Azuchi Castle and the Azuchi Screens

A low peak with ridges projecting into an inlet lake connected to Lake Biwa 琵琶湖, Mount Azuchi was plucked from obscurity in 1576 when Nobunaga determined the site would be developed as a new capital for his military dynasty. Riding a string of military victories, commanding the largest armies on the archipelago, and poised to expand his already unprecedented and unmatched territorial claims, Nobunaga required a home that could serve to both maintain his position and enhance his public image.⁸ Construction was finished in 1579, and the resulting castle and adjacent town remained a center of politics and culture until 1582, when the castle complex was destroyed in the chaos that followed Nobunaga’s assassination.

The building of Azuchi Castle represented a dramatic return to monumental architecture in Japan. Not since the construction of Shōkokuji 相国寺 and the raising of a massive seven-level pagoda at the end of the fourteenth century had such a large-scale timber frame structure been deployed as a means to project authority on the archipelago.⁹ The scale of construction at Azuchi, however, well surpassed this closest precedent; and in the construction challenges it embodied, Azuchi Castle was more comparable to the eighth-century temple complex of Tōdaiji 東大寺 and its Great Buddha Hall (daibutsuden 大仏殿), one of the largest timber frame buildings ever constructed.¹⁰ Unlike Shōkokuji and Tōdaiji, however, Azuchi Castle was an expression of the reach and authority of a leader unmediated by religious institutions.¹¹ This distinction was made possible by the choice of architectural forms used, namely military or defensive architecture such as towers (yagura 櫓), drystone walls (ishigaki 石垣), and gates (mon 門). Prior to Azuchi, these architectural typologies had been understood as temporary and purely functional.¹² Nobunaga’s achievement at Azuchi was to take advantage of the association between these structures and their warrior inhabitants that had been cultivated after a century of civil wars, to infuse this martial architecture with opulent decoration, and to position the castle and its many structures near a sizeable population. In this way, the structures became an inescapable reminder of Nobunaga’s identity, wealth, cultural sophistication, organizational capacity, and historical place. In turn, the castle served to legitimize Nobunaga’s past successes and, more importantly, to cultivate a sense that his ascendancy was an inevitable fact.

Around the time of Azuchi Castle’s completion in 1579, Nobunaga commissioned various distinguished persons to commemorate his architectural achievements there.¹³ Three such commissions are known: a floor-by-floor and room-by-room description of the castle’s tenshu 天主 (keep), a poetic encomium, and the Azuchi Screens.¹⁴ As noted above, the screens are attributed to Eitoku, the most famed artist of his day. Eitoku’s eminence derived in large part due to his work at Azuchi Castle, where, in 1576 and at the age of thirty-three, he was commissioned by Nobunaga to paint the interiors of several buildings within the castle’s inner bailey (honmaru 本丸). This commission would come to be seen as marking a critical moment in Eitoku’s career, when he transitioned from finely detailed works referred to as saiga 細画 (small [detailed] painting) to the formulation of a dynamic style of large-scale painting referred to as taiga 大画 (large painting), which mostly consisted of vibrantly painted subjects set against gold leaf backgrounds conceived to fill the spacious interiors of castles such as Azuchi.¹⁵

Produced at the tail end of the Azuchi Castle commission, the Azuchi Screens represent both a culmination of Eitoku’s time there, but also a return to his roots. Based on European eyewitness accounts that characterize the screens as on “panels”¹⁶ and “gilded and painted upon,”¹⁷ as well as the architectural subjects and receding diagonal ground plane that appear in Pages 569 and 570, it is clear that the Azuchi Screens were executed in the style of a popular sixteenth- and seventeenth-century painting genre exemplifying the saiga mode: rakuchū rakugai-zu 洛中洛外図 (lit., scenes in and around the capital). As the name suggests, rakuchū rakugai-zu are defined by their subject, the ancient Japanese capital of Kyoto and its environs. Significantly, Azuchi is thought to be one of the first locales of import beyond the capital and its outskirts to be depicted in this mode.¹⁸ As such, the Azuchi Screens represent an important beginning whereupon the genre of rakuchū rakugai-zu evolved into a subcategory of the broader grouping of cityscape screens 都市図屏風 (toshizu byōbu).

Serendipitously, two such urban portraits by Eitoku still exist. Typifying the form and style of the genre as well as Eitoku’s skill in rendering minutiae, the first of these works consists of a pair of six-panel screens known as the Uesugi Screens 上杉本洛中洛外図屏風 (Uesugi-bon rakuchū rakugai-zu byōbu; fig. 3). The second example, a pair of four-panel screens, is known as Scenes of Amusements Around the Capital (Rakugai meisho yūrakuzu 洛外名所遊楽図; fig. 4). The two pairs of screens offer panoramic, bird’s-eye views with a wide array of finely detailed temples, shrines, homes, and shops, as well as courtier and elite warrior estates. In front of and within these backdrops, thousands of figures go about their daily lives. The Uesugi Screens are organized around the layout of the capital’s grid plan. While broadly conforming to the topography of the capital in its day, billowing clouds and gold leaf ground serve to structure the pictorial space to feature a range of prominent institutions. Scenes of Amusements likewise employs golden clouds to organize space, but as compared to the Uesugi Screens, distance is even more compressed as its subjects are topographically situated well beyond the west and south of the capital’s grid.

FIGURE 3. Kanō Eitoku, Rakuchū rakugai, Uesugi Screens, ca. 1565. Pair of six-panel folding screens; ink, colors, and gold on paper, each: 160 x 364 cm. Yonezawa City Uesugi Museum

FIGURE 4. Kanō Eitoku, Scenes of Amusements Around the Capital (Rakugai meisho yūrakuzu), 16th century. Pair of four-panel folding screens; ink, colors, and gold on paper, each: 85.4 x 177.0 cm. Private collection

Based on the similarity of these screens to the sources noted above, it is evident that the Azuchi Screens contained a mixture of gold clouds and ground corralling the castle and its environs in a panoramic manner. A distribution of subjects can also be tentatively proposed using a 1660 description of the screens, presumably based on now-lost accounts, that relates that there were “two decorative screens . . . one of which was adorned with a painted portrait of the new City and the other the unbreachable Fortress of Azuchi.”¹⁹ In other words, one screen depicted the castle and the other the town of Azuchi extending from the base of Mount Azuchi (fig. 5).

