Wednesday, April 1, 2009

Great East Temple

The Great East Temple, Tōdai-ji, is located in Nara, a couple hour's train ride from Osaka, Japan. Within the Tōdai-ji compound lies the largest wooden structure under one roof in the world, pictured above. It's called the Daibutsu-den, 大仏殿, which means "Big Buddha Hall". It's called that because inside the walls it houses an enormous bronze casting of a seated Buddha, 16m/52' high and weighing some 500 tons, composed of eight separate pieces. The casting of the statue alone took 3 years, used up the nation's supply of bronze and nearly bankrupted the Japanese economy. Here's an image (courtesy of Tomas Tam) of the Daibutsu:

Following a period of great social and political instability in Japan, connected to both warfare, disasters (earthquakes, typhoons, poor crops) and epidemics, the Emperor Shōmu issued an edict in 741 AD to promote the construction of provincial temples. This, it was hoped, would help bring about order. The complex of Tōdai-ji was intended to be the headquarters for not only for Buddhism in the province of Yamato, but the central facility for the six schools of Buddhism flourishing in Japan at that time. All of these sects were from the Vinaya lineage, which ultimately died out in Japan.

After years of construction, and the conscription of some 2,600,000 workers (at peak, 1000 chefs were employed for providing food to the carpenters), the 'eye-opening' ceremony was held in the year 752, with some 10,000 people in attendance.

The years have not been kind to the Daibutsu-den, or the rest of the compound. Originally, the Daibutsu-den was flanked by a pair of pagodas, each some 100m tall, at that time the second tallest structures in the world after the Great Pyramid at Giza. An earthquake destroyed them (which is unusual in itself since pagodas are among the most stable structures in seismic events). Further, warfare resulted in the complex being torched on two separate occasions. The buildings have been rebuilt many times. In fact, the current version (principally erected in 1709), pictured at the top, is about 3/5 the size of the original. Here's a sketch of the original:

You can see that the original had 9 bays in the core structure, or moya, whereas the current version has but 5 bays. The lower eaves on this structure project nearly 10 meters (33') from the walls.

The building survives today largely as a result of the late 19th century rebuilding carried out by the Meiji administration. By that point in time, the building was in danger of structural failure. The huge upper roof was collapsing under the sheer weight of the roof tiles, some 2000 tons. Here's a Meiji-period photograph showing the decaying structure:

You can readily see the 'drunken' line of the eaves, and the fact that posts had been inserted at each hip to prop up the roof from what was by then certain collapse. The bracket complexes at the corners were bent nearly 20˚ from the horizontal, and to further compound the problems that beset the now unstable roof, rainwater had begun to seep in at various points thus causing accelerated decay. Here's a drawing showing the same view as in the above photograph and in which the sagging hips are more clearly seen:

The repairs and insertion of struts to support the roof, done around 1882, were by and large unsuccessful. Part of the problem too was that carpenters skilled in such large reconstruction projects were no longer to be found, since no buildings of the scale of the Daibutsu-den had been undertaken for generations.

This is a cross-section of the structure of the Daibutsu-den, showing clearly how it was principally framed:

It is readily apparent - and this may be surprising to some readers - that diagonal bracing is employed in Japanese traditional architecture, though in a discrete manner, placing the diagonal elements where they cannot be seen so as not to interfere with the preferred visual continuity of horizontal and vertical lines.

When it was rebuilt, the Japanese brought a different approach to the process than we see in the West these days. Timber buildings in the West that have aged and deformed to the point where eave lines are wavy, walls are bulging, and so forth, are often restored in a manner that might be best described as the 'frozen in time' approach. Instead of correcting the problems that have led to the deformations, the major thrust of work is to stabilize things so they do not worsen. Many buildings can be seen in England and Germany that are like this, and many people in fact consider the bowed surfaces and sagging roof-lines as something 'quaint'.

The Japanese do not take this approach. They completely dismantle and rebuild the structure, making correction to the structural problems, and in many cases making changes in the outward appearance as well. They try to keep as much of the original structural material as possible.

Since these structures are rebuilt on a frequency of 50~150 years, it is often the case with many that the sum of the various rebuilds over the centuries has resulted in the structure and appearance of the building having undergone extensive transformation from the original construction.

When such structures are dismantled, it is very often the case that considerable detective work is needed to discover how the building may have been built originally and what changes have been done in the intervening years. In modern times, it has become normal for the restoration specialists to uncover sufficient detail from the de-construction, along with whatever material may be available in the historical record, to come to a conclusion about how the building must have originally been configured. And it is often the case that the rebuild is done to conform to the original design intent as opposed to the most recent version.

Of course, to rebuild the Daibutsu-den in the original 9-bay wide version would have meant increasing it in size some 60%, and that was pretty much off the table from the start, both due to cost concerns and the fact that it was no longer possible to obtain large enough timbers, anywhere in Japan, to reconstruct the building as it used to be. In fact, even at the reduced size of the 1709 version, the architects overseeing the rebuild discovered that they could no longer find adequately sized timbers for some key structural components, most notably the ridgepole and the main support pillars. Probably the same was true in the prior rebuild, thus contributing to the decision to reduce the size.

