JAPANESE MINKA XV - FLOOR STRUCTURE 1: EARTHEN FLOORS 1

As covered back at the start of this series, the very earliest dwellings in the Japanese archaeological record had earthen floors (doza-yuka, 土座床). Somewhat later, the raised timber floor (taka-yuka, 高床) appeared, but this latter type never completely supplanted the former; the two co-existed, both as broadly separate streams and literally side-by-side, up until the 20th century, and even today the sunken entry area (genkan, 玄関) of Japanese homes is a vestigial reminder of the earthen-floored doma (土間) utility spaces that were once ubiquitous in minka, even those whose main living areas were raised-floor.

Before we go on, it is important to distinguish the doma from the subject of this post: doza-yuka-sumai (土座床住まい), or ‘earthen-floor living’, where not only utility and work activities but also seated social activities such as eating take place on the doza. The do 土 of doza means earth, and za 座 means ‘sit’ or ‘seat’.

It is reasonable to assume that the ‘modern’ doza-yuka is the direct ancestor of the ancient pit dwellings (tate-ana jūkyo 竪穴住居) of the Jо̄mon era. Broadly speaking, the traditional territory of the doza-yuka up until modern times stretched from the mountainous northern parts of the Kinki/Kansai region (the area of western Honshū encompassing the prefectures of Nara, Wakayama, Kyо̄to, О̄saka, Hyо̄go, and Shiga, and generally taken to include Mie, Fukui, Tokushima, and Tottori prefectures), through the Hokuriku region (the coastal prefectures of north western Honshū, i.e. Fukui, Toyama, Ishikawa, and Niigata), the northern parts of the old Shinshū Province (modern-day Nagano prefecture), to the Tо̄hoku region of northern Honshū, consisting of Akita, Aomori, Fukushima, Iwate, Miyagi, and Yamagata prefectures.

The fact that ‘earthen-floor living’ spread so widely and survived so long, despite obvious shortcomings such as dampness and proximity to vermin, is testament to its chief advantage: it is very effective against the bitterly cold winters experienced by all of the regions listed above. Of course, the inhabitants of earthen-floor dwellings did not sit or sleep directly on the bare earth. In both doza-yuka and taka-yuka dwellings, there is a clear, material differentiation between the doma utility area, with its bare earth floor where people would generally only stand in the course of cooking or other work, and the ‘living’ area, for eating, socialising and sleeping. Whereas in the taka-yuka dwelling this differentiation is marked by the ‘step up’ from the doma onto the timber board or bamboo covered floor of the living area, in the earthen-floored dwelling, with all areas at the same level, the boundary was often delineated with a timber sill, and the living area was differentiated from the doma by putting down layers of woven straw or reed mats called mushiro (莚), on which people could sit. On such a floor there are no drafts from below, and the mushiro are warm, soft, and pleasant underfoot. It was said that they were also very comfortable to sleep on, especially for children and the elderly, and there were many examples of minka where even after other living areas of the dwelling had been ‘upgraded’ to taka-yuka, the bedrooms remained as doza-yuka.

Image showing an earthen-floored doma in the foreground, in the midground the doza-yuka living area, spread with mushiro mats and separated from the doma by a timber sill, and in the background a taka-yuka raised-floor area separated from the doza-yuka by sliding partitions.

The same minka, here showing the doma with its posts set directly on foundation stones.

The persistence of doza-yuka might also be attributable in part to legal constraints: taka-yuka were often subject to the kind of sumptuary laws that were widespread in feudal Japan until they were lifted after the Meiji Restoration. In the Tо̄hoku region, for example, the use of board-laid floors was limited to the the formal room (zashiki, 座敷) of the village ‘officer’ or head-man’s house.

 

JAPANESE MINKA XIV - THE GROUND SILL

Continuing on with our ground-up (literally) examination of the structural systems of the minka, today I would like to build on the post-on-stone method covered in last week’s post to consider a later development: the dodai (土台), or ground sill.

The ground sill (or sill plate, sole plate or ground plate, as it is variously known in English) is the horizontal timber member that sits between the foundation (be that foundation stones, a stone or brick stem wall, etc.) and the posts, and transfers the load of the latter down into the former.

Image showing dodai resting on foundation stones below and supporting posts above.

While the practice of setting each post on its own foundation stone represented a significant improvement over planting the posts directly into the ground, it also has several disadvantages. For one, variability in the height of foundation stones means that the posts are not aligned at their bases and thus the posts will be of variable lengths; given the irregularity of the bearing surface of the stones, the posts are also difficult to set plumb; additionally, the open-grain post ends, though they aren’t in direct contact with the ground, still tend to draw moisture up from the stones, speeding their decay. By the use of a ground sill set on a line of foundation stones, post lengths can be made uniform, post bases can be simply cut square, and posts can be somewhat offset, i.e. placed at locations not directly over foundation stones, with the dodai acting essentially as a beam. While it is true that the dodai must still be worked somewhat so it sits level on the line of foundation stones, the degree of precision required in this isn’t as great as that needed when shaping posts to sit directly onto the stones.