FIGURE 5. Map of Azuchi by the author, based on Kido Masayuki, “Azuchiyama to Azuchiyama shitamachi—Oda Nobunaga ga mezashi shita katachi—” (Mount Azuchi and the town of Azuchiyama—the form of Oda Nobunaga’s aims), in Nobunaga no jōkamachi (Nobunaga’s castle towns), edited by Niki Hiroshi and Matsuo Nobuhiro (Tokyo: Koshi Shoin, 2008), 113–35; and Shiga-ken Azuchi-jōkaku chōsa kenkyūjo, ed., Hakkutsu chōsa 15 nen no kiseki zusetsu Azuchi-jō wo horu (Excavation survey: illustrated history of Heisei 15 excavation of Azuchi Castle) (Hikone: Sunrise, 2004), 62–67

The latter “town-screen” almost certainly resembled the Uesugi Screens. Like the capital, the town of Azuchi was topographically flat and composed of a grid of straight, wide streets within which rice paddies, the homes and shops of merchants and artisans, the estates of Nobunaga’s vassals, as well as various area temples and shrines were situated.²⁰ Included among these was the Jesuits’ Azuchi Seminary.²¹ Little is known about the exact location of this structure or building complex other than that it was constructed on reclaimed land and included a three-story tower.²² Although the composition of the screens was likely adapted, as in the Scenes of Amusements, to respond to the unique quirks of Azuchi’s topography and urban plan, the seminary and other subjects very likely served, as in the Uesugi Screens, as backdrops for bustling scenes of commerce, daily life, and festival celebrations.²³

Clues to Eitoku’s approach to rendering the “castle-screen” may be deduced via another analogous cityscape: the Mitsui Memorial Museum’s Jurakutei Screen (Jurakutei-zu byōbu 聚楽第図屏風, fig. 6).²⁴ Jurakutei 聚楽第 was a castle complex constructed in 1585 by Nobunaga’s successor Toyotomi Hideyoshi 豊臣秀吉 (1536–1598) in the heart of the capital and was one of the first post-Azuchi castles. Like Azuchi, the palatial castle estate served as a base of operations and residence.²⁵ The Jurakutei Screen was executed slightly later, in the early 1590s, by an unidentified artist from the Kanō atelier. Like the Uesugi and Scenes of Amusements screens, the painting renders its subject panoramically, employing billowing gold clouds to both organize space and selectively reveal a sprawling urban landscape beyond.²⁶ The composition is arranged around a central apex, the Jurakutei tenshu, situated at the top of the fourth panel from the right. Spreading out below it and corralled by a wall and moat are the buildings of the castle’s inner bailey.

FIGURE 6. Jurakutei Screen, late 16th century. Six-panel folding screen; ink, colors, and gold on paper, 156 x 355 cm. Mitsui Memorial Museum, Tokyo

Primary sources and excavations on Mount Azuchi suggest that many similar subjects were included within the castle-screen. Ascending the southern face of Mount Azuchi is the ōtemichi 大手道, a six-meter-wide staircase-path that was guarded at its base by a 110-meter-long wall,²⁷ characterized by three gates, two at either end and one in its center that checked the ōtemichi’s start. Beyond this wall and situated above it on the mountain were several precincts that have been presumed to be the ruins of the estates of Nobunaga’s vassals and allies.²⁸ The second path, known as the Dodobashi-michi 百々橋道, follows the mountain’s southwest ridge. Shortly up its ascent, this path arrives at the temple complex of Sōkenji. Although unusual for Japanese castles constructed after Azuchi, the inclusion of a temple was a common feature of large pre-Azuchi castles.²⁹ Nobunaga’s decimation of various religious institutions that opposed him can be credited, at least in part, for their later exclusion from castle precincts including Jurakutei. However, prior to this moment, castle-temples functioned as sites for warriors praying for success in battle, for longevity, and to ward off pre-battle curses. Past Sōkenji and halfway up the mountain, the ōtemichi and Dodobashi-michi met and ascended to the peak of the mountain, where the outer and inner baileys of the castle were situated.

As they constituted the tallest and most visible parts of Azuchi Castle, the various buildings and surrounding walls of the inner bailey were almost certainly included within Eitoku’s screens. However, given the limited nature of primary sources and the archaeological record, it is impossible to state with any certainty the layout or exact character of any of these structures.³⁰ The only building confirmed in both the primary and archaeological records is the tenshu.³¹ This structure stood seven stories high atop a drystone pedestal base, had sweeping gables, and was sided with a combination of plaster and black clapboard wall.³² Above these stories and emerging from and contrasting with a tiled, sloping roof that covered the third and fourth floors was a crowning two-story belvedere. The lower story of this belvedere was a red and white octagonal floor.³³ Atop it sat a story characterized by three-by-three bays and covered in part or whole with gold leaf. This pinnacle of both castle and mountain is widely thought to have been capped with a red-yellow tiled, hip-and-gable roof with shachihoko 鯱鉾, ornamental finials placed at either end of the ridge in the shape of an imaginary animal, a type of dragon with a face similar to a fierce tiger and an arched, carp-like body.³⁴

Across the World: From Screens to Sketch to Print

The tale of how the screens crossed the oceans and came to be reproduced in xylographic form is one that has been extensively studied.³⁵ Nobunaga presented the Azuchi Screens to Alessandro Valignano (1539–1606), Jesuit Visitor of Missions in the Indies, as a farewell gift when Valignano was visiting Azuchi around August 1581. Valignano subsequently incorporated them in an endeavor retroactively named the Tenshō Embassy (Tenshō no shisetsu 天正の使節). The embassy, comprising a delegation of four young sons of Christian daimyo, was planned by Valignano to promote the Jesuits’ work in Japan by parading them around southern Europe and, upon the boys’ return to Japan, having them act as educators on the glory of Christian Europe. The embassy departed Nagasaki on February 20, 1582, and, via the Cape of Good Hope, arrived in Lisbon in August 1584. From Portugal, they continued through Spain and into Italy, met with dignitaries and rulers including Philip II (1527–1598), and were greeted with much fanfare in each city they entered. The climax of the boys’ tour was a formal audience with Pope Gregory XIII in Rome held on March 23, 1585. Soon after this encounter, they formally presented the pope with the Azuchi Screens, after which the boys were taken on a tour of parts of the Vatican, during which they were promised that the screens would be put on display in a gallery near the pope’s apartments.³⁶ The boys departed Rome on June 3 of the same year and returned to Japan on July 21, 1590.

