In  the academic realm, there has been a  continuing  debate over two proposed correlations. The first we may call the  584283 correlation - named after the julian day number of its  beginning date  -  and the second we may call the 584285  correlation.  The first  implies that the first day of the 13-baktun cycle  of  the Long  Count (written 0.0.0.0.0) was August 11th, 3114  B.C.;  the second,  of course, is two days later. The former has the  advantage of being in line with the count still being followed  today. The  latter was initially the best bet when formulated  in  1930, but after a reappraisal of historical documents and  ethnographic data,  the date was shifted. Today, the 285 correlation is  still being championed by Floyd Lounsbury, with Linda Schele  following suit.  I've examined Lounsbury's papers on this topic,  and  they are  not  convincing. Dennis Tedlock has been the  best  advocate pointing  out the faults in Lounsbury's original paper but,  with someone of Schele's stature lending support - apparently  without any  close  scrutiny of the matter - has created  a  distressing situation.  A note from my article "The How and Why of the  Mayan End Date in A.D. 2012" (Dec '94) clarifies the problem:

"Linda  Schele  and David Freidel,  unlike  most  Mayanists, continue to support the work of Floyd Lounsbury in promoting  the 584285 correlation. This is 2 days off from the Thompson correlation  that I use. The decisive factor in supporting the  Thompson correlation  of 584283 is the fact that it corresponds  with  the tzolkin  count still followed in the highlands of  Guatemala.  To account for this discrepency in his correlation, Lounsbury claims that  the  count was shifted back two days  sometime  before  the conquest  (not  likely), thus explaining its  present  placement. This  means that either correlation will give the  December  21st end date. Nevertheless, Schele and Freidel still report that  the end date is December 23rd, 2012 rather than Dec. 21st, an  unfortunate faux pas understandable only because they aren't  particularly interested in the specifics of the correlation debate.  For a  detailed discussion of this topic, refer to my  book Tzolkin: Visionary Perspectives and Calendar Studies (59).

The  entire  problem  is complex, yet  clearly  solvable.  I explored  the  correlation  question in Chapter Two  of  my  book Tzolkin: Visionary Perspectives and Calendar Studies (1994),  and an excerpt follows this brief intro. In an unpublished paper from 1993,  I  also  exposed the fallacy of Lounsbury's 1992  argument  which appeared in The Sky in Mayan Literature.

There  is good software available with which the student  of Mayan  calendrics  can track the tzolkin, Long  Count,  haab  and other  Mayan time cycles. For the convenience of visitors  to  my website,  I  provide simple tzolkin correlations  on  a  monthly basis.  These dates correspond to the count still be followed  by daykeepers  among  the  Quiche, Ixil, Tzutujil  and  other  Mayan groups today.

The  correlation question is probably the  most  fundamental and  difficult one that Mayan scholars have struggled with.  Once answered, an exact comparison between Mayan dates and the Gregorian calendar can be made. But the problem is complex.

Most of the stelae inscriptions in the archeological  record contain Long Count dates alongside tzolkin/haab dates. For  example, the ruler Chan-Bahlum of Palenque dedicated the Group of the Cross  Complex  and performed rituals on 9.12.18.5.16 2  Cib  14 Mol.  Knowing this, we can calculate backwards to find the  tzolkin/haab  date  which corresponds to the base date  of  the  Long Count   (the   days  previous  to  the  date  given   above   are 9.12.18.5.15  1  Men  13  Mol,  9.12.18.5.14  13  Ix  12   Mol, 9.12.18.5.13  12 Ben 11 Mol, and so on). In this way, the  tzolkin/haab date which corresponds to 0.0.0.0.0 in the Long Count is 4  Ahau  8  Cumku. In addition, the majority of  dates  from  the archeological  record  are internally consistent; they  all  lead back  to 4 Ahau 8 Cumhu. This is the mythological starting  point of the Great Cycle of 13 Baktuns. But we still haven't correlated this base date with a European time-frame. We don't know  whether Chan-Bahlum  dedicated those temples before the Battle  of  Hastings,  or before the Roman Empire. Archeological  dating  methods alone have helped to estimate the time of a kingdom's  existence, yet the urge to pinpoint an exact correlation has belabored  many scholars.  What they were looking for became known as  "the  Ahau equation"  or  the  "correlation constant." This  would  be  the Julian Day number which corresponds to the base date of the  Long Count, thereby providing a link with the Gregorian calendar. For the  sake of brevity, I will usually refer to this as  the  "corr #".

