider all participants and structures to truly reconceptualize the market. Scholars have usefully divided art markets into three sectors that evolved in complexity over time as markets matured. The primary sector is the direct sale of paintings from producers to clients. The secondary sector represents indirect sales, or the mediation of dealers. Finally, the tertiary sector is the resale market, a development of the late seventeenth century.22 In this section, we provide some insights into the development of each segment and suggest useful avenues for future research.

Regarding the primary sector, we have seen that the evidence refutes there being only a small group of producers working primarily for the church. A reconsideration of the regulatory environment helps us understand how large numbers arose and were maintained. Here must move away from the traditional assumption that guilds stifled growth, often adopted by art historians because the rise of painting to liberal art, a transition that necessitates freedom from the guild, is a topos in baroque art history. Thus, scholars have taken the institution’s existence in Seville to suggest a “medieval spirit.”23 But painters’ guilds were a reality in early modern Europe; they were even reinforced in the northern cities usually credited with the development of the mass-market for paintings.24 A more generous interpretation, espoused by Falomir for Spain, contrasts the severity of guild ordinances with the flexibility of their historical application.25 However, future research should also consider the guild’s potential for positive interventions in the market, as has been established by forty years of debate in the field of economics, where scholars have established that guilds were institutions with both potentials and drawbacks.26

For one, we must clarify that, though the Sevillian guild did control who became a master, it did not limit the number of painters in the city. The rules were actually openly disinterested in limiting supply, stating that “being a good examined official, [anyone] should be able to take any work they wish, even if they are foreigners, because it is in the benefit of the republic that we have many officials, as long as they are good.”27 And the documents evidence that a significant minority of painters were not local. From 1650-1699, we find that 77 out of 345 painters were Sevillian; 8 were Northern (Flemish and German); and 47 were from other areas of Castile [Figure 3]. Though Zurbarán’s presence as an unexamined master was contested by the guild and his case has often been used as proof of its restrictiveness, it should be reinterpreted as an exception. Fuente de Cantos, his hometown, was at the edge of a broad Castilian catchment area from which Seville received many workers [Figure 4].

Fig. 3: Painter Origins, 1650-1699

Elaborated by the author based on documents from Duncan Kinkead, Pintores y Doradores en Sevilla, 1650-1699: Documentos (Bloomington: Authorhouse, 2007).

Instead, the ordinances of the Sevillian painters’ guild sought to guarantee a minimum product quality in a market plagued by uncertainty. Clauses twice forbade the sale of old paintings as new and established the use of a stamp by guild overseers for marking acceptable canvases. Restrictions of rights to examined masters – the right to open a shop, to train apprentices, to hire journeymen and accept commissions – did not limit the labor force, but rather assigned the transfer of skills and supervision of production to qualified members. And the guild could also directly stimulate the market. The actions of Dutch and Flemish guilds are well-studied in this regard: they provided venues for the sale of art by setting up exhibitions, lotteries and even permanent sales locations, such as the panden of Antwerp.28 No research on this topic has been conducted for Seville, no doubt due to the loss of the guild archive. The fact that the painting academy, active from 1660 to 1674, was set up in the Lonja (the bourse) is evocative of a further connection with this venue, but none can yet be established. And though we do not know whether the guild rented spaces from religious institutions, examples of this are known for other cities in Spain.29 The guild’s attitude towards export, however, is clear; in 1678 they sent a representative to Madrid to defend them against the taxation of the devotional paintings that “the masters of this city make in their houses and workshops and which they send for sale to the ports of sea and of the Indies.”30 But this defense was made for painters exporting of their own accord; we must also gauge the guild’s attitudes towards the action of intermediaries.

