One of the most interesting presentations of the conference, I thought, was one by Mark Erdmann, on “The Chinese Roots of the Azuchi Castle Donjon.” Now, I am by no means an expert on castles, let alone on Azuchi, and so I’m sure that a lot of what I found really new and exciting in this presentation might be old hat for some of my friends at the Samurai-Archives, who are more well-read, and more focused, on such topics. But, precisely because I know relatively little about Azuchi, and as it relates to artistic display, and performance of legitimacy (performativity), and intertextuality – on top of the basic fact that castles are cool – I found it a really fresh, exciting presentation.
Before we get into it, though, I just have to take a moment to say how much I hate the word “donjon.” Okay, maybe hate is too strong a word. But, I really don’t understand why we should ever be using a French word – which is hard to be sure you’re pronouncing it correctly, which sounds too much like “dungeon”1, and which is just a tad too obscure to be sure that your readers/listeners know what you’re talking about – when we have the perfectly good English word “keep,” and the even more precise Japanese term tenshu.
The modern, post-war reconstruction of Azuchi Castle. I’m not sure which interpretation / version this was based on. But, the gold structure at top, and red mingtang-like structure below it, are clearly visible.
Anyway, that brief aside aside, Erdmann’s talk focused chiefly on two points: (1) the origin of the term tenshu, and (2) a new theory as to the symbolism of an octagonal section near the top of Azuchi’s tenshu tower, which he suggests played an important role in conveying discursive symbolism of Nobunaga’s legitimacy.
Perhaps we should start at the beginning. Azuchi Castle was built over the course of 1576-1579, by Oda Nobunaga, who had just secured his control over most of central Japan. As such, it was built not only as a residence and base of operations, but also as a monument to Nobunaga’s wealth and power, and was covered inside and out in elaborate architectural elements and ornate decorations. Its main tower keep, or tenshu, was decidedly unique and bizarre, and various major elements of its design were not emulated by any later structure; however, the very fact that it had this multi-story tower keep, built atop a considerable stone foundation, and decorated up with various sorts of gables and other architectural elaborations, set a groundbreaking precedent for what would soon afterwards become the standard form for Japanese castles – luxurious aristocratic residences posing as (or doubling as) military headquarters & fortifications.2 Perhaps indicative of how innovative a concept it was, Azuchi was not even called “Azuchi Castle” (安土城, Azuchi-jô) at the time, but rather, the “castle” and the town associated with it, were known as Azuchi-yama (安土山, lit. “Mt. Azuchi” or “Azuchi Mountain”).
The castle was destroyed in 1582 by Akechi Mitsuhide, the traitorous retainer who engineered Nobunaga’s demise at Honnôji. More or less all that survives, as I understand it, is what has been recovered through archaeological excavation – in other words, chiefly, the foundation stones. Based on this and various forms of textual and visual evidence, Naitô Akira, in the 1970s, proposed a certain understanding of the style and form of the castle; more recently, Miyakami Shigetaka has revised Naitô’s version, arguing that Naitô did not consider or corroborate enough different sources, and that his own (Miyakami’s) new version is more accurate to what the castle likely actually looked like. From what I gather, these two are the most prominent voices in this debate, and the most prominent competing conceptions of the structure.
What makes Azuchi so bizarre? Well, rather than having a tower of purely rectangular levels (stories), in a consistent, coherent architectural style & aesthetic, built atop a rectangular base, Azuchi included a couple of extra layers that, from the graphics Erdmann showed, look very much like two additional buildings simply stacked atop three stories of much more typical-looking tenshu architecture. The topmost story was three by three bays square, and covered in gold; Erdmann describes it as resembling quite closely the famous Kinkakuji, or Golden Pavilion, of Shogun Ashikaga Yoshimitsu, and in fact argues that it was directly, intentionally, based upon Kinkakuji, in order to draw a symbolic connection between Nobunaga and the Ashikaga shoguns, and to therefore bolster his legitimacy. The layer below that, octagonally shaped, and painted a bright red or vermillion, Erdmann argues, was meant to evoke the Chinese concept of the Mingtang.
The Mingtang (明堂, J: meidô), Erdmann explains, is a concept going back to the Duke of Zhou, an ancient Chinese figure who will come up again when we discuss the origins of the term tenshu. The Mingtang was a powerfully symbolic structure, roughly circular in shape, essentially just the circular space within a ring of columns, which would, like so much else in Chinese Imperial architecture, represent a re-creation of the cosmos. As the Emperor walked around the circle within the Mingtang, he would be symbolically passing through the four cardinal directions (plus center), and through the four seasons, as well as through the twelve zodiac signs, representing the hours of the day, the days, and the years. I don’t quite have the language to express it, and one of these days I really do need to learn the best way to express it – and thus also to understand it – but, traditionally, in China, the Emperor was seen as embodying, or reenacting, or simply existing at the center of, the functioning of the cosmos. And so, already, we can begin to understand what it would have meant, symbolically, discursively, for Nobunaga to walk around within a room resembling, recalling, the Mingtang of ancient China.