It is at this point that Philips van Winghe converges with the history of the screens. Van Winghe had come to Rome in 1589 from his home in the Flemish city of Louvain. While the motives for his relocation are unclear—he is believed to have been either escaping war or hoping to advance his research—his brief time in Rome would leave a considerable legacy.³⁷ Through his impressive connections, van Winghe surveyed the collections of great Roman families, became one of the first to explore and document newly discovered catacombs beneath the city, and in doing all this was an important pioneer of Christian archaeology. While he is best known for his work as an archaeologist, van Winghe’s interests were, like many of his time, broad in nature and extended to multiple cultures and curiosities that he encountered. In this vein, in early July 1592 van Winghe visited the Gallery of Maps (Galleria delle carte geografiche) in the Vatican to copy a fresco of the Lazio region of Italy for his friend the famed mapmaker Abraham Ortelius (1527–1598). Van Winghe’s copy of the Lazio map still exists, and it is due to the inclusion of a letter to Ortelius on its verso that these details are known.³⁸ For lack of any other record of his visiting the Papal City, this occasion is generally assumed to be the moment when he came across the Azuchi Screens and, during this encounter, made at least two sketches of their content. Less than a month later, van Winghe contracted malaria and he subsequently died in September 1592.

The process by which van Winghe’s sketches were translated into printed versions is only partially understood.³⁹ After his death, van Winghe’s belongings were sent to Jérôme (also called Hieronymus) van Winghe (1557–1637), Philips’s brother and canon of the cathedral in Tournai. Aware of the value of his brother’s work to scholars and keen to have it acknowledged, Jérôme advertised the notebooks within his network. Among the people he showed and eventually loaned them to was Nicolas-Claude Fabri de Peiresc (1580–1637), a French polymath and central figure in early-modern European intellectual history.⁴⁰ Peiresc is the critical link through which Philips van Winghe’s sketches found their way into Lorenzo Pignoria’s book. In a letter from Peiresc to Pignoria dated January 4, 1616, Peiresc makes clear that he had obtained a copy of Pignoria’s newly published 1615 edition of Imagini and, as his friend, offers a list of potentially useful objects for Pignoria for future inquiry.⁴¹ Included in the list among other objects is an image of the Mesoamerican god Quetzalcoatl and an image of a “temple of Japanese [people] at the peak of a mountain extracted from the paintings the Japanese ambassadors gave to Gregory XIII.”⁴² Only these subjects from the list (figs. 1, 2, 7) would subsequently be added to the 1624 edition of Imagini, and both are labeled as from van Winghe. A third addition to the 1624 edition—a set of four images showing different views of an Indonesian rakshasa carved on the handle of a kris—are not mentioned in the letter, but are credited in the accompanying text to Peiresc (see fig. 10).⁴³

FIGURE 7. Philip Esengren, Quetzalcoatl, copied from a sketch by Philips van Winghe, in Lorenzo Pignoria, “Seconda parte delle imagini de gli dei indiani,” in Le imagini de gli dei de gli antichi (1624), page 550. Woodblock print, 14.9 x 21.9 cm. Reproduction from the 1626 edition. Private collection

Why did Pignoria add these few images to the 1624 edition? This question will be revisited below, but suffice to say here that they had some relevance to Imagini’s greater thesis.⁴⁴ A native of Padua who studied under the Jesuits, was later ordained a priest, and had a diverse career as librarian, curate, and canon, Pignoria was a prolific scholar and active participant in the international network of scholars known as the Republic of Letters. His revised editions of Imagini represent one of his most important legacies.⁴⁵ Originally published in 1556 by Vincenzo Cartari (ca. 1531–after 1569), Imagini had long served as a highly regarded reference on Greco-Roman mythology.⁴⁶ With his reworking of this classic text, Pignoria had multiple goals. First, he sought to update the book with new methodologies including firsthand visual and philological analysis. The full title of the 1615 edition of Imagini encapsulates his underlying ethos: The true and new images of the Gods of the Ancients . . . readapted . . . to their real, and no more [only] observed resemblances. Obtained from marbles, bronzes, jewels and other antique memories, with deep study and particular diligence. . . .⁴⁷ This new focus on material culture was made manifest in the single largest change to the book: new illustrations produced by Philip Esengren (known as Filippo Ferroverde in the Italian milieu, active early seventeenth century). Pignoria states in his preface that, with these new images, he sought to correct a critical shortcoming in many similar volumes that, he believed, had failed to faithfully reproduce objects of study.⁴⁸ These images along with annotations to Cartari’s text and new illustrative materials transformed Cartari’s work from a popular reference into a scholarly research tool.⁴⁹

In line with this shift, the second major change to the volume would be the addition of the “Seconda parte.” With this section, Pignoria had an ambitious goal. He wanted to prove through comparative readings of iconographies that the exotic religions and icons of these distant lands all originated from and were derivations of ancient Egyptian polytheism.⁵⁰ To be sure, this thesis was not novel. The deluge of accounts and imports from foreign cultures around this time had forced many scholars to grapple with the nature of these cultures as well as their own. Situated within a compendium of classical gods and their iconographies, “Seconda parte” served as a point of contrast, highlighting Europe’s own relationship with this Egyptogenetic past. Herein lay a key point of intersection with van Winghe, whose main research interest centered on antiquity, particularly early Christian art—that is, the moment when Christian culture broke from its polytheistic past. As other cultures appeared to have stagnated, the customs and artifacts of these distant cultures, for van Winghe, Pignoria, and their peers, represented a new window into their own history and the evolution of Western culture.

