Chris’s Tea (mini) Castle Restoration: Asagi Shikkui

Typically during work, I would not get a chance to document the steps taken to remove and refinish plaster on a wall, since we’re moving quick and it’s not my place.  That is one reason why I am so thankful that my friend Chris bought the little house he had been renting for the last 12 (14?) years.  It’s an old gem.  Probably near or past 100 years standing, it was certainly a tea house at its initial creation.  It was certainly designed by the same person who designed Moan, a Kyoto destination tea house on the top of Yoshida-yama, a hill which asserts its presence in the east of Kyoto valley.  The roof design is complex, and mind-blowing.  A hop skip and a jump away from the Philosophers Walk by Ginkaku-ji (the Silver Pavillion) and also to Kyoto University, had Chris not taken the leap to purchase this ramshackle tea house, the owner would have very certainly seen to its demise.


Every now and then I’d run into Chris by the river and he would tell me about his place and ask if I could come by to assess his walls, which he wants to replaster.  I was so excited for him when I finally did make my way this past April.  It’s a teeny thing.  It’s so precious!  It’s in need of so much love, and here he is committing to give it.  Thanks to Chris, it will be beautiful again.  It is an honor to be a part of the restoration.  May it provide joy for another hundred years.

* * * * * * * * * * * * * * * * * * * *

Plasters used:



Aratsuchi: heavy clay-rice straw for patching on bamboo lathing (makes arakabe)

Mizuzuri:  Clay slip.  Mostly earth (no additional sand) and some rice straw. Very wet mix. (not pictured)

Nakatsuchi: brown coat mix, more sandy, shorter rice straw fibers (makes nakanuri)

Sunabai: lime, seaweed powder, manila hemp fibers, fine sand (0.3mm), used as a primer before the lime-based finish

Asagi shikkui:  shikkui (lime, seaweed, fine hemp fibers) with a touch of blue clay (Asagi) for finish.

* * * * * * * * * * * * * * * * * * * *

This blue tea house.  I’ve seen more than a few tea houses, but this is the first time I’ve seen one that is plastered in blue on the outside and inside.  (Typically, the inside of a tea house would be an earthen finish, as would be the outside, so this place was special from the get.)  The blue is a clay, called Asagi.  I’m fairly certain that the worn out surface is Asagi with a touch of shikkui.  The application is thin, and not as lasting as lime.  This may have been replastered once since the original build, or not.   It was so weathered that it was powdering off at the touch.


Asagi is hard to come by, and quite expensive.  As fate would have it, the company I’m with, Tanaka Akiyoshi Sakan, did a job on a temple perimeter wall which left a vat of material which was shikkui with a touch of Asagi.  Tanaka oyakata donated the material to Chris’s project, which was extremely kind and super duper lucky for us.  I had already been given dried nakatsuchi from Sato oyakata, who I worked for between last November and March this year.  His support in my education and growth has been phenomenal, and I am indebted to him for so much. (Sato Sakan is also where Tanaka oyakata trained, spending his initial ten years in this craft there.)

Phase One: Removal and prep for brown coat

Step 1: Remove



The drop below the wall was long and onto a neighborhood road.  Breathing in the dust of old plaster is not particularly pleasant, nor is creating a dusty workplace, so walls were wetted before scraping.  I held a dustpan under my scraping hand so as not to drop below, with a collection bag nearby.  There were a lot of hemp fibers in the mix.  It is necessary to get all the old finish off that’s patchy, but if it’s only vaguely coloring the brown coat beneath, it can stay.  I aimed to scrape as evenly as possible, and identify areas where the plaster was floating, removing just as much as minimally necessary.  Around the edges, this can often reach the bamboo lathing.

Step 2: Wash.  Bucket of water with horsehair brush (or equivalently soft brush) works well. (The photo above is right after the wash step, hence the darkened wall.)

Step 3: Patchwork

This is where aratsuchi comes in.  Ordinarily, you would let this application completely dry.  Here, time constraints meant only letting it dry til the next day.  Wet the edges adjacent to where patching will happen.  For better grip of the next layer, notching is necessary after aratsuchi is applied.



Part of the plaster had a severe crack, and proved to be floating.  Underneath was a nuki.  (Nuki are thin wooden boards, maybe up to 1cm thick, and up to 8cm wide.  Often they are included just horizontally, but here also supporting the walls vertically.)  There was no nuki-buse, a process of laying mesh to prevent cracks where board meets earth.  I really don’t understand how such a seemingly high-class structure didn’t get this treatment to begin with.  I’ve seen a lot of old walls revealing hemp fabric reinforcement (nuki-buse) on these parts.  I did not want to remove the whole wall to perform this step.  I did the best I could think of, in a situation quite less than ideal.



