Japanese Gate Typology (14)

Today marks the 600th post on the Carpentry Way!!

This is a series of posts describing the many varieties of Japanese gate, or mon (). If you look at the preceding kanji more closely,  you can think of it as a picture of a pair of swinging doors, reminiscent of saloon doors in a wild west movie. I thought that it would be a nice idea to bring more awareness to a western audience as to the tremendous diversity in forms and types of Japanese gates. This series intends to be a gateway to gates.

Previous posts in this series:

  1. Heijūmon 
  2. Kabukimon 
  3. Kōraimon 
  4. Yakuimon 
  5. Yotsu-ashimon 
  6. Munamon 
  7. Commentary
  8. Uzumimon 
  9. Yaguramon 
  10. Rōmon
  11. Shōrōmon
  12. Taikomon
  13. Nijūmon


Sanmon (三門/山門)

The triple () gate () or the mountain () gate (). I believe the use of the kanji for mountain () in this compound is actually incorrect, however it is commonly seen. At best it is a pseudonym for ‘三門‘.

The term sanmon, written as ‘三門‘ is actually an abbreviation for san-ge-datsu-mon (三解脱門). The word refers to the three spiritual gates one must pass through to reach enlightenment. Thus, a short and sweet translation of san-ge-datsu-mon would be ‘enlightenment gate’. A sanmon is a special type of nijūmon (二重門), as covered in the previous post. Sanmon associate to Buddhist temples.

And what might those three spiritual gates be? They all have their own special names:

  1. kūgedatsumon (空解脱門), usually abbreviated to kūmon ()
  2. musōgedatsumon (無相解脱門), usually abbreviated to musōmon (無相
  3. muganmon (無願門)

The term gedatsu (解脱) is the Japanese rendering of the Sanskrit term ‘moksha‘, which means, to quote wikipedia, 

…the final extrication of the soul or consciousness (purusha) from samsara and the bringing to an end of all the suffering involved in being subject to the cycle of repeated death and rebirth (reincarnation).

In terms of specific meanings associated to the three gates of enlightenment, wrapped up together in the san-ge-datsu-mon, we have kūmon () as the gate of emptinessmusōmon (無相) as the gate of formlessness, and muganmon (無願門) as the gate of inaction. As noted in an earlier post in this series, gates have a large symbolic as well as physical presence in Japanese culture. 

The sanmon gate has no doors, which itself makes it a marker of the boundary between the sacred and the profane, an opening which can never be closed. If someone passes through the gate, he can free himself from three passions ( Ton; “greed”, Shin; “hatred”, Chi; “foolishness”).

The archetype of this gate form is found at a former Zen temple located in Tokyo – all that remains of the temple complex, named Zōjōji, is the gate. It’s a massive structure, at 5×3 bays:


The temple that once occupied these grounds was one of the most important in Japanese history. Six of the fifteen Shoguns who ruled during the Tokugawa era are interred here. The temple was destroyed by US bombing in WWII. Nowadays, a golf practice course and hotel occupy a chunk of the land. Priorities appear to have shifted. The gate above is a reconstruction, and is made largely from concrete. We can probably leave off a closer look at the ‘framing’, eh?

Another imposing sanmon, at Zenkōji:

A glance up at the two eaves:

The sanmon at Chōshō-ji (長勝寺):

From the inside of the compound the two stair are clearly visible. The tightly packed bracket complexes make this gate clearly in the Zenshūyō style:

As you can see, a little repair work was going on when the picture was taken.

Another magnificent sanmon, this one at Tōfuku-ji (東福寺):

From the inside, you can see the ladder to the second story tucked in at the side of the gate:


The posts placed at the corners remind me of a picture from around 1840 of the great Tō-dai-ji in Nara, which had such a heavy tile roof (4000 tons or so), and such deep eaves, that the hips were drooping downward from the weight and need props put in for support. I think the above structure is suffering from a similar situation and that the struts at the corners are later than the original construction. In the case of Tōdai-ji, the roof tiles were re-designed to shed about 20% of their weight, so I wonder if they’ll do the same thing when they rebuild this gate at some future point?

Speaking of Tōdai-ji, the great southern gate, ‘Nandaimon’ as it is called, is in fact a fantastically huge and tall sanmon:

The people under the gate give a good idea as to the scale of the structure. Nandaimon, like several other structures in the compound, is a Japanese National Treasure, and is built in a late 12th century Song Dynasty style.