Many  noted Mayan scholars have worked on this  problem,  in addition  to their work in related fields. Tozzer, Spinden,  Morley, Teeple and Thomspon all contributed to the foundation  which eventually  enabled the "corr #" to be discovered. But it  was  a long and well scrutinized process.

The first approach was to identify eclipse glyphs  alongside Long Count and tzolkin/haab dates, and try to find likely  candidates  in  the Gregorian calendar by way of  modern  astronomical data. In his book Maya Civilization the great Mayanist J. Eric S. Thompson  had a more ambitious idea. He figured that if an  exact correlation could be made, by way of comparing inscription glyphs with  astronomical  events, he could identify  the  glyphs  which referred  to the planets and various phenomena such as  eclipses, conjunctions and solstices. Indeed, a precise correlation has  in recent years helped to decipher many hieroglyphs. But the  astronomical approach, by itself, was lacking.

Contributions  from other fields of research  soon  required any proposed correlation constant to accord with several  different considerations; ethnohistorical, archeological, and astronomical. The crucial ethnohistorical documents include the survivingMayan  books such as the Grolier, Madrid and Dresden Codices,  as well as post-conquest writings from the Yucatan. Over the  years, corr #'s have ranged from J.D. 394,483 (Bowditch:1910) to 774,083 (Vaillant:1935).  This  implies a beginning point  of  the  Great Cycle  as long ago as 3634 B.C and as recently as 2594  B.C.  By 1930,  almost a dozen proposals had already been forwarded,  when Thompson  came up with his 584285 corr #. It was based  primarily on  three cross-referenced documents from post conquest  Yucatan. These were: the Chronicle of Oxcutzcab, the Book of Chilam Balam, and  the  writings  of Bishop Diego de Landa.  The  Chronicle  of Oxcutzcab  states that a tun ended on 13 Ahau 8 Xul in  the  year 1539 (Teeple 1930:101). The book of Chilam Balam places an indigenous  calendar  next to the Julian one used  by  the  Spaniards, indicating  that  February 15th, 1544 = 11 Chuen 18  or  19  Zac. Landa's  records place July 16th, 1553 across from 12 Kan 2  Pop. (This  was later realized to have been off by one day,  as  Landa neglected to count the leap year day of Feb 29th 1552.)

In Skywatchers of Ancient Mexico, Anthony Aveni relates that this assembled evidence makes 11.16.0.0.0 (the tun ending  spoken of  above)  fall  on November 3rd,  1539  (Aveni  1980:207).  The 11.16.0.0.0 Long Count date indicates that it was not only a  tun ending, but a katun ending as well. The original Thompson corr  # was equal to J.D. 584285. Since then, ethnographic data from  the Highlands  of  Guatemala and the re-appraisal  of  the  documents reviewed above give us the slightly adjusted "Thompson 2"  correlation, otherwise known as the Goodman-Martinez-Thompson correlation  (the  GMT),  in which the base of the Long  Count  is  J.D. 584283 (Satterthwaite 1965:628).

It  seems that scholars had a free for all with  this  question,  and  that multitudes of justifiable "solutions"  could  be generated from the vast field of data.

