In the last post we looked at Chinese Classical corner leg tables in their wide variety of forms, and reiterated the point, as made in the first post, that these forms are dependent upon well fitted and precisely made joinery to be successful. This is most particularly true in those tables which have no additional support mechanisms - like giant's arm braces, spandrels, or stretchers (or a combination thereof) other than the corner leg joints. Today we turn our attention to looking at those joints, which are also of highly varied form and of intriguing evolution. Like many things in traditional carpentry, these joints reached their zenith of development quite a while ago (in this case, the 16~17th century), and have since entered a period of technical decline. Or abyss if you prefer.
The three way miter joint is sometimes called the 'Parson's Miter joint' in English, a reference to Parson's tables which have typically have the legs flush to the top. Some others call them a "showcase joint" as they are sometimes used in the framework of glassed display cases.
Tage Frid shows a method to make one using a tablesaw, joint assembled with floating splines and glued together. Well fitted, and his closing comment (p. 203) indicates that he "finally got lucky and made that joint fit". Less than inspiring commentary from a guy writing a book on joinery, and seems to be, when you read it, an example of cutting a joint for the first time so that he can then write a section in his book on how to do it - curious.
Frid's employ of three identical parts connected with floating tenons in fabricating the joint seems to be, or has become, very much the norm in western practice. Every year or two an article will come out in a woodworking magazine, and in virtually all cases what you will see is a three-way joint involving mirror-image table saw cuts, and the use of glued floating tenons, biscuits, or in recent times, Festool Domino® floating tenons:
In fact, it seems to be the case quite often that when you come across articles on this joint they make some mention of how Chinese versions of three-way mitered joints are incredibly complex, but the next comment is something along the lines of (though it's not always explicitly stated): "that's crazy (hard) to do and therefore I'm going to do something much simpler". Hence the article you are to be presented with. I'm not sure why more people aren't motivated to tackle difficult work and would rather seek the easiest way of accomplishing something, but such seems to be the common response. I would have thought more people would have taken their intrigue at the Chinese joint forms and sought to learn how to do them by trying to emulate what they see.
I'm also not sure why the magazines keep repeating what is essentially the same approach in each article that they churn out on this topic - where's all the dazzling innovation we clever Westerners all supposed to be famous for? Yea, I know, a rhetorical question, we'll move on...
Even in China the three-way miter joint was done with varying levels of sophistication over that country's long history of furniture making. The peak of development occurred in the Ming Dynasty (1368~1644) and carried into the early portion of the following dynasty, the Qing (pronounced 'ching', an era which spanned 1644~1912). The Chinese characters for Qing, '清朝' mean daybreak of purity, BTW.
As the eminent Chinese furniture scholar Wang Shixiang notes,
"The remaining examples of Ming and early Qing furniture display, in general, intelligent and well-executed construction. But by the middle to late Qing, cabinetmaking began to lose its integrity: joints became less exacting and complicated, or, in some cases disappeared entirely, and members were simply glued together. When the glue oxidizes, they fall apart."
In relation to the previous post's comments about waisted furniture being constructed by means of one-, two-, or three-piece construction. Ming pieces typically featured larger sections of wood and one-piece waists in those pieces which had waists. By the middle of the Qing, construction involving waists made with separate pieces had became commonplace as cabinetmakers economized on material and paid more attention to appearance than construction. As W. Shixiang notes in his masterwork 2-Volume set, Connoisseurship of Chinese Furniture,
"That Qing construction is generally inferior to Ming construction is clearly evident from its preference for aprons, waists, and additional mouldings made of separate members"
That is, elegant simplicity and robust design is replaced by insubstantial work that plays on effect and carefully masks it true character.
By the mid-Qing period, mass produced furniture began to appear in China, and in the late Qing China was invaded by foreigners, the "Eight Nation Alliance", following the Boxer Rebellion of 1900. The Qing was the last Chinese imperial era, drawing to a close a 2000 year process. During the last portion of the Qing, social cohesion was absent, poverty was widespread and workmanship in all crafts went into serious decline.
Anyway, the history in regards to the three way miter joint forms an interesting development to trace. The Chinese call the three way miter joint baojiansun, written '抱肩榫' The first character means embrace, hold, the second means shoulder, and the third character, quite rare in Japanese I might add, is tenon. So, we have 'embracing-shoulders-tenon', which makes good sense I think.
Whether the table has a high waist or a low one, the form of the joint is essentially the same in the classic Ming form:
Notice that an incorporated sliding dovetail is the primary means of mechanically connecting the apron to the leg, and the joint is further strengthened by a tenon on the apron which enters a triangular mortise pocket on the leg, underneath the miter. The apron is mechanically held against horizontal withdrawal by the sliding dovetail, and vertical loads are mechanically resisted by the mortise and tenon between apron and leg. The two smaller tenons atop the leg insert into corresponding mortises on the underside of the table top frame corner, fixing the table and anchoring the leg against lateral forces, thus completing the trio of primary mechanical interconnections. It's a brilliant joint!
