Visualization

I’ve been working on some tools to visualize data and see if the patterns allow for some new insights, or at least to confirm anecdotal evidence or gut level knowledge.

This below is a chart of sword lengths vs. year of production for all Juyo Token, Tokubetsu Juyo Token and Juyo Bijutsuhin swords. I cut the length off at 150 cm, as there are a scattering of outlier swords (odachi and onaginata) that will cause the chart to compress.

Anyway you should be able to look at this image and draw a few conclusions. What do you see?

Analysis after the break…

The first thing we see is some vertical banding. This is because I’ve had to estimate the year of production for a lot of things, like Ko-Bizen blades, that have no dates. Over time I’ll refine it further but it means going through and reading about 20,000 documents (not an exaggeration). Where swords are dated, they appear at the correct year and where they are not, they are scoped down to either NBTHK estimates or the generally accepted work period of the smith.

Past that there is a thin zone in the middle of the data, and we can see this corresponds to the Muromachi period.

The Muromachi period is kind of the dark ages of Nihonto, where mass production took over from high level manufacture.

It’s not completely without skill, rather we can see some data clustered right around 1500. This is the time where we have some master smiths working in Bizen, along with Mino smiths like Kanemoto and Kanesada and of course Muramasa. So we have kind of an island in the upper part of this zone. There are only very few blades that come up as Tokubetsu Juyo or Jubi in this period though.

At the left side of the thin zone there is one vertical band. This is the beginning of Oei and we can see that it plugs right into the Nanbokucho period that comes before and it comes before the slide into obscurity of the main part of the Muromachi. 

This topic came up recently when I was asked about the Oei smiths and the talent level.

We use the general periods of Heian, Kamakura, Nanbokucho, Muromachi, and then Edo (Shinto, Shinshinto) to subdivide most of sword production.

We can see from this data though that:

  1. Muromachi is simply too long a period of time to make generalized conclusions.
  2. Oei is more closely associated in time with the Nanbokucho period than the main body of Muromachi.
  3. Oei is more closely associated in skill with the Nanbokucho period than the main body of Muromachi.

Similarly to the right of the Muromachi gap we get a very strong signal from around 1600.

This signal is very dark blue which indicates a lot of Juyo Token from this production time. There are also green and red mixed in which indicate that Jubi and Tokuju swords are found here as well.

The Momoyama period has a lot of skilled swordsmiths and this is evident in this visualization. Since the bar has a strong presence vertically it shows skilled manufacture of all lengths of blades (tanto, ko-wakizashi, wakizashi, and katana). The smiths of Momoyama were in some cases copying earlier smiths such as Shizu and Sadamune, and so we see this in their output.

As we move to the right though we see tanto vanishing. This says that high quality work is not being done anymore in the tanto lengths. We do see another signal here though.

If you match this over the grid you can see that this island doesn’t appear so strongly in any other period. If you check the lengths what you see are blades from 48 to 60 cm, they are full sized wakizashi and the time period is around 1650-1700.

What this shows is the historical time frame where Japan has settled down and gotten into business without war. The result is an upwardly mobile merchant class and with that, swordsmiths responded by making excellent weapons. In some cases in this time period, a wakizashi by the same smith may be better than a katana he makes. The reason simply is that the customer for the wakizashi can pay more and so the smith puts in the best work.

Smiths in this time period are Sukehiro, Kotetsu, Terukane, Shinkai and so forth and their work and others like them, is what is causing this island. Around this time cutting tests got popular and I think the reason for that is to try to stir the market for long swords which may have been lax.

The empty zone under this wakizashi island seems to imply that tanto development is simply abandoned for 100 years.

The Shinshinto period appears with a roar after as swordsmiths basically by religion say “what we’re doing sucks” and attempt to replicate work of the past. So you start seeing tanto production again at high quality. But almost nowhere in this zone do we see any Tokuju or Jubi. It is also clear where production cut off at the sword ban and everything went down to a halt.

Going back earlier from the Muromachi nadir we see some interesting patterns with short blades.

These lines can be used to chart stocks too… but what we see are increasing maximums and increasing minimums. This means that for high quality swords, they are getting longer on the short end. This island has one foot in Kamakura and one foot in Nanbokucho and so you see the rise of the ko-wakizashi with this island. Also you see that these short blades are held in high regard as there are strong signals of red and green in there. Those are Tokuju and Jubi swords. But by middle 1450, not so much.

Heading North now on the chart we see its beating heart.

Kamakura is obviously the heart of the operation.

It leans into Nanbokucho and those early blades are the same as what came before them in Kamakura.

The mix of colors shows that it is the core area where we find high level blades. 

Now looking at the full set, we see a thickish zone here in Kamakura and this is because of Kamakura blades shortened by later periods because they no longer could make that great weapon. And to the left, we see a few bands. These are Ko-Bizen, Ko-Yamashiro and Ko-Hoki and Ko-Senjuin. We don’t have a lot left from these blades so the data is thin. It shows if you have one, you have something precious. And the strong red and yellow show that it is rare AND high skilled. Nice place to shop.