The  GMT has survived decades of  interdisciplinary  testing and seems to be the one with the most support, although there are still  scholars who question its validity. Judith  Ann  Remington writes in an article of hers which appeared in Archeoastronomy in the  Americas that, "actually, the GMT correlation does  not  fit the astronomical evidence very well. It usually requires a fairly major  "ammendment"  of the [Venus Table] text  [of  the  Dresden Codex]"(198).  She  continues, relating that when the  Carbon  14 dating process became available, it supported the Spinden  correlation  (489,383),  and that "the GMT  correlation  was  accepted perfunctorily  at a time when the Spinden correlation  was  being rejected because of Spinden's "ungentlemanly ways"" (200).

But years before this attack, Thompson had already  defended his  position against the initial Carbon 14 data. The  Carbon  14 method was applied to archeological dating soon after its discovery, and Thompson writes that, "The uncritical acceptance of  the new  [Carbon  14] process savored too much of dancing  round  the Golden  Calf  for my liking" (191). It was soon found  that  more extensive testing was needed to obtain accurate results. Finally, a  new Carbon 14 gas-analysis technique was developed. From  Maya Civilization we read:

"In 1959 the University of Pennsylvania ran 33 counts of  samples from  ten beams in a Tikal temple. The whole series averaged  out to A.D. 746 with a leeway of 34 years on either side. These beams were  carved with a Maya date which is equivalent of A.D. 741  in the GMT correlation... the equivalent in the Spinden  correlation is A.D. 481" (40-41).

Overwhelming support for the precise placement of the Thompson corr # (the GMT) came in the 40' and 50's, when newly discovered  calendar  counts  still being followed  among  the  Quiche, Kekchi  and  Ixil  of Guatemala all supported  the  584283.  Any suspected  break  in the Calendar count between the time  of  the conquest  and recent findings is highly unlikely; they all  accurately  project  backward  to dates from the  Aztec  and  Yucatec conquest.  As  far  as the surviving counts of  Guatemala  go  to support  proposed  correlations, they do indeed all  support  the GMT:584283. But they would also support any correlation that  was different  from  the GMT by a multiple of the 260-day  cycle.  It turns out the Bowditch (1910), the Vaillant (1935) and the  Spinden (1930) are. But the archeological evidence related by  Thompson  seems  to supercede this fact.  When  the  interdisciplinary approach  is fully considered, the GMT seems to be the  best  bet right  now.  A  review of my source material used  in  this  book demonstrates  the support for the 584283.

And yet, the debate still continues among an important group of  scholars (see the last 2 sources in the table listed  above). In  "The  Base of the Venus Table of the Dresen Codex,  and  it's Significance  for  the Calendar Correlation  Problem",  Floyd  G. Lounsbury  demonstrates  support for the original  value  of  the Thompson  correlation  (1930),  which is  584285.  Regardless  of flawless reasoning and detailed charts, the astronomical basis of his  argument does not provide an accuracy within a 2 day  range. Dennis  Tedlock  relates in the notes to his translation  of  the Popol Vuh that Lounsbury's astronomical argument could easily  be two days off. Astronomer John B. Carlson concurs.

Another study also refutes the implied 2-day shift. In 1988, Munro Edmonson published an exhaustive study of the calendar systems of Mesoamerica, showing the universal uniformity of all  the calendar  counts.  This  means that if it was  the  day  6  Coatl (serpent)  to  the Aztecs of Central Mexico, it  was  6  Chicchan (serpent) to the Maya of the Yucatec, and 6 Kan (serpent) to  the Quiche  of  Guatemala, something that  demonstrates  the  amazing sophistication  of  preconquest  Mesoamerican  civilization.   He discusses  the  Yucatec documents and concludes that  the  584283 correlation is the correct one. Furthermore, Edmonson goes on  to say that no other correlation is "No other solution is  ethnohistorically possible without postulating a break in the  continuity and uniformity of the universal Middle American day count" (249). Such a position is countered by every date in Edmonson's study.

And  this is exactly what Lounsbury suggests in  his  paper, which  has many other merits. He uses a brilliant argument  that, in  fact, applies to a completely different thesis,  to  validate his  support of the 584285. Let's look a little closer at  Lounsbury's paper, wade through the technical jargon and try to  decipher exactly what is said.