Mizuya (12): Total Makeover?

In the last post in this series discussing the design process for a kitchen sideboard, or hutch – what the Japanese call a mizuyapost 10 (post 11 relating to acquiring some wrought iron for potential hardware), I felt I had brought the design to a good spot for a pause in the process:

While the hardware left much to be desired at the spot where I had left off, I thought I had come to an arrangement which was satisfactory in terms of the wood bits. With other work on the slate, the sideboard  project was left to brew for a time. That was 10 months ago, and other than gradually acquiring some materials, I haven’t in truth given the project a heck of a lot of further thought. Until recently, that is…. What follows is a fresh look at design issues with this cabinet – forgive me if I ramble at times.

My design inspiration began with a Japanese mizuya, but as I worked through that design it has taken on, in my view at least, some Chinese furniture aspects, with a side trip also to consider Korean cabinets for design ideas.

Sizing Parts.

As mentioned in previous posts, a typical mizuya is constructed with softwood, and then lacquered or stained. The size of the posts on the carcase with this form of Japanese cabinet is governed largely by the interface of the post with the sliding door. I say this while setting aside for the moment aesthetics, and ‘tradition’.

The width of the sliding door tracks, given the presence of two joined door frames running within, and the possibility that one door can be slid by the other in either direction all the way over to the post, needs to be at least 2″ I would say. This track width requires a post wide enough to backstop the two sliding door thicknesses, and accept the track section, with some added width required if the track and post do not meet each other flush at one face. A post minimum size might be on the order of 2.25~2.5 inches. The current design has the post at 3.5″.

The requirement for accommodating the track produces a post, in softwood,  that is somewhat wider than it needs to be for strength alone. When using a tough hardwood such as the bubinga that I have intended from early days in this design, one ends up  with post sections which are probably double the width they need to be in terms of structural integrity.

Now, some allowance needs to be made for joinery when sizing members, so that suggests a slightly thicker post perhaps than the minimum that might work if there were no interrupting joinery. Given that, there is also the matter of the high strength of the bubinga, which would certainly allow for slim post sections if the joinery could be worked out. Slimmer post sections can be employed, and represent a somewhat wiser and economical use of material than thick sections. To move to slimmer posts though means a move away from sliding doors to hinged doors. The sliding doors and attendant tracks can only be slimmed down so far, especially in view of having decent joinery options at the door frame corners. So, not quite sure what direction to take there for the moment.

Hinged or Sliding doors?

I haven’t come across Chinese classical cabinets and shelves with sliding doors- they are all hinged. Sometimes metal hinges are used, other times the outer stile is formed with all-wood pin-like extensions which fit into corresponding sockets in the frame. This design is very pure in terms of its use of wood alone for the mechanism, however the doors cannot open much past 135˚ with that design. A pro and a con. I do kinda like sliding doors overall though – just like the feel of them when well set up.

I could see a way in which the all-wood hinge could be configured to allow for a greater door opening range. I then, however, can’t help but think that themodern Brusso type of L-shaped knife hinges, which are a heck of a lot more discrete might well be a better choice – and that goes as well when they are compared to the much larger metal hinges one usually sees on classic Chinese cabinets. That said, one ‘plus’ to a sliding door is that nothing is swung out into the room, so though one can only access half the cabinet at a time, one can do so without getting out of the way of the door swinging outward.

Breaking Down the Monolith.

There’s another issue that arises though with post sections, and other sections in the carcase for that matter: the weight of the complete cabinet. While portability is not part of the design consideration here, I can see that the current design is going to make for a cabinet on the heavier side, perhaps topping 200 lbs (90kg.). Thick posts and sills, and panels with raised fields, while connoting a pleasing chunkiness, a certain luxuriousness possibly, also make for a certain excessiveness as well. Depends how you look at it, certainly, but I do find myself admiring the combination of really hard tough woods and graceful lightness that are so much a part of classical Chinese furniture. I guess that is the inspiration which is most appealing to me, and the direction I would like to follow.