The dodai also acts to tie all the posts together, thus forming a stronger overall structure. The post-to-dodai joint is formed by cutting a through-tenon known as a naga-hozo (長枘) into the post, and opening a corresponding mortise (hozo-ana, 枘穴) in the dodai; the joint can then be pinned with a timber peg (komi-sen, 込み栓), or, in more recent times, a shorter stub-tenon (tan-hozo, 短枘) and blind mortise joint may be used, reinforced by nailing a t-plate to the outside of the joint.

Three types of post-groundsill joint: on the left, a long nagahozo tenon and through-mortise; in the middle, the same but with the addition of a peg through the tenon; on the right, a stub tenon (tan-hozo) and reinforcing steel t-plate.

Close-up showing the peg (komi-sen) pinning the tenoned post into the mortised dodai.

The gap between the ground and dodai is sometimes filled in with smaller stones known as 差し石 sashi-ishi ‘insert stones’ or 覗き石 nozoki-ishi ‘peep stones’.

Sashi-ishi or nozoki-ishi used to infill the gap between the dodai and the ground.

In more recent eras and in more ‘upmarket’ townhouses and the like, foundation stones were replaced with dressed-stone strip foundations known as nunoishi (布石), which provided a continuous, flat support for the dodai.

A dodai (a) bearing on a continuous nuno-ishi stone dressed stone foundation (b).

In modern construction, the dodai rests on top of a reinforced concrete strip stem-wall which forms part of the foundation; stone, concrete, or plastic risers of 20mm or so are used between the foundation and dodai, both to protect the timber against rising damp, and to provide a ventilation gap to the subfloor space.

A modern dodai bearing on a reinforced concrete foundation, and between them a recent innovation: a continuous perforated synthetic strip serving dual purpose as both a damp-proof course and a ‘vent’ providing the necessary airflow to the subfloor space.

Given the proximity of the dodai to the ground, durable rot- and insect-resistant timbers are preferred, especially the cypress species hinoki (檜, Chamaecyparis obtusa) and hiba (檜葉,Thujopsis dolabrata, also known as asunaro アスナロ), heartwood of sugi (杉, Cryptomeria japonica), or Japanese chestnut kuri (栗, Castanea crenata).

One disadvantage of the dodai is that, when it does eventually rot out, it is more trouble to repair or replace than it is to simply cut the bases off individual posts and replace them, while leaving the rest of the post in place, as is done in the case of posts bearing directly on foundation stones. Perhaps because of this, the individual post-bearing foundation stone system survived in many places long after the advent of the dodai, with the addition of a tie (nuki 貫) threaded through mortises cut into the posts, tying them together and giving much of the structural stability of the dodai system without the dodai itself.

Image showing posts founded on individual foundation stones and structurally tied together by the addition of a nuki run through mortises in the posts. Note that the post in the foreground has had its rotten base cut off and replaced - a relatively simple operation which can be done without disturbing any of the other posts.

 

JAPANESE MINKA XIII - FOUNDATIONS 3

In last week’s post on ishiba-date (石場建て), the practice of using foundation stones (礎石, soseki) under timber posts, I noted that these stones are often river stones, used in their natural state without any working or dressing. If this is the case, then a problem immediately becomes apparent: how are the timber post and foundation stone to be accurately mated? Ideally, the full cross-section of the post end must bear fully on the foundation stone, for several reasons: to spread the load transmitted by the post to the stone to the maximum extent, to avoid any stress concentrations and possible localised crushing of the timber; to give the post maximum ‘grip’ on the stone to prevent any sideways movement; and to eliminate any gaps or depressions where water could enter, remain, and eventually rot out the post base. Given that timber is easier to work than stone, it makes sense to have the timber conform to the stone and not the other way around. Additionally, the work can be done by a carpenter, who is already required to build the house, whereas working stone requires a mason, i.e. bringing in an extra trade.

The task of matching the timber post end to the surface of the foundation stone might sound simple, but giving it some further thought makes it clear that it isn’t so straightforward. The post will sit on a convex section of the stone, so the post end must be made concave, and not in one but in two dimensions, i.e. a compound curve must be formed into the post end. The means by which this is achieved, with only a compass and templates, is quite ingenious.

First (B-1), a centre point is marked on the stone, two axis lines are drawn (Japanese carpenters traditionally use India ink for marking, not chalk or pencil) through this point at right angles to one another, and the ends of the axes are marked with the four cardinal points of north, south, east and west (1).

Then a template made of a thin veneer of hinoki (Japanese cypress, Chamaecyparis obtusa) is placed on the stone to align with one of the axis lines (2) and an inked compass is used to transfer the profile of the stone along this axis to the template (3).

The ‘transfer template’ is cut along the marked line to create a concave cutout (5). This concave template is then placed against another template (6) and the cut line is traced onto this ‘final template’, which is then cut (7), giving a convex cutout that describes the profile required through the centre of the post end, along one axis.

The whole process is then repeated for the other axis, giving two ‘final’ templates.

