Fan of the Fan (VII)

In this series I have looked at the geometrical challenges facing a carpenter wishing to employ fan rafters, ōgi-daruki, in a roof structure. We've seen in this series how it is difficult to reconcile the aesthetic considerations of obtaining both a nice pattern of rafter tips at the eave edge, and an even division of rafter spaces when viewing the rafters from underneath. We've seen 4 different Japanese solutions to this problem, the last of which, as detailed in the previous post, provides the cleanest answer to the lay out challenge, in my opinion at least.

Fan rafters as they are seen today tend to associate to, and derive from, temple architecture, and started becoming popular in Japan somewhat late in history, around the 14th~15th century.

There are two distinct patterns of fan raftering employed on temples - either the fan rafters radiate from a central point in the structure, or they proceed parallel to one another until they get near to the hip rafters, at which point they begin their fan.

The type that radiate from a central point, the same design as I used on the recent bell tower design, associate strongly to Zen Buddhism style architecture, or zenshū-yō (禅宗様). If you want to read up a bit more on that architecture and an example of a structure employing fan raftering of that type, I suggest you take a look at a post I did a while back, titled "Peace: Enjoy the Zen" where a six-sided pagoda with fan rafters is featured, Anraku-ji. Fan rafters originating geometrically at a central point are a natural choice for structures which are square or regular polygons in their plan.

The other type of fan raftering, in which the rafters along the eave remain parallel to one another until they approach the hip, is associated to "Big Buddha" style, dai-butsu-yō (大仏様), architecture. For an example, see my post from last year titled "Great East Temple"). This type of fan raftering is called sumi-ōgi-daruki (隅扇垂木), and is the more commonly seen variant as most buildings on the temple grounds are rectangular in plan, not square.

The earliest extant example in Japan of dai-butsu-yō is seen here at Jōdōji (浄土寺) in Nara, constructed in 1192:


Here's a classic example a two-tier eave with sumi-ōgi-daruki:


One more example of that style of raftering, for good luck:


While I stated that the use of fan rafters associates strongly to temple architecture, there are however residences that employ it, especially in the entry foyer, or genkan. Here's one such residence:


The portion of the building which projects out towards the viewer in the above photo contains the formal entry, and that is one area of the house which tends to receive more decorative attention (similarly to the guest room inside the house) than other parts of the structure. Here's a close up of that genkan eave:


Another residential example:


Nice copper gutters and rafter caps!

Even tea house (sukiya) architecture makes use of fan rafters occasionally:


So at this point, after all my droning on, the reader hopefully will have a pretty good idea as to the way the Japanese deal with fan rafters. Both the Zen style and Big Buddha style of temple architecture, however, derive from continental precedents, so now I'd like to turn to China and Korea to see how temples in those countries make use of fan rafters. I suspect there may be some people out there who would have trouble distinguishing Chinese, Korean, and Japanese temple architecture from one another, but believe me, they are as different as English, French and German traditional wooden architecture are from one another, and, similarly, share common points as well.

I'm not as into Chinese and Korean temple architecture as I am Japanese, so the depth of my analysis of those architectural styles and building methods here is necessarily limited. I do not read Chinese or Korean, so I cannot access source material from those cultures, and that is a major limiting factor on my understanding.

With those caveats in mind, let's have a look at a Chinese eave (and I thank R. Wiborg of San Francisco for allowing me access to his materials some 6 years back):


This eave, like virtually all of the examples of Chinese and Korean temples I looked at, is like the sumi-ōgi-daruki variant, in that the rafters do not begin to fan until they get close to the hip rafter. Notice that the composite hip rafter is marked out along its upper section with the positions of the 13 fan rafters.

Here's the plan view:


One observes a difference right away - in the above Chinese method at least, the fan rafters terminate directly onto the hip rafter. You won't see that in Japanese work. Note as well the steep climb of the curve, and how that climb is accomplished - by a scalloped filler board. This board forms then a wedge-shaped space in the eave.

Here's a Korean example showing the visual consequence of that tapered board:


Chinese/Korean fan raftering, which I will choose to lump together as they are very similar to my eyes at least, very commonly employ round rafters. Round rafters are much easier to deal with than rectilinear ones in fan raftering, as no extensive shaping is necessary to get them to conform to the other eave members. The gate house at the Ellison Residence employed pole fan rafters, so I have had some direct experience with that method.

In cases where a two-tier eave is used in China and Korea , a common system for dealing with deep eaves, the upper, or flying rafters will usually be of the rectilinear variety while the base rafters are round:


That was another Korean building - the choice of paint colors and patterns is one giveaway.

One of the things that strikes me about Chinese/Korean employment of fan rafters (not in all cases mind you) is that the rafters nearest the hip appear to be sunk half-way into the hip rafter. You won't see this in Japanese temple architecture, at least not so far as I have seen.

Another continental example, where the fan rafters at the hip are sliced length-wise and directly abutting the hip:


If you trace lines along the lower and upper tiers of that eave, you will see that the alignment of the rafters is not consistent with one another. Again, this is unlike Japanese work.

