Mayan New Year ceremonies. Reproduced, with permission, from Plate 28 in A Commentary on the Dresden Codex: A Maya Hieroglyphic Book, by 1 B. S. Thompson, American Philosophical Society, Philadelphia, PA, 1972.10 The Mayan Calendars: The invention of the Central American calendar in the Seventh century before Christ may be described with all propriety as one of the outstanding intellectual achievements in the history of man. This calendar solved with conspicuous success the great problem of measuring and defining time which confronts all civilized nations. Moreover il required the elaboration of one of the four or five original systems of writing the parts of speech in graphic symbols, and it conjoined with this supplementary invention of hieroglyphs the earliest discovery of the device of figures with place values in the notation of numbers. This time machine of ancient America was distinctly a scientific construction, the product of critical scrutiny of various natural phenomena by a master mind among the Mayas. It permitted a school of astronomer-priests to keep accurate records of celestial occurrences over a range of many centuries, with the ultimate reduction of the accumulated data through logical inferences to patterns of truth. —Herbert J. Spinden’ The Reduction of Mayan Dates (1924) The Mayans, developers of an ancient Amerindian civilization in Central America, employed three separate, overlapping calendrical systems called by schol- ars the long count, the haab, and the tzolkin. Their civilization reached its zenith during the period 250-900 C.E., and the Mayans survive to this day in Guatemala and in the Yucatan peninsula of Mexico and Belize; some groups have preserved parts of the calendar systems. What is known today has been recovered through astroarcheological and epigraphic research (see, for example, [8]). There is gen- eral agreement on the Mayan calendrical rules and the correspondence between the three Mayan calendars; however, the exact correspondence between the Mayan calendars and Western calendars is still a matter of some slight dispute. Correspon- dences are proposed by date equivalences in Spanish sources and by interpreting Mayan recordings of astronomical phenomena, such as new moons. In this book, we give the details for the most popular (and nearly universally accepted) of the corre- spondences, the Goodman-Martinez-Thompson correlation {18}. Another correlation was used by Spinden [15], [16], [17].? A superb discussion of Mayan mathematics, ' Stewart M Clamen wrote an early version of the Mayan calendar functions given here, ? Some of Spinden’s date calculations are wrong Here are three examples on page 46 of “Maya Dates and What They Reveal” [17], he gives the equivalence 1D 1,785,384 = February 10, 176 (Gregorian), but it should be February 11, 176 (Gregorian), on top of page 55 several Gregorian dates are off by 137138 10 The Mayan Calendars astronomy, and calendrical matters is given by Lounsbury [7] (see also [6]). Other good general sources are [9] and [14]. [4] contains extensive discussions of the regional variations of the calendars of Mesoamerica; however, its correlations are considered speculative, not authoritative. The Aztecs had calendars analogous to the haab and tzolkin, borrowed from the Mayans who long predated them. We discuss these two calendars in Section 10.3. 