Overview of the re-assembled tablet found in Coligny, France

Drawing by de Ricci^[1][a]

The Coligny calendar is a second century Celtic calendar found in 1897 in Coligny, France.^[3] It is a lunisolar calendar with a five-year cycle of 62 months. It has been used to reconstruct the ancient Celtic calendar. The letters on the calendar are Latin and the language is Gaulish.

The calendar features "weeks" that consist of 5 days. Each month has six weeks and either 29 or 30 days. There are twelve such months in a year, totaling 354 days. A calendar cycle consisted of five years of this type, sixty regular months plus two intercalary months. For the calendar to remain in sync with the lunar phases, the five-year cycle must have been 1,831 days long. This would have made the calendar drift out of sync with the seasons by almost a day every year. Roman sources suggest that the Celtic calendar had a thirty-year cycle. The solar drift issue could have been dealt with by dropping a month once every 30 years.^[4] Each Celtic month started on the sixth day of the lunar cycle, according to Pliny the Elder.^[5] This is the date of the quarter moon, the easiest lunar date to confirm by direct observation.^[4]

The calendar is now held at the Gallo-Roman Museum of Lyon-Fourvière. It was engraved on a bronze tablet, preserved in 73 fragments, that was originally 1.48 metres (4 ft 10 in) wide by 0.9 metres (2 ft 11 in) tall.^{[2][6]: 111} Based on the style of lettering and the accompanying objects, it probably dates to the end of the second century.^{[2][6]: 111} A similar calendar found nearby at Villards d'Heria is preserved only in eight small fragments. It is now preserved in the Musée d'Archéologie du Jura at Lons-le-Saunier.^[2]

List of months

The names of the twelve regular months are recorded as Samonios, Dumannios, Rivros, Anagantios, Ogronios, Cutios, Giamonios, Simivisonnios, Equos, Elembivios, Edrinios, and Cantlos. There were two intercalary months, Quimonios, which appeared in first year of the five year calendar cycle, and Sonnocingos, which appeared in the third year.

Samonios refers to summer (samo-) while Giamonios refers to winter (giamos). The meanings of the other names are less clear. The Celtic year was divided into halves: "summer" from May 1 to October 31, and "winter" from November 1 to April 30.

A five-year cycle would have drifted by nearly a day every year compared to the seasons. So it was probably adjusted somehow. If an intercalary month was dropped every thirty years, the solar and calendar cycles could be brought back into alignment.

The names of the twelve regular months can be reconstructed with some certainty in spite of the fragmentary state of the calendar, as each of them was repeated five times. The two intercalary months occur only once each, the first intercalary month happens on year one of five and happens between Cantlos and Samonios and contains 29 days. The second intercalary moon happens on year three of five and contains 30 days between Cutios and Giamonios. The intercalary month names are consequently reconstructed with much less certainty.^[d]

Reconstruction

The Coligny calendar was a lunisolar calendar. It attempted to synchronize the solar year and the lunar month. The common lunar year contained 354 or 355 days.

Detail of Samonios (year 1), with Quimon- visible at the top.

The first month of the year was Samonios. The name was based on samo-, Gaulish for summer.^[6]: 267 This suggests that the Celtic year once started on the summer solstice, as argued by le Contel & Verdier.^[8] However, Monard^[9] argues for an autumn equinox start by comparison with Irish Samhain.^[e]

An intercalary month every two and a half years. The additional months were placed before Samonios in the first year, and between Cutios and Giamonios in the third year. The name of the first intercalary month is not known with certainty, the text being fragmentary.^[f] The name of the second intercalary month is reconstructed as Rantaranos or Bantaranos, based on the reading of the fifth line in the corresponding fragment.^[g]

The months were divided into two halves, the beginning of the second half marked with the term atenoux or "renewal".^[h] The basic unit of the Celtic calendar was thus the fortnight or half-month, as is also suggested in traces in Celtic folklore. The first half was always 15 days, the second half either 14 or 15 days on alternate months (similar to Hindu calendars).

Months of 30 days were marked MAT, months of 29 days were marked ANM(AT). This has been read as "lucky" and "unlucky", respectively, based on comparison with Middle Welsh mad and anfad, but there is no indication of any religious or ritual content,^[13] and the meaning could also be merely descriptive: "complete" and "incomplete", or "full" and "partial".^[14]

The Coligny calendar as reconstructed consisted of 16 columns and 4 rows, with two intercalary months given half a column (spanning two rows) each, resulting in a table of the 62 months of the five-year cycle, as follows (numbered 1–62, with the first three letters of their reconstructed names given for ease of reference; intercalary months are marked in yellow):