Thinking more about the idea of a 7′ tall cabinet which weighs a few hundred pounds to begin with, and then is loaded up with plates, bowls, silverware., etc, well, it starts to become really quite heavy. While I’m sure the floor can take it, I’m not sure my back can! The solutions to the portliness are some form of component ‘ultra-slim fast’ or, re-design to make the entire cabinet actually two smaller cabinets stacked up. This is hardly an uncommon solution to the problem for overly-large cabinets, however the Japanese examples of mizuya which are built from two framed cabinets end up with a doubled band of wood framing elements at the point they stack, and this doubled layer has always looked like an ugly thick band around the cabinet to my eyes. Another strike, I suppose, against drawing inspiration from mizuya in terms of a stacked pair of cabinets. There are Chinese cabinets comprised of two smaller cabinets, and I’m giving those a closer look. With both the Chinese and Japanese versions, when the two cabinets are separated, to one degree or another the two components can be used independently, which is an aspect which appeals to me a great deal.

Western hutches have often been built from two smaller assemblies, however in most cases I have seen, when the two assemblies are separated from one another, only the lower case remains functional as an independent cabinet.

The idea of having a larger cabinet which can be broken down into two smaller fully-functional cabinets seems like a wise design considering that 50 or 100 years from now a later owner of the cabinet may use the piece in an entirely different manner. Maybe sideboards will be totally out of fashion, or people have their dining rooms set up quite differently, who knows? If a furniture piece has a certain amount of ‘generic functionality’ built into it, then if can be more readily re-purposed than say, that hand-dovetailed cabinet you built to hold 500 VHS tapes.

Cabinet Construction Ruminations.

Chinese cabinets are generally built as demountable structures, whereas Japanese cabinets generally making use of adhesives, though in the better pieces this was using the starch of rice paste-based glue, which is reversible. These days most Japanese furniture makers seems satisfied to use ‘bondo’ which is a Japanese brand of white aliphatic resin glue, and is not nearly so reversible as the plant- or animal-based glues. Glue was never something to be slathered on the joints anyway – that would indicate poor joinery, full stop. The point is, if the joinery is technically good, just like if a grappler’s pin is technically good in jū-jutsu, then not a lot of added strength should be required to make it work.

It seems like the Japanese idea was to make the entire cabinet quite light so that if there was some need to move it, there was a possibility to lug the entire thing out in one piece, or possibly two pieces. The Chinese idea seems to have been that the cabinet, made from tough elastic woods in the best examples, knocks apart with a mallet and is transported in knocked-down fashion. A cabinet might knock down into 8~10 frame and panel assemblies, for instance.

Another look at the Bottom End.

Another issue has recently come to occupy my thoughts with the design as it stands at present is the ‘post on sill’ aspect. A lot of Japanese cabinets are post on sill, as indeed is the case for some of their wooden architecture. Chinese Ming and Qing cabinets, on the other hand, invariably put the posts right to the ground and fit the horizontals in between the posts. This, it is said, was due to the tile and earthen floors in Chinese houses tending to be damp at times. The Japanese house also had an earthen floor section, however a raised floor characterized most of the space, and the furniture was generally placed on the raised wooden floor. In my case, the floor in my house is wooden, and there is no issue with dampness, however it seems to me that having the cabinet bearing on a sill invites certain problems. For one thing, in such an arrangement the sills need to be at least as wide in section as the posts, so they tend to be kind of chunky. The sill resting on the floor has good air circulation on two sides at best, poor air circulation on the inside, under the floor panel, and no air circulation with the portion which is directly against the floor. Even Japanese timber buildings which use posts on sills rest those sills on the foundation such that there are places for air to circulate under the sills in places, and under the entire floor structure for that matter. Wood which gets an even air circulation on all faces tends to remain stable, whereas wood which gets uneven degrees of air exposure is more likely to have issues with bowing and warping in the long run. And the last thing I would want is a cabinet in which the sill had become uneven in respect to the floor, producing a cabinet that rocked slightly when pushed or when a door was opened perhaps.