10.1 The Long Count The long count is a strict counting of days from the beginning of the current cycle, each cycle containing 2,880,000 days (about 7885 solar years); the Mayans believed that the universe is destroyed and recreated at the start of every cycle. The units of the long count are | kin = I day 1 uinal 20 kin (20 days) ltun = 18uinal (360 days) Ikatun = 20tun (7200 days) Ibaktun = 20katun (144,000 days) Thus, the long count date 12.16.11.16.6 means 12 baktun, 16 katun, 11 tun, 16 uinal, and 6 kin, for a total of 1,847,486 days from the start of the Mayan calendar epoch. (It is uncertain when the Mayan day began; there is evidence that the tzolkin day began at sunset and the haab day at sunrise, or in any case that they began at different times of day.) Although not relevant here, the Mayans used the following larger units for longer time periods: 20 baktun —_—_ (2,880,000 days) 1 calabtun 20 pictun (57,600,000 days) 1 kinchiltun 20 calabtun (1,152,000,000 days) 1 alautun = 20kinchiltun (23,040,000,000 days) 1 pictun An alautun is about 63,081,377 solar years! The starting epoch of the long count, according to the Goodman-Martinez- Thompson correlation, is taken as Monday, August 11, —3113 (Gregorian). This date equals September 6, 3114 B.C.E. (Julian),? which was (at noon) JD 584,283, that is, R.D. —1,137,142:4 mayan-epoch “! fixed-from-jd (584283) 0.1) In other words, our R.D. 0 is long count 7.17.18.13.2. 1 day; on page $7 he gives the equivalence 1D 2,104,772 = August 30, 1050 (Gregorian), but it should be July 27, 1050 (Gregorian). 3 Thompson [18] errs in referring to this date as “3113 B C..” confusing the two systems of dealing with years before the common era (see page 12). His error has been reproduced by many scholars. * Almost all experts believe this correlation, or possibly 30 $84,285, is correct; [8] discusses the sources of these two main contenders Spinden’s value, now no longer used, is 30 489,38410.2 The Haab and Tzolkin Calendars 139 Thus, to convert from a Mayan long count date to an R.D. date we need only compute the total number of days given by the long count and subtract the number of days before R.D. 0 by adding the epoch: fixed-from-mayan-long-count (10.2) de ( [atm kawn | tun | uinal | kin ) e mayan-epoch + baktun x 144000 + katun x 7200 + tun x 360 + uinal x 20 + kin In the opposite direction, converting an R.D. date to a Mayan long count date, we need to add the number of days in the long count before R.D. 0 and then divide the result into baktun, katun, tun, uinal, and kin: mayan-long-count-from-fixed (date) %f (10.3) battun | katun | tun | uinat | kin | where long-count = date — mayan-epoch long-count ee day-of-baktun = long-count mod 144000 EE Ht | The haab month names and tolkin day names are transliterated from the Yucatan (Yucatec) Mayan language The Guatemalan (Quiché) Mayans used slightly different names. The translations here are according to [4]Zip Zotz Xul Yaxkin Mol Chen 10.2 The Haab and Tzolkin Calendars 141 @a2 «~ 8 ® 8B ~ 8B 6B « 8 8 Be ~ § So ~— 2 @ & Muan 5) & Sa -~ 68 g @ kow § iz} om OB vo BB a Figure 10.1- The haab month signs. Adapted from Spinden [16, Fig. 3]. It is not possible to convert a haab date to an R.D. date because without a “year” there is no unique corresponding R.D. date. We can ask, though, for the R.D. date of the Mayan haab date on or before a given R.D. date: mayan-haab-on-or-before (haab, date) - date — ( ( date — mayan-haab-epoch — mayan-haab-ordinal (haab) ) mod 365 ) (10.7)142 10 The Mayan Calendars This is an instance of formula (1.54) with the ordinal position of R.D. 0 being A = (0 — mayan-haab-epoch) mod 365. ‘The third Mayan calendar, the tzolkin (or sacred) calendar, was a religious calen- dar consisting of two cycles: a 13-day count and a cycle of 20 names: (1) Imix (Alligator) (11) Chuen (Monkey) (2) Ik (Wind) (12) Eb (Tooth) (3) Akbal (Night) (13) Ben (Cane) (4) Kan (Iguana) (14) Ix (Jaguar) (5) Chicchan (Serpent) (15) Men (Eagle) (6) Cimi (Death) (16) Cib (Owl) (7) Manik (Deer) (17) Caban (Quake) (8) Lamat (Rabbit) (18) Etznab (Flint) (9) Muluc (Rain) (19) Cauac (Storm) (10) Oc (Foot) (20) Ahau (Lord) (The translations are