We would like to thank Mrs. Alison Boyle, Curator Astronomy & Modern Physics at the Science Museum, London, for her assistance in the preparation of this article.
Oriental astronomy is subject to a wide increasing interest among institutions and private collectors. The present map lays the base for a late 17th century bronze celestial globe, sold at Christie’s Books sale, May 1998, for £139,000.
Shibukawa was the official astronomer to the Japanese Edo court and, based on his own observations, was responsible for the first major reform of the Japanese lunar calender. It has been suggested that his were the first systematic observations to be made in Japan.
He was one of the first people to use a telescope in Japan after the instrument was introduced by European traders.
Shibukawa Shunkai, also called Shibukawa Harumi, was one of the greatest Japanese astronomers. From 1650 to 1667 he travelled Japan studying Chinese and Japanese classics, and soon became interested in astronomy and science. Between the ages of 14 and 15 he studied Shinto and Confucianism with Yamazaki Ansai, and at 18, worked on calendar science with Okanoi Gentei, the physician of the Emperor and under Tsuchimakado Taifuku, the master of the Chinese Yin-yang (onyô – dô) Science at the Imperial Court. We know that in 1667 he stayed for a short time with Hoshuna Masayuki in Aizu Province, Hoshina was the uncle of the 4th Shogun Ietsuna, and a pupil of Yamazaki Ansai. From 1786 Shibukawa settled in Edo as the official astronomer of the shogunate. Here he concentrated on his astronomical work, publishing several celestial maps and books in particular in 1698 his Tenmon Kaitô (Book of Knowledge of heaven and the constellations).
His earliest surviving celestial work is the bronze celestial globe, the base dated 1673, now in the Eisei Bunko, Toykyo. In 1677 he published his first important star map the Tenmon Bun’ya no su (Map showing divisions of the heaven and regions they govern), a circular star map based on the ancient Korean star map of 1395 the Ch’onsang yolch’a punyajido. In this new star map Shibukawa added a Japanese adaption of Chinese field allocation astrology in which the terrestrial regions were associated with 9 celestial divisions based on various groupings of lunar lodges. Shibukawa also put forward a new calendrical system to replace the Senmyo (Chinese Xuamning) calendar. This new calendar the Jokyo calendar was compiled from Shibukawa’s own astronomical observations rather that Chinese theory. This, the first Japanese calendar, was adopted in 1684, and earned Shibukawa his appointment as official astronomer. In this highranking position Shibukawa became a dominant influence on Japanese astronomy. Later in his life he reworked his 1677 starchart, the Tenmon seisho zu (Map of the arrangement of stars and constellations), making various amendments to star positions. This chart was published by his son Hisatada. Shibukawa’s influence was considerable; the celestial maps of Iguchi Tsunenori 1689 and 1698 and Namara Johaku, 1692 are all copies of the Tensho retsuji no zu, and even as late as 1779, Shibukawa’s work formed the basis of celestial mapping.
Barbarians who may have theories but cannot prove methods. On Western astronomers
In 1684 he resigned his position as go master but remained at court as head of a new department, the Tenmongata (or Bureau of Astronomy). In this office he was largely responsible for the annual production of the civil calendar. His new system, the Jujireki, was based upon the Shou-shih calendar of the Yuan dynasty (1279-1368) corrected for the differences in latitude and longitude between Peking and Kyoto. However, as Shibukawa’s instruments were inferior to those used by the Yuan astronomers, discrepancies soon arose and Shibukawa’s system remained in force only until 1754.
A RARE AND IMPORTANT JAPANESE STAR MAP, compiled from Shibukawa’s own systematic star observations, the map combining a new star map with concepts from Chinese field-allocation astrology. In China the twelve Jupiter stations were correlated with the twelve ancient States of China, and used to interpret celestial omens. Shibukawa applied this field allocation concept to the regions of Japan, allowing scholars to interpret stellar activity and predict events in various regions of the country.
One of four documented copies.
In 1677 he published his first important star map the Tenmon Bun’ya no su (Map showing divisions of the heaven and regions they govern), a circular star map based on the ancient Korean star map of 1395 the Ch’onsang yolch’a punyajido. In this new star map Shibukawa added a Japanese adaption of Chinese field allocation astrology in which the terrestrial regions were associated with 9 celestial divisions based on various groupings of lunar lodges.
SHIBUKAWA SHUNKAI – THE CELESTIAL CARTOGRAPHY
The characters are classical Chinese and the celestial cartography is based on Chinese and Korean models. This attractive celestial globe depicts the night sky in a typical East Asian style down to within 35 degrees of the South Pole, representing those stars visible during the course of the year from about latitude 36 deg N. (about the latitude of Kyoto or Tokyo). In the south polar region the sky is unmapped, which is not unusual as only globes showing Jesuit influence depict the far southern constellations in any detail. These stars are of course invisible from the mid northern latitudes. The globe represents the stars and constellations as viewed from outside the universe, a customary concept, and a direct reversal of the forms used on star maps which depict the stars as they would appear looking up into the night sky. Unlike European globes which give orders of brightness to stars by using different sized symbols, the East Asia globe denotes all stars as a standard circle. Traditionally, based in early Chinese work the standard number of constellations was 283, totalling 1464 stars (Ptolemy defined 48 constellations with 1022 stars) however the total number of stars visible to the unaided eye in the course of a year is perhaps some 4000. The circles shown on the globe are standard, the celestial equator, ecliptic, circle of perpetual visibility and invisibility. The longitudinal division of the globe into lunar lodges is standard, however on this globe 27 lodges are depicted rather than the normal 28, the missing lodge being no. 20 (Zuixi) which is left out because although in ancient times it had a width of 1 degree, by AD1300 the star on which it was based had disappeared. The situation was rectified in the famous star map by Schall von Bell in 1640 which reversed the adjacent lodges. All later Chinese globes and maps follow this pattern. But the maker of this globe was clearly unaware of Schall’s work. A report on this globe by Professor F. Richard Stephenson, Professor of Physics at Durham Univeristy, indicates that although the relative positions of the equator, ecliptic and width of lunar lodges are tolerably accurate, some of the star positions are in error by as much as 10 degrees. For example, the lunar lodge constellation Tei [Chinese di] is placed 10° further north than marked on this globe, and again the lunar lodge constellation Gyu [Chinese Niu] is shown south of the ecliptic; its true position is north of the ecliptic. The number of such errors is considerable, however many of these mistakes are to be seen in the depiction of the stars on the Eisei-Bunko globe dated 1673, and there is a close correlation of the star positions between the two globes although there is a variation of the position of the characters, namely the stars. With regard to the omission of the 20th lunar lodge it is interesting to compare the two starcharts by Shibukawa. On the 1677 chart the 20th lodge is present but by 1699, this lodge has been removed altogether, a feature unrelated to any Chinese forms from this period or later, suggesting that this globe was made in the later decades of the 17th century. Interestingly Kazuhiko Miyajima, Professor of Astronomy at Doshisha University in his article, Japanese Celestial Cartography before the Meiji Period, History of Cartography, vol.2, pt.2, pp.579-603, discusses defects in Shibukawa’s approach to star positioning, “To fix the positions of constellations, the position of one star was measured and entered on the map; the other stars were then added by the eye.” Such an approach made by converting a flat star map onto a 3-dimensional surface would undoubtedly lead to some errors appearing.