A sill means that a good fit to the floor involves a flat floor too. Our house is modern and on a good foundation, so the floor should be decently flat. But 50 years from now – and 50 years is nowhere near the design lifespan I am aiming towards – the cabinet may be in another house without such flat floors. And dealing with an uneven fit of cabinet to floor, whether caused by the sills moving or the floor, is generally accomplished by shimming with folded bits of card stock, etc., which is generally not what you would call an aesthetic ‘plus’ to the desired overall effect, now is it? Knowing that may happen down the line, perhaps there is a way to design from the get-go to forestall or eliminate that outcome altogether.

With a cabinet mounted on four legs/points instead of a framed sill, there remains the possibility that the cabinet will not fit the floor perfectly, however with integral levelers on at least two of the legs, then solving any rocking problem is a piece of cake, and if the cabinet is moved to another spot where the floor is different (a near certainty with a wood-framed floor), the fit can be re-tuned, with at worst the minor hassle of emptying of cabinet contents beforehand.

Now, one can relieve the underside of the sills in such a way that the sills only bear at their ends, directly under the posts, however the trade off there is that you create a narrow crevice under the sill for dust and dirt to accumulate, and it is an area that can’t be easily cleaned without moving the cabinet. If this direction were taken, then it would make sense to have the cabinet on wheels in some fashion so that it could be rolled out of the way for cleaning that area from time to time. The plan for this cabinet however is not to create something akin to a Japanese wheeled chest, or kuruma dansu, so any use of wheels in this case would mean something altogether more discrete – it would likely involve modern heavy duty metal and rubber/plastic ball-bearing swiveling wheels, which represents a move away from ‘traditional’. Ditto for the Brusso type brass knife hinges, but those are inherently simple things, and quite discrete. A cabinet with posts directly on the floor allows for the floor of the cabinet to be, say, 6~8″ off the ground, which means that cleaning underneath it is no big deal, so my thoughts for the design are starting to consider moving things in that direction.


A final consideration involves my desire to employ joinery which is demountable. As a reader pointed out on post 10 of this series, my use of wedged through tenons in places on the existing design is not exactly what one would call ‘demountable’, at least not conveniently. So, I’d like to revisit those areas and see what alternatives might work just as well, all the while also striving to not overload the piece with too many exposed tenons and pegs, ala Greene and Greene. I’m looking to make a piece with, hopefully, more quiet repose than that. I think I’m getting past the point where a concern about making a piece express its joinery, so that the constructional system is ‘clear’ to the viewer, has been a little misguided on my part. The fact is, in this day and age, the vast majority of people viewing a cabinet are not going to get that exposed pegs and joinery even relates to a structural logic, conditioned as they are to view objects primarily in terms of their decoration. So, I am thinking to discard that concern altogether and make a piece so that the joinery is doing what it needs to and is toned down. Try to make the lines of the piece speak, and the wood, and put the construction in the background.

Just because the constructional aspect is being calmed in expression does not mean that the “if you can’t see it who cares?” idea is somehow going to come into play, and the next step is to use biscuits, dowels or some similar system. I believe in the virtues of demountable joinery, and that forms a bedrock of how I design things, and always will.

Conclusions, however tentative.

So many things to think about it seems! I’m definitely at a juncture with the design. I’m inclined to revamp the design, and with that idea comes a feeling that I’ve got to ‘throw it all out’ and start again. But, lest the baby get tossed out with the bathwater, I recognize that there are a few aspects to the cabinet which I will definitely want to retain. Among them, the fabrication in bubinga, the frame and panel system, the fastener- and glue-free drawer design (which I think should work, though it is a design that’s very much ‘experimental’ in nature), and the shippō-gumi lattice pattern of overlapping circles in the upper glassed doors. I also like the small cornice I have made in the ‘ceiling’ of the cabinet. The basic arrangement of drawers and doors below also works in terms of what will go in the cabinet.