This next sanmon has a similar enclosed curved skirt as we saw in a previous post in this thread when looking at bell tower gates, or shōrōmon:

The above gate is located at Sōfuku-ji (崇福寺) in Nagasaki, and the gate has been designated as a national treasure. It is not a bell tower. Built by a Chinese monk in 1629, the gate and other buildings in the compound are artifacts of Chinese Ming dynasty architecture.

While many sanmon are imposing, grand gates, the main thing to remember about them is simply that they are the main gate in a temple. Thus, a small temple may have a small sanmon, as in this example, Shōgyō-ji (正行寺) located in Hokkaidō:

You may notice the above gate, while quite pretty, is not grand or imposing at all, and doesn’t even look like a sanmon. You see, even if the gate doesn’t have the normal attributes of a sanmon, which is a type of nijūmon and supposed to have two stories and two roofs, if it is the main gate of the temple then almost anything goes. 

Here’s a bell-tower gate at Jōganji (浄願寺) in Hyōgo prefecture which is called a sanmon:


Here’s a single story gate called a sanmon, at Hōryū-ji (法隆寺):

So, that’s how it goes with classification schemes sometimes – it can be a bit confusing and some gates fit into multiple categories. Some seem to be clearly of a given type based on the structural configuration but get named something else based on where they are located within a temple precinct or how they are used or by whom they are used. We’ll look at another case in the same vein with the next post. Thanks for coming by the Carpentry Way. On to post 15

4 Replies to “Japanese Gate Typology (14)”

  1. Hi Chris,
    Your threads on gates have been a very interesting read as always. So much information organized neatly and concisely almost like a well-written textbook.

    I have some questions regarding restoration work and practices being carried out in East Asia in general (China, Japan, Korea). I know, from my own research, that the architectural practices in these countries places emphasis on wood as the primary building material for most structures. Specifically speaking, I am confused as to why some recent restoration work uses concrete instead of wood? Wouldn't this undermine the spirit of the original building as well as the original craftsmen who have made the originals possible? I have heard that the Shiteno-ji Temple at Osaka uses concrete. This practice is not only evident in Japan. As it seems, China is using materials other than wood to restore some sites such as part of the Daming Palace at Chang'an/Xian.

    I know that proper restoration work using original building techniques is possible using only wood, as evident in Hong Kong's Chi Lin Nunnery (a beautiful Tang Dynasty Chinese Reconstruction using only wood and no iron nails). In my opinion, I'd say that this method of restoration is what makes the architecture and its site worth visiting, even though it is newly reconstructed.

    You see, I have been heavily interested in restoration work and practices for East Asian Architecture for a while specializing in 3D computer rendering. I have studied quite a bit from all the three countries of East Asia. What I find astonishing is that China's Tang Architecture and Korea's Baekje Architecture looks identical to Japanese Architecture (discounting the ones utilizing the hidden roof features.) Take for example the Chi Lin Nunnery in Hong Kong. If I give a fresh-eyed look at it, it gives that “zen” feeling. Furthermore, take for example the Tang painting at Prince Yide's Tomb of an 8th century mural of Chang'an Towers and city gates. If I look closely at the fence as well as the wood that supports the “balcony”, that structural area is almost indistinguishable from the Yakushi Temple Pagoda's balcony structure. This made me wonder what would happen if the Chinese and Koreans chose not to change the way they build. That is what I think is exciting about restoration especially on architectural styles that are considered “endangered”.

    It would be awesome if you can share any other reconstruction that is carried out identically as Chi Lin Nunnery (or at least to some point), I am always interested in finding out more examples of restoration carried out just like how the ancient craftsmen have done.

    Thanks for putting together these articles, they have been very interesting and informative.


  2. Frank,

    thanks for the observations and questions.

    I guess the matter of using concrete, welded steel, etc., as temple frames boils down to cost. I'm not sure though how much of the cost calculation is based on up-front, or whether it factors in the cost of rebuild down the line to the equation. I think most Japanese would prefer a wooden temple, as it seems a bit more connected to nature than the alternates, but I also think a significant number of visitors might not notice that the temple is made from another material besides wood upon passage.