according to [4].) The pictographs for the tzolkin names are shown in Figure 10.2. According to [9], the tzolkin calendar’s length is roughly that of human gestation and approximates the crop cycle, hence it was used for prediction of human destiny and to determine planting/harvesting times. This calendar is still in use among the Guatemalan Mayans. Unlike the haab months and days, the counts and names cycle simultaneously, and thus, for example, 13 Etznab precedes 1 Cauac, which precedes 2 Ahau, which precedes 3 Imix, and so on. Because 20 and 13 are relatively prime, this progression results in 260 unique dates, forming the “divine” year. The long count date 0.0.0.0.0 is taken to be tzolkin date 4 Ahau. (The differ- ent correlations agree on this, too.) Representing tzolkin dates as pairs of positive integers number | name |, where number and name are integers in the ranges | to 13 and | to 20, respectively, we specify def mayan-tzolkin-epoch ‘= (10.8) mayan-epoch ~ mayan-tzolkin-ordinal ( 4 | 20 ) where the function mayan-tzolkin-ordinal is explained below. We can convert from an R.D. date to a tzolkin date with mayan-tzolkin-from-fixed (date)! (10.9) number | name where: count = date — mayan-tzolkin-epoch + 1 number = count amod 13 name = count amod 2010.2 The Haab and Tzolkin Calendars 143 Imix © Chuen @ - Oe + Be Akbal @ Ben @ Ken @ Ik @ Q Chicenan @) @ Men @ @ ~ 69 += Ba Manik © aban @ Lamat (3) Eznab @ wc §) EC Cauac @ ® 7. | Figure 10.2: The tolkin name signs. Adapted from Spinden [16, Fig. 1]. Just as with the haab calendar, it is impossible to convert a tzolkin date to an R.D. date. Unlike the haab calendar, however, because day numbers and day names cycle simultaneously, to calculate the number of days between two given tzolkin dates requires the solution to a pair of simultaneous linear congruences, as we did in Section 1.11. (See [11] for a general discussion of this topic and [7] for a specific discussion relating to the Mayan calendars.) Suppose we want to know the number of days x from tzolkin date [1 | 1 ] until the next occurrence of tzolkin date | m | n |. We apply formula (1.65) with a=m-1l1,b=n-1,c = 13,d = 20,f = A = 0. Because k = —3 is the144 10 The Mayan Calendars multiplicative inverse of 13 modulo 20, we get x = (m—1+413[-3(n — m)}) mod 260. Accordingly, we define mayan-tzolkin-ordinal ( rane | ) a (10.10) (number — 1 + 39 x (number — name)) mod 260 As with the haab calendar, this function can be used to compute the R.D. date of the Mayan tzolkin date on or before a given R.D. date: mayan-tzolkin-on-or-before (tzolkin, date) %f (10.11) date — ( ( date — mayan-tzolkin-epoch — mayan-tzolkin-ordinal (rzolkin) ) mod 260 ) This is another instance of formula (1.54). The Mayans referred to haab years by their “year bearer,” the tzolkin day name of 0 Pop (the first day of that haab year).° Because the haab year is 365 days and the tzolkin is 260, only tzolkin day names Ik, Manik, Eb, and Caban can occur as year bearers in this scheme. The year bearer for a given R.D. date is computed by mayan-year-bearer-from-fixed (date) ‘f (0.12) bogus if ( mayan-haab-from-fixed (date) ) month = 19 (mayan-tzolkin-from-fixed (x)) name otherwise where + = mayan-haab-on-or-before( [1 [ 0 J. date + 364 ) Dates in Uayeb are not in a haab year and hence have no year bearer; in such cases bogus is returned. A popular way for the Mayans to specify a date was to use the haab and tzolkin dates together, forming a cycle of the least common multiple of 365 and 260 days: 18,980 days or approximately 52 solar years. This cycle is called a calendar round, and we seek the latest date, on or before a given R.D. date, that falls on a specified date of the calendar round, with Haab date haab and Tzolkin date rzolkin. Again we apply