This post examines the illustrations of the Japanese translation of Johann Adam Kulmus’s Anatomischen Tabellen¹, Kaitai shinsho (New Book on Anatomy)², published in 1774. Kaitai shinsho is often hailed in scholarship and in more popular writing as a paradigm-changing event in the field of Rangaku, or Dutch studies, as well as medicine in Tokugawa Japan. Sachs, for example, outright states that ’the study of anatomy with Central European methods began with this translation’³. Sakula describes a memorial erected to the book in Tokyo in 1959⁴. Such concentrated attention and praise is somewhat reminiscent of Skinner’s ’mythology of coherence’ and ’mythology of foreshadowing’⁵.

Two German articles deal with the Anatomischen Tabellen and Kaitai shinsho in detail, those by Sachs⁶ and Wagenseil.⁷ In English, a brief overview is given by Sakula;⁸ most other scholarship treats it in the context of a bigger topic, such as Low⁹ or Horiuchi.¹⁰ There is one monograph that deals with the subject at hand (Gabor Lukacs’s Kaitai shinsho), however, this has received mixed reviews¹¹.

Kaitai shinsho was based on the Dutch edition of the German work¹², published in 1734 by Geradus Dicten in Amsterdam¹³. The original Anatomischen Tabellen served as a textbook for students of surgery, which is why it was written in German, with short explanatory texts and many clear illustrations, reaching 23 editions (14 German, 1 Dutch, 1 French and 5 Latin)¹⁴. As Santoni¹⁵ demonstrates, the Japanese edition wasn’t a precise translation of the Dutch work, with parts like the dedication or the preface missing next to various comments about the Christian context. They omitted the footnotes of the original that contained the long explanations, and some parts reveal their battle for a clear understanding¹⁶.

The Japanese translation was published in the form of five bands (257 x 178 mm large, 6-7 mm thick according to Sachs¹⁷), four of which are text and one illustration. The latter is composed of a preface by Yoshio Eisho 吉雄永章, Kulmus’s translated foreword, Sugita Genpaku’s 杉田玄白¹⁸ foreword, the illustrations (41 plates), as well as an afterword by the artist Odano Naotake 小田野直武¹⁹. These have been translated into German and can be found in Wagenseil²⁰.

There exists an oft-cited anecdote detailing the reasons for beginning this translation project. Maeno Ryōtaku 前野良沢²¹ and Sugita Genpaku witnessed the execution and dissection of a criminal named Aocha Baba in 1771 and compared what they saw with their copy of the Tabellen. The end result was:

Horiuchi argues that a point of emphasis is ’serving their lords’, that is, the Confucian social context of learning being valuable insofar as it serves the state. With the conviction that their current medical knowledge wasn’t the best they could thus provide, the commencement of the translation was a necessity in the Confucian gentlemen paradigm that Sugita and his companions defined themselves by. This emphasis on the utility of the new equaling its worth is also discussed by Kuriyama²³ when talking about the Western-style illustrations. Furthermore, this Confucian aspect is further underlined by the end text itself, which was written in kanbun, traditionally associated with the higher classes and the culturally elite.

There were several collaborators, some named much later by Sugita; the group consisted of at least Sugita Genpaku, Maeno Ryōtaku, Nakagawa Jun’an 中川淳庵²⁴, Ishikawa Genjō²⁵, Katsuragawa Hoshū 桂川甫周²⁶, and Odano Naotake. Much has been discussed about Sugita in other literatures, so this post will forgo that²⁷.

Maeno Ryōtaku (1723-1803)²⁸ was of samurai class, and a pupil of Yoshimasu Tōdō (1702-1773) and Aoki Kon’yo (1698-1769). Brought up by a physician uncle, he trained under Yoshimasu and was eventually adopted into the Maeno physician family and succeeded as the head. From Aoki, he began to learn Dutch, which led to two visits to Nagasaki, and Ryōtaku having the best grasp of the language in their translation circle. It is assumed that he didn’t think the translation was ready at the time of its publication, which is why his name was ommitted.

Nakagawa Jun’an (1739-1786)²⁹ was from a physician family, and a pupil of Yasutomi Kiseki, from whom he first learned of western medicine. He had close ties to Hiraga Gennai.

Katsuragawa Hoshū (1751-1809) was from a physician family with interests in Western medicine and topography. He was appointed in 1793 as a teacher at the Igakkan, the medical school of the shogunate, specializing in Western style surgery. According to Goodman³⁰, it was Katsuragawa who procured the government’s sanction for publication.

Odano Naotake (1749-1780)³¹ was a member of the Akita ranga school of painting. Upon his daimyo’s orders, Odano stayed with Hiraga Gennai in Edo for five years, from whom he learned Western-style painting and book illustration.