A Tower? A Gate? A Closer Reading of Pages 569 and 570

Pignoria’s inclusion of van Winghe’s sketches went largely unnoticed outside of Italian and Renaissance studies until a passing mention in a 1969 article by R. W. Littlebown.⁵¹ Donald Lach subsequently reproduced the prints the next year and would be the first modern scholar to rethink Pignoria’s assessment that these prints were of “the temples of some Japanese deity” and to connect the images to Azuchi.⁵² In doing this, Lach posited a revised guess regarding their subject matter: “The sketches of van Winghe show respectively the tower and the gate of Azuchi Castle.” The Page 569 building has since been regularly named as the tenshu or a yagura/tower of Azuchi Castle, while the Page 570 building has come to be known as the castle’s main gate—that is, buildings that were not, as both Peiresc and Pignoria note, those of a temple complex. Lach’s identifications can be traced back to two unstated assumptions: first, that van Winghe, uninformed about Azuchi, would have been drawn to the tenshu; and second, that Pignoria, in selecting van Winghe’s sketch and identifying it as a temple, could not have known any better. In other words, for lack of knowledge about Japan or Japanese culture, it is assumed that van Winghe’s gaze was directed to the tenshu as an obvious apex or focal point in the screens’ composition—that is, as in the Jurakutei Screen—and owing to its privileged position, it is what he sketched. Likewise uneducated about Japan, Pignoria then made the mistake of inserting an ostensibly secular building within a compendium of religious iconography. However, close analysis of these images, particularly Page 569, reveals there is good reason to believe that van Winghe and Pignoria were not so ignorant and that they indeed reproduced parts of a temple.

Before engaging in a close reading of the sketches-turned-prints, it is first necessary to posit one critical correction to Pignoria’s efforts. Van Winghe’s sketches, as they appear in Imagini, are almost certainly mirror images. Accompanying the Quetzalcoatl image (fig. 7) in Imagini that was introduced to Pignoria by Peiresc is the following text:

I [Pignoria] came across another image of Homopoca, or of a similar deity which others say is of Quetzalcoatl. And it was taken from certain papers which were of Winghe of Tournai [sic], learned youth, who claimed to have taken it from a large book, which is in the Vatican Library, compiled by F. Pietro de los Ríos.⁵³

Exactly as Pignoria notes, the original image is included in the Codex Ríos by Pietro del Ríos (dates unknown) and owned by the Vatican Library (fig. 8). Additionally, a copy of van Winghe’s notebook—attributed to Alfonso Chacòn (1530–1599), van Winghe’s friend and collaborator—containing van Winghe’s sketch of the Codex Ríos Quetzalcoatl also exists (fig. 9).⁵⁴ Placing the three images side by side, it is immediately clear that the Imagini version is reversed. Both the Codex Ríos Quetzalcoatl and the Chacòn-copy Quetzalcoatl face left. In contrast, the Imagini Quetzalcoatl faces right. Further corroborative evidence of this reversal exists in the rakshasa image included in Imagini (fig. 10). Exactly as with the Quetzalcoatl, each of Peiresc’s depictions of the rakshasa (fig. 11) are mirror images of their printed copies in Imagini. Accordingly, there is every reason to assume that Pages 569 and 570 were likewise printed with their subjects’ composition reversed. While this change is enlightening as it suggests that the Uesugi, Azuchi, as well as most of the Scenes of Amusements screens share a left-to-right receding ground plane, it does not tell us much about the screens themselves. Still, this reversal is important as it brings us closer to van Winghe’s originals— and, consequently, all discussion hereafter of the prints will refer to their reversed states (figs. 12, 13).

FIGURE 8. Pietro de los Ríos, Quetzalcoatl, in Codex Vaticanus A (Vat. Lat. 3738, also known as Codex Ríos), after 1566, fol. 35r. Vatican Apostolic Library

FIGURE 9. Attributed to Alfonso Chacòn (1530–1599), Folio 58r (Quetzalcoatl) in MS 1564, copied from sketches by Philips van Winghe. Biblioteca Angelica, Rome

FIGURE 10. Philip Esengren, Rakshasa on an Indonesian kris (first of three prints), in Lorenzo Pignoria, “Seconda parte delle imagini de gli dei indiani,” in Le imagini de gli dei de gli antichi (1624), page 586. Woodblock print, 14.9 x 21.9 cm. Reproduction from the 1626 edition. Private collection

FIGURE 11. Nicolas-Claude Fabri de Peiresc, Rakshasa on Indonesian kris, in Cabinet de Peiresc (ca. 1620), fol. 38r. Drawing on paper. AA-54, Cabinet des estampes, Bibliothèque nationale de France (BnF)

FIGURE 12. Page 569 reversed

FIGURE 13. Page 570 reversed

The Page 569 building can be divided into three parts (fig. 14): (1) a central, hip-and-gable roofed, two-level building depicted in a three-quarters view with its façade shaded and extending back to the right; and (2) background and (3) foreground structural elements, the details of which suggest that both are parts of a gate, wall (hei 塀), or roofed corridor (rō 廊). Each of these specifications requires qualification.

FIGURE 14. Folio 569 reversed with guidelines

The dimensions of the central building’s first level may be posited as three bays deep and, extrapolating from this dimension, three to five bays wide.⁵⁵ This reading of the building’s first-level depth is rooted in an essential clarification. On the right side of the building is an apparent copy error by van Winghe and/or Esengren in the form of a smaller “half-bay” (see fig. 14) that suggests the building’s side is four bays deep. This half-bay, however, should be read as the leftmost bay of the building’s façade. The origin of this copy error is readily apparent through comparison with Eitoku’s depiction of the Golden Pavilion (Kinkaku 金閣) of Rokuonji 鹿苑寺 (also known as Kinkakuji 金閣寺; fig. 15) in the Uesugi Screens. The angle at which the right side of the Golden Pavilion recedes, particularly on its second story, is so wide that if not copied precisely the foreshortening can easily be misread as an additional bay. That this is a copy error is further demonstrated by the second-level cusped windows (katōmado 火灯窓), decorative features typically situated in symmetrical fashion in the outermost bays of buildings.⁵⁶ The right window of the Page 569 building should, like the left side, have only a single column to its right. It follows that the building’s second level can be confidently posited as three-by-three bays.