Here's a high-waisted version of the same joint:
Observe that the top part of the leg is extended and resembles an exposed post on the waist. Additionally, a groove is provided on the leg extension to receive a floating panel.
Next one is very similar to the above, except we now have a two-piece waist:
Now, let the devolution begin. In the next joint, the male dovetail on the leg has been replaced by a mortise and floating tenon:
Although this joint is satisfactory, the reliance upon glue has increased and the mechanical interconnection we had with the sliding dovetail, in terms of surface area, has decreased and thus the joint will be a bit weaker. It is however easier and cheaper to make, and that will appeal to some. The tenon on the apron entering the leg remains unchanged.
The next one is also a step down, in an un-waisted table:
Here, the sliding dovetail is retained, which mechanically holds the apron to the leg and self-tightens with loading, however the mortise and tenon is lost. This is obviously weaker, and without the mortise and tenon there is nothing to restrain any tendency on the part of the apron miter tip to bend/warp outwards from the junction.
Next step down - long gone is the sliding dovetail, and now gone as well is the floating tenon:
At least the mortise and tenon remains which gives mechanical resistance to vertical loads on the table top, however only glue keeps the apron from horizontal withdrawal from the leg.
Next, same as above, but for high-waisted construction:
In the above example, at least the tenon on the apron end and triangular mortise on the leg are still hanging in there to provide resistance to vertical loads, but the joint is now utterly reliant upon glue for success.
The next joint retains the tenon, but that tenon has become atrophied, and therefore a weaker joint is the result:
The next step away from the lofty heights of sophistication is this:
The mortise and tenon have now been replaced by an tapered stub tenon, which imparts little strength but does at least retain the apron miter tip from any tendency to warp away from the leg. Here's a color photo of the same joint, showing the less-than-crisp joinery, which looks in places, I dare say, almost like it was done with a hatchet:
In the final stages of this moribund tale of joinery decay, we get this:
This joint obviously has zero mechanical integrity other than the one tenon at the top, and glue is all that holds it together. Without the mortise and tenon, any glue failure will result in the apron sliding down the miter. A degenerate joint. In the late Qing there were some tables of this inferior construction which ended up in the Beijing Palace, believe it or not.
Another type of three way mitered joint is found on some pieces of furniture where a leg or stile is joined by two rails, used on tables, bookshelves and, as one example, on the top of the frames in these display cabinets:
In these cases the leg/stile itself cannot extend through the rails, but must be covered over by the rails, so the form of the three-way mitered joint changes somewhat.
Here's a classic example of that joint:
The leg/stile has two tenons, one short, and one long. The long tenon goes right through one of the rails, while the short tenon is restricted in the other rail by that members horizontally-oriented tenon.
The Chinese term for this form of three-way construction is simianping, written '四面平', meaning, appropriately, 'four faces flat'. The joint itself is called zongjiaosun, which translates as rice dumpling joint. A strange name to be sure - the joint is thought to resemble a rice dumpling in some way. I have looked to see what sort of dumpling it might reference, and this is my best guess:
The rice dumpling joint presents a tidy appearance, however with so many pieces of wood vying for the same space, the joint will be weak unless adequately-large sizes of material are employed. Many pieces employing this joinery reinforce the connection with stretchers or giant's arm braces.
On table tops, the leg tenon should not pierce through the upper surface of the table, and thus a modified version of this joint is employed where both tenons are stubby:
This form of the joint is therefore a bit weaker. With bookcases and cabinets where the frame members connect above eye level, the long tenons on the stiles are better if longer and through -
and can be made more, er, decorative, if extended enough above the surrounding surface to be obvious.
Here's that same joint just shown in eye-searing color:
Lastly, here's a degenerate form of the above rice dumpling joint:
All the tenons are stub tenons, which will give a very clean appearance, at a cost of considerable loss of strength.
Right then, I believe I have given an adequate survey of the varieties of three way miter joint that furniture history serves up for us. I've come across other modern versions of the classical examples described above in this post, but none worthy of much attention, none with anything new or improved or, god forbid, revolutionary to offer, so I'll leave them out of the arena of consideration.
I still haven't delved into the table I introduced in that first post, to explain what is so neat about it, so that is next on the slate. Excuse the crass use of a dangling carrot like that in post one. I'm getting there! We will be looking at that table in some detail. That table has it's own unique solution to the three-way miter, as do I, and after all I needed to set the background for that explanation before plunging into it. I hope that makes good sense.
Thanks for coming by the Carpentry Way today. I hope you'll return for the next installment in this design/build thread for a Ming-inspired dining table. Go to post 4