The YouTube algorithm has been very good to me lately and led me to a channel for the 竹中大工道具館 (Takenaka Carpentry Tools Museum) which is based in my old home prefecture’s capitol, Kobe. Japanese carpentry is always a fun rabbit hole to dive down and for me it often borders on unbelievable. To see these intricate works of art, houses, shrines, temples, castles and to find out these impeccable structures are sometimes built entirely without nails stretches my imagination beyond what it can fathom.

I watched two expertly produced, subtitled videos over a Friday evening and it inspired me enough to share some of my own thoughts.


The first video takes us inside the modest workshop of Sakae Tategu Kōgei and follows Eiichi Yokota’s evolution from an all-purpose furniture craftsman to creating beautiful and intricate kumiko craftwork applied over shoji screens, room dividers, or even large format decorative installations. The designs resemble creations made on isometric graph paper; a series of latticed triangle shapes, filled in with even more intricate handcrafted pieces. Patterns and wood species are painstakingly combined to create utter masterpieces. The process is intensely manual and intricate down to the last detail. Even the glue is hand made on a piece of wood.

One striking scene (13:41) that remains with me is when Yokota-san looks over sixty or so custom tools he’s made over the years all arranged on a table.

These tools are all developed specifically for kumiko; not quite suited for mass production. Such items are invented and ordered into production by the user. Because of this, each tool involves customized work.

The phrase “not quite suited for mass production” stings me, because it seems like it’d never be profitable. The more I dwell on it, the more I appreciate the resignation that not everything can be automated or mass-produced. Some designs require customized tools, and some tools involve customized work… and that’s a great constraint to understand and abide in.


The second video follows construction of a kiguminoie (timber house), a house built entirely out of traditional Japanese carpentry methods. It’s a peaceful process to watch and the family is grateful for all the work put into their new home. There’s community, there’s ceremony, and a lot of hammering wooden beams with giant cartoon-like mallets. A good bulk of the video is specifically dedicated to discussing the handmade joints used in the process, enough that they get their own title cards.


One part of the video (12:53) explains the advantages of kanawa-tsugi (mortised rabbeted oblique scarf joint) method of joining two boards. This particular joint lends itself well for creating large beams (infrastructure) that is replaceable without dramatically disturbing the whole structure.

Just raising by 3 centimeters will allow you to replace the base plate. The kanawa-tsugi allows for easy replacements without having to raise the building a great deal. However, these days buildings are demolished in less than 30 or so years, so we live in an era where such measures aren’t that necessary. Because we’re living in an era where you demolish the house before things are replaced, we don’t have to do this. But in doing so, the two wood members fit together smoothly; the fit becomes incredibly good.

Here I realized I was not watching a video about carpentry, but a video about sustainability and a video about systems design. Crafting long-lasting, sustainable buildings isn’t truly necessary in an era of disposable houses, but they do it anyways. My brain made the jump to my profession of web development. Meanwhile, the video moves to a discussion of why this system is favorable.

This splicing joint, it’s surprisingly rigid once assembled; so even in an era where we have foundations like these, we still use the kanawa-tsugi. And well, if you consider the inheritance of skills — we can keep this heritage by having young people do the work properly, as they did in the past. If you don’t do this, or you become too practical, you’d only make simple things. It’s very low-tech, his inheritance of skills — using carpentry tools to do manual work. We deliberately put effort into these things; completing work properly in unseen areas.

Two moments stand out to me here. Rigidity (non-brittleness) is desirable in most systems, especially houses, and this provides strength to survive an earthquake. But the process gets tied to a human concern; the passing on an “inheritance of skills”. The assertion that “If… you become too practical, you’d only make simple things” wows me that such a short sentence can pack so much into it. Low-tech, but deliberately. Again, I know they’re talking about framing houses and training new hires, but I can only hear them talking about websites.


When talking about the kama-tsugi (gooseneck joint) method, we’re taken on a history of kama-tsugi over the centuries which itself evolves into a question of whether industrialization has replaced the need for hand hewn joints. Essentially, have machines replaced humans?

[R]ecently, mass production has made fabrication more practical by using machines and producing such joints that are even easier and faster. Currently, these [composite joints] are the most practical—economically as well. That’s probably how it became so simple. The precision is by no means bad; a prefabricated joint is more precise than one made by hand with poor skills. Of course, with craftsmen, some are good; others, not so good. It’s fine if only people with excellent skills make these joints and build houses; but occasionally, you’ll get someone who doesn’t. In that case, there’s a possibility that prefabrication will provide a more uniform, better outcome. With prefabrication, they don’t really have as much variety in types of joints. Most splicing joints are kama-tsugi, and most connecting joints are ari of a dovetail type; it’s even all the same dovetail. In comparison, we use roughly 10 times that in variety—perhaps more than 10 times.

In no way do I think my ability to style an HTML element correlates in anyway to someone who has spent 10,000 hours passing a chisel through wood to achieve the perfect joint, but I find this relatable. I feel this tension in web development. When someone reaches for a heavy prefabricated framework, I wonder why you wouldn’t prefer making some peices yourself. “The precision is by no means bad; a prefabricated joint is more precise than one made by hand with poor skills” weeps with perspective.

Sometimes it’s a question of skill, sometimes it’s about the desired outcome. Prefabrication may produce a more uniform outcome, but it inherently lacks variety in the joints and the design is then limited by the prefabricated pieces. It gets subtitled as “all the same” but the actual borrowed word in Japanese is “one pattern”; it’s all one pattern. As someone who daily deals with patterns, this is familiar.

The knowledge of materials and tools

We tend to throw the term “craftsmanship” around on the web, but it’s nice to see “craftsmanship” in the context of intricate Japanese carpentry where skills take a lifetime to develop and are generations of apprenticeships in the making. Traditions evolving over centuries, not sprint cycles. This is the perspective I need. And for decades Japan has embodied a culture stuck between a rich heritage, but constantly learning to adapt to innovation and the new modern.

In that evolving world, what is the meaning of craftsmanship? I don’t know if I got the answer, but one thing I took away from these videos are that true craftsmen are cognizant of both materials and tools. They know their materials, in this case wood, their relative strengths and which variety is best suited for which situation. They know how much material is waste (cost and bloat) and which cut or join (process) will produce the best outcome. And they know about their tools; to the extent they build their own specialty tools and spend years learning how to expertly wield them. They are proud of their tools and lay them out on the table in an unintentional knolling worthy of a million likes on Instagram. There’s a relationship there between the person and the tools.

So, back to me. Back to the web. If I am to don the mantle of a web craftsman — assuming that’s where the algorithm is leading me — I think I know what I need to do next. I need to know the materials and I need to have a relationship with my tools. Knowing when to leverage prefabrication and knowing when something custom is more suited. That’s probably not a nuance you can operate in when faced with more industrialized processes, but perhaps there’s room for craftsmanship in the modern world.