With all of the aforementioned elements in seeming flux, and presuming I make some of the changes I’ve been contemplating, the term ‘mizuya’ doesn’t make so much sense. I’m not sure it made sense to call the piece a mizuya earlier on either, but it was a starting point. As the design moves ever further away from the ‘classic’ mizuya pattern, I think I’ll simply be calling this a ‘sideboard’, with what you might call ‘various influences’. I’m less interested in adhering to a given pattern than I am in understanding at least some of the ‘whys’ of that pattern and adopting those aspects that work for my situation, and making a piece which is logical to the materials I am using. Duplicating a pattern generally expressed in a softwood can of course be done, but in the end it makes the most sense, when you consider it in detail, to move away a bit from the pattern when using a much harder and stronger material. Patterns come about for very good reasons, reflecting as they do something which has worked on several levels. It doesn’t mean there might not be some room in there for adding at least a well-considered footnote.

Just talkin’ here. Hopefully you managed to stay awake.

Thanks for coming by the Carpentry Way. In the next post in this series I will show the next stop in the design process. The re-draw has begun – we’ll see what gels and what settles out.

Next post on this topic is a new thread, “Sidling up to Sideboards

10 thoughts on “Mizuya (12): Total Makeover?

  1. “hand dovetailed vhs cab ” – ha! Out of pure curiosity do you plan on building this cab anytime soon? I've been enjoying your carpentry way blog for years, thanks, tAyLoR bLaChLy

  2. Taylor,

    glad you saw the humor in that VHS cabinet – at least how trapped in time it has become. All that work and craftsmanship, and an orphan results in the end. Similarly, pieces such as TV cabinets for smaller TV sets, and Newport Secretary cabinets, etc., are pieces of a certain technological age since passed or superseded. Thinking to design a cabinet which is more adaptable, that 'learns' as Stewart Brand termed it.One can not predict fashions to be sure, but I can't imagine that people will stop needing furniture to store things anytime in the future.

    This sideboard I'm describing here is going to be constructed starting this fall. This process has taken a while, though I'm glad I didn't build the original design. A little water under the bridge and time to reflect is generally a plus for design.

    I've got most of the raw materials now so it is a matter of working the design out once or twice more. This is a design AND build piece, you can rest assured.


  3. I have only recently followed this blog, so forgive me for not reading all the backstory. I enjoy the insight into your design process, so even though lengthy (like this comment), it's informative to see other's point of view. I like oriental influenced pieces, and while I believe you certainly know your stuff joinery wise, your model doesn't have the balance that I'd expect in such a piece. IMHO It stems from dividing the upper case into 2 widths but the lower into 3 + the offset doors. You might find something better if you centered the doors in the lower case and put a bank of narrower drawers on each side, dividing Into an even number whether they are all equal widths or not. Conversely, a 3piece wide sliding doors upper case could also find balance, but take some more thinking to be functional. The top most section feels a bit heavy to me. I like the sliding doors a lot, and even if not authentic it fits the vibe from my western perspective. I also really like the transition piece in the posts and the Far East look that creates there, although I'm not sure of its function; the “missing” lower rail in the panel adds exotic appeal.

  4. JMAW Works,

    your detailed comment and critique is most appreciated. I think you raise some good points about the design as it stood 10 month's back, and it is most generous of you to do so.

    The question of how to divide the cases up is one I continue to look at in designing this piece. I think the point about the upper and lower cases not relating to one another in how they are divided up on a vertical basis is most insightful. I hadn't thought about that aspect, obviously, though it is also perhaps the case of a preference for 'centered and balanced' over asymmetrical is being expressed here? I think in either case though the vertical division relationship between upper and lower cases is definitely worth further thought on my part. There are various directions one could head….