In spite of its fragmentary state, the calendar can be reconstructed with confidence due to its regular composition. An exception is the 9th month Equos, which in years 1 and 5 is a month of 30 days but in spite of this still marked ANM. MacNeill^[15] suggested that Equos in years 2 and 4 may have had only 28 days,^[15] while Olmsted^[16][17] suggested 28 days in year 2 and 29 days in year 4.^[16][17]

The following table gives the sequence of months in a five-year cycle, with the suggested length of each month according to Mac Neill^[15] and Olmsted:^[16][17]

The total of 1831 days is very close to the exact value of 62 × 29.530585 = 1830.90 days, keeping the calendar in relatively good agreement with the synodic month (with an error of one day in 50 years), but the aim of reconciling the lunar cycle with the tropical year is only met with poor accuracy, five tropical years corresponding to 5 × 365.24219052 = 1826.21 days (with an error of 4.79 days in five years, or close to one day per year).

As pointed out already by Ricci,^[18] based on the mention of a 30 year cycle used by the Celts in Pliny's Naturalis historia (book 16), if one intercalary month is dropped every thirty years, the error is reduced to 30 – (6 × 4,79) = 1.27 days in a 30 year period (or a shift of the seasons by one day in about 20 to 21 years).^[i]

Steinrücken^[19] has proposed that Pliny's statement that the Celtic month begins on the sixth day of the month^[j] may be taken as evidence for the age of this system: Assuming that the month was originally aligned with lunations, a shift of 5 days corresponds to a period of 975 years, suggesting a starting date in the 10th century BCE.^[19][k]

In the Coligny calendar, there is a hole in the metal sheet for each day, intended for a peg marking the current date. The middle of each month is marked atenoux, interpreted as the term for the night of the full moon.^[16]: 172

There is an additional marker prinni loudin in 30 day months (MAT), at the first day of the first month (Samonios), the second day of the second 30 day month, and so on. The same system is used for 29 day months (ANMAT), with a marker prinni laget. In Olmsted's interpretation, prinni is translated "path, course", loudin and laget as "increasing" and "decreasing", respectively, in reference to the yearly path of the Sun, prinni loudin in Samonios marking summer solstice and prinni laget in Giamonios marking winter solstice.^{[16]: 76, 176–177}

Sample month

The following table shows the arrangement of a complete month (Samonios of year 2, with TRINVX(TION)SAMO(NII) marked on the 17th day). This is the only month out of 62 that has been preserved without any gaps.^[12]: 182

M SAMON MAT ◎ I N DVMAN IVOS ◎ II ıƚı M D IVOS ◎ III ƚıı D DVM IVO ◎ IIII M D ◎ V D AMB ◎ VI M D ◎ VII PRIN     LOVDIN ◎ VIII D DVM ◎ VIIII ııƚ M D ◎ X M D ◎ XI D AMB ◎ XII M D ◎ XIII ƚıı M D ◎ XIIII ıƚı M D ◎ XV ııƚ M D ATENOUX ◎ I D DVMAN ◎ II ııƚ D TRINVXSAMO ◎ III D AMB ◎ IIII ƚıı M D ◎ V ıƚı D AMB ◎ VI ııƚ M D ◎ VII D AMB ◎ VIII N INIS R ◎ VIIII N INIS R ◎ X ƚıı M D ◎ XI ıƚı D AMB IVOS ◎ XII ııƚ M D IVOS ◎ XIII D AMB IVOS ◎ XIIII M D IVOS ◎ XV D AMB IVOS

Drawing of month 14 (Samonios of year 2) by de Ricci.[1]

Each month is divided into two half-months or "fortnights" divided by the word atenoux. Within each half-month, the arrangement is tabular, beginning with the double circle "◎" indicating the peg-hole for marking the current day in the first column, followed by a Roman numeral for the day's position in the half-month. In the second column are occasional "trigrams" of the form ƚıı, ıƚı, or ııƚ, usually in that order, and sometimes instead the letter M, occasionally in combination with it; their significance is not known. In the third column, each day is marked by the letter N or D (excepting days marked as prinni loudin or prinni laget). In the final column, days are marked with additional information, such as IVOS,^[l] INIS R,^[m] AMB (only found on odd days), among others.

In the month Samonios depicted above, the 17th day is marked TRINVXSAMO, corresponding to TRINOSAM SINDIV in Samonios of year 1.

The name of the following month, DVM(AN), is mentioned several times (on days 1, 3, 8 and 16). Conversely, the following month marks days 1, 8, 16 and 17 with SAMON(I). This "exchanging of days" in odd months with the following, and in even months with the preceding month is also found in other parts of the calendar.

Skribbatous^[24] offered a modern reconstruction of the Coligny calendar under a Creative Commons license.^[24] This version is being used in an experimental use of the calendar, by Larrouturou,^[25] taking as a starting point the full moon of 5 June 2020 and marking the dates of the eclipses over a period of 5 years.