The dealers and merchants that constituted the secondary sector have not received the same attention as painters. Falomir suggests that dealers specialized in trade emerged as a third phase after institutional commissions gave way to artists engaged in export via intermediaries,31 but this remains an untested proposition within an indeterminate timeline. In our database, 56 documents attest to 40 bulk sales of paintings, totaling more than 3,412 canvases in the latter half of the seventeenth century. Twenty-eight were executed by artists’ agents abroad, at the painter’s own risk; twelve were sold directly to merchants for posterior resale. Some specialized art dealers can already be identified. Eddy Stols found three shops selling Flemish paintings in the mid 1600s, one of them near the cathedral, led by the Fleming Salomon Paradis.32 Other dealers appear in the documents: Jerónimo Delgado was a “seller of canvases”, and Juan Matheso held a canvas shop in the Chopine-makers’ Arch (arquillo de los chapineros).33 We also find evidence of a progression towards specialization. Some painters became merchants: Alonso Pérez, for instance, began subcontracting paintings to other artists, then shipping them and other items such as pigments, linseed oil and cinnamon to the Indies.34 Shippers also took the initiative and purchased paintings for sale at their own risk.35 We also gain some insight on the networks through which trade functioned. One painter, Juan Simón Gutierrez, had Diego González, a resident of Mexico City, sell his paintings. Once the bulk of traffic shifted to Cádiz, this same painter relied on a second intermediary in that city, who was in contact with others in the Americas.36

Fig. 4: The Sevillian catchment area: local origins of painters, 1650-1699.

Elaborated by the author based on documents from Duncan Kinkead, Pintores y Doradores en Sevilla, 1650-1699: Documentos (Bloomington: Authorhouse, 2007).

The tertiary sector was still not developed. Regatones sold used goods, including paintings. The almoneda, or estate sale, often acted as a means through which consumers could acquire paintings, but as far as we know there existed no specialized venue for the sale of used canvases.37 This state of affairs, however, is in line with that of other areas of Europe at the time.

Conclusions

The archive holds only partial evidence of a market that must be reconstructed from fragments. Artists sold canvases at their own risk by exporting pre-fabricated images or selling them in their shops, at dealers’ stores, on the street or at fair. Only rarely were these activities documented. But archival documentation provides complementary evidence to that of extant paintings, and its agglomeration and analysis has proven a rich methodology for historical art market research.

In this paper, we gave clear evidence of Seville being a major center: our estimates of the number of painters refute the claim that Seville held a small population of artists working primarily for the church, revealing instead numbers on par with those of northern cities. These findings justify the use of an art markets approach to study not only painters, but also the intermediaries and structures that could sustain a mass market. Guild ordinances and data on artist origins suggest that the institution was permeable. Regarding the secondary sector, we gave evidence of the existence of specialized dealers, both Spanish and foreign; cases that suggest the transformation of both artists and outsiders into dealers; and evidence of networks of agents extending to Latin America. Though there was no specialized tertiary sector, it is clear is that the structures necessary to support an abundant export trade were well-developed.

Fig. 5: Boy painting a merchant’s mark onto a shipping crate. Detail of Acción de la Real Compañia de San Fernando de Sevilla (c.1738), engraving by Pedro Tortolero.

Biblioteca Nacional de España. This image is in the public domain.

The broad brushstrokes outlined above should give way to further research. A full chronology of painting exports, bridging the gap between 1600 and 1630 – the period in which total Indies trade peaked– is still necessary. And, like Peter Cherry recently suggested, we must “take[] account of the realities of workshop practices and artistic production of the time,”38 moving beyond studying painters as individuals to analyze workshops and their organization and practices. For this, we must identify masters; not through membership lists, but by their activities as seen in notarial documents, as accepting commissions and apprentices, hiring journeymen and holding a shop were all activities in which only masters could engage. Lists of dealers and agents should also be culled from these sources and from the shipping registers of the Archive of the Indies to test the propositions here suggested through anecdote.

The proposition that Seville held a mass market for painting from 1500 to 1700 questions long-upheld ideas about Spain’s role in the early modern economy: painting was a booming industry in a country usually viewed as a site of redistribution and consumption. Within the history of art markets, it suggests the existence of an understudied key player contemporaneous to the Bruges-Antwerp-Amsterdam axis. Above all, Seville provides yet another example that questions the preconceptions we hold about artistic production, bringing painters back into their surrounding economic reality, revealing the multitude of players that make artistic production possible, and challenging us to find the data and methods for their analysis.

Felipe Álvarez de Toledo López-Herrera is a Ph.D. candidate at the Art, Art History & Visual Studies Department at Duke University and a member of DALMI, the Duke Art, Law & Markets Initiative.