formula (1.65), this time with c = 365 and d = 260, and no shifts. The greatest © This is the Tikal custom, some scholars believe and some data is consistent instead with the tzolkin name of the last day of the previous year.10.3 The Aztec Calendars 145 common divisor of c and d is 5. The inverse of 365/5 = 73 modulo 260/5 = 52 is (by coincidence) also 5. Plugging these values into (1.65), we get (a + 365[b — a]) mod 18980, for the position of the pair of dates, a and b, in the calendar round. Using formula (1.54) to go back to the last occurrence of haab and tzolkin before date, with k and A determined in this way—once with @ = mayan-haab-ordinal(haab) and b = mayan-tzolkin-ordinal(tzolkin) and again with a = mayan-haab-epoch and b = mayan-tzolkin-epoch—and simplifying, we have: mayan-calendar-round-on-or-before (10.13) def (haab, tzolkin, date) = date — ((date — haab-count — 365 x diff) mod 18980) if (diff mod 5) =0 bogus otherwise where haab-count = mayan-haab-ordinal (haab) + mayan-haab-epoch tzolkin-count = mayan-tzolkin-ordinal (tzo/kin) + mayan-tzolkin-epoch diff = tzolkin-count — haab-count For impossible combinations bogus is returned. This function can be used to compute the number of days between a pair of dates ‘on the calendar round or to write a function mayan-calendar-round-on-or-after; we leave these to the reader. 10.3 The Aztec Calendars ‘The import of calendrics for Mesoamerican culture cannot be overstated. —Kay Read. Time and Sacrifice in the Aztec Cosmos (1998) The Aztecs (more properly called Mexica-Tenochca) used two calendars, the xi- huitl which is nearly identical in to the Mayan haab and the tonalpohualli which is akin to the Mayan tzolkin; in both cases the names are in Nahuatl, however. There are many idiosyncrasies in the Aztec calendar (associating dates, days, or years with colors, directions, patrons, auspiciousness, and so on), but these are computationally trivial and we ignore them.146 10 The Mayan Calendars The precise correlation between Aztec dates and our R.D. dates is based on the recorded Aztec dates of the fall of (what later became) Mexico City to Hernan Cortés, August 13, 1521 (Julian), Thus we define aztec-correlation “© (10.14) fixed-from-jutian ( [1521 [ august | 19 | ) which was R.D. 555,403. The xihuitl calendar approximated the solar year; like the Mayan haab, it was 365 days long, broken down into 18 “months” of 20 days each, followed by 5 unnamed worthless days called nemontemi. Scholars believe that the Aztecs used intercalation on the xihuitl calendar to keep it synchronized with the solar seasons, but the details of how they added days are a matter of speculation. The Nahuatl (Aztec) names (following [4, pp. 221-222]) of the xihuit] months are: (1) Quecholli (Macaw) (2) Panquetzaliztli (Flag raising) (3) Atemoztli (Falling water) (4) Tititl (Storm) (5) Izcalli (Sprout) (6) Cuahuitleua (Tree rises) (7) Tlacaxipehualiztli (Man flaying) (8) Tozoztontli (1-vigil) (9) Huei Tozoztli (2-vigil) (10) Toxcatl (Drought) (11) Etzalcualiztli (Eating bean soup) (12) Tecuilhuitontli (1-lords feast) (13) Huei Tecuilhuitl (2-lords feast) (14) Miccailhuitontli (1-dead feast) (15) Heui Miccailhuitl (2-dead feast) (16) Ochpaniztli (Road sweeping) (17) Pachtontli (1-moss) (18) Huei Pachtli (2-moss) (19) Nemontemi (Full in vain) But, while in the Mayan haab the day count is of elapsed days (that is, goes from 0 to 19), the xihuitl day count goes from | to 20. Thus we represent a xihuitl date as a pair where month and day are integers in the ranges | to 19 and I to 20, respectively, and we treat the nemontemi as a defective nineteenth month. We can count the number of elapsed days in the cycle of Aztec xihuitl dates as follows: aztec-xihuitl-ordinal (x-date) & (10.15) (month — 1) x 