FIGURE 15. Kanō Eitoku, detail of the Golden Pavilion (Rokuonji) (left screen, panel 2), from the Uesugi Screens

These dimensions represent a foundational data point as, when cross-referenced with the building’s fixtures, they clarify the structure to be a Buddhist hall. To explain this, it is best to start with the most clearly rendered decorative fixture: the cusped windows. Within the Japanese premodern context, two-level buildings of a size similar to the Page 569 building that possess windows of this type are limited to a few architectural typologies: pavilions (such as the Golden Pavilion), temple or castle gates, turrets, and Buddhist halls. The pavilion and temple gate typologies are easily eliminated for consideration. Pavilions as a norm possess pyramidal roofs with cypress shingles, not tiles;⁵⁷ temple gates invariably possess a one- or two-bay depth at most, not three.⁵⁸ The character of the Page 569 building’s first-level side bays further narrows this list down to a finalist. Rectangles in the two rightmost bays suggest paneled doors (sankarado 桟唐戸) and, above these doors, either non-penetrating tie beams (nageshi 長押), overhead walls (kokabe 小壁), a transom (ranma 欄間), or a combination of these fixtures. Critically, the backmost (i.e., leftmost) bay lacks the elongated rectangles suggestive of paneled doors. In other words, the building’s side possesses paneled doors in two front bays and a blank, presumably plaster, wall at the back. This layout is one commonly found in the side bays of Buddhist halls with a three-bay depth.⁵⁹ In contrast, the castle gates and towers lack, for functional defensive reasons, multiple entryways to their interior or, in the case of gates, doors on their sides.⁶⁰

Although the building’s context does not offer additional direct evidence to support this reading of the Page 569 building as a Buddhist hall, it is consistent with such an interpretation. The background structural element—crossing behind the ridge, gable, and rightmost hip of the central building—contains multiple hints that it is either the roof of a connecting corridor or surrounding wall. Within Eitoku’s other cityscapes, only two types of structures extend out from the backs of gables. The first of these is an adjacent overlapping gable such as those depicted in the Uesugi Screens rendering of the Manjuji Buddha Hall 万寿寺仏殿 (butsuden) and Main Gate 三門 (sanmon) (fig. 16). The second type is a roofed corridor, exemplified in the Uesugi Screens depiction of the Shōkokuji Buddha Hall (fig. 17). While the left projection of the Page 569 building shares the same shape as an overlapping gable, its patterning of parallel lines with half circles at their base and intermittent squares below clearly identifies the structural elements as tiled eaves and not a gable. This reading is corroborated by the rendering of the figures on top of the structure. Figures truncated from the waist up, such as those depicted in Page 569, are without exception in Eitoku’s other cityscapes situated within courtyards and beyond a roofed corridor or wall (fig. 18). Accordingly, the placement of another truncated figure on the projection to the right of the gable suggests that its upper trapezoidal member is a continuation of the leftmost projecting corridor or wall.

FIGURE 16. Kanō Eitoku, detail of Manjuji (right screen, panel 1), from the Uesugi Screens

FIGURE 17. Kanō Eitoku, detail of Shōkokuji (left screen, panels 3 and 4), from the Uesugi Screens

FIGURE 18. Kanō Eitoku, detail of the Imperial Palace (right screen, panel 6), from the Uesugi Screens

Although tangential to the discussion of the building’s context, it is worth pointing out that the right lower half of the background structural element is likely a corner eave. The basis for this assertion is twofold. First, the second-level eave and rightmost edge of the right projection are aligned in angle and orientation (see fig. 14). Second, such a rendering of a corner eave is consistent with Eitoku’s brushwork elsewhere. As visible in several details (figs. 15–18), the back-right corner eaves of buildings in Eitoku’s paintings, particularly those like the Page 569 building, invariably project out from a building’s walls.⁶¹

Returning to the matter of context, the foreground structural element remains the most difficult part of the image to read, but one key attribute suggests that it is also part of a gate, wall, or corridor. The lower register of the structure lacks any direct parallel in Eitoku’s cityscapes and, consequently, remains tantalizingly ambiguous. The upper register, however, contains two sets of three elongated and mirrored S-shaped forms. Their distinct shape corresponds to only two forms found in Eitoku’s other cityscapes: eburi-ita 柄振板 (lit., hoe board) and the left gables of hirakaramon 平唐門. Eburi-ita are decorative cusped wooden boards that are attached to the ends of wall roofs where the wall connects to a gate or where there is an end to a wall. A hirakaramon is a type of gate with undulating or cusped bargeboards on the sides of its roof. Eitoku’s treatment of eburi-ita represents another signature habit of brushwork. In contrast to cityscape artists who either entirely omitted or cursorily rendered eburi-ita, Eitoku invariably rendered this fixture as pronounced in height and rising well above the wall that it serves (fig. 19).⁶² Although the S/ shapes in the Page 569 building appear as more condensed and in greater number than in any single example of eburi-ita or hirakaramon within Eitoku’s other cityscapes, their distinct shared shape and the lack of anything even close in resemblance suggests that they are likewise cusped boards and positioned in a context where such fixtures are typically found, namely as part of gates, walls, or corridors. In sum, the central building appears to be situated between these architectural typologies. While Azuchi contained many such walled-off areas of varying function, one common place to find such structures is surrounding a temple complex and temple buildings—that is, the setting of a Buddhist hall.

