[Instrument]

The Lunar Calendar in Japan

By Steve Renshaw and Saori Ihara

Revised February, 2010

Lunar Calendars

Solar Terms

Names for Months

The Solar System and Names for Days

Calendar Reform in Japan

Matasaburou's Calendar


Lunar Calendars

Most people think of the Japanese lunar calendar as being basically the same as that used in China. The Schefflers argue in the written supplement to their calendar program CalMaster 2000 that the two are "virtually identical", with "the main difference being that besides the cyclical dating and chronology being tied to the reign of each emperor [Japanese emperor to Japanese calendar; Chinese emperor to Chinese calendar], a general year numbering system [was] used that dates from the Emperor [of Japan] Jimmu Tenno in 660 B.C." For the most part, this is true. However, while calendar reckoning in Japan closely followed developments in China, lapses in acquisition of improved methods often led to a difference of at least a day or two between the two systems and often much larger "gaps".

There are a number of good sources on means of reckoning a lunar calendar, and these should be consulted for more in-depth methods [see Reingold and Dershowitz (2001) for example; Helmar Aslaksen also has an excellent article on the Mathematics of the Chinese Calendar]. Basically, as you would expect, the primary unit of a lunar calendar is the lunar "month". This is supposed to be based on "true" periods of lunation and can range from 29 days 8 hours to 29 days 19 hours [average lunation period=29.53... days]. A "month" is determined by taking the closest number of actual days corresponding to the period of lunation. Thus, a month can be either "short" with 29 days or "long" with 30 days. For any particular region or culture, the "start" or first day of a lunar month would be the day that the new moon occurs.

Obviously, the first day of a lunar "month" could sometimes differ for cultures with different longitudes; thus a lunar calendar based on longitude in China would not necessarily be accurate for Japan. Even with an accurate reckoning of the start of the month, 12 such "months" do not add up to the requisite number of days (365.24...) of an actual solar year. Ancient calendar scholars in a number of cultures found that after three solar years, the lunar calendar would be behind the sun's actual progress by about 33 days. For cultures based on agriculture, this was an intolerable system. To correct for the discrepancy, a leap "month" was added in appropriate years. Finding common multiples in days, it was initially determined that a 19 year cycle (Metonic Cycle) with 7 appropriate intercalary leap months could keep both solar and lunar reckoning in relative congruence (Chinese scholars in later times used a cycle of 76 years, and some experimented with even larger cycles). To keep the solar calendar aligned with actual seasons, it was required that the winter solstice always occur in a non-intercalary 11th month. Thus, generally in early lunar calendars, the new year began on the second new moon following winter solstice. Normal years based on a lunar calendar may have 353-355 days. Leap years may have 383-385 days. [It should be noted that in more modern times, as Reingold and Dershowitz (2001) point out, that the Metonic cycle is not used to determine leap year. "Since 1645, true, not mean, behavior of the moon and sun is used in calculations, and as a consequence, months 11 snd 12 can followed by a leap month... Chinese New Year is not always the second new moon after the winter solstice..." (p. 259). Calculations are based on a pre-determined "reference" meridian for each country or territory.]

A lunar calendar can be quite precise and synchronized with the seasons so long as correct astronomical data are used. While both calendar and navigational needs were important in the advancement of astronomy in Europe, it was improvements in lunar calendar reckoning that drove most of Chinese and later Japanese advances in astronomy. For both China and Japan, early calendars served more of an astrological function than one in keeping with accurate observation of astronomical phenomena. Chinese practices in both observation and calculation did, however, steadily advance over the centuries. Before the 18th century, improvements and reforms of the lunar calendar seemed to occur in China long before they reached Japan. After Jesuit missionaries visited China in the 16th Century, reform there was quite rapid. Meanwhile, both Emperors and Shogunates in Japan continued to value knowing "good" days and "bad" days (with associated astrology) more than precise calendar reckoning.