20 + day — 110.3 The Aztec Calendars 147 where day aztec-xihuitl-day (x-date) month = aztec-xihuitl-month (x-date) The only difference from the haab computation is the subtraction of | from the day to compensate for the shift in range. According to [3, Table 3], the xihuitl date at the correlation point is 2 Xocotl- huetzi, so the start of a xihuitl cycle is aztec-xihuitl-correlation %! (10.16) aztec-correlation — aztec-xihuiti-ordinal (aztec-xihuitl-date (11, 2)) or R.D. 555202. Then we can compute the xihuitl date of an R.D. by aztec-xihuitl-from-fixed (date) (10.17) aztec-xihuitl-date (month, day) where count = ( date —aztec-xihuitl-correlation ) mod 365 (count mod 20) +1 count | 20 Again, the only difference from the haab computation is the addition of | to the day to compensate for the shift in range. As with the Mayan haab, because there is no count of the cycles we cannot invert this function and find the R.D. of a xihuitl date, but we can use equation (1.54) to find the R.D. of the xihuitl date on or before a given R.D.: day month aztec-xihuitl-on-or-before (xihuit!, date) © (10.18) date — ( ( date — aztec-xihuitl-correlation — aztec-xihuitl-ordinal (xihuit!) ) mod 365 )148 JO The Mayan Calendars The Aztec tonalpohualli (divinatory) calendar is, except for the names in the cycle of days, identical to that of the Mayan tzolkin: two simultaneous cycles run, a 13-day count and a cycle of 20 names [4, p. 221]: (1) Cipactli (Alligator) (11) Ozomatli (Monkey) (2) Ehecatl (Wind) (12) Malinalli (Grass) (3) Calli (House) (13) Acatl (Cane) (4) Cuetzpallin (Iguana) (14) Ocelot! (Jaguar) (5) Coatl (Serpent) (15) Quauhtli (Eagle) (6) Miquiztli (Death) (16) Cozcaquauhtli(Buzzard) (7) Mazatl (Deer) (17) Olin (Quake) (8) Tochtli (Rabbit) (18) Tecpatl (Flint) (9) Atl (Water) (19) Quiahuitl (Rain) (10) Itzcuintli (Dog) (20) Xochitl (Flower) The implementation is identical to the Mayan tzolkin; we represent tonalpohualli dates as where number and name are integers in the ranges | to 13 and | to 20, respectively, and we compute the ordinal number in the cycle of a given tonalpohualli date by: aztec-tonalpohualli-ordinal (t-datey %! (10.19) (number — 1 + 39 x (number — name)) mod 260 where number = aztec-tonalpohualli-number (t-date) name = aztec-tonalpohualli-name (t-date) According to [3, Table 3] the date at the correlation is 1 Coatl, so the start of a tonalpohualli cycle is aztec-tonalpohualli-correlation %! (10.20) aztec-correlation — aztec-tonalpohualli-ordinal (aztec-tonalpohualli-date (1, 5)) or R.D. 555299. Mimicking our tzolkin conversions we have: aztec-tonalpohualli-from-fixed (date) = (10.21) aztec-tonalpohualli-date (number, name) where count = date — aztec-tonalpohualli-correlation + | number = count amod 13 name = count amod 2010.3 The Aztec Calendars 149 aztec-tonalpohualli-on-or-before (10.22) (tonalpohualli, date) date — ( ( date — aztec-tonalpohualli-correlation. — aztec-tonalpohualli-ordinal (tonalpohualli) ) mod 260 ) According to (3] and (12], the Aztec, like the Maya, used “calendar rounds” of 52 xihuitl years—the time it takes for the xihuitl and tonalpohualli to realign; these were called xiuhmolpilli. The 52 xihuitl years of a calendar round were designated by names and numbers using four of the twenty tonalpohualli day signs, Calli (3), Tochtli (8), Acatl (13), Tecpatl (18), similar to the “year bearer” of the Mayan calen- dar, and numbers | through 13. Thus we represent xiuhmolpilli designations as [Fumber [name ], where number is an integer in the range | to 13 and name is an integer 3, 8, 13, or 18. The xihuit! year designation was taken from the tonalpohualli date of the last day of that xihuitl year (excluding the nemontemi, of course). The name of the xihuitl year containing a given R.D. is thus