    I think the Japanese would be the ones to preserve wooden building traditions the longest among their neighbors. In China, masonry wall temple buildings of all kinds are commonplace, so I suspect many folks there might not find the idea of a concrete temple building as anything other than a new wondrous thing and welcome it. Same for Korea perhaps.

    A lot of people are more attuned to the meaning of the temple as having a purely physical manifestation of a form – I mean, they have certain conceptions about what a temple building looks like – but are less concerned with what it might be made of.

    I'm glad you, as I do, appreciate wooden temple architecture for its very 'wood-ness'. I'd rather see wood than plaster or welded steel, stone rather than concrete. The right material needs to be put in the right place though.

    I think the connections you are noting between Chinese Tang, Korean Baekje, and some Japanese architectural examples are definitely there. The wide variety of temple and shrine architectural styles and systems you can run across in parts of Japan can be a bit confusing and overwhelming at first, at least I found. In certain places you can see, for example, a near new temple built in 1990, another rebuilt one from around 1950, another next to it from 1900 (a structure that has been rebuilt in the past 25 years), another across the way from 1850, and so on, and years on back, one after the other. It's easy to simply see a jumble of stylistic differences between buildings, and an assortment of framing systems, and not pay as much mind to the chronology of the builds and the differences that accrue between them, for various reasons of course. The Japanese cultural tendency to 'rei-nen dori' (do the same this year as last, maintain tradition), and willingness to spend on it, actually makes the country a fabulous place to see classic examples of old Chinese and Korean architecture, with examples of some types long vanished from the scene on the continent.

    I understand Chinese wooden architecture reached its peak around 1000~1100 AD, and went into some decline, gradually supplanted by masonry building, largely because of a large timber shortage I think. By that point in history, there had already been at least a couple of periods of significant contact between Japan and the continent, with Buddhism and temple-building long established about 300~400 years ago.

    Frank, you sound like you have an interesting background. Are you involved with Wikipedia by any chance?

    And, I will definitely think about future posts on temple reconstruction work – thanks for asking.


  3. Hi Chris, thanks for the awesome reply!

    No, I'm not affiliated with Wikipedia. However I have been studying East Asian Architecture for two years now specializing in classical forms. I read a lot of online sites including your blog to expand my knowledge base. Although I am not a hands on craftsmen, I pride myself in using 3D programs to create renders and models of these architectural types. Furthermore, I am also a natural born artist so I can really tell the difference between most East Asian architectural types. For example, the architectural style of Heian Palace and Heijo Kyo Palace looks like the Daming Palace of China. (Daming Palace, from my research, does not exist anymore due to the fall of the Tang, however great 3D render images done by 3D artist have brought it back to life. Some interesting picture links at the bottom for convenience.)

    There is obviously many more, however I do not wish to clutter the post with links.

    If the reconstruction of this palace can ever be done in the near future using only wood like the originals, I would have to say, in my humble opinion, that it will be the most beautiful palace complex in the world. It would also showcase Tang architectural influences in Asia.

    Korean Baekje Architecture also has seen much revival in the form of new reconstructions. This style is interesting because Baekje Carpenters is said to have designed Horyuji Temple and thus resembling closely with its Korean counterparts. Although due to warfare, Baekje Architectures have mostly disappeared leaving only written records of its previous grandeur.

    I do agree with using wood; most of the time I get frustrated hearing a new reconstruction being carried out using alternate means / materials. I think that this practice is partly because building in concrete is easier to maintain if cost is factored in.

    And yes, I do think a topic on reconstruction would be interesting. Will be looking forward to it.


  4. Hi Chris,

    Was going through a carpentrer's manual pdf layout of a sanmon and a question has bothered me the past couple of days. This question concerns the nature of a high rank 5 bay sanmon.

    The question is, the manual included a PDF layout of a Zen Style (5 bays X 2 bays) Sanmon from top view and it is as what i expected, the ken size (space between the pillars) is wider than than the second floor ken space. When I look at real pictures however, I observe that each bracket approximately holds 8 rafters (tarukiwari), both the top floor and the bottom floor.

    If that is the case, how does the Japanese achieve the wider ken spacing on the bottom bottom level of the sanmon in respect to the top? Do they do uniform scaling on the bottom floor, such as brackets and rafter tip dimensions are uniformly larger than the ones on the second floor? Or is it something else I'm missing on.


Anything to add?

error: Content is protected !!
%d bloggers like this: