1. Introduction – Antiquity and Exoticism in the Early 18th Century

1Early modern Europe was deeply interested both in antiquity and in the “exotic”, far-away lands and peoples perceived as different. The study of antiquity rested on a textual tradition, but it could also involve remnants of cities and buildings (Schnapp, 1993). The exotic could be experienced by traveling and through new traditions stemming from travel, formed by maps, books and pictures (Schmidt, 2015). These two areas of knowledge were not uniform or static. Antiquity could be clarified, consolidated, or criticized (Grafton, 1992), as is shown by the far-flung querelle of the Ancients and the Moderns (Fumaroli, 2001). The exotic could change its meaning, expand indefinitely, its parts be rearranged (Gallegos Gabilondo, 2018). What is more, antiquity and the exotic could overlap and question assumptions on history and identity: what if far-away people have an antiquity of their own? What if this exotic antiquity does not correspond to the most trustworthy classical authors, nor to the sacred history of the Bible (Rossi, 1979)? In this paper I will try to show how a place, or better still, an environment was materially assembled within the early 18th-century Republic of Letters, where one of these unsettling intersections took place.

3This story unfolds in the context of early modern Eurasian circulation of goods and peoples and in that of European shifting experiences of Safavid Iran (Matthee, 2009, 2016). The interactions between Europe and Iran were motivated by diplomacy and commerce (Floor & Herzig, 2012); however, as was the case for other parts of Asia, an interest in religion, history and “Oriental” knowledge had a role to play too (Russell, 1994; Hamilton & al., 2005; Brentjes, 2010; Loop & al., 2017; Bevilacqua, 2018). The enthusiasm of travelers for the ruins led to an explosion of contradictory information and armchair scholars started to pay attention. Some questions arose: who built those once grandiose structures? When? And for what purpose? Nobody could agree on their historical meaning (Sancisi-Weerdenburg, 1991).

4These questions are interesting because they involved a familiar type of object which nonetheless resisted a definitive interpretation. Europeans understood ruins in the light of a textual tradition and/or by reading the inscriptions they may carry. In the case of Persepolis, no interpretation based on texts was conclusive, since evidence from Greek and Latin authors, or even from the Old Testament, was incomplete or conflicting. In addition, nobody could decipher the cuneiform inscriptions found at the site (Mousavi, 2012, p. 113-122). This predicament increased the interest in the ruins within the Republic of Letters and stimulated a demand for better descriptions (e.g. Baudelot de Dairval, 1686, vol. 2, p. 721-722).

6This coincidence caused a controversy. De Bruijn, after a discussion involving his patron Nicolaes Witsen, scholar and mayor of Amsterdam, and the humanist Gijsbert Cuper (Chen, 2009; Peters, 2010), published a pamphlet debating the differences between Chardin’s and Kaempfer’s descriptions and his own (de Bruijn, 1714). He tried to demonstrate that his competitors were imprecise or wrong, because they spent too little time on the site and did not master the art of drawing. Only his relation was a truthful reproduction of the ruins:

• 3  “Ten dien einde heb ik aldaer omtrent drie maenden doorgebragt, met dat inzigt, om alle berispinge (...)

That is why I have spent there about three months, with the intention to contradict any reproach and to deserve the trust of my fellow countrymen, as the first to have made the World aware, according to the truth, of the Rests of such a famous and 2000-years-old Work, so well described — so I think — in its proper original form according to Art, as is represented from life in the Prints as accurately as possible. (de Bruijn, 1714, p. 50)3

7These sources and debates are well-known (Drijvers, 1989; Peters, 1989; Drijvers, 1991; Drijvers, 1993; Sancisi-Weerdenburg, 1997), but it is worth going back to them. In fact, the literature discussing the early modern European reception of ancient Persian ruins displays two weak points. First, most of this literature has been produced by archeologists and/or ancient history scholars. As a result, despite its pioneering character and a clear awareness of the problem (Kuhrt, 1991), early modern descriptions of the ruins are often placed in a somewhat teleological history moving from error to truth: the emphasis mainly falls on what travelers “misunderstood” or “got right”. In my opinion, this approach does not allow us to fully understand the contexts in which the ruins were experienced, reproduced, and circulated. Secondly, this literature tends to concentrate on single historical figures, undermining to some extent the study of the wider networks in which these events took place. The approach adopted in this paper — which is but a modest attempt at avoiding these pitfalls — seeks to put the travelers’ experiences in the context of early modern knowledge building. A broad methodological model can be found in recent works about early modern perceptions of the history of philosophy (Levitin, 2015), Hinduism and Buddhism (App, 2010), or “exotic geography” (Schmidt, 2015, p. 145-161). At a different level, the concept of milieu de savoir, understood as an attempt at underlining the ultimately fluctuating and unattainable nature of lieux de savoir (Jacob, 2014), will help me to bring attention to the following points.

8Between the 17th and the 18th centuries, fierce debates raged through Europe about the reliability of classical authors, the value of the Biblical account as an infallible source for the history of mankind in comparison to “exotic” ones (Rossi, 1979), and more broadly about the standing of historiography as a form of valid knowledge (Borghero, 1983). It may be useful to take this context into account when considering the efforts made by the travelers to represent the ruins as accurately as possible (DaCosta Kaufmann, 1999; Burioni, 2013). In fact, although they deployed partially different strategies and showed different degrees of confidence in their claims to truth, all three travelers shared the aim of giving the ruins a definite material and historical identity. In fact, the travelers tried to establish the ruins as a lieu de savoir and then to make it accessible in the form of a book. This portable Persepolis could be used to safely enquire about the history, art, and customs of ancient Persia. However, such a project was doomed to fail. As I will try to show, the travelers’ attempts at crystallizing the ruins were undermined by their fundamentally ecological nature. This was most apparent in the case of the most confident traveler, de Bruijn.

9In fact, the site changed over time due to natural or man-made damages and was experienced by travelers in different ways. The ruins are best understood not as a regulated and immutable place for the development of knowledge, but as a complex environment where the result varies according to the shifting conditions of information gathering, classifying, and processing. I will analyze these issues in the first section of my work. When travelers reconstructed Persepolis at home in order to offer it to the public, the set of textual and visual information retrieved on the field was affected by the social and material conditions of the travelogue’s editing and publishing process. I will address these problems in the second section. To sum up, the on-field experience, as well as the material and intellectual tactics deployed to support the reconstructions of the ruins, were fragile, mobile, and interdependent. In the end, individual variability was such that it was impossible to form a universally acceptable knowledge in and about the ruins.

10As an operative solution for the writing of this paper, I will discuss separately the field experience and the back-home reconstructions of the ruins as a milieu de savoir. In fact, both stages were connected and influenced each other, even if they contributed differently to the potential crystallization of Persepolis as a history laboratory.

2. On the Field

13Another influencing factor was indeed the surrounding network of human settlements, from the local village of “Mircasgoon” (Chardin, 1811, vol. 8, p. 410) to the nearby city of Shiraz. The local inhabitants provided the travelers with food, tools, manpower and information about the ruins. The importance of local actors for the development of “European” knowledge has been widely discussed (Raj, 2007; Schaffer & al., 2009) and thus is hardly surprising. It is worth stressing all the same. First, the ability to tap into local resources could be an essential part of a traveler’s way of materially dealing with the ruins: de Bruijn had to send for a stonecutter from Shiraz when he broke all his chisels trying to detach the sculptures which were to confirm the veracity of his description (de Bruijn, 1711, preface, p. 219; figure 1).

2.1. Acquiring: The Senses and Their Enhancement

15The most basic task carried out by visitors was to acquire information through the senses: travelers experienced Persepolis first and foremost with their eyes and hands. Sight, often supported by touch, appears to have had two functions.

16The first one would be to supply the material perception of the ruins: identifying colors, recognizing shapes and figures, roughly appreciate dimensions and proportions. Touch seemed to be instrumental in getting information about the nature and quality of the materials. However, the outcome of these operations was not uniform. For example, by analyzing their hardness and different colors, Chardin states that the building materials were not taken from the close Kuh-e Rahmat mountain (Chardin, 1811, vol. 8, p. 285-286). De Bruijn does the same thing but maintains the opposite (de Bruijn, 1711, p. 223-224). This was not inconsequential. Chardin’s opinion led him to believe that ancient Persians had the ability to move heavy materials over long distances, while de Bruijn ended up thinking that the monuments had mainly been cut out of the mountain.