The time of the late nineteenth century was one of incredible change in Japan, as it raced from the feudal era straight into the 20th century in a matter of 40 years. The Japanese architects who oversaw the rebuild of the Daibutsu-den were trained in both Western architecture as well as engineering. These men performed detailed study on the building, producing western-style scale drawings, along with calculations of mechanical stress within the structure.

As a result, extensive structural changes were made to the frame of the Daibutsu-den: structurally weakened members were replaced with new timbers, as was customary, and in a new innovation, steel bracing was inserted to strengthen the roof truss and the eaves bracket complexes. The posts, formerly one-piece and from enormous trees, were now composite affairs, tied at intervals with steel bands. The ridgepole, for which a suitable replacement could not be found, was changed out in favor of a box truss of English 'Shelton' steel. Here's the rebuilt structure in cross-section:

The intrusion of such a foreign structural element was concealed from view using a dropped ceiling, an element which had been also employed in the former structure. By performing engineering analysis on the roof structure, architects were able to reduce the main roof load by some 12% simply by cutting down on the number of tiles by some 16,000 pieces. This reduction was accomplished by making changes to the configuration of the tiles themselves.

Thus the building survives to this day, a testament to the skill of the re-builders in the late nineteenth century and their wise adoption of western building techniques to solve problems in materials, design, and shortage of skilled labor in such large undertakings - while preserving the appearance of the 'original'. It was an eminently pragmatic solution.

What I hoped to show by this look at one of the most famous wooden structures in Japan, if not the world, is that 'purity' and 'tradition' are not always what they seem. While Tōdai-ji was completed initially in 752, and thus has an age today of 1257 years, its not really accurate to say that the temple is 1257 years old. There has been a temple on that site for 1257 years, but it has not been the same temple as we see today, in detail, size or material. In the current building, there may well be parts in the structure that are several hundred years old, re-used and re-used again with each re-build, but I suspect there are no parts of the structure that are 1257 years old. Even the bronze statue has been recast several times, for various reasons including earthquake damage - the hands on the current statue were cast in the Momoyama Period (1568-1615) and the head was cast in the Edo Period (1615-1867). The head cast in the Edo Period was in fact a replacement for a previous replacement - the head which had fallen off the statue and damaged in 855.

And while the building maintains the appearance of a very traditional Japanese temple, inside the structure are numerous innovations of a western nature. I admire the Japanese for their pragmatism, and both their strong tendency to preserve the old while not being afraid to adopt the new. I apply the pragmatic approach myself, in some ways, in how I approach my woodwork too - I love to use hand tools, and readily use most powered tools. I love the framing square (sashigane), and the pocket calculator as well.

I used to have a very idealistic, rose-tinted view of Japanese carpentry. That view is what I brought with me in 1999 when I visited a temple carpentry company in Osaka called Kongo Gumi, in business since 1400 and responsible for the upkeep and repair of such national treasure structures as the Osaka Castle - it doesn't get much more 'established' and 'traditional' than that company. When I arrived at the shop, instead of what might be imagined, you know - koto music playing in the background, the soughing of the wind in the pines as the carpenters pull gossamer shavings off the wood with their hand planes, etc. - instead I was greeted with the site of a worker running hundreds of blocks, large bracket complex components, through an oscillating spindle sander. All around I saw lots of machines, and the one set of chisels I glimpsed on a bench, did not look, to my eye at least, terribly well sharpened. It kinda blew the image right out of the water for me - and I'm thankful for that. They still of course do use all the traditional tools, but they have no hesitation about employing machinery - even sanding equipment.

I plan to take a look at various intriguing Japanese traditional wooden buildings from time to time on this blog, and share with you the ones I find the most intriguing and inspiring.


  1. Another great post. The last paragraphs are a nice eye-opener.
    I am always looking forward to your next post, giving me food for taught.

  2. I've been in that temple and it really is huge. They were rebuilding the gate when I was there in 90'

  3. Very interesting Chris. Any info on Ginkaku-ji, the Silver Pavilion in Kyoto? It was being re-built when I was there last November. I have some images if you are interested.


  4. @Woodjoint (Charlie),

    Yes it is really huge inside and out - it reminded me of the Vehicle Assembly Building at Kennedy Space Center (which of course is a fair bit taller than the Daibutsu-den).

    @shoji (Craig),

    ah yes, I do plan a post on Ginkaku-ji. Though I have lots of pictures of it, please send me any photos you have of the place.

  5. Hi Chris,
    How do I attach photos with this comment section. I have a folder for you.

  6. Craig,

    I don't believe you can attach them directly tot he 'Comments' section. Can you email them to me? I thought my address was on the top of the page, but in case it isn't:



  7. Kurisu

    Just for the record, I did pass through the hole in the large timber last time I was there at 21.. hard to say at 37 if it would happen again.



All comments are moderated:

Anonymous comments will be deleted

Spam will be deleted

Comments containing links unrelated to blog content will be deleted