    The three piece sliding door idea: this results in a wider upper and lower track, or would require some very thin doors – not sure I want to head in that direction. assuming a door panel thickness of 1/4″, the reasonable size for the door frame members becomes 3/4″. I could see a door as thin as 9/16″. The sliding door in three parts is worth further thought.

    As for the jog/transition along the posts about halfway up and its function: The posts are thicker down in their lower sections for strength and in relation to joinery requirements in a few spots. The posts are thinner above in an attempt to lighten the feel of the cabinet up high. Not sure it is succeeding as you mention finding the upper case 'heavy'. I wanted to thin the post out somewhere and I decided to make the transition be above where the 'girts' connect, as a fair amount of material is required at those connections. That joint, in fact, partly determines the post section.

    All that said, I had been looking at the idea of making the transition to a thinner post located within the girt connection, but for those joints. In my revamped design, so far drawn, the middle band of drawers has been removed altogether and there is now a shelf/working surface there instead…

    …which means no more band of girts. The upper and lower cases/tables are separate units also.

    …which means no more middle bank of drawers,

    …which has removed that jog and post thickness transition issue from the picture altogether.

    The matter of the narrow drawers around a central pair of doors: one of the things controlling the drawer widths the need for one about 12″ wide internally for silverware. That probably means a drawer width of 14″ given the drawer design. Add 3/4″ or so for a framing member on the side of the drawer bank, 2 banks of drawers means 14.75″ times 2, = 29.5″. I've got about 50″ total between posts, and with 29.5″ gone for the drawers and frame members, leaves 20.5″ for a central opening. Seems a bit narrow, but it would work with hinged doors. In the current version I'm looking at using sliding doors there instead of hinged.

    I am thinking about what individual items I want to store in the cabinet, at the same time I am wanting the storage compartments to be versatile, 'multi-purpose' so to speak. I want to make sure I can fit the requirements of my dining room stuff. Not entirely sure of all that will be going in the cabinet, and my own use of it over time may change.

    I will give the matter of the vertical division lines of the cabinet(s) more thought. Thanks for the insight once again. I have more to say this comment reply is getting a bit long. I'll perhaps reflect those points in the next blog entry on the topic of designing this cabinet.


  5. Chris

    Glad to see we are thinking about the mizuya again, even though the nomenclature is changing. I would strongly, strongly advise the use of sliding doors. Gravity works evenly on the joints, so they need not be robust. The track tends to keep winding in check. No metal fittings! On a hinged door, the top mortice on the hanging stile is under strain from the moment it is hung. If anyone can, you can combine the delicate proportions of Chinese hardwood furniture with the genius and playful asymmetry of Japanese mizuya.

    Can't wait to see how this unfolds!


  6. Tom,

    thanks for the encouragement. I do like the simplicity of the sliding doors, and your point about the tracks helping to keep them relatively flat is a good one. No metal fittings is also good, and they are easy to remove from the cabinet in a few seconds.

    The symmetry/asymmetry puzzle, finding the sweet spot between Chinese and Japanese approaches is the place I now return to at the present phase of design. And putting two cabinets together in such a way that when separated there is no hint of the mechanism putting them together is proving to be a tough nut to crack. Even deciding where to separate the cabinet masses is uncertain at this point.

    It's funny how it can go from a place of thinking the design is nearly there to a place where almost all of it is in total flux!


  7. Hey Chris,

    Also a big fan of sliding doors and really looking forward to following the design and construction process.

    One thing that I have found is that as much as I like them, sliding doors on the bottom of a cabinet are not nearly as practical as drawers, at least for me. You can bend over or crouch and open up a drawer to see the entire contents and remove or replace items as needed. I once built a cabinet with sliding doors on the bottom and found that, much to my displeasure, I wound up needing to sit on the floor to view and access the contents. Your experiences, obviously, may differ.

  8. Chris
    Can't wait for the build!! Tradition is building it your way!! Make that the tradition. Make it strong with function then the form will come!!You will figure it out!! Good luck!


Leave a Reply to Chris Hall Cancel reply