17This remark brings us to the second possible function of sight and touch, a reconstructive one. Just as they could process sensory information to understand construction techniques, the travelers could infer the past existence of lost structures from what was extant and hypothesize the initial appearance of the complex. This, in turn, was connected to its original function. According to de Bruijn, the columns had once supported arcades, an appropriate architectural feature for the palace of a king (de Bruijn, 1711, p. 227). For Chardin, instead, the columns had never supported any roof, the lack thereof being a suitable feature for a temple, his favorite hypothesis about Chilminar’s function (Chardin, 1811, vol. 8, p. 280-282). By applying this reconstructive practice, the travelers were trying to give a definite identity to the ruins, but the differences between their results suggest that the stability of Persepolis would be put in danger once a third party would go through its representations.

18However, experiencing the ruins was not just about sight and touch. In one case, Chardin uses an auditory sensation to formulate a hypothesis about the invisible inner structure of Naqsh-e Rostam’s tombs. Having sent two of his servants to explore the chambers,

• 7 “Je leur dis d’aller aussi avant que la lumière les pourroit conduire, et qu’étant là, ils criassen (...)

I told them to go on as much as the light could lead them, and being there, to shout with all their strength. When they did so, we heard their voices rolling in that cavity and saw countless storms of pigeons getting out from the three other openings. This led me to believe that the tombs, or recesses reciprocally separated, did not correspond to the façades and altars on the surface, but that there was just one vault from one end of the mountain to the other. (Chardin, 1811, vol. 8, p. 350)7

20As we have seen, sometimes the acquisition of information was mediated by the senses of the servants. While Kaempfer probably had very few, or no personal servants at all, apparently Chardin had many of them and used them for the exploration of the underground conduits of Persepolis and the tombs of Naqsh-e Rostam (Chardin, 1811, vol. 8, p. 328-332, 346-352). While it is not always clear if they were his own servants, or villagers and passers-by recruited on the spot, also de Bruijn availed himself of the help of others when the task at hand was demanding or unsafe (de Bruijn, 1711, p. 213, 222, 226). Here, money availability was paramount: a reward was needed to persuade others to take on dangerous tasks (Chardin, 1811, vol. 8, p. 331). In this sense, differences in individual conditions of travel may have determined what a traveler could or could not do, just as individual differences in managing the information retrieved through the senses influenced the process of knowledge building. We may say that, in fact, this individual variability put at risk the stability of the Marv Dasht ruins as a lieu de savoir in the very first steps of its construction.

2.2. Classifying: Measurements and Instruments

21The development of a historical knowledge about and through the ruins appeared to be possible only if they were accurately described. Accurateness meant that every bit of it had to be measured, in order to correctly reproduce dimensions and proportions. Every traveler used the units of measurement of his country of origin (table 1), so that the similarity between some of their names does not imply that they were the same. Apparently, they chose units according to the dimensions of the objects to measure.

Table 1 - Units of measurement used by the travelers, in decreasing order of size

	Chardin	Kaempfer	de Bruijn
Distance	Lieue, Pas	Parasanga, Passus	Uren, Schrede
Length	Toise, Pied, Pouce	Passus, Pes,	Schrede, Voet, Duim
Width	Toise, Pied, Pouce	Orgyia, Passus, Pes, Spithama	Schrede, Voet, Duim
Height Depth	Toise, Pied, Pouce	Orgyia, Passus, Pes, Palmus	Voet, Duim

Source: Author’s elaboration

22In practice, measurement was not a simple matter. Columns were high on the ground, rock tombs and Sassanid reliefs were difficult to reach, sometimes even the winter sunlight was unbearable for the eyes (Kaempfer, 1712, p. 332). Kaempfer sometimes admits that he could not take measurements (e. g. Kaempfer, 1712, p. 352): besides being in a hurry, we may argue that he did not have adequate tools. On the contrary, the homogeneity of de Bruijn’s measurements suggests that he was fully equipped.

23However, we find more details in Chardin’s travelogue. Chardin possessed simple tools like ropes and poles as well as complex ones. In particular, he took the height of the Apadana columns with a “quart de nonante” (Chardin, 1811, vol. 8, p. 278), a type of back staff. This instrument, used by engineers to take heights, was not difficult to build but its use requested some expertise (Manesson Mallet, 1702, vol. 2, p. 85-98). Another hint comes from one of Chardin’s plates (figure 2).

26However, the use of such tools may have affected the material result of the travelers’ efforts. Let’s look at how the tomb of Artaxerxes III is represented in Chardin’s plate 67 (figure 4) and in de Bruijn’s plate 158 (figure 5). While de Bruijn arranges faithfully the human figure on the left and the fire altar on the right, Chardin’s plate 67 shows them in reverse.

