Saturday, January 22, 2011

Matching the crescents with Rapa Nui lunar nights

Having Mamari tablet's most fascinating section opened up now, let's continue with a closer look on it.

Moons and the glyphs are clearly organised into eight sections. Each section starts with kind of "Heraldic glyphs" and is followed by moon looking glyphs, "Crescent glyphs", that face to the right, with some exceptions that we discuss soon.

Let us assume that the crescents represent nights during a lunar month. A total of three different lists of Rapa Nui lunar night names have survived, by William J. Thomson in 1886, Alfred Métraux in 1936 and Sebastian Englert in 1940s. Each list is quite the same, with some small differences.

The first section looks like this, first the Heraldic section followed by the Crescent glyph part, which seems to have just two glyphs:

Barthel traced the hook looking glyph (glyph number 7) as a crescent with a tongue-like appendix. Here the original and Barthel's tracing are presented next to each other:
Upon closer inspection his tracing looks suspicious. If compared to other crescents, the manner appendices are added to crescents is different, always clearly keeping the crescent form separate from additional elements, which would not be the case in the hook looking glyph, were it interpreted to be a crescent, too. Barthel's tracing has also both added and omitted lines from the glyph to make it better suit his own interpretation.

It appears possible that the first section has only one crescent, while the unknown hook looking glyph is actually a part of the Heraldic glyphs.

What ever the case, let's leave the beginning of the lunar month for a while and continue with the nights that follow.

The second section has six crescents (numbers 12-17). We could proceed by assuming that these match the six unnamed "kokore" nights of the lunar month.

The third section has three crescents (23, 24, 26). If the kokore nights were in proper position, the names of these nights would be Maharua, Hua and Atua.

The fourth section has three crescents (32-34) and a round glyph (35). The crescents would apparently match the nights of Maure, Ina-Ira and Rakau. This interpretation requires that Hotu, mentioned by both Métraux and Thomson preceding Maure, but omitted by Englert, would be left out from the list, at least as such; this needs a closer scrutiny once we are through with the basic material. After Rakau, the next night would be Motohi, the night of the full moon. If the analysis is correct so far, the round glyph with a humanoid figure sitting within its perimeter stands for the full moon. The glyph resembles a common Polynesian legend about a girl named Hina, who went to the moon, stayed there, and could be seen on the nights of the full moon. This legend must have been known in Easter Island: one of the lunar nights was even called "Ina-Ira" (perhaps the only truly original Rapanui night name), and the name of the moon in general was called "mahina".

It might appear as a minor detail, but it could be asked, why to use this specific glyph to represent the full moon. Is the seated humanoid figure in the moon the same person as the seated figure in the Heraldic sets (11, 19, 28, 37, 47, 55, 66)? Perhaps the glyphs are telling a story around Hina that repeats every month?

Continuing to the next section.

Following the full moon, the fifth Crescent glyph section has five crescents, apparently the second outing of unnamed kokore nights, totalling five (41-45).

The sixth section has three crescents. Tapume, Matua, Rongo (51-53).

The seventh section has five crescents. Rongo Tane, Mauri Nui, Mauri Kato, Mutu and Tireo (60-64). Last of these is given by both Thomson and Englert as the last night of the lunar month.

The eighth, and also the last, section in the tablet has two crescent glyphs (73-74). Up until now, we have fairly easily recognised 26 nights, and then we had (at least) one crescent glyph in the first section yet to be identified (8).

So, which nights are the remaining two at the end?

Métraux's calendar had Hiro following Tireo, so we could start by assuming that the first of these two stands for Hiro.  

Hiro was however omitted from the end by both Thomson and Englert, so we need to discuss that more later. How about the last crescent glyph? Now, if we go back to the beginning where we had possibly just one night, followed by the six kokore nights, that would make the first night in the list as Ari, leaving the night of the new moon, Ata, yet to be placed. Since we have an extra crescent left at the end of the list and one night remaining to be placed, it appears possible that the list actually ends with Ata, so that the night of the new moon is given in the tablet last, not first, which is the European custom. Thus, the tablet's lunar month would start with Ari, the night when the moon first appears and end with Ata, the night when the moon has become invisible. This interpretation excludes the hook glyph (7) from the list of nights.

So, the total number of recognised nights in the list is 29. This is the minimum number of nights in a lunar month. However, lunar months often need 30 nights instead of 29, since the length of a lunar month is about 29 and a half nights. Before dwelling there in more detail, we need to analyse the Heraldic glyphs next.

Sunday, January 16, 2011

Mamari tablet lunar calendar revisited

This is a post for anyone who is already familiar with Rapa Nui's old writing system, the so-called Rongo Rongo, which still goes on undeciphered.

One day I was reading Jacques Guy's innovative analysis of one of the best understood sections of Rongo Rongo in the Manari tablet, which is regarded to be a some sort of a "lunar calendar". Guy's analyses is here. Following Guy's work, the more I was trying to make sense of the glyph sequences myself, the more I was feeling that the details did not quite add up. Guy was using Barthel's tracings from the 1950s to analyse the contents of the tablet. So, I compared Barthel's glyphs to a photograph of the original tablet.

Closer examination of the photograph quickly revealed, that Barthel reproduced the glyphs quite carelessly, frequently resorting to over-interpreting and over-harmonising the glyphs, adding and removing elements so that the final tracings could only be used to give a general idea of the writing, with little to no use for detailed examination. This was especially surprising regarding the amount of work different scholars have invested in using his tracings in their futile attempts to make some sense of the tablets.

This gave me an idea to use the original photograph as the basis to revisit the Manari tablet's lunar calendar and see if some new ideas could be introduced once we are pass Barthel's work.

For better visibility of the glyphs, the tablet photograph's colours were first inverted. Then the contrast and brightness were manipulated to bring the glyphs clearly up front. Slight manual removal of remaining noise was performed as well. The reverse boustrophedon reading order was removed and glyphs organised left to right, top to bottom.

Since we are now interested in the "lunar calendar" part of the tablet, I next removed other content and left only the 88 glyphs of the calendar.

In order to further present the contents better, the moon crescents were grouped into 8 sections, each with at least one or more right facing crescents, headed by a set of preceding "heraldic" glyphs. Each separate glyph was numbered from 1 to 88.

Now, the resulting image for the "lunar calendar" appears like this (click the image to make it bigger):

Let's have here similarly organised Barthel's tracings so that similarities and differences are clearly comparable:

What do you think? I have some ideas to further analyse the lunar calendar, once its original glyphs are now readily available.