Another interesting difference relates to the employment of rectilinear section fan rafters in China and Korea: unlike the Japanese work, the sides of the rafter are not kept plumb. In many cases it would appear, the fan rafters are rotated along with the eave perimeter fascia:


Notice too the un-evenness of the fan rafter spacing on both sides of the lower hip. That would be considered a blunder by a Japanese carpenter. This rotation of the rafters is not attractive to my eye and conveys the sense, visually at least, that they could roll over from loading.

More examples of that fan rafter rotation, this one Chinese:


Another:


And another, this one with a curious staggered pattern of rafter spacing between the lower and upper tier:


I'm not, I will say, a fan of that fan. It looks disordered and choppy.

Returning now to the plan view of the fan rafters, Chinese-style:


Notice that the rafter tips for each fan rafter are cut square to the centerline of each rafter - the Japanese carpenters have improved upon this, I would say, by shaping each rafter end cut in a differential manner to achieve a smooth sweep of tips over to the hip.

Now, if the idea, as the previous illustration indicates, is to cram the rafters in one next to another, then the rafters cannot remain whole in section but must be tapered to fit in among their neighbors. The Chinese carpenters have an intriguing method for accomplishing that.

First off, they obtain their parallelogram rafter sections by slicing them out of a blank:


This looks economical, to be sure, however if there were any wood movement as a result of ripping the rafters out of the blank, you would not be able to obtain straight sections. perhaps they account for that, or perhaps they're not so fussy about it.

In the next illustration, for the round section rafters, a marking jig is constructed. This is a simple affair using a pair of notched boards to hold the rafter. On the end of these boards re marked a set of even spaces from a center point representing each rafter:


The bevel gauge is used, I believe, to set the rotation of the rafter relative to its position on the eave curve, given that the rafters rotate with the curve.

Back to those marks for the rafters - they are used with an ink line to snap the taper lines on the sides of the rafters, like so:


Here's the same procedure being employed on a round section rafters:


I am not optimistic that snapping a line from a single position across a round with a flat on it will result in a straight line, but I'll set that aside. The ink line is shifted to different positions for each respective rafter, and the result is the production of differentially tapered rafters which will fit against one another:


I presume the rafters are then fitted in place, adjusting as they go.

I'm not personally so interested in this aspect of Chinese and Korean wooden architecture, for it lacks, in my eyes at least, the sophistication and elegance I find characteristic of Japanese carpentry work. Some will be more drawn to the Chinese/Korean work because of their rusticity, color, and exuberance - of that I'm sure. Some will love the painted decoration. I tend to prefer the beauty of the natural material without augmentation or concealment, anything beyond hand planing and oiling is unnecessary, not to mention harder to maintain.

The extreme curves of the continental roof forms do not appeal to me, and I'm not much of a fan of those types of tile roofs for that matter with simple pipe-section-esque tiles. The Chinese roofs which sweep up to such an exaggerated manner at the eave remind me of photos of women from bygone days revealing their petticoats under their skirts or something like that, and while perhaps titillating or fun for the viewer at first glance, it becomes crass after you see it too many times. I can see however that if that vast majority of other structures in the built environment were drab, the Chinese temple architecture would be by contrast a relief.

The steep reverse pitches of those hip rafters seem to engender places where the roof tiles would be hard-pressed not to leak in certain spots, and any time where an aesthetic drive in architecture (curving that eave up more and more) subsumes considerations of practicality and durability I generally hop off the wagon. There are practical reasons for up-curving a hip, but beyond a certain point the structural argument becomes moot.

And finally, in terms of the fan rafters, their crowding together at the eave, which necessitates the tapering of the rafters, and the lack of attention paid to getting the spacing clean and well-aligned, well, I find to be a shortcoming in the Chinese and Korean buildings.

The Japanese, through working at solving these challenges for a few hundred years, have produced with their fan raftering an eave with the end cuts of the rafters in harmony with one another, cleanly aligned and spaced, and with a smoothly sweeping eave edge, regardless of where the eave curve begins (if there is a curve at all). That is, in my view, one of the unique triumphs of Japanese architecture over its continental antecedents.

Some people might put down the Japanese for 'copying' the work done on the continent, however in my view they have paid the highest compliment to the Chinese buildings of former times in their sincere emulation of the form, and later improvement of fan raftering. And, dear reader, this fan raftering but one narrow aspect of this ancient form of timber construction, and the Japanese have applied their process of continual improvement to all aspects of that form of construction. No stone, let me assure you, is left unturned.

So there is a lot more to see and learn yet in Japanese architecture - that's my conclusion for now. That's why I imitate the Japanese builders - their work is the best example going, in most cases, both in terms of technical sophistication and refined aesthetics. I know, without question, that the Japanese carpenters of the past had considered the details of everything they did in a most thorough manner, and it is only for me put the time in to discover what they did. While for furniture, I look to the Chinese Ming period masters, for architecture, it is the Japanese carpenters who I feel have truly mastered the art of building in wood.