• 13 “La perspective est dessinée, avec autant d’exactitude qu’on a pu le faire par le moyen de la lunet (...)

The perspective is drawn, as accurately as one could have done by means of the spyglass, and as accurately as the constraints of the place could have allowed; for the mountain being on a slope, and without a plate, one must either look at this work almost perpendicularly, or move away from it out of reach. There is only one figure, whose features neither I nor my painter could clearly discern. It is a mysterious figure that can be seen at the very top of the work. We looked at it several times, without being able to be sure which side it was facing, sometimes it was looking at the fire and the sun, and sometimes we thought it was looking the other way. With that being said, I am sure that the whole representation does not have a false trait. (Chardin 1811, vol. 8, p. 306-307)13

28In fact, the only other plate by Chardin to show such an erroneous arrangement is the general view of the tombs of Naqsh-e Rostam (plate 74), equally taken from far away (figure 6). At the same time, the close-up view of the southern tomb at Persepolis is correct (plate 68), as it is in the corresponding Grelot’s drawing for Bembo (figures 7 and 8). The draughtsman may have made these sketches by looking closer and with the naked eye. When Grelot left Chardin out of mistreatment, the traveler kept his drawings (Van der Cruysse, 1998, p. 175-177). For the tomb of Artaxerxes III and for Naqsh-e Rostam, he may have taken the drafts done through the spyglass’ eye. This question, which must remain open in the absence of conclusive evidence, allows us to focus on how instruments could affect knowledge construction on Persepolis.

29All three travelers will portrait themselves and their aides as interpreters of a practice of measurement and precision. They may have perceived this as helpful in forging a bond of trust with their readers, through the creation of a common frame of reference. At the same time, since they were not completely commensurable, these frames contributed to the locality and specificity of each traveler’s Persepolis (Bourguet, Licoppe & Sibum, 2002). All their tools may have been, to their eyes, instruments of precision: but precision for whom (Schaffer, 2015)? This may explain why de Bruijn will compare Chardin’s measurements to his own: it was the only way to make his criticism understandable (de Bruijn, 1714, p. 24-25).

2.3. Processing: Construction of the Site

30Our travelers had gathered information through senses and servants and had classified it through measurement and instruments. These two stages of the fieldwork prepared and partially intertwined with the processing of the information, namely the construction of knowledge about the ruins. In order to complete this last stage, the travelers adopted different strategies that I will analyze in the next sections. However, senses operated against the background of previous experiences and/or intellectual assumptions. Expectations about how an Oriental temple or royal palace should have looked, or how religious service could have been performed (Rubiès, 2007, 2012), were instrumental in directing the senses and shaping the knowledge that appeared to derive from them. A condition of intellectual mediation enveloped both the field experience and the preparation of the book (Brentjes, 2009).

31If we were to draw an analogy between the traveler-writer and the contemporary natural philosopher, we would underline that theory (historical tradition) and practice (visiting the ruins) did not enjoy a hierarchical relationship in the context of knowledge building (what is Chilminar?) and that instruments played a key role in bridging the two (Gauvin, 2008). It was essential, for example, to verify if the distance between the main terrace of Chilminar and the tombs of the Kuh-e Rahmat corresponded to the “four plethra” mentioned by Diodorus of Sicily in his Historical Library (xvii, 71, 7), the key source for the identification of Chilminar with the royal palace of Persepolis (Chardin, 1811, vol. 8, p. 388; de Bruijn, 1711, p. 229-230; Kaempfer, 1712, p. 312-313). However, figuratively speaking, instruments bridged different islets in a largely uncharted lagoon, not the two banks of a river.

32In addition, the use of instruments and the practice of measurement depended on habits of knowledge (Gauvin, 2011) whose correct deployment was not always guaranteed on the field. On the other hand, their output was calibrated on the specificity of the Marv Dasht ruins. When looking through the spyglass, Chardin never doubted that the blurred figure was looking at the fire and the sun. This corresponded both to other items among the ruins and to his text-based knowledge about the religion of ancient Persia.

33It is also useful to comment on de Bruijn’s ambiguous attitude towards Persepolis and his will to reconstruct it. In fact, something he did on the field put in jeopardy his project, contributing to the instability of the site. In the first place, as I mentioned, de Bruijn snapped some pieces of sculpture from Persepolis, in order to strengthen his claims to truth. However, as we will see concretely for his drawings, this kind of displacement could not guarantee the stability nor the hegemony of his narrative about Persepolis. In the second place, de Bruijn engraved on a wall his name and the year he was in Persepolis (Simpson, 2005, p. 27). The site was permanently modified by taking something away and by adding something. Now Persepolis could speak not only for ancient Persia, but for modern Netherlanders too. Or, at least, this is what may have thought another famous traveler to Persepolis, Carsten Niebuhr, when he recognized de Bruijn’s name engraved in red chalk (Niebuhr, 1778, p. 159).

2.4. Traveling Libraries: The Intellectual (Back)ground

34The potential connection between the experience of travel and that of assembling, possessing, and using a library is very important for our case. According to the auction catalogues of Chardin’s (Levi, 1713) and Kaempfer’s libraries (Catalogus, 1773; Merzbacher, 2004), they owned several classical sources that could have influenced their descriptions of the ruins. However, as we have seen, classical authors did not monopolize the stage. The Bible was an indispensable reference and texts from the Arabic and Persian geographical and historiographical traditions had a role to play too.

3. Back Home

37Chardin, Kaempfer and de Bruijn came back to Europe in different moments and contexts. Chardin, a Huguenot in the time of the Revocation of the Edict of Nantes, sought refuge in London. He was welcomed by members of the Royal Society such as John Evelyn and was elected Fellow in 1682 (Van der Cruysse, 1998, p. 275-280, 292-296). Kaempfer, after a brief stay in the Netherlands, came back to his hometown of Lemgo in Westphalia (Haberland, 1990, p. 85-89), while de Bruijn established himself in Amsterdam, close to his patron Witsen (de Hond, 1997). All three had notebooks and drawings reporting their experience.

38Now their paper Persepolis found itself in a very different environment from the Marv Dasht plain or the travelers’ train: the early modern European city. The impact of the urban milieus on knowledge production has recently been discussed as part of a history of entanglements (de Munck & Romano, 2020). This helps us stress how much these spaces influenced the outcome of the travelers’ attempts at transforming the ruins in a lieu de savoir: the construction of Persepolis depended upon the encounters made by travelers in such cities as London, Amsterdam, Isfahan, Shiraz, Batavia and so on. John Evelyn made Chardin show Grelot’s sketches to influential members of the London high society (Evelyn, 1959, p. 769), while Kaempfer elicited his fellow countrymen’s curiosity thanks to what he had brought back from the Orient (Haberland, 1990, p. 89-90). However, the most interesting case is that of de Bruijn, who transformed his painter’s cabinet in a window on Asia by showing and selling oil paintings and watercolors made during his voyage (Drijvers, 1991, p. 95; de Hond, 1994, p. 65). A part of them may have been about Persepolis (de Bruijn, 1711, p. 224; de Bruijn, 1714, p. 49). Some of the fragments of sculpture were also visible in his cabinet, while others were now in the hands of his patrons Witsen and Anton Ulrich, Duke of Brunswick-Lüneburg (de Bruijn, 1711, preface). As I mentioned, de Bruijn’s show of these items had a precise role in his claims to truth. However, arranging them in his book and circulating them was both an attempt at consolidating Persepolis as a lieu de savoir and a way of losing control over it.

39Considering the spatiality of the travelers’ experiences allows us to also discuss their connected temporalities. We know that Chardin told de Bruijn that he could not draw himself: de Bruijn could have had this conversation with Chardin during his trip to London between 1700 and 1701 (de Bruijn, 1714, p. 6; de Hond, 1997, p. 13). Just before leaving for Asia on 28th July 1701, de Bruijn could have learned about a key element of his future arguments. When he came back to Amsterdam, he had a great advantage, because the city was then the unrivalled center of (illustrated) book publishing (Kolfin & Van der Veen, 2011; Pettegree & Weduwen, 2019). However, he was also exposed to competing narratives about the ruins, which could not fail to arrive in such a place. In fact, Gijsbert Cuper had greatly appreciated de Bruijn’s work, until the (un)timely arrival of Chardin’s and Kaempfer’s books. It is striking that Witsen, who had so much helped de Bruijn, was instrumental on bringing Cuper’s attention to these works (Drijvers, 1991, p. 98-101). This was a first crack in de Bruijn’s Persepolis, because Kaempfer and Chardin enjoyed a scholarly reputation that the painter apparently could not match (Schmidt, 2015, p. 156-159). Only a personal intervention by de Bruijn, supported once more by his patron, resolved Cuper’s doubts (Chen, 2005, p. 18-37). The tensions between the efforts made by the most uncompromising author to crystallize Persepolis and the inevitable renegotiations of his mobile material representations symbolize the back-home part of Persepolis as a milieu de savoir.

3.1. Building the Book I: From the Notebook to the Text

42The transformation of travel notes into a published book involved the publishers as well as the network of acquaintances that the authors considered capable of discussing the interpretation of the ruins — even if most of them had never been there. This is true for Chardin (1811, vol. 1, p. xxix; vol. 8, p. 313-4) as well as for Kaempfer (2001, p. 59-73) and de Bruijn, as I mentioned. In fact, the preparation for the publication implied the reworking of travel notes in a form acceptable to a scholarly public. This was done by embedding the description of the ruins in a frame of reference deemed necessary for its understanding. Each traveler adopted different strategies. Chardin constantly made comparison between the ruins, a variety of textual sources and his first-hand experience of “Oriental” lands and peoples in order to demonstrate that Chilminar had been a temple, built half a millennium before the life of Moses; he also inserted in his description a dissertation on the religious minority that he saw as the heir to the ancient Persians, the “Guèbres” (Chardin, 1811, vol. 8, p. 354-382, 401). De Bruijn’s description is followed by three large historical sections aimed at demonstrating that Chilminar was instead the royal palace of Persepolis (de Bruijn, 1711, p. 228-316). It is likely that the author of these texts was not de Bruijn, but one of his acquaintances, the scholar Praetorius (Drijvers, 1991, p. 92). Kaempfer too gave some historical context, despite being the most cautious on this issue (Kaempfer, 1712, p. 297-304, 315-317). These historical addenda may have been sketched before or during the journey itself, but most likely they took their final form after it. However, these texts could support the travelers’ interpretations only if they were themselves supported by the verbal and visual descriptions.

3.2. Building the Book II: From the Sketch to the Engraving

44Sketches were thus a vital instrument for the representation of the ruins and that is why our travelers committed themselves to their production. Before being transformed into engravings, the sketches had a life of their own. De Bruijn showed his drawings of Persepolis to VOC’s employee and traveler François Valentijn during his stay in Batavia in 1706 (Valentijn, 1726, vol. 4.1, p. 241), just as Chardin showed Grelot’s sketches to his new friends in England and asked for their help in publishing them (Evelyn, 1959, p. 768-9).

47However, it was not possible, nor useful, to apply mimesis to every object. Therefore, some of the images were less mimetic than “characteristic”, in the sense given to this word in the context of what has been called the “epistemic virtue” of “truth-to-nature”. Characteristic images were a compromise between the peculiarity of individual objects and the clearly recognizable groups of which they were part. In other words, these images represented the typical in an individual (Daston & Galison, 2007, p. 70). Kaempfer offered a general view of the mountain façade of Naqsh-e Rostam, with every different relief or tomb marked with Roman numerals. Then he described each “Figura”, most of the time providing a corresponding plate. When he came to the Achaemenid tombs, he explained:

• 24 “Quo loco paulo retrocedit rupes : eo quod lacunis abundaret, nullum coelum experta, spatio 160 pas (...)

At this point the slope of the mountain recedes a little; here, untouched by the climate, it offers an abundance of cavities, for the space of 160 passus. After these, Figures IV, V, VI and VII follow, one after the other and separated by a few passus from each other; we present these in a single representation, since all are similar to each other, like the egg with the egg, and are distinguished only by the number. (Kaempfer, 1712, p. 312)24

48One of Kaempfer’s complaints against Brandshagen referred to this plate (Kaempfer, 1712, p. 313): it was a way to salvage his mimetic attitude from the technical inconveniences he stumbled against in the broader process of “drawing from nature” (Daston & Galison, 2007, p. 99).

50However, it is useful to point out that one of the things reproached by de Bruijn to Kaempfer is that the German spoke of stork nests as being on just one of the columns of Persepolis, while according to his experience there were plenty of them here and there (de Bruijn, 1711, p. 209; de Bruijn, 1714, p. 39). De Bruijn may have believed that only his own experience qualified as a starting point for a true representation of Persepolis, and that he tried to crystallize his own representation as the immutably true one. His Persepolis was outside the contingencies of time: the animals and people he inserted in his plates were meant to reproduce a precise historical moment — that of his visit — but also a regular feature of the ruins, often frequented by nomads and their flocks (de Bruijn, 1711, p. 224).

52What happened later was even more disrupting. A new French edition of de Bruijn’s work appeared in 1725, presumably edited by the savant Antoine Banier (Anonymous, 1725, p. 502). In the first place, publishers and translators found fitting to cut a significant part of de Bruijn’s historical addenda (1711, p. 208-318; 1718, vol. 2, p. 261-293; 1725, vol. 4, p. 301-408), thus undermining the overall force of his arguments. What is more, Banier inserted copious footnotes to de Bruijn’s materials, including his 1714 pamphlet. In this paratexts, Banier contested de Bruijn’s arguments and claims to superiority.

• 27 “Il est vray que M. Chardin ne sçavoit pas dessiner : mais il se servoit de M. Grelot, qui étoit fo (...)

It is true that M. Chardin could not draw, but he had at his service M. Grelot, who was a very clever & very honest man; after all M. le Bruyn does not have so much to complain about the difference between his drawings and those of Chardin, since they are very similar, and one does not see whence comes his animosity against him. Chardin was a very intelligent traveler, and he seems to have examined the ruins well. Everybody is allowed to express his conjectures. (de Bruijn, 1725, vol. 5, p. 308, n. a)27

53Anyway, this was still a small crack in de Bruijn’s personal Persepolis. A greater one came from a luxurious album published in London about a decade after his death, Persepolis Illustrata (1739), consisting of 32 images distributed in 13 plates. It was introduced by an historical dissertation taken from the English translation of de Bruijn’s travelogue (1737) and supporting de Bruijn’s identification of Chilminar with the royal palace of Persepolis. The anonymous person behind Persepolis Illustrata, however, mixed de Bruijn’s visual materials with some of Chardin’s and never specified who was the author of what. This gesture represented, at the same time, a deviation from Chardin’s interpretative stance — because his plates now illustrated a royal palace and not a temple — and an attitude of total negligence towards de Bruijn’s claims. A new paper Persepolis was born, which escaped completely the control of both authors and eluded their attempts at crystallization, while at the same time offering a new one. Some years later, the visual materials found in Persepolis Illustrata were published in the in-4° edition (1747-1748) of one of the most successful editorial projects of the Enlightenment, the London Universal History, and once more de Bruijn’s claims were denied (Universal History, 1747, vol. 5, p. 96-102, n. S-T). From there, the verbal and visual descriptions of the ruins of Marv Dasht so carefully crafted and defended by Chardin and de Bruijn would experience a joint journey through multiple Europeans languages and locations (Abbattista, 1989).

4. Conclusions. The Impossible Construction of Persepolis

54Reproducing the ruins of Marv Dasht was a difficult operation and our travelers knew that well. That is why, while on the field, they took great care in the observation of the monuments. In order to gather and classify information, they raced against the clock, put a strain on their senses and used every resource available: money, tools, manpower. They transferred information to paper in the hope of preserving it unspoiled and of communicating it to the reading public. However, they were not only limited by the conditions of travel and thus by the availability of those resources. They were also entangled in a cultural background that (re)organized their sensible experience, partially determining how they could access and process the information retrieved on the field in order to construct a knowledge about and through Persepolis. Hence, each traveler developed his own representation of the ruins, because the conditions of their visits and the very nature of the site made it impossible to construct each time, by each point of view, the same knowledge of the monuments. That is why Persepolis could be considered a milieu de savoir: the material and intellectual results of early modern travelers’ interest in it was at best fluctuating, at worst conflicting, and always subjected to factors that were impossible to completely control. Yet, when each traveler left the Marv Dasht, Persepolis was like an open book to him (with some unreadable pages, though). Ironically, the very attempt at crystallizing Persepolis in a stabilized entity through the power of the printing press was likely to put in jeopardy the traveler’s will of setting the record straight once and for all. The complexity of the material operations requested by this process, but most of all the circulation of knowledge it implied, made it impossible to keep Persepolis unaltered or to channel in one direction the historical knowledge stemming from (the representations of) it. In this sense, it is striking that the traveler the most committed to claim for his Persepolis the status of truth was also the most successful in spreading materials that could undermine those claims: circulation, readings and re-employments prevented the paper Persepolis from becoming a stable history laboratory and confirmed its identity as a milieu de savoir.