Step 4: Clay slip (mizuzuri) / mesh where necessary

No photos of this, but the consistency of the plaster is quite similar to clay slip, plus some soft straw.  The idea is to make a layer that absolutely binds the new plaster to the old.  You severely wet down the old plaster, where your spray water (or brush water) is basically cascading down the wall.  Only do small sections at a time, because you want your substrate really really wet.  Use a small trowel, to get as much pressure as you can (reasonably without breaking the wall underneath of course) to really rub in the new plaster to the wet old substrate.  This might be considered the most important step.  (Although I’m not exactly sure which is NOT the most important step : )

This is also where mesh goes over any cracks.  In this case, I also meshed some areas where the aratsuchi patch was long/large, because of its shrinkage factor.

Phase Two: Brown coat (nakanuri)  [warning:  I skipped chiri-mawari / go-around and soko-ume / filler coat due to time constraints.  I have commented on why that was an unfortunate move at the end of the post]

Much of Japanese architecture aesthetically requires reveal of posts, beams, and any structural wooden elements that make up the edges around the sections of walls.  In the case of these walls, the reveal on the sections after removing just the finish left me room to plaster another brown coat and finish, and still have the requisite reveal (called “chiri“).  This is not always the case, and often you have to remove the previous brown coat for appropriate restoration.  In fact, now that I think about it, I don’t think these walls had ever been replastered except for an odd patch here and there.   (Plaster archeology can be mysterious and revealing, at times confounding, at times wondrous.)

Step 1: Taping — makes for easier cleaning and doubles as an ink line

The size particles of your nakanuri mix determines how thick your brown coat will apply.  In this case the plaster materials had gone through a 3mm sieve.  Ideally, this would mean a 5-6mm brown coat, as the plaster goes on in two passes.  However, in old buildings, foundations move and wood moves.  The wall shifts slightly over decades, but the bamboo lattice framework is still solid.  The best thing to do for an aesthetic result is to tape according to reveal.  This means that some areas will get a thicker application than others.  Where the application will be thicker than 5mm (in this case), it is necessary to do a filler pass before the standard two-pass process.  At times, a part of the wall is bulging and only allows for one pass, in which case that part is skipped at the first pass and plastered only on the second pass.  (aside — Japanese tape is so varied and well designed, in degrees of stickiness for all kinds of situations.  Anyone want to do an export/import business?)

higeko and ready for nakanuri

Walls were taped and higeko-d here on day one, so nakanuri could begin right away on day two. From this photo, Aratsuchi patches got another day to dry. Not ideal, not absolutely horrible.

Step 2: Higeko at corners

Tanaka Sakan likes to do at least “sumi-san (corner-three)” with higeko.  Higeko are the little brad nails with hemp fiber tied in lengths.  These are nailed into the wood at the level of substrate, close enough that the threads, spread in a V-shape from the nail, can overlap the thread of the next higeko, forming an “X.”  On site, they embed these threads into the first pass of the chiri-mawari, or go-around — a sandy mix beveled around the edges.  At this project, I did not have the time to do that step, so instead I chose to embed the threads after the first pass of the nakanuri, or brown coat.  Also, I did not have enough higeko, so some areas only got one or two at the corner.  Also, I used higeko along edges where excessive aratsuchi was needed, as a measure against cracks where shrinkage was more likely to occur.

Step 3: Plaster nakanuri brown coat

Step 4: Clean edges

Step 5: When walls are leather hard, pass over surface with a trowel that is not jigane made of iron (hanyaki works well, forged with a little steel)  This helps catch any minor undulations, which are slight enough to be worked out by passing over the area a few times.  This ensures a flat surface for the finish pass.



Phase Three : Finish

Step 1: Tape

Shikkui goes on in two passes, after one pass of sunabai.  All said, the thickness of this combined process amounts to about 1.5mm (about 1/16 of an inch).  This material, with the splash of clay, is a bit of a different creature.  I found I was better off taping to 2mm, because at 1.5mm plaster was still covering the tape, causing me more work fixing edges after tape removal.  Because the plaster has some clay, I was advised to hit it only once after evening, rather than my customary three times, which would compress it a little more.  Less compression = thicker application!

Step 2: Sunabai 



This material is scraped on thin with a stiffish trowel.  After covering the surface, you go back and even out the moisture in the layer. The weather (aka the time of year) determines the amount of seaweed in the mix.  Hotter climate makes a faster drying time, so you want a little more seaweed glue.  The weather and the angle of the sun affects the wicking of the moisture, as does the substrate, so how you work the surface changes slightly in each situation.   Sometimes you have to step away to let moisture wick out.  The goal is to have a less-than-wet-but-definitely-not-dry, evenly moist surface with zero dry-out spots (which are visually noticeable because they turn bright white). Once evening the layer begins, this is a “fast and furious” step, evening out the moisture, scraping where the material has extra “bulk” and moving it to where it could use more. It’s good to clean the tape of sunabai when done.