Return to Top


Solar Terms

In China (and later Japan), the solar year was divided into 24 periods called "solar terms" ("ki" in Japanese). These were based upon the sun's path around the ecliptic and were/are used to determine "where" to place appropriate leap months. Early "calendar" scholars in China determined 12 "principle" solar terms corresponding to multiples of 30 degrees along the ecliptic. Each of these principles was further divided in half, thus making 24 terms of 15 degrees each. The "mid" terms, or those between principle terms were called "sectional" terms. The winter solstice, for example, is a principle term at 270 degrees of solar longitude. "Pure Brightness", which occurs about two weeks after Spring Equinox (0 degrees), is a sectional term at 15 degrees. Early scholars learned that leap months (intercalary months) cannot be months with principle terms. Otherwise, exact alignment of the solar year with seasons would soon fail.

PRINCIPLE AND SECTIONAL TERMS
NUMBER
SOLAR
LONGITUDE
APPROXIMATE
GREGORIAN
DATE
JAPANESE
ENGLISH
1 315 Feb 4 RisShun Spring Begins
2 330 Feb 19 AmaMizu Rain Water
3 345 Mar 6 KeiChitsu Insects Awaken
4 0 Mar 21 ShunBun Vernal Equinox
5 15 Apr 5 SeiMyou Pure Brightness
6 30 Apr 20 KoKuu Grain Rain
7 45 May 6 RikKa Summer Begins
8 60 May 21 KoMitsu (ShoMitsu) Grain Fills
9 75 Jun 6 BouShu Awn Appears
10 90 Jun 21 GeShi Summer Solstice
11 105 Jul 7 ShouSho Moderate Heat
12 120 Jul 23 Daisho (Taisho) Great Heat
13 135 Aug 8 RisShuu Autumn Begins
14 150 Aug 23 ShoSho Heat Ends
15 165 Sep 8 HakuRo White Dew
16 180 Sep 23 ShuuBun Autumn Equinox
17 195 Oct 8 KanRo Cold Dew
18 210 Oct 23 ShimoOri Frost Descends
19 225 Nov 7 RiTou Winter Begins
20 240 Nov 22 ShouSetsu Light Snow
21 255 Dec 6 TaiSetsu Heavy Snow
22 270 Dec 22 Touji Winter Solstice
23 285 Jan 5 ShouKan Moderate Cold
24 300 Jan 20 TaiKan Severe Cold

Return to Top


Names for Months

In more ancient times, names for months did not follow the Chinese model of simple numbering; each month had a distinct name related to agriculture and the seasons. However, in modern times, even before the Meiji Restoration and consequent adoption of the Gregorian system, months were numbered sequentially [note the Kanji (Chinese Character) for moon is used for "month"]:

MONTHS IN THE JAPANESE CALENDAR
[Old and New Names for Months]

EXPLANATION: JAPANESE NAMES FOR MONTHS
ENGLISH MODERN
JAPANESE
NAME
OLD NAME OLD NAME'S
MEANING
January IchiGatsu MuTsuki Harmony, Happy Spring
February NiGatsu KisaRagi Seasonal Change of Dress
March SanGatsu YaYohi Grass Grows Dense
April ShiGatsu UTzuki Summer, Plant Rice
May GoGatsu SaTsuki Rice Sprouts
June RokuGatsu MiNaTzuki Watering Month
(put water in the field)
July ShichiGatsu FuTzuki Month of Letters
August HachiGatsu HaTzuki Month of Leaves
September KuGatsu NagaTsuki Autumn Long Month
October JuuGatsu KaNaTzuki Month of Gods*
November JuuIchiGatsu ShimoTsuki Month of Falling Frost
December JuuNiGatsu ShiHasu "Poor Looking" Winter

Return to Top


The Solar System and Names for Days

For some time following adoption of a lunar calendar, Japan used Chinese sexagenary cycles for naming days and years. However, there is evidence that from 807, a seven day week with names related to the "planets" had found its way to Japan. Before this year, as in China, Sunday had no special significance (workers did not get any days "off"), and the most important aspect of each day was determining whether it was "good" or "bad" (a practice that continues in "unofficial" lunar calendar use to this day).