FIGURE 19. Kanō Eitoku, detail of Hosokawa kanrei mansion (left screen, panel 3), from the Uesugi Screens

Before moving on to the implications of this reading of Page 569, a brief word is required about Page 570. Although it is the easier of the two images to read, identification of its subject remains elusive. Depicted is a two-level gate capped with a tiled hip-and-gable roof and with walling characterized by exposed beams and lintels (shinkabe-zukuri 真壁造), similar to the exterior of the Maruoka Castle tenshu 丸岡城天守 in Fukui 福井 (fig. 20).⁶³ Four kneeling figures sit on the staircase leading up to the gate’s threshold. Beyond them and within the gate stands a wall that corrals the path at a sharp right angle—an important detail that clarifies the gate as a masugata 枡形 type, a standard gate typology in Japanese castles characterized by perpendicularly set doors. In the foreground and flanking the staircase is a drystone wall with a board roof (naga-itabuki 長板葺き). While multiple masugata gates have been found within the Azuchi Castle ruins, excavations to date have yet to identify traces of foundation stones or walls that correspond to the structures depicted.⁶⁴ While frustrating, this absence of a clear parallel is also revealing as it disqualifies multiple masugata gates, such as the ōtemichi and several other excavated gates on Mount Azuchi, and in turn tentatively opens the possibility for determining its identity based on an identification of the Page 569 building.

FIGURE 20. Tenshu pedestal and first story, Maruoka Castle, Fukui, constructed 1576. Photo by Nakamura Yasuo

Nobunaga/Nebuchadnezzar’s Draw

The above reading of the Page 569 building as a Buddhist hall reveals that Lach’s unspoken assumptions of naïveté merit a second look and raise the question of how Pignoria, Peiresc, and presumably van Winghe (as discussed below) might have concluded that these buildings were part of a temple, and selected that temple out of all the subjects in the Azuchi Screens for sketching, sharing, and publication.

This question is one that was first tackled by the historian Takemoto Chizu.⁶⁵ In her study of van Winghe, Takemoto discovered that van Winghe had some knowledge of Azuchi prior to encountering the screens. Within the pages of the only surviving original copy of van Winghe’s notebooks, owned by the Royal Library in Brussels, is a list of seminaries and colleges founded during the Papacy of Gregory XIII. At the top of this list is “Azuchiyama.”⁶⁶ While only a brief mention, the appearance of Azuchi in the notebook is significant in that it proves van Winghe had both access and opportunity to learn about Azuchi from either Jesuit accounts or published records on the Japanese mission. The reference therefore disproves the notion that van Winghe was a blank slate drawn to the screens for no other reason than their beauty or exoticism. This knowledge of Azuchi also may help to explain why, of all the objects he encountered during his visit, van Winghe might have spent time with the screens. In his aforementioned letter to Ortelius, van Winghe mentions the heat and discomfort caused by his having to stand to copy the Lazio map in the Gallery of Maps.⁶⁷ Under such strained conditions, and as many a scholar may be able to attest, a little knowledge of a subject does much to captivate and sustain interest even on a hot day.

Takemoto was also the first to propose that Pignoria’s characterization of these images as “a temple of a Japanese deity” may be related to one of the most widely known and repeated myths about Nobunaga: his alleged apotheosis.⁶⁸ The myth of Nobunaga demanding to be worshiped as a god derives from a report authored by the missionary Luís Fróis (1532–1597) that details the events leading to Nobunaga’s assassination. In one of its key passages, the founding of Azuchi Castle’s dedicated temple, Sōkenji, is described:

. . . [Nobunaga] finally determined to break out with the same imprudence and insolence of Nebuchadnezzar, pretending to be worshiped by all, not as an earthly, and mortal man, but as if he was divine, or lord of immortality, and to put into effect his nefarious, and abominable desire, he had a temple built nearby his mansions, on a hill that is separated from the fortress, where he wrote the intent of his venomous ambition, in which he said in this way, translated from Japanese into our language: In these great kingdoms of Japan in the fortress of Azuchi, in this land that seen from afar causes joy, and contentment to those who see it. Nobunaga, lord of all Japan made this temple, with the name of Sōkenji. . . .⁶⁹

Fróis’s description of Sōkenji’s founding as a site of worship for Nobunaga was meant to serve as an example of his hubris and, in turn, serve as an explanation for the sudden, dramatic death of a leader whom Fróis had for over a decade lauded as a critical ally in the Jesuits’ proselytizing efforts.⁷⁰ While more conjecture than fact, the report is nonetheless critical to the discussion at hand as it links Sōkenji to Azuchi. Further research is required to determine whether van Winghe could have seen Fróis’s account reproduced.⁷¹ Yet, even if he had not read it, van Winghe could have learned about its contents from Jesuit friends.⁷² Alternatively, the gallery’s caretakers could have learned of Nobunaga and Sōkenji from any number of sources and conveyed the information to van Winghe in a passing remark. All this is to say that there are any number of conduits that would have allowed van Winghe to connect Fróis’s letter to the subject of the screens. Accordingly, he would have had good reason to search for and then focus his interest on the Icarian temple Sōkenji.

The above reading of the Page 569 building allows Takemoto’s ideas to be taken a step farther.⁷³ Indeed, it may be posited that van Winghe and Pignoria knew the precise subject of Page 569—a building within the grounds of Sōkenji, specifically, a structure known as the Bishamon Hall. Most importantly, both Pignoria’s and Peiresc’s descriptions of the images point to this conclusion. Their characterizations of the temple as “placed above some steep rocks” and “at the peak of a mountain,” respectively, narrow the search down to subjects within the castle-screen, which would have included many steep, drystone-wall foundations, and disqualifies subjects on the flat terrain of the town-screen. Among the temples likely depicted in the Azuchi Screens, only Sōkenji meets this essential qualification.

Assuming that van Winghe wanted to find Sōkenji, he likely did not have trouble doing so. Fróis explicitly notes in the same letter that the temple was near the fortress. As a religious institution, Sōkenji would have stood out among residential estates, fortifications, and the main castle complex as the coloring of these buildings distinguished them from other building groups on the mountain. Later depictions of Sōkenji depict it as containing white and vermillion buildings (fig. 21). Standard for Buddhist architecture, this color scheme would not have been used on residences or other secular buildings. Moreover, based on the depictions of temples in Eitoku’s other cityscapes, it can be assumed that Sōkenji would have been populated by robed and tonsured monks (see figs. 15–17). The resemblance of these monks to tonsured Franciscans or black-robed Jesuits makes it possible that van Winghe was able to deduce the nature of Sōkenji’s buildings based on their inhabitants alone.⁷⁴

FIGURE 21. Detail of Sōkenji from Gō-shū Gamō-gun Toyoura-mura to Suda-mura sanron tachiai ezu (Illustration of mountain [territory] dispute in Toyoura Village and Suda Village, Gamō District, Ōmi Province), 1695. Ink and color on paper. Private collection

Fróis’s account would have allowed van Winghe to narrow his search even further. Shortly after the quote above, Fróis mentions the enshrinement of a stone, named as a “Bonção” (J: 盆山 bonsan), that contained Nobunaga’s essence (“Xintay;” J: 神体 shintai). Thereupon, the placement of this stone is described: “[Nobunaga] ordered to be made in the most eminent place of the temple above all the Buddhas a certain kind of [procession] float, or tiny closed chapel, where he had [the stone] placed.”⁷⁵ Van Winghe almost certainly did not know the form of a Japanese float (mikoshi 神輿) or “closed chapel” (zushi 厨子), but the contours of this description—that is, a retrofitted structure that was the “most eminent place … above all,” within Sōkenji—would have been sufficient.

Only five structures are definitively known to have been part of the Sōkenji precinct around the time of the creation of the screens: a niō gate (niōmon 仁王門), a three-level pagoda (sanjū-no-tō 三重塔), a noh stage (butai 舞台), a viewing gallery (osajiki 御桟敷) for the stage, and the Bishamon Hall.⁷⁶ The gate and pagoda still exist, and they along with the Bishamon Hall can be seen in an 1805 print depicting Sōkenji (fig. 22). The gate sits below the temple on the Dodobashi-michi. Notably, its typology is wholly distinct from the gate depicted in Page 570, and consequently, there is little chance that it is the subject of that print. The pagoda is likewise situated on the periphery of the temple complex, positioned at a level below the other temple buildings. As with the gate, the pagoda’s form is so distinctly different from the subjects of Page 569 or 570 that there is no possibility that it is one of the buildings depicted. Not appearing in the print are the stage and gallery. Both structures are mentioned in passing in Nobunaga’s posthumous biography, and the only clue as to their location within Sōkenji comes from the stage’s full name as it appears in that text: the “Bishamon Hall stage” (Bishamon-dō gobutai 毘沙門堂御舞台), suggesting that the stage, along with its audience gallery, were situated in some proximity to the Bishamon Hall. Owing to the unique nature of these architectural types, and a lack of comparative examples in Eitoku’s other paintings that might help us to reimagine their forms, it is difficult to be sure how these structures were rendered. Thus, their forms may hold a critical “missing link” that might explain the unusual character of the back- and foreground structural elements in Page 569.⁷⁷

FIGURE 22. Akisato Ritō, “Azuchiyama Sōkenji,” from Kisoji meisho zue (A collection of images of famous places on the Kisō Road), 1805. Private collection

Only one building remains: the Bishamon Hall. Easily identifiable in the 1805 print as the largest, highest-reaching, and most central structure, its importance within Sōkenji is immediately appreciable. While it is impossible to know the manner in which it was depicted by Eitoku, the building’s core attributes—that is, central, multistoried—also likely made it stand out from its neighbors. In sum, the precise reason that van Winghe was attracted to the Azuchi Screens and specifically the two buildings he sketched is unknowable, but both motive and means exist.

As the Bishamon Hall burned down after a lightning strike in 1854, the primary and archaeological records are unfortunately ambiguous and inconsistent. Written descriptions and illustrations prior to its destruction suggest that it was a building that underwent several significant renovations.⁷⁸ The original building is generally thought to have been transported to Sōkenji during Azuchi’s construction around 1576, and as noted in Fróis’s report, it was then refurbished to serve Nobunaga’s designs.⁷⁹ Some time after Nobunaga’s death in 1582, Sōkenji changed its sect affiliation to Rinzai Zen and was reinvented as a caretaker of the Azuchi Castle ruins and Nobunaga’s grave there, a role that the institution has maintained until the present day. Contradictory records from the Edo period (1616–1868), including the 1805 print cited above, suggest that the building was renovated or underwent significant structural changes, possibly at two separate times.⁸⁰ Indeed, its original name of Bishamon Hall—known via the quote above about the adjacent stage—likewise appears to have changed twice, to Abbot’s Quarters (hōjō 方丈) and then later to Main Hall (hondō 本堂). Some Edo-period records confirm that the hall was five-by-five bays and possessed a second level or story.⁸¹ The archaeological records likewise confirm the dimensions of the hall as five-by-five bays. However, in the aforementioned 1695 image (fig. 21), the structure appears as an abbreviated two-by-two-bay structure; and in the 1805 print, the building layout, with five-by-three bays, is cosmetically consistent with the Page 569 building. In sum, owing to the multiple renovations and the loss of the building, the exact form of the hall in Eitoku and Nobunaga’s day is impossible to know.

Stronger evidence that the Page 569 building is the Bishamon Hall, however, may be found in the reasoning behind Pignoria’s selection of these images for publication. Van Winghe’s sketches of the Azuchi Screens stand out within Pignoria’s “Seconda parte” as the only examples of architecture within the text and as one of only three subjects from Peiresc’s offerings that would be added to the 1624 and subsequent editions. In these respects—that is, inconsistent with the book’s focus on iconography and as selected among several options—their inclusion appears to be very deliberate. This considered approach is consistent with research by Paola von Wyss-Giascosa on Pignoria’s process in compiling the 1615 edition of Imagini.⁸² In her analysis of the “Seconda parte,” she additionally points out the difficulty that Pignoria had with the Japanese section, specifically a lack of iconographic examples from Asia. This issue was so acute, and undermined his Egyptogenetic argument to such a degree, that he strayed from his own stated scholarly principles of relying on careful, first-person visual analysis and instead had Esengren rely on written accounts to fill a gap in the visual record.⁸³ Wyss- Giascosa only briefly considers the additions to the 1624 edition, and van Winghe’s prints are not mentioned in her discussion, but they fall perfectly in line with her assessment of Pignoria’s struggles and methodology. Lacking visual sources from Asia and eager to prove his main point on the Egyptogenetic origins of non-Christian religions, Pignoria found van Winghe’s sketches, precisely because they were of Sōkenji, ideal for inclusion even though they strayed from the book’s focus on iconography.

It is the draw of Nebuchadnezzar that, I contend, accounts for Pignoria’s selection of van Winghe’s sketches of the Azuchi Screens. It may be safely assumed, based on the similarity of Peiresc’s and Pignoria’s descriptions and the nature of van Winghe’s surviving notebook—filled with a mix of images and notes—that the identification of the building as a temple on steep mountains came directly from notes accompanying van Winghe’s sketches. Pignoria, as a self-proclaimed expert on the “Indies” and dependent on Jesuit accounts of Japan to compile the “Seconda parte,” was almost certainly aware of Fróis’s account. Consequently, when he encountered van Winghe’s notes, he could have easily and independently connected the dots and identified Nobunaga as a “Japanese deity.”⁸⁴ Worth noting also is that the names of Azuchi, Nobunaga, Sōkenji, or even Fróis may have been included in van Winghe’s notes. The aforementioned Chacòn copy of Quetzalcoatl (fig. 9) reveals that while Pignoria borrowed from van Winghe’s notes, he did so selectively. More information regarding Quetzalcoatl, such as the god’s Mesoamerican origins, is included, but Pignoria did not transcribe everything available to him. Such omissions are consistent with the Japanese section of the “Seconda parte.”⁸⁵ In contrast to the section on Mesoamerican gods, Pignoria limits his discussions to descriptions and, thereby, on its surface sidesteps his overarching Egyptogenetic argument. Yet while less actively promoted, this argument still represented the core organizing pillar. Although an artificially drawn parallel, Fróis’s comparison between Nobunaga and Nebuchadnezzar perfectly complemented his thesis of the new world containing echoes of the ancient past, and thereby justified the addition of van Winghe’s Azuchi Screens sketches.⁸⁶

Conclusion

For all of the efforts by Nobunaga and the Jesuits, it is a strange irony that the only vestiges of the Azuchi Screens are a pair of awkward prints based on sketches by an early Christian archaeology specialist and included in a book on religious iconography published a continent away. Nonetheless and while their story and character are complex, these prints represent critical windows onto the Azuchi Screens, its castle subject, and the reception and dissemination of information on Japan in European circles in the late sixteenth and seventeenth centuries.

The proposition made here that the subject of Page 569 is the Sōkenji Bishamon Hall is tentative, but it has multiple implications for future research. As related to the Azuchi Screens and their content, identifying a single point such as the Page 569 building has the potential to open multiple new lines of inquiry. These include comparative analysis with the topography as well as determining the position of the viewer and the arrangement of castle and town on the left or right screen. Such inquiries may appear minor but have the potential to shed light on Nobunaga and the Oda clan’s perception of the capital vis-à-vis Azuchi, and on the degree to which Hideyoshi relied on Nobunaga as a model for consolidating his authority. In other words, this identification may help in understanding critical unspoken assumptions about the definition of power and authority at a key moment in Japan’s formation as it moved from the medieval to the early modern era. Such an identification would also have implications for our understanding of the development of Azuchi Castle and Sōkenji, in particular. Research into the main hall of Tsukubusuma Shrine 都久夫須麻神社 on the island of Chikubushima 竹生島 has revealed that it was originally a three-by-three-bay structure, conceived as a memorial for Hideyoshi’s son, but then transplanted and expanded to five-by-five bays after being moved to Chikubushima.⁸⁷ This transformation is eye-catching as it seems to have echoes in the Bishamon Hall: as a transplanted building, as a structure that might have been expanded from three to five bays deep during the refurbishment Fróis describes, and even as a site of self-deification. In other words, it could shed light on the evolution of memorialization and self-deification during this era. In this new identification also lie the grounds for expansion of the excavations on Mount Azuchi. The form of any gate at the base of Mount Azuchi checking the start of the Dodobashi-michi is entirely unknown, and the site has never been excavated. Yet while the existence of a gate at this location is unknown, it is an obvious spot for a gate and as such, it represents a solid candidate for the subject of Page 570. As the initial entryway to the path leading up to Sōkenji, a pairing of gate and main hall is a logical one for van Winghe to have illustrated—that is, a start and a climax. Notably, such an arrangement is one that finds an echo in the Jurakutei Screen. If one assumes that the Azuchi Screens were similar in character to the Jurakutei Screens, then the comparatively low angle of the Page 570 gate suggests that it occupied a spot in the lower register—that is, exactly where one would presume such a gate, located at the base of the mountain, to be found.

Finally, it is my hope that this reading of van Winghe’s images reveals that, despite the Eurocentric views of their authors, their research into Japan and Asia and that of their peers merits further inquiry. To be sure, the time that the three authors of Page 569 and Page 570—van Winghe, Pignoria, and Esengren—likely spent considering the Azuchi Screens was not greater than a few hours or days at most. Moreover, Pignoria and Esengren never saw the screens “in the flesh.” Yet, the two principal actors, van Winghe and Pignoria, were not passive or unaware observers. Both were well-read scholars with some knowledge of Japan and Japanese culture; both placed a premium on firsthand viewing and careful formal analysis; and their attention to detail shines through even in a copy of a copy. Interested in making sense of the new worlds and not yet weighed down by later, more fully formed colonialist attitudes, the work done by Pignoria and his peers represents a rich frontier for an understanding of European engagement with Asia, and accordingly merits reappraisal.