According to Chinese belief, only a proper rightful Sage King (C/J: ??) can build a Mingtang; therefore, the very ability of Nobunaga to construct one serves as a sign of his legitimacy. Ones built in China over the centuries have varied dramatically, but all follow certain common forms – namely, much like Azuchi Castle, the Mingtang is composed of squares topped with circles, as seen in the diagram above. This octagonal hall at Azuchi further resembles, or recalls, Chinese architecture with its inclusion of red roofing tiles, a relative rarity in Japan compared to the grey tiles seen on the lower levels of Azuchi, and thus very much evocative of China. Furthermore, the Azuchiyama-no-ki (安土山記, “Record of Mt. Azuchi”) explicitly describes Nobunaga as a “Sage King,” and as a genius for his choice of Azuchi as the site for his castle, recognizing that Azuchi-yama was just as great as Taishan (Mount Tai), the famous Chinese mountain where, incidentally, the first Mingtang was erected. Erdmann questions if Nobunaga’s welcoming of Jesuit missionaries at Azuchi was intentionally, consciously, intended to mirror the Duke of Zhou’s welcoming of “people of the Four Quadrants.” Going beyond the mere architectural forms, Nobunaga also installed within these top two stories (the Mingtang-esque level, and the golden pavilion-style level) series or systems of wall paintings, by great Kanô artists, depicting Confucian and Buddhist themes related to discourses of rightful, virtuous kingship.
There are a few problems with this system of symbolism, however, as Erdmann points out. Firstly, what are the proper dimensions for a Mingtang, and does Azuchi match these dimensions, and the arrangement of circles and squares, well enough to properly qualify, and function, as a Mingtang? Second, sometimes the same thing can have very different meanings in different contexts. The structure Nobunaga placed on this fourth story of Azuchi Castle may have been intended to resemble a Mingtang, but this is also the form of the octagonal halls (円堂, J: endô) seen at Japanese Buddhist temples, where they are associated with memorial functions. Erdmann gives the examples of the Yumedono at Hôryû-ji, dedicated to the memory of Shôtoku Taishi, and the Hoku’endô at Kôfuku-ji, dedicated to the memory of Fujiwara no Fuhito. In Japan, this form reminds people of memorial functions, when in order to serve the discursive purpose of the Mingtang, Nobunaga needs it to evoke ideas of his living power and righteousness.
Turning to another side of Erdmann’s talk, there was the issue of the meaning and origin of the term tenshu (天守), which refers to the castle’s tower keep. Erdmann traces the origins of the term to 1579, and identifies it with Nobunaga’s efforts at evoking discourses of legitimacy, by tying himself to complex and ancient discourses related to the Mandate of Heaven (天命). One of the first steps in his discursive schemes was the renaming of Inabayama castle to “Gifu” (岐阜), employing characters connecting him to the Qishan (岐山) of the Kings of Wen & Wu of Zhou, and to Qufu (曲阜), the home of Confucius, and of the Duke of Zhou. It would seem that the origin of the term tenshu is often associated with Gifu, but Erdmann points out that there was no tower keep at Gifu when Nobunaga first renamed it that, and that the term tenshu in fact only came into more widespread usage later.
In the 1570s, Nobunaga also began to employ a seal reading Tenka fubu (天下布武), which might be translated many ways, but which Erdmann, quoting another eminent scholar, translates as “overspread the realm with military might.” A rather awkward translation, in my humble opinion, but the important part is the use of the term tenka, meaning “All Under Heaven,” or, simply, the Realm. By invoking “Heaven,” he recalls connections to Tentô (天道, C: tian dao), a concept very closely related to the Mandate of Heaven, and to the Chinese concept of the Emperor as the “Son of Heaven” (天子, C: tian zi). Nobunaga further pushes his association with rightful rule by having the Imperial era name changed in 1573 to Tenshô (天正), meaning “Right with Heaven.”
Sakugen Shûryô (策彦周良, 1501-1579), apparently the last Japanese ambassador to Ming China, is attributed with coining the term tenshu (天主, “Heavenly Master” or “Master of Heaven” – note the different characters), to refer to a Buddhist temple at the foot of a mountain. This term was employed at Sakamoto Castle in 1573. Erdmann argues there is a connection to be drawn between the tenshu (天主) at the bottom of a mountain, and the tenshu (天守) at the top.
I wouldn’t be surprised if there’s a lot more scholarship & debate out there on the origins, and meaning, of the term, but, for me, this was all quite new, and quite interesting.
Mark Erdmann is a PhD student at Harvard and, near as I can tell, has yet to publish anything. A shame, considering how fascinating his presentation was. I eagerly look forward to articles he might publish on these subjects, so as to fill in the gaps, learn more about these fascinating concepts, and have something concrete to cite. Best of luck with your dissertation, sir, and thank you for an excellent presentation.
1) Probably because they’re related etymologically.
2) Someone’s going to tear me a new one if I don’t make a point of being clear that the tenshu, the impressive tower keeps we most associate with Japanese castles, were not the residences; residential buildings were located elsewhere within the castle compound, though, clearly, still nearby somewhere, within the walls.