In 806, the Buddhist monk and scholar Koubou Daishi wrote that Japanese calendar scholars had trouble determining good and bad days because they did not know "Mitsubi" (the secret day). Having just returned from China, he wrote of hearing of a day called "Mitsu" and apparently determined that this day must be the "secret day". It turned out to be the day "Sunday". Actually, though sources are somewhat obscure, he may have heard the Samarkand equivalent to Sunday pronounced "mee-ruu" and mistook this for "mitsu". At any rate, from 807, "Mitsubi" was written into calendars as "Sun" day or "NichiYoubi".

It seems evident that by 1007, seven day weeks with nomenclature similar to that in Western cultures were common in Japan. In the diary of Michinaga Fujiwara, September 23, 1007 is recorded as a "Kayoubi" (Tuesday); simulating in Gregorian form, this is indeed the case. The seven "sei" (solar system bodies including the Sun, Moon, Mars, Mercury, Jupiter, Venus, and Saturn) were as important to early calendar scholars in Japan as they were anywhere else, regardless of the inadequacy of accurate methods for determining their motion. As in the West, similarity of names for days and names for the then known solar system gained acceptance with the latter being "cemented" by the Meiji Restoration.

JAPANESE KANJI (CHINESE CHARACTERS) FOR "PLANETS" AND DAYS
[Symbols, Planets and Days]

EXPLANATION: "PLANETS" AND DAYS
JAPANESE
ROOT
ENGLISH JAPANESE
"PLANET"
ENGLISH JAPANESE
DAY
ENGLISH
Taiyou, Hi
Nichi
Sun Taiyou Sun NichiYoubi Sunday
Tsuki, Gatsu
Getsu
Moon Tsuki Moon GetsuYoubi Monday
Honoo, Hi
Ka
Fire KaSei Mars KaYoubi Tuesday
Mizu
Sui
Water SuiSei Mercury SuiYoubi Wednesday
Ki
Moku
Tree MokuSei Jupiter MokuYoubi Thursday
Kin Gold KinSei Venus KinYoubi Friday
Tsuchi
Dou
Land DouSei Saturn DouYoubi Saturday

Return to Top


Calendar Reform in Japan

The lack of a perceived need for precision as well as isolationist practices in Japan often kept calendar reforms adopted in China and elsewhere in the world from finding their way here. While a lunar calendar was "officially" established in Japan in 692 by the Emperor "Jitou", accurate methods of astronomical observation and calculation found little acceptance. A good example of the difficulty of Japanese "astronomers" may be seen in the calendar reform of 1684. For the first time, Japan got a lunar calendar based on Japanese longitudinal coordinates. This reform was the work primarily of the "calendar scholars" Harumi Shibukawa and Anbu Yasutomi with their student and later colleague, Jinzan Tani.

In many ways, theirs is a sad story. Shibukawa and Yasutomi seem to have been more theoretical while Tani held sway in empirical investigation and calculation. Seeing the current calendar to be in error by up to two days, they had an uphill struggle against a general obsession with astrology, mystic assumptions of the emperor, disinterest by the Shogunate, and their own lack of information from the "outside" world. By this time, Jesuit missionaries, well aware of events in Europe, were helping Chinese calendar scholars to develop accurate ephemerides and thus more accurate calendars. Unaware of these developments, yet seeing "consistent" inconsistencies in their own observations and calculations, the three scholars of Japan, especially the empirically oriented Tani, worked hard to develop an accurate means of reckoning. While they were able to get a more accurate version of an older (1282) calendar from China adopted, they were nevertheless restricted by time and place from making advances that would have truly made their observations meaningful. While other reforms were introduced in the later Edo era, it was not until the Meiji Restoration that Japan joined the world stage in having a consistent and relatively accurate method for recording dates. This, of course, was the Gregorian calendar.