Next (B-2), the square post, with the centreline of each of its sides marked, is placed onto the foundation stone so that these centreline markings align with the two axis lines on the stone (1). A compass is used to mark the profile of the foundation stone at each face of the post onto these faces (2). The end of the post is cut square close to the profile lines (3), and a chisel is then used to remove the final material from the four faces up to the profile lines (4).

Then the first concave centreline template is placed against the post end and material is carefully chiselled out from the ‘interior’ of the post end until its profile along the relevant axis matches that of the template (5). The procedure is then repeated for the other axis using the other template (6).

Picture (7) shows a dowel inserted into holes drilled into the centrepoints of the foundation stone and post end. This is to accurately locate the post on the stone and to ensures that the post doesn’t shift off centre during construction; it likely wouldn’t be of much structural use in an earthquake.

The same procedure is used for round posts, except that four axis lines and four final templates are required instead of just two (B-3).

A carpenter, inkpot in hand, using a split-bamboo compass to transfer the profile of the foundation stone to one of the faces of the post.

 

JAPANESE MINKA XII - FOUNDATIONS 2

Last week’s post covered the most primitive method of foundation used in minka, the horidate-bashira method, in which timber posts are set directly into shallow excavations. It also briefly touched on a later improvement over that method: ishiba-date (石場建て), the practice of using foundation stones (礎石, soseki) under the timber posts. This week I would like to look at this method in more detail.

In many cases, the foundation stones used in minka are used in their natural state, without any working or dressing. Typically these are large, attractive river stones known as tama-ishi (玉石) or gorotaishi (ごろた石), whose edges have been rounded smooth by centuries of water action.

Natural, unworked river stones used as post foundations in a minka.

Another example, here below a raised floor.

On important buildings such as temples, the head or ‘column seat’ of the stone, that part visible above ground, would be finely worked into a circular pad, resulting in a composition very similar in appearance to the base of one of the simpler orders of classical Greek or Roman columns. With the passing of time, the soseki of abandoned temples, known as ‘temple stones’ (伽藍石, garan-ishi), became particularly prized for use in landscape gardens. Worked soseki can also occasionally be found amongst more recent and ‘high-end’ minka, where they are called ‘shoe stones’ (沓石, kutsu-ishi).

A foundation stone, presumably in a temple, with its ‘column seat' worked into a disc shape.

In order to lay the foundation stone, the first stage of foundation construction is no different for minka than it is for concrete foundations in modern buildings: the ground is excavated down to the depth of the bearing layer (jiban 地盤 or jiyama 地山), i.e. the level at which the soil is deemed hard enough or well-structured enough to support the weight of the building. This stage is known as ne-giri (根伐り) or ‘root cutting’. In minka, foundation construction in general is called chigyou (地形); isolated pad footings for individual posts are called tsubo-gata-gyou (壺型形) or ‘pot-form’ footings, and strip footings are known as nuno-chigyou (布地形) or ‘bolt-form’ (literally ‘cloth-form’) footings.

The foundation stone does not bear directly on the soil at the base of the excavation: a layer of large, split stones known as wariguri-ishi (割栗石) are first laid in the pit, oriented in a standing position, i.e. with their long axis vertical (koba-date, 小端建) so that their pointy ends penetrate into the bearing layer. These stones are then usually covered by a layer of sharp gravel. Again, this practice is strikingly similar to that followed in modern concrete foundation construction.

Various methods of founding stones on wariguri-ishi. Top: an isolated timber post on a foundation stone. Bottom left: a continuous timber groundsill on intermittent foundation stones. Bottom right: a continuous timber groundsill on a continuous dressed stone strip footing.

Compaction (chizuki 地搗き ‘earth pounding’ or touzuki 胴搗き ‘trunk pounding’) is achieved by the use of various implements: at the smaller scale ranging from a simple disc-shaped ‘mortar stone’ (usu-jou no ishi 臼状の石) with ropes tied around it, called a ‘turtle pounder’ (kame-no-ko-zuki 亀の子搗き), ‘flat turtle’ (hira-game平亀) or ‘flat octopus’ (平蛸 hira-dako); or a hard timber ‘trunk’ (tou 胴) with two or four wooden handles, for use by as many men, called variously an ‘octopus trunk pounder’ (tako-tou-tsuki 蛸胴搗き) ‘small octopus’ (ko-dako 子蛸 or ‘big octopus’ ou-dako 大蛸) On larger projects, a method known as yoitomake (ヨイトマケ) was employed: either ‘oar trunk pounding’ (yaguratou-tsuki 櫓胴搗き) or ‘pole trunk pounding’ (shinboutou-tsuki 真棒胴搗き), where a large timber trunk is suspended from a tripod or scaffold by means of pulleys and ropes.

An illustration of the various implements used for foundation compaction.

The shinboutou-tsuki method required the participation of the whole village: the villagers would raise the trunk by pulling on the ropes, then release the ropes in unison, dropping it into the hole. Naturally work songs and chants arose to aid the villagers in the co-ordination of their actions and to relieve the monotony of the work; these songs and chants show great variation across the different regions of the country.

Approximately 40 villagers engaged in ‘pole trunk pounding’ or shinboutou-tsuki of a stone foundation.