Step 3: Two-pass shikkui

Considering how thin this layer ultimately becomes, I’m always surprised at the thickness of the application each pass needs to pull it off.  It requires a well-worn jigane (iron) trowel, to prevent the trowel from catching into the material.  It’s such a pleasure to work with this combination of material and trowel.



The first and second passes go on in perpendicular directions from each other.  This is to help the hemp fibers in the plaster cross over each other and make a stronger finish.  Lately I’ve taken to plastering the first pass vertically, and the second pass horizontally.  (In the past I went horizontal, then vertical.)



After the two passes are applied, it’s time to even irregularities, filling all the cat faces and making sure you’re not leaving any holes or high spots.  Since this is shikkui with a touch of clay, working it too much will cause variations in color.  I was advised to level everything out, then just hit it once with a finish trowel.  I think what I ended up doing was:

After applying the second pass,  I went over the surface vertically-ish from left to right, followed by horizontally-ish from top to bottom, to level out the variations of the plaster.  The “ish” is because there’s a need to go diagonal in places as well, dictated to you by the surface as you move the trowel.  Below is a video intended to give you a sense of that bit of the process.  If you were watching someone who has done this longer, you would see more controlled finger positions, faster movements.  I’m still in training, and can pull it off, but I do wish I could show you close up of someone who’s been at it longer, for a better example.   What you can see is where the trowel passes over the same spot multiple times, the “spot” (aka cat face) goes from visible to invisible (aka, surface variations are evened out) :


heap of fibers

heap o fiber clumps removed from material

Sometimes, moving plaster around isn’t enough to fill tiny areas missing plaster.  Tanaka Oyakata gave me a great tip about using the plaster that slightly cakes to the back edge of the trowel to fill in any pocks on the wall.  This is preferable to using un-used plaster because the amount of moisture in the material is similar to that which is now on the wall.  If you used plaster off your hawk, it’s comparatively pretty wet, and the slight variation at that spot can prevent an even finish.

By the way, as you go, you run into little hunks of hemp fibers that got knotted.  These have to come out, or they cause trouble.  It’s typical to end up with a little pile on the hawk.


Once all seems well, you step away for a bit to let moisture wick away before switching trowels for the final hits.  I’ve become accustomed to using a 0.4 stainless/honyaki? trowel (it’s not stainless, cuz it rusts if it’s not cared for, but it’s not honyaki cuz it’s too stainless…I don’t know what it is).  At Shikkui Asahara I used this trowel, but at Tanaka everyone uses honyaki for the final pass on shikkui, and I need to do the same.  I’m on my way to figuring that out.  Here, I stuck with the tool I was accustomed to, since I hadn’t applied a shikkui+clay finish before.  Using the 0.4 to go over the surface from left to right, then right to left, methodically from top to bottom.  Then, remove tape, fix edges.  Then, a final stroke over just from right to left, methodically from top to bottom.  Done.

dried and not

In the photo above, you can see the surface varies in color where the trowel has “hit” and “not hit.”  This is because the surface is not entirely even, with the trowel missing the slightly lower spots.  This indicates that I have a lot of room for improvement.

For starters, I know that I didn’t do all the steps to build the nakanuri up.  I did not have the time to do soko-ume.  Soko-ume is done after chiri-mawari (the go-around).  If your nakanuri is designed/mixed to go on 5mm, then the soko-ume, done with basically the same nakanuri material over the aratsuchi, is applied as needed so that the nakanuri goes on evenly, and then shrinks evenly as water evaporates.  Because I didn’t do the “soko-ume” filler layer, the variations in depth of the nakanuri layer created undulations upon drying, because thickly applied areas shrank more than thinly applied areas.  I could tell when I was applying the sunabai over the nakanuri that my substrate was not flat, and I understood then why soko-ume is very important, and always performed.  Had I had the time and materials to do everything “right” (chiri-mawari plus soko-ume) I would have gotten a better substrate for the finish, hence, a better result in the end.

For here, at this project, under these circumstances, I am not unsatisfied with the result, and Chris is super happy, which is nice. There will be slight undulations, visible if you stare, far far above the road where people and cars commute.

The plastering on this house will be an ongoing project as I can make time and as Chris can plug away at areas he feels confident he can work on, like the arakabe, and anywhere else he wants to play with.   I’m glad to have been “forced” to do the back of the house, the part least seen, first.  Now I know I won’t be skipping any steps to make the front finish a little prettier.