In 1774, the preliminary text of Kaitai Yakuzu was published. This served to acquire the approval of the authorities before the publication of the full work.

Existing literature dealing with Kaitai shinsho often talks about the quality of the images as worse than the originals; by Wagenseil: “Alle Bilder des Kaitai Shinsho sind qualitativ wesentlich schlechter als ihre Vorlagen: ...” (p. 78); by Sachs: "[...] die qualitativ - im Vergleich zu den "holländischen" (eigentliche deutschen) Kupferstichvorlagen - schlechteren Holzschnitte des "Kaitai Shinsho"[...]" (p.81). However, these discussions omit the examination of the fundamental difference between intaglio and that of relief printing.

The Anatomischen Tabellen and almost all other Western anatomical books that rangaku scholars of the age came into contact with used the intaglio technique to create their illustrations. In this process, the reverse image is etched or engraved into a metal plate (usually copper), with the ink remaining in the depressions at the time of printing, from where it transfers onto the paper as the paper is pushed into the cavities by the press.

In contrast, relief printing necessitates that everything but the image be carved down, leaving the higher surface with the ink that is transferred to the paper, often by pressing down with hand. This was the technique utilized in the making of Kaitai shinsho, and all other Japanese books of the age.

This difference in the position of the ink is substantial when it comes to the density of the line and the possibility of details such as shading and crosshatching. Form and depth could be shown through shadows masterfully in intaglio, with very thin and closely drawn lines. This fine line is not possible with the softer woodcuts; in fact, figures are usually depicted in a flat manner, "completely devoid of shading or shadows"³². The Japanese woodblock’s strong suit is its freedom of line, colors, and integration with text.

Despite this technical given, the artist Odano Naotake did try to replicate the use of shadow for depth in many cases. But there was only so far this could be taken in a woodcut; a comparision of Bidloo’s images and those of Kaitai shinsho make it clear that such soft transitions were impossible to achieve, and further addition of the discontinous shadows would have more than likely muddled the clarity of the illustration. This was later achieved with the intaglio print of the book, Jutei Kaitai shinsho.

The etching or intaglio technique as used for illustrations was introduced by Shiba Kōkan³³ from 1783. Kornicki writes that this was particulary associated with the Rangaku school’s vision, utilized among other subject matters for medical illustrations³⁴.

To declare these illustrations qualtitatively lesser, despite their precise nature as argued above, is to favor one technique over the other without argument. Of course, it may be argued that the intaglio technique is better suited to medical illustrations, but this assumption begs the question: is it a simple western cultural habit that we associate this technique with classical scientific illustration? What exactly about it is more precise?

Western visual culture has put much emphasis on shadow, light, and form since the Renaissance, when the use of perspective became the leading paradigm of depiction. Since then, the subjects of the illustration are depicted from one viewpoint, as though gazing through a window, with the goal of the artist being the mastery of life-like depiction, the ultimate imitation. The technique of intaglio serves these needs very well; woodblock techniques were quickly sidelined in the repertoire.

This obsession with perspective and form cannot be said for the Japanese context at this time. Movement and clarity was much more important, as demonstrated by Tinios³⁵. Epistemic images of this tradition rely on few lines and annotations within, the merging of text and image to make their points clear.

Next to the original tables from Kulmus, five works serve as sources of the illustrations. These foreign images are clearly marked with seal script kanjis from one of the five elements, making attribution simpler; although, as Wagenseil³⁶ notes, the authors have been mistakenly named. The following table lists the sources as based on my comparisions.

As clear from above, I propose a different attribution than Wagenseil. Sugita isn’t wrong when he gives Ambroise Paré as a source, as the images of the hand and foot do come from his work, and not from Bidloo’s, as Wagenseil and Sachs both suggest. They also name Coiter as a source, in place of which Wagenseil suggests Vesal(ius), (Sachs merely states the latter as a source). I could find no match to Coiter and Vesal, instead, many images match the work of Valverde, from whom the title page has also been copied. Mestler notes that "the Japanese had no direct knowledge of Vesalius’ work," but that Valverde copied parts of his De Fabrica³⁷. Which parts exactly is not elaborated on.

As visible from the comparison images below, identification of the sources can clearly be made. I propose that Wagenseil makes these mistaken attributions out of the conviction that the images of Kaitai shinsho are of a worse quality (he clearly states this), and thus are also inaccurate. This is not the case. What he mistakes as a change in quality is in fact a change in printing technique, one that necessitates different pictorial techniques (which he admits), but doesn’t mean a lesser understanding of the images or their imprecise copying.

There is only one place where imprecision in copying seems to occur, on page four. The coccyx and the sacrum seem to be at first sight very crude copies of the originals. This doesn’t fit the overall accuracy of the book, and closer inspection reveals that in fact, the designation is false; they are copies from Bartholin, not Valverde.