given by aztec-xiuhmolpilli-from-fixed (date) (10.23) bogus if month = 19 aztec-tonalpohualli-from-fixed (x) otherwise where x = aztec-xihuitl-on-or-before (aztec-xihuitl-date (18, 20) , date + 364) month = aztec-xihuitl-month (aztec-xihuitl-from-fixed (date)) This returns bogus for the nemontemi. We can determine the combination of xihuitl and tonalpohualli dates on or before an R.D. by using (1.65): aztec-xihuitl-tonalpohualli-on-or-before (10.24) (xihuitl, tonalpohualli, date) “2 date — (date — xihuitl-count — 365 x diff) mod 18980) if (diff mod 5)=0 bogus otherwise150 10 The Mayan Calendars where xihuitl-count = aztec-xihuitl-ordinal (xihuitl) + aztec-xihuitl-correlation tonalpohualli-count. = aztec-tonalpohualli-ordinal (tonalpohualli) + aztec-tonalpohualli-correlation diff = tonalpohualli-count — xihuitl-count References [1] A. F. Aveni, Empires of Time: Calendars, Clocks, and Cultures, Basic Books, Inc., New York, 1989. Republished by Kondasha America, Inc., New York, 1995. [2] C. P. Bowditch, The Numeration. Calendar Systems and Astronomical Knowledge of the Mayas, Cambridge University Press, Cambridge. 1910. [3] A. Caso, “Calendrical Systems of Central Mexico,” Handbook of Middle American In- dians, R. Wauchope, general ed., vol. 10, pt. 1 (G. F. Ekholm and 1. Bernal, vol. eds.), chap. 13, pp. 333-348, University of Texas Press, Austin, Texas, 1971. [4] M. S. Edmundson, The Book of the Year: Middle American Calendrical Systems, Uni- versity of Utah Press, Salt Lake City, Utah, 1988. [5] J. T. Goodman, The Archaic Maya Inscriptions. Appendix to vol. VIII of Biologia Centrali-Americanna, ed. by F. D. Godman and O. Salvin, R. H. Porter and Dulau & Co,, London, 1897. 16] J. S. Justeson, “Ancient Mayan Ethnoastronomy, An Overview of Epigraphic Sources,” World Archeoastronomy (Selected Papers from the 2nd Oxford International Conference on Archaeoastronomy, Merida, Yucatan, Mexico, January 13-17 1986), A. F. Aveni, ed, chap. 8 (pp. 76-129), 1989. 17] F. G. Lounsbury, “Maya Numeration, Computation, and Calendrical Astronomy,” Dic- tionary of Scientific Biography, vol. 15, suppl. 1, pp. 759-818, Charles Scribner's Sons, New York, 1978, [8] F. G. Lounsbury, “A Derivation of the Mayan-to-Julian Calendar Correlation from the Dresden Codex Venus Chronology,” The Sky in Mayan Literature, pp. 184-206, A. F. Aveni, ed., Oxford University Press, Oxford, 1992 [9] S. Milbrath, Star Gods of the Maya: Astronomy in Art, Folklore, and Calendars, Univer- sity of Texas Press, Austin, Texas, 1999, [10] S. G. Morley, The Ancient Maya, revised by G. W. Brainerd, Stanford University Press, Stanford, CA, 1963. [11] ©. Ore, Number Theory and Its History, McGraw-Hill Book Co., Inc., New York, 1948. Reprinted by Dover Publications, Inc., Mineola, NY, 1987. [12] K. Read, Time and Sacrifice in the Aztec Cosmos, Indiana University Press, Bloomington Indiana, 1998 [13] L. Satterwaite, “Concepts and Structures of Maya Calendrical Arithmetics,” Ph.D. The- sis, University of Pennsylvania, Philadelphia, 1947 [14] L. Satterwaite, “Calendarics of the Maya Lowlands,” Handbook of Middle American Indians, R. Wauchope, general ed., vol. 3, pt. 2 (G. R. Willey,, vol. ed.), chap. 24, pp. 603-631, University of Texas Press, Austin, Texas, 1965. [15] H. J. Spinden, “Central American Calendars and the Gregorian Day,” Proceedings of the National Academy of Sciences (USA), vol. 6, pp. 56-59, 1920.References 151 [16] H. J. Spinden, “The Reduction of Maya Dates,” Peabody Museum Papers, vol. VI, no. 4, 1924. [17] H. J Spinden, “Maya Dates and What They Reveal,” Science Bulletin (The Museum of the Brooklyn Institute of Arts and Sciences), vol. IV, no. 1, 1930. [18] J. E. S. Thompson, Maya Hieroglyphic Writing, 3rd ed., University of Oklahoma Press, Norman, OK, 1971.