Data Visualization and Data Quality

The inestimable Rob MacDougall is running a course on Digital History, and even better, he’s running it more or less publicly! I’m getting all kinds of ideas here. On the other hand, it sometimes raises surprising problems. The unit on Data Visualization includes an assigned reading that looked like something I might use for historiography, David Staley, Computers, Visualization, and History: How New Technology Will Transform Our Understanding of the Past (M.E. Sharpe, 2003). But when I started looking through it, the first ‘data visualization’ presented was an illustration of Japanese history from William McNeill’s 1963 The Rise of the West that made my teeth clench. Rob asked me to explain what’s wrong with it, which is fair.

The caption reads

In addition to information about costume, architecture, and other forms of material culture, the figures in the diagram convey meaningful information through gesture and body language, the shading of figures, their relative sizes, and their location in the diagram

That’s all true, as far as it goes. The problem, of course, is whether the diagram is conveying accurate and clear information, and on both accounts it fails. I realize that I’m being a little unfair: the McNeill book was a survey text written almost a half-century ago, and the diagram is being used as an example of potential; it’s not being cited as an up-to-date description of Japanese history that would be acceptable today. Still, it’s worth talking about.

Here’s the image:

It turns out that the McNeill book was chock-full of these little gems, and there is a substantial section – given that the book is Western-oriented – devoted to Japanese history. Even for 1963, there are clear errors and overstatements – “the vigorous vulgarity of Hideyoshi’s age” (649) is probably my favorite line, especially as McNeill blames it for the later development of Kabuki, Geisha, Bunraku and Genroku culture – and this diagram does not really help. The focus is on European influence in the world, so the political structure of Japan in this period is really secondary: the diagram is, I think, intended to fill in some of the structural gaps left by the text, which seems like a very weak use of a complex diagram. It probably also should be viewed in the context of the other diagrams of the book: there’s a visual vocabulary here that I’m not going to be able to read the way that McNeill’s readers do.

Still, flaws in the diagram jump out at me, and here’s a short list, in no particular order, of problems

  • The headgear. The Emperor wears a papal mitre, probably because of the early (incorrect) descriptions of the Emperor as being a sort of religious potentate, in contrast to the secular authority of shoguns. The samurai and pirates (!) wear courtiers’ hats, and the Buddhist monks have hair (and beards, I think) which are more Fu Manchu than … well, Buddhist monks were bald and samurai rarely wore hats over their topknots.
  • The Villages. First, why do the Tokugawa villages have trees (and mountains in the background, I think) and pre-Tokugawa villages have only houses (on flat plains)? Even if there’s a real distinction being made here, it’s completely obscure. Second, the lines of authority in the Tokugawa period are questionable — there’s no visual distinction between the trade/migration links with towns on the one hand, and the authority lines from the samurai, unless those arrows mean more to McNeill’s readers than they do to me — but at least the villages are actually connected to something, whereas the pre-Tokugawa villages appear to be ahistorically independent, or irrelevant.
  • Buddhist Monks. The pre-Tokugawa diagram casts them as a fundamentally disruptive element, equal in weight to samurai – though Samurai appear twice in the diagram, perhaps to emphasize those who borrowed technology and religion from the West (though those weren’t either identical nor exclusive categories). It also puts their label in a place where it could apply to either the Buddhist Monk figure or the Emperor. The Tokugawa diagram eliminates them entirely when, in fact, Buddhism remained integral to Japan’s religious culture after it’s political/military defeats and, more importantly, Buddhist temples’ family and religious registration systems were a critical institution of rule connecting the Shogunate to localities (mediated through the daimyo)
  • Civil War. Aside from the problems above, though the diagram does suggest hierarchy in Tokugawa-era samurai (by size, as the caption says), there’s no indication in the pre-Tokugawa diagram that the samurai were fighting against each other: all the forces at work in the Civil War era appear to be attacking the Emperor, including the pirates who didn’t even participate in the civil wars at all, primarily being focused (as the text says, actually) on attacking Chinese shipping. There doesn’t seem to be a distinction between the dancing pirates and the towns, either, which is exceedingly odd.
  • Towns. Aside from being pirate havens (not!), Japan appears to have no actual cities in this period, but they did have city walls (not!). Also, towns come under direct control of the Shogun in the Tokugawa era, represented oddly by their appearing on one side of a balance scale held by the Shogun along with the samurai, rather than distinguishing between world-class cities under shogunal authority (Edo, Osaka, Kyoto) and the hundreds of other castle towns and other urban areas under daimyo control.
  • Jesuits. In this diagram, Jesuits stand in for all Westerners – at least the Spanish and Portuguese, Jesuits and Franciscans expelled from Japan by the Shogunal exclusion edicts of the 1630s.

That’s most of it, I think. The diagram, in the end, conveys almost no true information that isn’t better handled in the text. There’s a reason why history textbooks, once the technology allowed cheaper reproduction at higher quality, have gotten away from these kind of pseudo-artistic schematics and replaced them with maps, artwork and photographs: the danger of interposing inauthentic material is less. Not nonexistent, obviously, but at least students have a chance to see the evidence for themselves, instead of having it filtered through an artist who may or may not know anything about the history or culture being portrayed at the behest of an historian who many not have a strong sense of the visual aspects of the history or culture being portrayed. It’s actually very hard to convey a lot of accurate information visually without a great deal of textual background and explanation. Don’t get me wrong: I love good visuals. One of my prize possessions is a Japanese book of historical diagrams and images for Japanese history: it’s intended as a supplement to a middle-school or high-school history text, but many of the diagrams have never been reproduced in English as far as I know. It covers economic, military and cultural history, and I use it in my Japanese history classes quite regularly, but (aside from the fact that the labels are in Japanese) what makes the diagrams work is that they don’t try to convey everything in a single image. It’s too much.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.