In modern times, many of the aspects of the "old" lunar system remain in Japanese culture including determination of some festivals and observances (such as "Tanabata" on the seventh day of the seventh month), names for years after the current Emperor (1996 is Heisei 8), and the old Chinese sexagesimal and astrological associations (familiar animals such as mouse, cow, cat, rabbit, etc.). While many Japanese faithfully continue to plan events with their "supplemental" lunar calendars, the Gregorian system is officially recognized for all legal transactions.

Return to Top


Matasaburou's Calendar

Several errors appear to arise in Matasaburou's Diary of 1664 and 1665 with regard to dates. In names, these are generally on the order of one or two days in most places. For example, the lunar October, 1664 is recorded with 30 days, yet there was an actual new moon on the "30th"; thus it should have become lunar November 1st. Of course, the dates become congruent by the next lunar New Year, but Matasaburou's recorded dates in lunar November and December of 1664 are "behind" a lunar calendar based on actual astronomical phenomena by at least a day.

Discrepancy between solar terms in the calendar Matasaburou used and actual astronomical phenomena may be seen in another example. 1664 was indeed a leap year, and in Matasaburou's December 12th (Gregorian) entry, he mentions that "May" (an extra 5th month) was the leap month for 1664. However, if this were so, then the extra "May" would have included the principle term "Great Heat" (120 degrees in solar longitude); as we know, this would not be an accurate intercalary month. In fact, using truer astronomical observations, the extra month would have been inserted after "June" (an extra 6th month). It is possible that Matasaburou simply forgot which month was a leap month, but an equally probable "cause" for this inconsistency could be the lack of accurate reckoning mentioned above.

Finally, in Matasaburou's February 6, 1665 (Gregorian) entry, he mentions that day as the first day of Spring. Using actual astronomical measures, February 3, 1665 would have been the more accurate date for this sectional term. In our assessment of Matasaburou's diary, we show the lunar dates that he recorded. The Gregorian dates listed were obtained by matching known phenomena (such as the lunar eclipse on February 11, 1664 and the solar eclipse that occurred on January 16, 1665) and "interpolating" intervening dates.

Return to Top


References

Reingold, E.M. and N. Dershowitz (2001) Calendrical Calculations: The Millenium Edition. Cambridge University Press.

Izuru, S. (Ed.) (1994) Koujien. (Japanese Etymological Dictionary). Iwanami Shoten, Tokyo.

Kasahara, K. (Ed.) (1995) Nippon no Rekishi (The History of Japan). Kodansha, Tokyo.

Masuda, K. (Ed.) (1995) Kenkyusha's New Japanese-English Dictionary. Kenkyusha, Tokyo.

North, J. (1995) The Norton History of Astronomy and Cosmology. W.W. Norton and Co., New York.

Pannekoek, A. (1961) A History of Astronomy. Interscience Publishers, Inc., New York.

Scheffler, A.O. and P.P. (1991) CalMaster2000: Dates, Holidays, and Astronomical Events. Zephyr Services, Pittsburgh, PA.

Shunichi, U. et al (Eds.) (1991) Nihon Zenshi; Japan Chronik. Kodansha, Tokyo.

Sugimoto, M. and D.L. Swain (1989) Science and Culture in Traditional Japan. Charles E. Tuttle Co. Inc., Tokyo.

Uchida, M. (1990) Koyomi to Tenmon: Ima Mukashi. (Calendars and Astronomy: Now and Then). Maruzen Co. Ltd., Tokyo.

Watanabe, T. (1984) Kinsei Nihon Tenmon Gakushi; Vol I. (A View of Japanese History of Astronomy in Modern Times; Vol I). Koseisha Kosaikaku, Tokyo.


Please send comments to

Steven L. Renshaw

Return to Astronomy in Japan Home Page